Entrenamiento con ejercicios para la bronquiectasia
Resumen
Antecedentes
La bronquiectasia se caracteriza por una producción excesiva de esputo, tos crónica y exacerbaciones agudas, y se asocia con síntomas de disnea y cansancio, que reducen la tolerancia al ejercicio y deterioran la calidad de vida. El entrenamiento con ejercicios, de forma aislada o junto con otras intervenciones, es beneficioso para las personas con otras enfermedades respiratorias, pero sus efectos en la bronquiectasia no han sido bien definidos.
Objetivos
Determinar los efectos del entrenamiento con ejercicios en comparación con la atención habitual sobre la tolerancia al ejercicio (desenlace principal), la calidad de vida (desenlace principal), la incidencia de exacerbación aguda y hospitalización, los síntomas respiratorios y de salud mental, la función física, la mortalidad y los eventos adversos en personas con exacerbación estable o aguda de bronquiectasia.
Métodos de búsqueda
Se identificaron ensayos del Registro especializado del Grupo Cochrane de Vías respiratorias (Cochrane Airways), ClinicalTrials.gov y el portal de ensayos de la Organización Mundial de la Salud, desde su creación hasta octubre de 2020. Se revisaron los resúmenes de congresos relacionados con temas respiratorios y las listas de referencias de todos los estudios primarios y artículos de revisión en busca de referencias adicionales.
Criterios de selección
Se incluyeron los ensayos controlados aleatorizados en los que el entrenamiento con ejercicios de al menos cuatro semanas de duración (u ocho sesiones) se comparó con la atención habitual en personas con bronquiectasia estable o que presentan una exacerbación aguda. Se permitieron las cointervenciones con el entrenamiento con ejercicios que incluyeron la capacitación, el entrenamiento de los músculos respiratorios y el tratamiento de depuración de las vías respiratorias, si también se aplicaban como parte de la atención habitual.
Obtención y análisis de los datos
Dos autores de la revisión, de forma independiente, examinaron y seleccionaron los ensayos para inclusión, extrajeron los datos de los desenlaces y evaluaron el riesgo de sesgo. Se estableció contacto con los autores de los estudios para obtener los datos faltantes. Se calcularon las diferencias de medias (DM) con un modelo de efectos aleatorios. Se utilizó el método GRADE para evaluar la certeza de la evidencia.
Resultados principales
Se incluyeron seis estudios, dos de ellos publicados como resúmenes, con un total de 275 participantes. Cinco estudios se realizaron con personas con bronquiectasia clínicamente estable y un estudio piloto se realizó después de una exacerbación aguda. Todos los estudios incluyeron cointervenciones como instrucciones para el tratamiento de depuración de las vías respiratorias y estrategias de respiración, entrega de un folleto informativo y realización de sesiones educativas. La duración del entrenamiento varió entre seis y ocho semanas, con una mezcla de sesiones supervisadas y no supervisadas realizadas en ámbitos ambulatorios o domiciliarios. En la revisión no se incluyeron estudios con niños; sin embargo, se identificaron dos estudios actualmente en curso. No se dispone de datos sobre los niveles de actividad física o los eventos adversos.
En las personas con bronquiectasia estable, la evidencia indica que el entrenamiento con ejercicios comparado con la atención habitual mejora la tolerancia al ejercicio funcional medida por la distancia incremental de caminata en cinta rodante, con una diferencia de medias (DM) entre los grupos de 87 metros (intervalo de confianza [IC] del 95%: 43 a 132 metros; cuatro estudios, 161 participantes; evidencia de certeza baja). La evidencia también indica que el entrenamiento con ejercicios mejora la distancia de caminata en seis minutos (6MWD) (DM entre los grupos de 42 metros; IC del 95%: 22 a 62; un estudio, 76 participantes; evidencia de certeza baja). La magnitud de estos cambios medios observados parece ser clínicamente relevante, ya que superan los umbrales de diferencia mínima clínicamente importante (DMCI) en las personas con enfermedad pulmonar crónica. La evidencia indica que la calidad de vida mejora después del entrenamiento con ejercicios según la puntuación total del St George's Respiratory Questionnaire (SGRQ) (DM ‐9,62 puntos; IC del 95%: ‐15,67 a ‐3,56 puntos; tres estudios, 160 participantes; evidencia de certeza baja), que supera la DMCI de 4 puntos para este desenlace. Se observó una reducción de la disnea (DM 1,0 puntos; IC del 95%: 0,47 a 1,53; un estudio, 76 participantes) y del cansancio (DM 1,51 puntos; IC del 95%: 0,80 a 2,22 puntos; un estudio, 76 participantes) tras el entrenamiento con ejercicios según estos dominios del Chronic Respiratory Disease Questionnaire. Sin embargo, no hubo cambios en la calidad de vida relacionada con la tos, medida por el Leicester Cough Questionnaire (LCQ) (DM ‐0,09 puntos; IC del 95%: ‐0,98 a 0,80 puntos; dos estudios, 103 participantes; evidencia de certeza moderada), ni en la ansiedad o la depresión. Dos estudios informaron desenlaces a más largo plazo hasta 12 meses después de la finalización de la intervención; sin embargo, el entrenamiento con ejercicios no pareció mejorar la capacidad de ejercicio ni la calidad de vida más que la atención habitual. El entrenamiento con ejercicios redujo el número de exacerbaciones agudas de bronquiectasia durante 12 meses en las personas con bronquiectasia estable (odds ratio 0,26; IC del 95%: 0,08 a 0,81; un estudio, 55 participantes).
Después de una exacerbación aguda de bronquiectasia, los datos de un estudio individual (n = 27) indican que el entrenamiento con ejercicios comparado con la atención habitual da lugar a poco o ningún efecto sobre la capacidad de ejercicio (DM 11 metros; IC del 95%: ‐27 a 49 metros; evidencia de certeza baja), la puntuación total del SGRQ (DM 6,34 puntos; IC del 95%: ‐17,08 a 29,76 puntos) o la puntuación del LCQ (DM ‐0,08 puntos; IC del 95%: ‐0,94 a 0,78 puntos; evidencia de certeza baja) y no reduce el tiempo hasta la primera exacerbación (cociente de riesgos instantáneos 0,83; IC del 95%: 0,31 a 2,22).
Conclusiones de los autores
Esta revisión proporciona evidencia de certeza baja que sugiere una mejoría en la capacidad de ejercicio funcional y en la calidad de vida inmediatamente después del entrenamiento con ejercicios en personas con bronquiectasia estable; sin embargo, los efectos del entrenamiento con ejercicios sobre la calidad de vida relacionada con la tos y los síntomas psicológicos parecen ser mínimos. Debido a la información insuficiente de los métodos, el escaso número de estudios y la variación entre los resultados de los estudios, la evidencia es de certeza muy baja a moderada. Se dispone de evidencia limitada para mostrar los efectos a largo plazo del entrenamiento con ejercicios sobre estos desenlaces.
PICO
Resumen en términos sencillos
Entrenamiento con ejercicios para la bronquiectasia
Pregunta de la revisión
Se quería saber si el entrenamiento con ejercicios mejora la tolerancia al ejercicio, la calidad de vida o los síntomas, y si reduce el número de brotes ("exacerbaciones") futuros, en las personas con bronquiectasia en comparación con las personas que no hicieron entrenamiento con ejercicios. Se analizaron los estudios que incluían a personas con enfermedad estable y a pacientes en el período posterior a un brote reciente. El objetivo era incluir evidencia relacionada con niños y adultos con bronquiectasia.
Antecedentes
Las personas con bronquiectasia presentan tos crónica y producción de esputo. Tienen un mayor riesgo de desarrollar exacerbaciones agudas que contribuyen a una tolerancia al ejercicio y una calidad de vida deficientes. Cuando lo realizan personas con otras afecciones pulmonares crónicas, el entrenamiento con ejercicios mejora la tolerancia al ejercicio y reduce los síntomas. Sin embargo, se sabe poco sobre el efecto del entrenamiento con ejercicios específicamente en la bronquiectasia.
Características de los estudios
La evidencia está actualizada hasta octubre de 2020. Se incluyeron seis estudios con un total de 275 participantes; cinco estudios estaban relacionados con personas con enfermedad estable. No se encontraron estudios con niños. El entrenamiento con ejercicios se realizó en combinación con otros tratamientos como el tratamiento de depuración de las vías respiratorias, el entrenamiento de los músculos respiratorios y la formación. Los participantes se asignaron de forma aleatoria a entrenamiento con ejercicios o ningún entrenamiento con ejercicios. El entrenamiento con ejercicios se realizó por al menos seis semanas, ya fuera en un ámbito de grupo o en el domicilio. Ninguno de los estudios incluidos fue financiado por compañías con intereses comerciales en los resultados del estudio.
Resultados clave
Después de la finalización del entrenamiento con ejercicios, los participantes en un estado clínico estable caminaron más lejos que los que no realizaron el entrenamiento con ejercicios (una media de 87 metros más lejos), pero la certeza de la evidencia es baja. Los participantes también informaron una mejor calidad de vida (evidencia de certeza baja) y menor dificultad para respirar y cansancio. Se encontró evidencia de certeza moderada que muestra que el entrenamiento con ejercicios podría no mejorar específicamente los síntomas relacionados con la tos, aunque la incidencia de exacerbaciones agudas fue menor. La evidencia no fue suficiente para mostrar si los efectos del entrenamiento con ejercicios durarían más allá del período de entrenamiento con ejercicios, y no hubo evidencia disponible para determinar si el entrenamiento con ejercicios ayuda a las personas a ser físicamente activas. No se observaron efectos beneficiosos en las personas que realizaron el entrenamiento con ejercicios poco después de un brote agudo de bronquiectasia.
Certeza de la evidencia
La certeza de la evidencia fue muy baja a moderada debido a la incertidumbre sobre la verdadera magnitud de los efectos beneficiosos observados, a los estudios mal realizados y a la falta general de datos suficientes. Se necesitan más estudios con un mayor número de participantes para determinar los efectos a largo plazo del entrenamiento con ejercicios, independientemente del estado clínico.
