Scolaris Content Display Scolaris Content Display

Opiáceos para el síndrome de piernas inquietas

Contraer todo Desplegar todo

Resumen

disponible en

Antecedentes

El síndrome de piernas inquietas (SPI) es un trastorno neurológico angustiante y frecuente que puede tener una gran repercusión en la calidad de vida de los pacientes con síntomas habituales e intensos. Los pacientes se quejan de sensaciones desagradables en las piernas, al dormir o antes de irse a dormir, y sienten el deseo de mover las piernas, que mejora con movimientos como caminar. Los síntomas comienzan con el paciente en reposo (p.ej., sentado o acostado) y siguen un patrón circadiano que aumenta durante la tarde o la noche. Para el SPI están disponibles muchas intervenciones farmacológicas que incluyen fármacos utilizados para tratar la enfermedad de Parkinson (L‐Dopa y agonistas dopaminérgicos), la epilepsia (anticonvulsivos), la ansiedad (benzodiazepinas) y el dolor (opiáceos). Los fármacos dopaminérgicos son los que se utilizan con mayor frecuencia para el tratamiento del SPI, pero algunos pacientes no responden de forma efectiva y requieren otros fármacos. Los opiáceos, una clase de fármacos utilizados para tratar el dolor intenso, parecen ser efectivos para tratar los síntomas del SPI y se recomiendan en los pacientes con síntomas graves, porque el SPI y el dolor parecen compartir el mismo mecanismo en el sistema nervioso central. Todos los fármacos disponibles se asocian hasta cierto punto con efectos secundarios, que pueden obstaculizar el tratamiento. Los opiáceos se asocian con eventos adversos como el estreñimiento, la tolerancia y la dependencia. Lo anterior justifica la realización de una revisión sistemática para evaluar si los opiáceos son seguros y efectivos para el tratamiento del SPI.

Objetivos

Evaluar los efectos de los opiáceos en comparación con placebo para el síndrome de piernas inquietas en pacientes adultos.

Métodos de búsqueda

Se hicieron búsquedas en el Registro Cochrane Central de Ensayos Controlados (Cochrane Central Register of Controlled trials), CENTRAL 2016, número 4 y en MEDLINE, EMBASE y LILACS hasta abril de 2016, utilizando una estrategia de búsqueda adaptada por Cochrane para identificar ensayos clínicos aleatorizados. Se verificaron las referencias de cada estudio y se estableció comunicación personal con otros autores para identificar cualquier estudio adicional. Se consideraron publicaciones en todos los idiomas.

Criterios de selección

Ensayos clínicos controlados aleatorizados de tratamiento con opiáceos en pacientes adultos con SPI idiopático.

Obtención y análisis de los datos

Dos autores de la revisión, de forma independiente, examinaron los artículos, extrajeron los datos en un formulario estándar y evaluaron el riesgo de sesgo. Si fue necesario, se discutieron las discrepancias con un tercer investigador para resolver cualquier duda.

Resultados principales

Se incluyó un ensayo clínico aleatorizado (n = 304 asignados al azar; 204 finalizaron; 276 analizados) que evaluó los opiáceos (oxicodona / naloxona de liberación prolongada) versus placebo. Después de 12 semanas, los síntomas del SPI habían mejorado más en el grupo de tratamiento farmacológico que en el grupo placebo (con el uso de la IRLSSS: DM ‐7,0; IC del 95%: ‐9,69 a ‐4,31 y el CGI: DM ‐1,11; IC del 95%: ‐1,49 a ‐0,73). Más pacientes del grupo de tratamiento farmacológico que del grupo placebo respondieron al fármaco (con el uso de la IRLSSS: RR 1,82; IC del 95%: 1,37 a 2,42 y el CGI: RR 1,92; IC del 95%: 1,49 a 2,48). La proporción de pacientes con remisión fue mayor en el grupo de tratamiento farmacológico que en el grupo placebo (con el uso de la IRLSSS: RR 2,14; IC del 95%: 1,45 a 3,16). Las puntuaciones de calidad de vida también mejoraron más en el grupo de tratamiento farmacológico que en el grupo placebo (DM ‐0,73; IC del 95%: ‐1,1 a ‐0,36). La calidad del sueño mejoró más en el grupo de fármacos medida según lo adecuado del sueño (DM ‐0,74; IC del 95%: ‐1,15 a ‐0,33) y la cantidad de sueño (DM 0,89; IC del 95%: 0,52 a 1,26).

No hubo diferencias entre los grupos en la somnolencia diurna, la dificultad para permanecer despiertos durante el día o las siestas durante el día. En el grupo de tratamiento farmacológico se informaron más eventos adversos (RR 1,22; IC del 95%: 1,07 a 1,39). Los principales eventos adversos fueron problemas gastrointestinales, fatiga y cefalea.

Conclusiones de los autores

Los opiáceos parecen ser efectivos para tratar los síntomas del SPI, pero no hay datos definitivos con respecto a este importante problema de seguridad. Esta conclusión se basa en un solo estudio con una alta tasa de abandono (evidencia de calidad baja).

PICO

Population
Intervention
Comparison
Outcome

El uso y la enseñanza del modelo PICO están muy extendidos en el ámbito de la atención sanitaria basada en la evidencia para formular preguntas y estrategias de búsqueda y para caracterizar estudios o metanálisis clínicos. PICO son las siglas en inglés de cuatro posibles componentes de una pregunta de investigación: paciente, población o problema; intervención; comparación; desenlace (outcome).

Para saber más sobre el uso del modelo PICO, puede consultar el Manual Cochrane.

Resumen en términos sencillos

Opiáceos para el síndrome de piernas inquietas

Antecedentes

El síndrome de piernas inquietas (SPI) es un trastorno neurológico muy frecuente en el que los pacientes se quejan de una necesidad intensa de mover las piernas y de que sienten sensaciones desagradables en las piernas, y todo lo anterior se presenta durante el sueño, principalmente a la hora de acostarse. El número de pacientes que se quejan del SPI varía según la raza, el sexo, la edad, el país y el estado de salud. Alrededor del 5% al 10% de las personas están afectadas; y, entre ellas, del 2% al 5% necesitan un tratamiento farmacológico continuo (fármaco). Cuando el SPI no responde a los fármacos generalmente utilizados para la enfermedad de Parkinson y la epilepsia, los médicos a menudo prescriben opiáceos.

