Scolaris Content Display Scolaris Content Display

Intervenciones conservadoras para el tratamiento de la incontinencia urinaria en mujeres: una revisión global de revisiones sistemáticas Cochrane

Contraer todo Desplegar todo

Antecedentes

La incontinencia urinaria (IU) es la pérdida involuntaria de orina y puede estar causada por varias afecciones diferentes. Los tipos más comunes de IU son la de esfuerzo (IUE), la de urgencia (IUU) y la mixta (IUM). Existe una amplia variedad de intervenciones disponibles para reducir los síntomas de la IU en las mujeres. Las intervenciones conservadoras se recomiendan generalmente como primera línea de tratamiento.

Objetivos

Resumir las revisiones Cochrane que evaluaron los efectos de las intervenciones conservadoras para el tratamiento de la IU en mujeres.

Métodos

Se realizaron búsquedas en la Biblioteca Cochrane hasta enero de 2021 (CDSR; 2021, número 1) y se añadió cualquier revisión Cochrane que incluyera estudios con mujeres de 18 años o más con un diagnóstico clínico de IUE, IUU o IUM, y que investigaran una intervención conservadora destinada a mejorar o curar la IU. Se incluyeron las revisiones que compararon una intervención conservadora con un "control" (que incluyó placebo, ningún tratamiento o atención habitual), otra intervención conservadora u otra intervención activa, pero no conservadora. Un grupo de personas interesadas realizó la selección y la síntesis de la evidencia.

Dos autores de esta revisión global aplicaron de forma independiente los criterios de inclusión, extrajeron los datos y evaluaron la calidad de la revisión, resolviendo los desacuerdos mediante debate. Los desenlaces principales de interés fueron la curación o la mejoría notificada por los pacientes y la calidad de vida específica de la enfermedad. El riesgo de sesgo en las revisiones incluidas se evaluó mediante la herramienta ROBIS. La certeza de la evidencia dentro de las revisiones se evaluó según el método GRADE. La evidencia relacionada con la IUE, la IUU o con todos los tipos de IU combinados (TIU) se resumió por separado. El grupo de TIU incluyó evidencia relacionada con participantes con IUM, así como de estudios que combinaron mujeres con diferentes diagnósticos (es decir, IUE, IUU y IUM) y estudios en los que el tipo de IU no estaba claro.

Resultados principales

Se incluyeron 29 revisiones Cochrane relevantes. Siete se centraron en terapias físicas; cinco en educación y asesoramiento sobre la conducta y el estilo de vida; uno en dispositivos mecánicos; uno en acupuntura y uno en yoga. Catorce se centraron en intervenciones no conservadoras, pero comparaban con una intervención conservadora. Ninguna revisión resumió evidencia relacionada con las terapias psicológicas. Hubo 112 ensayos únicos (con 8975 mujeres) que contaban con datos de los desenlaces principales incluidos en al menos un análisis.

Incontinencia urinaria de esfuerzo (14 revisiones)

Intervención conservadora versus control: hubo evidencia de certeza moderada o alta de que el entrenamiento muscular del suelo pélvico (EMSP), el EMSP junto con biorretroalimentación y los conos fueron más beneficiosos que el control para curar o mejorar la IU. El EMSP y los dispositivos intravaginales mejoraron la calidad de vida en comparación con el control.

Una intervención conservadora versus otra intervención conservadora: para la curación y la mejoría de la IU hubo evidencia de certeza moderada o alta de que: el pesario para la incontinencia junto con el EMSP fue más beneficioso que el pesario solo; el EMSP más la intervención educativa fue más beneficioso que los conos; el EMSP más intensivo fue más beneficioso que el menos intensivo; y el EMSP más una estrategia de adherencia fue más beneficioso que el EMSP solo. No hubo evidencia de certeza moderada ni alta sobre la calidad de vida.

Incontinencia urinaria de urgencia (cinco revisiones)

Intervención conservadora versus control: hubo evidencia de certeza moderada a alta que mostró que el EMSP más retroalimentación, el EMSP más biorretroalimentación, la estimulación eléctrica y el entrenamiento vesical fueron más beneficiosos que el control para curar o mejorar la IU. Las mujeres que utilizaron estimulación eléctrica más EMSP tuvieron una mayor calidad de vida que las mujeres del grupo control.

Una intervención conservadora versus otra intervención conservadora: para la curación o la mejoría, hubo evidencia de certeza moderada de que la estimulación eléctrica fue más eficaz que la acupuntura con láser. Hubo evidencia de certeza alta o moderada de que el EMSP dio lugar a una mayor calidad de vida que la estimulación eléctrica y que la estimulación eléctrica más el EMSP dio lugar a una mejor curación o mejoría y a una mayor calidad de vida que el EMSP solo.

Todos los tipos de incontinencia urinaria (13 revisiones)

Intervención conservadora versus control: hubo evidencia de certeza moderada a alta de una mejor curación o mejoría con el EMSP, la estimulación eléctrica, la pérdida de peso y los conos en comparación con el control. Hubo evidencia de certeza moderada de mejoría de la calidad de vida con el EMSP en comparación con el control.

Una intervención conservadora versus otra intervención conservadora: hubo evidencia de certeza moderada o alta de una mejor curación o mejoría con el EMSP con entrenamiento vesical que con el entrenamiento vesical solo. Asimismo, el EMSP con más supervisión individual por parte del profesional sanitario fue más eficaz que con menos contacto/supervisión y el EMSP más intensivo fue más beneficioso que el menos intensivo. Hubo evidencia de certeza moderada de que el EMSP más el entrenamiento vesical dio lugar a una mayor calidad de vida que el entrenamiento vesical solo.

Conclusiones de los autores

Hubo evidencia de certeza alta de que el EMSP es más beneficioso que el control para todos los tipos de IU con respecto a desenlaces de curación o mejoría y calidad de vida. Se tiene una confianza moderada en que, si el EMSP es más intenso, más frecuente, con supervisión individual, combinado o no con intervenciones conductuales cono sin una estrategia de adherencia, la efectividad mejora. Se tiene una confianza alta en que, para la curación o la mejoría, los conos son más beneficiosos que el control (pero no que el EMSP) en las mujeres con IUE, la estimulación eléctrica es beneficiosa para las mujeres con IUU y la pérdida de peso da lugar a más curación y mejoría que el control en las mujeres con TIU.

La mayoría de la evidencia incluida en las revisiones Cochrane es certeza baja. Es importante que las futuras revisiones Cochrane nuevas y actualizadas desarrollen preguntas que sean más útiles desde el punto de vista clínico, eviten revisiones superpuestas y consulten a las mujeres con IU para identificar mejor los desenlaces importantes.

Intervenciones conservadoras para la incontinencia urinaria en mujeres: una revisión global de revisiones Cochrane

¿Qué es la incontinencia urinaria?

Los problemas de vejiga son frecuentes en las mujeres. A tener que ir al baño con frecuencia, de forma urgente y a veces no llegar a tiempo se llama incontinencia urinaria de urgencia. La incontinencia urinaria de urgencia se produce cuando, por alguna razón, las señales que indican a la mujer que debe vaciar la vejiga son mucho más fuertes y se producen con más frecuencia de la necesaria. Las pérdidas al estornudar o hacer ejercicio se denominan incontinencia urinaria de esfuerzo, y se pueden producir si los músculos que controlan la salida de la vejiga son más débiles de lo que deberían. Las mujeres también pueden tener una mezcla de estas dos afecciones, lo que se denomina incontinencia urinaria mixta.

Los síntomas de los problemas de vejiga pueden causar mucha angustia. Por ejemplo, las mujeres a menudo pueden ser reacias a salir y pueden temer salir a caminar o participar en clases de ejercicios. A menudo, estas mujeres se pueden sentir aisladas y su calidad de vida es significativamente peor en comparación con las mujeres sin síntomas de vejiga.

¿Cómo se trata la incontinencia urinaria?

Las opciones de tratamiento para la incontinencia urinaria incluyen principalmente el "tratamiento conservador" (que evita los métodos invasivos), la medicación y la cirugía. En primer lugar se deben ofrecer tratamientos conservadores, que incluyen el entrenamiento del músculo del suelo pélvico (músculo situado entre el coxis y el hueso púbico que sostiene la vejiga, el intestino, la vagina y el útero) (con y sin tratamientos complementarios como la estimulación eléctrica), el entrenamiento de la vejiga y el uso de dispositivos. Los suelen proporcionar fisioterapeutas o enfermeras con formación especializada.

¿Cuáles fueron los objetivos?

Existe un número creciente de revisiones Cochrane relacionadas con el tratamiento conservador de los diferentes tipos de incontinencia urinaria, y el objetivo de esta revisión fue reunir los resultados de estos estudios de investigación en un documento general accesible, con aportaciones de médicos y de las mujeres afectadas por la incontinencia.

¿Qué grado de actualización tiene esta revisión global?

Esta revisión global está actualizada hasta el 18 de enero de 2021.

¿Qué se hizo?

Se buscaron revisiones Cochrane relacionadas con el tratamiento conservador de la incontinencia urinaria en mujeres y se encontraron 29 revisiones relevantes. A partir de ellas, se compilaron en tablas los datos sobre el tipo de intervención (tratamiento) y con lo que se comparó. El tratamiento de comparación podría haber sido un control (como un tratamiento simulado o la atención habitual), otra intervención conservadora o una intervención no conservadora. Se identificaron dos desenlaces clave que eran importantes para las mujeres: si se curaban o mejoraban y si su calidad de vida había mejorado. Se evaluó la calidad de las revisiones incluidas y la certeza de los datos dentro de estas revisiones (el grado de confianza en que los resultados de la revisión son correctos para apoyar o rechazar un hallazgo).

Resultados clave

Existe evidencia de certeza alta de que realizar un entrenamiento de los músculos del suelo pélvico puede curar los síntomas y mejorar la calidad de vida en todos los tipos de incontinencia urinaria. Existe evidencia de certeza moderada o alta de que estos ejercicios para los músculos del suelo pélvico funcionan mejor si son más intensos, cuentan con más respaldo por parte de un profesional sanitario y se combinan con estrategias de apoyo para el uso continuado. Las modificaciones del estilo de vida, como perder peso e intentar controlar la frecuencia con la que se vacía la vejiga, también podrían ser beneficiosas en algunos tipos de incontinencia urinaria. El uso de complementos, como la estimulación eléctrica, también podría ser beneficioso, sobre todo en las personas con incontinencia urinaria mixta o de urgencia.

Calidad de la evidencia

Aproximadamente la mitad de los resultados aportaron evidencia de certeza moderada o alta. Sin embargo, el 81% de los resultados de esta revisión a partir de los análisis de las revisiones incluyeron datos de un solo ensayo. Estas revisiones no habían podido recopilar los resultados de varios ensayos. No fue posible identificar revisiones Cochrane de algunos tratamientos utilizados con frecuencia, como las terapias psicológicas. En general, faltó un seguimiento a largo plazo y el uso de desenlaces múltiples y diversos limitó la posibilidad de combinar los resultados para obtener evidencia valiosa.

Conclusiones de los autores

Existe mucha evidencia del tratamiento conservador de la incontinencia urinaria en las mujeres y el uso de ejercicios para los músculos del suelo pélvico cuenta con un apoyo robusto para la mayoría de las pacientes, independientemente del tipo de incontinencia. Sin embargo, la evidencia actual sobre el tratamiento conservador de la incontinencia urinaria tiene muchas limitaciones y, a menudo, la evidencia no respalda las decisiones clínicas claras. Se necesitan con urgencia más estudios de investigación para establecer evidencia de calidad alta que aborden las cuestiones que interesan a las mujeres afectadas por la incontinencia urinaria.

Authors' conclusions

There is a relatively large number of overlapping Cochrane Reviews with small numbers of trials related to interventions to improve urinary incontinence (UI) in women and this overview serves to signpost clinicians and policymakers toward relevant systematic reviews to support clinical decisions, providing a single, accessible and comprehensive document that brings together all relevant reviews. This overview should also play a key role in research prioritisation, ensuring effective use of resources, promoting collaborative working toward shared priorities and avoiding duplication of effort.

Implications for practice

A diverse range of conservative interventions are aimed at curing or improving UI in women. In general, evidence is of low‐certainty and does not support clear clinical decisions. However, there is certainty that pelvic floor muscle training (PFMT) is effective for most types of UI, at least in the short‐term. The dose or intensity or the benefits of using adjuncts is less clear. However, clinical application of evidence will depend on specific details of an individual patient or setting, or both, and clinical decisions will require expert clinical reasoning and judgement if available evidence is to be interpreted and applied effectively.

It would also seem that, at present, there are many interventions used routinely with little evidence of effect. Often it is a package of interventions that is offered to the patient. Expert opinion cannot be ignored and clinical reasoning is important when selecting and implementing interventions on the basis of individual patient assessment. However, there may at times be a danger of continuing to do things because it has always been done and they seem to work for some patients, or of starting to use techniques or devices due to anecdotal evidence about how good they are. The conservative interventions, and indeed most of the devices described in the overview, are unlike drugs, which go through rigorous trials before being licensed. However, it is important that we aspire to appropriate evidence‐based and cost‐effective conservative interventions that are attractive to commissioners, policymakers and patients.

Implications for research

Further research is urgently required to establish high‐certainty evidence related to interventions to cure or improve urinary incontinence in women. This includes new and updated Cochrane Reviews and randomised controlled trials (RCTs). Care should be taken to ensure that both reviews and RCTs are of the highest possible methodological quality, with comprehensive reporting (see EQUATOR 2021 network).

In order to avoid research waste and ensure that evidence relating to the effectiveness of conservative interventions for UI is easily accessible to end‐users, any new and updated Cochrane Reviews should be carefully planned. New reviews should be planned in order to synthesise evidence relating to interventions that have been identified as clinically relevant and important by key stakeholders. For example, there are no reviews on manual therapy techniques as a treatment, or on the use of ultrasound as a biofeedback tool.

Careful consideration should be given to issues of lumping and splitting, ensuring that reviews address questions that are clinically useful and that the same evidence is not synthesised within multiple overlapping reviews. Outcomes should focus on what matters to women: quality of life, health economic outcomes and long‐term benefits. Behavioural change techniques, such as the use of apps, should be explored to improve adherence in the short and long term.

Further research is clearly required to establish high‐certainty evidence of effectiveness and cost‐effectiveness related to interventions to improve UI in women, as is effective collaboration to support large, robust RCTs of interventions currently used routinely within clinical practice. In particular, arising from but not limited to, the results of this overview, we believe that the following research is justified and important.

  • High‐quality RCTs related to the intensity and dose of PFMT. Dose should always be considered when primary and secondary research is planned and performed.

  • Full‐scale (Phase 3) RCTs to establish the benefits of electrical stimulation in its various forms and on all populations.

  • High‐quality RCTs to establish effectiveness of lifestyle advice interventions (e.g. habit training, timed voiding, bladder training and suppression techniques, weight reduction and fluid advice).

  • High‐quality RCTs on the benefits of behavioural change techniques and the long‐term adherence to PFMT.

  • High‐quality RCTs on the use of manual therapy techniques for the relief of urinary symptoms.

Background

At least one‐quarter of all women have urinary incontinence (UI), with prevalence increasing with age (Sandvik 2000). Around 20% of women with urinary problems seek professional help. This percentage increases with advancing age and is higher among women with other concomitant urogenital problems (Morrill 2007).

Evidence suggests that, for the majority of women affected, UI impacts significantly on daily living. It interferes with women's physical, psychological and social activities, reducing general health, well‐being and quality of life (NICE 2019). It is associated with an increased prevalence of major depression and, in older women, is linked to social isolation and psychological distress (Bogner 2002; Melville 2009). UI can cause several serious medical conditions (such as perineal rash, pressure ulcers and urinary tract infections) and increases the risk of admission to long‐term residential care (Hunskaar 2002).

The annual cost to the UK National Health Service (NHS) of treating clinically significant female UI has been estimated to be GBP 233 million in 1999/2000 (equivalent to approximately GBP 351 million in 2021, using the EPPI‐Centre Cost Converter) (Turner 2004). This does not include the personal costs borne by the women affected, which have been estimated to be GBP 178 million (equivalent to approximately GBP 248 million in 2021 using the EPPI‐Centre Cost Converter) (Turner 2004). Therefore, UI is prevalent and costly to healthcare providers, as well as burdensome to women both financially and in terms of negative impacts on physical and mental well‐being.

UI can result from damage to the neural control of the bladder or the pelvic floor muscles or from direct mechanical trauma to the pelvic floor (Glazener 2017a). The risk is increased by vaginal (particularly assisted) delivery, increasing age and parity, obesity and the menopause (Saraswat 2020). UI may also be caused by trauma or disease to the bladder.

Incidence figures depend on the definition used and the population investigated, with reported annual incidence rates (numbers of new cases) of UI ranging from 1% to 11% and the annual remission rate from 6% to 11% (Hunskaar 2005).

Description of the condition

UI is the involuntary loss of urine and can be caused by a number of different conditions (Blaivas 1997; Haylen 2010).

Continence is achieved through an interplay of the normal anatomical and physiological properties of the bladder, urethra, sphincter and pelvic floor, as well as the nervous system co‐ordinating these areas. The active relaxation of the bladder, coupled with the ability of the urethra and sphincter to contain urine within the bladder by acting as a closure mechanism during filling, allows storage of urine until an appropriate time and place to void is reached. The role of the pelvic floor in providing support to the bladder and urethra and allowing normal abdominal pressure transmission to the proximal urethra is also considered essential in the maintenance of continence. Crucial to the healthy functioning of the bladder, urethra, sphincter and pelvic floor is the co‐ordination between them, facilitated by an intact nervous system control. Incontinence occurs when this normal relationship between the lower urinary tract components is disrupted, resulting from nerve damage or direct mechanical trauma to the pelvic organs. Advancing age, higher parity, vaginal delivery, obesity and menopause are associated with an increase in risk (Saraswat 2020).

There are three main types of UI. Stress urinary incontinence (SUI) is the complaint of involuntary loss of urine on effort or physical exertion (e.g. sporting activities) (Haylen 2010). SUI is a symptom, rather than a condition. Research has shown that about 50% of the women below the age of 65 years with UI had SUI (Milsom 2012). Urgency urinary incontinence (UUI) is the complaint of involuntary loss of urine associated with urgency (Haylen 2010). Isolated UUI is the least common type, accounting for 10% of women who have UI (Milsom 2012). Mixed urinary incontinence (MUI) is the complaint of involuntary loss of urine associated with urgency and also with effort or physical exertion or on sneezing or coughing (Haylen 2010). It occurs in around 30% of women (Milsom 2012).

In addition, loss of urine may occur at night (known as nocturnal enuresis, the complaint of loss of urine occurring during sleep) or the interruption of sleep because of the need to urinate (with loss of urine if the toilet is not reached in time to void) and during intercourse (known as coital incontinence, the complaint of involuntary loss of urine with coitus, occurring with penetration, intromission or at orgasm).

Description of the interventions

A wide range of interventions can be delivered to reduce the symptoms of UI in women, including conservative, pharmacological and surgical interventions (see Figure 1). The type of intervention selected for an individual woman will depend on an assessment of their symptoms, types of incontinence, factors contributing to UI, associated medical conditions, and clinician and individual choice. Conservative interventions are generally recommended as the first line of treatment for women with UI and are, therefore, the focus of this overview (NICE 2019). These include, but are not limited to, the following.


ADL: activities of daily living; EMG: electromyography; TENS: transcutaneous electrical nerve stimulation.

ADL: activities of daily living; EMG: electromyography; TENS: transcutaneous electrical nerve stimulation.

  • Mechanical devices to prevent or reduce urinary leakage. These include pessaries (urethral and vaginal inserts) and mechanical plugs/patches (Lipp 2014).

  • Physical therapies, where for women with SUI, the aim is to improve muscle control. This principally includes pelvic floor muscle training (PFMT), which can be delivered with or without the use of assistive devices such as weighted vaginal cones, biofeedback or electrical stimulation (Dumoulin 2018).

  • Educational, behavioural and lifestyle advice to enhance management of UI. These commonly include methods of toileting assistance, such as prompted voiding, habit or bladder retraining and timed voiding, and advice about lifestyle factors, such as weight loss, management of fluid intake, caffeine and alcohol intake, and physical activity and exertion (Eustice 2000; Imamura 2015; Ostaszkiewicz 2004a; Ostaszkiewicz 2004b).

  • Psychological interventions, a range of which can be used to help a woman cope with her UI symptoms and improve her quality of life, based on a number of different philosophical or theoretical approaches. These include the Health Belief Model Theory of Planned Behaviour and the Social Cognitive Theory (self‐efficacy) (Alewijnse 2003a; Alewijnse 2003b; Chiarelli 1999; Whitford 2011).

  • Complementary therapies, which Cochrane's Complementary Medicine Field defines as "practices and ideas which are outside the domain of conventional medicine in several countries" and which are defined by its users as "preventing or treating illness, or promoting health and wellbeing" (Smith 2006). Therapies that are considered complementary practices in one country or culture may be considered conventional in another. For the purpose of this overview, we defined complementary therapies as complementary interventions (such as acupuncture or electroacupuncture, reflexology) but excluding medicines or consumed remedies (i.e. herbal medicines, traditional Chinese medicines, homeopathic remedies) (Bø 2013).

In addition to these groups of interventions, there are a growing number of digital health interventions that use new technologies and media to support and enhance the delivery of conservative management of UI. In particular, digital health interventions can help support the delivery of behavioural‐based interventions and may be used as part of bladder training or voiding programmes, or both.

The above conservative interventions are the focus of the overview. The following interventions were only included if they were used as comparators in the included evidence.

  • Pharmacological therapies for treating UI, including oestrogen (Cody 2012), anticholinergic drugs (Rai 2012), adrenergic drugs (Alhasso 2005), and botulinum toxin (Duthie 2011). These interventions may have had local or systemic effects.

  • Surgical procedures that aimed to treat UI by lifting and supporting the urethrovesical junction. There is disagreement about the precise mechanisms achieved by surgery, and the choice of procedures is often influenced by several different factors, including coexistent problems, a surgeon's specialty, and preference and the physical features of the woman affected (Glazener 2017a). Surgical methods principally include open abdominal retropubic suspension (Lapitan 2016), laparoscopic retropubic suspension (Dean 2017), mid‐urethral sling procedures (Ford 2015), traditional suburethral sling procedures (Saraswat 2020), anterior vaginal repair (Glazener 2017a), bladder neck needle suspensions (Bakali 2019), periurethral injections (Kirchin 2017), and artificial sphincters (Islah 2013).

  • Specialised products, such as special pads and bedsheets, and catheters, sheaths and bags.

  • Different techniques for the diagnosis of the cause of UI, including urodynamic investigations, diaries, pad tests, and imaging techniques such as x‐rays and ultrasound (Clement 2013; Groutz 2000).

How the intervention might work

Conservative interventions can work in a variety of ways, and the mechanism of action may be mechanical, physical, behavioural, psychological or a combination of these.

Mechanical devices

These are physical devices designed to stop or control urinary leakage, and work in a number of different ways (Lipp 2014). Intravaginal devices (or 'internal vaginal devices', also known as pessaries) are inserted into the vagina with the aim of supporting the bladder neck to improve SUI. Some devices are shaped with a bump that compresses the urethra, which also helps to reduce SUI. Intraurethral devices are inserted into the urethra, acting like a plug to prevent leakage. They are inserted and removed by the individual as required. External urethral devices are applied like a seal to the outer surface of the urethral opening (external placement) to stop leakage of urine from the urethra.

Physical therapies

Physical therapies are provided by rehabilitation professionals using specially designed exercises, delivered with or without the use of assistive devices, to help individuals regain or improve physical control of their bladder. These include the following.

PFMT involves repetitive selective voluntary contraction and relaxation of specific pelvic floor muscles. PFMT exercises can be taught to women by rehabilitation professionals but are then carried out independently by the woman on a regular basis, with or without supervision. PFMT can improve the strength, endurance and co‐ordination of these muscles (Alves 2015; Dumoulin 2018). For women with UUI, the biological rationale is based on Godec's observation that a detrusor muscle contraction can be inhibited by a pelvic floor muscle contraction induced by electrical stimulations (Godec 1975). Furthermore, de Groat 1997 demonstrated that during urine storage there is an increased pudendal nerve outflow response to the external urethral sphincter increasing intraurethral pressure, representing what he termed a "guarding reflex" for incontinence (de Groat 1997; de Groat 2001). Additionally, Morrison 1995 demonstrated that Barrington's micturition centre excitatory loop switches on when bladder pressures are between 5 mmHg and 25 mmHg, while the inhibitory loop is predominantly active above 25 mmHg. Inhibition involves an automatic, unconscious increase in tone for both the pelvic floor muscle and the urethral striated muscle. Thus, voluntary pelvic floor muscle contractions may be used to control UUI. After inhibiting the urgency to void and the detrusor contraction, the woman can reach the toilet in time to avoid urine leakage.

Biofeedback is used to supplement or enhance PFMT. Information about a normally unconscious physiological process is presented to the individual and the therapist as a visual, auditory or tactile signal (Sandweiss 1985). Such feedback enables a person to identify and modify a bodily function of which they may be unaware. Typically this may involve digital palpation or the use of a device to record the biological signals (e.g. squeeze pressure, electrical activity, pelvic floor morphometry using ultrasound) during a voluntary pelvic floor muscle contraction and presentation of this information back to the woman in auditory or visual form. Examples of this feedback are: verbal encouragement; a louder sound with a stronger squeeze or an increasing number of lights on a visual display as the strength of the squeeze increases; and visual display of levator ani contraction on an ultrasound screen. Thus, for a muscle that cannot be seen, the women receive some type of signal about their ability to use their pelvic floor muscle. Biofeedback may also be provided using weighted vaginal cones, which are small weights placed in the vagina that require contraction of the pelvic floor muscle to prevent them from slipping out. The cones provide a form of biofeedback, as the sensation of one slipping out induces a pelvic floor muscle contraction that may both strengthen muscles and help to synchronise muscle contraction with increases in abdominal pressure (Herbison 2013).

Several different types of stimulation, including electrical and magnetic stimulation, can be delivered either through surface electrodes (transcutaneous) or via direct stimulation (percutaneous), with the aim of stimulating the nerve supply and altering nerve activity. Stimulation of nerve supply is thought to improve muscle tone and sensation of the pelvic floor muscles, enhancing muscle control. It also aims to reduce detrusor contraction in the case of UUI. Electrical stimulation therapy can be used to treat overactive bladder (OAB) via different routes, such as implantable or internal electrodes (sacral neuromodulation) and non‐implantable or external electrodes. The latter can be subclassified as endocavitary electrodes (rectal or intravaginal) or percutaneous electrodes (tibial nerve stimulation). Cadwell 1963 was the first to report the use of intravaginal electrical stimulation (IES) in the treatment of UI. Subsequently, Messelink 1999 also used it with satisfactory results. IES using frequencies below 12 Hz stimulates the pudendal nerve, which may inhibit the detrusor muscle, reduce involuntary contractions and, consequently, reduce the number of micturitions in 24 hours (Messelink 1999). Electrical stimulation also works in passively, helping women to become conscious of the perineal muscle contraction and this may, in turn, help to inhibit detrusor involuntary contractions (Amaro 2003). IES can be used alone or in association with pelvic floor muscle exercises, often indicated in SUI and OAB. Percutaneous tibial nerve stimulation (PTNS) is a form of neuromodulation that delivers retrograde stimulation to the sacral nerve plexus through percutaneous electrical stimulation of the posterior tibial nerve via a needle electrode inserted cephalad to the medial malleolus, an anatomical area recognised as the bladder centre (Hajebrahimi 2015). Magnetic stimulation appears to induce inhibitory effects on detrusor overactivity in a similar manner to electrical stimulation, with the significant clinical advantage of being non‐invasive (Takahashi 2003).

Bladder training encourages people to extend the time between voiding so that continence might be regained. This can take months to achieve but may help people who are physically and mentally able to use this method. For women with UUI, the biological rationale is based on Godec's observation that a detrusor muscle contraction can be inhibited by a pelvic floor muscle contraction induced by electrical stimulation (Godec 1975). Furthermore, de Groat 1997 demonstrated that during urine storage there is an increased pudendal nerve outflow response to the external urethral sphincter increasing intraurethral pressure and representing what he termed a "guarding reflex" for incontinence (de Groat 1997; de Groat 2001). Additionally, Morrison 1995 demonstrated that Barrington's micturition centre excitatory loop switches on when bladder pressures are between 5 mmHg and 25 mmHg, while the inhibitory loop is predominantly active above 25 mmHg. Inhibition involves an automatic, unconscious increase in tone for both the pelvic floor muscle and the urethral striated muscle. Thus, voluntary pelvic floor muscle contractions may be used to control UUI. After inhibiting the urgency to void and the detrusor contraction, the woman can reach the toilet in time to avoid urine leakage (Wallace 2004).

Manual therapy is defined as a clinical physical approach utilising specific hands‐on techniques. It may include massage, soft tissue mobilisation, various connective tissue techniques, myofascial release, mobilisation of joints, joint manipulation or mobilisation of nerve tissue. It is used to diagnose and treat soft tissues and joint structures for the purpose of modulating pain, increasing range of motion, reducing soft tissue oedema, inducing relaxation, improving contractile and non‐contractile tissue extensibility or stability (or both), facilitating movement and improving function (personal communication: Bø 2017).

Educational, behavioural and lifestyle advice

Several lifestyle factors are thought to play a role either in the onset or later in the resolution or management of UI. These include the following.

  • Diet: many dietary factors are thought to aggravate urinary urgency and may also relate to weight gain or constipation, or both (see below). Therefore, dietary advice can be beneficial to the management of UI (Imamura 2015).

  • Exercise and activities of daily living (ADL) advice: weakened pelvic floor support structures and raised intra‐abdominal pressure caused by heavy lifting and strenuous activity may result in UI. Strenuous activity alone may also increase incontinence in the short term. Appropriate advice can help women to manage the impact of exercise and daily physical activity on UI, while maintaining a healthy lifestyle (Bø 2013).

  • Fluid/caffeine intake: worsening of urinary urgency, frequency and incontinence is often reported after consuming caffeine, alcohol, fizzy drinks, sweetened diet drinks or excessive fluids. Caffeine can increase bladder muscle contractility, whereas alcohol or excessive fluids may have a diuretic effect (Imamura 2015).

  • Voiding interventions: this is a broad term used to describe any type of scheduled toileting intervention, which can include programmes of scheduled bladder voiding and bladder training (aimed at trying to correct faulty habit patterns of frequent urination if present, improve control over bladder urgency, prolong voiding intervals, increase bladder capacity, reduce incontinent episodes and restore women's confidence in controlling bladder function) (Eustice 2000).

  • Weight loss/obesity: obesity and UI are common problems in women. Obese women have higher intra‐abdominal pressure than non‐obese women and it is thought that this chronically elevated pressure may predispose to incontinence by weakening pelvic floor support structures and by raising intra‐abdominal pressure (Imamura 2015).

  • Smoking cessation: there is evidence of a relationship between cigarette smoking and UI, although the mechanism is not fully understood (Bump 1992). Chronic coughing among smokers may also contribute to UI by raising intra‐abdominal pressure (Imamura 2015).

  • Healthy bowel management: constipation can obstruct the bladder, preventing adequate voiding and resulting in urine leakage. Chronic straining may also be a risk factor in the development of UI. Advice that avoids or limits constipation or chronic straining may improve or prevent UI (Imamura 2015).

  • Anatomy and physiology education: educational interventions to teach women about the causes of their UI may improve understanding of the condition and may therefore help women manage their symptoms (Imamura 2015).

Psychological therapies

There are many different types of psychological therapies, which are based on a range of theoretical and philosophical standpoints. These are often forms of talking therapy with individuals or in groups but may also include interventions such as telephone or internet‐based support. Psychological therapies are generally aimed at helping people change the way they think and behave. Psychological therapies may help women with UI to manage and maintain a sense of well‐being and enhance quality of life. For the purposes of this overview, we considered psychological therapies within the categories proposed and described by Shinohara 2013.

  • Behavioural therapies include behavioural therapy, behavioural activation, social skills and assertiveness training, and relaxation therapy.

  • Cognitive‐behavioural therapies include cognitive therapy, rational emotive behavioural therapy, problem‐solving therapy, self‐control therapy and courses aimed at coping with depression.

  • Mindfulness‐based 'third wave' cognitive and behavioural therapies include acceptance and commitment therapy, compassionate mind training, functional analytical psychotherapy, extended behavioural activation, meta‐cognitive therapy, mindfulness‐based cognitive therapy and dialectical behavioural therapy.

  • Psychodynamic therapies include interventions based on the Drive/structural model, Relational model and Integrative analytical model (Malan 1963; Mann 1973; Strupp 1984).

  • Humanist therapies include person‐centred therapy, Gestalt therapy, experiential therapies, transactional analysis, existential therapy, and non‐directive and supportive therapies.

  • Interpersonal, cognitive analytic and other integrative therapies include interpersonal therapy, cognitive‐analytic therapy, psychodynamic‐interpersonal therapy, cognitive‐behavioural analysis system of psychotherapy, counselling and motivational interviewing.

Complementary therapies

Several alternative therapies, such as hypnotherapy and acupuncture, may be used, often alongside other conservative interventions. These interventions are generally provided to help people feel better and to promote health and well‐being. Complementary therapies used to reduce symptoms and promote well‐being in women with UI may include (but are not limited to) the following.

  • Acupuncture or electroacupuncture is the practice of inserting a needle or needles into certain points in the body for therapeutic purposes (Wang 2013).

  • Hypnotherapy is a form of psychotherapy that can be used to create subconscious change in an individual in the form of new responses, thoughts, attitudes, behaviours or feelings (Komesu 2011).

  • The Bowen Technique is a hands‐on therapy in which very gentle pressure is applied to specific points on the body (Wilks 2007).

  • Reflexology is a massage used to relieve tension and treat illness, based on the theory that there are reflex points on the feet, hands and head linked to every part of the body (Mak 2007).

Other conservative interventions

There are several other conservative interventions that may be used for women with UI that do not fit within the above categories. These can include (but are not limited to) the following.

  • Core‐stability training involves specific exercises, comprising stretching and strengthening exercises that are adapted to the condition of the intervention, aimed at improving muscle strength and control around the pelvic area. These exercises may be delivered by an exercise instructor or person who is not a rehabilitation professional. Increased muscle strength and control around the pelvic area may improve the symptoms of UI by changing intra‐abdominal pressure and increasing pelvic floor muscle control (Bø 2013).

  • Breathing exercises and hypopressive exercises generally aim to complement PFMT by changing the pressure on the abdominal wall and improving the overall quality of PFMT exercises (Bø 2013).

  • Modern Pilates exercise programmes incorporate exercises that involve breathing and contraction of pelvic floor muscles. The pelvic floor muscles are not specifically trained but they are trained incidentally during exercise and movement. The co‐contraction of pelvic floor muscles that occurs incidentally during Pilates exercises will counteract increases in intra‐abdominal pressure that occur during exercise, preventing leakage and strengthening pelvic floor muscles (Bø 2013).

  • Yoga is a physical, mental and spiritual practice, which may benefit UI through changes to physical (e.g. muscle stretching, control) and psychological mechanisms (Bø 2013).

  • In the Paula Method, all sphincters in the body work simultaneously so exercising the ring muscles of the mouth, eyes, or nose may result in co‐contraction and strengthening of the pelvic floor muscles (Bø 2013).

  • Tai Chi is an ancient exercise regimen originating in China and has widespread use as exercise for general health in China. Chang 1986 describes an exercise called "the deer" involving contraction of the anal sphincter. The exercise is recommended for both men and women for conditions related to the pelvic area (Bø 2013).

  • Additionally, Carriere 2006 has claimed that "poor posture" can lead to pain and dysfunction in the pelvic floor. It is thought that optimal strategies for transferring loads will balance control of movement while maintaining optimal joint axes, maintain sufficient intra‐abdominal pressure without compromising the organs (preserve continence, prevent prolapse or herniation) and support respiration. Suboptimal strategies for posture, movement and breathing, or combinations thereof, create failed load transfer which can lead to pain, incontinence and breathing disorders (Bø 2013).

Why it is important to do this overview

Conservative management is recommended as a first‐line treatment for women with UI (NICE 2019). However, identifying the most effective rehabilitation interventions is not always easy. Given the importance of curing, improving or managing UI symptoms to allow women to have an active lifestyle and good quality of life, there are a substantive and growing number of randomised controlled trials (RCTs) and systematic reviews relating to the effectiveness of conservative interventions for UI. Despite this growing body of evidence, current clinical practice often does not reflect the available evidence‐base, and this important area of practice receives little attention in undergraduate physiotherapy education (Francis 2012; McClurg 2013). Lack of sufficient time to identify and synthesise evidence is cited as the key barrier to evidence‐utilisation within UI rehabilitation (McClurg 2013).

It has been recognised that a large and growing body of systematic reviews can be overwhelming for decision‐makers, including women and healthcare practitioners, who do not have time to keep up‐to‐date with this evidence‐base (Bastian 2010). In one accessible, comprehensive document, a Cochrane Overview of conservative interventions for women with UI will synthesise all high‐quality evidence about UI conservative interventions, assess the limitations of current best evidence and enable indirect comparisons of the effects of different interventions on UI. This overview will support evidence‐based management of UI among key decision‐makers (such as clinicians, policymakers or informed health service users) and educators of allied health professionals.

Objectives

To summarise Cochrane Reviews that assess the effects of conservative interventions for treating urinary incontinence in women.

Methods

Criteria for considering reviews for inclusion

Types of studies

We included Cochrane Reviews published in the Cochrane Database of Systematic Reviews that assessed conservative interventions for treating UI in women.

Types of participants

We included reviews of studies in which the participants were women aged 18 years or older with a clinical diagnosis of SUI, UUI or MUI, regardless of cause or comorbidities.

Types of interventions

We included reviews of studies that investigated a conservative intervention for which the primary aim was to improve or cure UI. Conservative interventions included those listed in How the interventions might work and are illustrated in Figure 1.

Where the above inclusion criteria were met, we included reviews of studies in which the participants were recruited from any setting, including community, hospital or care home environments. We considered reviews that included both male and female participants but only included reviews in which we could extract data relating specifically to the female participants.

We excluded reviews of surgical or pharmacological interventions, products to manage leakage of urine and investigative techniques unless these were compared with a conservative intervention. We included reviews in which a conservative intervention was considered a control intervention.

We included reviews that compared a conservative intervention with any other intervention. We categorised these comparison interventions as either 'control' (which included placebo, no treatment or usual care), another conservative intervention or another active, but non‐conservative, intervention. We documented the definitions of comparison interventions provided by review authors, and explored and described the comparison interventions according to our prestated taxonomy of interventions.

Stakeholder participation

We convened a stakeholder group comprising 14 purposively selected people, including clinicians, service users and commissioners. Members of this group are listed and acknowledged in the Acknowledgements section. We used formal group consensus methods based on nominal group techniques, as this method enables the pooling of decisions and judgements from a group of informed experts, leading to votes on a range of options until ultimately group consensus is reached (Pollock 2014; Pollock 2016). Stakeholders contributed to the protocol development using consensus methodologies to identify key areas of clinical priority to incorporate into the overview and identified key outcomes for the overview (McClurg 2016). The process of stakeholder group involvement is outlined in Appendix 1.

Types of outcome measures

This overview focussed on two critical (primary) outcomes.

  • Symptomatic cure or improvement of UI, as reported by the woman (including through self‐report or bladder diaries). This outcome is based on the woman's observations (i.e. is a participant‐reported measure), as described/reported in the review.

  • Condition‐specific quality of life, as measured by specific instruments designed to assess the impact of UI symptoms on the life of a woman, such as King's Health Questionnaire, Incontinence Quality of Life (I‐QOL) and Bristol Female Lower Urinary Tract Symptoms (BFLUTS) questionnaire (Jackson 1996; Kelleher 1997; Wagner 1996)

See Differences between protocol and review for further information.

Timing of outcome assessment

We considered outcomes at three time periods: the end of treatment, up to one year after end of treatment and more than one year after end of treatment.

We categorised outcomes presented within analyses (forest plots) as either 'immediate' (i.e. at the end of intervention) or 'follow‐up', documenting and reporting within tables the time point of the data pooled, as reported in the included review.

We identified information relating to all outcomes synthesised within the included reviews but only extracted data relating to effect size from relevant analyses of comparisons relating to these stated outcomes of interest.

Search methods for identification of reviews

We identified relevant reviews from Cochrane Incontinence's list of published Cochrane Reviews and searched the Cochrane Library (Cochrane Database of Systematic Reviews; 2021, Issue 1) using the strategy given in Appendix 2 on 18 January 2021.

We also noted titles and protocols registered with Cochrane Incontinence for consideration in future versions of the overview.

Data collection and analysis

During the process of data collection and analysis, evidence relating to SUI, UUI or MUI were separated according to these three subgroups. With the exception of 'Data analysis', the description of methods within subsequent sections refers to synthesis of data as presented within the included reviews and not to any re‐analysis or pooling of data.

Selection of reviews

Two overview authors (two of CH, DM, AP, PC) independently considered the titles and abstracts from the identified reviews and applied the inclusion criteria (see Criteria for considering reviews for inclusion). We resolved disagreements through consideration and discussion of the full paper, involving a third overview author where necessary.

We contacted review authors of any titles or protocols that appeared to meet our selection criteria, identifying those that authors indicated should be completed within three months of our initial search date. We also contacted authors of all completed reviews meeting our selection criteria for which the search date was more than 12 months ago and asked if an update was anticipated within this three‐month period. Initial contact with review authors was made via Cochrane Incontinence. When authors indicated that a review should be finished or updated within this timeframe, we sent a reminder email in advance of this date to check on progress and to gain access to relevant prepublication data where possible.

