Effects of Home-Based Pulmonary Rehabilitation on Dyspnea, Exercise Capacity, Quality of Life and Impact of the Disease in COPD Patients: A Systematic Review

Abstract

Conventional pulmonary rehabilitation programs are used as therapies for the treatment of chronic obstructive pulmonary disease (COPD). However, this modality presents barriers that make rehabilitation difficult. For this reason, home-based pulmonary rehabilitation (HBPR) has been used to overcome these barriers. The objective was to systematically compare a structured program with HBPR or a control group for participants with COPD. The primary outcome was an improvement in symptoms in the level of dyspnea and secondary outcomes were parameters in lung function, exercise capacity, health-related quality of life (HRQoL) and the impact of the disease on the individual. The Medline (via PubMed), Virtual Health Library and Cochrane Library databases were searched until May 10, 2021. Randomized controlled trials were included without restrictions on the year of publication or language. The risk of bias was evaluated using the Cochrane risk-of-bias tool for randomized trials (RoB). Our results showed that there was a significant decrease in the level of dyspnea, (MD: 5.46; 95% CI: 1.97 to 8.96), increased distance covered (MD: 61.75; 95% CI: 42, 94 to 80.56, significant improvement in HRQoL (MD: −11.30; 95% CI: −19.81 to −2.79) and reduction in the impact of the disease (DM: −4.71; 95% CI: −7.95 to −1.47). All results found were comparing the intervention group versus the control group. To conclude we found a reduction in the levels of dyspnea, an increase in the distance covered on the six-minute walk test, improving HRQoL and decreasing the impact of the disease in COPD patients in home-based pulmonary rehabilitation.

Supplemental data for this article is available online at https://doi.org/10.1080/15412555.2021.2020234 .

Keywords:

• Physical therapy modalities
• pulmonary disease
• chronic obstructive
• exercise tolerance

Introduction

Chronic obstructive pulmonary disease (COPD) is a common, preventable and treatable disease characterized by persistent respiratory symptoms and airflow limitation caused by airway and/or alveolar abnormalities [1]. The prevalence of COPD has increased worldwide, and the disease is now considered the third leading cause of death [2] and represents 3.64% of the total years lived with disability [3].

Dyspnea is the most common clinical manifestation of COPD [4]. In addition, symptoms promote progressive impairment of the ability to perform activities of daily living and physical activities, impairing sleep quality, and are associated with a decline in health-related quality of life (HRQoL) [5]. One should use the best tools available to diagnose and evaluate COPD and combine pharmacological and no pharmacological measures, such as pulmonary rehabilitation, for effective management [6].

Previous studies have shown that at any stage of COPD, physical training performed in the context of a pulmonary rehabilitation program is effective in improving exercise tolerance, muscle strength, physical capacity, HRQoL, reducing levels of dyspnea and fatigue, and presents the potential to restore mental and social functioning [7–10]. In the context of pulmonary rehabilitation, several items are recommended, such as warm-up exercises, strengthening of the upper limbs and lower limbs, aerobic exercises, and stretching. In addition, energy conservation techniques and information about the disease and health self-management are necessary [11–13].

One of the major obstacles to pulmonary rehabilitation is the difficulties that patients have in accessing conventional pulmonary rehabilitation (CPR) programs [12]. Such difficulties occur around trips to rehabilitation centers, lack of transport, diseases and comorbidities, scarcity of programs, mainly in rural areas and insufficient number of health professionals [14–16].

To eliminate these obstacles, home-based pulmonary rehabilitation (HBPR) has been used, since it is an acceptable, easy and accessible method that is addressed at home. The activities are the same as those performed in ambulatorial rehabilitation and can result in clinical benefits equivalent to those obtained through CPR for COPD patients [17].

From this perspective, in 2014, a systematic review showed that HBPR programs are effective in relieving respiratory symptoms in COPD patients, improving HRQoL and exercise capacity [18]. However, some methodological issues related to the quality of the included studies with small sample sizes, study randomization process and descriptions of the allocation concealment unclear highlight the importance of a new systematic review in this field. Determining responses to HBPR will help us to assist with an individualized prescription recommendation for exercise in COPD patients.