Pregunta

¿Los opiáceos son efectivos y seguros para los pacientes con SPI?

Métodos

Se buscaron en la bibliografía los estudios en cualquier idioma, publicados o no, que consideraran los opiáceos para el tratamiento del SPI

Resultados

Se incluyó un ensayo clínico controlado aleatorizado con riesgo moderado de sesgo que probó una combinación de oxicodona y naloxona versus cápsulas placebo, administradas dos veces al día en pacientes que no respondieron a los fármacos más habituales. Los investigadores utilizaron la International RLS severity scale para determinar si los pacientes mejoraron después de 12 semanas de tratamiento. Los participantes que recibieron la combinación de oxicodona y naloxona informaron de una mejora en los síntomas del SPI, la calidad de vida y la calidad del sueño; el 42% del grupo del fármaco no presentaba síntomas.

Discusión

El estudio en general estuvo bien diseñado, pero tuvo un alto riesgo de sesgo debido al porcentaje alto de participantes que se retiraron del tratamiento (sesgo de deserción). El 84% del grupo del fármaco desarrolló eventos adversos, que en su mayoría estuvieron relacionados con el sistema gastrointestinal, dolor de cabeza, fatiga y somnolencia; el 9,8% abandonó el estudio debido a los eventos adversos.

Conclusión

El uso de los opiáceos para el tratamiento del SPI en los pacientes resistentes al tratamiento convencional está respaldado por evidencia de calidad moderada. La prescripción de estos fármacos se debe basar en la experiencia clínica y se debe tener precaución debido a la posibilidad de abuso, dependencia y eventos adversos. Ningún paciente que recibió opiáceos se quejó de que empeoraran los síntomas.

Authors' conclusions

Implications for practice

This systematic review found low evidence, based on one trial, in favour of prolonged release oxycodone/naloxone to treat RLS symptoms in patients who had failed previous treatment. The benefits of opioids included amelioration of subjective RLS symptoms and improved quality of life. Adverse events were common during opioid treatment. They were mostly mild to moderate in severity and included gastrointestinal complaints (obstipation, ileus, sub‐ileus, nausea, vomiting, flank pain), fatigue, headache, somnolence, dizziness, dry mouth, and pruritus. Withdrawal symptoms were also reported.

Further large, well‐designed, randomised clinical trials, that address issues such as minimal clinically important change in RSL symptoms, adverse events, and respiratory polysomnographic data, are still required before clinicians should consider prescribing opioids to treat symptoms of RLS.

Implications for research

Many opioid drugs are available to use in RLS, and it would be of interest to perform randomised clinical trial to assess the effect of these agents in this very distressing condition. Even though opioids have demonstrated effectiveness for the treatment of RLS symptoms, this drug class still has to be compared to other agents recommended for RLS. Furthermore, the impact of opioid treatment on sleep variables as assessed by polysomnography should be evaluated.

Summary of findings

Open in table viewer
Summary of findings for the main comparison. Opioids compared to placebo for RLS

Opioid treatment compared to placebo for patients with RLS

Patient or population: RLS
Setting: 55 hospitals and specialised private neurology practices in Austria, Germany, Spain, and Sweden.
Intervention: opioids
Comparison: placebo

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Quality of the evidence
(GRADE)

Comments

Risk with placebo

Risk with opioids

RLS symptoms
assessed with: IRLSSS
Scale from: 0 to 40
follow up: mean 12 weeks

The mean RLS symptoms was 22.1 points

The mean RLS symptoms in the intervention group was 7 points lower (9,69 lower to 4,31 lower)

270
(1 RCT)

⊕⊕⊝⊝
LOW 1

Moderate effect size of difference in mean response (0.57) 2

RLS symptoms
assessed with: CGI
Scale from: 0 to 7
follow up: mean 12 weeks

The mean RLS symptoms was 4.1 points

The mean RLS symptoms in the intervention group was 1,11 points lower (1,49 lower to 0,73 lower)

276
(1 RCT)

⊕⊕⊝⊝
LOW 1

Moderate effect size of difference in mean response (0.64) 2

Drug responders
assessed with: IRLSSS ‐ Reduced score at least 50%
follow up: mean 12 days

Study population

RR 1.82
(1.37 to 2.42)

276
(1 RCT)

⊕⊕⊝⊝
LOW 1

313 per 1.000

569 per 1.000
(428 to 756)

Drug responders
assessed with: CGI ‐ Self reported "Much improved" or "Very much improved"
follow up: mean 12 weeks

Study population

RR 1.92
(1.49 to 2.48)

276
(1 RCT)

⊕⊕⊝⊝
LOW 1

347 per 1.000

667 per 1.000
(517 to 861)

Remitters
assessed with: IRLSSS ‐ Scored 10 or less
follow up: mean 12 weeks

Study population

RR 2.14
(1.45 to 3.16)

276
(1 RCT)

⊕⊕⊝⊝
LOW 1

194 per 1.000

416 per 1.000
(282 to 614)

Adverse Events
assessed with: clinical assessment
follow up: mean 12 weeks

Study population

RR 1.22
(1.07 to 1.39)

304
(1 RCT)

⊕⊕⊝⊝
LOW 1

688 per 1.000

840 per 1.000
(736 to 957)

Quality of life
assessed with: RLS Quality of Life Questionnaire (RLS‐QoL)
Scale from: 0 to 7
follow up: mean 12 weeks

The mean quality of life was 3.64 points

The mean quality of life in the intervention group was 0,73 points lower (1,1 lower to 0,36 lower)

276
(1 RCT)

⊕⊕⊝⊝
LOW 1

Small effect size of difference in mean response (0.43) 2

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; RR: Risk ratio; OR: Odds ratio;

GRADE Working Group grades of evidence
High quality: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate quality: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low quality: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low quality: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

1 The only trial included presentes high risk of attrition bias.

2 0.2 represents a small effect, 0.5 a moderate effect, and 0.8 a large effect of difference in mean response (Cohen 1988).

Background

Description of the condition

Restless legs syndrome (RLS) is a sensorimotor disorder characterised by a distressing urge to move the legs, and sometimes, other parts of the body as well, usually accompanied by a marked sense of discomfort or pain in the leg or other affected body parts (Allen 2003).

