A meta-analysis on the structure of pulmonary rehabilitation maintenance programmes on COPD patients' functional capacity.

Pulmonary rehabilitation (PR) improves functional capacity, health-related quality of life (HRQoL) in COPD patients, and maintenance programmes are relevant in preserving those improvements. However, little is known about the structure of maintenance programmes after PR. We performed a systematic review and meta-analysis of experimental and quasi-experimental studies evaluating individuals with COPD admitted to a maintenance PR programme, delivered after an initial PR programme. We reported functional capacity evaluation (6-minute-walking-test), HRQoL, dyspnoea and symptom control. Searches were performed on the 11th April 2021 using MEDLINE, Embase, EBSCO, CINAHL, Web of Science and Cochrane Library. We extracted summary-level data from trial publications and used a random-effects model, predicting that severe heterogeneity was detected. The protocol was registered in PROSPERO (CRD42021247724). Fifteen studies were included in the meta-analysis, with 1151 participants. Maintenance programmes were associated with a pooled mean increase of 27.08 meters in 6mWT (CI: 10.39 to 43.77; I2 = 93%; p < 0.0001), being better in supervised, long (>12 month) home-based programmes; and having a potential MD of -4.20 pts in SGRQ (CI: -4.49 to -3.91; I2 = 0%; p = 0.74). Regarding dyspnoea and exacerbations, we found a nonsignificant trend for improvement after maintenance PR programmes. Severe COPD patients showed smaller improvements in programmes up to a year. Overall, the strength of the underlying evidence was moderate. Despite limitations of risk of bias and heterogeneity, our results support that home-based, supervised, long-term maintenance PR programmes may significantly improve functional capacity in COPD patients and HRQoL.

Introduction

Chronic Obstructive Pulmonary Disease (COPD) is a common, preventable and treatable disease, and it is the third cause of death worldwide[1]. COPD patients often present a decline in functional capacity (FC) over time and persistent symptoms such as dyspnoea, low exercise tolerance and fatigue, leading to poor health-related quality of life and an increased risk for exacerbations and death[1].

Pulmonary Rehabilitation (PR) is one of the most cost-effective treatments for COPD[1]. Its standard core elements are physical exercise training, patient-directed education, smoking cessation support, disease self-management and behaviour change[1-4]. However, improvements in the functional capacity of COPD patients last less than one year[1,2,4].

Functional capacity tends to decrease over time, mainly if the patient does not change his behaviour concerning physical activity[1]. While initial studies showed a difference at six months favouring supervised maintenance exercises[5-7], the effect seems not to be sustained at 12 months. A more recent meta-analysis suggested that maintenance PR programmes might confer long-term efficacy (12 months) on functional capacity in patients with COPD[8], with a mean difference in the six-minute walk test (6mWT) of 27 meters. The 6mWT is a predictor of morbidity and mortality[9] among COPD patients, it is a simple clinical exercise test for the objective evaluation of functional exercise capacity improvement in COPD, which measures the distance a patient can quickly walk on a flat hard surface in 6 minutes.

It is well known that PR programmes include outpatient or inpatient (centre-based), community and home-based types. Most PR programmes are centre-based. Participants are prescribed personalized and structured programmes[3], usually lasting eight to twelve weeks[1,2]. Still, such programmes vary significantly depending on healthcare systems and local policies[2,10]. According to the consensus of PR, to maintain the benefits of this intervention, health behaviour change is crucial[1,2,4].

Malaguti et al. [11], in a recent Cochrane review that specifically included papers on supervised maintenance programs, states that these programmes have an impact on functional exercise capacity. Particularly, the optimal duration to achieve relevant benefits, the role of the supervision and setting of maintenance PR programmes is still unclear[12]. Thus, little is known about what components should constitute a maintenance programme[8].

This systematic review and meta-analysis aimed to synthesise’ evidence of interventional studies, including maintenance PR programmes on COPD patients, evaluate its impact on functional capacity and other disease-related secondary outcomes, and determine the components of those programmes that lead to better long term outcomes.

Methods

This systematic review and meta-analysis followed the Cochrane Handbook recommendations[13]. Our study protocol has been prospectively registered and published in the PROSPERO database, CRD42021247724.

Search strategy and selection criteria

The primary outcome of this study was the functional capacity measured by the 6mWT, and we selected experimental and quasi-experimental studies evaluating individuals with COPD submitted to a maintenance PR programme, delivered after an initial PR programme and reporting an evaluation of functional capacity measured by the 6mWT[14,15]. Quasi-experimental studies were considered in this systematic review because we aimed to acknowledge the components of the intervention.