Thus, the main objective of this systematic review and meta-analysis was to synthesize the efficacy of HBPR compared to CPR or without rehabilitation therapy in reducing the levels of dyspnea. The secondary outcomes were improvements in lung function, exercise performance, endurance, muscle strength, health-related quality of life and disease impact.

Methodology

Protocol and registry

The review methodology was registered in the International Prospective Register of Systematic Reviews (PROSPERO; CRD42020151669) and followed the recommendations of the Cochrane Handbook [19] and PRISMA statement [20].

Eligibility criteria

We considered randomized clinical trials (RCTs) that compared a structured HBPR with CPR or without rehabilitation therapy. Inclusion criteria: studies on individuals with a diagnosis of COPD (mild, moderate, or severe) confirmed by spirometry [21, 22] (American Thoracic Society [23], British Thoracic Society [24], or Global Initiative for Chronic Obstructive Lung Disease [6]), over the age of 20 years old; Studies that exposure included any type of activity, including but not limited to aerobic and resistance training/activities Exclusion criteria: Studies including a population diagnosed with COPD associated with another respiratory disease, such as asthma, were excluded if the data could not be obtained separately; study maintenance pulmonary rehabilitation; papers which clearly demonstrate that patients attended less than 70% of the total rehabilitation sessions. In addition, duplicate studies, conference abstracts, letters to the editor, case reports, studies with a design incompatible with the purpose of this review and those that did not clarify the interventions used were excluded.

Search strategy and selection of articles

Searches were performed from inception until October 28, 2019 (updated May 10, 2021), without date limits or language restrictions, in the following databases: Medline via PubMed (www.pubmed.gov), Lilacs via Virtual Health Library (www.bvsalud.org) and CENTRAL via Cochrane Library (www.cochranelibrary.com). The ClinicalTrials.gov was assessed to identify potential ongoing studies. Moreover, the initial search was complemented by a manual search of reference lists from retrieved articles. To ensure literature saturation, comprehensive searches were constructed of both index terming’s (MeSH terms), “free text” terms, and synonyms. The search strategies and filters applied are described in Appendix 1, Supplementary Material. To avoid publication bias, the gray literature (Google Scholar) was also consulted.

The electronic searches were performed by two independent authors (DMX and AAF). Articles found in the databases were exported to Rayyan, a web and mobile app for systematic reviews (http://rayyan.qcri.org), for management and the removal of duplicates and selection of manuscripts. First, the manuscripts were selected by titles and abstracts based on the eligibility criteria. Then, the selected studies were read in full to confirm their eligibility. Disagreements on the inclusion process were resolved by a third reviewer (VPL).

Data extraction

Data were extracted from the included studies by two independent reviewers (DMX and AAF). Any disagreements that arose between the reviewers were resolved through discussion with a third reviewer (VPL). Authors of papers were contacted to request missing or additional data when necessary. The following information was extracted: Condition of participants (COPD classification and smoking participants); Symptoms (subjective ratings of exertional symptoms; i.e. dyspnea, secretion, cough, wheezing); Cardiorespiratory and metabolic responses (heart rate (HR), breathing frequency, peripheral capillary oxygen saturation (SpO2%), VO2, carbon dioxide production (VCO2) and concentration of blood lactate. Forced expiratory volume in 1 s (FEV1), forced vital capacity (FVC), inspiratory and expiratory maximal pressure (MIP and MEP), inspiratory capacity, residual volume and total lung capacity (TLC)); Physical capacity (distance walked on the six-minute walk test (6MWT)); Composite mortality indices (BODE index (iBODE) variation or other indices with similar properties); Health status (health-related quality of life); and Impact of COPD (COPD Assessment Test (CAT)).

Risk of bias in individual studies

Assessment of the methodological quality of the studies was performed using the Cochrane risk-of-bias tool for randomized trials. Based on this tool, the following criteria were evaluated: random sequence generation, allocation concealment, blinding of outcome assessor, incomplete outcome data, and selective outcome reporting. The domain blinding of patients and personnel was not considered for analysis once it is not possible to blind patients and staff. The quality assessment was guided by Rev-Man 5.3 software (Cochrane Collaboration, Oxford, UK).