Data extraction and management

Two overview authors (CH, AP) independently extracted data. We resolved disagreements by discussion, with assistance from a third overview author where necessary. We used a data collection form specifically designed and piloted by the overview author team. Onto this form, we extracted and recorded key features of each review, including details of the aims and rationale, types of studies, participants, interventions, comparisons, outcomes assessed, date of last search and analyses (forest plots) completed.

Using a spreadsheet, we systematically synthesised the studies included within all identified reviews to explore whether any reviews covered the same studies. When there was overlap between reviews, two overview authors discussed the overlap with consideration of each review question and explored comparisons, the date of the last search and key aspects of methodological quality (e.g. types of studies included, risk of bias assessment). We used these details to reach an agreement regarding which data from which review comparisons were to be included within the overview.

Type of urinary incontinence

During the data extraction process, two overview authors independently noted whether each included review reported evidence relating to SUI, UUI, or MUI. We resolved disagreements through discussion, using a third overview author where necessary. We had anticipated that we may identify data relating to combined or unclear populations and had proposed at the protocol stage that, if necessary, we would group data relating to "combined" populations and "unclear" populations (where the population was undefined). However, during the process of categorising type of UI we found that we were unable to distinguish between 'mixed', 'combined' and 'unclear' populations. Therefore, we did not use the planned additional groups of 'combined' and 'unclear' but instead categorised all mixed, combined or unclear data into a group of 'all types of UI'. Thus, we categorised data into three separate populations: SUI, UUI and 'all types of UI' (AUI).

We completed all subsequent stages of the overview in triplicate, for:

  • conservative interventions for management of SUI;

  • conservative interventions for management of UUI;

  • conservative interventions for management of AUI.

Criteria for identifying relevant comparisons

We used extracted data to determine which reviews had analyses (comparisons within forest plots) of relevance to this overview according to the three populations of interest. Relevant comparisons evaluated the effect on the stated primary outcomes of interest to the overview by comparing the effects of:

  • any conservative intervention versus control, placebo or standard care;

  • any conservative intervention versus other active intervention (i.e. surgical or pharmacological intervention); or

  • one conservative intervention versus another conservative intervention.

Originally, we had also planned to extract data relating to analyses comparing different doses, intensities or timing of delivery of conservative intervention. However, due to lack of analyses focused on these comparisons, we extracted no relevant data.

Assessment of methodological quality of included reviews

Quality of included reviews

Two overview authors independently assessed the methodological quality of the included reviews using the ROBIS (Risk of Bias In Systematic Reviews) tool (Whiting 2016), with input from a third overview author (AE) where there were disagreements.

ROBIS is completed in three phases, with the first assessing relevance, identifying concerns with the review process and judging the risk of bias in the review. Phase 2 assesses four domains to help assess specific concerns about potential biases within the review.

  • Domain one: study eligibility criteria (i.e. review eligibility criteria were clear, appropriate and prespecified).

  • Domain two: identification and selection of studies (i.e. all relevant primary studies should have been identified and included in the review).

  • Domain three: data collection and study appraisal (i.e. judgement relating to bias that may have been introduced through the data collection and assessment of risk of bias of included studies).

  • Domain four: synthesis and findings (i.e. appropriate methods have been used for any meta‐analyses, and syntheses of results).

We completed all signalling questions for the assessment of these domains and used these to help judge overall risk of bias. We used the rating guidance published with the ROBIS tool in answering all signalling questions (Whiting 2016).

Two overview authors carried out assessments for domains one to three independently, discussing any disagreements and reaching consensus on a final judgement of risk of bias for each of these three domains. Due to an initial high level of disagreement between these two overview authors for domain four and lack of clarity relating to how to consistently judge this domain, a third overview author (AE) independently judged this domain for 50% of the reviews. The three overview authors met and discussed their independent judgements and reached consensus on how to judge certain scenarios. Following this, the original two overview authors (CH, AP) discussed and reached consensus over the judgements for this domain for the remaining 50% of the reviews.

Phase 3 involves a judgement of the overall risk of bias of each review, following assessment of three signalling questions. These prompt an answer of 'yes', 'probably yes', 'probably no', 'no' or 'unclear'. Two or three overview authors completed the three signalling questions independently and reached consensus through discussion. The final overall judgement of low, high or unclear risk of bias was not assigned to avoid the possibility of this being used as a summary 'quality score' for each review (Whiting 2016).

Quality of evidence in included reviews

We did not reassess the quality of individual studies included within reviews but reported the quality of individual studies as assessed by the review authors, who all used the Cochrane RoB 1 tool (Higgins 2011). We assessed our certainty in the evidence synthesised within each relevant comparison (i.e. all relevant analyses and forest plots from included reviews which contain data for one of our prestated primary or secondary outcomes of interest) using the GRADE approach (Guyatt 2011a). This includes:

  • risk of bias due to flawed design or conduct of studies (Guyatt 2011b);

  • imprecision (e.g. when confidence intervals (CIs) for treatment effect are wide) (Guyatt 2011c);

  • inconsistency (e.g. when point estimates vary widely, I² statistic is large) (Guyatt 2011d);

  • indirectness (e.g. variations in participants, interventions, comparisons and outcomes) (Guyatt 2011e);

  • publication bias (may be explored using funnel plots and classed as not suspected, suspected, strongly suspected or very strongly suspected) (Guyatt 2011f).

The GRADE approach provides a system for rating certainty of evidence and strength of recommendations that is explicit, comprehensive, transparent and pragmatic. It is increasingly being adopted by organisations worldwide. However, difficulties associated with the subjectivity involved in judging grade of evidence have previously been reported, while poor agreement has been found on grading strength of evidence within systematic reviews using GRADE, even among experienced systematic reviewers (Berkman 2013). One Cochrane Overview reported the difficulty in achieving agreement between independent overview authors for GRADE judgements when numerous comparisons needed to be assessed, proposing the use of an objective algorithm to enable transparent, reproducible assignment of GRADE levels of evidence (Pollock 2014; Pollock 2016).

We explored the use of the iterative methods reported by Pollock 2016 to develop a set of objective criteria for exploring the certainty of the specific body of evidence included within this overview. Using the GRADE guidance papers, we assessed the limitations, imprecision, inconsistency, indirectness and publication bias of the comparison in two reviews and all authors assigned a final GRADE rating assigned independently. We discussed these results to determine the key variables to consider and how these could be used to assign a GRADE level.

Overview authors agreed that, for this body of evidence, the algorithm should involve systematic assessment of:

  • risk of bias of trials contributing participants to the analysis, as reported by the review authors within risk of bias tables, and specifically considering the proportion of participants within the analysis that was judged at high risk of bias for blinding of outcome assessor and allocation concealment;

  • differential attrition;

  • publication bias, as assessed by domain 2 of the ROBIS tool;

  • imprecision, considering both the CIs and sample size;

  • inconsistency, based on heterogeneity within the analysis, as determined by the I² statistic; and

  • indirectness, determined by considering important issues of clinical heterogeneity, specifically whether data from participants with SUI and UUI were combined and data from women who were antenatal and postnatal were combined.

Details of all the criteria within the algorithm, the downgrades applied based on the data arising from each relevant analysis within the overview, and justification for the cut‐offs adopted within the algorithm are provided in Appendix 3. There were two reviews for which we did not use the algorithm, and instead we used the GRADE assessments as reported by review authors; reasons for this difference are outlined in Differences between protocol and review.

Data synthesis

Intervention categorisation

We categorised conservative interventions addressed by each review as shown in Figure 1. We categorised the interventions investigated in each included study at three levels, as described in Appendix 4. We categorised all interventions for each study within each relevant comparison using the described levels. Where this related to a primary outcome of interest, this was completed by two overview authors independently (CH, DM), with any differences resolved through discussion. Where this related to a secondary outcome of interest, this was performed by one overview author (DM) and checked by a second (CH). We had planned to use different layers of intervention categorisation to present and describe interventions at various points within analyses and maps but did not complete these (see Differences between protocol and review).

Exploration of subgroups

The objective of this overview was to systematically synthesise the results of data pooled within reviews relating to conservative interventions for different types of UI. As part of this objective, we planned to explore existing data relating to different subgroups of women. We did not plan to carry out any statistical subgroup comparisons ourselves but rather to extract and tabulate results of relevant analyses from the included reviews where these were presented for one or more of our predefined subgroups (see Differences between protocol and review). However, none of the included reviews reported any statistical subgroup analyses based on any of our predefined subgroups and, consequently, no exploration of subgroups is presented within this overview.

Statistical analysis

We planned to analyse available data relating to comparisons within the included reviews for the three populations of interest (SUI, UUI and AUI) and the two primary outcomes (symptomatic cure or improvement of UI and condition‐specific quality of life). We planned to summarise included data within forest plots and network maps but have not included these in the final overview to simplify and enhance accessibility, and promote sustainability, of this overview (see Differences between protocol and review).

Results

Description of included reviews

Results of the search

We considered 1944 titles of Cochrane Reviews and protocols after removal of duplicates and 57 full‐text articles (Figure 2). Sixteen full‐text articles did not meet our inclusion criteria and were excluded (see Appendix 5 for references and further characteristics of these excluded reviews).


PRISMA study flow diagram. AUI: all types of urinary incontinence; n: number of records; SUI: stress urinary incontinence; UUI: urgency urinary incontinence.

PRISMA study flow diagram. AUI: all types of urinary incontinence; n: number of records; SUI: stress urinary incontinence; UUI: urgency urinary incontinence.

The remaining 41 full texts met our inclusion criteria and were included. Twenty‐nine were completed reviews (Alhasso 2005; Ayeleke 2015; Cody 2012; Dumoulin 2018; Duthie 2011; Eustice 2000; Freites 2019; Glazener 2017a; Glazener 2017b; Hay‐Smith 2011; Herbison 2009; Herbison 2013; Herderschee 2011; Imamura 2015; Kang 2015; Kirchin 2017; Lapitan 2017; Lipp 2014; Mariappan 2005; Nambiar 2017; Ostaszkiewicz 2004a; Ostaszkiewicz 2004b; Rai 2012; Saraswat 2020; Stewart 2016; Stewart 2017; Wallace 2004; Wang 2013; Wieland 2019). Twelve were protocols (Chua 2015; Cotterill 2018; French 2010; Funada 2020; Hajebrahimi 2015; Hargreaves 2020; Khazali 2016; Lane 2020; Lins 2014; Ostaszkiewicz 2013; Reynard 2016; Yi 2014). However, two protocols were withdrawn prior to publication of this overview (Khazali 2016; Reynard 2016; see Appendix 6).

Characteristics of the 29 included reviews are provided in Table 1 (design and conduct) and Table 2 (results).

Open in table viewer
Table 1. Characteristics of included reviews: design and conduct

Study ID

Objective

Inclusion criteria

Population

Intervention

Planned comparisons

Study design

Databases searched

Alhasso 2005

To determine the effects of adrenergic agonists in the treatment of SUI.

Women with UI

Age range 18–90 years

Adrenergic agonist drug

Adrenergic agonist drug

  • vs placebo or no treatment

  • vs conservative therapies

  • vs surgery

  • higher dose vs a lower dose

  • vs another adrenergic agonist

  • vs alternative forms of pharmacotherapy

  • + another drug vs the other drug treatment alone

  • + another drug vs adrenergic agonist treatment alone

RCT
QRCT

Cochrane Incontinence's Specialised Register – 15 September 2010

Ayeleke 2015

To compare the effects of PFMT plus another active treatment vs the same active treatment alone in the management of women with UI.

Women with SUI, UUI or MUI

Age range 18–75 years

PFMT as a programme of repeated voluntary pelvic floor muscle contractions taught or supervised (or both) by healthcare professionals. All types of PFMT programmes

were considered for inclusion.

A: physical

  • PFMT + vaginal cones vs vaginal cones alone

B: behavioural

  • PFMT + lifestyle intervention (e.g. weight reduction) vs lifestyle intervention alone (lifestyle intervention must be structured or supervised)

  • PFMT + bladder training vs bladder training alone

C: electrical or magnetic

  • PFMT + ES vs ES alone

  • PFMT + magnetic stimulation vs magnetic stimulation alone

D: mechanical

  • PFMT + continence pessaries vs continence pessaries alone

E: drugs

  • PFMT + drug therapy vs drug therapy alone

F: surgery

  • PFMT before surgical intervention vs surgical intervention alone

G: other

  • PFMT + any other stand‐alone active treatment vs the same stand‐alone active treatment

RCT
QRCT

Cochrane Incontinence's Specialised Register – 5 May 2015 (update)t

Embase Classic and Embase 1947 – 7 March 2013 (original)
CINAHL – January 1982 and 6 May 2015 (update)
ClinicalTrials.gov. – 30 May 2013 (original)
WHO ICTRP 3 June 2013 (original)

Cody 2012

To assess the beneficial and harmful effects of oestrogen therapy used for the treatment of UI.

Postmenopausal women with UI and diagnosed as having SUI, UUI or MUI

Age not reported

Oestrogen therapy (different types of oestrogens, different doses and different routes of administration)

Oestrogen therapy

  • vs placebo or no treatment

  • vs other forms of treatment

  • + other therapy vs placebo or no treatment

  • + another treatment vs oestrogen

  • + another treatment vs that other treatment

  • 1 type of oestrogen vs another

  • 1 method of administration of oestrogen vs another

  • high‐dose vs lower dose

Cochrane Incontinence's Specialised Register – 21 June 2012

Dumoulin 2018

To assess the effects of PFMT for women
with UI vs no treatment, placebo or sham treatments, or other inactive control treatments.

All women with UI and diagnosed as having SUI, UUI or MUI

Overall age not addressed in overview

PFMT defined as a programme of repeated voluntary pelvic floor muscle contractions taught and supervised by a healthcare professional. All types of PFMT programmes were considered. Trials in which PFMT was combined with a single episode of biofeedback or advice on strategies for symptoms of urgency or frequency were eligible for inclusion.

Comparisons not listed explicitly. However, the objective was to determine the effects of PFMT for women with UI in comparison to no treatment, placebo

or sham treatments, or other inactive control treatments. In addition, 1 arm of all eligible trials included use of a PFMT programme to ameliorate symptoms of existing urine leakage, 1 received no treatment, 1 placebo, 1 sham treatment and 1 an inactive control treatment

RCT
QRCT

Cochrane Incontinence's Specialised Register – 12 February 2018

Duthie 2011

To compare intravesical botulinum toxin injection with other treatments for neurogenic and idiopathic OAB in adults with or without UI

Men and women diagnosed with idiopathic or neurogenic
OAB syndrome regardless of whether they also had SUI.

Intravesical botulinum toxin

Intravesical botulinum toxin

  • vs placebo

  • vs no treatment

  • vs pharmacological interventions

  • vs non‐pharmacological interventions

  • higher doses vs lower doses

  • + other treatments vs other treatments alone

  • 1 formulation vs another

  • 1 intravesical injection technique vs another

RCT
QRCT

Cochrane Incontinence's Specialised Register – 23 February 2010

Eustice 2000

To determine the effects of prompted voiding for the management
of UI.

Men and women with or without cognitive impairment diagnosed
as having UI

Mean age 84 years

Prompted voiding

Prompted voiding

  • vs no prompted voiding

  • vs other treatments

  • + another treatment vs that other treatment

RCT
QRCT

Cochrane Incontinence's Specialised Register – 31 January 2006

Freites 2019

To assess the effects of laparoscopic colposuspension for UI in women and summarise the principal findings of relevant economic evaluations of these interventions.

Women with SUI or MUI

Laparoscopic colposuspension

Laparoscopic colposuspension

  • vs no treatment or sham operation

  • vs conservative interventions (e.g. PFMT, ES, cones, biofeedback)

  • vs open colposuspension (abdominal surgery)

  • vs traditional sling procedures (abdominal and vaginal surgery)

  • vs mid‐urethral sling procedures (abdominal and vaginal surgery)

  • vs needle suspension (abdominal and vaginal surgery)

  • vs anterior vaginal repair

  • vs periurethral injections

  • 1 method of laparoscopic colposuspension vs another

RCT

QRCT

Cochrane Incontinence's Specialised Register – 22 May 2019

Glazener 2017a

To determine the effects of anterior vaginal repair (anterior colporrhaphy)
on SUI or MUI in women

Women with UI diagnosed as having: urodynamic UI (urodynamic diagnosis), SUI (clinical diagnosis) or MUI (any SUI + other urinary symptoms

Age not reported

Anterior vaginal repair (anterior colporrhaphy)

Anterior vaginal repair

  • vs no treatment or sham operation

  • conservative interventions

  • open abdominal retropubic suspensions for all women

  • vs abdominal surgery for women with co‐existent prolapse

  • bladder neck needle suspensions

  • suburethral sling procedures

  • laparoscopic colposuspensions

  • other methods of anterior vaginal repair

RCT
QRCT

Cochrane Incontinence's Specialised Register – 1 September 2009

Glazener 2017b

To determine the effects of needle suspension on SUI or MUI in comparison with other management options.

Women with SUI MUI

Needle suspension

Needle suspension

  • vs no treatment or sham operation

  • vs conservative interventions (e.g. PFMT, ES, cones, biofeedback)

  • vs open abdominal retropubic suspension (abdominal surgery)

  • vs suburethral sling procedures (abdominal and vaginal surgery)

  • vs anterior vaginal repair (anterior colporrhaphy, vaginal surgery)

  • vs laparoscopic retropubic suspension

  • vs periurethral injections

  • vs drug treatment (e.g. adrenergic agonists)

  • 1 method of needle suspension vs another

  • 1 suture type vs another for carrying out needle suspension operations

RCT

QRCT

Cochrane Incontinence's Specialised Register – 12 November 2014

Additional searches conducted for a brief economic commentary in April 2017

Hay‐Smith 2011

To compare the effects of different approaches to PFMT for women with UI.

Women with UI diagnosed as having SUI, UUI or MUI on the basis of symptoms, signs or urodynamic evaluation, as defined by the study authors.

PFMT defined as any programme of repeated voluntary pelvic floor muscle contractions, or 'indirect' voluntary pelvic floor muscle contraction irrespective of variations in purpose and training parameters

  • More vs less contact with health professionals

  • Group vs individual supervision of PFMT

  • Direct vs indirect methods of PFMT

  • Individualised vs generic PFMT

  • Near maximal vs submaximal contractions

  • Daily vs 3 times per week PFMT

  • Upright and supine vs supine exercise positions alone

  • Strength and motor learning vs motor learning PFMT alone

  • PFMT and abdominal muscle exercise vs PFMT alone

  • PFMT with intravaginal resistance device vs PFMT alone

  • PFMT and adherence strategy vs PFMT alone

  • More‐intensive vs less‐intensive PFMT programmes

RCT
QRCT

Cochrane Incontinence's Specialised Register ‐– 17 May 2011

Herbison 2009

To determine the effects of implantable ES devices in the treatment of urine storage and voiding problems.

People with

  • refractory UUI (non‐neurogenic)

  • OAB syndrome (symptoms of urgency or increased daytime frequency, or both, or nocturia)

  • urinary retention (voiding problem) with functional urethral sphincter

An implanted device to provide stimulation to the sacral nerve(s).

Sacral nerve stimulation

  • + implanted devices vs no treatment

  • + implanted devices vs other treatments

  • 1 method of providing the stimulation (e.g. surgical technique) vs another

RCT
QRCT

Cochrane Incontinence's Specialised Register (10 February 2009)

CENTRAL (March 2008)

MEDLINE (March 2008)

Embase (March 2008)

CINAHL (March 2008)

Herbison 2013

To determine the effectiveness of vaginal cones in the management of female SUI

Women whose predominant complaint was SUI, diagnosed either by symptom classification or urodynamics.

Premenopausal, postmenopausal or

3 months' postpartum

Weighted vaginal cones following a standardised (within trial) protocol.

Vaginal cones

  • vs no treatment

  • vs other conservative therapies

  • vaginal cones + another conservative therapy vs another conservative therapy alone or cones alone

  • vs non‐conservative methods, e.g. surgery or injectables

RCT
QRCT

Cochrane Incontinence's Specialised Register – 19 September 2012

MEDLINE –

February 2013

Embase – February 2013

Herderschee 2011

To determine whether feedback (including biofeedback) provides additional benefit to PFMT in women with UI (SUI, UUI or MUI), regardless of cause.

Women of all ages with SUI, UUI or MUI, diagnosed by symptoms

(reported by the woman), signs (as reported or observed by the healthcare professional) or urodynamics, regardless of cause.

Age: 20–80 years

≥ 1 PFMT arm had to include a form of feedback (or biofeedback) to teach, modulate or encourage pelvic floor muscle contractions

  • PFMT + biofeedback vs PFMT alone

  • PFMT + feedback vs PFMT alone

  • PFMT + feedback + biofeedback vs PFMT alone

  • PFMT + biofeedback vs PFMT + feedback

  • PFMT + 1 type of biofeedback vs PFMT + another type of biofeedback

RCT
QRCT

Cochrane Incontinence's Specialised Register – 13 May 2010

Imamura 2015

To determine the effectiveness of specific lifestyle interventions (i.e. weight loss; dietary changes; fluid intake; reduction in caffeinated, carbonated and alcoholic drinks; avoidance of constipation; stopping smoking; and physical activity) in the management of adult UI.

Adults with UI, diagnosed either by symptom classification (SUI; UUI; MUI) or by urodynamic investigation (urodynamic SUI where not all participants had UI at baseline

Age range (mean) 49–70 years

A community‐based lifestyle intervention following a standardised (within trial) protocol.

Not clearly stated, but the methods stated: "Comparison interventions included no (active) treatment, other conservative physical therapies such as pelvic floor muscle training
(PFMT) or bladder training, or pharmacological therapies."

RCT
QRCT

Cochrane Incontinence's Specialised Register – 3 July 2013

Note: a further update was performed on 27 October 2014 but data entered only into 'studies awaiting classification'.

Kang 2015

To evaluate the efficacy of transurethral radiofrequency collagen denaturation, compared with other interventions, in the treatment of

women with UI.

Women with SUI or MUI diagnosed clinically or
using urodynamics.

Transurethral radiofrequency collagen denaturation

Transurethral radiofrequency collagen denaturation

  • vs no treatment/sham treatment

  • vs conservative physical treatment

  • vs mechanical devices

  • vs drug treatment

  • vs injectable treatment for UI

  • vs other surgery for UI

RCT, QRCT

quote: "excluded cluster‐randomised and cross‐over trials."

Cochrane Incontinence Group Specialised

Register

Embase and Embase Classic

Google scholar

Contacted manufacturers

US FDA website

Current Controlled Trials

ClinicalTrials.gov

WHO International Clinical Trials

Registry Platform

Handsearches of relevant conference abstracts – 19 December 2014

Kirchin 2017

To determine the effects of periurethral and transurethral bulking agents on cure or improvement of UI in women

Women with UI.

Classification of diagnoses was as defined by the study authors.

Age not reported

Urethral injection therapy

Urethral injection therapy

  • vs no treatment

  • vs non‐surgical management

  • vs other surgical managements

  • 1 material for injectable treatment vs another

  • 1 route of injection vs another route

RCT
QRCT

Cochrane Incontinence's Specialised Register – 8 November 2010

MEDLINE – January 1996 to March 2017

Embase – January 1980 to April 2017

NHS EED – April 2017

Lapitan 2017

To determine the effects of open retropubic colposuspension for the treatment of UI in women.

Women with SUI or MUI

Open retropubic colposuspension

Open retropubic colposuspension

  • vs no treatment or sham operation

  • vs conservative interventions

  • vs drug therapy

  • vs other surgical techniques

RCT

QRCT

Cochrane Incontinence's Specialised Register – 5 May 2015

Additional searches conducted for a brief economic commentary in April 2017

Lipp 2014

To determine the effects of mechanical devices in the management of adult female UI, particularly SUI.

Women with UI diagnosed as having SUI, UUI or other incontinence either by symptom classification or by urodynamic diagnosis, as defined by the study authors.

Age not mentioned under participants. Individual trials reported mean or range (44–73 years)

Interventions using mechanical devices designed to control urinary leakage by being inserted: within the vagina; within the urethra or applied to the external surface of the urethra.

A mechanical device

  • vs no treatment

  • vs another mechanical device

  • vs other treatments

RCT
QRCT

Cochrane Incontinence's Specialised Register – 21 August 2014

Embase – 26 August 2014

CINAHL – 26 August 2014

Mariappan 2005

To determine the effects of SNRI in the management of SUI and MUI that includes SUI in adults.

Women and men with UI diagnosed as having SUI (clinical diagnosis), urodynamic SUI (urodynamic diagnosis) or MUI

Age 49–54 years

SNRIs

SNRI

  • vs placebo or 'open' no treatment

  • vs conservative non‐pharmacological therapies

  • vs surgery

  • vs other classes of drugs

  • vs another SNRI

  • SNRI given orally vs given via another route

  • higher dose vs lower dose

  • + another drug vs the other drug alone

  • + conservative non‐pharmacological therapy vs the conservative non‐pharmacological therapy alone

RCT
QRCT

Cochrane Incontinence Specialised Register – 5 March 2007

CENTRAL – the Cochrane Library 2006 – 14 November 2006

MEDLINE – 7 February 2007
MEDLINE In‐Process & Other Non‐Indexed Citations – 7 February 2007

Nambiar 2017

To assess the effectiveness of mini‐sling procedures in women with urodynamic clinical SUI or MUI in terms of improved continence status, quality of life or adverse events.

Women with SUI, or MUI if women had stress‐predominant symptoms

Single‐incision sling

Single‐incision slings

  • vs no treatment

  • vs conservative treatment

  • vs colposuspension

  • vs laparoscopic procedures

  • vs traditional suburethral slings

  • vs retropubic minimally invasive slings

  • vs obturator minimally invasive slings

  • 1 single‐incision sling vs another

RCT

QRT

Cochrane Incontinence's Specialised Register – 6 February 2013

ClinicalTrials.gov and WHO ICTRP – 20 September 2012

Additional searches conducted for a brief economic commentary in April 2017

Ostaszkiewicz 2004a

To assess the effects of habit retraining for the management of

UI in adults

Men and women with or without cognitive impairment diagnosed either by symptom classification or by urodynamic study as having UI.

Women mean age 80.3 years

Men mean age 83 years

Habit retraining

Habit retraining

  • vs no habit retraining

  • vs another intervention

  • + another intervention vs that intervention alone

  • + another intervention vs habit retraining alone

  • + another intervention vs usual care

RCT
QRCT

Cochrane Incontinence Specialised Register – 2 April 2009

MEDLINE – 15 February 2004

Embase – 9 May 2002

CINAHL – March 2001

PsycINFO – August 2002

Current Contents 2001

Ostaszkiewicz 2004b

To assess the effects of timed voiding for the management of UI in adults.

Men and women with or without cognitive impairment diagnosed either by symptom classification or urodynamic study as having UI.

Mean age 86 years

Timed voiding alone or + other interventions.

While acknowledging variability in the definition of timed voiding, this review of timed voiding applied the definitions proposed by Hadley 1986 and by the Agency for Healthcare Research and Quality (Fantl 1991). In this context, timed voiding is distinguished from bladder training, prompted voiding and habit training.

Timed voiding

  • vs no timed voiding

  • vs other interventions

  • + another intervention vs other intervention alone

  • + another intervention vs timed voiding alone

  • + another intervention vs usual care

RCT
QRCT

Cochrane Incontinence Specialised Register – 2 April 2009

MEDLINE – January 1966 to November
2003 (performed 15 November 2003)

Embase – 1980–2002 week 18 (9 May 2002)

PsycINFO – January 1972 to August 2002 (August 2002)

CINAHL – January 1982 to February 2001 (performed in March 2001)

Biological abstracts – January 1980 to December 2000 (performed March 2001)

Current Contents – January 1993–2001 (performed March 2001)

DARE – Issue 1, 2001

Rai 2012

To compare the effects of anticholinergic drugs with various non‐pharmacological therapies for non‐neurogenic OAB syndrome in adults.

Men and women with idiopathic OAB syndrome

or UUI (symptomatic or urodynamic

diagnosis, or both)

Age not provided in review and often not in individual trials

Anticholinergic drug

Anticholinergic drugs

  • vs bladder training alone

  • vs PFMT alone

  • vs external ES

  • vs surgery

  • + non‐drug therapies vs non‐drug therapies alone

  • + non‐drug therapies vs anticholinergic drugs alone

  • vs combination of non‐drug therapies

RCT
QRCT

Cochrane Incontinence Specialised Register – 4 September 2012

Saraswat 2020

To assess the effects of traditional suburethral sling procedures for treatment of urodynamic SUI (urodynamic diagnosis), or for symptoms of SUI or MUI (clinical diagnosis) in women.

Women with SUI due to hypermobility or intrinsic sphincter deficiency (or both, diagnosed clinically or with urodynamics, or with MUI.

Age not provided in review or often in individual trials

Traditional suburethral sling procedures to treat SUI or MUI

Traditional suburethral sling operation

  • vs no treatment or sham operation

  • vs conservative management

  • vs drugs

  • vs injectables

  • vs colposuspension

  • vs bladder neck needle suspension

  • sling operation vs anterior repair

  • vs laparoscopic procedures

  • vs a minimally invasive sling

  • 1 type of traditional sling operations vs another traditional sling operation

RCT
QRCT

Cochrane Incontinence Specialised Register – 27 February 2017

(quote: "A further updated search was conducted on 23 January 2019, the results of which were not fully incorporated into the review.")

Stewart 2016

To determine the effectiveness of ES

Adults aged ≥ 18 years, or according to study authors' definitions of adult) with:

  • symptomatic diagnosis of OAB, UUI or MUI, or

  • urodynamic diagnosis of detrusor overactivity in addition to OAB symptoms (urgency, frequency or episodes of urgency incontinence)

Age reported in review 46–70 years, though 9 trials did not report participants' ages.

Some trials included men and women but majority women.

Neuromodulation

with ES targeting specific

nerves in the sacral plexus that control pelvic floor function. Routes were rectal, vaginal and posterior tibial nerve

ES

  • vs no active treatment, placebo or sham treatment

  • vs conservative treatment (e.g. bladder training, PFMT, biofeedback, magnetic stimulation)

  • vs drugs (e.g. anticholinergics)

  • vs surgery (including botulinum toxin);

  • + another treatment vs other treatment alone

  • + another treatment vs no active treatment, placebo or sham treatment

  • 1 type of ES vs another

RCT

QRCT

Cochrane Incontinence Specialised Register – December 2014

Stewart 2017

To assess the effects of ES with non‐implanted devices, alone or in combination with other treatment, for managing SUI or stress‐predominant MUI in women.

Women with SUI or stress‐predominant MUI on the basis of symptoms, signs or urodynamic diagnosis.

ES with non‐implanted devices

ES

  • vs no active treatment

  • placebo or sham treatment

  • vs other conservative treatment (e.g. bladder training, PFMT, biofeedback, magnetic stimulation)

  • drugs (e.g. duloxetine)

  • vs surgery or injection of bulking agents

  • + another treatment vs the other treatment alone

  • 1 type of ES vs another

RCT

QRCT

Cochrane Incontinence's Specialised Register – 27 February 2017

Additional searches conducted for a brief economic commentary – 10 February 2016

Wallace 2004

To assess the effects of bladder training on UI, however that diagnosis was made

Men and women with UI

Age not reported in review

Range from individual trials 19–68 years

All participants were women except in 1 trial

Bladder training; the terms bladder retraining, bladder drill or bladder re‐education were assumed to be synonymous with bladder training.

Bladder training

  • vs no bladder training

  • vs other treatments

  • + another treatment (such as conservative or pharmacological) vs that other treatment alone

RCT
QRCT

Cochrane Incontinence Specialised Register – 15 March 2006

Wang 2013

To determine the effects and safety of acupuncture for SUI in adults.

People of any age or ethnicity, with a diagnosis of SUI (according to the definition of the International Continence Society) were eligible for inclusion.

Mean age 52 years

All women

Scalp acupuncture, body acupuncture, electroacupuncture, warm acupuncture, elongated needle, auricular acupuncture and fire needle.

Acupuncture

  • vs placebo or no treatment

  • vs any other treatment

RCT
QRCT

Cochrane Incontinence Specialised Register – 28 January 2013

Embase – 20 February 2013

AMED – 20 February 2013

Chinese Biomedical Literature Database – 20 February 2013
Chinese Acupuncture Trials Register – 20 February 2013

China National Knowledge Infrastructure – 20 February 2013

Wieland 2019

To assess the effects of yoga for treating UI in women.

Women aged ≥ 18 years diagnosed with UI on the basis of symptoms, signs or urodynamic evaluation.

Yoga (any yoga tradition but excluded studies in which the yoga intervention did not include a physical practice component).

Yoga

  • vs no specific active intervention (e.g. usual care, waiting list)

  • vs an active intervention (e.g. lifestyle intervention or PFMT) for which we considered different active comparators separately (e.g. yoga vs lifestyle advice, yoga vs PFMT)

  • + an intervention vs the same intervention without yoga (e.g. yoga as an add‐on intervention to PFMT vs PFMT alone)

RCT

Cochrane Incontinence's Specialised Register – 21 June 2018

Cochrane Complementary Medicine Field Specialised Register – 21 June 2018

Additional searches conducted for an economic evaluation – 22 January 2019

ES: electrical stimulation; FDA: Food and Drug Administration; MUI: mixed urinary incontinence; OAB: overactive bladder; PFMT: pelvic floor muscle training; QRCT: quasi‐randomised controlled trial; RCT: randomised controlled trial; SNRI: serotonin and noradrenaline reuptake inhibitors; SUI: stress urinary incontinence; UI: urinary incontinence; UUI: urge urinary incontinence; WHO: World Health Organization.

Open in table viewer
Table 2. Characteristics of included reviews: results

Study ID

Intervention focus

Included studies (number that contained a conservative intervention)

Participant details1, for participants within studies containing a conservative intervention

n

Age (years)

Sex

Pregnancy as an inclusion or exclusion criteria?

Type of UI

Alhasso 2005

To determine the effects of adrenergic agonists in the treatment of SUI

22 (2)

218

Range 30–90

100% F

No

SUI

Ayeleke 2015

To compare the effects of pelvic floor muscle training + another active treatment vs the same active treatment alone in the management of women with UI

13 (13)

1164

Range 18–75

100% F

No

SUI

MUI

UUI

Cody 2012

To assess the effects (both beneficial and harmful) of oestrogen therapy used for the treatment of UI

34 (3)

192

Not reported

100% F

Not reported

AUI

Dumoulin 2018

To determine the effects of PFMT for women with UI vs no treatment, placebo or sham treatments, or other inactive control treatments

31 (31)

1871

18 trials restricted participation based on ages. Lowest age limit 13 years. No upper age limit in several trials.

100% F

No

SUI 18 trials

MUI 1 trial

UUI 1 trial

AUI 11 trials

Duthie 2011

To compare intravesical botulinum toxin injection with other treatments for neurogenic and idiopathic OAB in adults with or without incontinence

19 (0)

0

Eustice 2000

To determine the effects of prompted voiding for the management of UI

9 (9)

674

Mean age 84

Range 69.3–92.4

100% female

No

Unclear

Freites 2019

To assess the effects of laparoscopic colposuspension for UI in women; and summarise the principal findings of relevant economic evaluations of these interventions

26 (0)

0

Glazener 2017a

To determine the effects of anterior vaginal repair (anterior colporrhaphy) on SUI or mixed UI in women

10 (1)

50

Mean 48

Range 31–66

100% female

Not reported

SUI

Glazener 2017b

To determine the effects of needle suspension on SUI or MUI in comparison with other management options

10 (0)

0

Hay‐Smith 2011

To compare the effects of different approaches to pelvic floor muscle training for women with UI

21 (1 trial with 3 arms)(21)

1490

Mean 35.9–58.4 (SD 10.8)

Range 36–80 (possibly)

100% female

No

SUI

UUI

MUI

Herbison 2009

To determine the effects of implantable ES devices in the treatment of urine storage and voiding problems

8 (0)

Herbison 2013

To determine the effectiveness of vaginal cones in the management of female SUI

23 studies

7 abstracts only, 1 ongoing (23)

1806

Not stated

100% female

No

SUI 22

Unclear 1

Herderschee 2011

To determine whether feedback (including biofeedback) provides additional benefit to PFMT in women with UI (SUI, UUI or MUI), regardless of cause

24 (24)

1583

All ages

Assume 100% female

No

SUI

MUI

UUI

Imamura 2015

To determine the effectiveness of specific lifestyle interventions (i.e. weight loss; dietary changes; fluid intake; reduction in caffeinated, carbonated and alcoholic drinks; avoidance of constipation; stopping smoking; and physical activity) in the management of adult UI

11 (11)

5974

49–58

(quotes: "unclear if this was a mean or median"), "except for two trials with means of 62.6 and 70.25 years"

100% female

Not an exclusion criteria

SUI

MUI

UUI

Kang 2015

To evaluate the efficacy of transurethral radiofrequency collagen denaturation, compared with other interventions, in the treatment of women with UI

1 (0)

Kirchin 2017

To determine the effects of periurethral and transurethral bulking agents on cure or improvement of UI in women.

35 (1)

47

Not reported

100% female

Not reported

SUI

Lapitan 2017

To determine the effects of open retropubic colposuspension for the treatment of UI in women

55 (2)

120

31–66 years in 1 study; not reported in 1 study

100% female

Not reported

SUI

Lipp 2014

To determine the effects of mechanical devices in the management of adult female UI, particularly SUI

8 (8)

787

Range 30–75

100% female

No

SUI

UUI

Other

Mariappan 2005

To determine the effects of SNRI in the management of SUI and MUI that includes SUI in adults

10 (1)

201 (possibly)

Range 18–75

100% female

(note: selection criteria included both sexes but data only available for female)

Exclusion

MUI

SUI

Nambiar 2017

To assess the effectiveness of mini‐sling procedures in women with urodynamic clinical stress or MUI in terms of improved continence status, quality of life or adverse events.

31 (0)

0

Ostaszkiewicz 2004a

To assess the effects of habit retraining for the management of UI in adults

4 (4)

378

Mean 80

Analysis on females

No

UUI

MUI

Unclear

Ostaszkiewicz 2004b

To assess the effects of timed voiding for the management of urinary incontinence in adults

2 (2)

298

Mean 86.7

100% female

No

UUI

MUI (possibly)

Rai 2012

To compare the effects of anticholinergic drugs with various non‐pharmacological therapies for non‐neurogenic OAB syndrome in adults

23 (23)

3695

6 had male and female 17 had only female

Not reported

Not reported

UUI

Saraswat 2020

To assess the effects of traditional suburethral sling procedures for treatment of urodynamic SUI

34 (0)

0

Stewart 2016

To determine the effectiveness of ES with non‐implanted electrodes in comparison with placebo or any other active treatment in adults with OAB, with or without UUI

51 (51)

3443

All trials included adults only. 1 trial included only participants aged > 65 years; mean age reported in this trial was 84 years. 9 trials did not report participants' age. Across the remaining trials, mean age ranged from 46 to 70

36/51 female only

11/51 male and female

1/51 male only

3/51 not reported

Not an exclusion criteria

OAB

UUI

MUI

Stewart 2017

To assess the effects of ES with non‐implanted devices, alone or in combination with other treatment, for managing SUI or stress‐predominant MUI in women

56 (56)

3781

Mean range 41–69

100% female

Not reported

SUI

MUI

AUI

Wallace 2004

To assess the effects of bladder training on UI, however that diagnosis was made

12

1473

5 trials no age restriction

Remaining trials 35–65

11/12 female

No

SUI

UUI

MUI

Wang 2013

Determine the effects and safety of acupuncture for SUI in adults

1

60

Mean 53.8

100% female

Not reported

SUI

Wieland 2019

To assess the effects of yoga for treating UI in women.

2 (2)

49

Range 22–79

100% female

1 study – pregnancy in last year was an exclusion criterion.

1 study – most were postmenopausal

UUI

AUI

AUI: all types of urinary incontinence; ES: electrical stimulation; F: female; M: male; MUI: mixed urinary incontinence; n: number of participants; OAB: overactive bladder; PFMT: pelvic floor muscle training; QRCT: quasi‐randomised controlled trial; RCT: randomised controlled trial; SNRI: serotonin and noradrenaline reuptake inhibitors; SUI: stress urinary incontinence; UUI: urgency urinary incontinence.
Number of participants are as reported in review text by review authors.1

Fifteen of the 29 included reviews were primarily focused on conservative interventions, while 14 were primarily focused on non‐conservative interventions but contained data (from 'control' groups) that were relevant to the criteria for this overview.

Interventions and comparisons

Physical therapies

Seven reviews focused on interventions categorised as physical therapies. PFMT was the intervention of interest in four reviews (Ayeleke 2015; Dumoulin 2018; Hay‐Smith 2011; Herderschee 2011). Vaginal cones was the intervention in one (Herbison 2013). In Dumoulin 2018, the comparison of interest was PFMT compared to no treatment, placebo or inactive control. In two of the PFMT‐focused reviews, the intervention of interest was different types or ways of delivering PFMT. Hay‐Smith 2011 compared different approaches to the delivery of PFMT, while Herderschee 2011 explored the addition of feedback or biofeedback to PFMT, compared to PFMT alone or PFMT plus a different type of feedback. Ayeleke 2015 compared PFMT combined with another active treatment with the same active treatment. Herbison 2013 explored the effect of vaginal cones delivered alone or in combination with other treatments compared with no treatment or other active interventions (conservative or non‐conservative). Stewart 2016 focused on the effect of electrical stimulation with non‐implanted devices for OAB syndrome and Stewart 2017 focused on electrical stimulation with non‐implanted devices for the management of SUI or predominantly SUI. In both reviews, the electrical stimulation intervention was compared to no active treatment, placebo or sham; to conservative treatments (PFMT); to drug treatments or a combination of these. Herbison 2009 focused on the effects of implantable electrical stimulation for UUI, OAB syndrome and urinary retention.