The prevalence of RLS is estimated at 5% to 15% in adults. It is more common in women, and can affect children as well (Picchietti 2005; Picchietti 2007[ Yeh 2012). When frequency, severity, or a combination of symptoms is added to the diagnostic criteria, the prevalence of RLS ranges from 2.2% to 7.9%. If the diagnosis is based on a clinical interview, taking into account all possible differential diagnoses, the prevalence declines to between 1.9% and 4.6% (Ohayon 2012).

Other features commonly found in adults with RLS include sleep disturbance, daytime fatigue, and decreased quality of life ratings, mostly in patients who also have iron deficiency anaemia (Allen 2013; Picchietti 2005).

The physical examination is typically normal. restless leg syndrome may be either idiopathic (primary RLS, which often has a familial component) or secondary, occurring in conjunction with other medical conditions, particularly iron deficiency anaemia, pregnancy, or end‐stage renal disease. Secondary RLS tends to remit without evidence of reoccurrence when the secondary condition is resolved, for example, after renal transplantation in patients with end‐stage renal disease, and postpartum in women with RLS during pregnancy (Lee 2001; Winkelmann 2002).

Periodic leg movements in sleep (PLMS) are characterized by brief (0.5‐ to 5.0‐second) lower‐extremity movements during sleep, which typically occur at 20‐ to 90‐second intervals, most commonly during the first three hours of sleep. The affected individual is usually not aware of the movements or of the associated transient partial arousals (Picchietti 2005; Trenkwalder 2005). Overall, 80% to 90% of individuals with RLS have PLMS, but PLMS are not necessarily associated with RLS (Hornyak 2006; Rye 2012). The prevalence of PLMS in the general population is 3.9% (Ohayon 2002).

Description of the intervention

The treatment of RLS includes pharmacological and non‐pharmacological therapies. The most common pharmacological agents used in clinical practice are levodopa, dopamine agonists, opioids, benzodiazepines, and anticonvulsants (Garcia‐Borreguero 2013; Trenkwalder 2005 ).

The precise mechanisms by which opiates ameliorate RLS are not well understood. Opiate receptors have been identified in the dorsal horn, where they are believed to participate in the regulation of incoming nociceptive sensory information. Opiates are also highly concentrated in brainstem areas, around the periaqueductal grey and in the basal ganglia (striatum, substantia nigra); each of these areas could be sites in which the opioids act to improve symptoms of RLS (Sandyk 1987). Involvement of the dopamine system in RLS pathophysiology seems probable, but involvement of the opiate system is less clear.

Narcotic medications generally have a relatively low potential for addiction, and cause little tolerance in the RLS population (Winkelmann 2002). Side effects include nausea, sedation, dizziness, and constipation. There are still concerns about the potential for abuse, addiction, and practical problems, so the treatment of RLS with opioids remains controversial.

The essentially normal presynaptic dopaminergic binding studies using F‐dopa PET or B‐CIT‐SPECT (diagnostic imaging tools) in patients with RLS lend support to the hypothesis that dopaminergic neurons and spinal pathways could be involved more in the pathophysiological mechanisms of the syndrome than the nigrostriatal system (Wetter 2004).

Althoug dopamine can be increased in the synaptic cleft of patients with RLS, the strongest evidence for a dopaminergic role in the pathophysiology of RLS comes from the pharmacological response to medications that increases dopamine function (Allen 2004; Earley 2013). Functional imaging studies have shown reduced fluoro dopa uptake or reduced D2 receptor binding in the corpus striatum (Wetter 2004).

Iron is a cofactor in dopamine production, and RLS patients exhibit a deficiency of iron in the brain. Studies with iron‐deprived rats indicate that low CSF ferritin and high transferrin can be expected to occur with reduced brain iron (Allen 2004; Rizzo 2013).

How the intervention might work

The endogenous opioid system plays a role in pain transmission , and there is evidence of opioid receptors involvement in the pathogenesis of RLS (Hening 1986; Mizoguchi 2014). A post‐mortem study of the brains of patients with RLS showed deficiencies of beta‐enkephalin and met‐endorphin in the thalamus (Walters 2009), which also suggests a direct implication of this system in the pathogenesis of RLS symptoms. Furthermore, mu‐receptors in rats presented signs suggestive of RLS, and interestingly, iron deficiency, which is somehow related to the clinical picture of RLS (DeAndrade 2013; Earley 2014). Hence, there is a body of data supporting the hypothesis that opioid drugs can improve endogenous opioid system function, thus providing improvement or even complete relief of RLS symptoms.

Why it is important to do this review

Many pharmacological intervention are available for RLS, including drugs used to treat Parkinson's disease (L‐Dopa, dopaminergic agonists), epilepsy (anticonvulsants), anxiety (benzodiazepines), and pain (opioids). Dopaminergic drugs are the most frequent treatment used for RLS, but some patients do not respond effectively and require other medication. Opioids, a class of medications used to treat severe pain, seem to be effective in treating RLS symptoms and are recommended for patients with severe symptoms, because RLS and pain appear to share the same mechanism in the central nervous system. All available drugs are associated to some degree with side effects, which can impede treatment. Dopaminergic drugs can cause worsening of symptoms, which is known as augmentation, anticonvulsant drugs are associated with somnolence, and benzodiazepines are beneficial only for minor symptoms of RLS. Opioids are associated with adverse events such as constipation, tolerance, and dependence. This justifies the conduct of a systematic review to ascertain whether opioids are safe and effective for the treatment of RLS.

Objectives

To evaluate the efficacy and safety of opioid treatment for idiopathic RLS.

Methods

Criteria for considering studies for this review

Types of studies

All randomised controlled trials (RCTs) and quasi‐RCTs (defined as trials using inadequate allocation assignment such as date of birth, day of the week or month of the year, person's medical record number, or simply allocating every alternate person). We considered studies with a parallel or a cross‐over design.