Summarized measures and synthesis of results

When appropriate, we performed a meta-analysis using R software, version 3.6.2 (meta package). The meta-analysis was considered when at least two studies reported sufficient data for the same outcome in the exposure and control groups.

The data referring to mean difference and standard deviation (SD) from baseline to post-intervention follow-up were collected. Effect sizes were expressed as weighted (or standardized) final post-intervention mean differences (for continuous data) and their 95% confidence intervals. Heterogeneity was assessed statistically using the I² tests: random effects meta-analysis was used if I² ≥ 30% and fixed effects models if I² < 30%. Subgroup analyses were conducted where there were sufficient data to investigate the impact between the types of rehabilitation programs (HBPR, CPR and no rehabilitation) regarding the follow-up period. Funnel plots to assess publication bias were not generated once there were 10 or fewer studies included in the meta-analyses. In addition, descriptive analyses are reported in tables.

Reliability of cumulative evidence

The Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach for grading the certainty of evidence and a Summary of Findings (SoF) were created using GRADEPro GDT (McMaster University, ON, Canada) for each outcome that allowed meta-analyses. The ranking of the quality of the evidence was based on five criteria: risk of bias, inconsistency, indirect evidence, imprecision, and publication bias. The SoF tables were generated at the end of the analysis according to the certainty of the evidence among the studies and classified as high, moderate, low, or very low certainty of the evidence.

Results

In total, 1570 articles were found from electronic databases, manual searches and gray literature. After screening and selection process based on eligibility criteria, 31 studies [12, 25–54] were included in qualitative analysis involving 2531 COPD patients. However, only 17 articles (including 898 COPD patients) provided enough information to be included in the meta-analyses [26, 28–33, 35, 37–39, 41, 42, 44, 46, 52, 53]. Figure 1 presents the flowchart of the study selection process.

Figure 1. Flow chart of study selection.

Description of the studies

The studies included were published between 1996 and 2021. The pulmonary rehabilitation program duration ranged from five weeks to a maximum of 18 months. The sample size of the included studies ranged from 20 to 305 patients, the mean age ranged from 56.69 to 76.70, and the mean FEV1 ranged from 23.21% predicted to 68.00% predicted (moderate, severe, and very severe). Among the studies, 27 reported data on the participants’ sex [12, 25–27, 29–34, 36, 37, 40–43, 45–54] 1159 males and 765 females with COPD (proportion of 1.65:1). In addition, 16 studies [26, 30, 32, 36, 42–49, 52, 54] reported the number of participants with COPD (n = 612) who still had smoking habits.

HBPR was considered a therapy based on any type of strength and resistance exercise for upper limbs, lower limbs, trunk and also cardiopulmonary exercises and specific exercises for respiratory muscles. As the activities were carried out at home, the HBPR could be supervised through calls and/or visits by the study team, as a way to verify compliance with what was proposed to the participants. In addition, health education and self-management of the disease can also be part of the HBPR, as it contributes to the participants’ autonomy process. Participants who were allocated to the CPR group perform the same strength and/or resistance exercises proposed for the HBPR group, however, the place of performance was in a clinic/specialized outpatient clinic. The control group received health education sessions and standard medical follow-up.

The main characteristics of the included studies, characteristics of the participants, interventions and clinical characteristics are presented in Tables 1–4, respectively.

Table 1. Main methodological characteristics of the included studies.