Thirteen reviews focused on non‐conservative interventions but included (or planned to include) trials in which the comparison comprised physical therapies (Alhasso 2005; Cody 2012; Duthie 2011; Freites 2019; Glazener 2017a; Glazener 2017b; Kang 2015; Kirchin 2017; Lapitan 2017; Mariappan 2005; Nambiar 2017; Rai 2012; Saraswat 2020). Rai 2012 also included trials in which groups received electrical stimulation and combinations of different physical therapies.

Educational, behavioural and lifestyle advice

Five reviews focused on interventions categorised as educational, behavioural or lifestyle advice. Three focused on habit training (Eustice 2000; Ostaszkiewicz 2004a; Wallace 2004). One focused on timed voiding (Ostaszkiewicz 2004b). One focused on lifestyle changes (Imamura 2015). Eustice 2000 compared prompted voiding to no prompting, while Ostaszkiewicz 2004a reported on three trials that tested habit retraining combined with other approaches against usual care and one compared habit retraining alone with habit retraining plus an electronic monitoring device. Wallace 2004 compared bladder training with no bladder training, drugs, PFMT and combinations of these. Ostaszkiewicz 2004a compared timed voiding plus additional interventions with usual care. Imamura 2015 compared weight loss programmes with a control intervention, change in fluid intake with no change and reduction in caffeinated drinks with no change, with one trial comparing a soy‐rich diet with a soy‐free diet.

One review focused on non‐conservative interventions but included some trials in which the comparison was bladder training (Rai 2012).

Mechanical devices

One review focused on the mechanical devices intervention category (Lipp 2014). This review compared a mechanical device with no treatment, one mechanical device with another, and one trial compared three groups: a mechanical device alone, behavioural therapy (PFMT) alone and behavioural therapy combined with a mechanical device.

Complementary interventions

One review focused on an intervention categorised as complementary, where the intervention was acupuncture versus midodrine, a drug for treating hypotension (Wang 2013).

Psychological therapies

None of the reviews focused on an intervention categorised as a psychological therapy.

Other interventions

One reviews focused on yoga (Wieland 2019).

Methodological quality of included reviews

Risk of bias of included reviews

See Table 3 and Figure 3 for a summary of the ROBIS assessments for the included reviews.

Open in table viewer
Table 3. Risk of bias

Reviews

ROBIS Phase 2

ROBIS Phase 3 signalling questions

1. Study

eligibility

criteria

2. Identification and selection of

studies

3. Data collection and study

appraisal

4. Synthesis and findings

Did the interpretation of findings address all the concerns identified in Domains 1 to 4

Was the relevance of identified studies to the review's research question appropriately considered?

Did the reviewers avoid emphasising results on the basis of their statistical significance?

Comments relating to Phase 3 signalling questions

Alhasso 2005

L

L

?4

L

PY

PY

PY

Ayeleke 2015

L

L

L

H9

PY

PY

PY

Cody 2012

L

L

?5

H10

PN

PY

PY

The concerns relating to domain 4 were not covered in the discussion.

Dumoulin 2018

L

L

L

H11

PY

PY

PY

Duthie 2011

L

L

L

H12

PN

PY

PY

The concerns relating to study variation have not been adequately discussed.

Eustice 2000

L

?3

H6

H13

PN

PY

PY

The issues relating to the search (Domain 3) and synthesis (Domain 4) were not addressed in the discussion.

Freites 2019

L

L

L

L

PY

PY

PY

Glazener 2017a

L

L

L

H14

PN

PY

PY

Domain 4 concerns were not adequately addressed in the discussion.

Glazener 2017b

L

L

L

H15

PY

PY

PY

Hay‐Smith 2011

L

L

L

H16

PY

PY

PY

Herbison 2009

L

L

L

L

PY

PY

PN

The discussion somewhat overemphasises the findings of the review. While authors noted the difficulties in study reporting and the possibility of duplicate publications, the implications of this are not conveyed within the discussion.

Herbison 2013

L

L

L

H17

PY

PY

PY

Herderschee 2011

L

L

L

H18

PN

PY

PY

Domain 4 concerns were not adequately addressed in the discussion

Imamura 2015

H1

L

?7

L

PN

PY

PY

The concerns relating the data‐driven nature of the review, and the change of criteria noted in Domain 1 were not adequately addressed.

Kang 2015

L

L

L

L

PY

PY

PY

Kirchin 2017

L

L

L

H19

PN

PY

PY

Domain 4 concerns were not addressed in the discussion.

Lapitan 2017

L

L

L

L

PY

PY

PY

Lipp 2014

L

L

L

H20

PY

PY

PY

Mariappan 2005

L

L

L

H21

PY

PY

PY

Nambiar 2017

L

L

L

L

PY

PY

PY

Ostaszkiewicz 2004a

L

L

?8

L

PY

PY

PY

Ostaszkiewicz 2004b

H2

L

?8

L

PY

PY

PY

Rai 2012

L

L

L

H22

PY

PY

PY

Saraswat 2020

L

L

L

L

PY

PY

PY

Stewart 2016

L

L

L

L

PY

PY

PY

Stewart 2017

L

L

L

L

PY

PY

PY

Wallace 2004

L

L

L

?23

PY

PY

PY

Wang 2013

L

L

L

L

PY

PY

PY

Wieland 2019

L

L

L

L

PY

PY

PY

TOTALS

L = 27

H = 2

? = 0

L = 28

H = 0

? = 1

L = 23

H = 1

? = 5

L = 14

H = 14

? = 1

PY = 22

PN = 7

PY = 29

PN = 0

PY = 28

PN = 1

Abbreviations:?: unclear risk of bias; H: high risk of bias; L: low risk of bias; PFMT: pelvic floor muscle training; PN: probably no; PY: probably yes.
1The review authors reported changes to inclusion criteria that were made after the review had started, made because only a low number of studies met the original inclusion criteria.
2Authors identified that the eligibility criteria were difficult to apply: quote: "The review was based on an a priori definition of timed voiding which was developed after an extensive review of the literature; nevertheless we struggled to categorize the interventions used from the descriptions given. Overlap between interventions was found with some interventions having features of more than one approach."
3The process of selection of studies was not fully described.
4There was a lack of clarity about the risk of bias assessment; it was unclear what criteria were used, and only one was reported.
5There was a lack of clarity about the included studies, with poorer details of non‐pharmacological interventions included as comparators.
6Concerns related to the risk of bias assessment – there was a lack of detail, and it appeared to be very limited in nature.
7Only two of the 11 included studies contributed to the analysis. The review authors reported that they tried to obtain the study results but these were not provided for some studies.
8There was a lack of information relating to identification and inclusion of studies (i.e. results of the search), and how risk of bias was assessed and by whom (what criteria were assessed).
9The authors identified trials at high risk of bias, but did not integrate this information into the narrative text describing and supporting the results.
10It appeared that relevant studies had been excluded from synthesis inappropriately – excluded studies appeared to include studies that would have been relevant but which data were unavailable, or outcomes not relevant. There was also a lack of information about how heterogeneity was explored, and no reported sensitivity analyses to explore risk of bias within trials.
11No sensitivity analyses were conducted in this synthesis. GRADE levels did not seem to have been downgraded in response to small sample size; and often only downgraded one level despite more than one identified risk.
12Between‐study variation and study risk of bias were not considered within the findings.
13The analysis that was conducted was correct, but not complete: it failed to address a number of important issues. It contained very little assessment of heterogeneity or risk of bias, and no attempts to ensure robust analyses.
14Between‐study variation and study risk of bias did not appear to have been considered within the findings.
15No GRADE judgements of certainty of evidence, or systematic exploration of study variation.
16The analysis did not fully explore heterogeneity, neither did it ensure the robustness of findings through funnel plots, etc. There also appeared to be some duplication in the subgroup analyses.
17There was high risk of bias due to lack of blinding and high attrition, but these were not addressed. Although authors stated it was not appropriate to perform sensitivity analyses as there were low numbers of included studies, there were analyses containing five trials that were arguably suitable.
18Heterogeneity and study quality did not appear to have been explored systematically within the synthesis and findings.
19The synthesis process did not address the different trial populations, and no sensitivity analyses were performed.
20There were three included studies with high attrition (Cornu 2012; Nielson 1995; Robinson 2003). This appeared to have been dealt with inconsistently and incompletely, with only data from Nielson 1995 being excluded from part of the synthesis. There was no explanation as to why sensitivity analyses to address risk of bias and heterogeneity were not performed.
21Although sensitivity analyses were described within the review methods, these were not conducted, and no rationale was provided.
22Although the grouping of data within the analysis did function as a sensitivity analysis, the quality of studies, especially in relation to attrition bias was not fully explored.
23The authors described a rigorous analysis plan; however, this was not conducted, which may have been an overly conservative approach. Clear information describing and justifying this decision was not provided.


Summary of ROBIS results for Phase 2, identifying concerns about bias in the review process

Summary of ROBIS results for Phase 2, identifying concerns about bias in the review process

Risk of bias relating to study eligibility criteria (ROBIS: phase 2, domain 1)

The risk of bias relating to the prespecification and application of study eligibility criteria was low across 27/29 reviews. Only two reviews were judged at high risk of bias in relation to study eligibility criteria (Imamura 2015; Ostaszkiewicz 2004b). Imamura 2015 changed the eligibility criteria after an initial search to increase the number of trials included in the review. In Ostaszkiewicz 2004b, the review authors encountered difficulties applying the eligibility criteria.

Risk of bias relating to identification and selection of studies (ROBIS: phase 2, domain 2)

The risk of bias relating to the identification and selection of studies for inclusion within the review was low across 28/29 reviews. In one review, there was insufficient information about the process with which they selected studies (Eustice 2000).

Risk of bias relating to data collection and study appraisal (ROBIS: phase 2, domain 3)

The risk of bias relating to the method of data collection and study appraisal was low in 23/29 reviews. There were concerns in Eustice 2000 about the comprehensiveness of the assessment of risk of bias of the included trials, and this was judged at high risk of bias. Imamura 2015 reported being unable to obtain all relevant trial data from trial authors, and we judged this at unclear risk of bias. The risk of bias was also unclear for Alhasso 2005, Cody 2012, Ostaszkiewicz 2004a, and Ostaszkiewicz 2004b, with lack of clear details reported in relation to data collection and study appraisal.

Risk of bias relating to synthesis and findings (ROBIS: phase 2, domain 4)

Fourteen reviews were at high risk of bias in relation to the methods and reporting of the syntheses. For the reviews judged at high risk of bias, concerns for 12/14 reviews largely related to lack of exploration of heterogeneity or consideration of risk of bias assessments using subgroup or sensitivity analyses (Ayeleke 2015; Dumoulin 2018; Duthie 2011; Eustice 2000; Glazener 2017a; Glazener 2017b; Hay‐Smith 2011; Herbison 2013; Herderschee 2011; Kirchin 2017; Mariappan 2005; Rai 2012), while concerns for 2/14 reviews related to the studies that were included in (or excluded from) analyses (Cody 2012; Lipp 2014). Of the remaining 15 reviews, 14 were judged at low risk of bias for this domain and, for one, there was insufficient information to make a judgement (Wallace 2004).

Risk of bias of the systematic review (ROBIS phase 3)

As discussed in the Methods, overview authors were unable to reach consensus on judgements relating to the interpretation of the review findings. The agreed responses to the ROBIS phase 3 signalling questions are reported rather than an overall judgement of risk of bias for each review (see Table 3). However, the signalling questions identified that there were no serious concerns for 21 reviews (Alhasso 2005; Ayeleke 2015; Dumoulin 2018; Freites 2019; Glazener 2017b; Hay‐Smith 2011; Herbison 2013; Kang 2015; Lapitan 2017; Lipp 2014; Mariappan 2005; Nambiar 2017; Ostaszkiewicz 2004a; Ostaszkiewicz 2004b; Rai 2012; Saraswat 2020; Stewart 2016; Stewart 2017; Wallace 2004; Wang 2013; Wieland 2019).

There were some concerns relating to the interpretation of the review findings for the remaining eight reviews. Of these, there were concerns in seven reviews that the interpretation of the findings did not address all of the concerns relating to bias in the review process (Cody 2012; Duthie 2011; Eustice 2000; Glazener 2017a; Herderschee 2011; Imamura 2015; Kirchin 2017). In one, there were concerns that the review authors may have emphasised results on the basis of their statistical significance (Herbison 2009; see Table 3).

Quality and quantity of evidence in included reviews

Seven reviews contained no trials with data relevant to the criteria for this overview (Duthie 2011; Freites 2019; Glazener 2017b; Herbison 2009; Kang 2015; Nambiar 2017; Saraswat 2020). Four reviews pooled data from relevant trials within analyses but did not have any analyses relating to the primary outcomes for this overview (Cody 2012; Eustice 2000; Ostaszkiewicz 2004a; Ostaszkiewicz 2004b). From the remaining 18 reviews, we extracted data on every relevant comparison for which there was an analysis relating to one of our primary outcomes of woman's observations of symptomatic cure or improvement of UI or condition‐specific quality of life.

We extracted data from 192 analyses relating to one of our primary outcomes: 134 analyses relating to woman's observation of symptomatic cure or improvement and 58 analyses relating to condition‐specific quality of life. Eighty‐one per cent of these analyses only contained one trial. These analyses contained data from 112 unique trials (which included 8975 women); data from 17 of these trials were included in more than one review (see Table 4).

Open in table viewer
Table 4. Trials included within reviews

Trial ID

Number of participants in comparison (total)

Number of reviews trial is cited in

Reviews trial is cited in

1st review

2nd review

3rd review

Amaro 2006

40

1

Stewart 2016

Aaronson 1995

38

2

Rai 2012

Stewart 2017

Arruda 2008

77

1

Stewart 2016

Arvonen 2001

37

1

Herbison 2013

Arvonen 2002

17

1

Herbison 2013

Asklund 2017

123

1

Dumoulin 2018

Baker 2014

20

1

Wieland 2019

Barroso 2004

36

1

Stewart 2016

Berglund 1993

20

1

Herbison 2013

Bertotto 2017

49

1

Dumoulin 2018

Beuttenmuller 2010

50

2

Dumoulin 2018

Stewart 2017

Bezerra 2009

34

1

Ayeleke 2015

Bi 2007

60

1

Wang 2013

Bø 1990

52

1

Hay‐Smith 2011

Bø 1999

122

3

Dumoulin 2018

Herbison 2013

Stewart 2017

Boos 1998

101

1

Lipp 2014

Bourcier 1994

102

1

Stewart 2017

Bridges 1988

69

1

Stewart 2017

Burgio 2002

122

1

Herderschee 2011

Burgio 2008

307

1

Rai 2012

Burgio 2010

58

2

Ayeleke 2015

Rai 2012

Burns 1993

82

1

Herderschee 2011

Cammu 1998

60

1

Herbison 2013

Carneiro 2010

50

1

Dumoulin 2018

Castro 2008

50

3

Dumoulin 2018

Herbison 2013

Stewart 2017

Celiker Tosun 2015

121

1

Dumoulin 2018

Colombo 1995

75

2

Rai 2012

Wallace 2004

Cornu 2012

55

1

Lipp 2014

Correia 2013

30

1

Stewart 2017

Correia 2014

48

1

Stewart 2017

de Oliveira 2009

60

1

Hay‐Smith 2011

Delgado 2009

40

1

Hay‐Smith 2011

Diniz Zanetti 2007

44

1

Hay‐Smith 2011

Diokno 2010

41

1

Dumoulin 2018

Eyjólfsdóttir 2009

24

1

Stewart 2017

Fantl 1991

123

1

Wallace 2004

Felicíssimo 2010

59

1

Hay‐Smith 2011

Finazzi‐Agrò 2010

35

1

Stewart 2016

Firra 2013

63

1

Stewart 2016

Franzén 2010

72

1

Stewart 2016

Goode 2002

125

2

Dumoulin 2018

Rai 2012

Goode 2003

87

2

Herderschee 2011

Stewart 2017

Hahn 1991

20

1

Stewart 2017

Haig 1995

58

1

Stewart 2017

Hay‐Smith 2002

123

1

Hay‐Smith 2011

Henalla 1989

51

1

Stewart 2017

Herbison 2004

34

1

Wallace 2004

Hofbauer 1990

21

2

Ayeleke 2015

Stewart 2017

Huang 2014

18

1

Wieland 2019

Hung 2010

64

1

Hay‐Smith 2011

Ishiko 2000

37

1

Alhasso 2005

Johnson 2000

20

1

Herderschee 2011

Kargar Jahromi 2015

48

1

Dumoulin 2018

Kaya 2011

46

1

Stewart 2016

Kaya 2014

108

1

Ayeleke 2015

Klarskov 1986

52

2

Glazener 2017a

Lapitan 2017

Knight 1998

70

1

Stewart 2017

Konstantinidou 2007

22

1

Hay‐Smith 2011

Kosilov 2013

229

1

Stewart 2016

Lagro‐Janssen 1991

18

1

Wallace 2004

Lagro‐Janssen 1992

66

1

Dumoulin 2018

Laycock 1988

36

1

Stewart 2017

Laycock 1993

40

1

Stewart 2017

Leong 2015

55

1

Dumoulin 2018

Lopès 2014

163

1

Stewart 2017

Luber 1997

54

1

Stewart 2017

Macaulay 1987

29

1

Rai 2012

McLean 2013

35

1

Dumoulin 2018

Milani 1986

75

1

Rai 2012

Millard 2004

475

1

Rai 2012

Morkved 2003

94

1

Herderschee 2011

Nascimento‐Correia 2012

45

1

Dumoulin 2018

Oláh 1990

54

2

Herbison 2013

Stewart 2017

Oldham 2013

124

2

Stewart 2016

Stewart 2017

Pages 2001

40

1

Herderschee 2011

Park 2002

50

1

Rai 2012

Patil 2010

110

1

Stewart 2017

Peattie 1988

33

1

Herbison 2013

Pereira 2012

14

1

Stewart 2017

Phelan 2012

738

1

Imamura 2015

Pieber 1995

46

1

Herbison 2013

Preyer 2007

31

1

Stewart 2016

Ramsay 1990

44

1

Hay‐Smith 2011

Richter 2007

299

2

Ayeleke 2015

Lipp 2014

Sancaktar 2010

40

1

Stewart 2016

Santos 2009

45

2

Herbison 2013

Stewart 2017

Sar 2009

34

1

Dumoulin 2018

Schagen van Leeuwen 2004

101

3

Ayeleke 2015

Hay‐Smith 2011

Mariappan 2005

Schmidt 2009

32

1

Stewart 2017

Schreiner 2010

52

1

Stewart 2016

Seo 2004

120

2

Herbison 2013

Stewart 2017

Solberg 2016

20

1

Dumoulin 2018

Song 2006

58

1

Rai 2012

Souto 2014

75

1

Stewart 2016

Sran 2016

48

1

Dumoulin 2018

Sriboonreung 2011

40

1

Hay‐Smith 2011

Subak 2005

40

1

Imamura 2015

Sugaya 2003

41

1

Hay‐Smith 2011

Tapp 1989

68

1

Stewart 2017

ter Meulen 2009

45

1

Kirchin 2017

Terlikowski 2013

102

1

Stewart 2017

Vecchioli‐Scaldazza 2013

40

1

Stewart 2016

Wang 2004

120

2

Stewart 2016

Herderschee 2011

Wang 2006

47

1

Rai 2012

Wells 1991

157

1

Alhasso 2005

Wells 1999

286

1

Hay‐Smith 2011

Whitmore 1995

52

1

Stewart 2017

Williams 2006

158

1

Herbison 2013

Wing 2010

304

1

Imamura 2015

Wilson 1987

30

2

Herderschee 2011

Stewart 2017

Wyman 1998

128

2

Ayeleke 2015

Wallace 2004

Yoon 2003

25

1

Wallace 2004

Total = 112 trials

Total participants = 8975

Mean participants per study = 80

Range 14–738, standard deviation 94

Trials in 1 review = 93

Trials in 2 reviews = 16

Trials in 3 reviews = 3

We assessed the certainty of evidence synthesised for each of these extracted analyses using the GRADE approach, and summarise these results below.

Stress urinary incontinence

Fourteen reviews presented analyses relating to primary outcome data in populations of women with SUI (Alhasso 2005; Ayeleke 2015; Dumoulin 2018; Glazener 2017a; Hay‐Smith 2011; Herbison 2013; Herderschee 2011; Kirchin 2017; Lapitan 2017; Lipp 2014; Mariappan 2005; Stewart 2017; Wallace 2004; Wang 2013). These 14 reviews contained 84 analyses relating to a primary outcome relevant to this overview: 17 compared a conservative intervention with control; 54 compared one conservative intervention with another conservative intervention; and 13 compared a conservative intervention with a non‐conservative intervention.

Table 5 illustrates the number of relevant analyses within these reviews and the judgement of certainty of evidence for each of these analyses. Of the 84 analyses, seven were high‐certainty, 42 were moderate‐certainty, 30 were low‐certainty and one was very low‐certainty. For four of the analyses, the effect was not estimable and we did not assign a certainty of evidence grade.

Open in table viewer
Table 5. Stress urinary incontinence: quantity and certainty of evidence

Comparison

Outcome

Certainty of the evidence

Not estimable

Total analyses

Table

High

Moderate

Low

Very low

Conservative vs

control

Cure or

improvement

2

1

2

0

0

5

Table 8

Quality of life

1

5

6

0

0

12

Table 9

Conservative vs

conservative

Cure or

improvement

3

22

15

1

2

43

Table 10

Quality of life

0

6

4

0

1

11

Table 11

Conservative vs

non‐conservative

Cure or

improvement

0

6

3

0

0

9

Table 12

Quality of life

1

2

0

0

1

4

Table 13

TOTALS

7

42

30

1

4

84

Urgency urinary incontinence

Five reviews presented analyses relating to a primary outcome in populations of women with UUI (Ayeleke 2015; Herderschee 2011; Rai 2012; Stewart 2016; Wallace 2004). However, Stewart 2016 was the only review to present any data relating to quality of life.

These five reviews contained 47 analyses relating to a primary outcome relevant to this overview: eight comparing a conservative intervention with control; 14 comparing one conservative intervention with another conservative intervention; and 25 comparing a conservative intervention with a non‐conservative intervention.

Table 6 illustrates the number of relevant analyses within these reviews and the judgement of certainty of evidence for each of these analyses. Of the 47 analyses, four were high‐certainty evidence, 25 were moderate‐certainty evidence and 18 were low‐certainty evidence.

Open in table viewer
Table 6. Urgency urinary incontinence: quantity and certainty of evidence

Comparison

Outcome

Certainty of the evidence

Not estimable

Total analyses

Table

High

Moderate

Low

Very low

Conservative vs

control

Cure or

improvement

2

4

1

0

0

7

Table 14

Quality of life

0

1

0

0

0

1

Table 15

Conservative vs

conservative

Cure or

improvement

0

7

3

0

0

10

Table 16

Quality of life

1

2

1

0

0

4

Table 17

Conservative vs

non‐conservative

Cure or

improvement

0

9

13

0

0

22

Table 18

Quality of life

1

2

0

0

0

3

Table 19

TOTALS

4

25

18

0

0

47

All types of urinary incontinence

Thirteen reviews presented analyses relating to a primary outcome in populations of women with AUI (Alhasso 2005; Ayeleke 2015; Dumoulin 2018; Hay‐Smith 2011; Herbison 2013; Herderschee 2011; Imamura 2010; Lipp 2014; Rai 2012; Stewart 2016; Stewart 2017; Wallace 2004; Wieland 2019). These 13 reviews contained 61 relevant analyses: 16 comparing a conservative intervention with control; 40 comparing one conservative intervention with another conservative intervention and five comparing a conservative intervention with a non‐conservative intervention.

Table 7 illustrates the number of relevant analyses within these reviews and the judgement of certainty of evidence for each. Of the 61 analyses, six were high‐certainty evidence, 14 were moderate‐certainty evidence, 32 were low‐certainty evidence and nine were very low‐certainty evidence.

Open in table viewer
Table 7. All types of urinary incontinence: quantity and certainty of evidence

Comparison

Outcome

Certainty of the evidence

Not estimable

Total analyses

Table

High

Moderate

Low

Very low

Conservative vs

control

Cure or

improvement

3

4

3

1

0

11

Table 20

Quality of life

2

2

0

1

0

5

Table 21

Conservative vs

conservative

Cure or

improvement

1

3

14

6

0

24

Table 22

Quality of life

0

3

12

1

0

16

Table 23

Conservative vs

non‐conservative

Cure or

improvement

0

1

2

0

0

3

Table 24

Quality of life

0

1

1

0

0

2

Table 25

TOTALS

6

14

32

9

0

61

Effect of interventions

This section reports the analyses identified to provide evidence relating to the effects of interventions for the primary outcomes of patient‐reported cure or improvement and condition‐specific quality of life for different types of UI and different comparators. This presents the results of the extracted data relating to SUI (84 analyses; see Table 5), UUI (47 analyses; see Table 6), and AUI (61 analyses; see Table 7).

For each of the outcome subsections, corresponding data are presented in a table. Where analyses were judged to provide low‐ or very low‐certainty evidence, a narrative description of the intervention comparisons was not provided in the text and readers are directed to the tables for this level of information.

Stress urinary incontinence

Any conservative intervention versus control, placebo or standard care
Symptomatic cure or improvement

For more information, see Table 8.

Open in table viewer
Table 8. Stress urinary incontinence: primary outcome participant‐reported cure or improvement; conservative versus control

Analysis code (in review)

Study ID

Comparisons

included

Timing of measure

Included trials

(n)

Effect

measure

Effect estimate

(CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades

(reason for downgrade)

High‐certainty evidence

1.1.1

Herbison 2013

Cones vs control

IM

Bø 19991

(n = 57)

RR

0.4 (0.2 to 0.6)

> 1 favours control

YES for cones

0

1.2.2

Herderschee 2011

PFMT + BF vs

PFMT alone (control)

IM

Goode 2003

(n = 87)

RR

0.6 (0.4 to 0.9)

> 1 favours PFMT alone

YES for PFMT and BF

0

Moderate‐certainty evidence

1.2.1

Dumoulin 2018

PFMT vs no treatment, placebo or control

IM

Asklund 2017; Bø 19991; Lagro‐Janssen 1992

(n = 242)

RR

6.3

(3.9 to 10.3)

> 1 favours PFMT

YES for PFMT

1(H)

Low‐certainty evidence

3.2.1

Hay‐Smith 2011

Direct vs indirect methods of PFMT

IM

Schagen van Leeuwen 2004; Ramsay 1990

(n = 138)

RR

0.7

(0.47 to 1.0)

> 1 favours indirect

NO

6(E,F,H×2)

2.2

Stewart 2017

ES vs sham treatment

IM

Hofbauer 1990Laycock 1993; Luber 1997; Terlikowski 2013; Whitmore 1995 (n = 236)

RR

2.03 (1.02 to 4.07)

> 1 favours ES

YES for ES

GRADE as assessed by review authors

1Bø 1999 was a four‐arm randomised controlled trial in which the groups were: control, PFMT, cones and ES. Data from PFMT and control arms were included in Dumoulin 2018, and data from cones and control arms were included in Herbison 2013.
Abbreviations: BF: biofeedback; CI: confidence interval; ES: electrical stimulation; IM: immediate; n: number of participants assessed NOT to have been cured or improved using a variety of different assessment methods; PFMT: pelvic floor muscle training; RR risk ratio.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Five analyses from five Cochrane Reviews present data on symptomatic cure and improvement in analyses comparing a conservative intervention with a control intervention (Dumoulin 2018; Hay‐Smith 2011; Herbison 2013; Herderschee 2011; Stewart 2017). There were 11 unique trials that compared a conservative intervention with a control intervention, with a cure or improvement outcome, at an immediate time point.

High‐ or moderate‐certainty evidence

Three analyses provided moderate‐ or high‐certainty evidence that PFMT had a beneficial effect on outcome compared to control (moderate‐certainty) and PFMT plus biofeedback had a beneficial effect on outcome compared to control (high‐certainty). Cones had a beneficial effect on outcome compared to control (high‐certainty).

Low‐ or very low‐certainty evidence

Two analyses provided low‐certainty evidence. Hay‐Smith 2011 described this analysis as a comparison of direct and indirect methods of delivering PFMT. Overview authors considered that the indirect methods delivered within the included trials met the criteria for being a control (or no treatment) intervention.

Condition‐specific quality of life

For more information, see Table 9.

Open in table viewer
Table 9. Stress urinary incontinence: quality of life, conservative versus control

Analysis code (in review)

Study ID

Comparisons included

Timing of measure

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

High‐certainty evidence

1.12.1

Dumoulin 2018

PFMT vs no treatment, placebo or control (I‐QoL)

IM

Castro 2008

(n = 50)

MD

−24.6 (−37.7 to −11.4)

> 0 favours control

YES favours PFMT

0

Moderate‐certainty evidence

1.5.1

Dumoulin 2018

PFMT vs no treatment, placebo or control

IM

Beuttenmuller 2010; Carneiro 2010; Nascimento‐Correia 20122

(n = 145)

MD

−11.89 (−20.5 to −3.2)

> 0 favours control

YES favours PFMT

2(H×2)

1.5

Lipp 2014

Intravaginal mechanical device vs no treatment

UC

Cornu 20123

(n = 55)

MD

−2.2 (−3.5 to −0.9)

> 0 favours no treatment

YES favours device

1(C)

1.6

Lipp 2014

Intravaginal mechanical device vs no treatment

UC

Cornu 20123

(n = 55)

MD

−1.67 (−2.8 to −0.5)

> 0 favours no treatment

YES favours device

1(C)

1.7

Lipp 2014

Intravaginal mechanical device vs no treatment

UC

Cornu 20123

(n = 55)

MD

−0.5 (−0.9 to −0.1)

> 0 favours no treatment

YES favours device

1(C)

1.3

Stewart 2017

ES vs no active treatment

IM

Correia 2013; Correia 2014; Lopès 2014; Pereira 2012 (n = 250)

SMD

−0.72 (−0.99 to −0.45)

< 0 favours ES

YES for ES

GRADE as assessed by review authors

Low‐certainty evidence

1.3.1

Dumoulin 2018

PFMT vs no treatment, placebo or control

IM

Beuttenmuller 2010; Carneiro 2010; Nascimento‐Correia 20122

(n = 145)

MD

−13.1 (−21.1 to −5.2)

> 0 favours control

YES favours PFMT

4(F×2, H×2)

1.4.1

Dumoulin 2018

PFMT vs no treatment, placebo or control

IM

Beuttenmuller 2010; Carneiro 2010; Nascimento‐Correia 20122

(n = 145)

MD

−11.8 (−20.8 to −2.7)

> 0 favours control

YES favours PFMT

4(F×2, H×2)

1.8.1

Dumoulin 2018

PFMT vs no treatment, placebo or control

IM

Beuttenmuller 2010; Carneiro 2010; Nascimento‐Correia 20122

(n = 145)

MD

1.8 (−3.4 to 7.0)

> 0 favours control

NO

4(E×2, H×2)

1.6.1

Dumoulin 2018

PFMT vs no treatment, placebo or control

IM

Bø 19991

(n = 55)

RR

0.8 (0.6 to 1.1)

> 1 favours PFMT

NO

3(C, E×2)

1.10.1

Dumoulin 2018

PFMT vs no treatment, placebo or control

IM

McLean 2013 (n = 35)

MD

−19.70 (−30.63 to −8.77)

> 0 favours control

YES favours PFMT

3 (E×2, H)

1.8

Lipp 2014

Intravaginal mechanical device vs no treatment

UC

Cornu 20123

(n = 46)

MD

−10.3 (−20.77 to 0.17)

> 0 favours no treatment

NO

3(C, E×2)

1Bø 1999 used the Bristol Female Lower Urinary Tract Symptoms questionnaire and this was reported as event data within the review (Dumoulin 2018).
2Beuttenmuller 2010; Carneiro 2010; Nascimento‐Correia 2012 – these three trials each present individual subscales for Kings Health Questionnaire (subscales of severity, impact, physical limitation and general health) and these pooled totals are presented in the tables.
3Cornu 2012 – the time point at which outcomes were measured was unclear.
Abbreviations: CI: confidence interval; ES: electrical stimulation; I‐QoL: Incontinence Quality of Life; IM: immediate; MD: mean difference; n: number of participants; n: number of participants; PFMT: pelvic floor muscle training; RR: risk ratio; SMD: standardised mean difference; UC: unclear.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Twelve analyses from three reviews assessed quality of life in analyses comparing a conservative intervention with a control intervention (Dumoulin 2018; Lipp 2014; Stewart 2017). Eight analyses measured the outcome immediately after the intervention. For the remaining four, the timing of outcome assessment was unclear. Across the three reviews, there were 10 unique trials assessing quality of life that compare a conservative intervention with a control intervention.

High‐ or moderate‐certainty evidence

One analysis provided high‐certainty evidence and five analyses provided moderate‐certainty evidence of a beneficial effect.

PFMT was more beneficial than control using the I‐QoL scale (high‐certainty) and the Kings Health Questionnaire (Physical Limitation subscale) (moderate‐certainty).

Intravaginal mechanical device was more beneficial than control using the Urinary Symptoms Profile – SUI subscore, Urinary Symptom Profile Questionnaire OAB subscore and the Urinary Symptom Profile Questionnaire Dysuria subscore (all moderate‐certainty).

Electrical stimulation using non‐implanted electrodes was more beneficial than no active treatment for a pooled measure of incontinence‐specific quality of life (moderate‐certainty).

Low‐ or very low‐certainty evidence

Six analyses provided low‐certainty evidence.

One conservative intervention versus another conservative intervention
Symptomatic cure or improvement

For more information, see Table 10.

Open in table viewer
Table 10. Stress urinary incontinence: cure or improved, conservative versus conservative

Analysis code (in review)

Study ID

Comparisons included

Timing of measure

Included trials

(n)

Effect

measure

Effect estimate

(CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades

(reason for downgrade)

High‐certainty evidence

2.1.1

Herbison 2013

Cones vs PFMT

IM

Bø 19991

(n = 52)

RR

4.6 (1.1 to 19.1)

> 1 favours PFMT

YES favours PFMT

0

8.1.1

Lipp 2014

Intravaginal pessary alone vs pessary + PFMT

IM

Richter 20071

(n = 299)

RR

0.7

(0.6 to 0.9)

> 1 favours pessary alone

YES favours PFMT + pessary

0

7.1.2

Lipp 2014

Intravaginal pessary vs behavioural therapy (PFMT) alone

IM

Richter 20071

(n = 295)

RR

0.7

(0.5 to 0.9)

> 1 favours device

YES favours PFMT

0

Moderate‐certainty evidence

6.1.2

Ayeleke 2015

PFMT + continence pessary vs continence pessary alone

FU1

Richter 2007

(n = 207)

RR

0.9 (0.7 to 1.1)

> 1 favours PFMT

NO

2(E×2)

6.3.3

Ayeleke 2015

PFMT + continence pessary vs continence pessary alone

FU2

Richter 2007

RR

0.9 (0.7 to 1.2)

> 1 favours PFMT

NO

2(E×2)

11.1.1

Hay‐Smith 2011

PFMT and adherence strategy vs PFMT alone

IM

Sugaya 20032

(n = 41)

RR

0.6 (0.38 to 0.9)

> 1 favours PFMT + adherence

YES favours PFMT + adherence

1(B)

12.2.2

Hay‐Smith 2011

More‐intensive vs less‐intensive PFMT programmes

IM

Diniz Zanetti 20072

(n = 44)

RR

0.3 (0.2 to 0.7)

> 1 favours more‐intense PFMT

YES favours more‐intense PFMT

1(H)

8.2.2

Hay‐Smith 2011

Strength and motor learning vs motor learning PFMT alone

IM

Hay‐Smith 20022

(n = 123)

RR

0.6 (0.3 to 1.4)

> 1 favours less intensive

NO

2(E×2)

2.1.1

Herbison 2013

Cones vs PFMT

IM

Cammu 19983

(n = 60)

RR

0.9 (0.5 to 1.6)

> 1 favours PFMT

NO

2(E×2)

7.1.1

Lipp 2014

Intravaginal pessary vs behavioural therapy (PFMT

IM

Richter 20071

(n = 295)

RR

0.8 (0.6 to 1.0)

> 1 favours device

NO

1(E)

7.1.3

Lipp 2014

Intravaginal pessary vs behavioural therapy (PFMT

FU1

Richter 2007

(n = 295)

RR

1.0 (0.8 to 1.3)

> 1 favours device

NO

1(E)

7.1.4

Lipp 2014

Intravaginal pessary vs behavioural therapy (PFMT

FU2

Richter 2007

(n = 295)

RR

0.9 (0.7 to 1.3)

> 1 favours device

NO

1(E)

7.1.6

Lipp 2014

Intravaginal pessary vs behavioural therapy (PFMT

FU2

Richter 2007

(n = 295)

RR

0.9 (0.7 to 1.2)

> 1 favours device

NO

1(E)

8.1.3

Lipp 2014

Intravaginal pessary alone vs pessary + PFMT

FU1

Richter 2007

(n = 299)

RR

0.9 (0.7 to 1.1)

> 1 favours pessary alone

NO

1(E)

8.1.4

Lipp 2014

Intravaginal pessary alone vs pessary + PFMT,

FU2

Richter 2007

(n = 299)

RR

0.9 (0.7 to 1.3)

> 1 favours pessary alone

NO

1(E)

8.1.6

Lipp 2014

Intravaginal pessary alone vs pessary + PFMT,

FU2

Richter 2007

(n = 299)

RR

1.0 (0.7 to 1.3)

> 1 favours pessary alone

NO

1(E)

9.1.1

Lipp 2014

Pessary + PFMT vs behavioural therapy (PFMT) alone

IM

Richter 20071

(n = 296)

RR

1.0 (0.5 to 1.3)

> 1 favours pessary + PFMT

NO

1(E)

9.1.3

Lipp 2014

Pessary + PFMT vs behavioural therapy (PFMT) alone

FU1

Richter 2007

(n = 296)

RR

1.1 (0.91 to 1.4)

> 1 favours pessary + PFMT

NO

1(E)

9.1.4

Lipp 2014

Pessary + PFMT vs behavioural therapy (PFMT) alone

FU2

Richter 2007

(n = 296)

RR

1.0 (0.7 to 1.4)

> 1 favours pessary + PFMT

NO

1(E)

9.1.6

Lipp 2014

Pessary + PFMT vs behavioural therapy (PFMT) alone

FU2

Richter 2007

(n = 296)

RR

0.9 (0.7 to 1.3)

> 1 favours pessary + PFMT

NO

1(E)

9.1.2

Lipp 2014

Pessary + PFMT vs behavioural therapy (PFMT) alone

IM

Richter 20071

(n = 296)

RR

0.9 (0.7 to 1.1)

> 1 favours device

NO

1(E)

7.1.5

Lipp 2014

Intravaginal pessary vs behavioural therapy (PFMT) alone

FU1

Richter 2007

(n = 295)

RR

0.9 (0.6 to 1.1)

> 1 favours pessary alone

NO

1(E)

8.1.2

Lipp 2014

Intravaginal pessary alone vs pessary + PFMT,

IM

Richter 20071

(n = 299)

RR

0.7 (0.5 to 1.0)

> 1 favours pessary alone

NO

1(E)

8.1.5

Lipp 2014

Intravaginal pessary alone vs pessary + PFMT

FU1

Richter 2007

(n = 299)

RR

1.1 (0.8 to 1.5)

> 1 favours pessary + PFMT

NO

1(E)

9.1.5

Lipp 2014

Pessary + PFMT vs behavioural therapy (PFMT) alone

FU1

Richter 2007

(n = 296)

RR

0.8 (0.6 to 1.1)

> 1 favours pessary + PFMT

NO

1(E)

Low‐certainty evidence

4.2

Ayeleke 2015

PFMT + ES vs ES alone (excluding implanted electrodes)

IM

Bezerra 20093; Hofbauer 19903

(n = 56)

RR

2.1 (0.8 to 5.4)

> 1 favours PFMT + pessary

NO

4(E×2, H×2)

6.1.1

Ayeleke 2015

PFMT + continence pessary vs continence pessary alone

IM

Richter 20071

(n = 242)

RR

1.0 (0.8 to 1.2)

> 1 favours PFMT + pessary

NO

3(C, E×2)

6.3.1

Ayeleke 2015

PFMT + continence pessary vs continence pessary alone

IM

Richter 20071

(n = 242)

RR

1.1 (0.9 to 1.4)

> 1 favours PFMT + pessary

NO

3(C, E×2)

6.3.2

Ayeleke 2015

PFMT + continence pessary vs continence pessary alone

FU1

Richter 2007

(n = 225)

RR

1.0 (0.8 to 1.3)

> 1 favours PFMT + pessary

NO

3(C, E×2)

1.2.1

Hay‐Smith 2011

More vs less contact with health professionals

IM

Bø 19902; Diniz Zanetti 20072; Felicíssimo 20102; Konstantinidou 20072

(n = 177)

RR

0.3 (0.14 to 0.5)

> 1 favours more contact

YES favours more contact

3(F×2, H)

2.2.1

Hay‐Smith 2011

Group vs individual supervision of PFMT

IM

Bø 19902; Felicíssimo 20102; Konstantinidou 20072

(n = 133)

RR

0.1 (0.0 to 0.4)

> 1 favours group

YES, favours group

3(F×2, H)

2.2.2

Hay‐Smith 2011

Group vs individual supervision of PFMT

IM

de Oliveira 20092

(n = 60)

RR

1.2 (0.6 to 2.3)

> 1 favours individual

NO

3(E×2, H)

4.1.1

Hay‐Smith 2011

Individualised vs generic PFMT

IM

de Oliveira 20092

(n = 60)

RR

0.8 (0.4 to 1.6)

> 1 favours generic

NO

3(E×2, H)

2.1.2

Herbison 2013

Cones vs PFMT

IM

Arvonen 20013; Arvonen 20023

(n = 54)

RR

0.8 (0.5 to 1.3)

> 1 favours PFMT

NO

3(E×2, H)

4.1.1

Herbison 2013

Cones + PFMT vs PFMT

IM

Pieber 19953

(n = 46)

RR

1.4 (0.8 to 2.4)

> 1 favours PFMT

NO

3(E×2, H)

4.2.1

Herbison 2013

Cones + PFMT vs PFMT

O

Pieber 1995

(n = 46)

RR

0.9 (0.5 to 1.6)

> 1 favours PFMT

NO

3(E×2, H)

5.1

Herbison 2013

Cones + PFMT vs ES

IM

Berglund 19933; Seo 20043 (n = 160)

RR

1.5 (0.8 to 2.6)

> 1 favours ES

NO

3(E×2, H)

3.2

Stewart 2017

ES vs PFMT

IM

Bø 1999; Castro 2008; Hahn 1991; Henalla 1989; Hofbauer 1990; Laycock 1988; Aaronson 1995 (n = 244)

RR

0.85 (0.70 to 1.03)

> 1 favours ES

NO

GRADE as assessed by review authors

4.2

Stewart 2017

ES vs vaginal cones

IM

Bridges 1988; Bø 1999; Castro 2008; Oláh 1990; Seo 2004 (n = 331)

RR

1.09 (0.97 to 1.21)

> 1 favours ES

NO

GRADE as assessed by review authors

6.2

Stewart 2017

ES + PFMT vs PFMT

IM

Eyjólfsdóttir 2009; Goode 2003; Hofbauer 1990; Knight 1998; Knight 1998; Tapp 1989; Wilson 1987 (n = 308)

RR

1.10 (0.95 to 1.28)

> 1 favours ES + PFMT

NO

GRADE as assessed by review authors

Very low‐certainty evidence

5.2

Stewart 2017

ES vs PFMT + vaginal cones

IM

Bourcier 1994; Laycock 1993 (n = 68)

RR

1.53 (1.08 to 2.18)

> 1 favours ES

YES for ES

GRADE as assessed by review authors

Other

6.2.1

Hay‐Smith 2011

Daily vs 3 times per week PFMT

IM

Sriboonreung 20112 (n = 40)

RR

NE

NE

9.2

Hay‐Smith 2011

PFMT and abdominal muscle exercise vs PFMT alone

IM

Sriboonreung 20112 (n = 40)

RR

NE

NE

1Trials comparing the effect of conservative interventions from different intervention categories, with an immediate outcome measure.
2Bø 1999, de Oliveira 2009, Diniz Zanetti 2007, Hay‐Smith 2011, Sriboonreung 2011, Sugaya 2003, Felicíssimo 2010, Konstantinidou 2007 compared different ways of delivering the same intervention.
3Arvonen 2001Arvonen 2002, Bezerra 2009, Cammu 1998, Hofbauer 1990, Pieber 1995, Laycock 1993, and Seo 2004 compared the effect of conservative interventions from within the same intervention categories (physical therapy interventions in all cases).
Abbreviations: CI: confidence interval; ES: electrical stimulation; FU1: follow‐up one; IM: immediate; PFMT: pelvic floor muscle training; RR: risk ratio; n: number of participants assessed NOT to have been cured or improved using a variety of different assessment methods; NE: not estimable; O: other.

Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

A total of 43 analyses from five reviews assessed symptomatic cure or improvement by comparing one conservative intervention with another conservative intervention (Ayeleke 2015; Hay‐Smith 2011; Herbison 2013; Lipp 2014; Stewart 2017). Of these, 27 related to immediate outcome assessment. Thirty‐two analyses compared two different conservative interventions and nine compared different ways of delivering the same conservative interventions. A total of 31 trials compared different conservative interventions with a measure of cure or improvement.

High‐ or moderate‐certainty evidence

Three analyses provided high‐certainty evidence and 22 analyses provided moderate‐certainty evidence. Five analyses provided moderate‐ or high‐certainty evidence of a difference between treatment groups, at an immediate time point.

  • Continence pessary plus PFMT was more beneficial compared to continence pessary alone using the PGI‐I, and PFMT was more beneficial than continence pessary alone, as measured by the Urinary Distress Inventory (UDI) subscale of the Pelvic Floor Distress Inventory (PFDI). Two analyses, both high‐certainty (note: analyses used data from same single trial that did not find a difference between other treatment groups, or using other measures or at follow‐up outcomes – see below) (Richter 2007).

  • PFMT plus educational/behavioural intervention was more beneficial than cones (high‐certainty).

  • More‐intensive PFMT was beneficial compared to less‐intensive PFMT (moderate‐certainty).

  • PFMT plus an adherence strategy was beneficial compared to PFMT alone (moderate‐certainty).

Twenty analyses, providing moderate‐certainty evidence, showed no evidence of a difference between groups.

  • Strength plus motor learning intervention versus motor learning intervention (moderate‐certainty).

  • PFMT plus continence pessary versus continence pessary alone, and continence pessary versus PFMT for outcomes assessed using a bladder diary, the PGI‐I and UDIS subscale of the PFDI using both immediate and follow‐up data (18 analyses, all with data from the same single trial (Richter 2007)) (moderate‐certainty).

  • PFMT versus cones (moderate‐certainty).

Low‐ or very low‐certainty evidence

Eighteen analyses provided low‐ or very low‐certainty evidence, or were not estimable.

Condition‐specific quality of life

For more information, see Table 11.

Open in table viewer
Table 11. Stress urinary incontinence: quality of life, conservative versus conservative

Analysis code (in review)

Study ID

Comparisons included

Timing of measure

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

Moderate‐certainty evidence

6.2.2

Ayeleke 2015

PFMT added to continence pessary vs continence pessary alone,

FU1

Richter 2007

(n = 207)

RR

0.8 (0.6 to 1.1)

> 1 favours PFMT + pessary

NO

2(E×2)

7.4

Lipp 2014

Intravaginal pessary vs behavioural therapy (PFMT)

IM

Richter 2007

(n = 295)

MD

0.7 (−9.5 to 10.9)

> 0 favours PFMT

NO

1(E)

7.5

Lipp 2014

Intravaginal pessary vs behavioural therapy (PFMT)

FU1

Richter 2007

(n = 295)

MD

−3.2 (−11.4 to 5.0)

> 0 favours PFMT

NO

1(E)

7.6

Lipp 2014

Intravaginal pessary vs behavioural therapy (PFMT)

FU1

Richter 2007

(n = 295)

MD

−0.2 (−1.3 to 0.9)

> 0 favours PFMT

NO

1(E)

7.7

Lipp 2014

Intravaginal pessary vs behavioural therapy (PFMT)

FU1

Richter 2007

(n = 295)

MD

0.3 (−0.7 to 1.3)

> 0 favours PFMT

NO

1(E)

2.3

Mariappan 2005

SNRI + conservative non‐pharmacological therapy vs conservative non‐pharmacological therapy alone

IM

Schagen van Leeuwen 20041

(n = 100)

OR

1.3 (0.5 to 2.9)

> 1 favours combination

NO

2(E×2)

Low‐certainty evidence

6.2.1

Ayeleke 2015

PFMT + continence pessary vs continence pessary alone

IM

Richter 20071

(n = 242)

RR

1.1 (0.8 to 1.5)

> 1 favours PFMT +

NO

3(C, E×2)

4.3

Stewart 2017

ES vs vaginal cones

IM

Castro 2008; Santos 2009 (n = 96)

MD

1.59 (−3.72 to 6.90)

> 1 favours ES

NO

GRADE as assessed by review authors

10.12.3

Wallace 2004

Bladder training + non‐pharmacological treatment vs non‐pharmacological treatment alone

IM

Wyman 19983

(n = 89)

MD

−18 (−36.6 to 0.6)

> 0 favours control

NO

4(E×2, G, H)

5.14.3

Wallace 2004

Bladder training vs other behavioural/physical/psychological treatments

IM

Wyman 19983

(n = 92)

MD

18 (−1.4 to 37.3)

> 0 favours control

NO

3(E×2, G)

Other

9.3

Mariappan 2005

SNRI + conservative non‐pharmacological therapy vs conservative non‐pharmacological

IM

Schagen van Leeuwen 2004 (n = 100)

MD

NE

NO

NE

1Data presented as number of events.
2Data for I‐QOL, effect size was not estimable as the standard deviations are provided as 0.
3These Wyman 1998 data comprised a subgroup of the participants from Wyman 1998: Wyman 1998 included participants with combined SUI, UUI and MUI and all data is presented within reviews for the combined group of participants, except for the two subgroups of data here, which related to participants with SUI only. These Wyman 1998 data related to the Urinary Distress Inventory measure of quality of life, for two pairs from this three‐arm trial (the physical therapy treatment (control) group data are the same for both pairs).
Abbreviations: CI: confidence interval; FU1: follow‐up one; IM: immediate; I‐QOL: Incontinence Quality of Life; MD: mean difference; MUI: mixed urinary incontinence; n: number of participants; NE: not estimable; O: other; OR: odds ratio; RR: risk ratio; SNRI: serotonin–noradrenaline reuptake inhibitors; SUI: stress urinary incontinence; UUI: urgency urinary incontinence.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Eleven analyses from five reviews assessed quality of life comparing one conservative intervention with another conservative intervention (Ayeleke 2015; Lipp 2014; Mariappan 2005; Stewart 2017; Wallace 2004). Seven of these analyses related to an immediate assessment and four to a follow‐up outcome. One of these analyses contained data that meant it was not possible to calculate an effect size. Five unique trials contributed data to these analyses.

High‐ or moderate‐certainty evidence

Six analyses provided moderate‐certainty evidence of no evidence of a difference between treatment groups.

  • PFMT plus serotonin–noradrenaline reuptake inhibitors (SNRI) was no different from PFMT alone (moderate‐certainty).

  • PFMT plus continence pessary was no different from continence pessary alone at follow‐up assessment (moderate‐certainty).

  • Continence pessary was no different from PFMT at immediate or follow‐up assessment (four analyses, moderate‐certainty).

Low‐ or very low‐certainty evidence

Four analyses provided low‐certainty evidence and one was not estimable.

Any conservative intervention versus another active intervention (non‐conservative intervention)
Symptomatic cure or improvement

For more information, see Table 12.

Open in table viewer
Table 12. Stress urinary incontinence: primary outcome participant‐reported cure or improvement; conservative versus non‐conservative

Analysis code (in review)

Study ID

Comparisons included

Timing of measure

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

Moderate‐certainty evidence

7.2.1

Ayeleke 2015

PFMT added to drug therapy vs drug therapy alone

IM

Schagen van Leeuwen 20041,2

(n = 90)

RR

1.1 (0.8 to 1.5)

> 1 favours PFMT + drug

NO

2(E×2)

7.4.1

Ayeleke 2015

PFMT added to drug therapy vs drug therapy alone

IM

Schagen van Leeuwen 20041,3

(n = 101)

RR

1.3 (0.9 to 1.8)

> 1 favours acupuncture

NO

2(E×2)

2.1

Alhasso 2005

Adrenergic agonist vs conservative therapy

UC

Ishiko 2000

(n = 34)

RR

1.3 (0.7 to 2.2)

> 1 favours drug

NO

2(E×2)

2.1

Alhasso 2005

Adrenergic agonist vs conservative therapy

UC

Ishiko 2000

(n = 37)

RR

0.9 (0.6 to 1.3)

> 1 favours drug

NO

2(E×2)

1.1

Lapitan 2017

Open retropubic colposuspension vs conservative interventions

FU1

Klarskov 19864 (n = 29)

RR

0.2 (0.1 to 0.7)

< 1 favours surgery

YES favours surgery

1(H)

2.1

Wang 2013

Acupuncture vs any other treatment

IM

Bi 20071

(n = 60)

RR

2.2 (1.3 to 3.8)

> 1 favours acupuncture

YES favours acupuncture

2(A, B)

Low‐certainty evidence

2.3

Glazener 2017a

Anterior repair vs PFMT

FU1

Klarskov 19864

(n = 16)

RR

0.3 (0.0 to 2.3)

> 1 favours PFMT

NO

2(E×2)

2.2

Glazener 2017a

Anterior repair vs PFMT

FU2

Klarskov 19864

(n = 16)

RR

2.6 (0.6 to 10.2)

> 1 favours PFMT

NO

2(E×2)

2.2

Kirchin 2017

Urethral injection therapy vs conservative management

IM

ter Meulen 20091

(n = 45)

RR

0.2 (0.0 to 1.8)

> 1 favours conservative

NO

3(E×2, H)

1Trials with an immediate outcome measure.
2Schagen van Leeuwen 2004 determined cure or improvement using two methods. These data related to data collected via a paper diary.
3Schagen van Leeuwen 2004 determined cure or improvement using two methods. These data are based on the Patient Global Impression scale – Improvement.
4Klarskov 1986 included participants who received different surgical interventions, Lapitan 2017 included all surgical groups, while Glazener 2017a included the subgroup of participants receiving anterior repair – therefore, 16 participants were presented twice.
Abbreviations: CI: confidence interval; FU1: follow‐up one; FU2: follow‐up two; IM: immediate; n: number of participants assessed NOT to have been cured or improved using a variety of different assessment methods; PFMT: pelvic floor muscle training; RR: risk ratio; UC: unclear.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Nine analyses from six reviews assessed symptomatic cure or improvement comparing one conservative intervention with a non‐conservative intervention (Alhasso 2005; Ayeleke 2015; Glazener 2017a; Kirchin 2017; Lapitan 2017; Wang 2013). Of these, four analyses related to immediate outcome assessment. Five trials compared conservative and non‐conservative interventions (Glazener 2017a and Lapitan 2017 both reported data from the same trial).

High‐ or moderate‐certainty evidence

Six analyses provided moderate‐certainty evidence. Two provided moderate‐certainty evidence of a beneficial effect of treatment.

  • Acupuncture led to high symptomatic cure or improvement compared to 'any other treatment', which comprised a pharmacological intervention.

  • Open retropubic colposuspension surgery led to higher symptomatic cure or improvement than a conservative intervention within the first year.

Four analyses provided moderate‐certainty evidence of no evidence of a difference between groups.

  • PFMT plus SNRI versus SNRI alone: two comparisons with different methods of assessing cure or improvement.

  • Adrenergic agonist versus conservative therapy.

  • Adrenergic agonist versus conservative therapy plus adrenergic agonist.

Low‐ or very low‐certainty evidence

Three analyses provided low‐certainty evidence.

Condition‐specific quality of life

For more information, see Table 13.

Open in table viewer
Table 13. Stress urinary incontinence: quality of life, conservative versus non‐conservative

Analysis code (in review)

Study ID

Comparisons included

Timing of measure

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

High‐certainty evidence

2.2

Mariappan 2005

SNRI vs conservative non‐ pharmacological therapies

IM

Schagen van Leeuwen 20042

(n = 92)

OR

2.4 (1.0 to 5.6)

> 1 favours duloxetine

YES favours

medication

1(H)

Moderate‐certainty evidence

7.3.1

Ayeleke 2015

PFMT + drug therapy vs drug therapy alone

IM

Schagen van Leeuwen 20041

(n = 101)

MD

5.8 (−2.1 to 13.7)

> 0 favours PFMT + drug

NO

2(E×2)

2.3

Kirchin 2017

Urethral injection therapy vs conservative management (PFMT)

IM

ter Meulen 20091

(n = 45)

MD

0.5 (0.1 to 0.9)

> 0 Favours PFMT

YES favours PFMT

1(H)

Not estimable

2.3

Mariappan 2005

SNRI vs conservative non‐ pharmacological therapies

IM

Schagen van Leeuwen 20042

(n = 99)

MD

NE

> 0 favours PFMT

NO

0

1Data from the I‐QOL measure of quality of life.
2The review Mariappan 2005 also reported data from Schagen van Leeuwen 2004, reporting the Physical Quality of Life Index (as event data,), and reporting the Incontinence Quality of Life questionnaire (but reporting the standard deviations as 0, and therefore not enabling an effect size estimate).
Abbreviations: CI: confidence interval; IM: immediate; MD: mean difference; n: number of participants; NE: not estimable; OR: odds ratio; PFMT: pelvic floor muscle training; RR: risk ratio; SNRI: serotonin–noradrenaline reuptake inhibitor.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Four analyses from three reviews assessed quality of life comparing one conservative intervention with a non‐conservative intervention (Ayeleke 2015; Kirchin 2017; Mariappan 2005). All four analyses relate to an immediate assessment. One analysis contained data that meant it was not possible to calculate an effect size. Two trials compared a conservative intervention with a non‐conservative intervention at an immediate time point (Schagen van Leeuwen 2004; ter Meulen 2009).

High‐ or moderate‐certainty evidence

One analysis provided high‐certainty evidence and two moderate‐certainty evidence. One analysis provided high‐certainty and one moderate‐certainty evidence of improved quality of life with non‐conservative interventions.

  • SNRI was more beneficial than conservative intervention (PFMT) (high‐certainty).

  • Urethral injection was more beneficial than conservative intervention (PFMT) (moderate‐certainty).

One analysis provided moderate‐certainty evidence of no difference between PFMT plus SNRI and SNRI alone.

Urgency urinary incontinence

Any conservative intervention versus control, placebo or standard care
Symptomatic cure or improvement

For more information, see Table 14.

Open in table viewer
Table 14. Urgency urinary incontinence: primary outcome participant‐reported cure or improvement; conservative versus control

Analysis code (in review)

Study ID

Comparisons included

Timing of measure

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

High‐certainty evidence

2.2.2

Herderschee 2011

PFMT + feedback vs PFMT alone (control)

IM

Burgio 2002b1

(n = 122)

RR

0.5 (0.4 to 0.8)

> 1 favours PFMT

YES favours PFMT + feedback

0

1.9

Stewart 2016

ES vs no active treatment, placebo or sham treatment

IM

Finazzi‐Agrò 2010²

(n = 35)

RR

0.3 (0.1 to 0.6)

< 1 favours ES

YES favours ES

0

Moderate‐certainty evidence

1.2.2

Herderschee 2011

PFMT + biofeedback vs PFMT alone

IM

Burgio 20021

(n = 118)

RR

0.5 (0.4 to 0.8)

> 1 favours PFMT alone

YES favours PFMT + BF

1(C)

1.19.2

Stewart 2016

ES vs no active treatment, placebo or sham treatment

IM

Kosilov 20132

(n = 107)

MD

−1.1 (−1.8 to −0.4)

< 0 favours

ES

YES favours ES

1(H)

1.8

Stewart 2016

ES vs no active treatment, placebo or sham treatment

IM

Amaro 2006

(n = 40)

RR

0.5 (0.1 to 1.7)

< 1 favours

ES

NO

2(E×2)

1.4.1

Wallace 2004

Bladder training vs no treatment

FU1

Lagro‐Janssen 1991

(n = 18)

RR

17.0 (1.1 to 256.5)

> 1 favours BT

YES favours BT

1(H)

Low‐certainty evidence

5.2

Stewart 2016

ES + PFMT vs no active treatment

IM

Firra 20132

(n = 12)

MD

−1.6 (−3.6 to 0.4)

< 0 favours ES + PFMT

NO

3(A, E×2)

1Burgio 2002 was a three‐arm trial, in which the arms were PFMT (n = 75), PFMT plus feedback (n = 74) and PFMT plus biofeedback (n = 73). Overview authors judged that the PFMT only group met the definition to be classed as a control intervention. Therefore, there were two comparisons categorised within the overview as 'conservative versus control', one that was PFMT plus biofeedback versus control (Burgio 2002), and one that was PFMT plus feedback versus control (Burgio 2002b).
2Data from these trials are presented as 'number of incontinence episodes in 24 hours'.
Abbreviations: CI: confidence interval; ES: electrical stimulation; FU1: follow‐up one; IM: immediate; MD: mean difference; n: number of participants assessed NOT to have been cured or improved using a variety of different assessment methods; PFMT: pelvic floor muscle training; RR: risk ratio.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Seven analyses from three reviews assessed symptomatic cure or improvement and compared a conservative intervention with a control intervention (Herderschee 2011; Stewart 2016; Wallace 2004). Six outcomes were at an immediate time point and one at a follow‐up time point. Six unique trials compared a conservative intervention with a control intervention, with five reporting data on immediate cure or improvement.

High‐ or moderate‐certainty evidence

Two analyses provided high‐certainty and two moderate‐certainty evidence of a beneficial effect demonstrating the following.

  • PFMT plus feedback had better cure or improvement compared to control (high‐certainty) and PFMT plus biofeedback had a beneficial effect on outcome compared to control (moderate‐certainty). In both these comparisons, the overview authors categorised the 'PFMT' intervention as a no treatment control intervention. This in effect adds a bias against the treatment interventions that make it less likely that any evidence of a difference favouring treatment would be found.

  • Electrical stimulation had better cure or improvement compared to control (two comparisons, one high‐certainty, one moderate‐certainty).

  • Bladder training had a beneficial effect compared to no treatment at a follow‐up time point (moderate‐certainty).

One analysis provided moderate‐certainty evidence of no difference when electrical stimulation was compared to control. This was in contrast to the findings of other comparisons, which demonstrated a beneficial effect (see above).

Low‐ or very low‐certainty evidence

One analysis provided low‐certainty evidence.

Condition‐specific quality of life

For more information, see Table 15.

Open in table viewer
Table 15. Urgency urinary incontinence: quality of life, conservative versus control

Analysis code (in review)

Study ID

Comparisons included

Timing of measure

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

Moderate‐certainty evidence

5.1.1

Stewart 2016

ES + PFMT vs no active treatment

IM

Firra 2013

(n = 12)

MD

−12.4 (−20.9 to −3.9)

< 0 favours ES + other

YES favours ES + PFMT

1(A)

Abbreviations: CI: confidence interval; ES: electrical stimulation; IM: immediate; MD: mean difference; n: number of participants; PFMT: pelvic floor muscle training.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

One analysis from one review assessed quality of life and compared a conservative intervention with control (Stewart 2016).

High‐ or moderate‐certainty evidence

One analysis provided moderate‐certainty evidence of a beneficial effect. Combined electrical stimulation plus PFMT was more beneficial than control for a measure of quality of life using the York Incontinence Perception Scale score. There was one unique trial.

One conservative intervention versus another conservative intervention
Symptomatic cure or improvement

For more information, see Table 16.

Open in table viewer
Table 16. Urgency urinary incontinence: cure or improvement, conservative versus conservative

Analysis code (in review)

Study ID

Comparisons included

Timing of measure

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

Moderate‐certainty evidence

4.2.1

Herderschee 2011

PFMT + BF vs PFMT + F

IM

Burgio 2002c1

(n = 110)

RR

1.0 (0.6 to 1.7)

> 1 favours PFMT + feedback

NO

2(E×2)

2.8.2

Stewart 2016

ES vs laseropuncture

IM

Kosilov 20132

(n = 114)

MD

−1.8 (−2.3 to −1.3)

< 0 favours ES

YES favours ES

1(H)

4.2.1

Stewart 2016

ES + PFMT vs PFMT alone

IM

Schreiner 20102

(n = 51)

RR

0.56 (0.3 to 0.9)

< 1 favours ES + other

YES favours ES + PFMT

1(H)

4.3.1

Stewart 2016

ES + PFMT vs PFMT alone

IM

Schreiner 20102

(n = 51)

RR

0.3 (0.1 to 0.7)

< 1 favours ES + other

YES favours ES + PFMT

1(H)

2.3.1

Stewart 2016

ES vs PFMT

IM

Wang 20042

(n = 69)

RR

0.8 (0.5 to 1.2)

< 1 favours ES

NO

2(E×2)

2.3.2

Stewart 2016

ES vs PFMT +

biofeedback

IM

Wang 20042

(n = 69)

RR

0.9 (0.6 to 1.6)

< 1 favours ES

NO

2(E×2)

2.8.1

Stewart 2016

ES vs PFMT

IM

Arruda 20082

(n = 42)

MD

0.1 (−8.7 to 8.9)

< 0 favours ES

NO

2(E×2)

Low‐certainty evidence

5.2

Rai 2012

Anticholinergic drugs + non‐drug therapies vs non‐drug therapies alone

IM

Park 2002

(n = 50)

RR

0.6 (0.3 to 1.3)

< 1 favours anticholinergics + non‐drug therapies

NO

4(E×2, H×2)

5.2

Rai 2012

Anticholinergic drugs + non‐drug therapies vs non‐drug therapies alone

IM

Song 2006

(n = 57)

RR

0.6 (0.3 to 1.2)

< 1 favours anticholinergics + non‐drug therapies

NO

3(E×2, H)

4.9.2

Stewart 2016

ES + PFMT vs PFMT

IM

Firra 20132

(n = 12)

MD

0.2 (−1.2 to 1.6)

< 0 favours ES + other

NO

3(A, E×2)

1Burgio 2002 was a three‐arm trial in which two of the groups received PFMT, combined with either feedback or biofeedback. Both groups received active treatment, each comprising a different way of delivering the same intervention.
2These trials compared ES with another conservative intervention. However, in all cases both conservative interventions were categorised as 'physical therapy' interventions.
Abbreviation: CI: confidence interval; ES: electrical stimulation; IM: immediate; MD: mean difference; n: number of participants assessed NOT to have been cured or improved using a variety of different assessment methods; RR: risk ratio.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Ten analyses from three reviews compared one conservative intervention with another conservative intervention for a cure or improvement outcome (Herderschee 2011; Rai 2012; Stewart 2016). All analyses related to immediate outcome assessment. Two analyses compared the effect of anticholinergic drugs plus conservative intervention versus the same conservative intervention, one analysis compared different ways of delivering the same conservative intervention (PFMT) and seven analyses compared different physical therapy interventions. Eight unique trials compared one conservative intervention with another conservative intervention for a measure of cure or improvement with an immediate outcome.

High‐ or moderate‐certainty evidence

Seven analyses provided moderate‐certainty evidence. Three analyses provided evidence of a more beneficial effect.

  • Electrical stimulation was more effective than laseropuncture (moderate‐certainty).

  • Electrical stimulation plus PFMT was more effective than PFMT alone (two analyses, moderate‐certainty).

Four analyses provided moderate‐certainty evidence of no difference between groups.

  • PFMT plus biofeedback versus individualised PFMT (moderate‐certainty).

  • Electrical stimulation versus PFMT (two analyses, moderate‐certainty).

  • Electrical stimulation versus PFMT plus biofeedback (one analysis, moderate‐certainty).

Low‐ or very low‐certainty evidence

Three analyses provided low‐certainty evidence.

Condition‐specific quality of life

For more information, see Table 17.

Open in table viewer
Table 17. Urgency urinary incontinence: quality of life, conservative versus conservative

Analysis code (in review)

Study ID

Comparisons included

Timing of measure

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

High‐certainty evidence

2.5.1

Stewart 2016

ES vs PFMT

IM

Wang 20041,2

(n = 69)

MD

129.8 (47.8 to 211.8)

< 0 favours ES

YES favours PFMT

0

Moderate‐certainty evidence

4.5.1

Stewart 2016

ES + PFMT vs PFMT alone

IM

Schreiner 20101 (n = 51)

MD

−2.7 (−5.1 to −0.2)

< 0 favours ES + other

YES favours ES + PFMT

1(H)

2.5.2

Stewart 2016

ES vs PFMT +

biofeedback

IM

Wang 20041,2 (n = 69)

MD

−5.8 (−89.0 to 77.4)

< 0 favours ES

NO

2(E×2)

Low‐certainty evidence

4.8.1

Stewart 2016

ES + PFMT vs PFMT

IM

Firra 20131 (n = 12)

MD

5.8 (−3.5 to 15.1)

< 0 favours ES + other

NO

3(A, E×2)

1Studies compare two conservative interventions both categorised as 'physical therapy'.
2Wang 2004 was a three‐arm trial comparing PFMT, PFMT plus biofeedback and ES.
Abbreviations: CI: confidence interval; ES: electrical stimulation; IM: immediate; MD: mean difference; n: number of participants; PFMT: pelvic floor muscle training.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Four analyses from one review presented quality of life data from participants with UUI (Stewart 2016). All analyses related to immediate outcome assessment. There were three unique studies.

High‐ or moderate‐certainty evidence

One analysis provided high‐certainty evidence and two moderate‐certainty evidence. Two analyses provided evidence of a difference between groups.

  • PFMT was more effective than electrical stimulation (high‐certainty).

  • Electrical stimulation plus PFMT was more effective than PFMT alone (moderate‐certainty).

One analysis provided moderate‐certainty evidence of no difference; there was no evidence that electrical stimulation had a different quality of life to PFMT plus biofeedback.

Low‐ or very low‐certainty evidence

One analysis provided low‐certainty evidence.

Any conservative intervention versus another active intervention (non‐conservative intervention)
Symptomatic cure or improvement

For more information, see Table 18.

Open in table viewer
Table 18. Urgency urinary incontinence: primary outcome participant‐reported cure or improvement; conservative versus non‐conservative

Analysis code (in review)

Study ID

Comparisons included

Timing of measure

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

Moderate‐certainty evidence

6.2.3

Rai 2012

Anticholinergic drugs + behavioural interventions vs anticholinergic drugs alone

IM

Burgio 2008; Burgio 20103

(n = 364)

RR

0.6 (0.4 to 0.9)

> 1 favours drugs alone

YES favours drugs + behavioural

interventions

2(F×2)

6.2

Rai 2012

Anticholinergic drugs + non‐drug therapies vs anticholinergic drugs alone

IM

Millard 2004

(n = 475)

RR

1.3 (0.8 to 1.9)

> 1 favours drugs

NO

1(H)

3.3

Rai 2012

Anticholinergic drugs vs external electrostimulation

IM

Wang 2006

(n = 47)

RR

1.5 (0.8 to 2.6)

> 1 favours ES

NO

2(E×2)

3.10.3

Stewart 2016

ES vs trospium +

solifenacin

IM

Kosilov 20131

(n = 110)

MD

2.2 (1.8 to 2.6)

< 0 favours ES

YES favours drugs

1(H)

4.9.1

Stewart 2016

ES + drugs vs drugs

IM

Sancaktar 20101

(n = 38)

MD

−0.9 (−1.1 to −0.7)

< 0 favours ES + other

YES favours ES + drugs

1(H)

3.10.2

Stewart 2016

ES vs oxybutynin

IM

Arruda 20081

(n = 43)

MD

0.9 (−6.4 to 8.2)

< 0 favours ES

NO

2(E×2)

3.6.1

Stewart 2016

ES vs tolterodine

IM

Franzén 2010

(n = 64)

RR

0.9 (0.4 to 2.0)

< 1 favours ES

NO

2(E×2)

4.9.3

Stewart 2016

ES + PFMT + drugs vs drugs

IM

Kaya 20111

(n = 30)

MD

−1 (−2.1 to 0.1)

< 0 favours ES + other

NO

2(E×2)

3.14.1

Stewart 2016

ES vs solifenacin succinate

IM

Vecchioli‐Scaldazza 20131

(n = 30)

MD

−0.9 (−2.0 to 0.2)

< 0 favours ES

NO

2(E×2)

Low‐certainty evidence

7.4.2

Ayeleke 2015

PFMT + drug therapy vs drug therapy alone

IM

Burgio 20103

(n = 58)

RR

0.9 (0.7 to 1.1)

> 1 favours PFMT + other

NO

3(E×2, H)

7.2

Rai 2012

Anticholinergic drugs vs combination non‐drug therapies

IM

Goode 2002

(n = 128)

RR

2.4 (1.0 to 5.8)

< 0 favours anticholinergics

NO

4(C, E×2, H)

1.3

Rai 2012

Anticholinergic drugs vs bladder training

IM

Colombo 19954

(n = 27)

RR

0.5 (0.4 to 4.5)

> 1 favours bladder training

NO

3(E×2, H)

1.3

Rai 2012

Anticholinergic drugs vs bladder training

IM

Milani 1986

(n = 75)

RR

0.8 (0.6 to 1.1)

> 1 favours bladder training

NO

3(E×2, H)

1.3

Rai 2012

Anticholinergic drugs vs bladder training

IM

Park 20022

(n = 48)

RR

0.8 (0.4 to 1.5)

> 1 favours bladder training

NO

4(E×2, H×2)

6.2

Rai 2012

Anticholinergic drugs + non‐drug therapies vs anticholinergic drugs alone

IM

Park 20022

(n = 50)

RR

0.7 (0.4 to 1.6)

> 1 favours drugs alone

NO

4(E×2, H×2)

3.3

Rai 2012

Anticholinergic drugs vs external electrostimulation

IM

Aaronson 1995

(n = 38)

RR

1.3 (0.7 to 2.3)

> 1 favours ES

NO

3(E×2, H)

1.3

Rai 2012

Anticholinergic drugs vs bladder training

IM

Song 20062

(n = 58)

RR

0.8 (0.4 to 1.5)

> 1 favours bladder training

NO

3(E×2, H)

6.2

Rai 2012

Anticholinergic drugs + non‐drug therapies vs anticholinergic drugs alone

IM

Song 20062

(n = 63)

RR

0.8 (0.4 to 1.6)

> 1 favours drugs alone

NO

3(E×2, H)

3.11

Stewart 2016

ES vs tolterodine

IM

Preyer 20071

(n = 31)

MD

1.3 (−1.6 to 4.2)

< 0 favours ES

NO

3(E×2, H)

3.5.1

Stewart 2016

ES vs oxybutynin after 12 weeks

IM

Souto 2014

(n = 37)

RR

0.3 (0.1 to 1.5)

< 1 favours ES

NO

3(E×2, H)

3.5.2

Stewart 2016

ES vs oxybutynin after 24 weeks

FU1

Souto 2014

(n = 37)

RR

0.6 (0.1 to 1.8)

< 1 favours ES

NO

3(E×2, H)

2.3

Wallace 2004

Bladder training vs anticholinergic drugs

IM

Colombo 19954

(n = 75)

RR

1.1 (0.9 to 1.3)

> 1 favours bladder training

NO

3(E×2, H)

1Data were presented as mean differences within these trials.
2Song 2006 and Park 2002 were both three‐arm trials, both with groups: bladder training, anticholinergic drug and bladder training plus anticholinergic drug.
3These are the same data for Burgio 2010. This was the same trial, included in two different reviews (the trial name within Ayeleke 2015 was "Burgio 2010a").
4Data from Colombo 1995 were included within both Kirchin 2017 and Wallace 2004. The methods used to determine number cure or improved differed within the two reviews, resulting with different effect sizes; however, both reviews presented data derived from the same participant groups.
Abbreviations: CI: confidence interval; ES: electrical stimulation; FU1: follow‐up one; IM: immediate; MD: mean difference; n: number of participants assessed NOT to have been cured or improved using a variety of different assessment methods; PFMT: pelvic floor muscle training; RR: risk ratio.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

A total of 22 analyses from four reviews presented data on symptomatic cure and improvement comparing one conservative intervention with a non‐conservative intervention (Ayeleke 2015; Rai 2012; Stewart 2016; Wallace 2004). Twenty‐one were at immediate time points. There were 18 unique trials comparing a conservative with a non‐conservative intervention, with an immediate measure of cure of improvement.

High‐ or moderate‐certainty evidence

Nine analyses provided moderate‐certainty evidence. Three of these provided moderate‐certainty evidence of a difference between groups.

  • Conservative intervention (PFMT plus behavioural interventions) plus anticholinergic drugs had better cure or improvement than anticholinergic drugs alone.

  • Pharmacological intervention (trospium and solifenacin) was more effective than electrical stimulation.

  • Electrical stimulation plus drugs was more effective than drugs alone.

Six analyses provided moderate‐certainty evidence of no difference between groups.

  • Anticholinergic drugs versus external electrical stimulation.

  • Anticholinergic drugs plus PFMT versus anticholinergic drugs alone.

  • Electrical stimulation versus drugs (four analyses).

  • Electrical stimulation plus PFMT plus drugs versus drugs.

Low‐ or very low‐certainty evidence

Thirteen analyses provided low‐certainty evidence.

Condition‐specific quality of life

For more information, see Table 19.

Open in table viewer
Table 19. Urgency urinary incontinence: quality of life, conservative versus non‐conservative

Analysis code (in review)

Study ID

Comparisons included

Timing of measure

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

High‐certainty evidence

4.6.1

Stewart 2016

ES + PFMT + drugs vs drugs

IM

Kaya 2011

(n = 30)

MD

−16.2 (−28.0 to −4.4)

< 0 favours ES + other

YES favours ES + PFMT + drugs

0

Moderate‐certainty evidence

4.4.1

Stewart 2016

ES + drugs vs drugs alone

IM

Sancaktar 2010

(n = 38)

MD

−2.2 (−3.5 to −0.9)

< 0 favours ES + other

YES favours ES + drugs

1(H)

3.8.1

Stewart 2016

ES vs solifenacin succinate

IM

Vecchioli‐Scaldazza 2013

(n = 30)

MD

−0.6 (−1.3 to 0.1)

< 0 favours ES

YES favours ES + drugs

2(E×2)

Abbreviation: CI: confidence interval; ES: electrical stimulation; IM: immediate; MD: mean difference; n: number of participants; PFMT: pelvic floor muscle training.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Three analyses in one review included quality of life data from participants with UUI comparing conservative intervention with a non‐conservative intervention (Stewart 2016). There were three unique studies.

High‐ or moderate‐certainty evidence

One analysis provided high‐certainty evidence of a beneficial effect. Electrical stimulation plus drugs had better quality of life when compared to drugs alone.

Two analyses provided moderate‐certainty evidence of a beneficial effect. Electrical stimulation plus PFMT plus drugs had better quality of life compared to drugs alone.

All type of urinary incontinence

Any conservative intervention versus control, placebo or standard care
Symptomatic cure or improvement

For more information, see Table 20.

Open in table viewer
Table 20. All types of urinary incontinence: primary outcome participant‐reported cure or improvement; conservative versus control

Analysis code (in review)

Study ID

Comparisons included

Timing of measure

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

High‐certainty evidence

1.2.4

Dumoulin 2018

PFMT vs no treatment, placebo or control

IM

Diokno 2010; Goode 2002 (n = 166)

RR

2.4 (1.6 to 3.5)

> 1 favours PFMT

YES favours PFMT

0

1.6

Imamura 2015

Weight loss vs no active intervention

IM

Subak 2005 (n = 40)

RR

16.5 (1.0 to 270.8)

At 3 months; > 1 favours weight loss

YES favours weight loss

0

11.19.1

Stewart 2016

ES vs no active treatment, placebo or sham treatment

IM

Barroso 20042 (n = 36)

MD

−1.7 (−2.3 to −1.0)

< 0 favours ES

YES favours ES

0

Moderate‐certainty evidence

1.1

Herbison 2013

Cones vs control

IM

Bø 1999; Williams 20061 (n = 215)

RR

0.7 (0.5 to 1.0)

> 1 favours control

YES favours cones

2(E×2)

1.1

Imamura 2015

Weight loss vs no active intervention

FU1

Wing 2010 (n = 304)

RR

1.4 (1.1 to 1.7)

> 1 favours weight loss

YES favours weight loss

1(C)

1.6

Imamura 2015

Weight loss vs no active intervention

FU1

Wing 2010 (n = 304)

RR

1.8 (1.2 to 2.8)

At 6 months; > 1 favours weight loss

YES favours weight loss

1(C)

1.6

Imamura 2015

Weight loss vs no active intervention

FU2

Wing 2010 (n = 287)

RR

1.1 (0.9 to 1.5)

At 18 months; > 1 favours weight loss

NO

2(C, E)

Low‐certainty evidence

1.6

Imamura 2015

Weight loss vs no active intervention

O

Phelan 2012;

Wing 2010 (n = 1032)

RR

1.2 (1.0 to 1.4)

> 1 favours weight loss

YES favours weight loss

3(C, E×2)

1.2

Stewart 2017

ES vs no active treatment

IM

Bø 1999; Castro 2008; Henalla 1989; Lopès 2014; Oldham 2013 (n = 347)

RR

1.73 (1.41 to 2.11)

> 1 favours ES

YES for ES

GRADE as assessed by review authors

1.20.4

Wallace 2004

Bladder training vs no treatment

IM

Fantl 1991 (n = 123)

RR

3.1 (2.0 to 5.0)

> 1 favours Bladder Training

YES favours BT

3(G, H×2)

Very low‐certainty evidence

1.1

Wieland 2019

Yoga vs waiting list control

IM

Huang 2014 (n = 18)

RR

6.33 (1.44 to 27.88)

> 1 favours yoga

YES for yoga

GRADE as assessed by review authors

1Bø 1999 contained only participants with SUI, but when pooled with Williams 2006 the pooled total contained a mixed population. Results from Williams 2006 are presented within the forest plot, but the individual study results from Bø 1999 are presented alongside the SUI data.
2Barroso 2004 presented data as a mean difference, rather than number cure or improved, and, therefore, these data are not included in the forest plot.
3Only data relating to immediate outcome measures were considered for inclusion in forest plots.
Abbreviations: CI: confidence interval; ES: electrical stimulation; FU1: follow‐up one; FU2: follow‐up two; IM: immediate; MD: mean difference; n: number of participants; O: other; PFMT: pelvic floor muscle training; RR: risk ratio.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Eleven analyses from seven reviews assessed symptomatic cure and improvement and compared a conservative intervention with a control intervention (Dumoulin 2018; Herbison 2013; Imamura 2015; Stewart 2016; Stewart 2017; Wallace 2004; Wieland 2019). Seven outcomes were assessed at an immediate time point, three at follow‐up time points and one was unclear. There were 13 unique trials contained within the analyses.

High‐ or moderate‐certainty evidence

Three analyses provided high‐certainty and four moderate‐certainty evidence. Six analyses provided high‐ or moderate‐certainty evidence for conservative intervention compared to control, for measures of cure or improvement.

  • PFMT had better cure or improvement compared to no treatment, control or placebo (high‐certainty).

  • Electrical stimulation had better cure or improvement compared to no treatment (high‐certainty).