Authors' conclusions
Summary of findings
| Exercise training compared to control in people with stable bronchiectasis | |||||||
| Patient or population: people with stable bronchiectasis | |||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | №. of participants | Certainty of the evidence | Comments | ||
|---|---|---|---|---|---|---|---|
| Risk with usual care | Risk with exercise training | ||||||
| Change in incremental shuttle walk distance | Follow‐up: range 6 weeks to 8 weeks | Mean change in incremental shuttle walk distance ranged from ‐70 to 2.5 metres | MD 87 metres higher | ‐ | 161 | ⊕⊕⊝⊝ | |
| Follow‐up: range 3 months to 12 months | Mean change in incremental shuttle walk distance ranged from 5.2 to 343 metres | MD 6.19 metres higher | ‐ | 82 | ⊕⊕⊝⊝ | ||
| Change in 6‐minute walk distance | Follow‐up: mean 8 weeks | Mean change in 6‐minute walk distance was ‐10.9 metres | MD 42 metres higher | ‐ | 76 | ⊕⊕⊝⊝ | |
| Follow‐up: mean 12 months | Mean change in 6‐minute walk distance was ‐8.26 metres | MD 6.74 metres lower | ‐ | 55 | ⊕⊕⊝⊝ LOWc,e | ||
| Change in endurance shuttle walk time | Follow‐up: mean 8 weeks | Mean change in endurance shuttle walk time was 0.2 minutes | MD 5.4 minutes higher | ‐ | 39 | ⊕⊕⊝⊝ | |
| Change in endurance shuttle walk distance | Follow‐up: mean 8 weeks | Mean change in endurance shuttle walk distance was ‐36.4 metres | MD 311.6 metres higher | ‐ | 27 | ⊕⊝⊝⊝ | |
| Follow‐up: mean 3 months | Mean change in endurance shuttle walking distance was 964.3 metres | MD 385.7 metres higher | ‐ | 27 | ⊕⊝⊝⊝ | ||
| Change in walking distance during endurance walking test assessed with constant load endurance test | Follow‐up: mean 8 weeks | Mean change in walking distance during endurance walking test was ‐112.6 metres | MD 505.4 metres higher | ‐ | 19 | ⊕⊝⊝⊝ | |
| Change in peak oxygen uptake | Follow‐up: mean 8 weeks | Mean change in peak oxygen uptake was ‐1.91 L/min | MD 3.87 L/min higher | ‐ | 19 | ⊕⊝⊝⊝ | |
| Change in quality of life | Follow‐up: range 6 weeks to 8 weeks | Mean change in quality of life ranged from 4 to 39.2 points | MD 9.62 points lower | ‐ | 110 | ⊕⊕⊝⊝ | Lower scores post intervention are favourable, indicating improvement in quality of life |
| Follow‐up: range 3 months to 12 months | Mean change in quality of life ranged from 33.6 to 45.2 points | MD 6.78 points fewer | ‐ | 65 | ⊕⊕⊝⊝ | Lower scores at follow‐up are favourable, indicating improvement in quality of life | |
| Change in cough‐related quality of life | Follow‐up: range 6 weeks to 8 weeks | Mean change in cough‐related quality of life ranged from 14.6 to 15.5 points | MD 0.09 points lower | ‐ | 103 | ⊕⊕⊕⊝ | Higher scores post intervention are favourable, indicating improvement in quality of life |
| Follow‐up: range 3 months to 12 months | Mean change in cough‐related quality of life ranged from 13.6 to 17.9 points | MD 0.97 points lower | ‐ | 82 | ⊕⊝⊝⊝ | Higher scores at follow‐up are favourable, indicating improvement in quality of life | |
| *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). | |||||||
| GRADE Working Group grades of evidence. | |||||||
| aWide confidence interval limits exceed (or likely exceed) a clinically relevant threshold for this patient group (imprecision ‐1). bAllocation bias for 3 studies was unclear, and only 1 study included blinded assessors (bias ‐1). cLimited data were available for meta‐analysis, reducing its broad representativeness of important factors (e.g. disease severity, settings) (indirectness ‐1). dAllocation bias in 1 study was unclear (bias ‐1). eConfidence interval limits exceed minimally important difference for this outcome in people with bronchiectasis (imprecision ‐1). fBlinding of assessors was not reported and attrition or reporting bias was unclear (bias ‐1). gAllocation bias and attrition and reporting bias for 1 study were unclear (bias ‐1). hEffect estimates contrast between benefit and harm in the two included studies (inconsistency ‐1). | |||||||
| Exercise training compared to control for people in the post‐acute period following acute exacerbation of bronchiectasis | ||||||
| Patient or population: people in the post‐acute period following acute exacerbation of bronchiectasis | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | №. of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with usual care | Risk with exercise training | |||||
| Change in 6‐minute walk distance | Mean change in 6‐minute walk distance was 15 metres | MD 11 metres higher | ‐ | 27 | ⊕⊕⊝⊝ | |
| Change in cough‐related quality of life assessed with Leicester Cough Questionnaire | Mean change in cough‐related quality of life was 0.91 units | MD 0.08 units lower | ‐ | 27 | ⊕⊕⊝⊝ | Higher scores post intervention are favourable, indicating improvement in quality of life |
| *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). | ||||||
| GRADE Working Group grades of evidence. | ||||||
| aRandom sequence generation was unclear and outcome data were incomplete in the sole included study (bias ‐1). bLimited data were available for meta‐analysis, reducing its broad representativeness of important factors (e.g. disease severity, settings) (indirectness ‐1). | ||||||
Background
Description of the condition
Bronchiectasis is a chronic and progressive respiratory condition characterised clinically by chronic cough, sputum production, and bronchial infection, and radiologically by abnormal and permanent dilation of the bronchial lumen (Polverino 2017). Peripheral muscle dysfunction is a common feature of the disease associated with muscle weakness, reduced endurance, and high levels of fatigue and dyspnoea (De Carmargo 2018; Inal‐Ince 2014; Ozalp 2012). Difficulty with activities of daily living, as reflected by functional measures of dyspnoea such as the Medical Research Council Dyspnoea Scale, has been reported in people with bronchiectasis (Martinez‐Garcia 2007). Although the causes of dyspnoea in bronchiectasis are multi‐factorial, key factors include altered respiratory mechanics and insufficient gas exchange (Ozalp 2012). Expiratory airflow limitation has been identified in people with moderate to severe bronchiectasis, with commonly recognised functional abnormalities such as air trapping (Radovanovic 2018), along with a corresponding increase in dynamic hyperinflation and heightened levels of dyspnoea (Koulouris 2003; Ozalp 2012). Respiratory muscle weakness has been reported in people with bronchiectasis compared to age‐matched healthy controls (Liaw 2011; Moran 2010; Newall 2005), as has reduced exercise tolerance (Koulouris 2003; Ozalp 2012). Children with bronchiectasis have demonstrated reduced maximal exercise capacity (Swaminathan 2003). Reduced quadriceps strength is common (Ozalp 2012), and fatigue has been reported in 27% to 74% of people with bronchiectasis (Hester 2011; King 2005; King 2006). People with bronchiectasis have been shown to be highly physically inactive, with lower proportions of physical activity undertaken each day compared to healthy controls (Bradley 2015; De Carmargo 2018; Gale 2012).
Although the aetiology of bronchiectasis is heterogeneous and includes severe infection, immune deficiency, autoimmune disorder, and ciliary disorder (Chalmers 2015), a proportion of cases of adults with bronchiectasis are classified as idiopathic (Araugo 2017; Kelly 2003). In children, common aetiologies are immunodeficiency, aspiration, and primary ciliary dyskinesia (Li 2005). Despite the unclear global prevalence of bronchiectasis, various reports provide an estimate according to country. In the USA, between 139 and 1106 cases per 100,000 population have been reported from data collected between 2000 and 2013 (Seitz 2012; Weycker 2017). In the UK, prevalence in 2013 was approximately 566 per 100,000 females and 485 per 100,000 males with a diagnosis of bronchiectasis (Quint 2016), with incidence increasing with age (Quint 2016). Prevalence was slightly lower in Germany in 2013, with an estimated 67 cases per 100,000, but a higher rate has been reported for people over 75 years of age (Ringshausen 2015). Prevalence of bronchiectasis in children has been reported as 1 in 5800 in northeast England (Eastham 2004), and as 1 in 1700 in New Zealand (Twiss 2005). Among some indigenous populations, prevalence is higher. In Australia, an estimated 1470 per 100,000 indigenous children are diagnosed with bronchiectasis (Chang 2002); in New Zealand, between 4.8 and 7.9 per 100,000 in the Maori population; and among the Pacific Islander population, between 17.8 and 18.3 per 100,000 (Edwards 2003). Bronchiectasis is associated with significant mortality, accounting for between 1438 and 1914 deaths per 100,000 people with bronchiectasis in the UK (Quint 2016). In Belgium, over a five‐year follow‐up period, the mortality rate was 20.4% (Goeminne 2014). With bronchiectasis characterised by recurrent acute exacerbations, the rate of hospitalisation is ever increasing, particularly among the older population (Ringshausen 2013; Seitz 2010; Seitz 2012). Acute exacerbations, peripheral and respiratory muscle dysfunction, and respiratory and psychological symptoms of anxiety and depression contribute to reductions in health‐related quality of life (HRQoL) (Giron Moreno 2013; O'Leary 2002; Olveira 2013), as observed in people with bronchiectasis (Chalmers 2018; Martinez‐Garcia 2005; Pifferi 2010).
Description of the intervention
International and national guidelines for managing bronchiectasis have highlighted the importance of minimising inflammation and infection, optimising airway clearance, and addressing structural lung disease (Al‐Jahdali 2017; Chang 2015; Hill 2019; Martinez‐Garcia 2018; Pasteur 2010; Pereira 2019; Polverino 2017). Several interventions are applied to achieve optimal management of bronchiectasis, including antibiotics, anti‐inflammatory agents, mucolytics, airway clearance therapy, and exercise training. Exercise training refers to structured programmes of activities that involve physical exertion and skeletal muscle contractions targeting improvements in physical function or exercise tolerance (or both). Exercise training may be undertaken in isolation or as part of a pulmonary rehabilitation programme. Pulmonary rehabilitation has been defined as "comprehensive intervention based on a thorough patient assessment followed by patient‐tailored therapies that include, but are not limited to, exercise training, education, and behavioural change, designed to improve the physical and psychological conditions of people with chronic respiratory disease and to promote the long‐term adherence to health‐enhancing behaviours" (Spruit 2013). It is well recognised that exercise training is a critical component of pulmonary rehabilitation; this may be complemented by formal educational sessions focusing on self‐management, behavioural modification, and counselling (Spruit 2013). Regardless of the circumstances in which exercise training is provided for people with bronchiectasis, any individually tailored exercise training programme prescribed for people with bronchiectasis may consist of lower and upper limb endurance exercise (of low or high intensity) and strength training (Spruit 2013). Exercise training may be completed in a hospital, in the community, or in a home‐based environment (Jose 2017; Lee 2008; O'Neill 2002; Spruit 2013); may or may not be undertaken in a group setting; and may be prescribed for a person who is in a stable clinical state or is experiencing an acute exacerbation (Greening 2014). Exercise training may also be completed under the supervision of a suitably trained healthcare professional or conducted unsupervised across any of these settings.
The majority of research in exercise training for people with chronic respiratory conditions has been undertaken in those diagnosed with chronic obstructive pulmonary disease (COPD) (Nici 2006; Spruit 2013). For this patient group, clinically significant improvements in respiratory symptoms, functional ability, exercise tolerance, exacerbation frequency, and HRQoL have been reported (Spruit 2013). As many symptoms are commonly seen in people with the two conditions, it has been postulated that exercise training may offer equivalent effects in people with bronchiectasis (Rochester 2015).
How the intervention might work
The theoretical rationale for performing exercise training in people with bronchiectasis relates to the respiratory and peripheral skeletal muscle manifestations of bronchiectasis. Exercise training targets improvements in physical function or exercise tolerance (or both). These are commonly associated with improvements in respiratory symptoms. Endurance and strength exercise training has been associated with improvement in peripheral muscle strength and aerobic capacity; reduced symptoms of dyspnoea and fatigue; and improved HRQoL in other chronic respiratory conditions such as COPD (McCarthy 2015; Spruit 2013). It is hypothesised that a similar effect may occur in bronchiectasis, although the precise mechanisms are unclear. Despite this, clinical guidelines support the inclusion of people with respiratory conditions other than COPD into rehabilitation programmes (Alison 2017; Spruit 2013). Although comparisons of the effects of pulmonary rehabilitation in people with COPD or bronchiectasis have demonstrated similar improvements in exercise tolerance and health status outcomes (Patel 2019), the longevity of these effects is unclear. Exercise training has not been previously associated with modification of the disease process in chronic respiratory conditions (Spruit 2013), so this is not anticipated to be a likely mechanism of action in bronchiectasis (Mandal 2012). Exercise training may additionally benefit respiratory symptoms such as dyspnoea, chronic cough, and sputum expectoration in people with bronchiectasis due to its effects on breathing patterns and sputum clearance. These have been demonstrated in adults with cystic fibrosis during exercise (Dwyer 2017), and they may be evaluated through measures of symptoms or quality of life domains.
Why it is important to do this review
An international policy statement for pulmonary rehabilitation supported the inclusion of people with bronchiectasis within pulmonary rehabilitation programmes (Rochester 2015). Although an earlier version of this review found that inspiratory muscle training improved endurance exercise capacity and quality of life, evidence for the effects of other types of physical training was not available (Bradley 2002). The 2017 Australian and New Zealand pulmonary rehabilitation guidelines also state that people with bronchiectasis can achieve improvements in exercise capacity and HRQoL following pulmonary rehabilitation compared to usual care (Alison 2017). Surveys of clinical practice have indicated that clinicians prescribe exercise training for people with bronchiectasis or refer people to pulmonary rehabilitation programmes (or both) (Lee 2008; O'Neill 2002). Although review authors previously completed a systematic review and meta‐analysis comparing the effects of pulmonary rehabilitation in bronchiectasis to usual care (Lee 2017), this was isolated to pulmonary rehabilitation and did not include broader definitions of exercise training that may be completed in other environments. With lack of ready access to pulmonary rehabilitation for people with chronic respiratory disease (Rochester 2015), including those with diagnosed bronchiectasis, it is important to consider a broad range of options for exercise training and its effects on clinical parameters compared to usual care, to guide future clinical practice.
Objectives
To determine effects of exercise training compared to usual care on exercise tolerance (primary outcome), quality of life (primary outcome), incidence of acute exacerbations and hospitalisation, respiratory and mental health symptoms, physical function, mortality, and adverse events in people with stable or acute exacerbation of bronchiectasis.