Secondary outcomes considered for inclusion were dyspnoea, measured by the modified Medical Research Council scale (mMRC);[16] exacerbations, measured by hospital admissions during the maintenance rehabilitation programme;[17] and health-related quality of life (HRQoL), measured by the St George’s Respiratory Questionnaire (SGRQ)[18], the EuroQoL5[19] or SF-36[20].

We excluded non-interventional studies and studies without an initial PR programme. In addition, studies in which the initial PR programme included only physical exercise or only education were also excluded because they are not actual PR programmes according to the accepted definition of PR[2,4]. We did not exclude studies based on language.

The search was performed on the 11th April 2021 in the Scopus database (MEDLINE and Embase), EBSCO, the Cumulative Index to Nursing and Allied Health Literature [CINAHL] Complete, Web of Science and Cochrane Library. In addition, we reviewed the references of the included literature and correlated systematic reviews. See “Supplementary Information 1 – Search strategy”.

Two researchers (LS, TM) conducted the study selection independently, using the RYYAN QCRI app (available at: https://www.rayyan.ai/), and all disagreements were resolved by consensus or through consultation with a third investigator (MP) in the review team. Duplicated and irrelevant studies were rejected first by examining titles and abstracts. After that, full texts of potentially eligible studies were obtained and reviewed according to inclusion and exclusion criteria. When the full text was not available, we contacted the authors to obtain raw data, and if the authors did not answer in two weeks, the study was excluded from the selection.

Data analysis

For selected articles, two researchers (LS and TM) extracted the following study characteristics: study design, country of study, year of publication, sample size and participant characteristics [such as age, sex and severity of COPD assessed by percentage (%) of predicted Forced Expiratory Volume in the 1st second (FEV1)], duration of the previous/initial PR, setting/design and duration of the maintenance PR programme, supervision and follow-up. Estimates of the association between the intervention parameters (PR programmes) and the study outcomes were measured as mean and standard deviation (SD) for all continuous outcomes. Missing SD was calculated as recommended in the 16.1.3.2 chapter of Cochrane Handbook for Systematic Reviews of Interventions[13]. Studies that do not report group means or MD were not included in the meta-analysis. Agreement between researchers was calculated with kappa statistics[13].

The meta-analysis was conducted using RevMan software v.5.4[21]. Due to the expected heterogeneity, we applied a random-effect model. For continuous outcome measures, data are presented as the mean and SD, and the effect size was estimated by the mean difference (MD) with 95% confidence intervals (CI). Missing SDs were arithmetically calculated[13]. Studies not reporting group means or MD were not included in the meta-analysis. Data not accessed through meta-analyses was summarised and described in the text.

In addition, there was, as we expected, heterogeneity due to the lack of evidence concerning the structure of maintenance programmes. Therefore, we performed sensitivity (leave-one-out) and subgroup analyses defined as a priori. We compared the effectiveness of the interventions between the duration of the initial PR programme, the severity of COPD (assessed by % of predicted FEV1, type of professional supervision and setting, risk of bias and publication year[2,4,11]. The guidelines recommend that these programmes last between eight and twelve weeks, therefore, we wanted to evaluate if the effects of a maintenance PR program were different in a twelve, eight or less than eight weeks programme[2,4]

The duration of the maintenance program and supervision by a health professional are aspects that vary substantially between existing programs and settings. We also wanted to adjust for the severity of COPD, according to % of predicted FEV1, and, finally, if the studies published before the 2013 guidelines[2] had different results. The methodological quality of the primary studies was also a reason for the analysis.

We used R[22-25] to perform a mixed-effects meta-regression to analyze the continuous variables in consideration of the presence of “residual heterogeneity”[26]. In addition, subgroup analysis was performed for all the categorical variables.

Assessment of the risk of bias

Two independent researchers (LS and TM) assessed the quality of the evidence for the collected outcomes of interest using the Grading of Recommendations, Assessment, Development and Evaluations (GRADE) system[13]. We appraised each study’s components, including selection, performance, detection, attrition, reporting, and other biases. For each study, either for the grading of each component and for the global study rating, we assigned categories of risk of bias: low, unclear, and high. The global grading involved taking an average of all individual components[13]. As blinding of allocation to patient and PR therapist was not possible, studies without any other source of bias were considered as being at the lowest risk possible for this type of intervention. We assessed publication bias with a funnel plot created in RevMan software v.5.4[21].