Year, Author	Country	Randomized (Y/N)	Follow-up	Inclusion criteria	Exclusion criteria
1996, BAULDOFF; ZULLO; HOFFMAM	USA	Y	8 weeks	Diagnosis of COPD (FEV1 / FVC <60% of predicted) Clinically stable Not having been hospitalized for 6 weeks or more before data collection Never having received training with arm exercises Without unstable heart disease (eg: recent myocardial infarction, uncontrolled hypertension, severe congestive heart failure) Not having musculoskeletal disability	NR
1996, STRIJBOS; ALTENA; KOETER	Netherlands	Y	6, 12 and 18 months	COPD, diagnosed by history, physical examination, radiography and results of pulmonary function tests; Dyspnea on exertion, which limits activities of daily living; PaCO2 at rest <6.5 kpa, and Pa02 at rest> 7.5 kpa; Post-bronchodilation FEV 1 between 600 and 1800 ml and <65% of predicted	Ischemic heart disease, musculoskeletal disorders, or other disabling diseases that may restrict therapy rehabilitation.
1996, WIJKSTRA et al	Netherlands	Y	3 months	FEV1 < 60% predicted; FEV1 / vital inspiratory capacity (VCI) <50%, both after two inhalations of 40 μg of ipratropium bromide; Severe airway obstruction (mean FEV1 44% pred, mean FEV1 / CVI 38%) Reversibility less than 10% (mean increase in FEV1 of 0.15 L, 5% predicted)	Patients with evidence of myocardial ischemia, intermittent claudication, musculoskeletal disorders or other disabling diseases.
2000, HERNANDEZ et al	Spain	Y	3 months	Diagnosis of COPD Stable disease with drug optimization FEV1 < 60% of predicted time Being a former smoker or having previously been included in a smoking cessation program that resulted in abstinence	Patients with evidence of severe or uncontrolled ischemic heart disease, systemic arterial hypertension, changes in the rib cage, neuromuscular disorders or intermittent claudication or osteoarticular lesions in the lower extremity that could affect normality Patients with acute exacerbation during the program
2003, SINGH et al	India	Y	1 month	Stable COPD Dyspnea in three or more daily activities been Never involved in a pulmonary rehabilitation program Had given up smoking for at least two months	Patients with right ventricular failure, unstable ischemic heart disease, oxygen saturation less than 88% at rest, musculoskeletal diseases, acute exacerbation and pneumothorax
2003, OH	Korea	Y	2 months	Diagnosis COPD Less than 85% of predicted FEV1 No exacerbation evidenced by physician No evidence of ischemic heart disease, musculoskeletal disorders, or other disabling diseases that could restrict the rehabilitation therapy	NR
2005, BOXALL et al	Australia	Y	3 months	Diagnosed with COPD Age> 60 years, Dyspnea on exertion Having no worsening of the symptoms of the disease during the last 2 weeks Motivated to run the program daily without supervision. Patients functionally confined at home	Patients who are already part of an outpatient program Patients with restricted shoulder movement Nursing home residents Patients with previous lung volume reduction surgery Patients with pain limiting mobility
2005, MURPHY; BELL; COSTELLO	USA	Y	6 weeks	Moderate to severe COPD (FEV1 < 60%) Long-standing smoking	Congestive heart failure, pneumonia, pneumothorax, pleural effusion, pulmonary embolism and arrhythmia
2006, KOPPERS et al	Netherlands	Y	5 weeks	Chronic airflow obstruction, defined as FEV1/FVC ratio ≤70%, FEV1 of 30 to 80% predicted, after bronchodilatation Stable clinical condition for at least 6 weeks	hypoxemia at rest or during exercise cardiac or orthopedic disease body mass index (BMI)≥30 kg/m2
2007, RESQUETI et al	Spain	Y	2 months–6 months	Age <80 years Dyspnea: MRC> 3 Difficulty going to the hospital due to severe shortness of breath or problems related to the place of residence	Heart disease or any other type of disease that limits exercise tolerance Mental disability that prevents participation
2007, SHEA; TAYLOR; PARATZ	Australia	Y	3 months	Diagnosis of COPD Written consent Not having participated in pulmonary rehabilitation in the last 12 months	Breathing problems other than COPD, or unstable medical conditions that limit the performance of progressive resistance exercise
2008, MOULIN et al	Spain	Y	3 weeks	Diagnosis of COPD Not having participated in pulmonary rehabilitation in the last 12 months	Patients unable to perform physical training due to heart problems, lung or orthopedic problems