Methods
Criteria for considering studies for this review
Types of studies
We included randomised controlled trials (RCTs) of a parallel‐group design. We included studies reported in full text, published as an abstract only, and provided as unpublished data. Studies published in languages other than English were eligible for inclusion, with translations sought via the Cochrane Airways network. We recorded studies that are included but are lacking available data as 'awaiting classification'.
Types of participants
We included people of any age with a diagnosis of bronchiectasis according to high‐resolution computed tomography (HRCT) or physician diagnosis (Pasteur 2010). Studies comprising patient groups of mixed respiratory pathology must include at least 75% with a primary diagnosis of bronchiectasis or available data on a bronchiectasis subgroup. No participants were excluded on the basis of coexisting respiratory disease (e.g. COPD). However, people with bronchiectasis due to cystic fibrosis were not eligible for inclusion. Participants were eligible for inclusion irrespective of whether they were experiencing an acute exacerbation of their bronchiectasis or were in a period of disease stability.
Types of interventions
We included studies comparing exercise training with usual care. Exercise training was defined as any structured exercise programme that targets improvements in physical function or exercise tolerance (or both). The intervention must have been applied for a minimum duration of four weeks or eight sessions and may have been undertaken as part of an inpatient, outpatient, community, or home‐based programme, in an individual or group setting. Both supervised and unsupervised exercise training interventions were allowed. Co‐interventions such as respiratory muscle training, airway clearance techniques, and patient education were permitted (Chang 2015; Polverino 2017), as these interventions may be integrated into clinical care for people with bronchiectasis and may be associated with clinical benefit (Martin‐Valero 2020). Such co‐interventions must, however, have been provided to both intervention and usual care groups. The effects of exercise training may endure for differing lengths of time depending upon the duration of the initial intervention. Therefore, we distinguished between studies of 12 weeks' duration or less and longer than 12 weeks' duration. In a previous large Cochrane systematic review of pulmonary rehabilitation for people with COPD (McCarthy 2015), 55 of 64 (86%) included studies involved training programmes of 12 weeks' duration or less, hence the use of this threshold as a marker of 'conventional' versus 'long‐term' interventions appeared justified.
Usual care was defined as treatment that did not include a structured physical exercise training programme. Usual care may have included adjunct therapies, such as medical interventions (i.e. antibiotic prescription), a regimen of airway clearance therapy, respiratory muscle training, or a combination of these.
Types of outcome measures
We evaluated the effects of exercise training on the following outcomes.
Primary outcomes
-
Exercise tolerance: measured via field walking tests (e.g. incremental shuttle walk test (ISWT), 6‐minute walk test (6MWT), endurance shuttle walk test (ESWT)) or cardiopulmonary exercise testing (e.g. maximal incremental treadmill or cycle ergometer cardiopulmonary exercise test (CPET), constant‐load exercise test (CLET)). The principal units of analysis for these tests were distance (metres) for ISWT and 6MWT; time (minutes) for endurance or CLET; and peak oxygen uptake (VO₂ peak) for maximal incremental CPET. We reported these outcomes separately. Assessment occurred upon completion of the exercise training intervention and at the longest time point available up to 12 months after intervention completion
-
Health‐related quality of life (HRQoL): measured via disease‐specific questionnaires for bronchiectasis (i.e. Quality of Life‐Bronchiectasis) or respiratory quality of life questionnaires (e.g. St George's Respiratory Questionnaire, Chronic Respiratory Disease Questionnaire (CRDQ), symptom‐specific questionnaires (e.g. Leicester Cough Questionnaire (LCQ)) or generic health questionnaires (e.g. Short Form‐36, EuroQol). Both total scores and symptom‐specific domain scores were used but were reported separately. Data from both disease‐specific and generic instruments were pooled for analysis; however, disease‐specific quality‐of‐life total scores were considered the principal analysis of interest. These were assessed upon completion of the exercise training intervention and at the longest time point available up to 12 months after intervention completion
Secondary outcomes
-
Exacerbations/hospitalisations: measured as incidence, rate, or time to first acute exacerbation or respiratory‐related hospitalisation, with each defined according to study authors. For this outcome, data were sourced from the longest time point available up to 12 months after intervention completion
-
Peripheral skeletal muscle force: may have included measures of muscle strength (kilograms), power (Newtons), or torque (Newton.metres). Data from muscle groups of the upper limb were pooled together, and data from muscle groups of the lower limb were pooled together. Upper limb muscle force was analysed separately from lower limb muscle force. This was assessed upon completion of the exercise training intervention and the longest time point available up to 12 months after intervention completion
-
Physical activity: comprising objectively measured outcomes of movement (e.g. steps, time spent in light/moderate/vigorous activity) but not sedentary behaviour. This was assessed upon completion of the exercise training intervention and the longest time point available up to 12 months after intervention completion
-
Mental health: comprising measures of anxiety and depression (e.g. Hospital Anxiety and Depression Scale (HADS), Beck Depression Inventory, Hamilton Anxiety/Depression Rating Scale). Anxiety data were analysed distinct from depression data. They were assessed upon completion of the exercise training intervention and the longest time point available up to 12 months after intervention completion
-
Clinical symptoms: comprising symptoms such as dyspnoea, cough, or fatigue, with all measures of symptoms eligible for inclusion. Symptoms measured at rest were the principal unit of interest; however, data obtained at the end of exercising were accepted when resting data were unavailable provided the outcome was measured in the same manner for each group within individual trials. These were assessed upon completion of the exercise training intervention and the longest time point available up to 12 months after intervention completion
-
Mortality: measured as the incidence or rate of death, assessed at the longest time point available up to 12 months after intervention completion
-
Adverse events: comprising events such as falls or injury, measured upon completion of the exercise training intervention
Reporting one or more of the outcomes listed here in the study was not an inclusion criterion for the review.
Search methods for identification of studies
Electronic searches
We identified studies from searches of the following databases and trial registries.
-
Cochrane Airways Trials Register (Cochrane Airways 2019), via the Cochrane Register of Studies, all years to 7 October 2020.
-
Cochrane Central Register of Controlled Trials (CENTRAL; 2020, Issue 9), via the Cochrane Register of Studies, all years to 7 October 2020.
-
MEDLINE (Ovid SP) ALL, 1946 to 7 October 2020.
-
Embase (Ovid SP), 1974 to 7 October 2020.
-
US National Institutes of Health Ongoing Trials Register ClinicalTrials.gov (www.clinicaltrials.gov), all years to 7 October 2020.
-
World Health Organization International Clinical Trials Registry Platform (apps.who.int/trialsearch), all years to 9 September 2019.
-
PEDro (Physiotherapy Evidence Database), all years to 7 October 2020.
The database search strategies are listed in Appendix 1. The Cochrane Airways Information Specialist developed the search strategies, in collaboration with the review authors.
All databases and trial registries were searched from their inception to September 2019, with no restriction on language or type of publication. An additional search was then conducted to update the search yield to 7 October 2020. Handsearched conference abstracts and grey literature were identified through the Cochrane Airways Trials Register and the CENTRAL database.
Searching other resources
We checked the reference lists of all primary studies and review articles as well as respiratory conference abstracts for additional references. We contacted authors of identified trials and experts in the field to identify other published or unpublished studies when possible. We searched for errata or retractions from included studies published in full text on PubMed, on 24 August 2020.
Data collection and analysis
Selection of studies
We used Cochrane's Screen4Me workflow to help assess the search results. Screen4Me comprises three components.
-
Known assessments: a service that matches records in the search results to records that have already been screened in Cochrane Crowd (Cochrane's citizen science platform, by which the Crowd help to identify and describe health evidence) and labelled as 'RCT' or 'not an RCT'.
-
RCT classifier: a machine‐learning model that distinguishes RCTs from non‐RCTs.
-
Cochrane Crowd, if appropriate (crowd.cochrane.org).
More detailed information about the Screen4Me components can be found in the following publications: McDonald 2017; Thomas 2017; Marshall 2018; and Noel‐Storr 2018.
Following this, two review authors (AL, CG) screened the titles and abstracts of the search results independently and coded them as 'retrieve' (eligible or potentially eligible/unclear) or 'do not retrieve.' We retrieved the full‐text study reports of all potentially eligible studies, and two review authors (AL, CG) independently screened them for inclusion, recording the reasons for exclusion of ineligible studies. They used Covidence software (Covidence 2018). We resolved any disagreement through discussion, or, if required, we consulted a third person (CO). We identified and excluded duplicates and collated multiple reports of the same study, so that each study, rather than each report, is the unit of interest in the review. We recorded the selection process in sufficient detail to complete a PRISMA flow diagram and Characteristics of excluded studies table (Moher 2009). Any records identified through the search that involve members of the review team were handled by team members who were not involved with the relevant study to avoid perceived conflicts of interest.
Data extraction and management
We used a data collection form that had been piloted on two studies in the review to record study characteristics and outcome data. Two review authors (CG, AL) independently extracted the following study characteristics from included studies.
-
Methods: study design, total duration of study, details of any 'run‐in' period, number of study centres and locations, study setting, withdrawals, and dates of study.
-
Participants: number, mean age, age range, gender, severity of condition, diagnostic criteria, baseline lung function, smoking history, and inclusion criteria and exclusion criteria.
-
Interventions: intervention, comparison, concomitant medications, and excluded medications.
-
Outcomes: primary and secondary outcomes specified and collected, and time points reported.
-
Notes: funding for studies, and notable conflicts of interest of trial authors.
Two review authors (CG, AL) independently extracted outcome data from included studies. We noted in the Characteristics of included studies table if outcome data were not reported in a usable way. We resolved disagreements by reaching consensus or by involving a third person (CO). One review author (CG) transferred data into Review Manager Web (RevMan Web 2019). A second review author (CO) spot‐checked study characteristics for accuracy against the study report. We contacted authors of included studies to verify data extracted from their study when required, and to request details of missing data when applicable.
No study data were extracted or analysed by review members directly involved with included studies.
Assessment of risk of bias in included studies
Two review authors (AL, CG) assessed risk of bias independently for each study using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2019). We resolved disagreements by discussion or by consultation with another review author (CO). We assessed risk of bias according to the following domains.
-
Random sequence generation.
-
Allocation concealment.
-
Blinding of participants and personnel.
-
Blinding of outcome assessment.
-
Incomplete outcome data.
-
Selective outcome reporting.
-
Other bias.
We judged each potential source of bias as having high, low, or unclear risk and provided a quote from the study report together with a justification for our judgement in the 'Risk of bias' table. We summarised risk of bias judgements across different studies for each of the domains listed. We considered blinding separately for outcomes that were self‐reported or were not self‐reported by participants due to the different impact this may have on findings. For example, knowledge of group allocation may influence self‐reported measures of respiratory symptoms but would have plausibly less impact upon an outcome such as all‐cause mortality. For consistency, we considered risk of bias related to blinding for self‐reported outcomes to be uniformly high across all such instances, as is commonly unpreventable in rehabilitation studies. When information on risk of bias related to unpublished data or correspondence with a trial author, we noted this in the 'Risk of bias' table.
When considering treatment effects, we took into account the risk of bias for studies that contributed to that outcome.
We conducted the review according to this published protocol and justified any deviations from it in the Differences between protocol and review section of the systematic review.
Measures of treatment effect
We planned to conduct analyses and to report findings from people younger than 18 years of age separately from people 18 years of age or older. However, no studies of people younger than 18 years met the inclusion criteria.
We conducted analyses and reported findings from studies describing interventions commencing during or within two weeks of discharge from an acute exacerbation or diagnosis of an acute exacerbation separately from those applicable to the stable disease state. We accepted study authors' definitions of acute exacerbations or stable disease state.
We conducted analyses and reported findings from interventions of a 'conventional' (12 weeks or less) duration separate from those of a 'long‐term' (greater than 12 weeks) duration.
We reported findings from outcome data collected at more than one time point (e.g. upon completion of the exercise training intervention, at the longest time point available up to 12 months after intervention completion) separately to avoid issues associated with participant double‐counting.
We analysed dichotomous data as odds ratios (ORs) and continuous data as mean differences (MDs) or standardised mean differences (SMDs) with 95% confidence intervals (CIs). We used SMDs when outcome data were reported via different metrics but were deemed clinically homogenous (e.g. data from different field walking tests or from different quality‐of‐life instruments). We did not use SMDs when such outcome data comprised a combination of both endpoint and change data. When SMDs were used for outcome data expressed as change from baseline, we used the standard deviation (SD) of baseline values as the unit of measurement to calculate the SMD, and adjusted standard errors to take correlation into account, when appropriate data were available. Results from analyses using SMDs were transformed back to native metrics for ease of interpretation. If data from rating scales were combined in a meta‐analysis, we ensured that they were entered with a consistent direction of effect (e.g. lower scores always indicate improvement).