Types of participants

Inclusion criteria

We considered children and adults who met any clinical criteria for idiopathic RLS (ICSD 2014; Walters 1995). A recent version of the criteria for clinical diagnostic lists four essential features (Allen 2003):

  1. An urge to move the legs, usually accompanied or caused by uncomfortable and unpleasant sensations in the legs (sometimes the urge to move is present without the uncomfortable sensations, and sometimes the arms or other body parts are involved in addition to the legs);

  2. The urge to move or unpleasant sensations begin or worsen during periods of rest or inactivity, such as lying or sitting;

  3. The urge to move or unpleasant sensations are partially or totally relieved by movement, such as walking or stretching, at least as long as the activity continues;

  4. The urge to move or unpleasant sensations are worse in the evening or night than during the day, or only occur in the evening or night (when symptoms are very severe, the worsening at night may not be noticeable but must have been previously present).

Some of the trials identified included patients with Periodic Limb Movement Disorder (PLMD), which is commonly associated with RLS. Studies exclusively examining patients with PLMD (without symptoms of restless legs) were excluded.

Exclusion criteria

We excluded studies that included patients with secondary forms of RLS, such as metabolic, neuropathic, or renal disease.

Types of interventions

We included trials that compared opioid drugs to placebo, to no treatment, or to other drug treatments.

Types of outcome measures

Primary outcomes

Improvement of restless legs symptoms, as assessed by a validated scale (Allen 2001; IRLSSG 2003).

Secondary outcomes

  1. Subjective sleep quality (any description about sleep quality, i.e., well‐being, improvement of fatigue);

  2. Sleep quality, as measured by overnight polysomnography (sleep efficiency, total sleep time, arousal index, PLMS index);

  3. Quality of life, as measured by a validated scale, such as the SF‐36;

  4. Adverse events, described in terms of:

    1. Number of withdrawals due to adverse events;

    2. Number of patients with any adverse events associated with interventions.

Search methods for identification of studies

Electronic searches

  1. Cochrane Central Register of Controlled Trials, (CENTRAL) 2016, Issue 4, in The Cochrane Library (accessed April 2016; Appendix 1).

  2. MEDLINE (1966 to April 2016), using the optimally sensitive strategy developed for the Cochrane Collaboration for the identification of randomised controlled trials ;(Appendix 2).

  3. EMBASE (1980 to April 2016), using a search strategy adapted from one developed for the Cochrane Collaboration for the identification of randomised controlled clinical trials (Appendix 3).

  4. LILACS (1982 to April 2016), using a search strategy adapted from one developed for the Cochrane Collaboration for the identification of randomised controlled clinical trials (Appendix 4).

Searching other resources

We assessed references from original papers and abstracts, reviews, systematic reviews, and meta‐analysis to identify any additional studies.
We contacted authors of the included studies to ask if they knew of any relevant unpublished material.

Data collection and analysis

Selection of studies

The search strategies described above were used to obtain titles and abstracts of relevant studies, which were independently screened by LC and KC. They initially retained review articles that might include relevant data or information on trials. The review authors independently assessed the retrieved abstracts, and if necessary, the full text of these studies to determine which studies met the inclusion criteria.

Data extraction and management

The same review authors independently extracted data, using standard data extraction forms. The two review authors entered the data into Review Manager software once all disagreements had been addressed. They had studies reported in non‐English language journals translated before being assessed. Where more than one publication of a trial existed, the papers were grouped, and for each available outcome, results were extracted from the publication with the most complete data. We requested further information from the original author by written correspondence as required, and any relevant information obtained in this manner was included in the review. We resolved all disagreements by consensus, with a third review author if needed.

Assessment of risk of bias in included studies

The same two review authors independently assessed the risk of bias of the included studies, without blinding to authorship or journal. They resolved any disagreement by discussion.

We incorporated the 'risk of bias' assessments into the 'risk of bias' tables, as described in the Cochrane Hankbook for Systematic Reviews of Interventions (Higgins 2011). We considered these criteria: adequate sequence generation, allocation concealment, blinding of participants, personnel, and outcome assessment, incomplete outcome data, selective outcome reporting, and other biases. We assigned 'low risk', 'high risk', or 'unclear risk' judgements to each criterion. 'Unclear risk' indicated there was insufficient information to permit a clear judgement.

We used the GRADE approach to assess quality of the evidence for each outcome across trials, and presented the results in a 'Summary of findings' table. We categorised the primary outcomes at the highest level, and downgraded them due to the following study limitations (risk of bias): limitations in design, inconsistency of results, indirectness of evidence, imprecision of results, and publication bias. We decreased the quality of the evidence by one point if there were serious problems with the risk of bias criteria, or two points if there were very serious problems.

For the 'Summary of findings' table, we analysed the primary outcome: RLS Symptoms, using the International RLS Severity Scale (IRLSSS) and the Clinical Global Impression severity

scale (CGI); drug responders, using the IRLSSS and CGI, and Remitters (decreased symptoms), using the IRLSSS; and the secondary outcomes: adverse events and quality of life (Summary of findings table 1).

Measures of treatment effect

We entered and analysed data in Review Manager 5.3 (RevMan 2014) software. We constructed the 'Summary of findings' tables using the GRADE profiler 3.2.2 software (GRADEpro 2014). For dichotomous variables, risk ratios (RR) with 95% confidence intervals (95% CI) were calculated using the fixed‐effects model. Mean differences (MD) with 95% confidence intervals were calculated for continuous outcome variables, using the fixed‐effects model.

Unit of analysis issues

Only one trial was included, in a simple parallel group design. In future updates, we may include cross‐over trials, only if they allow pooling of data and analysis according to the recommendations of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We will evaluate the two cross‐over periods with a paired analysis, and the carry‐over effect

Dealing with missing data

In suspected cases of missing data, we contacted the primary investigator of the study.

Assessment of heterogeneity

As only one study was included, we did not assess heterogeneity. In future updates, we will assess heterogeneity (with the Chi² test), which we will assume to be present when the significance level is lower than 0.10 (P < 0.10). When significant heterogeneity is present, we will attempt to explain the differences based on clinical characteristics of the included studies.

Assessment of reporting biases

To reduce reporting bias, we contacted as many authors who are involved in RLS research as possible, and asked about any unpublished trials of which they might be aware. We also search the International Committee of Medical Journal Editors for trials that were registered and not published. We had planned to use a funnel plot (trial effect versus trial size) to explore the possibility of publication bias, but did not found sufficient studies (10 or more) for any of the primary analyses.