2008, MALTAIS et al	Canada	Y	3 months	Stable COPD, that is, no change in medication and symptoms (dyspnea, volume, or color of sputum) for at least 4 weeks before the study; were 40 years or older; were current or former smokers of at least 10 pack-years (20 cigarettes per pack); had an FEV1 less than 70% of the predicted value and FEV1–FVC ratio less than 0.70; and had a Medical Research Council dyspnea score of at least 2.	A previous diagnosis of asthma, congestive left heart failure as the primary disease, a terminal disease, dementia, or an uncontrolled psychiatric illness. We sought to study a broad COPD population and did not exclude patients with oxygen dependence or other comorbid conditions.
2009, FERNANDEZ et al	Spain	Y	12 months	Diagnosis of very severe COPD Younger than 80 years of age Stable COPD The correct administration of pharmacological treatment according to the recommendations of the Global Initiative for Chronic Obstructive Lung Disease Guide2 Home treatment with oxygen for at least 6 months prior to the commencement of the study absence of severe cardiovascular pathology, unstable angina or acute myocardial infarction, cerebral vascular accident, or physical or psychological disorder that impedes the practice of physical exercise	NR
2010, OLIVEIRA et al	Brazil	Y	2 months	COPD based on the GOLD classification Clinical stability in the eight weeks prior to the study	Hospitalization COPD instability Presence of neuromuscular disease, other associated respiratory disease, orthopedic or neurological disease that affects gait Recent impairment due to comorbidities, such as myocardial infarction, heart failure, stroke or neoplasia Previous pneumonectomy or other thoracic surgery
2010, GHANEM et al	Egypt	Y	2 months	Arabic as the first language Age> 40 years Local district resident Ability to complete the CRQ in one session	Patients unable to read or write; Patients with locomotor problems Cognitive impairment, ischemic heart disease, aortic valve disease, cancer or lung diseases other than COPD
2012, AKINCI, OLGUM	Turkey	Y	3 months	COPD GOLD 3 or 4 Clinically stable No history of infections or exacerbation of respiratory symptoms No change in medication during the 2 months prior to the study	Myocardial infarctions in the previous 4 months Unstable angina, severe congestive heart disease, severe hypertension problems, diabetes mellitus with complications and muscle and joint problems Cancer and asthma patients
2013, DIAS et al	Brazil	Y	2 months	Both sexes Age> 40 years COPD diagnosed by GOLD guidelines Clinically stable No signs of disease exacerbation in the past 30 days Do not participate in any program that involves simultaneous physical activity	Patients with severe comorbidities, such as previous heart disease, orthopedic diseases in the upper and lower limbs, motor sequelae of neurological or visual disorders that could interfere with physical exercise, Uncontrolled SAH Other respiratory diseases Patients who did not agree to participate in the study
2014, PINTO et al	Spain	Y	12 weeks	Patients with severe or very severe COPD with or without the use of oxygen Age <80 years of age Ex-smokers; Never having participated in any rehabilitation	Long period of exacerbation Cognitive problems, musculoskeletal disorders, heart disease or other disabling diseases that could restrict exercise Communication problems
2014, TABAK et al	Netherlands	Y	9 months	No exacerbation in the month prior to enrolling the program Three or more exacerbations or hospitalization for respiratory problems in the 2 years prior to entering the study Clinical diagnosis of COPD according to GOLD. No exacerbation in the month prior to enrollment > 3 exacerbations, defined as respiratory problems that required a course of oral corticosteroids and / or antibiotics, or> 1 hospitalization for respiratory problems in the 2 years prior to the start of the study Ex-smoker Age 40 to 75 years Post-bronchodilator forced expiratory volume in 1 s (FEV 1) 25 to 80% of predicted Able to understand and read Dutch Written informed consent of the subject before participation Having a computer with Internet access at home	Another serious illness with a low survival rate; Diseases with respiratory symptoms or that compromise lung function (for example, heart failure or sarcoidosis); erious psychiatric illness Uncontrolled diabetes mellitus during a past COPD exacerbation or hospitalization for diabetes mellitus in the 2 years preceding the study; Need for regular oxygen therapy (> 16 h / day or oxygen tension <7.2 kpa); Maintenance therapy with antibiotics; Deficiency of α −1 antitrypsin, disorders or progressive diseases that seriously influence the ability to walk (eg, amputation, paralysis, progressive muscle disease)