We undertook meta‐analyses only when this was meaningful, that is, when treatments, participants, and the underlying clinical question were similar enough for pooling to make sense.
We described skewed data narratively (e.g. as medians and interquartile ranges for each group).
When multiple trial arms were reported in a single study, we included only the relevant arms. If two comparisons (i.e. exercise training approach one versus usual care and exercise training approach two versus usual care) were combined in the same meta‐analysis, we combined the active arms or halved the control group to avoid double‐counting.
If adjusted analyses were available (ANOVA or ANCOVA), we used these as a preference in our meta‐analyses. If both change from baseline and endpoint scores were available for continuous data, we used change from baseline unless a low correlation between measurements in individuals was reported. If a study reported outcomes at multiple time points, we used the data closest to the primary time point of interest, as defined in the Types of outcome measures section.
We used intention‐to‐treat (ITT) or 'full analysis set' analyses when they were reported (i.e. those in which data had been imputed for participants who were randomly assigned but did not complete the study) instead of completer or per‐protocol analyses.
Unit of analysis issues
For continuous outcomes, we used endpoint, rather than change, data as the principal unit of analysis. Change data were included in pooled meta‐analyses only when endpoint data were not reported, with discussion provided regarding the potential for exaggerated weighting given to such studies.
For dichotomous outcomes, we used the number of people experiencing an event as the unit of analysis (e.g. number of exacerbations). However, if a study reported rate ratios, we analysed them on this basis. We meta‐analysed data from cluster‐RCTs only when available data were adjusted (or could be adjusted), to account for clustering.
Dealing with missing data
We contacted investigators or study sponsors to verify key study characteristics and to obtain missing numerical outcome data when possible (e.g. when a study is identified as an abstract only). When this was not possible, and missing data were thought to introduce serious bias, we took this into consideration in the GRADE rating for affected outcomes.
Assessment of heterogeneity
We used the I² statistic to measure heterogeneity among the studies in each analysis, using a random‐effects or fixed‐effect model, depending on assessment of heterogeneity. When substantial heterogeneity was identified, we reported this but were unable to explore the possible causes by pre‐specified sensitivity analysis due to the absence of subgroups.
Assessment of reporting biases
If we had identified more than 10 studies, we planned to create an example of a funnel plot and to analyse this for small‐study and publication biases (Egger 1997). However, we identified a total of only six studies for this review, which precluded the creation of funnel plots.
Data synthesis
We meta‐analysed data using a random‐effects model, and we performed a sensitivity analysis with a fixed‐effect model.
Subgroup analysis and investigation of heterogeneity
We planned to carry out the following subgroup analyses.
-
Exercise training interventions characterised as unicomponent (e.g. exercise training alone) versus those characterised as multi‐component (e.g. exercise training plus at least one adjunct therapy).
We planned to use the following outcomes in subgroup analyses.
-
Exercise tolerance.
-
Disease‐specific or generic HRQoL (total scores only).
We identified only multi‐component studies for inclusion in this review; therefore there was no analysis of unicomponent interventions and no possibility of comparisons between types of interventions.
Sensitivity analysis
We planned to carry out the following sensitivity analysis while removing the following from the primary outcome analyses.
-
Studies identified as being at high risk of bias for domains other than performance bias, considering blinding of participants and personnel to knowledge of group allocation as inherently challenging in studies of exercise interventions.
We planned to compare results from the principal random‐effects model with those from a fixed‐effect model. However, the small number of studies and the abstract reporting of one study precluded sensitivity analysis. If in future updates, a greater number of studies are included, we will perform a sensitivity analysis.
Summary of findings and assessment of the certainty of the evidence
We constructed 'Summary of findings' tables to present the main findings of this review. We reported primary outcomes and presented anticipated absolute effects and 95% CIs. Review author AL performed an assessment of the certainty of evidence for each outcome using the GRADE approach (Schunemann 2019). Review author CO checked GRADE assessments and 'Summary of findings' tables and revised tables to reflect discussions between AL and CO. Analysis findings were interpreted in line with existing minimally important difference threshold for outcomes for which bronchiectasis‐specific guidance was available.
We planned to create separate 'Summary of findings' tables for each of our separate analyses (paediatric and adult populations; acute and stable disease states; intervention duration up to 12 weeks or lasting longer than 12 weeks). However, due to lack of available data for these comparisons, we could present 'Summary of findings' tables only for acute and stable disease. We planned to generate no tables for the level of the two subgroups (as defined in Subgroup analysis and investigation of heterogeneity). We reported on the following primary outcomes for each table.
-
Exercise tolerance.
-
Disease‐specific or generic HRQoL (total scores only).
We used the five GRADE considerations (risk of bias, inconsistency, imprecision, indirectness, and publication bias) to assess the quality of a body of evidence as it relates to studies that contributed data for the pre‐specified outcomes. We used the methods and recommendations described in Section 8.5 and Chapter 14 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2019), along with GRADEpro software (GRADEpro 2015). We justified any decisions to downgrade the quality of studies by using footnotes, and we made comments to aid the reader's understanding of the review when necessary.
Results
Description of studies
See Characteristics of included studies, Characteristics of excluded studies, and Characteristics of studies awaiting classification for complete details.
Results of the search
The search of databases and clinical trials registries up to 9 October 2020 yielded a total of 1660 records. After removal of 398 duplicates and 525 records via Cochrane Crowd Known Assessments and Screen4Me (Figure 1), a total of 737 records remained. We excluded 715 records on the basis of title and abstract and evaluated 22 records (15 studies) for eligibility via full text. Six studies were excluded, as they did not meet the review criteria. Four citations pertaining to three studies were identified as suitable for inclusion but are currently ongoing; these were not included in the analysis. A total of six studies were included (Figure 2). Full details of excluded and ongoing studies are outlined in the Excluded studies and Ongoing studies sections.
Overview of Cochrane Crowd Known Assessments and Screen4Me workflows for original and updated searches.
Figure 2. Study flow diagram
Included studies
Design
All studies included in this review were parallel‐group, randomised, controlled trials (Chalmers 2019; Dal Corso 2017; Kumar 2017; Lee 2014; Mandal 2012; Newall 2005). Two of the included studies were described as pilot studies (Chalmers 2019; Mandal 2012).
Participants
The six included studies involved 275 participants, with sample sizes ranging from 20 to 85. Five studies included clinically stable adult participants (Dal Corso 2017; Kumar 2017; Lee 2014; Mandal 2012; Newall 2005), and one study included adult participants who undertook the intervention following an acute exacerbation of bronchiectasis (Chalmers 2019). None of the included studies involved children with bronchiectasis; however two of the three ongoing studies involve children (Joschtel 2014 is a pilot study of Trost 2019, but with different participants, aims, and outcome endpoints (confirmed via email correspondence)). Bronchiectasis was diagnosed on the basis of HRCT in five studies (Chalmers 2019; Dal Corso 2017; Lee 2014; Mandal 2012; Newall 2005), and on the basis of physician diagnosis in two studies (Dal Corso 2017; Kumar 2017). The age range of participants was 63 to 72 years in five studies (Chalmers 2019; Dal Corso 2017; Lee 2014; Mandal 2012; Newall 2005); one study did not report participant age (Kumar 2017). Of the five studies including those who undertook the intervention when clinically stable, disease severity according to spirometry ranged from a forced expiratory volume of one second (FEV₁) of 45% to 77% predicted (Dal Corso 2017; Kumar 2017; Lee 2014; Mandal 2012; Newall 2005). The study in which the intervention was undertaken following an acute exacerbation reported participant FEV₁ between 52% and 96% predicted (Chalmers 2019). Acute exacerbations in this study were community‐managed (Chalmers 2019).
Interventions
All studies compared multi‐component exercise training programmes involving adjunct interventions (ranging from instruction or review of airway clearance therapy to a mix of educational sessions) versus no exercise training. Five studies investigated exercise training in an outpatient setting (Chalmers 2019; Kumar 2017; Lee 2014; Mandal 2012; Newall 2005), and one study evaluated a home‐based exercise training programme (Dal Corso 2017). The duration of exercise training programmes varied from six to eight weeks for outpatient rehabilitation and was eight weeks for home‐based training.
Four studies examined the effects of combined aerobic and resistance training (Chalmers 2019; Dal Corso 2017; Lee 2014; Mandal 2012); one study examined the effects of aerobic training only (Newall 2005), and one study did not specify the modality of training applied (Kumar 2017). Five studies added co‐interventions to exercise training, which were also offered to the control group; these included instruction in airway clearance therapy and/or breathing strategies (Chalmers 2019; Lee 2014; Mandal 2012), provision of an educational booklet (Dal Corso 2017), and delivery of educational sessions (Mandal 2012; Newall 2005). One study did not specify which co‐interventions were offered in conjunction with exercise training (Kumar 2017). A summary of key characteristics of interventions is presented in Table 1.
| Study | Setting | Intervention | Control | Duration (weeks) | Frequency | Follow‐up (weeks) |
|---|---|---|---|---|---|---|
| OP Supervised | Aerobic: treadmill, bike, walking (80% VO₂ max) Resistance: UL, LL (8 to 10 reps at 1 RM). Education including instruction in daily chest PT | Guideline concordant ongoing management including instruction in daily chest PT | 6 | 2 OP + 2 HB sessions/week | 12 | |
| HB Unsupervised | Aerobic: 20 minutes stepping on platform (60% to 80% maximal stepping rate in IST) Resistance: 20 minutes UL, LL (8 reps, 3 sets at 70% MVIC). Education. Weekly phone call. Home visit every 15 days | Education Recommended to walk at moderate intensity for 30 minutes 3 times/week. Weekly phone call (no exercise or PA discussion) | 8 | 3 times/week Average session duration: 50 minutes | Nil | |
| OP | No details provided | Standard care | 8 | No details provided | Nil | |
| OP Supervised | Aerobic: 15 minutes treadmill or walking (75% ISWT speed) and bike (60% maximal work rate) Resistance: 15 minutes UL, LL (10 reps, 1 to 3 sets at 10 RM) Encouraged to maintain HEP over follow‐up period via monthly phone call. Reviewed usual ACT. Taught ACBT If no usual ACT | Education Twice‐weekly phone call (no exercise or PA discussion) Reviewed usual ACT. Taught ACBT if no usual ACT | 8 | 2 OP + 3 to 5 HB sessions/week | 52 | |
| OP Supervised | Aerobic: 10 minutes treadmill, bike, ski machine (85% VO₂ max) Resistance: UL, LL (10 reps, 3 sets at 60% 1 RM, progressed to 70% at Week 3, 80% at Week 5). Twice‐daily chest PT. Education. Self‐management plan. Encouraged to undertake gym programme for 6 months at end of intervention | Twice‐daily chest PT. Education. Self‐management plan | 8 | 2 OP + 1 HB session/ week | 12 | |
| OP Supervised | OP aerobic: 15 minutes treadmill, cycle ergometry, stair climbing (80% peak HR on maximal test) HB aerobic: 45 minutes walking at target intensity Education. Sham IMT at sub‐therapeutic load (7 cmH₂O) for 15 minutes twice daily at home | Education | 8 | 2 OP + 1 HB sessions/week Average session duration 45 minutes | 12 |
Abbreviations: ACBT: active cycle of breathing technique; ACT: airway clearance technique; HB: home‐based; HEP: home exercise programme; IMT: inspiratory muscle training; IST: incremental step test; ISWT: incremental shuttle walk test; LL: lower limb; MVIC: maximum voluntary isometric contraction; OP: outpatient; PA: physical activity; PR: pulmonary rehabilitation; PT: physiotherapy; RM: repetition maximum; UL; upper limb; VO₂ max: maximal oxygen uptake.
Outcome measures
All studies reported upon a measure of functional exercise tolerance; most used the six‐minute walk test (Chalmers 2019; Kumar 2017; Lee 2014), or the incremental shuttle walk test (Dal Corso 2017; Lee 2014; Mandal 2012; Newall 2005). Two studies also performed cardiopulmonary exercise testing (Kumar 2017; Newall 2005), two studies applied an endurance shuttle walk test (Dal Corso 2017; Mandal 2012), and one applied a sub‐maximal treadmill‐based exercise test (Newall 2005). Quality of life was assessed by St George's Respiratory Questionnaire (Chalmers 2019; Dal Corso 2017; Kumar 2017; Lee 2014; Mandal 2012; Newall 2005), the Chronic Respiratory Disease Questionnaire (Lee 2014), and the Leicester Cough Questionnaire (Chalmers 2019; Kumar 2017; Lee 2014; Mandal 2012). Health status was measured in one study (Chalmers 2019). Peripheral muscle strength was measured by a dynamometer in one study (Dal Corso 2017), and psychological symptoms of anxiety and depression were measured on the Hospital Anxiety and Depression Scale and the Depression Anxiety Stress Scale in two studies (Kumar 2017; Lee 2014). Time to next exacerbation was measured in two studies (Chalmers 2019; Lee 2014); exacerbation rate over 12 months and mortality were measured in one study only (Lee 2014).