Data synthesis

As there was only one trial, we did not pool the data with a meta‐analysis. In futures updates, for clinically homogeneous studies, we will pool outcomes in meta‐analyses, using the fixed‐effect model as a default, and the inverse variance method.

Subgroup analysis and investigation of heterogeneity

We did not performed subgroup analysis in this review. In future updates, if a sufficient number of studies (more than 10) are eligible, we will perform subgroup analyses according to age, gender, and duration of treatment. We will categorize periods of treatment as short‐term (up to four weeks), or long‐term (more than four weeks).

Sensitivity analysis

We did not perform sensitivity analyses, as only one trial was included. Assuming we have sufficient trials in future updates, we will perform sensitivity analyses by omitting trials that include participants with different clinical characteristics, or trials with higher risk of bias.The following strategies will be used for the sensitivity analyses:

  1. Separating RCTs published as abstracts.

  2. Separating RCTs of lower risk of bias as assessed by allocation concealment.

  3. Separating RCTs without an intention‐to‐treat analysis.

  4. Separating cross‐over studies from the analysis.

Results

Description of studies

Results of the search

Our search of electronics databases yielded a total of 2935 publications with 45 duplicate records: 246 records from CENTRAL, 2204 from MEDLINE, 483 from EMBASE, none from LILACS, and two from manual sources (Figure 1).. After analysing the title and abstracts, four of these publications were selected for full‐text analysis; their eligibility was subsequently confirmed. One study was published first as an abstract in 1991, but we retained only the complete publication for analysis (Kaplan 1993). Allen 1992, Kaplan 1993 and Walters 1993 were excluded because there were insufficient individual cross‐over data available to analyse the variance of the two periods, the wash‐out period, or the first period. We ultimately included one study(Trenkwalder 2013).


Diagram about search, excluded and included researches.

Diagram about search, excluded and included researches.

Included studies

Trenkwalder 2013 was a randomised, double‐blind, parallel, multicentre trial that enrolled 304 adults with RLS. The patients were recruited from 55 sites in Austria, Germany, Spain, and Sweden. They received progressive doses of prolonged‐release oxycodone (maximum 40 mg taken twice daily) and naloxone (maximum 20 mg taken twice daily), or placebo. The investigators used the IRLSSS to assess RLS symptoms, responders and remitters, and the Clinical Global Impression (CGI) scales to assess RLS symptoms, adverse events, and proportion of responders. Subjective sleep change was measured with the Medical Outcome Study (MOS). Change in disease‐specific quality of life was assessed with the RLS‐QoL questionnaire. Authors considered responders those patients that scored 50% less in the IRLSSS and those who declared themselves "much improved" or "very much improved" in the CGI scale; remitters those patients that scored 10 or less in the IRLSSS at the end of treatment; reduced symptoms those patients that scored any value less compared to the baseline score.

One hundred patients dropped out of the study due to adverse events, lack of therapeutic effect, patient choice, or administrative reasons. Analysis was completed on a sample of 276 participants according to the intention‐to‐treat principle, which included all participants who had received at least one dose of the medication or placebo, and for whom follow‐up data were available. The included patients had a high mean IRLSSS at baseline (31.6 ± 4.7). We contacted the first author (Dr Claudia Trenkwalder) on 8 November 2013 to ask her how she dealt with the missing data.

Excluded studies

Allen 1992: a randomised, double‐blind, cross‐over trial of six adult patients. The patients had PLMS and were all diagnosed with RLS on retrospective review. The patients were recruited from The Johns Hopkins Sleep Center, Baltimore, MD, USA. They received progressive doses of propoxyphene (maximum 300 mg), carbidopa/levodopa (maximum 100mg/200mg), or placebo for two weeks prior to polysomnographic studies. We contacted the first author (Dr. Richard Allen) on 20 October 2013, who provided some important information, but the available data were not sufficient or sufficiently detailed to enable inclusion of the study in this review.

Kaplan 1993: a randomised, double‐blind, cross‐over trial of six adult patients with PLMS who were all diagnosed with RLS on retrospective review. This new sample of patients was recruited from The Johns Hopkins Sleep Center, Baltimore, MD, USA. They received progressive doses of propoxyphene (maximum 200 mg), carbidopa/levodopa (maximum 100 mg/200 mg), or placebo for two weeks prior to polysomnographic studies. We contacted one of the authors (Dr. Richard Allen) on 20 October 2013, who provided some important information, but the available data were not sufficient or sufficiently detailed to enable inclusion of the study in this review.

Walters 1993: a randomised, double‐blind, cross‐over trial of 11 adult patients with RLS and PLMS. The diagnosis of RLS was established clinically by the presence of: abnormal sensations, primarily in the legs, motor restlessness, and worsening of paraesthesias and motor restlessness at night and at rest. The patients were recruited from the Lyons VA Medical Center, UMDNJ‐Robert Wood Johnson University Hospital, and the Sleep Disorders Center of Columbia Presbyterian Medical Center. They received progressive doses of oxycodone or placebo under guidance of a member of the research group, to relieve symptoms. Oxycodone or placebo capsules were tapered off in three days, and the second‐phase dose titration started immediately. Paresthesias, motor restlessness, and daytime alertness were rated on a symptom severity scale of zero to four, for two weeks prior to polysomnographic studies, and on the night of polysomnography. We contacted the first author (Dr Arthur Walters) on 16 October 2013, who provided some important information, but again, the available data were not sufficient or sufficiently detailed to enable inclusion of the study in this review.

Risk of bias in included studies

The 'Risk of bias' assessments for the included study can be found in Figure 2.


Risk of bias summary: review authors' judgements about each risk of bias criterion for each included study.

Risk of bias summary: review authors' judgements about each risk of bias criterion for each included study.

Allocation

Low risk: the randomisation was conducted with a validated interactive response technology system that automated the random assignment of treatment groups to randomisation numbers by site, in blocks of four.

Blinding

Low risk for performance bias: matching placebo tablets (identical in appearance, colour, and taste).

Low risk for detection bias: during the double‐blind phase, patients and all personnel involved in the conduct and interpretation of the study (including investigators, site personnel, and sponsor staff) were masked to treatment assignment.