  • Weight loss interventions had a beneficial effect compared to no active treatment at immediate outcome (high‐certainty) and six‐month follow‐up outcome measured by self‐report and by a seven‐day voiding diary (two analyses, moderate‐certainty).

  • Cones had better cure or improvement compared to control (moderate‐certainty).

One analysis provided moderate‐certainty evidence of no difference in cure or improvement for weight loss interventions compared to no active treatment at a longer term (18 months) follow‐up (moderate‐certainty).

Low‐ or very low‐certainty evidence

Four analyses provided low‐ or very low‐certainty evidence.

Condition‐specific quality of life

For more information, see Table 21.

Open in table viewer
Table 21. All types of urinary incontinence: primary outcome quality of life; conservative versus control

Analysis code (in review)

Study ID

Comparisons included

Outcome measure (timing of measure)

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

High‐certainty evidence

1.11.4

Dumoulin 2018

PFMT vs no treatment, placebo or control

Incontinence Impact Questionnaire Long Form (IM)

Sran 2016 (n = 48)

MD

−52.67 (−95.00 to −10.34)

> 0 favours control

YES favours PFMT

0

1.16.4

Dumoulin 2018

PFMT vs no treatment, placebo or control

Incontinence Impact Questionnaire Long Form (FU)

Sran 2016 (n = 48)

MD

−41.91 (−83.20 to −0.62)

> 0 favours control

YES favours PFMT

0

Moderate‐certainty evidence

1.12.4

Dumoulin 2018

PFMT vs no treatment, placebo or control

Incontinence of Quality of Life Questionnaire (IM)

Sar 2009

(n = 34)

MD

28.9 (35.1 to 22.7)

> 0 favours control

YES favours PFMT

1(A) no change

1.10.4

Dumoulin 2018

PFMT vs no treatment, placebo or control

Incontinence Impact Questionnaire Short Form (IM)

Celiker Tosun 2015; Leong 2015 (n = 176)

MD

−7.54 (−14.70 to −0.39)

> 0 favours control

YES favours PFMT

2(F×2)

Very low‐certainty evidence

1.2

Wieland 2019

Yoga vs waiting list control

IM

Huang 2014 (n = 18)

MD

1.7 (−33.0 to 36.5)

> 0 favours control

NO

GRADE as assessed by review authors

Abbreviations: CI: confidence interval; FU: follow‐up; IM: immediate; MD: mean difference; n: number of participants; PFMT: pelvic floor muscle training.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Five analyses from two reviews presented quality of life data from participants with AUI (Dumoulin 2018; Wieland 2019). These analyses contained data from five unique studies.

High‐ or moderate‐certainty evidence

Two analyses provided high‐certainty and two moderate‐certainty evidence that PFMT was more beneficial than no treatment, control or placebo on measures of quality of life at an immediate time point.

Very low‐certainty evidence

One analysis provided very low‐certainty evidence.

One conservative intervention versus another conservative intervention
Symptomatic cure or improvement

For more information, see Table 22.

Open in table viewer
Table 22. All types of urinary incontinence: primary outcome participant‐reported cure or improvement; conservative versus conservative

Analysis code (in review)

Study ID

Comparisons included

Outcome measure (timing of measure)

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

High‐certainty evidence

1.2.31

Hay‐Smith 2011

Home PFMT programme (unsupervised) vs PFMT (Sapsford style2) with 8 clinic visits

IM

Hung 2010(n = 64)

RR

0.1 (0 to 0.7)

> 1 favours less contact/supervision

YES favours more contact/supervision

0

Moderate‐certainty evidence

3.2.1

Ayeleke 2015

PFMT + bladder training vs bladder training alone

IM

Wyman 1998(n = 128)

RR

1.3 (1.0 to 1.8)

> 1 favours PFMT + bladder training

YES favours PFMT + BT

2(C, H)

12.2.3

Hay‐Smith 2011

More‐intensive vs less‐intensive PFMT programmes

IM

de Oliveira 20093; Delgado 2009; Hay‐Smith 20023; Sriboonreung 20113; Sugaya 20033; Wells 1999(n = 405)

RR

0.7 (0.6 to 0.9)

> 1 favours less intensive

YES favours more intensive

1(H)

10.3

Wallace 2004

Bladder training + non‐pharmacological treatments vs non‐pharmacological treatments (PFMT)

IM

Wyman 1998(n = 124)

RR

1.2 (1.0 to 1.4)

> 1 favours bladder training + non‐pharmacological

YES favours BT + PFMT

2(G, H)

Low‐certainty evidence

3.5.1

Ayeleke 2015

PFMT + bladder training vs bladder training alone

IM

Kaya 2014; Wyman 1998

(n = 235)

RR

1.3 (1.1 to 1.4)

> 1 favours PFMT + bladder training

YES favours PFMT + BT

4(C, E×2, H)

3.2.2

Ayeleke 2015

PFMT + bladder training vs bladder training alone

FU1

Wyman 1998

(n = 120)

RR

1.3 (0.9 to 1.8)

> 1 favours PFMT + bladder training

NO

3(E×2, H)

3.5.2

Ayeleke 2015

PFMT + bladder training vs bladder training alone

FU1

Wyman 1998

(n = 119)

RR

1.2 (0.9 to 1.5)

> 1 favours PFMT + bladder training

NO

3(E×2, H)

12.2.1

Hay‐Smith 2011

More‐intensive vs less intensive PFMT programmes

IM

Bø 19903; Felicíssimo 20103; Schagen van Leeuwen 20043; Hung 2010; Konstantinidou 20073; Ramsay 19903

(n = 335)

RR

0.4 (0.2 to 0.8)

> 1 favours less intensive

YES more intense

3(E×2, H)

10.2

Hay‐Smith 2011

PFMT with intravaginal resistance device vs PFMT alone

IM

Delgado 2009; Wells 1999(n = 120)

RR

0.9 (0.6 to 1.2)

> 1 favours no device

NO

4(A, E×2, H)

1.2.2

Herderschee 2011

PFMT + biofeedback vs PFMT alone

IM

Burgio 20024; Goode 2003; Pages 2001; Wang 20044; Wilson 1987

(n = 343)

RR

0.7 (0.6 to 0.8)

> 1 favours PFMT alone

YES favours PFMT + BF

4(C, E×2, H)

1.2

Herderschee 2011

PFMT + biofeedback vs PFMT alone

IM

Burgio 20024; Goode 20033; Pages 20013; Wang 20044; Wilson 19873; Burns 1993; Morkved 20033

(n = 520)

RR

0.7 (0.7 to 0.9)

> 1 favours PFMT alone

YES favours BF

4(C, E×2, H)

1.2.1

Herderschee 2011

PFMT + biofeedback vs PFMT alone

IM

Burns 1993; Morkved 20033

(n = 177)

RR

0.9 (0.7 to 1.0)

> 1 favours PFMT alone

NO

4(E×2, H×2)

4.2.2

Herderschee 2011

PFMT + biofeedback vs PFMT + feedback

IM

Johnson 2000(n = 20)

RR

1.0 (0.2 to 5.8)

> 1 favours PFMT + feedback

NO

3(B, E×2)

4.2

Herderschee 2011

PFMT + biofeedback vs PFMT + feedback

IM

Burgio 20024; Johnson 2000

(n = 130)

RR

1.0 (0.6 to 1.6)

> 1 favours PFMT + feedback

NO

4(E×2, H×2)

6.2

Lipp 2014

Intraurethral Reliance device vs intraurethral FemAssist device

IM

Boos 1998(n = 101)

RR

1.0 (0.5 to 2.0)

> 1 favours reliance

NO

3(B, E×2)

5.3

Wallace 2004

Bladder training + behavioural/physical/psychological treatments

IM

Wyman 1998(n = 129)

RR

0.9 (0.7 to 1.1)

> 1 favours bladder training

NO

4(C, E×2, G)

5.4

Wallace 2004

Bladder training + behavioural/physical/psychological treatments

FU1

Wyman 1998 (n = 124)

RR

0.9 (0.7 to 1.1)

> 1 favours bladder training

NO

4(E×2, G, H)

10.4

Wallace 2004

Bladder training + non‐pharmacological treatments vs non‐pharmacological treatments

FU1

Wyman 1998 (n = 122)

RR

1.1 (0.9 to 1.3)

> 1 favours bladder training + non‐pharmacological

NO

4(E×2, G, H)

Very low‐certainty evidence

2.1

Herbison 2013

Cones vs PFMT

IM

Bø 19993; Cammu 19983; Peattie 1988; Williams 2006; Arvonen 20013; Arvonen 20023

(n = 358)

RR

0.9 (0.7 to 1.2)

> 1 favours PFMT

NO

5(C, F×2, G, H)

3.1

Herbison 2013

Cones vs ES

IM

Bø 19993; Oláh 1990; Santos 2009(n = 151)

RR

1.3 (0.8 to 1.9)

> 1 favours ES

NO

6(E×2, F×2, G, H)

3.2

Herbison 2013

Cones v ES

IM

Bø 19993; Castro 20083; Oláh 1990

(n = 154)

RR

1.2 (0.9 to 1.6)

> 1 favours ES

NO

6(E×2, F×2, G, H)

2.1

Wieland 2019

Yoga vs mindfulness stress reduction

IM

Baker 2014 (n = 24)

RR

0.09 (0.01 to 1.43)

> 1 favours yoga

NO

GRADE as assessed by review authors

2.2

Wieland 2019

Yoga vs mindfulness stress reduction

FU1

Baker 2014 (n = 20)

RR

0.2 (0 to 1.4)

> 1 favours yoga

NO

GRADE as assessed by review authors

2.3

Wieland 2019

Yoga vs mindfulness stress reduction

FU2

Baker 2014 (n = 21)

RR

0.2 (0 to 1.5)

> 1 favours yoga

NO

GRADE as assessed by review authors

1The data from Hung 2010 for this comparison and outcomes were included in both Analysis 1.2.3 and Analysis 3.2.3 in Hay‐Smith 2011. While the study data were the same in these analyses, the reported risk ratio differed. We are uncertain about the risk ratio presented in Analysis 3.2.3 and, therefore, chose to present the result from Analysis 1.2.3.
2Sapsford‐style PFMT involved diagrammatic and transversus abdominus co‐ordination to 'activate' pelvic floor muscle contraction and no direct pelvic floor muscle contractions.
3Trials with participants who had stress urinary incontinence only, where these were pooled with trials with other populations to give a pooled total relating to a combined participant population.
4Trials with participants who had urgency urinary incontinence only, where these were pooled with trials with other populations to give a pooled total relating to a combined participant population.
Abbreviations: CI: confidence interval; ES: electrical stimulation; FU: follow‐up; IM: immediate; n: number of participants assessed NOT to have been cured or improved using a variety of different assessment methods; PFMT: pelvic floor muscle training; RR: risk ratio.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

A total of 25 analyses from seven reviews presented data assessing symptomatic cure or improvement comparing one conservative intervention with another conservative intervention (Ayeleke 2015; Hay‐Smith 2011; Herderschee 2011; Herbison 2013; Lipp 2014; Wallace 2004; Wieland 2019). Of these, 19 related to immediate outcome assessment and six to follow‐up assessment. Eleven analyses compared the effect of different conservative interventions and 11 analyses compared different methods of delivering the same conservative intervention (in 10 analyses this was different ways of delivering PFMT and in one analysis this was different types of IU devices). These analyses pooled data from 32 unique trials. Of these 32 trials, 13 had participants with mixed, combined or unclear types of UI. The remaining 19 trials included participants with either SUI or UUI.

High‐ or moderate‐certainty evidence

Two analyses provided high‐certainty evidence and three moderate‐certainty evidence of a benefit of one conservative intervention compared to another.

Where this was comparing different types of conservative interventions, PFMT plus bladder training had better cure or improvement than bladder training alone or PFMT alone (two analyses, both containing the same single trial, moderate‐certainty).

Where this was comparing different ways of delivering a conservative intervention:

  • a Sapsord style PFMT programme with eight clinic visits had better cure or improvement than an unsupervised home PFMT programme (high‐certainty). Sapsford‐style PFMT involved diagrammatic and transversus abdominus co‐ordination to 'activate' pelvic floor muscle contraction and no direct pelvic floor muscle contractions;

  • 'more‐intensive' PFMT programmes had better cure or improvement than 'less‐intensive' PFMT programmes (moderate‐certainty).

Low‐ or very low‐certainty evidence

Fourteen analyses provided low‐certainty evidence and six very low‐certainty evidence.

Condition‐specific quality of life

For more information, see Table 23.

Open in table viewer
Table 23. All types of urinary incontinence: quality of life, conservative versus conservative

Analysis code (in review)

Study ID

Comparisons included

Outcome measure (timing of measure)

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

Moderate‐certainty evidence

3.3.1

Ayeleke 2015

PFMT + bladder training vs bladder training alone

IM

Wyman 19981

(n = 127)

MD

−25.5 (−49.9 to −1.0)

> 0 favours bladder training

YES favours PFMT + bladder training

1(H)

3.4.1

Ayeleke 2015

PFMT + bladder training vs bladder training alone

IM

Wyman 19982

(n = 128)

MD

−31.1 (−50.0 to −13.2)

> 0 favours bladder training

YES

favours PFMT + bladder training

2(C, H)

10.12

Wallace 2004

Bladder training + PFMT + biofeedback vs PFMT + biofeedback

IM

Wyman 19982

(n = 35)

MD

−47.2 (−87.0 to −7.4)

> 0 favours control

YES favours bladder training + PFMT +

biofeedback

2(G, H)

Low‐certainty evidence

3.3.2

Ayeleke 2015

PFMT + bladder training vs bladder training alone

FU1

Wyman 1998

(n = 118)

MD

−5.9 (−35.5 to 23.7)

> 0 favours bladder training

NO

3(E×2, H)

3.4.2

Ayeleke 2015

PFMT + bladder training vs bladder training alone

FU1

Wyman 1998

(n = 118)

MD

−18.9 (−38.0 to 0.1)

> 0 favours bladder training

NO

3(E×2, H)

6.7

Lipp 2014

Intraurethral Reliance device vs intraurethral FemAssist device

IM

Boos 19983

(n = 67)

MD

−1.0 (−12.6 to 10.6)

> 0 favours FemAssist

NO

3(B, E×2)

6.7

Lipp 2014

Intraurethral Reliance device vs intraurethral FemAssist device

IM

Boos 19983

(n = 67)

MD

−6.9 (−15.5 to 1.7)

> 0 favours FemAssist

NO

3(B, E×2)

6.7

Lipp 2014

Intraurethral Reliance device vs intraurethral FemAssist device

IM

Boos 19983

(n = 67)

MD

−7.4 (−16.7 to 1.9)

> 0 favours FemAssist

NO

3(B, E×2)

6.7

Lipp 2014

Intraurethral Reliance device vs intraurethral FemAssist device

IM

Boos 19983

(n = 67)

MD

5.2 (−5.0 to 15.4)

> 0 favours FemAssist

NO

3(B, E×2)

6.7

Lipp 2014

Intraurethral Reliance device vs intraurethral FemAssist device

IM

Boos 19983

(n = 67)

MD

−1.0 (−12.6 to 10.6)

> 0 favours FemAssist

NO

3(B, E×2)

6.7

Lipp 2014

Intraurethral Reliance device vs intraurethral FemAssist device

IM

Boos 19983

(n = 67)

MD

1.1 (−9.5 to 11.8)

> 0 favours FemAssist

NO

3(B, E×2)

6.7

Lipp 2014

Intraurethral Reliance device vs intraurethral FemAssist device

IM

Boos 19983

(n = 67)

MD

1.9 (−6.9 to 10.7)

> 0 favours FemAssist

NO

3(B, E×2)

10.13

Wallace 2004

Bladder training + non‐pharmacological treatment vs non‐pharmacological treatment

FU1

Wyman 1998

(n = 122)

MD

−12.2 (−30.4 to 6.0)

> 0 favours control

NO

4(E×2, G, H)

5.14.4

Wallace 2004

Bladder training vs other behavioural/physical/psychological treatments

IM

Wyman 19982

(n = 38)

MD

−28.0 (−68.4 to 12.4)

> 0 favours control

NO

4(E×2, G, H)

5.15

Wallace 2004

Bladder training vs other behavioural/physical/psychological treatments

FU1

Wyman 1998

(n = 124)

MD

6.7 (−12.2 to 25.6)

> 0 favours control

NO

4(E×2, G, H)

Very low‐certainty evidence

6.3

Stewart 2017

ES + PFMT vs PFMT

IM

Beuttenmuller 2010; Haig 1995; Patil 2010; Schmidt 2009 (n = 193)

SMD

−0.35 (−0.64 to −0.05)

< 0 favours ES + PFMT

YES for ES + PFMT

GRADE as assessed by review authors

1Outcome measured by the Incontinence Impact Questionnaire – Revised.
2Outcome measured by the Urinary Distress Inventory, for different pairs of treatment groups from this three‐arm trial.
3Boos 1998 compared two different types of mechanical devices.
Abbreviations: CI: confidence interval; FU1: follow‐up one; IM: immediate; MD: mean difference; n: number of participants; PFMT: pelvic floor muscle training; SMD: standardised mean difference.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Sixteen analyses from four reviews presented data on quality of life for comparisons of one conservative intervention with another conservative intervention (Ayeleke 2015; Lipp 2014; Stewart 2017; Wallace 2004). Of these, 12 related to an immediate measure of quality of life and four to follow‐up measures. Nine analyses compared the effects of two different types of conservative interventions and seven compared two similar types of interventions. Six unique trials contributed data to these analyses.

High‐ or moderate‐certainty evidence

Three analyses provide moderate‐certainty evidence of higher quality of life with one conservative intervention compared to another conservative intervention.

  • PFMT plus bladder training resulted in better quality of life than bladder training alone at an immediate time point measured using the IIQ‐R and the UDI (two analyses, moderate‐certainty).

  • PFMT plus bladder training resulted in better quality of life than PFMT alone at an immediate time point measured using the UDI (moderate‐certainty).

Low‐ or very low‐certainty evidence

Thirteen analyses provide low‐ or very low‐certainty evidence.

Any conservative intervention versus another active intervention (non‐conservative)
Symptomatic cure or improvement

For more information, see Table 24.

Open in table viewer
Table 24. All types of urinary incontinence: primary outcome participant‐reported cure or improvement; conservative versus non‐conservative

Analysis code (in review)

Study ID

Comparisons included

Outcome measure (timing of measure)

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

Moderate‐certainty evidence

2.1.2

Alhasso 2005

Adrenergic agonist vs conservative therapy

UC

Wells 1991

(n = 157)

RR

1.4 (1.1 to 1.8)

> 1 favours drug

YES favours drugs

1(H)

Low‐certainty evidence

2.3.2

Alhasso 2005

Adrenergic agonist vs conservative therapy

UC

Wells 1991

(n = 86)

RR

0.9 (0.5 to 1.4)

> 1 favours drug; differential dropout from groups, more from PFMT group (P < 0.01)

NO

4(C, E×2, H)

1.3.1

Rai 2012

Anticholinergic drugs vs bladder training

IM

Macaulay 1987

(n = 29)

RR

1.0 (0.7 to 1.3)

> 1 favours bladder training

NO

3(E×2, H)

Abbreviations: CI: confidence interval; IM: immediate; n: number of participants; RR: risk ratio; UC: unclear.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Three analyses from two reviews assessed symptomatic cure or improvement comparing one conservative intervention with a non‐conservative intervention (Alhasso 2005; Rai 2012). One outcome was at an immediate time point and two were unclear. There were two unique trials comparing a conservative intervention with a non‐conservative intervention but only one with an immediate measure of cure of improvement.

High‐ or moderate‐certainty evidence

One analysis provided moderate‐certainty evidence of better cure or improvement with a non‐conservative compared with conservative intervention. An adrenergic agonist (phenylpropanolamine) was more effective than conservative intervention (PFMT), although the timing of the outcome measure was unclear.

Low‐ or very low‐certainty evidence

Two analyses provided low‐certainty evidence.

Condition‐specific quality of life

For more information, see Table 25.

Open in table viewer
Table 25. All types of urinary incontinence: quality of life; conservative versus non‐conservative

Analysis code (in review)

Study ID

Comparisons included

Outcome measure (timing of measure)

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

Moderate‐certainty evidence

2.13

Wallace 2004

Bladder training vs anticholinergic drugs

IM

Herbison 20041

(n = 34)

MD

−8.0 (−18.8 to 2.8)

> 0 favours bladder training

NO

2(E×2)

Low‐certainty evidence

9.13

Wallace 2004

Bladder training + pharmacological treatments vs pharmacological treatment alone

IM

Herbison 20041

(n = 28)

MD

2.0 (−6.8 to 10.8)

> 0 favours anticholinergic + bladder training

NO

3(C, E×2)

1Herbison 2004 was a three‐arm trial, where the groups were: bladder training, anticholinergic drug (oxybutynin) and bladder training + anticholinergic drugs. The anticholinergic drug group data within both analyses are the same.
Abbreviations: CI: confidence interval; IM: immediate; MD: mean difference; n: number of participants.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Two analyses from one review compared one conservative intervention with a non‐conservative intervention assessing quality of life (Wallace 2004). Both analyses had outcome measures at an immediate time point. The two analyses included only one unique trial (Herbison 2004).

High‐ or moderate‐certainty evidence

One analysis provided moderate‐certainty evidence of no difference between groups.

  • Bladder training versus anticholinergic drugs measured using the OAB Questionnaire.

Low‐ or very low‐certainty evidence

One analysis provided low‐certainty evidence.

Discussion

Evidence relating to conservative management of urinary incontinence in women

For summary of findings tables by type of UI, see Table 26, Table 27, and Table 28.

Open in table viewer
Table 26. Stress urinary incontinence: summary of findings – high‐ or moderate‐certainty evidence of beneficial effect on primary outcomes

Comparison group

Intervention of greater benefit

Comparison intervention

High‐ or moderate‐certainty evidence relating to cure or improvement

High‐ or moderate‐certainty evidence relating to quality of life

Review

Conservative intervention vs control

(see Table 8 and Table 9)

Cones

Control

Herbison 2013

PFMT + biofeedback

PFMT alone

Herderschee 2011

PFMT

Control

Dumoulin 2018

Intravaginal mechanical device

No treatment

Lipp 2014

ES

No treatment

Stewart 2017

1 conservative intervention vs another conservative intervention

(see Table 10 and Table 11)

Pessary + PFMT

Pessary alone

Lipp 2014

PFMT

Cones

Herbison 2013

PFMT

Intravaginal pessary

Lipp 2014

PFMT + adherence strategy

PFMT alone

Hay‐Smith 2011

More‐intensive PFMT

Less‐intensive PFMT

Hay‐Smith 2011

Conservative intervention vs non‐conservative intervention

(see Table 12 and Table 13)

Acupuncture

Any other treatment

Wang 2013

SNRI

Conservative

Mariappan 2005

PFMT

Urethral injection therapy

Kirchin 2017

Open retropubic colposuspension

Conservative

Lapitan 2017

ES: electrical stimulation; PFMT: pelvic floor muscle training; SNRI: serotonin–noradrenaline reuptake inhibitor.

Open in table viewer
Table 27. Urgency urinary incontinence: summary of findings – high‐ or moderate‐certainty evidence of beneficial effect on primary outcomes

Comparison group

Intervention of greater benefit

Comparison intervention

High‐ or moderate‐certainty evidence relating to cure or improvement

High‐ or moderate‐certainty evidence relating to quality of life

Review

Conservative intervention vs control

(see Table 14 and Table 15)

PFMT + feedback/biofeedback

Control

Herderschee 2011

ES

Control

Stewart 2016

Bladder training

Control

Wallace 2004

ES + PFMT

No active treatment

Stewart 2016

1 conservative intervention vs another conservative intervention

(see Table 16 and Table 17)

ES

Laseropuncture

Stewart 2016

ES + PFMT

PFMT alone

Stewart 2016

ES

PFMT

Stewart 2016

Conservative intervention vs non‐conservative intervention

(see Table 18 and Table 19)

Trospium + solifenacin

ES

Stewart 2016

ES + drugs

Drugs

Stewart 2016

ES + PFMT + drugs

Drugs

Stewart 2016

Anticholinergic drugs + behavioural interventions (PFMT)

Anticholinergic drugs

Rai 2012

ES: electrical stimulation; PFMT: pelvic floor muscle training.

Open in table viewer
Table 28. All types of urinary incontinence: summary of findings – high‐ or moderate‐certainty evidence of beneficial effect on primary outcomes

Comparison group

Intervention of greater benefit

Comparison intervention

High‐ or moderate‐certainty evidence relating to cure or improvement

High‐ or moderate‐certainty evidence relating to quality of life

Review

Conservative intervention vs control

(see Table 20 and Table 21)

PFMT

Control

Dumoulin 2018

Weight loss

No active intervention

Imamura 2015

ES

Control

Stewart 2016

Cones

Control

Herbison 2013

1 conservative intervention vs another conservative intervention

(see Table 22 and Table 23)

Home PFMT programme (no health professional supervision)

Structured programme of 'indirect' PFMT (every 2 weeks, individual, health professional contact)

Hay‐Smith 2011

PFMT + bladder training

Bladder training

Ayeleke 2015

More‐intensive PFMT

Less‐intensive PFMT

Hay‐Smith 2011

PFMT + bladder training

PFMT

Wallace 2004

PFMT + biofeedback + bladder training

PFMT + biofeedback

Wallace 2004

Conservative intervention vs non‐conservative intervention

(see Table 24 and Table 25)

Adrenergic agonist

Conservative interventions

Alhasso 2005

ES: electrical stimulation; PFMT: pelvic floor muscle training.

There are 29 Cochrane Reviews reporting evidence relevant to the conservative management of UI in women (see Table 1). There were 12 Cochrane Protocols relevant to this topic but two have been withdrawn, leaving a further 10 potential reviews (see Appendix 6).

Of the 29 included Cochrane Reviews, 18 contained data and analyses relating to the primary outcomes and comparisons of relevance to this overview (including multiple additional analyses relating to a wide range of secondary outcomes (see Appendix 7), data from which have not been synthesised within this overview). Seventy per cent of the analyses (134/192) related to the effect of conservative management of UI on symptomatic cure or improvement, and 30% (58/192) related to condition‐specific quality of life. Eighty‐one per cent of these analyses included data from only one trial. Therefore, this body of evidence comprises a relatively large number of Cochrane Reviews, synthesising a relatively small number of trials within a relatively large number of analyses.

There was a lack of high‐certainty evidence relating to conservative management of UI. We judged only 9% (17/192) of analyses to provide high‐certainty evidence relating to one of our primary outcomes. Most of the evidence relating to the effect of conservative interventions was low‐ to moderate‐certainty. We found that analyses from populations with UUI were more likely to be of moderate‐certainty, while those from populations with AUI were more likely to be of low‐certainty. However, analyses from populations with UUI were more likely to include data from only one trial.

For SUI, the main focus of research was PFMT. PFMT was primarily compared to a control intervention or to the addition of biofeedback. Further comparisons included adding electrical stimulation or intravaginal devices to a PFMT programme. There was a limited amount reported on mechanical devices, electrical stimulation, Macroplastique injections or drugs. For UUI, pharmaceutical interventions were the most frequently studied and these were primarily compared with electrical stimulation, education, behavioural or lifestyle advice, and PFMT in various combinations. There was a particular lack of evidence relating to quality of life outcomes for women with UUI, with the only analyses relating to this outcome focused on comparisons with electrical stimulation. For AUI, similar to SUI, the main focus of research was PFMT delivered with and without a variety of adjuncts. There was some evidence relating to the effect of educational and lifestyle interventions on quality of life for women with AUI, delivered both with and without PFMT and adjuncts.

Assessment of the quality of the reviews and the certainty of evidence within the Cochrane Reviews highlighted several methodological limitations relating to reporting, quality and definition of key parameters. This arguably presents challenges to clinicians and policymakers, who will have to be familiar with a large number of reviews and aware of the overlapping evidence between reviews in order to make informed evidence‐based clinical decisions. Therefore, this overview has an important role in synthesising the best evidence on conservative interventions for UI into a single, accessible, comprehensive document to signpost clinicians and policymakers towards relevant Cochrane Reviews to support clinical decisions.

Role of the stakeholder group

Conservative interventions of relevance to this overview were identified at the protocol stage by a stakeholder group, using consensus decision‐making techniques based on the nominal group technique. Six broad categories of conservative intervention were identified, each with multiple subcategories (see Figure 1). The Cochrane Reviews included in this overview assessed trials of interventions from four of these categories, with most focusing on physical therapy, educational, behavioural or lifestyle interventions, or mechanical devices (see Table 29). There was no evidence specifically relating to the predefined subcategory of psychological therapies within the included Cochrane Reviews. Moreover, exploration of intervention descriptions revealed that interventions were often delivered in combination, rather than singularly, and that reporting within the trials of the details of the delivered intervention and the 'control' were often poor. This causes significant challenges in the synthesis and interpretation of the available evidence.

Open in table viewer
Table 29. Identification of relevant reviews mapped to six primary conservative categories

Primary conservative intervention (categories)

Review

Educational/behavioural/lifestyle

Imamura 2015 – lifestyle intervention
Wallace 2004 – bladder training
Ostaszkiewicz 2004a – habit training
Ostaszkiewicz 2004a – timed voiding
Eustice 2000 – prompted voiding

Physical therapies

Ayeleke 2015 – pelvic floor muscle training + another active treatment vs the same active treatment alone
Dumoulin 2018 – pelvic floor muscle training vs no treatment, or inactive control treatments
Hay‐Smith 2011 – comparisons of approaches to pelvic floor muscle training
Herderschee 2011 – feedback or biofeedback
Herbison 2013 – cones
Stewart 2016 and Stewart 2017 – electrical stimulation

Psychological therapies

Mechanical devices

Lipp 2014 – mechanical devices

Complementary therapies

Wang 2013 – acupuncture

Others

Wieland 2019 – yoga

The stakeholder group also highlighted the need to present evidence according to populations of women with SUI, MUI or UUI, and we have used these as far as possible following the International Continence Society definitions (Haylen 2010 and, if appropriate, D'Ancona 2019). However, there are substantial inconsistencies within the literature in terms of diagnosis and definition of these conditions and, subsequently, some trials and Cochrane Reviews include mixed populations while others are limited to a population with a specific diagnosis or set of symptoms. In practice, we found that we were often unable to distinguish between 'mixed', 'combined' and 'unclear' populations, and, therefore, brought all data relating to mixed, combined or unclear populations into a group for all types of UI (AUI) (see Differences between protocol and review). These challenges and variations within the current evidence base again highlight the need for an accessible overview that summarises the evidence in a comprehensive, consistent and coherent manner.

Summary of main results

Evidence of effectiveness of conservative interventions

Evidence from the trials in the included Cochrane Reviews has been collated within numerous analyses. Around 25% of the reported analyses related to our first primary outcome of cure or improvement but very few (only around 10%) related to our second primary outcome of participant‐reported quality of life. Interpretation of the results of the high‐certainty and moderate‐certainty evidence was challenging due to a number of factors, in particular the investigation of multimodal (rather than singular) interventions, poorly defined and described interventions and control group interventions, and the wide variety of outcome measures and methods of reporting and analysis.

Based on our assessment of the evidence, we are highly certain that PFMT is more beneficial than control for all types of UI for the primary outcomes of cure or improvement and quality of life. We are moderately certain that if the PFMT is more intense, more frequent, with individual supervision, or combined with behavioural interventions or an adherence strategy effectiveness is improved (or a combination of these) effectiveness is improved. We are also highly certain that for our primary outcome of cure or improvement there was benefit associated with the use of cones compared to a control for women with SUI (but not compared to PFMT), electrical stimulation was beneficial for women with UUI and weight loss reduced symptoms in women with AUI compared to control.

Conservative interventions with little or no evidence

Of the six primary conservative intervention categories identified by our stakeholders, four were covered by relevant Cochrane Reviews (see Table 29). The Cochrane Reviews identified a lack of evidence relating to the use of psychological therapies for the conservative management of UI, despite the fact that these approaches are increasingly used within clinical practice. The only Cochrane Review of complementary therapies in this topic area focused on acupuncture, and identified and included only one trial.

We identified five Cochrane Reviews relating to education, behaviour or lifestyle interventions. There was no high‐ or moderate‐certainty evidence for most lifestyle interventions, despite the fact that anecdotal evidence suggests that interventions such as fluid reduction or change in caffeine, diet, exercise smoking cessation and education are frequently delivered within clinical practice. Although four Cochrane Reviews synthesised evidence relating to voiding interventions, they all had very limited evidence, perhaps highlighting the difficulty in undertaking relevant research in these areas where often there are multicausative factors, multimorbidities and participants may be older and potentially in residential care.

We found a lack of evidence relating to a number of physical therapy interventions, including the use of manual therapies, ultrasound, and dynamometry biofeedback and magnetic stimulation. Evidence relating to the effect of electrical stimulation was only considered in the population of women with OAB.

Summary of evidence of effect of key interventions

Stress urinary incontinence

  • Fourteen Cochrane Reviews synthesised evidence relating to SUI. These contained 84 analyses relevant to the primary outcomes of this overview.

  • Most evidence was moderate‐ or low‐certainty.

  • There was some high‐ or moderate‐certainty evidence relating to cones, PFMT and intravaginal mechanical devices.

  • There was relatively little evidence relating to primary outcomes, particularly quality of life, and virtually no long‐term follow‐up data.

Table 26 summarises where there was moderate‐ or high‐certainty evidence of a benefit of one intervention, for primary outcomes of cure of improvement or quality of life, and signposts the reviews within which this evidence was synthesised.

Urgency urinary incontinence

  • Seven Cochrane Reviews synthesised evidence relating to UUI, but only five contained analyses relevant to the primary outcomes of this overview. These contained 47 relevant analyses relevant to the primary outcomes of this overview.

  • Most evidence was moderate‐ or low‐certainty.

  • There was some moderate‐ and high‐certainty evidence indicating effectiveness of PFMT, electrical stimulation and bladder training when compared to no treatment or a control treatment.

  • There was relatively little evidence relating to primary outcomes, particularly quality of life, and virtually no long‐term follow‐up data.

Table 27 summarises where there was moderate‐ or high‐certainty evidence of a benefit of one intervention, for primary outcomes of cure of improvement or quality of life, and signposts the Cochrane Reviews within which this evidence was synthesised.

All types of urinary incontinence

  • Thirteen Cochrane Reviews synthesised evidence relating to AUI. These contained 61 analyses relevant to the primary outcomes of this overview.

  • Most evidence was moderate‐ or low‐certainty, but there was some high‐certainty evidence relating to weight loss and PFMT.

  • There was some moderate‐ or high‐certainty evidence that PFMT, weight loss, electrical stimulation and cones may be more beneficial than control (no treatment or placebo treatment), and that PFMT that was more intense, more frequent, with individual supervision, or combined with behavioural interventions or an adherence strategy effectiveness (or a combination of these) may be more beneficial than less‐intense or frequent PFMT, or PFMT that has no individual supervision, and is not combined with behavioural interventions or adherence strategies.

  • There was relatively little evidence relating to primary outcomes, particularly quality of life, and virtually no long‐term follow‐up data.

Table 28 summarises where there was moderate‐ or high‐certainty evidence of a benefit of one intervention, for primary outcomes of cure of improvement or quality of life, and signposts the Cochrane Reviews within which this evidence was synthesised.

Overall completeness and applicability of evidence

Interventions covered by Cochrane Reviews

Despite a relatively large number of Cochrane Reviews synthesising evidence relating to interventions for UI (29) and multiple analyses presenting data relating to the effectiveness of conservative interventions, review evidence relating to conservative interventions for UI was not complete.

The trials included within the reviews included 8975 participants with data for one of our primary outcomes, with a mean of about 80 participants per trial (standard deviation 94; range 14 to 738). Given the number and variety of potential conservative interventions for UI, the number of trials and participants is arguably relatively low, while the number of Cochrane Reviews and analyses are high. The high number of Cochrane Reviews is primarily because of 'splitting' of reviews, with many of the included Cochrane Reviews addressing highly specific and narrow research questions (Weir 2012). For example, four included reviews were focused on PFMT, with each review addressed a very narrow question: PFMT versus no treatment, placebo or sham treatment; PFMT combined with other active treatment; PFMT combined with feedback and different approaches to PFMT (Ayeleke 2015; Dumoulin 2018; Hay‐Smith 2011; Herderschee 2011). There is an argument for first demonstrating the effectiveness of an intervention compared to no treatment or placebo treatment, before determining whether reviews comparing different interventions or approaches to an intervention can be justified. While it can be argued that such narrow reviews will have some advantages, such as greater homogeneity of studies, the disadvantage may be that the evidence is less accessible to clinical decision‐makers when split across multiple reviews (Weir 2012). Conversely, Cochrane Reviews addressing broad research questions, which are not supported by network meta‐analyses that explore differences between interventions, may become large and unwieldy and, therefore, also difficult to access. Furthermore, this pattern of splitting reviews results in some trials being included in more than one review, which can potentially mislead readers (who are unlikely to spend time checking the overlap between reviews) into believing that the volume of evidence is greater than it is. Of the 112 unique trials reporting primary outcome data and summarised within analyses relevant to this overview, 17 were included in two or more reviews.

Despite the high number of Cochrane Reviews included in this overview, there were a number of conservative interventions identified as important by our stakeholder group for which no Cochrane Reviews were identified. Our searching identified 10 published protocols, some of which will address conservative interventions or populations for which there is no existing Cochrane Review, which may fill some of the current gaps in evidence (e.g. Khazali 2016; Ostaszkiewicz 2013; Reynard 2016; see Appendix 6). However, some are likely to overlap with existing reviews, including trials that have already been included within reviews. Therefore, it is unclear whether these new reviews will add clarity to this already complex body of synthesised evidence or will instead reduce the accessibility of information further.

Commonly, the planning and publication of Cochrane Reviews is dependent on the drive and enthusiasm of individual review authors, with editorial oversight that aims to ensure coverage without excessive overlap. Despite concerted efforts to focus reviews on distinct pairwise comparisons, issues of lumping, splitting and overlapping have arisen within the topics covered. Systematic reviews can play a key role in the avoidance of research waste, ensuring that there is full knowledge of what is already known but poorly planned, overlapping reviews can contribute to research waste (Berge 2017; Chalmers 2009; Chalmers 2014; Pollock 2014), and Cochrane has an important role in determining how best to achieve optimal accessibility of evidence with minimal overlap.

With our stakeholder group highlighting conservative interventions for which no Cochrane Review evidence exists, there is a clear need for careful consideration of this body of Cochrane Reviews and the introduction of clear strategies for prioritisation of future reviews. Cochrane has identified the need for robust prioritisation methods and processes (Bero 2012), and the Cochrane Knowledge Translation Priority Setting Working Group has defined a set of standards and materials to support priority setting for Cochrane Reviews (Cochrane Priority Setting Guidance). In order to optimise research efficiency and enhance accessibility of evidence, we believe that there is a need to continue to develop and implement clear, strategic plans to ensure that Cochrane Reviews address what is of greatest importance to the end‐users of the reviews in a manner that is easily accessible. It may be beneficial to use enhanced strategies to ensure meaningful involvement of key stakeholders in all aspects of review prioritisation, planning, conduct and reporting (Pollock 2018).

Outcomes within included reviews

The inclusion of relevant outcomes, which matter to people affected by UI, is of key importance to the content of individual reviews.

As well as ensuring that outcomes within trials and reviews are validated and relevant, it is important that the long‐term effect of interventions is investigated and reported. This overview highlights that less than 20% of the analyses reported in the reviews related to follow‐up outcomes for our primary outcomes of interest. Furthermore, the timing of the outcome assessment was often unclear. Clearly it is essential that future trials are designed to incorporate appropriate long‐term follow‐up in order that the clinical and cost‐effectiveness of interventions can be truly understood. Therefore, we would recommend that researchers make plans to follow up participants longer term, both through data linkage and obtaining consent for contact with the participant over several years. Timing of outcome assessment should be clearly reported within trials and Cochrane Reviews. Health economics data, although often identified as an outcome in the reviews, is only briefly reported within four trials, investigating conservative interventions, and one trial exploring surgical interventions. None attempted to identify and summarise cost and cost‐effectiveness data.

There are limitations within the current evidence base in relation to the outcomes that are reported in the trials and pooled within analyses within the Cochrane Reviews. The challenges in pooling data from the completed trials when they have used and reported different measures of outcomes arguably provides an explanation for the fact that we found that 81% of the relevant analyses only included data from one trial. The inability to pool outcome data from trials investigating the same or similar interventions is a key limitation of the current evidence relating to conservative interventions for UI, and we urge trialists to ensure that outcome measures are validated and reflect outcomes of importance. It is also essential that review authors adhere to Cochrane standards by prestating relevant outcomes and measures, clearly stating how these will be incorporated into meta‐analyses (including the structure of comparisons and subgroups, and decisions on combining of measures assessing similar domains).

Our stakeholder group identified that how women feel about and cope with incontinence is of key importance and that consequently patient‐reported quality of life is a high priority outcome. However, there is currently a lack of evidence relating to this outcome, with predominantly objective measures analysed within current Cochrane Reviews. These objective measures do not necessarily correlate with, or reflect, what is of greatest importance to women with UI. Additionally, the stakeholder group identified that a primary outcome for this overview should be "cure or improvement as reported by the woman" (i.e. participant‐reported cure or improvement). We later identified that Cochrane Reviews commonly report outcomes of both participant‐reported cure or improvement and participant‐reported cure. In order to enable comprehensive synthesis of the available Cochrane Review data, we added 'participant‐reported cure' as an additional secondary outcome. However, it is important to note that our focus on 'cure or improvement' as the primary outcome means that data relating to 'cure' only has been incorporated only as a secondary outcome and is therefore omitted from key tables and figures. These key tables and figures focus only on primary outcomes. It is important to note that for outcomes of both participant‐reported 'cure or improvement' (primary outcome) and 'cure' (secondary outcome), we based our categorisation of data on definitions and descriptions included within the Cochrane Reviews. However, there are a range of different definitions used (both within trials and within reviews) and, consequently, measures reported from different reviews, or from different trials within these reviews, may be based on a variety of different definitions.