Excluded studies
We excluded six studies after review via full text. Reasons for exclusion related to incorrect study population (investigation of a control group that had a different diagnosis (Choe 1996; Finnerty 1999), incorrect comparator (use of an adjunct co‐intervention that was not applied equally to the usual care arm) (Bradley 2006; Greening 2014), and use of a study design that was not consistent with the inclusion criteria (Kokura 2016; Taylor 2019)).
Risk of bias in included studies
Risk of bias was completed for all six studies. An overview of risk of bias for the domains listed below is outlined in Figure 3 and Figure 4.
Figure 3. Risk of bias summary
Figure 4. Risk of bias graph
Allocation
All studies described participants as randomly allocated to groups; however only four studies described their methods as computer‐generated random sequences (Dal Corso 2017; Lee 2014; Mandal 2012; Newall 2005), resulting in low attributed risk of bias due to sequence generation. Three studies revealed the use of sealed, opaque envelopes as the method of concealment for allocation (Chalmers 2019; Dal Corso 2017; Lee 2014), and three studies did not provide sufficient information to show the method used (Kumar 2017; Mandal 2012; Newall 2005), resulting in unclear ratings of bias due to allocation concealment.
Blinding
The physical nature of exercise training interventions imposes limits to the success of blinding strategies for participants and study personnel. We rated the risk of bias related to performance and detection bias separately for outcomes that were self‐reported or were not self‐reported due to the potentially different impact that participant blinding may have on outcomes such as self‐reported symptoms compared to exacerbations, for example.
All studies included self‐reported outcomes and were consequently rated as having high risk of performance and detection bias for these outcomes due to the risk of participant knowledge of group allocation impacting self‐reported results.
One study stated that participants and personnel were not blinded to group allocation (Dal Corso 2017), and another study inferred that participants and personnel were not blinded to group allocation (Lee 2014). Lee 2014 used an independent, blinded assessor to measure all post‐treatment outcome measures and was therefore rated as having low risk of performance and detection bias for non‐self‐reported outcomes. Dal Corso 2017 stated that due to limited resources, use of a blinded assessor was not possible; this study was therefore rated as having high risk of performance and detection bias for non‐self‐reported outcomes. The remaining four studies did not provide sufficient information regarding blinding of participants and personnel delivering the intervention and undertaking outcome measurement, resulting in unclear risk of bias for these non‐self‐reported outcomes (Chalmers 2019; Kumar 2017; Mandal 2012; Newall 2005). No studies reported whether data analysts were blinded to group allocation.
Incomplete outcome data
Three studies stated withdrawal of participants and reasons for attrition (Lee 2014; Mandal 2012; Newall 2005). One study stated the impact of attrition on the method of statistical analysis applied (Lee 2014), two reported attrition but did not include the reasons or impact on outcomes (Dal Corso 2017; Kumar 2017), and one did not provide sufficient detail on attrition (Chalmers 2019). These latter three studies were rated as possessing uncertain risk of bias related to attrition.
Selective reporting
Four studies were reported prospectively on a clinical trial registry (Chalmers 2019; Dal Corso 2017; Lee 2014; Mandal 2012). Results were reported for all outcomes at each time point for four studies (Chalmers 2019; Lee 2014; Mandal 2012; Newall 2005). Two studies were presented in abstract form only; one had an additional outcome of maximal exercise capacity, for which the findings were not reported (Dal Corso 2017), and one provided insufficient detail to determine selective reporting (Kumar 2017). Both were rated as having uncertain risk for reporting bias.
Other potential sources of bias
Two studies were available in abstract form only (Dal Corso 2017; Kumar 2017), limiting our ability to accurately determine the presence of some sources of bias. It is unclear in one study why the applied method of randomisation resulted in an uneven group allocation (Chalmers 2019). These three studies were consequently rated as having unclear risk of other sources of potential bias. The remaining three studies did not demonstrate any other potential sources of bias (Lee 2014; Mandal 2012; Newall 2005).
Effects of interventions
See: Summary of findings 1 Exercise training compared to control in people with stable bronchiectasis; Summary of findings 2 Exercise training compared to control for people in the post‐acute period following acute exacerbation of bronchiectasis
Refer to the 'Summary of findings' tables for an overview of the main findings related to primary outcome comparisons (summary of findings Table 1; summary of findings Table 2). We were able to include data from six studies in a quantitative and narrative synthesis; all studies were conducted in the adult population (Chalmers 2019; Dal Corso 2017; Kumar 2017; Lee 2014; Mandal 2012; Newall 2005).
Exercise training versus usual care in the stable clinical state
Primary outcomes
Exercise tolerance
Five studies involving 234 participants reported findings related to different metrics of exercise tolerance. Pooled data from four studies involving 161 participants with stable bronchiectasis suggest improvements in incremental shuttle walk distance (ISWD) immediately following exercise training (mean difference (MD) in change from baseline 87.12 metres, 95% confidence interval (CI) 42.65 to 131.58 m; I² = 64%; n = 161; low‐certainty evidence; Analysis 1.1; Figure 5). One study did not demonstrate meaningful between‐group differences following training completion (Mandal 2012), but investigators noted large within‐group changes in the intervention group only (which started with a ‐55.8 m mean lower baseline ISWT level). The meta‐analysis for this outcome was considerably influenced by Lee 2014 (weighting 36.1%), which was the only included study to contribute change from baseline rather than endpoint data. Exploratory step‐by‐step removal of studies from the meta‐analysis revealed that Dal Corso 2017 contributed the most to the high degree of statistical heterogeneity; its removal resolved heterogeneity but did not meaningfully alter findings (MD 68.74 metres, 95% CI 44.47 to 93.07 m; I² = 0%; n = 122). For the 6‐minute walk distance (6MWD), evidence suggests that considerable improvement was observed immediately following intervention (MD in change from baseline 42.1 m, 95% CI 21.9 to 62.4 m; n = 76; low‐certainty evidence; Analysis 1.2), with a similar magnitude of change evident for 6MWD (36.9 metres between groups at the conclusion of eight weeks of training) in favour of the intervention (P < 0.001) in another study that could not be included in the meta‐analysis (Kumar 2017). Although improvement in endurance shuttle walk test was noted in minutes in favour of exercise training (MD 5.4 minutes, 95% CI 2.71 to 8.09 minutes; n = 39; low‐certainty evidence; Analysis 1.3), no difference in distance (metres) was evident (very low‐certainty evidence; Analysis 1.4; Mandal 2012). Two studies reported that testing of maximal exercise capacity was performed following exercise training (Kumar 2017; Newall 2005). Exercise training may improve constant‐load exercise test performance, but the evidence is very uncertain (MD 505.4 metres, 95% CI 136.51 to 874.29 m; n = 19; very low‐certainty evidence; Analysis 1.5). No improvement in VO₂ peak was noted between baseline and follow‐up (Analysis 1.6; Newall 2005). In contrast, a mean change in VO₂ peak of 81.7 mL/min between groups post intervention in favour of exercise training (P = 0.006) was reported in one study that could not be included in the meta‐analysis (Kumar 2017).
Note: Data are expressed as endpoint (end‐intervention) for Mandal 2012, and as change from baseline for all others.
Two studies whose data were available reported results of the ISWD at three months' follow‐up (Mandal 2012), and at 12 months' follow‐up (Lee 2014). Evidence suggests that exercise training had no effect at either time point (low‐certainty evidence; Analysis 1.7). Similar findings were noted for the 6MWD at 12 months' follow‐up (Analysis 1.8; Lee 2014), and for the constant‐load endurance test, with improvement not maintained at three months' follow‐up in the exercise training group (Newall 2005). In contrast, improvements in endurance shuttle walk distance may have been maintained at three months' follow‐up; however the evidence is very uncertain (very low‐certainty evidence; Analysis 1.9).
Health‐related quality of life
Five studies involving 234 participants with stable bronchiectasis reported findings related to different metrics for HRQoL. Of the five studies that used the St George's Respiratory Quotient (SGRQ) total score (Dal Corso 2017; Kumar 2017; Lee 2014; Mandal 2012; Newall 2005), data were available for pooling from three studies. Pooled data involving 110 participants suggest improvements in quality of life immediately following exercise training (MD in change from baseline ‐9.62 points, 95% CI ‐15.67 to ‐3.56 points; I² = 16%; n = 101; low‐certainty evidence; Analysis 1.10; Figure 6). Similarly, one study that could not be included in the meta‐analysis demonstrated a mean improvement of 8.43 points between groups at end intervention in favour of exercising training (P = 0.002) (Kumar 2017). In contrast, a mean change in SGRQ total score of 2.3 points (95% CI ‐2.9 to 7.4 points) was demonstrated between groups at end intervention for the other study that could not be included in the meta‐analysis (Newall 2005). Data for longer‐term effects on SGRQ total score were available at three months and at 12 months in two studies (Lee 2014; Mandal 2012). Despite a trend towards maintained improvement in quality of life, evidence suggests this may not occur (low‐certainty evidence; Analysis 1.11). For the CRDQ, improvement in the domains of dyspnoea and fatigue was evident in favour of exercise training, but this was not noted for emotional function and mastery (Analysis 1.12; Figure 7; Lee 2014), nor did any CRDQ domains retain improvement at 12 months (Analysis 1.13).
Note: Data are expressed as change from baseline for Dal Corso 2017, and as endpoint (end intervention) for all others.
Figure 7: Forest plot Analysis 1.12 ‐ Health‐related quality of life (Chronic Respiratory Disease Questionnaire) at end‐intervention
Pooled data from two studies involving 103 participants demonstrated that exercise training likely had little to no impact on cough‐related quality of life immediately following treatment as measured by the total LCQ score (MD in change from baseline ‐0.09 points, 95% CI ‐0.98 to 0.80 points; I² = 0%; n = 103; moderate‐certainty evidence) or on LCQ domains of physical, psychological, and social scores (Analysis 1.14; Figure 8). Similar findings were apparent at three months' and at 12 months' follow‐up; however the evidence is very uncertain (very low‐certainty evidence; Analysis 1.15). In contrast, one study that could not be included in the meta‐analysis demonstrated improvement in all LCQ domains with exercise training (all P < 0.001), with only psychological score (P = 0.034) and total score (P = 0.009) improved in the control group (Kumar 2017).
Figure 8: Forest plot Analysis 1.14 ‐ Health‐related quality of life (Leicester Cough Questionnaire) at end‐intervention
Secondary outcomes
Exacerbation and hospitalisation
The rate of exacerbation in those undertaking exercise training was reduced compared to that in the control group (odds ratio (OR) 0.26, 95% CI 0.08 to 0.81; Analysis 1.16). Time to first exacerbation was longer in those with stable bronchiectasis (log rank 0.49, 95% CI 0.01 to 0.97) (Lee 2014).
Peripheral muscle strength
Quadriceps muscle strength was measured in one study (Dal Corso 2017). Exercise training improved quadriceps muscle force compared to usual care (MD 7.4 kg, 95% CI 2.81 to 11.99 kg; n = 39) (Analysis 1.17).
Physical activity
Physical activity levels were not measured in any studies.
Mental health
Two studies measured anxiety and/or depression, using the Hospital Anxiety and Depression Scale (Lee 2014), or the Depression Anxiety Stress Scale (Kumar 2017). One study demonstrated no effect of exercise training on anxiety or depression immediately post intervention (Analysis 1.18; Analysis 1.19), or at 12 months' follow‐up (Analysis 1.20; Analysis 1.21). In contrast, a second study found that the sub‐scale score for depression and the total score improved, regardless of whether exercise training was undertaken, and anxiety improved with exercise training only (P = 0.01) (Kumar 2017).
Clinical symptoms
Clinical symptoms of dyspnoea, cough, and fatigue were assessed only as part of HRQoL questionnaires.
Mortality
The incidence of mortality was evaluated in Lee 2014, which reported no differences between groups (OR 0.27, 95% CI 0.01 to 6.87; Analysis 1.22).