Incomplete outcome data

High risk: a total of 306 patients was randomised, but only 174 completed the 12 week study; intention‐to‐treat analysis was provided.

Selective reporting

Low risk: all relevant clinical outcomes were analysed; no subgroup analysis was performed; supplementary data was also available.

Other potential sources of bias

Unclear risk: study was proposed by Mundipharma; the principle investigator and three investigators declared conflict of interest with this pharmaceutical company.

Effects of interventions

See: Summary of findings for the main comparison Opioids compared to placebo for RLS

Primary outcome measure

RLS Symptoms

In the Trenkwalder 2013 study (N = 276), RLS symptoms, measured on the IRLSSS were reduced more in the opioids group than in the placebo group (MD ‐7.0; 95% CI ‐9.69 to ‐4.31; Analysis 1.1; Figure 3). Symtoms measured on the scale also improved more in the opioid group than in the placebo group (MD ‐1.11; 95% CI ‐1.49 to ‐0.73; Analysis 1.2; ).The proportion of drug responders measured on the IRLSSS was greater in the opioid group than in the placebo group (RR 1.82; 95% CI 1.37 to 2.42; Analysis 1.3; Figure 4). Measured on the CGI scale, the proportion of participants who responded in the opioid group was higher than in the placebo group (RR 1.92; 95% CI 1.49 to 2.48; Analysis 1.4; ).The proportion of remitters, measured by the IRLSSS was greater in the opioid group than in the placebo group (RR 2.14; 95% CI 1.45 to 3.16; Analysis 1.5; Figure 5). The quality for these three outcomes was downgraded to low because of attrition bias (summary of findings Table for the main comparison).


Forest plot of comparison: 1 opioids and placebo, outcome: 1.1 RLS symptoms ‐ IRLSSS.

Forest plot of comparison: 1 opioids and placebo, outcome: 1.1 RLS symptoms ‐ IRLSSS.


Forest plot of comparison: 1 opioids and placebo, outcome: 1.3 Drug responders ‐ IRLSSS.

Forest plot of comparison: 1 opioids and placebo, outcome: 1.3 Drug responders ‐ IRLSSS.


Forest plot of comparison: 1 opioids and placebo, outcome: 1.5 Remitters ‐ IRLSSS.

Forest plot of comparison: 1 opioids and placebo, outcome: 1.5 Remitters ‐ IRLSSS.

Secondary outcome measures

Quality of Sleep

Quality of sleep was improved in the drug group more than in the placebo group, measured by sleep adequacy (MD ‐0.74; 95% CI ‐1.15 to ‐0.33; Analysis 1.6); and sleep quantity (MD 0.89; 95% CI 0.52 to 1.26; Analysis 1.7). There was no significant difference between groups for daytime somnolence (MD 0,21; 95% CI ‐0.13 to 0.55; Analysis 1.8), trouble staying awake during the day (MD 0,13; 95% CI ‐0.19 to 0.45; Analysis 1.9), or naps during the day (MD 0,01; 95% CI ‐0.32 to 0.34; Analysis 1.10).

Quality of LIfe

Quality of life improved more in the opioid group than in the placebo group (MD ‐0.73; 95% CI ‐1.10 to ‐0.36; Analysis 1.11;). The quality of evidence for this outcome was downgraded from to low because of attrition bias (summary of findings Table for the main comparison).

Adverse Events

Trenkwalder 2013 (N = 304) reported more adverse events (gastrointestinal problems, fatigue, and headache) in the opioid group than in the placebo group (RR 1.22; 95% CI 1.07 to 1.39; Analysis 1.12; Figure 6). Thirty patients (9.8%) left the study for drug‐related adverse events. The quality of the evidence for this outcome was downgraded to low because of attrition bias (summary of findings Table for the main comparison).


Forest plot of comparison: 1 opioids and placebo, outcome: 1.12 Adverse Events.

Forest plot of comparison: 1 opioids and placebo, outcome: 1.12 Adverse Events.

.

Discussion

Summary of main results

In the Trenkwalder 2013 study, patients in the oxycodone group showed improvements in RLS symptom and quality of life at week 12. Adverse events were significantly more common in the oxycodone group. The symptoms score mean difference was seven points less (CI: ‐9.69 to ‐4.31) in the IRLSSS, and one point less (CI: ‐1.49 to ‐0.73) in the CGI in the opioid group after 12 weeks of treatment. Although the quality of life improved only marginally, this finding could be related to the observational period, which may have been too short to provide an accurate perception of that impact. All patients in the opioid group reduced symptoms and improved quality of life and quality of sleep.

Despite the large number of participants who did not complete the planned 12 weeks of treatment, this study provided essential information about the use of opioids to relieve RLS symptoms. Although only 174 patients completed the planned 12 weeks of treatment (100 patients, or 33%, discontinued), analysis was done according to the intention‐to‐treat principle, and all participants who took at least one dose of medication (or placebo) and provide data during the first week of follow‐up were analysed. It is important to state that the reduction of RLS symptoms was significant as early as the first week of treatment, suggesting that this intervention could relieve patients' symptoms quickly. The number of drug responders in the oxycodone/naloxone group was greater than in the placebo group, and there were twice as many drug remitters (patients who became symptoms‐free) in the oxycodone/naloxone group than in the placebo group, resulting in a low effect of the intervention. The adverse events reported by Trenkwalder 2013 low the effect of the intervention, which is an important contribution to our understanding of the impact of opioid treatment on patients. Opioids are classically used to treat pain, and the most important concern about their use is the potential for abuse and dependency, although it might not be an issue for RLS patients (Aurora 2012; Silver 2011; Walters 2001). Opioids are associated with constipation, and have the potential to worsen or trigger central sleep apnoea (Randerath 2012). In this regard, it should be noted that only one patient had withdrawal symptoms after 12 weeks in the Trenkwalder 2013 study.