Furthermore, this overview highlights that there are multiple, diverse secondary outcomes, which are often poorly defined. Initial plans to extract and summarise data from analyses relating to a range of secondary outcomes were not completed due to the number and diversity of secondary outcome data identified (see Differences between protocol and review). In many cases, trials reported the results of unvalidated, poorly defined questionnaires and the results of these have been inputted into several analyses within Cochrane Reviews. Arguably, the production of multiple analyses from one outcome measure, often with only results from a single trial, can mislead readers in relation to the quantity and quality of evidence, creating barriers to the interpretation of evidence.

Quality of the evidence

Assessment of quality of included reviews

Overview authors independently assessed the quality of included reviews using the ROBIS tool (Whiting 2016). The authors involved in the independent assessment of risk of bias using the ROBIS tool all reported challenges in the application of this tool. There were considerable disagreements in the independent assessments. Consensus was reached through discussion between two independent authors for Domains 1 to 3 but could not be reached for Domain 4. Given these challenges, a third independent author (AE) with statistical expertise was involved. After independent assessment and discussion between the three authors to reach a consensus on 50% of the included reviews, many of the original issues that were causing uncertainty and disagreement had been clarified and the original two authors completed the assessments for the remaining 50%. The difficulties in applying and reaching consensus over the ROBIS assessments were a limitation of this overview. However, in order to be transparent about our decision making, wherever a review was judged to have unclear or high risk of bias, we reported our justification for this decision.

Furthermore, we sought clarification around any particular issues that arose relating to the application of the ROBIS tool. For example, initially there was substantial disagreement between independent overview authors in their assessment of risk of bias relating to the search for studies, as many of the included reviews only searched a single database (the Cochrane Incontinence Specialised Register). Guidance was sought from the experts involved in the development of the ROBIS tool through personal communication as to whether a review that only searched this one register should be judged as having a high or low level of concern about the identification of all relevant studies (Whiting 2015 [pers comm]). This led authors to agree that the use of only the Cochrane Incontinence Specialised Register should be judged as a low level of concern as, despite the use of only one database, the register does involve systematic searching of a wide number of databases using comprehensive search strategies.

The ROBIS tool includes a final judgement of overall risk of bias within Phase 3 of the assessment process. We had originally planned to implement this final phase of the ROBIS tool. However, the lack of agreement between independent authors and challenges in reaching consensus between authors within Phase 2 of the ROBIS tool raised concerns among the overview authors in relation to applying one overall judgement. Given the complex discussion and decision‐making involved in reaching consensus over the Phase 2 domains, overview authors considered that it would be challenging to reach one overall judgement and also that this risked implementing a reductionist approach. Overview authors considered that just one overall judgement may fail to recognise subtle differences in the strengths and limitations between the included reviews and may encourage readers of this overview to use this as a summary 'quality score' (Whiting 2016). Rather than report this final judgement, we therefore opted to report the agreed responses to the three signalling questions, believing that this provides a more transparent and open reporting of our judgements. Where we answered a signalling question as 'probably no' or 'unclear', we provided a reason for this answer. We did not use the responses 'yes' or 'no' in our answers to any of the signalling questions, as we did not consider it to be possible for us to be completely certain in our responses.

There clearly are a number of challenges associated with the use of the ROBIS tool to assess risk of bias of reviews included in an overview and further work and guidance to address the issues that we experienced would be beneficial. However, we believe that our transparent reporting of our judgements using this tool, and by avoidance of a final overall, potentially reductionist judgement that may mask differences between reviews, are strengths of our approach.

Quality of included reviews

We have provided a detailed, transparent assessment of the quality of included reviews using the ROBIS tool (Table 3; Figure 3). Notwithstanding our difficulties with using the ROBIS tool, around half of the included reviews were at high risk of bias in relation to the methods and approach to the method and reporting of the syntheses. The concerns related to both limitations in the information provided within trials and the methods of synthesising trials within the reviews. As discussed earlier, methodological limitations in the conduct or reporting (or both) of trials and reviews relating to conservative interventions for UI present challenges to interpretation and the conclusions that can be drawn. This includes paying particular attention to the outcomes used within trials and synthesised within reviews in order that future research is relevant and meaningful to people affected by UI. It is essential that future research in this field addresses these methodological limitations. It is particularly important that future research has adequate reporting and we urge researchers to adhere to relevant reporting guidelines for trials and reviews (Liberati 2009; Moher 2005).

Assessment of certainty of evidence in included reviews

We planned to use an objective algorithm as, despite uncertainty about the appropriateness of this approach to GRADE, we anticipated having a large quantity of analyses to quality assess and considered the use of a transparent, objective process would be beneficial (Pollock 2016). Building on previous work, we developed an objective algorithm to apply GRADE levels of evidence (Pollock 2014; Pollock 2016). The criteria that we used within the objective algorithm were developed within detailed discussions among the overview authors. The agreed criteria are presented in Appendix 3 and our method of applying these criteria outlined in the methods section. However, we do acknowledge the concerns that have been voiced in relation to this approach and recognise that both this approach and the particular cut‐offs and 'weightings' selected for use within the algorithm require systematic exploration (Gionfriddo 2016; Murad 2016). The application of rigid cut‐offs for objective criteria potentially means that two analyses, which may subjectively appear fairly similar in relation to the certainty of evidence synthesised, could result in differing objectively applied GRADEs.

A complex relationship exists between the criteria contributing to our judgement of certainty of evidence. For example, in an attempt to address concerns relating to the use of sample size as the sole criterion to judge imprecision used by Pollock 2014, our algorithm considered both the CIs and the sample size. If the 95% CIs did not include an RR of 1.0 (or MD of 0) then there was no downgrade (regardless of sample size). However, if the 95% CIs did include an RR of 1.0 (or MD of 0) then there was a downgrade if the sample size was fewer than 344 participants. In this way, fewer downgrades are applied to analyses that demonstrate statistical significance and analyses that do not demonstrate statistical significance will be downgraded at least once (twice if the sample size is fewer than 258 participants). Many of the analyses within this overview did have small sample sizes, meaning that analyses with statistically non‐significant findings were downgraded twice. This system may potentially result in us overemphasising and reporting greater certainty in the statistically significant findings. Further exploration into appropriateness of this is important.

The lack of indepth exploration of the impact of the criteria used within this algorithm is clearly a limitation of this overview. For example, our algorithm will also downgrade to low certainty any analyses where the CI rules out a clinically important difference in favour or one or other of the interventions. This is unlikely to be an issue in this overview but may be in others. However, until further guidance is available relating to the use of GRADE within overviews, we believe that this objective application and determination of GRADE levels of evidence enhance the transparency and consistency of this overview. There is a new 'GRADE in overviews' project group, exploring these issues and aiming to develop relevant guidance, and we await the recommendation of this group with interest.

Potential biases in the overview process

One of the limitations of this overview was that we only included Cochrane Reviews. However, we made this decision based on our knowledge of Cochrane methodology and that only RCTs would be included. We are relatively certain that we were able to identify all Cochrane Reviews relevant to the conservative management of UI. However, many of these are now out‐of‐date and consequently, there could be included trials that have not been incorporated in this evidence base (Alhasso 2005; Cody 2012; Duthie 2011; Eustice 2000; Hay‐Smith 2011; Herbison 2009; Herderschee 2011; Kirchin 2017; Mariappan 2005; Ostaszkiewicz 2004a; Ostaszkiewicz 2004b; Saraswat 2020; Wallace 2004).

However, even with our restriction to Cochrane Reviews, the quality of the reviews included in this overview and the quality of the trials included within these reviews, have varied substantially. Additional sources of potential bias includes the fact that there was overlapping trial evidence between reviews, which we systematically explored and included and deleted as relevant. Our decision to exclude analyses that included men also presents a possible source of bias, as does the decision to exclude analyses were the outcomes synthesised were unclear. The methodological limitations within both the reviews and the studies included within the reviews mean that all evidence within this overview should be interpreted with caution, as several biases may exist.

As identified above, there may be bias within our use of ROBIS and GRADE but we have been transparent about our processes and the use of the objective algorithm for GRADE has made it consistent and reproducible. The use of thresholds within our approach to GRADE may have introduced biases to our assessment of certainty of evidence, as there are important limitations relating to thresholds. For example, our algorithm considered the proportion of studies at risk of bias, rather than the impact of studies with a high risk of bias on the size and direction of effect. Also, our algorithm places emphasis on the statistical significance of results, rather than on a judgement of what is clinically important or relevant. We recommend further research to explore the validity and impact of an algorithmic approach to applying GRADE within overviews.

Agreements and disagreements with other studies or reviews

The main finding of this overview was that PFMT is effective for curing or improving the symptoms of most types of UI, with some evidence that more 'intense' training provides an even better outcome. However, there is a particular need to establish evidence related to 'intensity' or the dose of intervention as this has widespread implications for clinical practice, organisation of services and future research, and indeed, adherence to the intervention by women.

These findings are supported by a systematic review and economic modelling of the effectiveness and cost‐effectiveness of non‐surgical treatments for women with SUI, which was a Health Technology Assessment monograph, which synthesised evidence from a range of sources, including Cochrane Reviews (Imamura 2010). It included a survey of 188 women with SUI to identify outcomes of importance to them (activities of daily living, sex, hygiene and lifestyle issues, emotional health and availability of services), as well as a systematic review considering five generic interventions (PFMT, ES, vaginal cones, bladder training and SNRI). This review identified that for women with SUI, there was evidence that more‐intense PFMT appeared to be the most effective treatment in terms of cure, while basic PFMT was better than no treatment in terms of improvement. We agree with these conclusions, although there is little evidence to support an optimal form of more‐intense PFMT (e.g. in the amount of supervision, individual assessment, type of supervision, use of adjuncts or aids de memoir, frequency of exercise) and current data for most comparisons were sparse.

Interpretation needs to be carefully considered, as confounders are frequent. For example, trials comparing biofeedback to no biofeedback often have widely different levels of contact with the clinician. These findings are also supported by the included Hay‐Smith 2011 review, which reported women receiving regular (e.g. weekly) supervision were more likely to report improvement than women doing PFMT with little or no supervision. A more recent review compared individual supervision, group supervision and home unsupervised PFMT (Paiva 2017). Despite considerable heterogeneity between protocols and low methodological quality, it was reported that PFMT was an efficient technique for the improvement of the symptoms of female UI, both in groups and individually with no significant differences between groups. However, group intervention was more efficient than home PFMT.

Other important issues regarding implications

Women need to be consulted about the type of treatment and setting they would like, including whether supported self‐management is an option. What is not in doubt is that, with increasingly ageing demographics, the number of women with UI is set to rise and there may not be sufficient trained clinicians to deliver some treatments to everyone. More specialists need to be trained, which may be costly, but options such as training generalists or the use of apps (applications for smartphones, etc.) need to be explored. Many of the conservative management options described in this overview involve long‐term changes to daily life activities and adherence to long‐term PFMT. Therefore, it is important that we explore how to maintain women's long‐term adherence to therapy.

ADL: activities of daily living; EMG: electromyography; TENS: transcutaneous electrical nerve stimulation.

Figuras y tablas -
Figure 1

ADL: activities of daily living; EMG: electromyography; TENS: transcutaneous electrical nerve stimulation.

PRISMA study flow diagram. AUI: all types of urinary incontinence; n: number of records; SUI: stress urinary incontinence; UUI: urgency urinary incontinence.
Figuras y tablas -
Figure 2

PRISMA study flow diagram. AUI: all types of urinary incontinence; n: number of records; SUI: stress urinary incontinence; UUI: urgency urinary incontinence.

Summary of ROBIS results for Phase 2, identifying concerns about bias in the review process

Figuras y tablas -
Figure 3

Summary of ROBIS results for Phase 2, identifying concerns about bias in the review process

Table 1. Characteristics of included reviews: design and conduct

Study ID

Objective

Inclusion criteria

Population

Intervention

Planned comparisons

Study design

Databases searched

Alhasso 2005

To determine the effects of adrenergic agonists in the treatment of SUI.

Women with UI

Age range 18–90 years

Adrenergic agonist drug

Adrenergic agonist drug

  • vs placebo or no treatment

  • vs conservative therapies

  • vs surgery

  • higher dose vs a lower dose

  • vs another adrenergic agonist

  • vs alternative forms of pharmacotherapy

  • + another drug vs the other drug treatment alone

  • + another drug vs adrenergic agonist treatment alone

RCT
QRCT

Cochrane Incontinence's Specialised Register – 15 September 2010

Ayeleke 2015

To compare the effects of PFMT plus another active treatment vs the same active treatment alone in the management of women with UI.

Women with SUI, UUI or MUI

Age range 18–75 years

PFMT as a programme of repeated voluntary pelvic floor muscle contractions taught or supervised (or both) by healthcare professionals. All types of PFMT programmes

were considered for inclusion.

A: physical

  • PFMT + vaginal cones vs vaginal cones alone

B: behavioural

  • PFMT + lifestyle intervention (e.g. weight reduction) vs lifestyle intervention alone (lifestyle intervention must be structured or supervised)

  • PFMT + bladder training vs bladder training alone

C: electrical or magnetic

  • PFMT + ES vs ES alone

  • PFMT + magnetic stimulation vs magnetic stimulation alone

D: mechanical

  • PFMT + continence pessaries vs continence pessaries alone

E: drugs

  • PFMT + drug therapy vs drug therapy alone

F: surgery

  • PFMT before surgical intervention vs surgical intervention alone

G: other

  • PFMT + any other stand‐alone active treatment vs the same stand‐alone active treatment

RCT
QRCT

Cochrane Incontinence's Specialised Register – 5 May 2015 (update)t

Embase Classic and Embase 1947 – 7 March 2013 (original)
CINAHL – January 1982 and 6 May 2015 (update)
ClinicalTrials.gov. – 30 May 2013 (original)
WHO ICTRP 3 June 2013 (original)

Cody 2012

To assess the beneficial and harmful effects of oestrogen therapy used for the treatment of UI.

Postmenopausal women with UI and diagnosed as having SUI, UUI or MUI

Age not reported

Oestrogen therapy (different types of oestrogens, different doses and different routes of administration)

Oestrogen therapy

  • vs placebo or no treatment

  • vs other forms of treatment

  • + other therapy vs placebo or no treatment

  • + another treatment vs oestrogen

  • + another treatment vs that other treatment

  • 1 type of oestrogen vs another

  • 1 method of administration of oestrogen vs another

  • high‐dose vs lower dose

Cochrane Incontinence's Specialised Register – 21 June 2012

Dumoulin 2018

To assess the effects of PFMT for women
with UI vs no treatment, placebo or sham treatments, or other inactive control treatments.

All women with UI and diagnosed as having SUI, UUI or MUI

Overall age not addressed in overview

PFMT defined as a programme of repeated voluntary pelvic floor muscle contractions taught and supervised by a healthcare professional. All types of PFMT programmes were considered. Trials in which PFMT was combined with a single episode of biofeedback or advice on strategies for symptoms of urgency or frequency were eligible for inclusion.

Comparisons not listed explicitly. However, the objective was to determine the effects of PFMT for women with UI in comparison to no treatment, placebo

or sham treatments, or other inactive control treatments. In addition, 1 arm of all eligible trials included use of a PFMT programme to ameliorate symptoms of existing urine leakage, 1 received no treatment, 1 placebo, 1 sham treatment and 1 an inactive control treatment

RCT
QRCT

Cochrane Incontinence's Specialised Register – 12 February 2018

Duthie 2011

To compare intravesical botulinum toxin injection with other treatments for neurogenic and idiopathic OAB in adults with or without UI

Men and women diagnosed with idiopathic or neurogenic
OAB syndrome regardless of whether they also had SUI.

Intravesical botulinum toxin

Intravesical botulinum toxin

  • vs placebo

  • vs no treatment

  • vs pharmacological interventions

  • vs non‐pharmacological interventions

  • higher doses vs lower doses

  • + other treatments vs other treatments alone

  • 1 formulation vs another

  • 1 intravesical injection technique vs another

RCT
QRCT

Cochrane Incontinence's Specialised Register – 23 February 2010

Eustice 2000

To determine the effects of prompted voiding for the management
of UI.

Men and women with or without cognitive impairment diagnosed
as having UI

Mean age 84 years

Prompted voiding

Prompted voiding

  • vs no prompted voiding

  • vs other treatments

  • + another treatment vs that other treatment

RCT
QRCT

Cochrane Incontinence's Specialised Register – 31 January 2006

Freites 2019

To assess the effects of laparoscopic colposuspension for UI in women and summarise the principal findings of relevant economic evaluations of these interventions.

Women with SUI or MUI

Laparoscopic colposuspension

Laparoscopic colposuspension

  • vs no treatment or sham operation

  • vs conservative interventions (e.g. PFMT, ES, cones, biofeedback)

  • vs open colposuspension (abdominal surgery)

  • vs traditional sling procedures (abdominal and vaginal surgery)

  • vs mid‐urethral sling procedures (abdominal and vaginal surgery)

  • vs needle suspension (abdominal and vaginal surgery)

  • vs anterior vaginal repair

  • vs periurethral injections

  • 1 method of laparoscopic colposuspension vs another

RCT

QRCT

Cochrane Incontinence's Specialised Register – 22 May 2019

Glazener 2017a

To determine the effects of anterior vaginal repair (anterior colporrhaphy)
on SUI or MUI in women

Women with UI diagnosed as having: urodynamic UI (urodynamic diagnosis), SUI (clinical diagnosis) or MUI (any SUI + other urinary symptoms

Age not reported

Anterior vaginal repair (anterior colporrhaphy)

Anterior vaginal repair

  • vs no treatment or sham operation

  • conservative interventions

  • open abdominal retropubic suspensions for all women

  • vs abdominal surgery for women with co‐existent prolapse

  • bladder neck needle suspensions

  • suburethral sling procedures

  • laparoscopic colposuspensions

  • other methods of anterior vaginal repair

RCT
QRCT

Cochrane Incontinence's Specialised Register – 1 September 2009

Glazener 2017b

To determine the effects of needle suspension on SUI or MUI in comparison with other management options.

Women with SUI MUI

Needle suspension

Needle suspension

  • vs no treatment or sham operation

  • vs conservative interventions (e.g. PFMT, ES, cones, biofeedback)

  • vs open abdominal retropubic suspension (abdominal surgery)

  • vs suburethral sling procedures (abdominal and vaginal surgery)

  • vs anterior vaginal repair (anterior colporrhaphy, vaginal surgery)

  • vs laparoscopic retropubic suspension

  • vs periurethral injections

  • vs drug treatment (e.g. adrenergic agonists)

  • 1 method of needle suspension vs another

  • 1 suture type vs another for carrying out needle suspension operations

RCT

QRCT

Cochrane Incontinence's Specialised Register – 12 November 2014

Additional searches conducted for a brief economic commentary in April 2017

Hay‐Smith 2011

To compare the effects of different approaches to PFMT for women with UI.

Women with UI diagnosed as having SUI, UUI or MUI on the basis of symptoms, signs or urodynamic evaluation, as defined by the study authors.

PFMT defined as any programme of repeated voluntary pelvic floor muscle contractions, or 'indirect' voluntary pelvic floor muscle contraction irrespective of variations in purpose and training parameters

  • More vs less contact with health professionals

  • Group vs individual supervision of PFMT

  • Direct vs indirect methods of PFMT

  • Individualised vs generic PFMT

  • Near maximal vs submaximal contractions

  • Daily vs 3 times per week PFMT

  • Upright and supine vs supine exercise positions alone

  • Strength and motor learning vs motor learning PFMT alone

  • PFMT and abdominal muscle exercise vs PFMT alone

  • PFMT with intravaginal resistance device vs PFMT alone

  • PFMT and adherence strategy vs PFMT alone

  • More‐intensive vs less‐intensive PFMT programmes

RCT
QRCT

Cochrane Incontinence's Specialised Register ‐– 17 May 2011

Herbison 2009

To determine the effects of implantable ES devices in the treatment of urine storage and voiding problems.

People with

  • refractory UUI (non‐neurogenic)

  • OAB syndrome (symptoms of urgency or increased daytime frequency, or both, or nocturia)

  • urinary retention (voiding problem) with functional urethral sphincter

An implanted device to provide stimulation to the sacral nerve(s).

Sacral nerve stimulation

  • + implanted devices vs no treatment

  • + implanted devices vs other treatments

  • 1 method of providing the stimulation (e.g. surgical technique) vs another

RCT
QRCT

Cochrane Incontinence's Specialised Register (10 February 2009)

CENTRAL (March 2008)

MEDLINE (March 2008)

Embase (March 2008)

CINAHL (March 2008)

Herbison 2013

To determine the effectiveness of vaginal cones in the management of female SUI

Women whose predominant complaint was SUI, diagnosed either by symptom classification or urodynamics.

Premenopausal, postmenopausal or

3 months' postpartum

Weighted vaginal cones following a standardised (within trial) protocol.

Vaginal cones

  • vs no treatment

  • vs other conservative therapies

  • vaginal cones + another conservative therapy vs another conservative therapy alone or cones alone

  • vs non‐conservative methods, e.g. surgery or injectables

RCT
QRCT

Cochrane Incontinence's Specialised Register – 19 September 2012

MEDLINE –

February 2013

Embase – February 2013

Herderschee 2011

To determine whether feedback (including biofeedback) provides additional benefit to PFMT in women with UI (SUI, UUI or MUI), regardless of cause.

Women of all ages with SUI, UUI or MUI, diagnosed by symptoms

(reported by the woman), signs (as reported or observed by the healthcare professional) or urodynamics, regardless of cause.

Age: 20–80 years

≥ 1 PFMT arm had to include a form of feedback (or biofeedback) to teach, modulate or encourage pelvic floor muscle contractions

  • PFMT + biofeedback vs PFMT alone

  • PFMT + feedback vs PFMT alone

  • PFMT + feedback + biofeedback vs PFMT alone

  • PFMT + biofeedback vs PFMT + feedback

  • PFMT + 1 type of biofeedback vs PFMT + another type of biofeedback

RCT
QRCT

Cochrane Incontinence's Specialised Register – 13 May 2010

Imamura 2015

To determine the effectiveness of specific lifestyle interventions (i.e. weight loss; dietary changes; fluid intake; reduction in caffeinated, carbonated and alcoholic drinks; avoidance of constipation; stopping smoking; and physical activity) in the management of adult UI.

Adults with UI, diagnosed either by symptom classification (SUI; UUI; MUI) or by urodynamic investigation (urodynamic SUI where not all participants had UI at baseline

Age range (mean) 49–70 years

A community‐based lifestyle intervention following a standardised (within trial) protocol.

Not clearly stated, but the methods stated: "Comparison interventions included no (active) treatment, other conservative physical therapies such as pelvic floor muscle training
(PFMT) or bladder training, or pharmacological therapies."

RCT
QRCT

Cochrane Incontinence's Specialised Register – 3 July 2013

Note: a further update was performed on 27 October 2014 but data entered only into 'studies awaiting classification'.

Kang 2015

To evaluate the efficacy of transurethral radiofrequency collagen denaturation, compared with other interventions, in the treatment of

women with UI.

Women with SUI or MUI diagnosed clinically or
using urodynamics.

Transurethral radiofrequency collagen denaturation

Transurethral radiofrequency collagen denaturation

  • vs no treatment/sham treatment

  • vs conservative physical treatment

  • vs mechanical devices

  • vs drug treatment

  • vs injectable treatment for UI

  • vs other surgery for UI

RCT, QRCT

quote: "excluded cluster‐randomised and cross‐over trials."

Cochrane Incontinence Group Specialised

Register

Embase and Embase Classic

Google scholar

Contacted manufacturers

US FDA website

Current Controlled Trials

ClinicalTrials.gov

WHO International Clinical Trials

Registry Platform

Handsearches of relevant conference abstracts – 19 December 2014

Kirchin 2017

To determine the effects of periurethral and transurethral bulking agents on cure or improvement of UI in women

Women with UI.

Classification of diagnoses was as defined by the study authors.

Age not reported

Urethral injection therapy

Urethral injection therapy

  • vs no treatment

  • vs non‐surgical management

  • vs other surgical managements

  • 1 material for injectable treatment vs another

  • 1 route of injection vs another route

RCT
QRCT

Cochrane Incontinence's Specialised Register – 8 November 2010

MEDLINE – January 1996 to March 2017

Embase – January 1980 to April 2017

NHS EED – April 2017

Lapitan 2017

To determine the effects of open retropubic colposuspension for the treatment of UI in women.

Women with SUI or MUI

Open retropubic colposuspension

Open retropubic colposuspension

  • vs no treatment or sham operation

  • vs conservative interventions

  • vs drug therapy

  • vs other surgical techniques

RCT

QRCT

Cochrane Incontinence's Specialised Register – 5 May 2015

Additional searches conducted for a brief economic commentary in April 2017

Lipp 2014

To determine the effects of mechanical devices in the management of adult female UI, particularly SUI.

Women with UI diagnosed as having SUI, UUI or other incontinence either by symptom classification or by urodynamic diagnosis, as defined by the study authors.

Age not mentioned under participants. Individual trials reported mean or range (44–73 years)

Interventions using mechanical devices designed to control urinary leakage by being inserted: within the vagina; within the urethra or applied to the external surface of the urethra.

A mechanical device

  • vs no treatment

  • vs another mechanical device

  • vs other treatments

RCT
QRCT

Cochrane Incontinence's Specialised Register – 21 August 2014

Embase – 26 August 2014

CINAHL – 26 August 2014

Mariappan 2005

To determine the effects of SNRI in the management of SUI and MUI that includes SUI in adults.

Women and men with UI diagnosed as having SUI (clinical diagnosis), urodynamic SUI (urodynamic diagnosis) or MUI

Age 49–54 years

SNRIs

SNRI

  • vs placebo or 'open' no treatment

  • vs conservative non‐pharmacological therapies

  • vs surgery

  • vs other classes of drugs

  • vs another SNRI

  • SNRI given orally vs given via another route

  • higher dose vs lower dose

  • + another drug vs the other drug alone

  • + conservative non‐pharmacological therapy vs the conservative non‐pharmacological therapy alone

RCT
QRCT

Cochrane Incontinence Specialised Register – 5 March 2007

CENTRAL – the Cochrane Library 2006 – 14 November 2006

MEDLINE – 7 February 2007
MEDLINE In‐Process & Other Non‐Indexed Citations – 7 February 2007

Nambiar 2017

To assess the effectiveness of mini‐sling procedures in women with urodynamic clinical SUI or MUI in terms of improved continence status, quality of life or adverse events.

Women with SUI, or MUI if women had stress‐predominant symptoms

Single‐incision sling

Single‐incision slings

  • vs no treatment

  • vs conservative treatment

  • vs colposuspension

  • vs laparoscopic procedures

  • vs traditional suburethral slings

  • vs retropubic minimally invasive slings

  • vs obturator minimally invasive slings

  • 1 single‐incision sling vs another

RCT

QRT

Cochrane Incontinence's Specialised Register – 6 February 2013

ClinicalTrials.gov and WHO ICTRP – 20 September 2012

Additional searches conducted for a brief economic commentary in April 2017

Ostaszkiewicz 2004a

To assess the effects of habit retraining for the management of

UI in adults

Men and women with or without cognitive impairment diagnosed either by symptom classification or by urodynamic study as having UI.

Women mean age 80.3 years

Men mean age 83 years

Habit retraining

Habit retraining

  • vs no habit retraining

  • vs another intervention

  • + another intervention vs that intervention alone

  • + another intervention vs habit retraining alone

  • + another intervention vs usual care

RCT
QRCT

Cochrane Incontinence Specialised Register – 2 April 2009

MEDLINE – 15 February 2004

Embase – 9 May 2002

CINAHL – March 2001

PsycINFO – August 2002

Current Contents 2001

Ostaszkiewicz 2004b

To assess the effects of timed voiding for the management of UI in adults.

Men and women with or without cognitive impairment diagnosed either by symptom classification or urodynamic study as having UI.

Mean age 86 years

Timed voiding alone or + other interventions.

While acknowledging variability in the definition of timed voiding, this review of timed voiding applied the definitions proposed by Hadley 1986 and by the Agency for Healthcare Research and Quality (Fantl 1991). In this context, timed voiding is distinguished from bladder training, prompted voiding and habit training.

Timed voiding

  • vs no timed voiding

  • vs other interventions

  • + another intervention vs other intervention alone

  • + another intervention vs timed voiding alone

  • + another intervention vs usual care

RCT
QRCT

Cochrane Incontinence Specialised Register – 2 April 2009

MEDLINE – January 1966 to November
2003 (performed 15 November 2003)

Embase – 1980–2002 week 18 (9 May 2002)

PsycINFO – January 1972 to August 2002 (August 2002)

CINAHL – January 1982 to February 2001 (performed in March 2001)

Biological abstracts – January 1980 to December 2000 (performed March 2001)

Current Contents – January 1993–2001 (performed March 2001)

DARE – Issue 1, 2001

Rai 2012

To compare the effects of anticholinergic drugs with various non‐pharmacological therapies for non‐neurogenic OAB syndrome in adults.

Men and women with idiopathic OAB syndrome

or UUI (symptomatic or urodynamic

diagnosis, or both)

Age not provided in review and often not in individual trials

Anticholinergic drug

Anticholinergic drugs

  • vs bladder training alone

  • vs PFMT alone

  • vs external ES

  • vs surgery

  • + non‐drug therapies vs non‐drug therapies alone

  • + non‐drug therapies vs anticholinergic drugs alone

  • vs combination of non‐drug therapies

RCT
QRCT

Cochrane Incontinence Specialised Register – 4 September 2012

Saraswat 2020

To assess the effects of traditional suburethral sling procedures for treatment of urodynamic SUI (urodynamic diagnosis), or for symptoms of SUI or MUI (clinical diagnosis) in women.

Women with SUI due to hypermobility or intrinsic sphincter deficiency (or both, diagnosed clinically or with urodynamics, or with MUI.

Age not provided in review or often in individual trials

Traditional suburethral sling procedures to treat SUI or MUI

Traditional suburethral sling operation

  • vs no treatment or sham operation

  • vs conservative management

  • vs drugs

  • vs injectables

  • vs colposuspension

  • vs bladder neck needle suspension

  • sling operation vs anterior repair

  • vs laparoscopic procedures

  • vs a minimally invasive sling

  • 1 type of traditional sling operations vs another traditional sling operation

RCT
QRCT

Cochrane Incontinence Specialised Register – 27 February 2017

(quote: "A further updated search was conducted on 23 January 2019, the results of which were not fully incorporated into the review.")

Stewart 2016

To determine the effectiveness of ES

Adults aged ≥ 18 years, or according to study authors' definitions of adult) with:

  • symptomatic diagnosis of OAB, UUI or MUI, or

  • urodynamic diagnosis of detrusor overactivity in addition to OAB symptoms (urgency, frequency or episodes of urgency incontinence)

Age reported in review 46–70 years, though 9 trials did not report participants' ages.

Some trials included men and women but majority women.

Neuromodulation

with ES targeting specific

nerves in the sacral plexus that control pelvic floor function. Routes were rectal, vaginal and posterior tibial nerve

ES

  • vs no active treatment, placebo or sham treatment

  • vs conservative treatment (e.g. bladder training, PFMT, biofeedback, magnetic stimulation)

  • vs drugs (e.g. anticholinergics)

  • vs surgery (including botulinum toxin);

  • + another treatment vs other treatment alone

  • + another treatment vs no active treatment, placebo or sham treatment

  • 1 type of ES vs another

RCT

QRCT

Cochrane Incontinence Specialised Register – December 2014

Stewart 2017

To assess the effects of ES with non‐implanted devices, alone or in combination with other treatment, for managing SUI or stress‐predominant MUI in women.

Women with SUI or stress‐predominant MUI on the basis of symptoms, signs or urodynamic diagnosis.

ES with non‐implanted devices

ES

  • vs no active treatment

  • placebo or sham treatment

  • vs other conservative treatment (e.g. bladder training, PFMT, biofeedback, magnetic stimulation)

  • drugs (e.g. duloxetine)

  • vs surgery or injection of bulking agents

  • + another treatment vs the other treatment alone

  • 1 type of ES vs another

RCT

QRCT

Cochrane Incontinence's Specialised Register – 27 February 2017

Additional searches conducted for a brief economic commentary – 10 February 2016

Wallace 2004

To assess the effects of bladder training on UI, however that diagnosis was made

Men and women with UI

Age not reported in review

Range from individual trials 19–68 years

All participants were women except in 1 trial

Bladder training; the terms bladder retraining, bladder drill or bladder re‐education were assumed to be synonymous with bladder training.

Bladder training

  • vs no bladder training

  • vs other treatments

  • + another treatment (such as conservative or pharmacological) vs that other treatment alone

RCT
QRCT

Cochrane Incontinence Specialised Register – 15 March 2006

Wang 2013

To determine the effects and safety of acupuncture for SUI in adults.

People of any age or ethnicity, with a diagnosis of SUI (according to the definition of the International Continence Society) were eligible for inclusion.

Mean age 52 years

All women

Scalp acupuncture, body acupuncture, electroacupuncture, warm acupuncture, elongated needle, auricular acupuncture and fire needle.

Acupuncture

  • vs placebo or no treatment

  • vs any other treatment

RCT
QRCT

Cochrane Incontinence Specialised Register – 28 January 2013

Embase – 20 February 2013

AMED – 20 February 2013

Chinese Biomedical Literature Database – 20 February 2013
Chinese Acupuncture Trials Register – 20 February 2013

China National Knowledge Infrastructure – 20 February 2013

Wieland 2019

To assess the effects of yoga for treating UI in women.

Women aged ≥ 18 years diagnosed with UI on the basis of symptoms, signs or urodynamic evaluation.

Yoga (any yoga tradition but excluded studies in which the yoga intervention did not include a physical practice component).

Yoga

  • vs no specific active intervention (e.g. usual care, waiting list)

  • vs an active intervention (e.g. lifestyle intervention or PFMT) for which we considered different active comparators separately (e.g. yoga vs lifestyle advice, yoga vs PFMT)

  • + an intervention vs the same intervention without yoga (e.g. yoga as an add‐on intervention to PFMT vs PFMT alone)

RCT

Cochrane Incontinence's Specialised Register – 21 June 2018

Cochrane Complementary Medicine Field Specialised Register – 21 June 2018

Additional searches conducted for an economic evaluation – 22 January 2019

ES: electrical stimulation; FDA: Food and Drug Administration; MUI: mixed urinary incontinence; OAB: overactive bladder; PFMT: pelvic floor muscle training; QRCT: quasi‐randomised controlled trial; RCT: randomised controlled trial; SNRI: serotonin and noradrenaline reuptake inhibitors; SUI: stress urinary incontinence; UI: urinary incontinence; UUI: urge urinary incontinence; WHO: World Health Organization.

Figuras y tablas -
Table 1. Characteristics of included reviews: design and conduct
Table 2. Characteristics of included reviews: results

Study ID

Intervention focus

Included studies (number that contained a conservative intervention)

Participant details1, for participants within studies containing a conservative intervention

n

Age (years)

Sex

Pregnancy as an inclusion or exclusion criteria?

Type of UI

Alhasso 2005

To determine the effects of adrenergic agonists in the treatment of SUI

22 (2)

218

Range 30–90

100% F

No

SUI

Ayeleke 2015

To compare the effects of pelvic floor muscle training + another active treatment vs the same active treatment alone in the management of women with UI

13 (13)

1164

Range 18–75

100% F

No

SUI

MUI

UUI

Cody 2012

To assess the effects (both beneficial and harmful) of oestrogen therapy used for the treatment of UI

34 (3)

192

Not reported

100% F

Not reported

AUI

Dumoulin 2018

To determine the effects of PFMT for women with UI vs no treatment, placebo or sham treatments, or other inactive control treatments

31 (31)

1871

18 trials restricted participation based on ages. Lowest age limit 13 years. No upper age limit in several trials.

100% F

No

SUI 18 trials

MUI 1 trial

UUI 1 trial

AUI 11 trials

Duthie 2011

To compare intravesical botulinum toxin injection with other treatments for neurogenic and idiopathic OAB in adults with or without incontinence

19 (0)

0

Eustice 2000

To determine the effects of prompted voiding for the management of UI

9 (9)

674

Mean age 84

Range 69.3–92.4

100% female

No

Unclear

Freites 2019

To assess the effects of laparoscopic colposuspension for UI in women; and summarise the principal findings of relevant economic evaluations of these interventions

26 (0)

0

Glazener 2017a

To determine the effects of anterior vaginal repair (anterior colporrhaphy) on SUI or mixed UI in women

10 (1)

50

Mean 48

Range 31–66

100% female

Not reported

SUI

Glazener 2017b

To determine the effects of needle suspension on SUI or MUI in comparison with other management options

10 (0)

0

Hay‐Smith 2011

To compare the effects of different approaches to pelvic floor muscle training for women with UI

21 (1 trial with 3 arms)(21)

1490

Mean 35.9–58.4 (SD 10.8)

Range 36–80 (possibly)

100% female

No

SUI

UUI

MUI

Herbison 2009

To determine the effects of implantable ES devices in the treatment of urine storage and voiding problems

8 (0)

Herbison 2013

To determine the effectiveness of vaginal cones in the management of female SUI

23 studies

7 abstracts only, 1 ongoing (23)

1806

Not stated

100% female

No

SUI 22

Unclear 1

Herderschee 2011

To determine whether feedback (including biofeedback) provides additional benefit to PFMT in women with UI (SUI, UUI or MUI), regardless of cause

24 (24)

1583

All ages

Assume 100% female

No

SUI

MUI

UUI

Imamura 2015

To determine the effectiveness of specific lifestyle interventions (i.e. weight loss; dietary changes; fluid intake; reduction in caffeinated, carbonated and alcoholic drinks; avoidance of constipation; stopping smoking; and physical activity) in the management of adult UI

11 (11)

5974

49–58

(quotes: "unclear if this was a mean or median"), "except for two trials with means of 62.6 and 70.25 years"

100% female

Not an exclusion criteria

SUI

MUI

UUI

Kang 2015

To evaluate the efficacy of transurethral radiofrequency collagen denaturation, compared with other interventions, in the treatment of women with UI

1 (0)

Kirchin 2017

To determine the effects of periurethral and transurethral bulking agents on cure or improvement of UI in women.

35 (1)

47

Not reported

100% female

Not reported

SUI

Lapitan 2017

To determine the effects of open retropubic colposuspension for the treatment of UI in women

55 (2)

120

31–66 years in 1 study; not reported in 1 study

100% female

Not reported

SUI

Lipp 2014

To determine the effects of mechanical devices in the management of adult female UI, particularly SUI

8 (8)

787

Range 30–75

100% female

No

SUI

UUI

Other

Mariappan 2005

To determine the effects of SNRI in the management of SUI and MUI that includes SUI in adults

10 (1)

201 (possibly)

Range 18–75

100% female

(note: selection criteria included both sexes but data only available for female)

Exclusion

MUI

SUI

Nambiar 2017

To assess the effectiveness of mini‐sling procedures in women with urodynamic clinical stress or MUI in terms of improved continence status, quality of life or adverse events.

31 (0)

0

Ostaszkiewicz 2004a

To assess the effects of habit retraining for the management of UI in adults

4 (4)

378

Mean 80

Analysis on females

No

UUI

MUI

Unclear

Ostaszkiewicz 2004b

To assess the effects of timed voiding for the management of urinary incontinence in adults

2 (2)

298

Mean 86.7

100% female

No

UUI

MUI (possibly)

Rai 2012

To compare the effects of anticholinergic drugs with various non‐pharmacological therapies for non‐neurogenic OAB syndrome in adults

23 (23)

3695

6 had male and female 17 had only female

Not reported

Not reported

UUI

Saraswat 2020

To assess the effects of traditional suburethral sling procedures for treatment of urodynamic SUI

34 (0)

0

Stewart 2016

To determine the effectiveness of ES with non‐implanted electrodes in comparison with placebo or any other active treatment in adults with OAB, with or without UUI

51 (51)

3443

All trials included adults only. 1 trial included only participants aged > 65 years; mean age reported in this trial was 84 years. 9 trials did not report participants' age. Across the remaining trials, mean age ranged from 46 to 70

36/51 female only

11/51 male and female

1/51 male only

3/51 not reported

Not an exclusion criteria

OAB

UUI

MUI

Stewart 2017

To assess the effects of ES with non‐implanted devices, alone or in combination with other treatment, for managing SUI or stress‐predominant MUI in women

56 (56)

3781

Mean range 41–69

100% female

Not reported

SUI

MUI

AUI

Wallace 2004

To assess the effects of bladder training on UI, however that diagnosis was made

12

1473

5 trials no age restriction

Remaining trials 35–65

11/12 female

No

SUI

UUI

MUI

Wang 2013

Determine the effects and safety of acupuncture for SUI in adults

1

60

Mean 53.8

100% female

Not reported

SUI

Wieland 2019

To assess the effects of yoga for treating UI in women.

2 (2)

49

Range 22–79

100% female

1 study – pregnancy in last year was an exclusion criterion.

1 study – most were postmenopausal

UUI

AUI

AUI: all types of urinary incontinence; ES: electrical stimulation; F: female; M: male; MUI: mixed urinary incontinence; n: number of participants; OAB: overactive bladder; PFMT: pelvic floor muscle training; QRCT: quasi‐randomised controlled trial; RCT: randomised controlled trial; SNRI: serotonin and noradrenaline reuptake inhibitors; SUI: stress urinary incontinence; UUI: urgency urinary incontinence.
Number of participants are as reported in review text by review authors.1

Figuras y tablas -
Table 2. Characteristics of included reviews: results
Table 3. Risk of bias

Reviews

ROBIS Phase 2

ROBIS Phase 3 signalling questions

1. Study

eligibility

criteria

2. Identification and selection of

studies

3. Data collection and study

appraisal

4. Synthesis and findings

Did the interpretation of findings address all the concerns identified in Domains 1 to 4

Was the relevance of identified studies to the review's research question appropriately considered?