Table 1

Description of included studies.

AUTHOR, YEAR	STUDY TYPE	SAMPLE SIZE	PRE PR PROGRAMME DURATION (WKS)	INTERVENTION	FOLLOW_UP (MONTHS)	SUPERVISION	SETTING	DISEASE SEVERITY
BAULDOFF, G.[31]	RCT	I: 12 C:12	Not reported	Use of music during exercise training and walks	2	SV	HB	NR
BROOKS, D.[32]	RCT	I: 37 C:48	inpatients: 6; outpatients: 8	Home programme and phone call between visits	12	USV	HB/CB	Moderate to Severe
BUTLER, S.J.[37]	RCT	I:49 C:48	Not reported	Community-based maintenance exercise programme and recommendations for home exercise.	12	AS	CB	Moderate to Severe
COCKRAM, J. 2006 [45]	QE	230	8	Community classes for exercise trainning	12	SV	CB	NR
CRUZ, J.[38]	RCT	I: 16 C:16	12	Participants were given a piezoelectric pedometer, a log diary to record daily steps, and a calendar to register their short-term step-count goals and daily steps. Individual feedback is provided weekly (in the first month) and fortnightly (in the second and third months) by telephone calls. The final aim was to achieve the long-term goal.	3	AS	HB/CB	Mild to severe
MOULIN, M.[39]	RCT	I:10 C:10	3	Individualized training plan. Patients were instructed to quickly walk a distance equivalent to 125% of their last 6MWT three times a day, with each training walk not exceeding 15 min. Patients were given a pedometer so that the training could be better incorporated into daily activities in a homebased setting, and with a training diary. Telephone contacts for motivation.	6	USV	HB	NR
GALDIZ, J B.[40]	RCT	I:48 C:46	8	Three training sessions a week and four educational sessions	12	USV	Tele	Moderate to severe
GUELL, M-R.[41]	RCT	I:68 C:70	8	Home-based programme. Calls to patients every 15 days using a standardized protocol. Alternate week supervised training session.	36	AS	HB	Moderate to severe
JIMÉNEZ-REGUERA[42]	RCT	I: 20 C: 24	8	The HappyAir app has an educational programme and an integrated plan as a model of a therapeutic programme based on communication that introduced the figure of the therapeutic educator.	10	USV	Tele	Moderate to severe
LI, Y.[43]	RCT	I:65 C:69	8	Step 1: Home-visit once every 2 weeks for 2 months; Step 2: Home-visit every 4 weeks and phone contact once a week for 4 months; Step 3: Phone contact once a week for 6 months.	12	SV	HB	Mild to severe
MOULLEC, G.[47]	QE	I:14 C:26	4	Individualized exercise training (3.5 hours/week; 72 sessions) supervised by a teacher of adapted physical; health education provided alternatively by all professionals of the health care network (2 hours/month; 12 sessions) and psychosocial support with discussion group (1 hour/month; 12 sessions) supervised by a psychologist in the same room.	12	SV	CB	Moderate to severe
RIES, A.L.[44]	RCT	I:64 C:74	8	1-weekly telephone calls; 2-monthly supervised reinforcement sessions.	12	AS	HB/CB	Mild to severe
ROMÁN, M.[33]	RCT	I:26 C:22	12	Weekly-session maintenance programme.	12	SV	CB	Mild to severe
SPENCER, L.M.[35]	RCT	I:24 C:24	8	Supervised, outpatient-based exercise 1 day per week plus unsupervised home exercise on four other days.	12	AS	HB/CB	Moderate
SOUZA, Y.[34]	RCT	I:25 C:25	12	Patients received the manual for activities and were instructed to use it.	6	USV	HB	Mild to severe
WETERING[36]	RCT	I:102 C:97	12	Patients were instructed to perform the same exercises twice a day in their home environment Monthly encouragments to programme adherence and home training. Six extra training sessions in 3 weeks after exacerbations.	20	AS	HB	Moderate to severe
ZANABONI[46]	QE	10	4	Two-year telerehabilitation programme consisting of home exercise, telemonitoring and self-management, weekly supervised.	24	ASV	HB	Mild to severe

RCT Randomized controlled trial, QE quasi-experimental, I intervention, C control, SV supervised, USV unsupervised, AS Alt. supervision, HB home-based, CB community-based, HB/CB Alternate Home-based and community-based, Tele Telerehabilitation, NR not reported