Overall completeness and applicability of evidence

This review included only one study (Trenkwalder 2013; N = 304), in which patients for whom previous treatment had failed were given oxycodone/naloxone twice daily at a mean dose of 20 mg oxycodone and 10 mg naloxone. The intervention was effective in relieving symptoms and improving quality of life, which are the most important outcomes when treating patients with RLS, measured according to the Clinical Global Impression scale and RLS Quality of Life questionnaire (RLS‐QoL). Although patients reduced RLS symptoms score at the end of the treatment compared to the placebo group, the authors did not report the minimum clinically important change to be expected for the IRLSSS. Trenkwalder 2013 did not address polysomnographic variables such as total sleep time or periodic leg movement in sleep. The authors carefully screened patients, excluding those with secondary RLS, those taking dopamine receptor blockers, those with co‐morbidities (apnoea syndrome, narcolepsy, myoclonus epilepsy, hallucinations or psychotic episodes, acute clinical augmentation, clinically evident respiratory disorders, clinically relevant constipation, or ileus), previous treatment with naloxone or naltrexone within 30 days of entry, contraindications or hypersensitivity to oxycodone, naloxone, related products, or other ingredients, drugs that potentially affected sleep architecture or motor manifestations during sleep, CNS depressants, current alcohol or drug misuse (including opioids), taking an investigational drug within 30 days of study entry, or taking monoamine oxidase inhibitors within two weeks of screening. Shift workers and patients with serum ferritin below 30 μg/L at screening were also excluded. Pregnant women were also not included, although this was not stated in the published text. All these restriction reflect the current caution towards the use of opioids to treat RLS symptoms, mostly derived from the medical experience of using opioids in the treatment of pain‐related syndromes. On the other hand, physicians should bear in mind that the external validity of oxycodone/naloxone treatment for RLS could be an issue in this review.

Quality of the evidence

The conclusions of this review are based on a single study of low evidence quality (summary of findings Table for the main comparison, Trenkwalder 2013), which was well‐designed and performed overall. This study was downgraded to a low evidence quality because of high attrition bias. As stated above, this study had a large number of dropouts before the planned 12 weeks of treatment, and thus may have underestimated the occurrence of adverse events (attrition bias), including potentially serious adverse events; this, however, may have been minimized by the use of intention‐to‐treat analysis (ITT). External validity is also an issue in this review, since the included study addressed patients with moderate or severe RLS in whom previous treatment had failed. Although the ITT analysis favours their conclusions about the benefits of the intervention, the high attrition bias precludes prediction of the entire range of possible adverse events for the oxycodone/naloxone treatment, which is an important issue when we are using opioids.

Potential biases in the review process

It is always possible that some clinical trials were performed and not published, published locally (e.g., as theses or dissertations), or in languages not included in the search strategy. Another potential source of bias was that we could not include three cross‐over studies, since the primary data required for standard analysis were not available (Allen 1992; Kaplan 1993; Walters 1993).

Agreements and disagreements with other studies or reviews

This systematic review suggests that opioids seem to be effective in the treatment of RLS. This finding is supported by one, multicentre trial (Trenkwalder 2013; N = 304), confirming what has been suggested by previous small, randomised, double‐blind, cross‐over studies (Allen 1992; Kaplan 1993; Walters 1993). Compared to this study, other available systematic reviews do not add further information, since they were based on those three small trials that we excluded from this review for methodological reasons (Hornyak 2014; Wilt 2012; Wilt 2013.

Diagram about search, excluded and included researches.
Figuras y tablas -
Figure 1

Diagram about search, excluded and included researches.

Risk of bias summary: review authors' judgements about each risk of bias criterion for each included study.
Figuras y tablas -
Figure 2

Risk of bias summary: review authors' judgements about each risk of bias criterion for each included study.

Forest plot of comparison: 1 opioids and placebo, outcome: 1.1 RLS symptoms ‐ IRLSSS.
Figuras y tablas -
Figure 3

Forest plot of comparison: 1 opioids and placebo, outcome: 1.1 RLS symptoms ‐ IRLSSS.

Forest plot of comparison: 1 opioids and placebo, outcome: 1.3 Drug responders ‐ IRLSSS.
Figuras y tablas -
Figure 4

Forest plot of comparison: 1 opioids and placebo, outcome: 1.3 Drug responders ‐ IRLSSS.

Forest plot of comparison: 1 opioids and placebo, outcome: 1.5 Remitters ‐ IRLSSS.
Figuras y tablas -
Figure 5

Forest plot of comparison: 1 opioids and placebo, outcome: 1.5 Remitters ‐ IRLSSS.

Forest plot of comparison: 1 opioids and placebo, outcome: 1.12 Adverse Events.
Figuras y tablas -
Figure 6

Forest plot of comparison: 1 opioids and placebo, outcome: 1.12 Adverse Events.

Comparison 1 opioids and placebo, Outcome 1 RLS symptoms ‐ IRLSSS.
Figuras y tablas -
Analysis 1.1

Comparison 1 opioids and placebo, Outcome 1 RLS symptoms ‐ IRLSSS.

Comparison 1 opioids and placebo, Outcome 2 RLS symptoms ‐ CGI.
Figuras y tablas -
Analysis 1.2

Comparison 1 opioids and placebo, Outcome 2 RLS symptoms ‐ CGI.

Comparison 1 opioids and placebo, Outcome 3 Drug responders ‐ IRLSSS.
Figuras y tablas -
Analysis 1.3

Comparison 1 opioids and placebo, Outcome 3 Drug responders ‐ IRLSSS.

Comparison 1 opioids and placebo, Outcome 4 Drug responders ‐ CGI.
Figuras y tablas -
Analysis 1.4

Comparison 1 opioids and placebo, Outcome 4 Drug responders ‐ CGI.

Comparison 1 opioids and placebo, Outcome 5 Remitters ‐ IRLSSS.
Figuras y tablas -
Analysis 1.5

Comparison 1 opioids and placebo, Outcome 5 Remitters ‐ IRLSSS.

Comparison 1 opioids and placebo, Outcome 6 Sleep adequacy ‐ MOS.
Figuras y tablas -
Analysis 1.6

Comparison 1 opioids and placebo, Outcome 6 Sleep adequacy ‐ MOS.

Comparison 1 opioids and placebo, Outcome 7 Sleep quantity ‐ MOS.
Figuras y tablas -
Analysis 1.7

Comparison 1 opioids and placebo, Outcome 7 Sleep quantity ‐ MOS.

Comparison 1 opioids and placebo, Outcome 8 Daytime somnolence ‐ MOS.
Figuras y tablas -
Analysis 1.8

Comparison 1 opioids and placebo, Outcome 8 Daytime somnolence ‐ MOS.