Did the reviewers avoid emphasising results on the basis of their statistical significance?

Comments relating to Phase 3 signalling questions

Alhasso 2005

L

L

?4

L

PY

PY

PY

Ayeleke 2015

L

L

L

H9

PY

PY

PY

Cody 2012

L

L

?5

H10

PN

PY

PY

The concerns relating to domain 4 were not covered in the discussion.

Dumoulin 2018

L

L

L

H11

PY

PY

PY

Duthie 2011

L

L

L

H12

PN

PY

PY

The concerns relating to study variation have not been adequately discussed.

Eustice 2000

L

?3

H6

H13

PN

PY

PY

The issues relating to the search (Domain 3) and synthesis (Domain 4) were not addressed in the discussion.

Freites 2019

L

L

L

L

PY

PY

PY

Glazener 2017a

L

L

L

H14

PN

PY

PY

Domain 4 concerns were not adequately addressed in the discussion.

Glazener 2017b

L

L

L

H15

PY

PY

PY

Hay‐Smith 2011

L

L

L

H16

PY

PY

PY

Herbison 2009

L

L

L

L

PY

PY

PN

The discussion somewhat overemphasises the findings of the review. While authors noted the difficulties in study reporting and the possibility of duplicate publications, the implications of this are not conveyed within the discussion.

Herbison 2013

L

L

L

H17

PY

PY

PY

Herderschee 2011

L

L

L

H18

PN

PY

PY

Domain 4 concerns were not adequately addressed in the discussion

Imamura 2015

H1

L

?7

L

PN

PY

PY

The concerns relating the data‐driven nature of the review, and the change of criteria noted in Domain 1 were not adequately addressed.

Kang 2015

L

L

L

L

PY

PY

PY

Kirchin 2017

L

L

L

H19

PN

PY

PY

Domain 4 concerns were not addressed in the discussion.

Lapitan 2017

L

L

L

L

PY

PY

PY

Lipp 2014

L

L

L

H20

PY

PY

PY

Mariappan 2005

L

L

L

H21

PY

PY

PY

Nambiar 2017

L

L

L

L

PY

PY

PY

Ostaszkiewicz 2004a

L

L

?8

L

PY

PY

PY

Ostaszkiewicz 2004b

H2

L

?8

L

PY

PY

PY

Rai 2012

L

L

L

H22

PY

PY

PY

Saraswat 2020

L

L

L

L

PY

PY

PY

Stewart 2016

L

L

L

L

PY

PY

PY

Stewart 2017

L

L

L

L

PY

PY

PY

Wallace 2004

L

L

L

?23

PY

PY

PY

Wang 2013

L

L

L

L

PY

PY

PY

Wieland 2019

L

L

L

L

PY

PY

PY

TOTALS

L = 27

H = 2

? = 0

L = 28

H = 0

? = 1

L = 23

H = 1

? = 5

L = 14

H = 14

? = 1

PY = 22

PN = 7

PY = 29

PN = 0

PY = 28

PN = 1

Abbreviations:?: unclear risk of bias; H: high risk of bias; L: low risk of bias; PFMT: pelvic floor muscle training; PN: probably no; PY: probably yes.
1The review authors reported changes to inclusion criteria that were made after the review had started, made because only a low number of studies met the original inclusion criteria.
2Authors identified that the eligibility criteria were difficult to apply: quote: "The review was based on an a priori definition of timed voiding which was developed after an extensive review of the literature; nevertheless we struggled to categorize the interventions used from the descriptions given. Overlap between interventions was found with some interventions having features of more than one approach."
3The process of selection of studies was not fully described.
4There was a lack of clarity about the risk of bias assessment; it was unclear what criteria were used, and only one was reported.
5There was a lack of clarity about the included studies, with poorer details of non‐pharmacological interventions included as comparators.
6Concerns related to the risk of bias assessment – there was a lack of detail, and it appeared to be very limited in nature.
7Only two of the 11 included studies contributed to the analysis. The review authors reported that they tried to obtain the study results but these were not provided for some studies.
8There was a lack of information relating to identification and inclusion of studies (i.e. results of the search), and how risk of bias was assessed and by whom (what criteria were assessed).
9The authors identified trials at high risk of bias, but did not integrate this information into the narrative text describing and supporting the results.
10It appeared that relevant studies had been excluded from synthesis inappropriately – excluded studies appeared to include studies that would have been relevant but which data were unavailable, or outcomes not relevant. There was also a lack of information about how heterogeneity was explored, and no reported sensitivity analyses to explore risk of bias within trials.
11No sensitivity analyses were conducted in this synthesis. GRADE levels did not seem to have been downgraded in response to small sample size; and often only downgraded one level despite more than one identified risk.
12Between‐study variation and study risk of bias were not considered within the findings.
13The analysis that was conducted was correct, but not complete: it failed to address a number of important issues. It contained very little assessment of heterogeneity or risk of bias, and no attempts to ensure robust analyses.
14Between‐study variation and study risk of bias did not appear to have been considered within the findings.
15No GRADE judgements of certainty of evidence, or systematic exploration of study variation.
16The analysis did not fully explore heterogeneity, neither did it ensure the robustness of findings through funnel plots, etc. There also appeared to be some duplication in the subgroup analyses.
17There was high risk of bias due to lack of blinding and high attrition, but these were not addressed. Although authors stated it was not appropriate to perform sensitivity analyses as there were low numbers of included studies, there were analyses containing five trials that were arguably suitable.
18Heterogeneity and study quality did not appear to have been explored systematically within the synthesis and findings.
19The synthesis process did not address the different trial populations, and no sensitivity analyses were performed.
20There were three included studies with high attrition (Cornu 2012; Nielson 1995; Robinson 2003). This appeared to have been dealt with inconsistently and incompletely, with only data from Nielson 1995 being excluded from part of the synthesis. There was no explanation as to why sensitivity analyses to address risk of bias and heterogeneity were not performed.
21Although sensitivity analyses were described within the review methods, these were not conducted, and no rationale was provided.
22Although the grouping of data within the analysis did function as a sensitivity analysis, the quality of studies, especially in relation to attrition bias was not fully explored.
23The authors described a rigorous analysis plan; however, this was not conducted, which may have been an overly conservative approach. Clear information describing and justifying this decision was not provided.

Figuras y tablas -
Table 3. Risk of bias
Table 4. Trials included within reviews

Trial ID

Number of participants in comparison (total)

Number of reviews trial is cited in

Reviews trial is cited in

1st review

2nd review

3rd review

Amaro 2006

40

1

Stewart 2016

Aaronson 1995

38

2

Rai 2012

Stewart 2017

Arruda 2008

77

1

Stewart 2016

Arvonen 2001

37

1

Herbison 2013

Arvonen 2002

17

1

Herbison 2013

Asklund 2017

123

1

Dumoulin 2018

Baker 2014

20

1

Wieland 2019

Barroso 2004

36

1

Stewart 2016

Berglund 1993

20

1

Herbison 2013

Bertotto 2017

49

1

Dumoulin 2018

Beuttenmuller 2010

50

2

Dumoulin 2018

Stewart 2017

Bezerra 2009

34

1

Ayeleke 2015

Bi 2007

60

1

Wang 2013

Bø 1990

52

1

Hay‐Smith 2011

Bø 1999

122

3

Dumoulin 2018

Herbison 2013

Stewart 2017

Boos 1998

101

1

Lipp 2014

Bourcier 1994

102

1

Stewart 2017

Bridges 1988

69

1

Stewart 2017

Burgio 2002

122

1

Herderschee 2011

Burgio 2008

307

1

Rai 2012

Burgio 2010

58

2

Ayeleke 2015

Rai 2012

Burns 1993

82

1

Herderschee 2011

Cammu 1998

60

1

Herbison 2013

Carneiro 2010

50

1

Dumoulin 2018

Castro 2008

50

3

Dumoulin 2018

Herbison 2013

Stewart 2017

Celiker Tosun 2015

121

1

Dumoulin 2018

Colombo 1995

75

2

Rai 2012

Wallace 2004

Cornu 2012

55

1

Lipp 2014

Correia 2013

30

1

Stewart 2017

Correia 2014

48

1

Stewart 2017

de Oliveira 2009

60

1

Hay‐Smith 2011

Delgado 2009

40

1

Hay‐Smith 2011

Diniz Zanetti 2007

44

1

Hay‐Smith 2011

Diokno 2010

41

1

Dumoulin 2018

Eyjólfsdóttir 2009

24

1

Stewart 2017

Fantl 1991

123

1

Wallace 2004

Felicíssimo 2010

59

1

Hay‐Smith 2011

Finazzi‐Agrò 2010

35

1

Stewart 2016

Firra 2013

63

1

Stewart 2016

Franzén 2010

72

1

Stewart 2016

Goode 2002

125

2

Dumoulin 2018

Rai 2012

Goode 2003

87

2

Herderschee 2011

Stewart 2017

Hahn 1991

20

1

Stewart 2017

Haig 1995

58

1

Stewart 2017

Hay‐Smith 2002

123

1

Hay‐Smith 2011

Henalla 1989

51

1

Stewart 2017

Herbison 2004

34

1

Wallace 2004

Hofbauer 1990

21

2

Ayeleke 2015

Stewart 2017

Huang 2014

18

1

Wieland 2019

Hung 2010

64

1

Hay‐Smith 2011

Ishiko 2000

37

1

Alhasso 2005

Johnson 2000

20

1

Herderschee 2011

Kargar Jahromi 2015

48

1

Dumoulin 2018

Kaya 2011

46

1

Stewart 2016

Kaya 2014

108

1

Ayeleke 2015

Klarskov 1986

52

2

Glazener 2017a

Lapitan 2017

Knight 1998

70

1

Stewart 2017

Konstantinidou 2007

22

1

Hay‐Smith 2011

Kosilov 2013

229

1

Stewart 2016

Lagro‐Janssen 1991

18

1

Wallace 2004

Lagro‐Janssen 1992

66

1

Dumoulin 2018

Laycock 1988

36

1

Stewart 2017

Laycock 1993

40

1

Stewart 2017

Leong 2015

55

1

Dumoulin 2018

Lopès 2014

163

1

Stewart 2017

Luber 1997

54

1

Stewart 2017

Macaulay 1987

29

1

Rai 2012

McLean 2013

35

1

Dumoulin 2018

Milani 1986

75

1

Rai 2012

Millard 2004

475

1

Rai 2012

Morkved 2003

94

1

Herderschee 2011

Nascimento‐Correia 2012

45

1

Dumoulin 2018

Oláh 1990

54

2

Herbison 2013

Stewart 2017

Oldham 2013

124

2

Stewart 2016

Stewart 2017

Pages 2001

40

1

Herderschee 2011

Park 2002

50

1

Rai 2012

Patil 2010

110

1

Stewart 2017

Peattie 1988

33

1

Herbison 2013

Pereira 2012

14

1

Stewart 2017

Phelan 2012

738

1

Imamura 2015

Pieber 1995

46

1

Herbison 2013

Preyer 2007

31

1

Stewart 2016

Ramsay 1990

44

1

Hay‐Smith 2011

Richter 2007

299

2

Ayeleke 2015

Lipp 2014

Sancaktar 2010

40

1

Stewart 2016

Santos 2009

45

2

Herbison 2013

Stewart 2017

Sar 2009

34

1

Dumoulin 2018

Schagen van Leeuwen 2004

101

3

Ayeleke 2015

Hay‐Smith 2011

Mariappan 2005

Schmidt 2009

32

1

Stewart 2017

Schreiner 2010

52

1

Stewart 2016

Seo 2004

120

2

Herbison 2013

Stewart 2017

Solberg 2016

20

1

Dumoulin 2018

Song 2006

58

1

Rai 2012

Souto 2014

75

1

Stewart 2016

Sran 2016

48

1

Dumoulin 2018

Sriboonreung 2011

40

1

Hay‐Smith 2011

Subak 2005

40

1

Imamura 2015

Sugaya 2003

41

1

Hay‐Smith 2011

Tapp 1989

68

1

Stewart 2017

ter Meulen 2009

45

1

Kirchin 2017

Terlikowski 2013

102

1

Stewart 2017

Vecchioli‐Scaldazza 2013

40

1

Stewart 2016

Wang 2004

120

2

Stewart 2016

Herderschee 2011

Wang 2006

47

1

Rai 2012

Wells 1991

157

1

Alhasso 2005

Wells 1999

286

1

Hay‐Smith 2011

Whitmore 1995

52

1

Stewart 2017

Williams 2006

158

1

Herbison 2013

Wing 2010

304

1

Imamura 2015

Wilson 1987

30

2

Herderschee 2011

Stewart 2017

Wyman 1998

128

2

Ayeleke 2015

Wallace 2004

Yoon 2003

25

1

Wallace 2004

Total = 112 trials

Total participants = 8975

Mean participants per study = 80

Range 14–738, standard deviation 94

Trials in 1 review = 93

Trials in 2 reviews = 16

Trials in 3 reviews = 3

Figuras y tablas -
Table 4. Trials included within reviews
Table 5. Stress urinary incontinence: quantity and certainty of evidence

Comparison

Outcome

Certainty of the evidence

Not estimable

Total analyses

Table

High

Moderate

Low

Very low

Conservative vs

control

Cure or

improvement

2

1

2

0

0

5

Table 8

Quality of life

1

5

6

0

0

12

Table 9

Conservative vs

conservative

Cure or

improvement

3

22

15

1

2

43

Table 10

Quality of life

0

6

4

0

1

11

Table 11

Conservative vs

non‐conservative

Cure or

improvement

0

6

3

0

0

9

Table 12

Quality of life

1

2

0

0

1

4

Table 13

TOTALS

7

42

30

1

4

84

Figuras y tablas -
Table 5. Stress urinary incontinence: quantity and certainty of evidence
Table 6. Urgency urinary incontinence: quantity and certainty of evidence

Comparison

Outcome

Certainty of the evidence

Not estimable

Total analyses

Table

High

Moderate

Low

Very low

Conservative vs

control

Cure or

improvement

2

4

1

0

0

7

Table 14

Quality of life

0

1

0

0

0

1

Table 15

Conservative vs

conservative

Cure or

improvement

0

7

3

0

0

10

Table 16

Quality of life

1

2

1

0

0

4

Table 17

Conservative vs

non‐conservative

Cure or

improvement

0

9

13

0

0

22

Table 18

Quality of life

1

2

0

0

0

3

Table 19

TOTALS

4

25

18

0

0

47

Figuras y tablas -
Table 6. Urgency urinary incontinence: quantity and certainty of evidence
Table 7. All types of urinary incontinence: quantity and certainty of evidence

Comparison

Outcome

Certainty of the evidence

Not estimable

Total analyses

Table

High

Moderate

Low

Very low

Conservative vs

control

Cure or

improvement

3

4

3

1

0

11

Table 20

Quality of life

2

2

0

1

0

5

Table 21

Conservative vs

conservative

Cure or

improvement

1

3

14

6

0

24

Table 22

Quality of life

0

3

12

1

0

16

Table 23

Conservative vs

non‐conservative

Cure or

improvement

0

1

2

0

0

3

Table 24

Quality of life

0

1

1

0

0

2

Table 25

TOTALS

6

14

32

9

0

61

Figuras y tablas -
Table 7. All types of urinary incontinence: quantity and certainty of evidence
Table 8. Stress urinary incontinence: primary outcome participant‐reported cure or improvement; conservative versus control

Analysis code (in review)

Study ID

Comparisons

included

Timing of measure

Included trials

(n)

Effect

measure

Effect estimate

(CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades

(reason for downgrade)

High‐certainty evidence

1.1.1

Herbison 2013

Cones vs control

IM

Bø 19991

(n = 57)

RR

0.4 (0.2 to 0.6)

> 1 favours control

YES for cones

0

1.2.2

Herderschee 2011

PFMT + BF vs

PFMT alone (control)

IM

Goode 2003

(n = 87)

RR

0.6 (0.4 to 0.9)

> 1 favours PFMT alone

YES for PFMT and BF

0

Moderate‐certainty evidence

1.2.1

Dumoulin 2018

PFMT vs no treatment, placebo or control

IM

Asklund 2017; Bø 19991; Lagro‐Janssen 1992

(n = 242)

RR

6.3

(3.9 to 10.3)

> 1 favours PFMT

YES for PFMT

1(H)

Low‐certainty evidence

3.2.1

Hay‐Smith 2011

Direct vs indirect methods of PFMT

IM

Schagen van Leeuwen 2004; Ramsay 1990

(n = 138)

RR

0.7

(0.47 to 1.0)

> 1 favours indirect

NO

6(E,F,H×2)

2.2

Stewart 2017

ES vs sham treatment

IM

Hofbauer 1990Laycock 1993; Luber 1997; Terlikowski 2013; Whitmore 1995 (n = 236)

RR

2.03 (1.02 to 4.07)

> 1 favours ES

YES for ES

GRADE as assessed by review authors

1Bø 1999 was a four‐arm randomised controlled trial in which the groups were: control, PFMT, cones and ES. Data from PFMT and control arms were included in Dumoulin 2018, and data from cones and control arms were included in Herbison 2013.
Abbreviations: BF: biofeedback; CI: confidence interval; ES: electrical stimulation; IM: immediate; n: number of participants assessed NOT to have been cured or improved using a variety of different assessment methods; PFMT: pelvic floor muscle training; RR risk ratio.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Figuras y tablas -
Table 8. Stress urinary incontinence: primary outcome participant‐reported cure or improvement; conservative versus control
Table 9. Stress urinary incontinence: quality of life, conservative versus control

Analysis code (in review)

Study ID

Comparisons included

Timing of measure

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

High‐certainty evidence

1.12.1

Dumoulin 2018

PFMT vs no treatment, placebo or control (I‐QoL)

IM

Castro 2008

(n = 50)

MD

−24.6 (−37.7 to −11.4)

> 0 favours control

YES favours PFMT

0

Moderate‐certainty evidence

1.5.1

Dumoulin 2018

PFMT vs no treatment, placebo or control

IM

Beuttenmuller 2010; Carneiro 2010; Nascimento‐Correia 20122

(n = 145)

MD

−11.89 (−20.5 to −3.2)

> 0 favours control

YES favours PFMT

2(H×2)

1.5

Lipp 2014

Intravaginal mechanical device vs no treatment

UC

Cornu 20123

(n = 55)

MD

−2.2 (−3.5 to −0.9)

> 0 favours no treatment

YES favours device

1(C)

1.6

Lipp 2014

Intravaginal mechanical device vs no treatment

UC

Cornu 20123

(n = 55)

MD

−1.67 (−2.8 to −0.5)

> 0 favours no treatment

YES favours device

1(C)

1.7

Lipp 2014

Intravaginal mechanical device vs no treatment

UC

Cornu 20123

(n = 55)

MD

−0.5 (−0.9 to −0.1)

> 0 favours no treatment

YES favours device

1(C)

1.3

Stewart 2017

ES vs no active treatment

IM

Correia 2013; Correia 2014; Lopès 2014; Pereira 2012 (n = 250)

SMD

−0.72 (−0.99 to −0.45)

< 0 favours ES

YES for ES

GRADE as assessed by review authors

Low‐certainty evidence

1.3.1

Dumoulin 2018

PFMT vs no treatment, placebo or control

IM

Beuttenmuller 2010; Carneiro 2010; Nascimento‐Correia 20122

(n = 145)

MD

−13.1 (−21.1 to −5.2)

> 0 favours control

YES favours PFMT

4(F×2, H×2)

1.4.1

Dumoulin 2018

PFMT vs no treatment, placebo or control

IM

Beuttenmuller 2010; Carneiro 2010; Nascimento‐Correia 20122

(n = 145)

MD

−11.8 (−20.8 to −2.7)

> 0 favours control

YES favours PFMT

4(F×2, H×2)

1.8.1

Dumoulin 2018

PFMT vs no treatment, placebo or control

IM

Beuttenmuller 2010; Carneiro 2010; Nascimento‐Correia 20122

(n = 145)

MD

1.8 (−3.4 to 7.0)

> 0 favours control

NO

4(E×2, H×2)

1.6.1

Dumoulin 2018

PFMT vs no treatment, placebo or control

IM

Bø 19991

(n = 55)

RR

0.8 (0.6 to 1.1)

> 1 favours PFMT

NO

3(C, E×2)

1.10.1

Dumoulin 2018

PFMT vs no treatment, placebo or control

IM

McLean 2013 (n = 35)

MD

−19.70 (−30.63 to −8.77)

> 0 favours control

YES favours PFMT

3 (E×2, H)

1.8

Lipp 2014

Intravaginal mechanical device vs no treatment

UC

Cornu 20123

(n = 46)

MD

−10.3 (−20.77 to 0.17)

> 0 favours no treatment

NO

3(C, E×2)

1Bø 1999 used the Bristol Female Lower Urinary Tract Symptoms questionnaire and this was reported as event data within the review (Dumoulin 2018).
2Beuttenmuller 2010; Carneiro 2010; Nascimento‐Correia 2012 – these three trials each present individual subscales for Kings Health Questionnaire (subscales of severity, impact, physical limitation and general health) and these pooled totals are presented in the tables.
3Cornu 2012 – the time point at which outcomes were measured was unclear.
Abbreviations: CI: confidence interval; ES: electrical stimulation; I‐QoL: Incontinence Quality of Life; IM: immediate; MD: mean difference; n: number of participants; n: number of participants; PFMT: pelvic floor muscle training; RR: risk ratio; SMD: standardised mean difference; UC: unclear.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Figuras y tablas -
Table 9. Stress urinary incontinence: quality of life, conservative versus control
Table 10. Stress urinary incontinence: cure or improved, conservative versus conservative

Analysis code (in review)

Study ID

Comparisons included

Timing of measure

Included trials

(n)

Effect

measure

Effect estimate

(CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades

(reason for downgrade)

High‐certainty evidence

2.1.1

Herbison 2013

Cones vs PFMT

IM

Bø 19991

(n = 52)

RR

4.6 (1.1 to 19.1)

> 1 favours PFMT

YES favours PFMT

0

8.1.1

Lipp 2014

Intravaginal pessary alone vs pessary + PFMT

IM

Richter 20071

(n = 299)

RR

0.7

(0.6 to 0.9)

> 1 favours pessary alone

YES favours PFMT + pessary

0

7.1.2

Lipp 2014

Intravaginal pessary vs behavioural therapy (PFMT) alone

IM

Richter 20071

(n = 295)

RR

0.7

(0.5 to 0.9)

> 1 favours device

YES favours PFMT

0

Moderate‐certainty evidence

6.1.2

Ayeleke 2015

PFMT + continence pessary vs continence pessary alone

FU1

Richter 2007

(n = 207)

RR

0.9 (0.7 to 1.1)

> 1 favours PFMT

NO

2(E×2)

6.3.3

Ayeleke 2015

PFMT + continence pessary vs continence pessary alone

FU2

Richter 2007

RR

0.9 (0.7 to 1.2)

> 1 favours PFMT

NO

2(E×2)

11.1.1

Hay‐Smith 2011

PFMT and adherence strategy vs PFMT alone

IM

Sugaya 20032

(n = 41)

RR

0.6 (0.38 to 0.9)

> 1 favours PFMT + adherence

YES favours PFMT + adherence

1(B)

12.2.2

Hay‐Smith 2011

More‐intensive vs less‐intensive PFMT programmes

IM

Diniz Zanetti 20072

(n = 44)

RR

0.3 (0.2 to 0.7)

> 1 favours more‐intense PFMT

YES favours more‐intense PFMT

1(H)

8.2.2

Hay‐Smith 2011

Strength and motor learning vs motor learning PFMT alone

IM

Hay‐Smith 20022

(n = 123)

RR

0.6 (0.3 to 1.4)

> 1 favours less intensive

NO

2(E×2)

2.1.1

Herbison 2013

Cones vs PFMT

IM

Cammu 19983

(n = 60)

RR

0.9 (0.5 to 1.6)

> 1 favours PFMT

NO

2(E×2)

7.1.1

Lipp 2014

Intravaginal pessary vs behavioural therapy (PFMT

IM

Richter 20071

(n = 295)

RR

0.8 (0.6 to 1.0)

> 1 favours device

NO

1(E)

7.1.3

Lipp 2014

Intravaginal pessary vs behavioural therapy (PFMT

FU1

Richter 2007

(n = 295)

RR

1.0 (0.8 to 1.3)

> 1 favours device

NO

1(E)

7.1.4

Lipp 2014

Intravaginal pessary vs behavioural therapy (PFMT

FU2

Richter 2007

(n = 295)

RR

0.9 (0.7 to 1.3)

> 1 favours device

NO

1(E)

7.1.6

Lipp 2014

Intravaginal pessary vs behavioural therapy (PFMT

FU2

Richter 2007

(n = 295)

RR

0.9 (0.7 to 1.2)

> 1 favours device

NO

1(E)

8.1.3

Lipp 2014

Intravaginal pessary alone vs pessary + PFMT

FU1

Richter 2007

(n = 299)

RR

0.9 (0.7 to 1.1)

> 1 favours pessary alone

NO

1(E)

8.1.4

Lipp 2014

Intravaginal pessary alone vs pessary + PFMT,

FU2

Richter 2007

(n = 299)

RR

0.9 (0.7 to 1.3)

> 1 favours pessary alone

NO

1(E)

8.1.6

Lipp 2014

Intravaginal pessary alone vs pessary + PFMT,

FU2

Richter 2007

(n = 299)

RR

1.0 (0.7 to 1.3)

> 1 favours pessary alone

NO

1(E)

9.1.1

Lipp 2014

Pessary + PFMT vs behavioural therapy (PFMT) alone

IM

Richter 20071

(n = 296)

RR

1.0 (0.5 to 1.3)

> 1 favours pessary + PFMT

NO

1(E)

9.1.3

Lipp 2014

Pessary + PFMT vs behavioural therapy (PFMT) alone

FU1

Richter 2007

(n = 296)

RR

1.1 (0.91 to 1.4)

> 1 favours pessary + PFMT

NO

1(E)

9.1.4

Lipp 2014

Pessary + PFMT vs behavioural therapy (PFMT) alone

FU2

Richter 2007

(n = 296)

RR

1.0 (0.7 to 1.4)

> 1 favours pessary + PFMT

NO

1(E)

9.1.6

Lipp 2014

Pessary + PFMT vs behavioural therapy (PFMT) alone

FU2

Richter 2007

(n = 296)

RR

0.9 (0.7 to 1.3)

> 1 favours pessary + PFMT

NO

1(E)

9.1.2

Lipp 2014

Pessary + PFMT vs behavioural therapy (PFMT) alone

IM

Richter 20071

(n = 296)

RR

0.9 (0.7 to 1.1)

> 1 favours device

NO

1(E)

7.1.5

Lipp 2014

Intravaginal pessary vs behavioural therapy (PFMT) alone

FU1

Richter 2007

(n = 295)

RR

0.9 (0.6 to 1.1)

> 1 favours pessary alone

NO

1(E)

8.1.2

Lipp 2014

Intravaginal pessary alone vs pessary + PFMT,

IM

Richter 20071

(n = 299)

RR

0.7 (0.5 to 1.0)

> 1 favours pessary alone

NO

1(E)

8.1.5

Lipp 2014

Intravaginal pessary alone vs pessary + PFMT

FU1

Richter 2007

(n = 299)

RR

1.1 (0.8 to 1.5)

> 1 favours pessary + PFMT

NO

1(E)

9.1.5

Lipp 2014

Pessary + PFMT vs behavioural therapy (PFMT) alone

FU1

Richter 2007

(n = 296)

RR

0.8 (0.6 to 1.1)

> 1 favours pessary + PFMT

NO

1(E)

Low‐certainty evidence

4.2

Ayeleke 2015

PFMT + ES vs ES alone (excluding implanted electrodes)

IM

Bezerra 20093; Hofbauer 19903

(n = 56)

RR

2.1 (0.8 to 5.4)

> 1 favours PFMT + pessary

NO

4(E×2, H×2)

6.1.1

Ayeleke 2015

PFMT + continence pessary vs continence pessary alone

IM

Richter 20071

(n = 242)

RR

1.0 (0.8 to 1.2)

> 1 favours PFMT + pessary

NO

3(C, E×2)

6.3.1

Ayeleke 2015

PFMT + continence pessary vs continence pessary alone

IM

Richter 20071

(n = 242)

RR

1.1 (0.9 to 1.4)

> 1 favours PFMT + pessary

NO

3(C, E×2)

6.3.2

Ayeleke 2015

PFMT + continence pessary vs continence pessary alone

FU1

Richter 2007

(n = 225)

RR

1.0 (0.8 to 1.3)

> 1 favours PFMT + pessary

NO

3(C, E×2)

1.2.1

Hay‐Smith 2011

More vs less contact with health professionals

IM

Bø 19902; Diniz Zanetti 20072; Felicíssimo 20102; Konstantinidou 20072

(n = 177)

RR

0.3 (0.14 to 0.5)

> 1 favours more contact

YES favours more contact

3(F×2, H)

2.2.1

Hay‐Smith 2011

Group vs individual supervision of PFMT

IM

Bø 19902; Felicíssimo 20102; Konstantinidou 20072

(n = 133)

RR

0.1 (0.0 to 0.4)

> 1 favours group

YES, favours group

3(F×2, H)

2.2.2

Hay‐Smith 2011

Group vs individual supervision of PFMT

IM

de Oliveira 20092

(n = 60)

RR

1.2 (0.6 to 2.3)

> 1 favours individual

NO

3(E×2, H)

4.1.1

Hay‐Smith 2011

Individualised vs generic PFMT

IM

de Oliveira 20092

(n = 60)

RR

0.8 (0.4 to 1.6)

> 1 favours generic

NO

3(E×2, H)

2.1.2

Herbison 2013

Cones vs PFMT

IM

Arvonen 20013; Arvonen 20023

(n = 54)

RR

0.8 (0.5 to 1.3)

> 1 favours PFMT

NO

3(E×2, H)

4.1.1

Herbison 2013

Cones + PFMT vs PFMT

IM

Pieber 19953

(n = 46)

RR

1.4 (0.8 to 2.4)

> 1 favours PFMT

NO

3(E×2, H)

4.2.1

Herbison 2013

Cones + PFMT vs PFMT

O

Pieber 1995

(n = 46)

RR

0.9 (0.5 to 1.6)

> 1 favours PFMT

NO

3(E×2, H)

5.1

Herbison 2013

Cones + PFMT vs ES

IM

Berglund 19933; Seo 20043 (n = 160)

RR

1.5 (0.8 to 2.6)

> 1 favours ES

NO

3(E×2, H)

3.2

Stewart 2017

ES vs PFMT

IM

Bø 1999; Castro 2008; Hahn 1991; Henalla 1989; Hofbauer 1990; Laycock 1988; Aaronson 1995 (n = 244)

RR

0.85 (0.70 to 1.03)

> 1 favours ES

NO

GRADE as assessed by review authors

4.2

Stewart 2017

ES vs vaginal cones

IM

Bridges 1988; Bø 1999; Castro 2008; Oláh 1990; Seo 2004 (n = 331)

RR

1.09 (0.97 to 1.21)

> 1 favours ES

NO

GRADE as assessed by review authors

6.2

Stewart 2017

ES + PFMT vs PFMT

IM

Eyjólfsdóttir 2009; Goode 2003; Hofbauer 1990; Knight 1998; Knight 1998; Tapp 1989; Wilson 1987 (n = 308)

RR

1.10 (0.95 to 1.28)

> 1 favours ES + PFMT

NO

GRADE as assessed by review authors

Very low‐certainty evidence

5.2

Stewart 2017

ES vs PFMT + vaginal cones

IM

Bourcier 1994; Laycock 1993 (n = 68)

RR

1.53 (1.08 to 2.18)

> 1 favours ES

YES for ES

GRADE as assessed by review authors

Other

6.2.1

Hay‐Smith 2011

Daily vs 3 times per week PFMT

IM

Sriboonreung 20112 (n = 40)

RR

NE

NE

9.2

Hay‐Smith 2011

PFMT and abdominal muscle exercise vs PFMT alone

IM

Sriboonreung 20112 (n = 40)

RR

NE

NE

1Trials comparing the effect of conservative interventions from different intervention categories, with an immediate outcome measure.
2Bø 1999, de Oliveira 2009, Diniz Zanetti 2007, Hay‐Smith 2011, Sriboonreung 2011, Sugaya 2003, Felicíssimo 2010, Konstantinidou 2007 compared different ways of delivering the same intervention.
3Arvonen 2001Arvonen 2002, Bezerra 2009, Cammu 1998, Hofbauer 1990, Pieber 1995, Laycock 1993, and Seo 2004 compared the effect of conservative interventions from within the same intervention categories (physical therapy interventions in all cases).
Abbreviations: CI: confidence interval; ES: electrical stimulation; FU1: follow‐up one; IM: immediate; PFMT: pelvic floor muscle training; RR: risk ratio; n: number of participants assessed NOT to have been cured or improved using a variety of different assessment methods; NE: not estimable; O: other.

Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Figuras y tablas -
Table 10. Stress urinary incontinence: cure or improved, conservative versus conservative
Table 11. Stress urinary incontinence: quality of life, conservative versus conservative

Analysis code (in review)

Study ID

Comparisons included

Timing of measure

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

Moderate‐certainty evidence

6.2.2

Ayeleke 2015

PFMT added to continence pessary vs continence pessary alone,

FU1

Richter 2007

(n = 207)

RR

0.8 (0.6 to 1.1)

> 1 favours PFMT + pessary

NO

2(E×2)

7.4

Lipp 2014

Intravaginal pessary vs behavioural therapy (PFMT)

IM

Richter 2007

(n = 295)

MD

0.7 (−9.5 to 10.9)

> 0 favours PFMT

NO

1(E)

7.5

Lipp 2014

Intravaginal pessary vs behavioural therapy (PFMT)

FU1

Richter 2007

(n = 295)

MD

−3.2 (−11.4 to 5.0)

> 0 favours PFMT

NO

1(E)

7.6

Lipp 2014

Intravaginal pessary vs behavioural therapy (PFMT)

FU1

Richter 2007

(n = 295)

MD

−0.2 (−1.3 to 0.9)

> 0 favours PFMT

NO

1(E)

7.7

Lipp 2014

Intravaginal pessary vs behavioural therapy (PFMT)

FU1

Richter 2007

(n = 295)

MD

0.3 (−0.7 to 1.3)

> 0 favours PFMT

NO

1(E)

2.3

Mariappan 2005

SNRI + conservative non‐pharmacological therapy vs conservative non‐pharmacological therapy alone

IM

Schagen van Leeuwen 20041

(n = 100)

OR

1.3 (0.5 to 2.9)

> 1 favours combination

NO

2(E×2)

Low‐certainty evidence

6.2.1

Ayeleke 2015

PFMT + continence pessary vs continence pessary alone

IM

Richter 20071

(n = 242)

RR

1.1 (0.8 to 1.5)

> 1 favours PFMT +

NO

3(C, E×2)

4.3

Stewart 2017

ES vs vaginal cones

IM

Castro 2008; Santos 2009 (n = 96)

MD

1.59 (−3.72 to 6.90)

> 1 favours ES

NO

GRADE as assessed by review authors

10.12.3

Wallace 2004

Bladder training + non‐pharmacological treatment vs non‐pharmacological treatment alone

IM

Wyman 19983

(n = 89)

MD

−18 (−36.6 to 0.6)

> 0 favours control

NO

4(E×2, G, H)

5.14.3

Wallace 2004

Bladder training vs other behavioural/physical/psychological treatments

IM

Wyman 19983

(n = 92)

MD

18 (−1.4 to 37.3)

> 0 favours control

NO

3(E×2, G)

Other

9.3

Mariappan 2005

SNRI + conservative non‐pharmacological therapy vs conservative non‐pharmacological

IM

Schagen van Leeuwen 2004 (n = 100)

MD

NE

NO

NE

1Data presented as number of events.
2Data for I‐QOL, effect size was not estimable as the standard deviations are provided as 0.
3These Wyman 1998 data comprised a subgroup of the participants from Wyman 1998: Wyman 1998 included participants with combined SUI, UUI and MUI and all data is presented within reviews for the combined group of participants, except for the two subgroups of data here, which related to participants with SUI only. These Wyman 1998 data related to the Urinary Distress Inventory measure of quality of life, for two pairs from this three‐arm trial (the physical therapy treatment (control) group data are the same for both pairs).
Abbreviations: CI: confidence interval; FU1: follow‐up one; IM: immediate; I‐QOL: Incontinence Quality of Life; MD: mean difference; MUI: mixed urinary incontinence; n: number of participants; NE: not estimable; O: other; OR: odds ratio; RR: risk ratio; SNRI: serotonin–noradrenaline reuptake inhibitors; SUI: stress urinary incontinence; UUI: urgency urinary incontinence.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Figuras y tablas -
Table 11. Stress urinary incontinence: quality of life, conservative versus conservative
Table 12. Stress urinary incontinence: primary outcome participant‐reported cure or improvement; conservative versus non‐conservative

Analysis code (in review)

Study ID

Comparisons included

Timing of measure

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

Moderate‐certainty evidence

7.2.1

Ayeleke 2015

PFMT added to drug therapy vs drug therapy alone

IM

Schagen van Leeuwen 20041,2

(n = 90)

RR

1.1 (0.8 to 1.5)

> 1 favours PFMT + drug

NO

2(E×2)

7.4.1

Ayeleke 2015

PFMT added to drug therapy vs drug therapy alone

IM

Schagen van Leeuwen 20041,3

(n = 101)

RR

1.3 (0.9 to 1.8)

> 1 favours acupuncture

NO

2(E×2)

2.1

Alhasso 2005

Adrenergic agonist vs conservative therapy

UC

Ishiko 2000

(n = 34)

RR

1.3 (0.7 to 2.2)

> 1 favours drug

NO

2(E×2)

2.1

Alhasso 2005

Adrenergic agonist vs conservative therapy

UC

Ishiko 2000

(n = 37)

RR

0.9 (0.6 to 1.3)

> 1 favours drug

NO

2(E×2)

1.1

Lapitan 2017

Open retropubic colposuspension vs conservative interventions

FU1

Klarskov 19864 (n = 29)

RR

0.2 (0.1 to 0.7)

< 1 favours surgery

YES favours surgery

1(H)

2.1

Wang 2013

Acupuncture vs any other treatment

IM

Bi 20071

(n = 60)

RR

2.2 (1.3 to 3.8)

> 1 favours acupuncture

YES favours acupuncture

2(A, B)

Low‐certainty evidence

2.3

Glazener 2017a

Anterior repair vs PFMT

FU1

Klarskov 19864

(n = 16)

RR

0.3 (0.0 to 2.3)

> 1 favours PFMT

NO

2(E×2)

2.2

Glazener 2017a

Anterior repair vs PFMT

FU2

Klarskov 19864

(n = 16)

RR

2.6 (0.6 to 10.2)

> 1 favours PFMT

NO

2(E×2)

2.2

Kirchin 2017

Urethral injection therapy vs conservative management

IM

ter Meulen 20091

(n = 45)

RR

0.2 (0.0 to 1.8)

> 1 favours conservative

NO

3(E×2, H)

1Trials with an immediate outcome measure.
2Schagen van Leeuwen 2004 determined cure or improvement using two methods. These data related to data collected via a paper diary.
3Schagen van Leeuwen 2004 determined cure or improvement using two methods. These data are based on the Patient Global Impression scale – Improvement.
4Klarskov 1986 included participants who received different surgical interventions, Lapitan 2017 included all surgical groups, while Glazener 2017a included the subgroup of participants receiving anterior repair – therefore, 16 participants were presented twice.
Abbreviations: CI: confidence interval; FU1: follow‐up one; FU2: follow‐up two; IM: immediate; n: number of participants assessed NOT to have been cured or improved using a variety of different assessment methods; PFMT: pelvic floor muscle training; RR: risk ratio; UC: unclear.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Figuras y tablas -
Table 12. Stress urinary incontinence: primary outcome participant‐reported cure or improvement; conservative versus non‐conservative
Table 13. Stress urinary incontinence: quality of life, conservative versus non‐conservative

Analysis code (in review)

Study ID

Comparisons included

Timing of measure

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

High‐certainty evidence

2.2

Mariappan 2005

SNRI vs conservative non‐ pharmacological therapies

IM

Schagen van Leeuwen 20042

(n = 92)

OR

2.4 (1.0 to 5.6)

> 1 favours duloxetine

YES favours

medication

1(H)

Moderate‐certainty evidence

7.3.1

Ayeleke 2015

PFMT + drug therapy vs drug therapy alone

IM

Schagen van Leeuwen 20041

(n = 101)

MD

5.8 (−2.1 to 13.7)

> 0 favours PFMT + drug

NO

2(E×2)

2.3

Kirchin 2017

Urethral injection therapy vs conservative management (PFMT)

IM

ter Meulen 20091

(n = 45)

MD

0.5 (0.1 to 0.9)

> 0 Favours PFMT

YES favours PFMT

1(H)

Not estimable

2.3

Mariappan 2005

SNRI vs conservative non‐ pharmacological therapies

IM

Schagen van Leeuwen 20042

(n = 99)