Adverse events
No studies reported on the occurrence of adverse events.
Exercise training versus usual care post acute exacerbation
Primary outcomes
Exercise tolerance
In one study of 27 participants undertaking exercise training following acute exacerbation of bronchiectasis, evidence suggests that treatment may result in little to no difference in the degree of improvement observed in 6MWD between groups following a 12‐week intervention (MD 11 m, 95% CI ‐26.29 to 48.79 m; n = 27; low‐certainty evidence) (Analysis 2.1; Chalmers 2019).
Health‐related quality of life
Exercise training did not appear to impact the degree of change in quality of life observed between groups immediately post intervention via SGRQ total or sub‐domain scores or by LCQ total score (low‐certainty evidence; Analysis 2.2; Analysis 2.3; Chalmers 2019). Health status was measured via the COPD Assessment Test (CAT), and although exercise training did not appear to impact the difference between groups at any time point, the magnitude of differences between groups at the 12‐week follow‐up time point (3.5 points) was noted as exceeding the minimally important difference threshold for this outcome.
Secondary outcomes
Exacerbation and hospitalisation
Time to first exacerbation did not change as a result of exercise training (Analysis 2.4).
Discussion
Summary of main results
This review included six studies comparing exercise training (all multi‐component) to usual care (comprising no exercise training) in adults with bronchiectasis who were in a stable disease state or during a period following an acute exacerbation. Multi‐component interventions comprised physical exercise, educational sessions, and specific instruction regarding airway clearance therapy and/or review of airway clearance therapy techniques. Exercise training was conducted predominantly in an outpatient setting, with programme duration ranging from six to eight weeks. Lack of clinical trial data pertaining to children with bronchiectasis was noted.
Exercise training led to improvement in measures of functional exercise capacity, as measured by the incremental shuttle walk test (ISWT) and the six‐minute walk test (6MWT) in people with stable bronchiectasis. The magnitude of improvement in ISWT following exercise training was 87.12 metres, which is greater than the minimal important difference of 47.5 metres for this outcome for chronic respiratory disease (Singh 2014), and between 35 and 70 metres for people with bronchiectasis (Lee 2014a; Walsh 2020). This reflects a similar magnitude of improvement in functional exercise capacity following pulmonary rehabilitation observed in people with chronic obstructive pulmonary disease (COPD) (Singh 2008). In contrast, the effect on maximal exercise capacity was variable, with contrasting findings between studies. Significant improvements in quality of life were also evident, with the degree of change following exercise training for St George's Respiratory Quotient (SGRQ) total score greater than the minimally important difference threshold of 4 points reported for this outcome in people with chronic respiratory disease (Jones 2005). This was complemented by small reductions in dyspnoea and fatigue immediately following exercise training. Data were insufficient to show the longer‐term effects of exercise training on exercise capacity and quality of life in people with bronchiectasis.
Exercise training appeared to have minimal impact on cough‐related quality of life, anxiety, or depression. Improvement in peripheral muscle strength for the quadriceps was reported in one study of home‐based training. Exercise training was associated with fewer acute exacerbations and longer time to first exacerbation over a 12‐month period in people with stable disease. A single study in people post acute exacerbation of bronchiectasis demonstrated that exercise training was not associated with changes in functional exercise capacity, cough‐related quality of life, or time to first exacerbation (Chalmers 2019). Physical activity was not measured by any studies, and no adverse events related to exercise training were reported.
Overall completeness and applicability of evidence
Five of the included studies involved individuals with stable bronchiectasis, with only one investigating the effects of exercise training in people post acute exacerbation of bronchiectasis. Although rarely specified, it is likely that a diverse range of bronchiectasis aetiologies were included in this review. Specific causes of bronchiectasis may result in more severe physiological impairments, which could impact the benefits of exercise training effectiveness. As the included studies in stable bronchiectasis demonstrated mild to moderate changes in spirometry as a reflection of disease severity, application of these findings to people with more severe disease in a stable clinical state may require further exploration. In the absence of broad application of bronchiectasis severity scores to classify disease severity, measures of airway obstruction may facilitate this interpretation. Guidelines for pulmonary rehabilitation recommend this intervention for people with chronic respiratory disease who demonstrate dyspnoea on exertion (Spruit 2013). In this review, the degree of exertional dyspnoea prior to the intervention was reported by two studies and was stated as ≥ 1 (minimum of 1 to maximum of 3) or ≥ 2 (minimum and maximum not reported) on the modified Medical Research Council Dyspnoea Scale (Kumar 2017; Lee 2014). Therefore caution should be exerted if findings observed in this review are extrapolated to patients with greater levels of exertional dyspnoea.
Pooled outcome data for ISWT reveal that exercise training likely confers clinically relevant benefit compared to usual care. Although this is reassuring and is clinically intuitive, we note the potential for this result to have been influenced by a considerable between‐group difference at baseline in one study (intervention: 287.5 metres; control: 343.3 metres) (Mandal 2012). The review authors noted that this baseline imbalance is not statistically significant; however it does exceed the minimally important difference threshold for people with chronic lung disease (Singh 2014). When considered in light of the lack of between‐group differences at the end of the intervention (mean difference (MD) of 5.5 metres) for this study, it is possible that the true magnitude of effect for this outcome may have been underestimated. The effect estimates for other measures of endurance capacity or maximal exercise testing were more variable. Although improvements were noted in endurance shuttle walk test time, no meaningful change in endurance shuttle walk test distance was noted (Mandal 2012). Similarly, although significant improvement in walking distance was observed during a constant load exercise test, the effect on peak oxygen uptake (VO₂ peak) showed contrast between studies. This may be explained by factors related to individual study methods, with sample size calculations specific for these outcomes not reported in either study, and the approach used for exercise prescription specific to exercise intensity lacking in one study (Kumar 2017). Therefore, it is difficult to determine whether lack of observed effect in such studies reflected true treatment effects or was a consequence of inadequate statistical power to detect appropriate changes. This may also account for one study reporting lack of difference in SGRQ total score at the end of the intervention (Newall 2005), in contrast to pooled improvement in this outcome and in measures of dyspnoea and fatigue immediately post intervention (Lee 2014).
Lack of observed effects of exercise training on psychological measures contrasts with reports of people with other respiratory conditions such as COPD (Gordon 2019). This finding could be attributable to the mean baseline levels of anxiety and depression in the single study that reported this outcome (Lee 2014). In this study, baseline Hospital Anxiety and Depression Scale (HADS) scores did not meet the minimum standard to satisfy diagnostic criteria for the presence of a mood disorder (Snaith 2003), meaning that the opportunity to demonstrate a treatment effect may have been diminished. The improvement in quadriceps muscle force concurs with findings observed in people with other respiratory conditions undergoing exercise training (Spruit 2013). However, given the small volume of data related to this outcome, further research is indicated to confirm this observation in people with bronchiectasis.
All studies in this review incorporated exercise training applied in conjunction with other co‐interventions. These included educational sessions covering a range of topics, regular use of airway clearance therapy, and review of a prescribed technique. It is important to note that all such co‐interventions were applied as part of usual care. Therefore, it is likely that observed effects in this review are attributable to the exercise training stimulus. Studies of self‐management in bronchiectasis, which have included educational sessions on causes of bronchiectasis, disease process, medical investigations, symptom management, exacerbations, health promotion, airway clearance techniques, and support, did not demonstrate beneficial effects on health‐related quality of life (HRQoL) compared to usual care in people with bronchiectasis (Kelly 2018). With some similarities in the educational topics included in the studies in this review, it is possible that their contributions to changes in HRQoL were minimal, but this should be confirmed in future work. In contrast, findings regarding airway clearance therapy, with a previous report of improvement in exercise capacity with this intervention (Murray 2009), are variable, and some studies show no change (Herrero‐Cortina 2019; Munoz 2018). Although the impact of airway clearance interventions on exercise capacity appears controversial and unclear, it cannot be excluded that some observed treatment effects may be partially influenced by the adjunct effect of this co‐intervention. Similarly, effective airway clearance therapy can improve dyspnoea, HRQoL (Lee 2015), and cough‐related quality of life (Murray 2009); the relative contribution of this co‐intervention is therefore somewhat difficult to determine.
Bronchiectasis‐specific quality of life and health status questionnaires, together with measures of bronchiectasis impact, have been recently developed but did not feature in the findings of our review (Crichton 2020; Quittner 2015; Spinou 2017). Application of these tools in evaluating the effects of exercise training in bronchiectasis may further clarify our observed effects on quality of life. The lack of effect on cough‐related quality of life may be related to the instructions provided during exercise training. Optimisation of exercise training to address cough symptoms may be achieved by incorporation of the forced expiratory technique during or immediately following training (Ward 2019). As it is unknown whether this strategy and/or advice was provided to participants in the included studies, it is difficult to determine whether this may account for the lack of change in cough‐related quality of life. Sub‐therapeutic, adjuvant sham respiratory muscle training was another co‐intervention reported in one study (Newall 2005). Although this may have been unlikely to have influenced outcomes related to exercise capacity, it was hypothesised that it may have facilitated improvements in airway clearance. This was not explicitly examined in this study, suggesting that its possible impact on changes in cough‐related quality of life is also unclear.
Although all studies in this review used aerobic exercise training, the inclusion of resistance training was more variable, and its use was not reported in two studies (Kumar 2017; Newall 2005). The use of aerobic exercise training aligns with current recommendations for pulmonary rehabilitation for people with chronic respiratory disease (Spruit 2013), but we are not able to draw consistent inferences regarding the relative contributions of resistance training. Only one study specifically measured peripheral muscle strength. All programmes adhered to current recommendations related to programme length (six to eight weeks), but training sessions varied from unsupervised to supervised, with sessions conducted three to four times per week. Clarity regarding the optimal exercise training prescription for people with bronchiectasis, regardless of clinical state, is currently lacking.
The observed reduction in the incidence of exacerbations following exercise training arose from a single study (Lee 2014; n = 30 included in analysis). This outcome is a common 'downstream' target of exercise training interventions, but data remain scarce for people with bronchiectasis. The ability to affect exacerbations is clinically important due to their negative impact upon subsequent aspects of disease progression such as future hospitalisations, poor quality of life, and mortality (Polverino 2018). It is furthermore interesting to note the magnitude of observed effect (75% reduced odds for experiencing an exacerbation), which is larger than for many other treatments (including pharmacological ones) that may seek to directly impact this outcome. It is challenging, however, to propose a mechanism that accurately explains our observed effect. It is perhaps unlikely attributable to exercise‐induced changes in respiratory‐related symptoms (e.g. dyspnoea, fatigue, cough‐related symptoms) in light of the small magnitude of observed changes noted over time. This may be unsurprising, as exercise training alone in this population is not hypothesised to alter sputum clearance; however further inquiry in this context would appear indicated.
The lack of clinically meaningful benefit of exercise training for people following an acute exacerbation of bronchiectasis seems unlikely to be attributable to intervention timing, as this aligned with international pulmonary rehabilitation commencement recommendations (Alison 2017; Puhan 2016). However, the lack of impact on time to first exacerbation is likely due to an insufficient follow‐up duration following the intervention (three months) (Chalmers 2019). The inclusion of only individuals who had completed a 14‐day course of antibiotics further limits the application of these findings. Although a mix of disease severity was noted for the baseline number of exacerbations or the Bronchiectasis Severity Index (mild to severe), no individuals with severe exacerbations that required hospitalisation were included. Further research is required to explore the effects of exercise training in alternate subgroups with respect to exacerbation severity and the need for hospitalisation.
The lack of data pertaining to children with bronchiectasis is a matter of concern. Exercise training undertaken in the context of pulmonary rehabilitation programmes is most commonly undertaken by adults with chronic lung disease, but this does not imply that children with bronchiectasis may not benefit from exercise training. We ensured that our intervention definition was not constrained to a pulmonary rehabilitation context for this specific reason, but we retrieved only one record of potential relevance (an ongoing clinical trial registered from Australia). It is important that the absence of clinical data for this patient group is not equated to lack of effectiveness of exercise training. Future clinical trials in this space are clearly indicated. Similarly, although no studies of telerehabilitation in people with bronchiectasis were presently identified for inclusion, we anticipate that clinical outcomes from this model of exercise training in this population will emerge in the future. The relatively small number of studies included within this review demonstrates an ongoing indication for additional research regarding the clinical effectiveness of exercise training for people with bronchiectasis across a broad range of clinical outcomes. Although withholding of exercise might pose ethical challenges to conducting such clinical trials in settings where this form of therapy is actively promoted, this is not yet usual care in many places and may be managed through wait‐list control groups (who receive the intervention in a delayed fashion). Additional research data in this area would equip healthcare providers with greater accuracy regarding the likelihood and magnitude of potential benefits from such treatments and, in turn, would assist people affected by bronchiectasis. This is quite different from the research landscape surrounding people affected by COPD, where further trials examining the effectiveness of pulmonary rehabilitation compared to usual care have been discouraged (Lacasse 2015). The modern era of personalised and flexible models of care (including rehabilitation) makes it challenging to identify any single direction of research that may confer the greatest benefit for the greatest number of people affected by bronchiectasis. Considerable opportunity clearly exists to extend inquiry into various aspects of patient care and precision medicine in the field of bronchiectasis rehabilitation.