Comparison 1 opioids and placebo, Outcome 9 Trouble staying awake during the day ‐ MOS.
Figuras y tablas -
Analysis 1.9

Comparison 1 opioids and placebo, Outcome 9 Trouble staying awake during the day ‐ MOS.

Comparison 1 opioids and placebo, Outcome 10 Naps during the day ‐ MOS.
Figuras y tablas -
Analysis 1.10

Comparison 1 opioids and placebo, Outcome 10 Naps during the day ‐ MOS.

Comparison 1 opioids and placebo, Outcome 11 Quality of life ‐ RLS QoL.
Figuras y tablas -
Analysis 1.11

Comparison 1 opioids and placebo, Outcome 11 Quality of life ‐ RLS QoL.

Comparison 1 opioids and placebo, Outcome 12 Adverse Events.
Figuras y tablas -
Analysis 1.12

Comparison 1 opioids and placebo, Outcome 12 Adverse Events.

Summary of findings for the main comparison. Opioids compared to placebo for RLS

Opioid treatment compared to placebo for patients with RLS

Patient or population: RLS
Setting: 55 hospitals and specialised private neurology practices in Austria, Germany, Spain, and Sweden.
Intervention: opioids
Comparison: placebo

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Quality of the evidence
(GRADE)

Comments

Risk with placebo

Risk with opioids

RLS symptoms
assessed with: IRLSSS
Scale from: 0 to 40
follow up: mean 12 weeks

The mean RLS symptoms was 22.1 points

The mean RLS symptoms in the intervention group was 7 points lower (9,69 lower to 4,31 lower)

270
(1 RCT)

⊕⊕⊝⊝
LOW 1

Moderate effect size of difference in mean response (0.57) 2

RLS symptoms
assessed with: CGI
Scale from: 0 to 7
follow up: mean 12 weeks

The mean RLS symptoms was 4.1 points

The mean RLS symptoms in the intervention group was 1,11 points lower (1,49 lower to 0,73 lower)

276
(1 RCT)

⊕⊕⊝⊝
LOW 1

Moderate effect size of difference in mean response (0.64) 2

Drug responders
assessed with: IRLSSS ‐ Reduced score at least 50%
follow up: mean 12 days

Study population

RR 1.82
(1.37 to 2.42)

276
(1 RCT)

⊕⊕⊝⊝
LOW 1

313 per 1.000

569 per 1.000
(428 to 756)

Drug responders
assessed with: CGI ‐ Self reported "Much improved" or "Very much improved"
follow up: mean 12 weeks

Study population

RR 1.92
(1.49 to 2.48)

276
(1 RCT)

⊕⊕⊝⊝
LOW 1

347 per 1.000

667 per 1.000
(517 to 861)

Remitters
assessed with: IRLSSS ‐ Scored 10 or less
follow up: mean 12 weeks

Study population

RR 2.14
(1.45 to 3.16)

276
(1 RCT)

⊕⊕⊝⊝
LOW 1

194 per 1.000

416 per 1.000
(282 to 614)

Adverse Events
assessed with: clinical assessment
follow up: mean 12 weeks

Study population

RR 1.22
(1.07 to 1.39)

304
(1 RCT)

⊕⊕⊝⊝
LOW 1

688 per 1.000

840 per 1.000
(736 to 957)

Quality of life
assessed with: RLS Quality of Life Questionnaire (RLS‐QoL)
Scale from: 0 to 7
follow up: mean 12 weeks

The mean quality of life was 3.64 points

The mean quality of life in the intervention group was 0,73 points lower (1,1 lower to 0,36 lower)

276
(1 RCT)

⊕⊕⊝⊝
LOW 1

Small effect size of difference in mean response (0.43) 2

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; RR: Risk ratio; OR: Odds ratio;

GRADE Working Group grades of evidence
High quality: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate quality: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low quality: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low quality: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

1 The only trial included presentes high risk of attrition bias.

2 0.2 represents a small effect, 0.5 a moderate effect, and 0.8 a large effect of difference in mean response (Cohen 1988).

Figuras y tablas -
Summary of findings for the main comparison. Opioids compared to placebo for RLS
Comparison 1. opioids and placebo

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 RLS symptoms ‐ IRLSSS Show forest plot

1

276

Mean Difference (IV, Random, 95% CI)

‐7.00 [‐9.69, ‐4.31]

2 RLS symptoms ‐ CGI Show forest plot

1

276

Mean Difference (IV, Fixed, 95% CI)

‐1.11 [‐1.49, ‐0.73]

3 Drug responders ‐ IRLSSS Show forest plot

1

276

Risk Ratio (M‐H, Fixed, 95% CI)

1.82 [1.37, 2.42]

4 Drug responders ‐ CGI Show forest plot

1

276

Risk Ratio (M‐H, Fixed, 95% CI)

1.92 [1.49, 2.48]

5 Remitters ‐ IRLSSS Show forest plot

1

276

Risk Ratio (M‐H, Fixed, 95% CI)

2.14 [1.45, 3.16]

6 Sleep adequacy ‐ MOS Show forest plot

1

276

Mean Difference (IV, Fixed, 95% CI)

‐0.74 [‐1.15, ‐0.33]

7 Sleep quantity ‐ MOS Show forest plot

1

276

Mean Difference (IV, Fixed, 95% CI)

0.89 [0.52, 1.26]

8 Daytime somnolence ‐ MOS Show forest plot

1

276

Mean Difference (IV, Fixed, 95% CI)

0.21 [‐0.13, 0.55]

9 Trouble staying awake during the day ‐ MOS Show forest plot

1

276

Mean Difference (IV, Fixed, 95% CI)

0.13 [‐0.19, 0.45]

10 Naps during the day ‐ MOS Show forest plot

1

276

Mean Difference (IV, Fixed, 95% CI)

0.01 [‐0.32, 0.34]

11 Quality of life ‐ RLS QoL Show forest plot

1

276

Mean Difference (IV, Fixed, 95% CI)

‐0.73 [‐1.10, ‐0.36]

12 Adverse Events Show forest plot

1

304

Risk Ratio (M‐H, Fixed, 95% CI)

1.22 [1.07, 1.39]

Figuras y tablas -
Comparison 1. opioids and placebo