2016, HOLLAND et al	Australia	Y	8 weeks–12 months	Diagnosis of COPD Smoking history ≥ 10 pack-years No exacerbations in four weeks Comorbidities that prevented physical training Not having undergone pulmonary rehabilitation within 2 years	NR
2016, TSAI et al	Australia	Y	2 months	Primary medical diagnosis of stable COPD (forced expiratory volume in 1 s (FEV1) / forced vital capacity (FVC) <70% and FEV1 < 80% post-bronchodilator predicted)	Not being interested in physical training; Not having access to a computer; Prefer to attend outpatient pulmonary rehabilitation; Mini mental status exam <24, poor hearing; Vision problems; Do not speak English
2017, HORTON et al	England	Y	7 weeks	Dyspnea (MRC) 2–5 Proficient level of English	Any condition that prohibits rehabilitation such as inability to use the lower limbs and other unstable comorbid conditions Participants who have completed another rehabilitation program in the past 12 months
2017, VASILOPOULOU et al	USA	Y	12 months	Written informed consent obtained before any assessment is carried out; Male and female patients ≥ 40 years’ old; Diagnosis of COPD [forced expiratory volume post-bronchodilator in the first second (FEV1) <80% of predicted and FEV1 / forced vital capacity (FVC) <75% without significant reversibility of the post-bronchodilator (<10% of FEV1% of normal foreseen); Ideal medical treatment according to GOLD, without regular use of systemic corticosteroids; Smokers or ex-smokers with a smoking history equivalent to at least 10 packs years (1 pack year = 20 cigarettes smoked per day for 1 year); Absence of other significant illnesses that may contribute to exercise limitation and at least 2 exacerbations of COPD in the year prior to entry time; No exacerbation of COPD in the year prior to the time of entry	Orthopedic, neurological or other complaints that significantly compromise normal biomechanical movement patterns, as judged by the investigator; Specifically, if the patient’s condition / comorbidities are such that physical activity cannot be increased, respiratory diseases other than COPD (for example, asthma), cognitive impairment of reading and / or difficulties in managing electronic devices, preventing interaction with the tablet, as judged by the investigator; Patients who do not use ideal pharmacotherapy
2018, COULTAS et al	USA	Y	5 months	Medical diagnosis of COPD confirmed by spirometry (FEV 1 / FVC <70% and FEV 1 < 70% predicted) Did not participate in pulmonary rehabilitation	NR
2019, LAHHAM et al	Australia	Y	2 months	Smokers or ex-smokers of at least 10 pack-years; Age> 40 years COPD diagnosis confirmed by spirometry	Previous diagnosis of asthma; Participated in a pulmonary rehabilitation program in the past two years; COPD exacerbation in the past four weeks; Has comorbidities that prevent participation in a physical training program
2019, BURGE et al	Australia	Y	8 weeks 12 months	NR	NR
2020, LAHHAM et al	Australia	Y	2 months	Potential participants were people diagnosed with spirometrically defined mild COPD (forced expiratory volume in one second (FEV1)/forced vial capacity (FVC) < 70%; FEV1 > 80%predicted), aged 40 years old or older, had a smoking history of at least 10 packet years and had no reported hospitalizations or exacerbations during the month before recruitment.	formal diagnosis of asthma or co-morbidities that precluded exercise participation
2020, JIANG et al	China	Y	6 months	aged 60 years and older, confirmed diagnosis of COPD according to the diagnosis and treatment guidelines for chronic obstructive pulmonary disease, forced expiratory volume in 1 second (FEV1)/forced vital capacity (FVC) ratio of <0.7, FEV1 < 80% predicted, and use of WeChat for effective communication	patients with mental disorders, cognitive disorders, and limb dysfunction; patients with unstable heart disease or arrhythmia requiring drug intervention; patients with a history of myocardial infarction or cerebral infarction in the previous year; patients too weak to perform the muscle strength test; patients with hypertension that could not be controlled with drugs; and patients with a history of syncope after exercise
2020, HANSEN et al	Denmark	Y	10 weeks	Adults with a clinical diagnosis of COPD defined as FEV1/FVC <0.70, FEV1 < 50%, Medical Research Council ≥2 and no participation in rehabilitation within 6 months of the start of intervention	NR
2021, KO et al	Hong Kong	Y	12 months	NR	NR

