Review Effect of pulmonary rehabilitation for patients with long COVID-19: a systematic review and meta-analysis of randomized controlled trials, 2025, Li+

Effect of pulmonary rehabilitation for patients with long COVID-19: a systematic review and meta-analysis of randomized controlled trials

Shige Li, Bing Dai, Yusheng Hou, Liang Zhang, Jie Liu, Haijia Hou, Dandan Song, Shengchen Wang, Xiangrui Li, Hongwen Zhao, Wei Wang, Jian Kang, Wei Tan

[Line breaks added]

Background
Pulmonary rehabilitation (PR) has demonstrated efficacy in managing long COVID-19, underscoring the need to refine and tailor PR strategies for optimal patient outcomes.

Objectives
To evaluate the impact of PR on patients with long COVID-19 and to compare the efficacy of different types and durations of PR interventions.

Design
Systematic review and meta-analysis.

Data sources and methods
We systematically searched randomized controlled trials (RCTs) of the effectiveness of PR in long COVID-19 patients published before April 2024. The primary outcomes were physical capacity assessed by the 6-minute walking test (6MWT), lung function measured by forced expiratory volume in the first second (FEV1) and forced vital capacity (FVC), health-related quality of life (HRQoL), and fatigue. Secondary outcomes were thirty-second sit-to-stand test (30STST), handgrip strength tests, maximal inspiratory pressure (MIP), maximal expiratory pressure (MEP), dyspnea, depression, anxiety, perceived effort, and adverse events.

Results
A total of 37 studies with 3363 patients were included. Compared to controls, PR improved physical capacity (6MWT, 30STST, handgrip), lung function (FEV1, FVC, MIP, MEP), HRQoL, fatigue, dyspnea, and anxiety but did not reach statistical significance for depression.

Subgroup analyses of PR duration indicated that programs of ⩽4 weeks improved 6MWT; those between 4 and 8 weeks significantly improved 6MWT, lung function (FEV1, FVC), HRQoL, and reduced fatigue; and programs over 8 weeks improved HRQoL and reduced fatigue.

Exercise type analysis revealed that breathing exercises improved 6MWT, lung function (FEV1, FVC), and HRQoL; multicomponent exercises enhanced 6MWT performance and reduced fatigue; the combination of both types improved 6MWT, FEV1 (L), FVC (%pred), HRQoL, and reduced fatigue.

Conclusion
PR improves physical capacity, lung function, and quality of life and alleviates dyspnea, fatigue, and anxiety in long COVID-19 patients. A 4- to 8-week PR program and a combination of both breath exercises and multicomponent training is most effective for managing long-term COVID-19 syndromes.

Trial registration
PROSPERO ID: CRD42024455008.

Link | PDF (Therapeutic Advances in Respiratory Disease) [Open Access]

 

The quality of RCTs was assessed using Cochrane Collaboration’s tool (RoB-2 tool)29 for risk of bias. This tool includes five domains: randomization process, deviations from intended interventions, missing outcome data, measurement of outcomes, and selection of reported results. Each domain was rated as “low,” “some concerns,” or “high.”

GRADEpro GDT was used to assess the certainty of the results32,33 according to the Grading of Recommendations, Assessment, Development, and Evaluations (GRADE) criteria. The evaluation included five domains: risk of bias in individual studies, inconsistency, indirectness, imprecision, and publication bias. The overall certainty of the evidence was rated as high, moderate, low, or very low.

(…)

Literature quality and bias assessment
The quality evaluation results of the RCTs are shown in Figure S1. Overall, 10 RCTs (27.0%)37,38,40,43,49,52,62,63,66,67 showed a high risk of bias, 19 (51.4%)21–24,26,39,42,44,45,48,51,54,56–60,64,65 showed some concern, and 8 (21.6%)25,41,46,47,50,53,55,61 had a low risk of bias. Among the included RCTs, 641,44,46,59,60,64 were performed with double blinding, and the others were performed with single blinding, no blinding, or had an unclear design; only 3 RCTs38,40,49 were inadequate for randomization; and 6 RCTs25,41,47,50,53,54included the intention to treat analyses of missing data, while the others did not.

(…)

There were several limitations to our review. First, the overall evidence GRADE was low to moderate, and bias in the included studies in terms of allocation concealment, blinding, and missing data may have led to an overestimation of the treatment outcomes. Second, due to heterogeneity among the included studies in terming of the exercise training types, PR duration, intensities, session frequencies, baseline disease severity levels, disease progression, comorbidities, and other factors, further subgroup analysis is limited. Third, as there were no standard measurement tools for HRQoL, fatigue, dyspnea, depression, or anxiety, SMD analysis was used, and TSA could not be performed on these results due to the large number of rating scales used in the RCTs. This may have led to an underestimation or overestimation of the treatment outcomes.

 

25 = REGAIN, unblinded and subjetive outcomes

46 = study on elderly hospitalised patients. ‘Blinding’ was not telling the patients if they were in the low (40-50 % of max HR) or moderate (50-70 % of max HR) intensity group, or the control group that did nothing in particular. In my opinion, that’s not adequate blinding.

I’m not sure I would trust their judgement if the assess REGAIN as low risk of bias..