MD

NE

> 0 favours PFMT

NO

0

1Data from the I‐QOL measure of quality of life.
2The review Mariappan 2005 also reported data from Schagen van Leeuwen 2004, reporting the Physical Quality of Life Index (as event data,), and reporting the Incontinence Quality of Life questionnaire (but reporting the standard deviations as 0, and therefore not enabling an effect size estimate).
Abbreviations: CI: confidence interval; IM: immediate; MD: mean difference; n: number of participants; NE: not estimable; OR: odds ratio; PFMT: pelvic floor muscle training; RR: risk ratio; SNRI: serotonin–noradrenaline reuptake inhibitor.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Figuras y tablas -
Table 13. Stress urinary incontinence: quality of life, conservative versus non‐conservative
Table 14. Urgency urinary incontinence: primary outcome participant‐reported cure or improvement; conservative versus control

Analysis code (in review)

Study ID

Comparisons included

Timing of measure

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

High‐certainty evidence

2.2.2

Herderschee 2011

PFMT + feedback vs PFMT alone (control)

IM

Burgio 2002b1

(n = 122)

RR

0.5 (0.4 to 0.8)

> 1 favours PFMT

YES favours PFMT + feedback

0

1.9

Stewart 2016

ES vs no active treatment, placebo or sham treatment

IM

Finazzi‐Agrò 2010²

(n = 35)

RR

0.3 (0.1 to 0.6)

< 1 favours ES

YES favours ES

0

Moderate‐certainty evidence

1.2.2

Herderschee 2011

PFMT + biofeedback vs PFMT alone

IM

Burgio 20021

(n = 118)

RR

0.5 (0.4 to 0.8)

> 1 favours PFMT alone

YES favours PFMT + BF

1(C)

1.19.2

Stewart 2016

ES vs no active treatment, placebo or sham treatment

IM

Kosilov 20132

(n = 107)

MD

−1.1 (−1.8 to −0.4)

< 0 favours

ES

YES favours ES

1(H)

1.8

Stewart 2016

ES vs no active treatment, placebo or sham treatment

IM

Amaro 2006

(n = 40)

RR

0.5 (0.1 to 1.7)

< 1 favours

ES

NO

2(E×2)

1.4.1

Wallace 2004

Bladder training vs no treatment

FU1

Lagro‐Janssen 1991

(n = 18)

RR

17.0 (1.1 to 256.5)

> 1 favours BT

YES favours BT

1(H)

Low‐certainty evidence

5.2

Stewart 2016

ES + PFMT vs no active treatment

IM

Firra 20132

(n = 12)

MD

−1.6 (−3.6 to 0.4)

< 0 favours ES + PFMT

NO

3(A, E×2)

1Burgio 2002 was a three‐arm trial, in which the arms were PFMT (n = 75), PFMT plus feedback (n = 74) and PFMT plus biofeedback (n = 73). Overview authors judged that the PFMT only group met the definition to be classed as a control intervention. Therefore, there were two comparisons categorised within the overview as 'conservative versus control', one that was PFMT plus biofeedback versus control (Burgio 2002), and one that was PFMT plus feedback versus control (Burgio 2002b).
2Data from these trials are presented as 'number of incontinence episodes in 24 hours'.
Abbreviations: CI: confidence interval; ES: electrical stimulation; FU1: follow‐up one; IM: immediate; MD: mean difference; n: number of participants assessed NOT to have been cured or improved using a variety of different assessment methods; PFMT: pelvic floor muscle training; RR: risk ratio.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Figuras y tablas -
Table 14. Urgency urinary incontinence: primary outcome participant‐reported cure or improvement; conservative versus control
Table 15. Urgency urinary incontinence: quality of life, conservative versus control

Analysis code (in review)

Study ID

Comparisons included

Timing of measure

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

Moderate‐certainty evidence

5.1.1

Stewart 2016

ES + PFMT vs no active treatment

IM

Firra 2013

(n = 12)

MD

−12.4 (−20.9 to −3.9)

< 0 favours ES + other

YES favours ES + PFMT

1(A)

Abbreviations: CI: confidence interval; ES: electrical stimulation; IM: immediate; MD: mean difference; n: number of participants; PFMT: pelvic floor muscle training.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Figuras y tablas -
Table 15. Urgency urinary incontinence: quality of life, conservative versus control
Table 16. Urgency urinary incontinence: cure or improvement, conservative versus conservative

Analysis code (in review)

Study ID

Comparisons included

Timing of measure

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

Moderate‐certainty evidence

4.2.1

Herderschee 2011

PFMT + BF vs PFMT + F

IM

Burgio 2002c1

(n = 110)

RR

1.0 (0.6 to 1.7)

> 1 favours PFMT + feedback

NO

2(E×2)

2.8.2

Stewart 2016

ES vs laseropuncture

IM

Kosilov 20132

(n = 114)

MD

−1.8 (−2.3 to −1.3)

< 0 favours ES

YES favours ES

1(H)

4.2.1

Stewart 2016

ES + PFMT vs PFMT alone

IM

Schreiner 20102

(n = 51)

RR

0.56 (0.3 to 0.9)

< 1 favours ES + other

YES favours ES + PFMT

1(H)

4.3.1

Stewart 2016

ES + PFMT vs PFMT alone

IM

Schreiner 20102

(n = 51)

RR

0.3 (0.1 to 0.7)

< 1 favours ES + other

YES favours ES + PFMT

1(H)

2.3.1

Stewart 2016

ES vs PFMT

IM

Wang 20042

(n = 69)

RR

0.8 (0.5 to 1.2)

< 1 favours ES

NO

2(E×2)

2.3.2

Stewart 2016

ES vs PFMT +

biofeedback

IM

Wang 20042

(n = 69)

RR

0.9 (0.6 to 1.6)

< 1 favours ES

NO

2(E×2)

2.8.1

Stewart 2016

ES vs PFMT

IM

Arruda 20082

(n = 42)

MD

0.1 (−8.7 to 8.9)

< 0 favours ES

NO

2(E×2)

Low‐certainty evidence

5.2

Rai 2012

Anticholinergic drugs + non‐drug therapies vs non‐drug therapies alone

IM

Park 2002

(n = 50)

RR

0.6 (0.3 to 1.3)

< 1 favours anticholinergics + non‐drug therapies

NO

4(E×2, H×2)

5.2

Rai 2012

Anticholinergic drugs + non‐drug therapies vs non‐drug therapies alone

IM

Song 2006

(n = 57)

RR

0.6 (0.3 to 1.2)

< 1 favours anticholinergics + non‐drug therapies

NO

3(E×2, H)

4.9.2

Stewart 2016

ES + PFMT vs PFMT

IM

Firra 20132

(n = 12)

MD

0.2 (−1.2 to 1.6)

< 0 favours ES + other

NO

3(A, E×2)

1Burgio 2002 was a three‐arm trial in which two of the groups received PFMT, combined with either feedback or biofeedback. Both groups received active treatment, each comprising a different way of delivering the same intervention.
2These trials compared ES with another conservative intervention. However, in all cases both conservative interventions were categorised as 'physical therapy' interventions.
Abbreviation: CI: confidence interval; ES: electrical stimulation; IM: immediate; MD: mean difference; n: number of participants assessed NOT to have been cured or improved using a variety of different assessment methods; RR: risk ratio.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Figuras y tablas -
Table 16. Urgency urinary incontinence: cure or improvement, conservative versus conservative
Table 17. Urgency urinary incontinence: quality of life, conservative versus conservative

Analysis code (in review)

Study ID

Comparisons included

Timing of measure

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

High‐certainty evidence

2.5.1

Stewart 2016

ES vs PFMT

IM

Wang 20041,2

(n = 69)

MD

129.8 (47.8 to 211.8)

< 0 favours ES

YES favours PFMT

0

Moderate‐certainty evidence

4.5.1

Stewart 2016

ES + PFMT vs PFMT alone

IM

Schreiner 20101 (n = 51)

MD

−2.7 (−5.1 to −0.2)

< 0 favours ES + other

YES favours ES + PFMT

1(H)

2.5.2

Stewart 2016

ES vs PFMT +

biofeedback

IM

Wang 20041,2 (n = 69)

MD

−5.8 (−89.0 to 77.4)

< 0 favours ES

NO

2(E×2)

Low‐certainty evidence

4.8.1

Stewart 2016

ES + PFMT vs PFMT

IM

Firra 20131 (n = 12)

MD

5.8 (−3.5 to 15.1)

< 0 favours ES + other

NO

3(A, E×2)

1Studies compare two conservative interventions both categorised as 'physical therapy'.
2Wang 2004 was a three‐arm trial comparing PFMT, PFMT plus biofeedback and ES.
Abbreviations: CI: confidence interval; ES: electrical stimulation; IM: immediate; MD: mean difference; n: number of participants; PFMT: pelvic floor muscle training.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Figuras y tablas -
Table 17. Urgency urinary incontinence: quality of life, conservative versus conservative
Table 18. Urgency urinary incontinence: primary outcome participant‐reported cure or improvement; conservative versus non‐conservative

Analysis code (in review)

Study ID

Comparisons included

Timing of measure

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

Moderate‐certainty evidence

6.2.3

Rai 2012

Anticholinergic drugs + behavioural interventions vs anticholinergic drugs alone

IM

Burgio 2008; Burgio 20103

(n = 364)

RR

0.6 (0.4 to 0.9)

> 1 favours drugs alone

YES favours drugs + behavioural

interventions

2(F×2)

6.2

Rai 2012

Anticholinergic drugs + non‐drug therapies vs anticholinergic drugs alone

IM

Millard 2004

(n = 475)

RR

1.3 (0.8 to 1.9)

> 1 favours drugs

NO

1(H)

3.3

Rai 2012

Anticholinergic drugs vs external electrostimulation

IM

Wang 2006

(n = 47)

RR

1.5 (0.8 to 2.6)

> 1 favours ES

NO

2(E×2)

3.10.3

Stewart 2016

ES vs trospium +

solifenacin

IM

Kosilov 20131

(n = 110)

MD

2.2 (1.8 to 2.6)

< 0 favours ES

YES favours drugs

1(H)

4.9.1

Stewart 2016

ES + drugs vs drugs

IM

Sancaktar 20101

(n = 38)

MD

−0.9 (−1.1 to −0.7)

< 0 favours ES + other

YES favours ES + drugs

1(H)

3.10.2

Stewart 2016

ES vs oxybutynin

IM

Arruda 20081

(n = 43)

MD

0.9 (−6.4 to 8.2)

< 0 favours ES

NO

2(E×2)

3.6.1

Stewart 2016

ES vs tolterodine

IM

Franzén 2010

(n = 64)

RR

0.9 (0.4 to 2.0)

< 1 favours ES

NO

2(E×2)

4.9.3

Stewart 2016

ES + PFMT + drugs vs drugs

IM

Kaya 20111

(n = 30)

MD

−1 (−2.1 to 0.1)

< 0 favours ES + other

NO

2(E×2)

3.14.1

Stewart 2016

ES vs solifenacin succinate

IM

Vecchioli‐Scaldazza 20131

(n = 30)

MD

−0.9 (−2.0 to 0.2)

< 0 favours ES

NO

2(E×2)

Low‐certainty evidence

7.4.2

Ayeleke 2015

PFMT + drug therapy vs drug therapy alone

IM

Burgio 20103

(n = 58)

RR

0.9 (0.7 to 1.1)

> 1 favours PFMT + other

NO

3(E×2, H)

7.2

Rai 2012

Anticholinergic drugs vs combination non‐drug therapies

IM

Goode 2002

(n = 128)

RR

2.4 (1.0 to 5.8)

< 0 favours anticholinergics

NO

4(C, E×2, H)

1.3

Rai 2012

Anticholinergic drugs vs bladder training

IM

Colombo 19954

(n = 27)

RR

0.5 (0.4 to 4.5)

> 1 favours bladder training

NO

3(E×2, H)

1.3

Rai 2012

Anticholinergic drugs vs bladder training

IM

Milani 1986

(n = 75)

RR

0.8 (0.6 to 1.1)

> 1 favours bladder training

NO

3(E×2, H)

1.3

Rai 2012

Anticholinergic drugs vs bladder training

IM

Park 20022

(n = 48)

RR

0.8 (0.4 to 1.5)

> 1 favours bladder training

NO

4(E×2, H×2)

6.2

Rai 2012

Anticholinergic drugs + non‐drug therapies vs anticholinergic drugs alone

IM

Park 20022

(n = 50)

RR

0.7 (0.4 to 1.6)

> 1 favours drugs alone

NO

4(E×2, H×2)

3.3

Rai 2012

Anticholinergic drugs vs external electrostimulation

IM

Aaronson 1995

(n = 38)

RR

1.3 (0.7 to 2.3)

> 1 favours ES

NO

3(E×2, H)

1.3

Rai 2012

Anticholinergic drugs vs bladder training

IM

Song 20062

(n = 58)

RR

0.8 (0.4 to 1.5)

> 1 favours bladder training

NO

3(E×2, H)

6.2

Rai 2012

Anticholinergic drugs + non‐drug therapies vs anticholinergic drugs alone

IM

Song 20062

(n = 63)

RR

0.8 (0.4 to 1.6)

> 1 favours drugs alone

NO

3(E×2, H)

3.11

Stewart 2016

ES vs tolterodine

IM

Preyer 20071

(n = 31)

MD

1.3 (−1.6 to 4.2)

< 0 favours ES

NO

3(E×2, H)

3.5.1

Stewart 2016

ES vs oxybutynin after 12 weeks

IM

Souto 2014

(n = 37)

RR

0.3 (0.1 to 1.5)

< 1 favours ES

NO

3(E×2, H)

3.5.2

Stewart 2016

ES vs oxybutynin after 24 weeks

FU1

Souto 2014

(n = 37)

RR

0.6 (0.1 to 1.8)

< 1 favours ES

NO

3(E×2, H)

2.3

Wallace 2004

Bladder training vs anticholinergic drugs

IM

Colombo 19954

(n = 75)

RR

1.1 (0.9 to 1.3)

> 1 favours bladder training

NO

3(E×2, H)

1Data were presented as mean differences within these trials.
2Song 2006 and Park 2002 were both three‐arm trials, both with groups: bladder training, anticholinergic drug and bladder training plus anticholinergic drug.
3These are the same data for Burgio 2010. This was the same trial, included in two different reviews (the trial name within Ayeleke 2015 was "Burgio 2010a").
4Data from Colombo 1995 were included within both Kirchin 2017 and Wallace 2004. The methods used to determine number cure or improved differed within the two reviews, resulting with different effect sizes; however, both reviews presented data derived from the same participant groups.
Abbreviations: CI: confidence interval; ES: electrical stimulation; FU1: follow‐up one; IM: immediate; MD: mean difference; n: number of participants assessed NOT to have been cured or improved using a variety of different assessment methods; PFMT: pelvic floor muscle training; RR: risk ratio.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Figuras y tablas -
Table 18. Urgency urinary incontinence: primary outcome participant‐reported cure or improvement; conservative versus non‐conservative
Table 19. Urgency urinary incontinence: quality of life, conservative versus non‐conservative

Analysis code (in review)

Study ID

Comparisons included

Timing of measure

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

High‐certainty evidence

4.6.1

Stewart 2016

ES + PFMT + drugs vs drugs

IM

Kaya 2011

(n = 30)

MD

−16.2 (−28.0 to −4.4)

< 0 favours ES + other

YES favours ES + PFMT + drugs

0

Moderate‐certainty evidence

4.4.1

Stewart 2016

ES + drugs vs drugs alone

IM

Sancaktar 2010

(n = 38)

MD

−2.2 (−3.5 to −0.9)

< 0 favours ES + other

YES favours ES + drugs

1(H)

3.8.1

Stewart 2016

ES vs solifenacin succinate

IM

Vecchioli‐Scaldazza 2013

(n = 30)

MD

−0.6 (−1.3 to 0.1)

< 0 favours ES

YES favours ES + drugs

2(E×2)

Abbreviation: CI: confidence interval; ES: electrical stimulation; IM: immediate; MD: mean difference; n: number of participants; PFMT: pelvic floor muscle training.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Figuras y tablas -
Table 19. Urgency urinary incontinence: quality of life, conservative versus non‐conservative
Table 20. All types of urinary incontinence: primary outcome participant‐reported cure or improvement; conservative versus control

Analysis code (in review)

Study ID

Comparisons included

Timing of measure

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

High‐certainty evidence

1.2.4

Dumoulin 2018

PFMT vs no treatment, placebo or control

IM

Diokno 2010; Goode 2002 (n = 166)

RR

2.4 (1.6 to 3.5)

> 1 favours PFMT

YES favours PFMT

0

1.6

Imamura 2015

Weight loss vs no active intervention

IM

Subak 2005 (n = 40)

RR

16.5 (1.0 to 270.8)

At 3 months; > 1 favours weight loss

YES favours weight loss

0

11.19.1

Stewart 2016

ES vs no active treatment, placebo or sham treatment

IM

Barroso 20042 (n = 36)

MD

−1.7 (−2.3 to −1.0)

< 0 favours ES

YES favours ES

0

Moderate‐certainty evidence

1.1

Herbison 2013

Cones vs control

IM

Bø 1999; Williams 20061 (n = 215)

RR

0.7 (0.5 to 1.0)

> 1 favours control

YES favours cones

2(E×2)

1.1

Imamura 2015

Weight loss vs no active intervention

FU1

Wing 2010 (n = 304)

RR

1.4 (1.1 to 1.7)

> 1 favours weight loss

YES favours weight loss

1(C)

1.6

Imamura 2015

Weight loss vs no active intervention

FU1

Wing 2010 (n = 304)

RR

1.8 (1.2 to 2.8)

At 6 months; > 1 favours weight loss

YES favours weight loss

1(C)

1.6

Imamura 2015

Weight loss vs no active intervention

FU2

Wing 2010 (n = 287)

RR

1.1 (0.9 to 1.5)

At 18 months; > 1 favours weight loss

NO

2(C, E)

Low‐certainty evidence

1.6

Imamura 2015

Weight loss vs no active intervention

O

Phelan 2012;

Wing 2010 (n = 1032)

RR

1.2 (1.0 to 1.4)

> 1 favours weight loss

YES favours weight loss

3(C, E×2)

1.2

Stewart 2017

ES vs no active treatment

IM

Bø 1999; Castro 2008; Henalla 1989; Lopès 2014; Oldham 2013 (n = 347)

RR

1.73 (1.41 to 2.11)

> 1 favours ES

YES for ES

GRADE as assessed by review authors

1.20.4

Wallace 2004

Bladder training vs no treatment

IM

Fantl 1991 (n = 123)

RR

3.1 (2.0 to 5.0)

> 1 favours Bladder Training

YES favours BT

3(G, H×2)

Very low‐certainty evidence

1.1

Wieland 2019

Yoga vs waiting list control

IM

Huang 2014 (n = 18)

RR

6.33 (1.44 to 27.88)

> 1 favours yoga

YES for yoga

GRADE as assessed by review authors

1Bø 1999 contained only participants with SUI, but when pooled with Williams 2006 the pooled total contained a mixed population. Results from Williams 2006 are presented within the forest plot, but the individual study results from Bø 1999 are presented alongside the SUI data.
2Barroso 2004 presented data as a mean difference, rather than number cure or improved, and, therefore, these data are not included in the forest plot.
3Only data relating to immediate outcome measures were considered for inclusion in forest plots.
Abbreviations: CI: confidence interval; ES: electrical stimulation; FU1: follow‐up one; FU2: follow‐up two; IM: immediate; MD: mean difference; n: number of participants; O: other; PFMT: pelvic floor muscle training; RR: risk ratio.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Figuras y tablas -
Table 20. All types of urinary incontinence: primary outcome participant‐reported cure or improvement; conservative versus control
Table 21. All types of urinary incontinence: primary outcome quality of life; conservative versus control

Analysis code (in review)

Study ID

Comparisons included

Outcome measure (timing of measure)

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

High‐certainty evidence

1.11.4

Dumoulin 2018

PFMT vs no treatment, placebo or control

Incontinence Impact Questionnaire Long Form (IM)

Sran 2016 (n = 48)

MD

−52.67 (−95.00 to −10.34)

> 0 favours control

YES favours PFMT

0

1.16.4

Dumoulin 2018

PFMT vs no treatment, placebo or control

Incontinence Impact Questionnaire Long Form (FU)

Sran 2016 (n = 48)

MD

−41.91 (−83.20 to −0.62)

> 0 favours control

YES favours PFMT

0

Moderate‐certainty evidence

1.12.4

Dumoulin 2018

PFMT vs no treatment, placebo or control

Incontinence of Quality of Life Questionnaire (IM)

Sar 2009

(n = 34)

MD

28.9 (35.1 to 22.7)

> 0 favours control

YES favours PFMT

1(A) no change

1.10.4

Dumoulin 2018

PFMT vs no treatment, placebo or control

Incontinence Impact Questionnaire Short Form (IM)

Celiker Tosun 2015; Leong 2015 (n = 176)

MD

−7.54 (−14.70 to −0.39)

> 0 favours control

YES favours PFMT

2(F×2)

Very low‐certainty evidence

1.2

Wieland 2019

Yoga vs waiting list control

IM

Huang 2014 (n = 18)

MD

1.7 (−33.0 to 36.5)

> 0 favours control

NO

GRADE as assessed by review authors

Abbreviations: CI: confidence interval; FU: follow‐up; IM: immediate; MD: mean difference; n: number of participants; PFMT: pelvic floor muscle training.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Figuras y tablas -
Table 21. All types of urinary incontinence: primary outcome quality of life; conservative versus control
Table 22. All types of urinary incontinence: primary outcome participant‐reported cure or improvement; conservative versus conservative

Analysis code (in review)

Study ID

Comparisons included

Outcome measure (timing of measure)

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

High‐certainty evidence

1.2.31

Hay‐Smith 2011

Home PFMT programme (unsupervised) vs PFMT (Sapsford style2) with 8 clinic visits

IM

Hung 2010(n = 64)

RR

0.1 (0 to 0.7)

> 1 favours less contact/supervision

YES favours more contact/supervision

0

Moderate‐certainty evidence

3.2.1

Ayeleke 2015

PFMT + bladder training vs bladder training alone

IM

Wyman 1998(n = 128)

RR

1.3 (1.0 to 1.8)

> 1 favours PFMT + bladder training

YES favours PFMT + BT

2(C, H)

12.2.3

Hay‐Smith 2011

More‐intensive vs less‐intensive PFMT programmes

IM

de Oliveira 20093; Delgado 2009; Hay‐Smith 20023; Sriboonreung 20113; Sugaya 20033; Wells 1999(n = 405)

RR

0.7 (0.6 to 0.9)

> 1 favours less intensive

YES favours more intensive

1(H)

10.3

Wallace 2004

Bladder training + non‐pharmacological treatments vs non‐pharmacological treatments (PFMT)

IM

Wyman 1998(n = 124)

RR

1.2 (1.0 to 1.4)

> 1 favours bladder training + non‐pharmacological

YES favours BT + PFMT

2(G, H)

Low‐certainty evidence

3.5.1

Ayeleke 2015

PFMT + bladder training vs bladder training alone

IM

Kaya 2014; Wyman 1998

(n = 235)

RR

1.3 (1.1 to 1.4)

> 1 favours PFMT + bladder training

YES favours PFMT + BT

4(C, E×2, H)

3.2.2

Ayeleke 2015

PFMT + bladder training vs bladder training alone

FU1

Wyman 1998

(n = 120)

RR

1.3 (0.9 to 1.8)

> 1 favours PFMT + bladder training

NO

3(E×2, H)

3.5.2

Ayeleke 2015

PFMT + bladder training vs bladder training alone

FU1

Wyman 1998

(n = 119)

RR

1.2 (0.9 to 1.5)

> 1 favours PFMT + bladder training

NO

3(E×2, H)

12.2.1

Hay‐Smith 2011

More‐intensive vs less intensive PFMT programmes

IM

Bø 19903; Felicíssimo 20103; Schagen van Leeuwen 20043; Hung 2010; Konstantinidou 20073; Ramsay 19903

(n = 335)

RR

0.4 (0.2 to 0.8)

> 1 favours less intensive

YES more intense

3(E×2, H)

10.2

Hay‐Smith 2011

PFMT with intravaginal resistance device vs PFMT alone

IM

Delgado 2009; Wells 1999(n = 120)

RR

0.9 (0.6 to 1.2)

> 1 favours no device

NO

4(A, E×2, H)

1.2.2

Herderschee 2011

PFMT + biofeedback vs PFMT alone

IM

Burgio 20024; Goode 2003; Pages 2001; Wang 20044; Wilson 1987

(n = 343)

RR

0.7 (0.6 to 0.8)

> 1 favours PFMT alone

YES favours PFMT + BF

4(C, E×2, H)

1.2

Herderschee 2011

PFMT + biofeedback vs PFMT alone

IM

Burgio 20024; Goode 20033; Pages 20013; Wang 20044; Wilson 19873; Burns 1993; Morkved 20033

(n = 520)

RR

0.7 (0.7 to 0.9)

> 1 favours PFMT alone

YES favours BF

4(C, E×2, H)

1.2.1

Herderschee 2011

PFMT + biofeedback vs PFMT alone

IM

Burns 1993; Morkved 20033

(n = 177)

RR

0.9 (0.7 to 1.0)

> 1 favours PFMT alone

NO

4(E×2, H×2)

4.2.2

Herderschee 2011

PFMT + biofeedback vs PFMT + feedback

IM

Johnson 2000(n = 20)

RR

1.0 (0.2 to 5.8)

> 1 favours PFMT + feedback

NO

3(B, E×2)

4.2

Herderschee 2011

PFMT + biofeedback vs PFMT + feedback

IM

Burgio 20024; Johnson 2000

(n = 130)

RR

1.0 (0.6 to 1.6)

> 1 favours PFMT + feedback

NO

4(E×2, H×2)

6.2

Lipp 2014

Intraurethral Reliance device vs intraurethral FemAssist device

IM

Boos 1998(n = 101)

RR

1.0 (0.5 to 2.0)

> 1 favours reliance

NO

3(B, E×2)

5.3

Wallace 2004

Bladder training + behavioural/physical/psychological treatments

IM

Wyman 1998(n = 129)

RR

0.9 (0.7 to 1.1)

> 1 favours bladder training

NO

4(C, E×2, G)

5.4

Wallace 2004

Bladder training + behavioural/physical/psychological treatments

FU1

Wyman 1998 (n = 124)

RR

0.9 (0.7 to 1.1)

> 1 favours bladder training

NO

4(E×2, G, H)

10.4

Wallace 2004

Bladder training + non‐pharmacological treatments vs non‐pharmacological treatments

FU1

Wyman 1998 (n = 122)

RR

1.1 (0.9 to 1.3)

> 1 favours bladder training + non‐pharmacological

NO

4(E×2, G, H)

Very low‐certainty evidence

2.1

Herbison 2013

Cones vs PFMT

IM

Bø 19993; Cammu 19983; Peattie 1988; Williams 2006; Arvonen 20013; Arvonen 20023

(n = 358)

RR

0.9 (0.7 to 1.2)

> 1 favours PFMT

NO

5(C, F×2, G, H)

3.1

Herbison 2013

Cones vs ES

IM

Bø 19993; Oláh 1990; Santos 2009(n = 151)

RR

1.3 (0.8 to 1.9)

> 1 favours ES

NO

6(E×2, F×2, G, H)

3.2

Herbison 2013

Cones v ES

IM

Bø 19993; Castro 20083; Oláh 1990

(n = 154)

RR

1.2 (0.9 to 1.6)

> 1 favours ES

NO

6(E×2, F×2, G, H)

2.1

Wieland 2019

Yoga vs mindfulness stress reduction

IM

Baker 2014 (n = 24)

RR

0.09 (0.01 to 1.43)

> 1 favours yoga

NO

GRADE as assessed by review authors

2.2

Wieland 2019

Yoga vs mindfulness stress reduction

FU1

Baker 2014 (n = 20)

RR

0.2 (0 to 1.4)

> 1 favours yoga

NO

GRADE as assessed by review authors

2.3

Wieland 2019

Yoga vs mindfulness stress reduction

FU2

Baker 2014 (n = 21)

RR

0.2 (0 to 1.5)

> 1 favours yoga

NO

GRADE as assessed by review authors

1The data from Hung 2010 for this comparison and outcomes were included in both Analysis 1.2.3 and Analysis 3.2.3 in Hay‐Smith 2011. While the study data were the same in these analyses, the reported risk ratio differed. We are uncertain about the risk ratio presented in Analysis 3.2.3 and, therefore, chose to present the result from Analysis 1.2.3.
2Sapsford‐style PFMT involved diagrammatic and transversus abdominus co‐ordination to 'activate' pelvic floor muscle contraction and no direct pelvic floor muscle contractions.
3Trials with participants who had stress urinary incontinence only, where these were pooled with trials with other populations to give a pooled total relating to a combined participant population.
4Trials with participants who had urgency urinary incontinence only, where these were pooled with trials with other populations to give a pooled total relating to a combined participant population.
Abbreviations: CI: confidence interval; ES: electrical stimulation; FU: follow‐up; IM: immediate; n: number of participants assessed NOT to have been cured or improved using a variety of different assessment methods; PFMT: pelvic floor muscle training; RR: risk ratio.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Figuras y tablas -
Table 22. All types of urinary incontinence: primary outcome participant‐reported cure or improvement; conservative versus conservative
Table 23. All types of urinary incontinence: quality of life, conservative versus conservative

Analysis code (in review)

Study ID

Comparisons included

Outcome measure (timing of measure)

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

Moderate‐certainty evidence

3.3.1

Ayeleke 2015

PFMT + bladder training vs bladder training alone

IM

Wyman 19981

(n = 127)

MD

−25.5 (−49.9 to −1.0)

> 0 favours bladder training

YES favours PFMT + bladder training

1(H)

3.4.1

Ayeleke 2015

PFMT + bladder training vs bladder training alone

IM

Wyman 19982

(n = 128)

MD

−31.1 (−50.0 to −13.2)

> 0 favours bladder training

YES

favours PFMT + bladder training

2(C, H)

10.12

Wallace 2004

Bladder training + PFMT + biofeedback vs PFMT + biofeedback

IM

Wyman 19982

(n = 35)

MD

−47.2 (−87.0 to −7.4)

> 0 favours control

YES favours bladder training + PFMT +

biofeedback

2(G, H)

Low‐certainty evidence

3.3.2

Ayeleke 2015

PFMT + bladder training vs bladder training alone

FU1

Wyman 1998

(n = 118)

MD

−5.9 (−35.5 to 23.7)

> 0 favours bladder training

NO

3(E×2, H)

3.4.2

Ayeleke 2015

PFMT + bladder training vs bladder training alone

FU1

Wyman 1998

(n = 118)

MD

−18.9 (−38.0 to 0.1)

> 0 favours bladder training

NO

3(E×2, H)

6.7

Lipp 2014

Intraurethral Reliance device vs intraurethral FemAssist device

IM

Boos 19983

(n = 67)

MD

−1.0 (−12.6 to 10.6)

> 0 favours FemAssist

NO

3(B, E×2)

6.7

Lipp 2014

Intraurethral Reliance device vs intraurethral FemAssist device

IM

Boos 19983

(n = 67)

MD

−6.9 (−15.5 to 1.7)

> 0 favours FemAssist

NO

3(B, E×2)

6.7

Lipp 2014

Intraurethral Reliance device vs intraurethral FemAssist device

IM

Boos 19983

(n = 67)

MD

−7.4 (−16.7 to 1.9)

> 0 favours FemAssist

NO

3(B, E×2)

6.7

Lipp 2014

Intraurethral Reliance device vs intraurethral FemAssist device

IM

Boos 19983

(n = 67)

MD

5.2 (−5.0 to 15.4)

> 0 favours FemAssist

NO

3(B, E×2)

6.7

Lipp 2014

Intraurethral Reliance device vs intraurethral FemAssist device

IM

Boos 19983

(n = 67)

MD

−1.0 (−12.6 to 10.6)

> 0 favours FemAssist

NO

3(B, E×2)

6.7

Lipp 2014

Intraurethral Reliance device vs intraurethral FemAssist device

IM

Boos 19983

(n = 67)

MD

1.1 (−9.5 to 11.8)

> 0 favours FemAssist

NO

3(B, E×2)

6.7

Lipp 2014

Intraurethral Reliance device vs intraurethral FemAssist device

IM

Boos 19983

(n = 67)

MD

1.9 (−6.9 to 10.7)

> 0 favours FemAssist

NO

3(B, E×2)

10.13

Wallace 2004

Bladder training + non‐pharmacological treatment vs non‐pharmacological treatment

FU1

Wyman 1998

(n = 122)

MD

−12.2 (−30.4 to 6.0)

> 0 favours control

NO

4(E×2, G, H)

5.14.4

Wallace 2004

Bladder training vs other behavioural/physical/psychological treatments

IM

Wyman 19982

(n = 38)

MD

−28.0 (−68.4 to 12.4)

> 0 favours control

NO

4(E×2, G, H)

5.15

Wallace 2004

Bladder training vs other behavioural/physical/psychological treatments

FU1

Wyman 1998

(n = 124)

MD

6.7 (−12.2 to 25.6)

> 0 favours control

NO

4(E×2, G, H)

Very low‐certainty evidence

6.3

Stewart 2017

ES + PFMT vs PFMT

IM

Beuttenmuller 2010; Haig 1995; Patil 2010; Schmidt 2009 (n = 193)

SMD

−0.35 (−0.64 to −0.05)

< 0 favours ES + PFMT

YES for ES + PFMT

GRADE as assessed by review authors

1Outcome measured by the Incontinence Impact Questionnaire – Revised.
2Outcome measured by the Urinary Distress Inventory, for different pairs of treatment groups from this three‐arm trial.
3Boos 1998 compared two different types of mechanical devices.
Abbreviations: CI: confidence interval; FU1: follow‐up one; IM: immediate; MD: mean difference; n: number of participants; PFMT: pelvic floor muscle training; SMD: standardised mean difference.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Figuras y tablas -
Table 23. All types of urinary incontinence: quality of life, conservative versus conservative
Table 24. All types of urinary incontinence: primary outcome participant‐reported cure or improvement; conservative versus non‐conservative

Analysis code (in review)

Study ID

Comparisons included

Outcome measure (timing of measure)

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

Moderate‐certainty evidence

2.1.2

Alhasso 2005

Adrenergic agonist vs conservative therapy

UC

Wells 1991

(n = 157)

RR

1.4 (1.1 to 1.8)

> 1 favours drug

YES favours drugs

1(H)

Low‐certainty evidence

2.3.2

Alhasso 2005

Adrenergic agonist vs conservative therapy

UC

Wells 1991

(n = 86)

RR

0.9 (0.5 to 1.4)

> 1 favours drug; differential dropout from groups, more from PFMT group (P < 0.01)

NO

4(C, E×2, H)

1.3.1

Rai 2012

Anticholinergic drugs vs bladder training

IM

Macaulay 1987

(n = 29)

RR

1.0 (0.7 to 1.3)

> 1 favours bladder training

NO

3(E×2, H)

Abbreviations: CI: confidence interval; IM: immediate; n: number of participants; RR: risk ratio; UC: unclear.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Figuras y tablas -
Table 24. All types of urinary incontinence: primary outcome participant‐reported cure or improvement; conservative versus non‐conservative
Table 25. All types of urinary incontinence: quality of life; conservative versus non‐conservative

Analysis code (in review)

Study ID

Comparisons included

Outcome measure (timing of measure)

Included trials (n)

Effect measure

Effect estimate (CI)

Direction of effect

Evidence of beneficial effect?

Number of downgrades (reason for downgrade)

Moderate‐certainty evidence

2.13

Wallace 2004

Bladder training vs anticholinergic drugs

IM

Herbison 20041

(n = 34)

MD

−8.0 (−18.8 to 2.8)

> 0 favours bladder training

NO

2(E×2)

Low‐certainty evidence

9.13

Wallace 2004

Bladder training + pharmacological treatments vs pharmacological treatment alone

IM

Herbison 20041

(n = 28)

MD

2.0 (−6.8 to 10.8)

> 0 favours anticholinergic + bladder training

NO

3(C, E×2)

1Herbison 2004 was a three‐arm trial, where the groups were: bladder training, anticholinergic drug (oxybutynin) and bladder training + anticholinergic drugs. The anticholinergic drug group data within both analyses are the same.
Abbreviations: CI: confidence interval; IM: immediate; MD: mean difference; n: number of participants.
Codes for GRADE assessment: A: blinding of outcome assessor; B: allocation concealment; C: differential attrition; D: ROBIS domain 2; E: CIs plus sample size; F: heterogeneity using I2 statistic; G: type of UI was unclear or not reported; H: unclear or not reported for blinding, allocation and differential attrition.

Figuras y tablas -
Table 25. All types of urinary incontinence: quality of life; conservative versus non‐conservative
Table 26. Stress urinary incontinence: summary of findings – high‐ or moderate‐certainty evidence of beneficial effect on primary outcomes

Comparison group

Intervention of greater benefit

Comparison intervention

High‐ or moderate‐certainty evidence relating to cure or improvement

High‐ or moderate‐certainty evidence relating to quality of life

Review

Conservative intervention vs control

(see Table 8 and Table 9)

Cones

Control

Herbison 2013

PFMT + biofeedback

PFMT alone

Herderschee 2011

PFMT

Control

Dumoulin 2018

Intravaginal mechanical device

No treatment

Lipp 2014

ES

No treatment

Stewart 2017

1 conservative intervention vs another conservative intervention

(see Table 10 and Table 11)

Pessary + PFMT

Pessary alone

Lipp 2014

PFMT

Cones

Herbison 2013

PFMT

Intravaginal pessary

Lipp 2014

PFMT + adherence strategy

PFMT alone

Hay‐Smith 2011

More‐intensive PFMT

Less‐intensive PFMT

Hay‐Smith 2011

Conservative intervention vs non‐conservative intervention

(see Table 12 and Table 13)

Acupuncture

Any other treatment

Wang 2013

SNRI

Conservative

Mariappan 2005

PFMT

Urethral injection therapy

Kirchin 2017

Open retropubic colposuspension

Conservative

Lapitan 2017

ES: electrical stimulation; PFMT: pelvic floor muscle training; SNRI: serotonin–noradrenaline reuptake inhibitor.

Figuras y tablas -
Table 26. Stress urinary incontinence: summary of findings – high‐ or moderate‐certainty evidence of beneficial effect on primary outcomes
Table 27. Urgency urinary incontinence: summary of findings – high‐ or moderate‐certainty evidence of beneficial effect on primary outcomes

Comparison group

Intervention of greater benefit

Comparison intervention

High‐ or moderate‐certainty evidence relating to cure or improvement

High‐ or moderate‐certainty evidence relating to quality of life

Review

Conservative intervention vs control

(see Table 14 and Table 15)

PFMT + feedback/biofeedback

Control

Herderschee 2011

ES

Control

Stewart 2016

Bladder training

Control

Wallace 2004

ES + PFMT

No active treatment

Stewart 2016

1 conservative intervention vs another conservative intervention

(see Table 16 and Table 17)

ES

Laseropuncture

Stewart 2016

ES + PFMT

PFMT alone

Stewart 2016

ES

PFMT

Stewart 2016

Conservative intervention vs non‐conservative intervention

(see Table 18 and Table 19)

Trospium + solifenacin

ES

Stewart 2016

ES + drugs

Drugs

Stewart 2016

ES + PFMT + drugs

Drugs

Stewart 2016

Anticholinergic drugs + behavioural interventions (PFMT)

Anticholinergic drugs

Rai 2012

ES: electrical stimulation; PFMT: pelvic floor muscle training.

Figuras y tablas -
Table 27. Urgency urinary incontinence: summary of findings – high‐ or moderate‐certainty evidence of beneficial effect on primary outcomes
Table 28. All types of urinary incontinence: summary of findings – high‐ or moderate‐certainty evidence of beneficial effect on primary outcomes

Comparison group

Intervention of greater benefit

Comparison intervention

High‐ or moderate‐certainty evidence relating to cure or improvement

High‐ or moderate‐certainty evidence relating to quality of life

Review

Conservative intervention vs control

(see Table 20 and Table 21)

PFMT

Control

Dumoulin 2018

Weight loss

No active intervention

Imamura 2015

ES

Control

Stewart 2016

Cones

Control

Herbison 2013

1 conservative intervention vs another conservative intervention

(see Table 22 and Table 23)

Home PFMT programme (no health professional supervision)

Structured programme of 'indirect' PFMT (every 2 weeks, individual, health professional contact)

Hay‐Smith 2011

PFMT + bladder training

Bladder training

Ayeleke 2015

More‐intensive PFMT

Less‐intensive PFMT

Hay‐Smith 2011

PFMT + bladder training

PFMT

Wallace 2004

PFMT + biofeedback + bladder training

PFMT + biofeedback

Wallace 2004

Conservative intervention vs non‐conservative intervention

(see Table 24 and Table 25)

Adrenergic agonist

Conservative interventions

Alhasso 2005

ES: electrical stimulation; PFMT: pelvic floor muscle training.

Figuras y tablas -
Table 28. All types of urinary incontinence: summary of findings – high‐ or moderate‐certainty evidence of beneficial effect on primary outcomes
Table 29. Identification of relevant reviews mapped to six primary conservative categories

Primary conservative intervention (categories)

Review

Educational/behavioural/lifestyle

Imamura 2015 – lifestyle intervention
Wallace 2004 – bladder training
Ostaszkiewicz 2004a – habit training
Ostaszkiewicz 2004a – timed voiding
Eustice 2000 – prompted voiding

Physical therapies

Ayeleke 2015 – pelvic floor muscle training + another active treatment vs the same active treatment alone
Dumoulin 2018 – pelvic floor muscle training vs no treatment, or inactive control treatments
Hay‐Smith 2011 – comparisons of approaches to pelvic floor muscle training
Herderschee 2011 – feedback or biofeedback
Herbison 2013 – cones
Stewart 2016 and Stewart 2017 – electrical stimulation

Psychological therapies

Mechanical devices

Lipp 2014 – mechanical devices

Complementary therapies

Wang 2013 – acupuncture

Others

Wieland 2019 – yoga

Figuras y tablas -
Table 29. Identification of relevant reviews mapped to six primary conservative categories