Quality of the evidence
The evidence provided in this review should be interpreted with regards to some identified sources of bias. Two of the six studies were available only in abstract form (Dal Corso 2017; Kumar 2017); this limits detail on data provided, as well as the ability to adequately assess all parameters of risk of bias for these studies. Due to unclear reporting, selection bias may be present in four studies (Chalmers 2019; Kumar 2017; Mandal 2012; Newall 2005). Lack of blinding of participants to knowledge of treatment allocation was confirmed in two studies (Dal Corso 2017; Lee 2014), but it was rated as unclear for all others. This was not unexpected, as blinding of participants in studies involving physical interventions is inherently difficult. Blinding of outcome assessors was applied in Lee 2014; for the remaining studies, this was not applied (Dal Corso 2017), or it was unclear (Chalmers 2019; Kumar 2017; Mandal 2012; Newall 2005). We attributed high risk of bias to all studies regarding self‐reported outcome measures, as we felt this potential influence was unavoidable. Data related to long‐term treatment effect estimates may have been affected by rates of participant attrition, but this appeared to be of modest magnitude, even for the 12‐month follow‐up data provided in Lee 2014.
We rated the overall quality of evidence according to the GRADE method as very low to low for different measures of exercise capacity, and as very low to moderate for HRQoL. This was predominantly attributed to increased risk of bias due to inadequate reporting of methods and lack of blinding of outcome measures, imprecision, and indirectness due to the small number of studies. Inconsistency was rarely observed across outcome data, with only one outcome (long‐term follow‐up of cough‐related quality of life) demonstrating effect estimates that contrasted between included studies.
Potential biases in the review process
All data were extracted independently by two review authors, and discrepancies were resolved through discussion. Risk of bias ratings were completed by two review authors independently. Data from Lee 2014 were extracted by two review authors who were not involved in this study.
A broad search included handsearching of conference abstracts and trial registries. Studies published in abstract form were included to ensure all available trials were incorporated within the review. When clarification of study characteristics or additional data were required, we attempted to obtain this information from study authors. This was successful on selected occasions, but in some cases, data were not available. This may have influenced the assessment of trial quality and some estimates of treatment effects.
Agreements and disagreements with other studies or reviews
These findings expand upon the conclusions of an earlier Cochrane Review, which found that inspiratory muscle training compared to sham or no training improved endurance exercise capacity and quality of life (Bradley 2002). In addition, results for exercise capacity, quality of life, and cough‐related quality of life are consistent with the overall findings of a previous systematic review of pulmonary rehabilitation for people with bronchiectasis (Lee 2017). The magnitude of change for these primary outcomes is similar to that in people with COPD who are undergoing pulmonary rehabilitation (McCarthy 2015). This supports the international statements and recent pulmonary rehabilitation guidelines for inclusion of people with bronchiectasis in pulmonary rehabilitation (Alison 2017; Spruit 2013). Our findings are also consistent with those of an earlier Cochrane Review of physical training in bronchiectasis, which demonstrated that the combination of exercise training, respiratory muscle training, and education improved endurance exercise capacity and quality of life compared to exercise training and sham respiratory muscle training (Bradley 2002). This previous review, however, incorporated findings from only two reports related to one study published in abstract form only, with limited participant numbers considerably limiting the ability to draw robust implications for clinical practice.
Overview of Cochrane Crowd Known Assessments and Screen4Me workflows for original and updated searches.
Figure 2. Study flow diagram
Figure 3. Risk of bias summary
Figure 4. Risk of bias graph
Note: Data are expressed as endpoint (end‐intervention) for Mandal 2012, and as change from baseline for all others.
Note: Data are expressed as change from baseline for Dal Corso 2017, and as endpoint (end intervention) for all others.
Figure 7: Forest plot Analysis 1.12 ‐ Health‐related quality of life (Chronic Respiratory Disease Questionnaire) at end‐intervention
Figure 8: Forest plot Analysis 1.14 ‐ Health‐related quality of life (Leicester Cough Questionnaire) at end‐intervention
Comparison 1: Exercise training vs usual care (stable disease), Outcome 1: Exercise tolerance ‐ ISWT (metres) at end intervention
Comparison 1: Exercise training vs usual care (stable disease), Outcome 2: Exercise tolerance ‐ 6MWT (metres) at end intervention
Comparison 1: Exercise training vs usual care (stable disease), Outcome 3: Exercise tolerance ‐ ESWT (mins) at end intervention
Comparison 1: Exercise training vs usual care (stable disease), Outcome 4: Exercise tolerance ‐ ESWT (metres) at end intervention
Comparison 1: Exercise training vs usual care (stable disease), Outcome 5: Exercise tolerance ‐ CLET (metres) at end intervention
Comparison 1: Exercise training vs usual care (stable disease), Outcome 6: Exercise tolerance ‐ CPET VO 2 peak at end intervention
Comparison 1: Exercise training vs usual care (stable disease), Outcome 7: Exercise tolerance ‐ ISWT (metres) at follow‐up
Comparison 1: Exercise training vs usual care (stable disease), Outcome 8: Exercise tolerance ‐ 6MWT (metres) at follow‐up
Comparison 1: Exercise training vs usual care (stable disease), Outcome 9: Exercise tolerance ‐ ESWT (metres) at follow‐up
Comparison 1: Exercise training vs usual care (stable disease), Outcome 10: HRQoL ‐ SGRQ total score at end intervention
Comparison 1: Exercise training vs usual care (stable disease), Outcome 11: HRQoL ‐ SGRQ total score at follow‐up
Comparison 1: Exercise training vs usual care (stable disease), Outcome 12: HRQoL ‐ CRDQ (all domains) at end treatment
Comparison 1: Exercise training vs usual care (stable disease), Outcome 13: HRQoL ‐ CRDQ (all domains) at follow‐up
Comparison 1: Exercise training vs usual care (stable disease), Outcome 14: HRQoL ‐ LCQ (all domains) at end intervention
Comparison 1: Exercise training vs usual care (stable disease), Outcome 15: HRQol ‐ LCQ (all domains) at follow‐up
Comparison 1: Exercise training vs usual care (stable disease), Outcome 16: Exacerbations / hospitalisations ‐ Number of exacerbations
Comparison 1: Exercise training vs usual care (stable disease), Outcome 17: Peripheral muscle force at end intervention
Comparison 1: Exercise training vs usual care (stable disease), Outcome 18: Anxiety ‐ HADS at end intervention
Comparison 1: Exercise training vs usual care (stable disease), Outcome 19: Depression ‐ HADS at end intervention
Comparison 1: Exercise training vs usual care (stable disease), Outcome 20: Anxiety ‐ HADS at follow‐up
Comparison 1: Exercise training vs usual care (stable disease), Outcome 21: Depression ‐ HADS at follow‐up
Comparison 1: Exercise training vs usual care (stable disease), Outcome 22: Mortality (all cause)
Comparison 1: Exercise training vs usual care (stable disease), Outcome 23: Sensitivity analysis for ISWT end intervention
Comparison 1: Exercise training vs usual care (stable disease), Outcome 24: Sensitivity analysis for SGRQ total end intervention
Comparison 1: Exercise training vs usual care (stable disease), Outcome 25: Heterogeneity for ISWT post intervention
Comparison 1: Exercise training vs usual care (stable disease), Outcome 26: Heterogeneity for SGRQ total score end intervention
Comparison 2: Exercise training vs usual care (post exacerbation), Outcome 1: Exercise tolerance ‐ 6MWT (metres) at end intervention
Comparison 2: Exercise training vs usual care (post exacerbation), Outcome 2: HRQoL ‐ SGRQ total score at end intervention
Comparison 2: Exercise training vs usual care (post exacerbation), Outcome 3: HRQoL ‐ LCQ total score at end intervention
Comparison 2: Exercise training vs usual care (post exacerbation), Outcome 4: Exacerbations/Hospitalisations ‐ Time to first exacerbation
| Exercise training compared to control in people with stable bronchiectasis | |||||||
| Patient or population: people with stable bronchiectasis | |||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | №. of participants | Certainty of the evidence | Comments | ||
|---|---|---|---|---|---|---|---|
| Risk with usual care | Risk with exercise training | ||||||
| Change in incremental shuttle walk distance | Follow‐up: range 6 weeks to 8 weeks | Mean change in incremental shuttle walk distance ranged from ‐70 to 2.5 metres | MD 87 metres higher | ‐ | 161 | ⊕⊕⊝⊝ | |
| Follow‐up: range 3 months to 12 months | Mean change in incremental shuttle walk distance ranged from 5.2 to 343 metres | MD 6.19 metres higher | ‐ | 82 | ⊕⊕⊝⊝ | ||
| Change in 6‐minute walk distance | Follow‐up: mean 8 weeks | Mean change in 6‐minute walk distance was ‐10.9 metres | MD 42 metres higher | ‐ | 76 | ⊕⊕⊝⊝ | |
| Follow‐up: mean 12 months | Mean change in 6‐minute walk distance was ‐8.26 metres | MD 6.74 metres lower | ‐ | 55 | ⊕⊕⊝⊝ LOWc,e | ||
| Change in endurance shuttle walk time | Follow‐up: mean 8 weeks | Mean change in endurance shuttle walk time was 0.2 minutes | MD 5.4 minutes higher | ‐ | 39 | ⊕⊕⊝⊝ | |
| Change in endurance shuttle walk distance | Follow‐up: mean 8 weeks | Mean change in endurance shuttle walk distance was ‐36.4 metres | MD 311.6 metres higher | ‐ | 27 | ⊕⊝⊝⊝ | |
| Follow‐up: mean 3 months | Mean change in endurance shuttle walking distance was 964.3 metres | MD 385.7 metres higher | ‐ | 27 | ⊕⊝⊝⊝ | ||
| Change in walking distance during endurance walking test assessed with constant load endurance test | Follow‐up: mean 8 weeks | Mean change in walking distance during endurance walking test was ‐112.6 metres | MD 505.4 metres higher | ‐ | 19 | ⊕⊝⊝⊝ | |
| Change in peak oxygen uptake | Follow‐up: mean 8 weeks | Mean change in peak oxygen uptake was ‐1.91 L/min | MD 3.87 L/min higher | ‐ | 19 | ⊕⊝⊝⊝ | |
| Change in quality of life | Follow‐up: range 6 weeks to 8 weeks | Mean change in quality of life ranged from 4 to 39.2 points | MD 9.62 points lower | ‐ | 110 | ⊕⊕⊝⊝ | Lower scores post intervention are favourable, indicating improvement in quality of life |
| Follow‐up: range 3 months to 12 months | Mean change in quality of life ranged from 33.6 to 45.2 points | MD 6.78 points fewer | ‐ | 65 | ⊕⊕⊝⊝ | Lower scores at follow‐up are favourable, indicating improvement in quality of life | |
| Change in cough‐related quality of life | Follow‐up: range 6 weeks to 8 weeks | Mean change in cough‐related quality of life ranged from 14.6 to 15.5 points | MD 0.09 points lower | ‐ | 103 | ⊕⊕⊕⊝ | Higher scores post intervention are favourable, indicating improvement in quality of life |
| Follow‐up: range 3 months to 12 months | Mean change in cough‐related quality of life ranged from 13.6 to 17.9 points | MD 0.97 points lower | ‐ | 82 | ⊕⊝⊝⊝ | Higher scores at follow‐up are favourable, indicating improvement in quality of life | |
| *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). | |||||||
| GRADE Working Group grades of evidence. | |||||||
| aWide confidence interval limits exceed (or likely exceed) a clinically relevant threshold for this patient group (imprecision ‐1). bAllocation bias for 3 studies was unclear, and only 1 study included blinded assessors (bias ‐1). cLimited data were available for meta‐analysis, reducing its broad representativeness of important factors (e.g. disease severity, settings) (indirectness ‐1). dAllocation bias in 1 study was unclear (bias ‐1). eConfidence interval limits exceed minimally important difference for this outcome in people with bronchiectasis (imprecision ‐1). fBlinding of assessors was not reported and attrition or reporting bias was unclear (bias ‐1). gAllocation bias and attrition and reporting bias for 1 study were unclear (bias ‐1). hEffect estimates contrast between benefit and harm in the two included studies (inconsistency ‐1). | |||||||
| Exercise training compared to control for people in the post‐acute period following acute exacerbation of bronchiectasis | ||||||
| Patient or population: people in the post‐acute period following acute exacerbation of bronchiectasis | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | №. of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with usual care | Risk with exercise training | |||||
| Change in 6‐minute walk distance | Mean change in 6‐minute walk distance was 15 metres | MD 11 metres higher | ‐ | 27 | ⊕⊕⊝⊝ | |
| Change in cough‐related quality of life assessed with Leicester Cough Questionnaire | Mean change in cough‐related quality of life was 0.91 units | MD 0.08 units lower | ‐ | 27 | ⊕⊕⊝⊝ | Higher scores post intervention are favourable, indicating improvement in quality of life |
| *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). | ||||||
| GRADE Working Group grades of evidence. | ||||||
| aRandom sequence generation was unclear and outcome data were incomplete in the sole included study (bias ‐1). bLimited data were available for meta‐analysis, reducing its broad representativeness of important factors (e.g. disease severity, settings) (indirectness ‐1). | ||||||
| Study | Setting | Intervention | Control | Duration (weeks) | Frequency | Follow‐up (weeks) |
|---|---|---|---|---|---|---|
| OP Supervised | Aerobic: treadmill, bike, walking (80% VO₂ max) Resistance: UL, LL (8 to 10 reps at 1 RM). Education including instruction in daily chest PT | Guideline concordant ongoing management including instruction in daily chest PT | 6 | 2 OP + 2 HB sessions/week | 12 | |
| HB Unsupervised | Aerobic: 20 minutes stepping on platform (60% to 80% maximal stepping rate in IST) Resistance: 20 minutes UL, LL (8 reps, 3 sets at 70% MVIC). Education. Weekly phone call. Home visit every 15 days | Education Recommended to walk at moderate intensity for 30 minutes 3 times/week. Weekly phone call (no exercise or PA discussion) | 8 | 3 times/week Average session duration: 50 minutes | Nil | |
| OP | No details provided | Standard care | 8 | No details provided | Nil | |
| OP Supervised | Aerobic: 15 minutes treadmill or walking (75% ISWT speed) and bike (60% maximal work rate) Resistance: 15 minutes UL, LL (10 reps, 1 to 3 sets at 10 RM) Encouraged to maintain HEP over follow‐up period via monthly phone call. Reviewed usual ACT. Taught ACBT If no usual ACT | Education Twice‐weekly phone call (no exercise or PA discussion) Reviewed usual ACT. Taught ACBT if no usual ACT | 8 | 2 OP + 3 to 5 HB sessions/week | 52 | |
| OP Supervised | Aerobic: 10 minutes treadmill, bike, ski machine (85% VO₂ max) Resistance: UL, LL (10 reps, 3 sets at 60% 1 RM, progressed to 70% at Week 3, 80% at Week 5). Twice‐daily chest PT. Education. Self‐management plan. Encouraged to undertake gym programme for 6 months at end of intervention | Twice‐daily chest PT. Education. Self‐management plan | 8 | 2 OP + 1 HB session/ week | 12 | |
| OP Supervised | OP aerobic: 15 minutes treadmill, cycle ergometry, stair climbing (80% peak HR on maximal test) HB aerobic: 45 minutes walking at target intensity Education. Sham IMT at sub‐therapeutic load (7 cmH₂O) for 15 minutes twice daily at home | Education | 8 | 2 OP + 1 HB sessions/week Average session duration 45 minutes | 12 | |
| Abbreviations: ACBT: active cycle of breathing technique; ACT: airway clearance technique; HB: home‐based; HEP: home exercise programme; IMT: inspiratory muscle training; IST: incremental step test; ISWT: incremental shuttle walk test; LL: lower limb; MVIC: maximum voluntary isometric contraction; OP: outpatient; PA: physical activity; PR: pulmonary rehabilitation; PT: physiotherapy; RM: repetition maximum; UL; upper limb; VO₂ max: maximal oxygen uptake. | ||||||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1.1 Exercise tolerance ‐ ISWT (metres) at end intervention Show forest plot | 4 | 161 | Mean Difference (IV, Random, 95% CI) | 87.12 [42.65, 131.58] |
| 1.1.1 Uni‐component | 0 | 0 | Mean Difference (IV, Random, 95% CI) | Not estimable |
| 1.1.2 Multi‐component | 4 | 161 | Mean Difference (IV, Random, 95% CI) | 87.12 [42.65, 131.58] |
| 1.2 Exercise tolerance ‐ 6MWT (metres) at end intervention Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.2.1 Uni‐component | 0 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.2.2 Multi‐component | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.3 Exercise tolerance ‐ ESWT (mins) at end intervention Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.3.1 Uni‐component | 0 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.3.2 Multi‐component | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.4 Exercise tolerance ‐ ESWT (metres) at end intervention Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.4.1 Uni‐component | 0 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.4.2 Multi‐component | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.5 Exercise tolerance ‐ CLET (metres) at end intervention Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.5.1 Uni‐component | 0 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.5.2 Multi‐component | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.6 Exercise tolerance ‐ CPET VO 2 peak at end intervention Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.6.1 Uni‐component | 0 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.6.2 Multi‐component | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.7 Exercise tolerance ‐ ISWT (metres) at follow‐up Show forest plot | 2 | 82 | Mean Difference (IV, Random, 95% CI) | 6.19 [‐15.51, 27.90] |
| 1.7.1 Uni‐component | 0 | 0 | Mean Difference (IV, Random, 95% CI) | Not estimable |
| 1.7.2 Multi‐component | 2 | 82 | Mean Difference (IV, Random, 95% CI) | 6.19 [‐15.51, 27.90] |
| 1.8 Exercise tolerance ‐ 6MWT (metres) at follow‐up Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.8.1 Uni‐component | 0 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.8.2 Multi‐component | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.9 Exercise tolerance ‐ ESWT (metres) at follow‐up Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.9.1 Uni‐component | 0 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.9.2 Multi‐component | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.10 HRQoL ‐ SGRQ total score at end intervention Show forest plot | 3 | 110 | Mean Difference (IV, Random, 95% CI) | ‐9.62 [‐15.67, ‐3.56] |
| 1.10.1 Uni‐component | 0 | 0 | Mean Difference (IV, Random, 95% CI) | Not estimable |
| 1.10.2 Multi‐component | 3 | 110 | Mean Difference (IV, Random, 95% CI) | ‐9.62 [‐15.67, ‐3.56] |
| 1.11 HRQoL ‐ SGRQ total score at follow‐up Show forest plot | 2 | 65 | Mean Difference (IV, Random, 95% CI) | ‐6.78 [‐14.98, 1.42] |
| 1.11.1 Uni‐component | 0 | 0 | Mean Difference (IV, Random, 95% CI) | Not estimable |
| 1.11.2 Multi‐component | 2 | 65 | Mean Difference (IV, Random, 95% CI) | ‐6.78 [‐14.98, 1.42] |
| 1.12 HRQoL ‐ CRDQ (all domains) at end treatment Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.12.1 CRDQ ‐ Dyspnoea | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.12.2 CRDQ ‐ Fatigue | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.12.3 CRDQ ‐ Emotional Function | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.12.4 CRDQ ‐ Mastery | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.13 HRQoL ‐ CRDQ (all domains) at follow‐up Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.13.1 CRDQ ‐ Dyspnoea | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.13.2 CRDQ ‐ Fatigue | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.13.3 CRDQ ‐ Emotional function | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.13.4 CRDQ ‐ Mastery | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.14 HRQoL ‐ LCQ (all domains) at end intervention Show forest plot | 2 | Mean Difference (IV, Random, 95% CI) | Subtotals only | |
| 1.14.1 LCQ ‐ Physical | 1 | 76 | Mean Difference (IV, Random, 95% CI) | 0.10 [‐0.28, 0.48] |
| 1.14.2 LCQ ‐ Psychological | 1 | 76 | Mean Difference (IV, Random, 95% CI) | ‐0.30 [‐0.84, 0.24] |
| 1.14.3 LCQ ‐ Social | 1 | 76 | Mean Difference (IV, Random, 95% CI) | 0.10 [‐0.33, 0.53] |
| 1.14.4 LCQ ‐ Total | 2 | 103 | Mean Difference (IV, Random, 95% CI) | ‐0.09 [‐0.98, 0.80] |
| 1.15 HRQol ‐ LCQ (all domains) at follow‐up Show forest plot | 2 | Mean Difference (IV, Random, 95% CI) | Subtotals only | |
| 1.15.1 LCQ ‐ Physical | 1 | 55 | Mean Difference (IV, Random, 95% CI) | ‐0.70 [‐1.45, 0.05] |
| 1.15.2 LCQ ‐ Psychological | 1 | 55 | Mean Difference (IV, Random, 95% CI) | ‐0.60 [‐1.18, ‐0.02] |
| 1.15.3 LCQ ‐ Social | 1 | 55 | Mean Difference (IV, Random, 95% CI) | ‐3.10 [‐4.11, ‐2.09] |
| 1.15.4 LCQ ‐ Total | 2 | 82 | Mean Difference (IV, Random, 95% CI) | ‐0.97 [‐8.27, 6.34] |
| 1.16 Exacerbations / hospitalisations ‐ Number of exacerbations Show forest plot | 1 | Odds Ratio (M‐H, Random, 95% CI) | Totals not selected | |
| 1.16.1 Uni‐component | 0 | Odds Ratio (M‐H, Random, 95% CI) | Totals not selected | |
| 1.16.2 Multi‐component | 1 | Odds Ratio (M‐H, Random, 95% CI) | Totals not selected | |
| 1.17 Peripheral muscle force at end intervention Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.17.1 Uni‐component | 0 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.17.2 Multi‐component | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.18 Anxiety ‐ HADS at end intervention Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.18.1 Uni‐component | 0 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.18.2 Multi‐component | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.19 Depression ‐ HADS at end intervention Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.19.1 Uni‐component | 0 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.19.2 Multi‐component | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.20 Anxiety ‐ HADS at follow‐up Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.20.1 Uni‐component | 0 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.20.2 Multi‐component | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.21 Depression ‐ HADS at follow‐up Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.21.1 Uni‐component | 0 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.21.2 Multi‐component | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.22 Mortality (all cause) Show forest plot | 1 | Odds Ratio (M‐H, Random, 95% CI) | Totals not selected | |
| 1.22.1 Uni‐component | 0 | Odds Ratio (M‐H, Random, 95% CI) | Totals not selected | |
| 1.22.2 Multi‐component | 1 | Odds Ratio (M‐H, Random, 95% CI) | Totals not selected | |
| 1.23 Sensitivity analysis for ISWT end intervention Show forest plot | 3 | 122 | Mean Difference (IV, Random, 95% CI) | 69.65 [45.32, 93.98] |
| 1.24 Sensitivity analysis for SGRQ total end intervention Show forest plot | 2 | 71 | Mean Difference (IV, Fixed, 95% CI) | ‐4.98 [‐13.61, 3.65] |
| 1.25 Heterogeneity for ISWT post intervention Show forest plot | 3 | 122 | Mean Difference (IV, Random, 95% CI) | 68.74 [44.42, 93.07] |
| 1.26 Heterogeneity for SGRQ total score end intervention Show forest plot | 2 | 71 | Mean Difference (IV, Random, 95% CI) | ‐4.98 [‐13.61, 3.65] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 2.1 Exercise tolerance ‐ 6MWT (metres) at end intervention Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 2.1.1 Uni‐component | 0 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 2.1.2 Multi‐component | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 2.2 HRQoL ‐ SGRQ total score at end intervention Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 2.2.1 Uni‐component | 0 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 2.2.2 Multi‐component | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 2.3 HRQoL ‐ LCQ total score at end intervention Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 2.3.1 Uni‐component | 0 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 2.3.2 Multi‐component | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 2.4 Exacerbations/Hospitalisations ‐ Time to first exacerbation Show forest plot | 1 | 27 | Hazard Ratio (IV, Random, 95% CI) | 0.83 [0.31, 2.22] |