NR: Not reported; NA: Not available; ^†Data extracted from abstracts; MRC: Medical Research Council; MMRC: modified Medical Research Council; COPD: chronic obstructive pulmonary diseases; GOLD: Global Initiative for Chronic Obstructive Lung Disease; USA: United States of America.

Table 2. Characteristic of the interventions and the population enrolled in included studies.

Year, Author	Age (mean, years)	Age (range)	Gender male/female	Smoker (Y/N)	Heart rate	Breathing frequency	Peripheral capillary oxygen saturation (SpO2%)	VO2	Carbon dioxide production (VCO2)	Concentration of blood lactate	COPD Gold classification	Symptoms: secretion (Y/N)	Symptoms: Cough (Y/N)	Symptoms: Wheezing (Y/N)	Description of Home-based rehabilitation programme intervention	Description of conventional supervised pulmonary rehabilitation
1996, BAULDOFF; ZULLO; HOFFMAM	62	47 − 76	9/11	NR	NR	NR	NR	NR	NR	NR	Grave	NR	NR	NR	Training program: upper extremity training 5 days a week for 8 weeks with the training level increased during weekly visits	standard medical treatment
1996, STRIJBOS; ALTENA; KOETER	61.4	53 − 67	38/7	Y: 17 N: 28	NR	NR	B: G1: 95.1 G2: 93.8 G3: 93.7 Post G1: 94.1 G2: 93.2 G3: 94.4 6 G1: 94.6 G2: 93.4 G3: 93.8 12 G1: 94.3 G2: 92.9 G3: 93.5 18 G1: 94.0 G2: 92.9 G3: 92.7		NR	NR	Moderate – severe - Very severe	NR	NR	NR	Training program: 24 sessions of 30 min in 12 consecutive weeks with 30 min on exercise days and at least 15 minutes on other days. monitoring three times a week for monitoring and motivation	standard medical treatment
1996, WIJKSTRA et al	63	57 − 69	37/6	NR	B G1: 120 G2: 126 P G1: 122 G3: 121	B G1:30 G2: 29 P G1: 29 G3:28	NR	B G1:1.0 (0.3) G2:1.1 (0.3) P G1:1.1 (0.4) G2: 1.0 (0.3)	B G1: 0.9 (0.4) G2: 1,0 (0.4) P G1: 1.0 (0.4) G2: 0.9 (0.3)	B G1:3,7 (1.7) G2:4,1 (1.9) P G1:3,1 (1.3) G2:3.4 (1.7)	Moderate– severe	NR	NR	NR	Training program: relaxation exercises, breathing techniques, UL training, target flow inspiratory muscle training (IMT) and physical training at home.	Did not participate in the rehabilitation program and received a diary to record the time spent on daily activities
2000, HERNANDEZ et al	64.3	56 − 72	NR	NR	NR	NR	NR	NR	NR	NR	Moderate – severe - Very severe	NR	NR	NR	Training Program: exercises for LL, through walking. With intensity set at 70% of the maximum speed reached in the swt. Sessions lasting 1 h, 6 days / week, at home, with checkup every 2 weeks	standard medical treatment
2003, SINGH et al	59.37	48–75	32/8	NR	NR	NR	NR	NR	NR	NR	Severe	NR	NR	NR	Training program: in removal of secretions, lower extremity exercise, breathing strategies, energy conservation and work simplification. Duration: 30 minutes twice a day Follow-up: supervised weekly	No intervention
2003, OH	65.80	54–79	14/9	Y: 2 N: 21	NR	NR	NR	NR	NR	NR	Moderate - severe –	NR	NR	NR	Training program: education, inspiratory muscle training, exercise training, and psychosocial components that include relaxation and telephone calls.	received only educational advice at the initial interview.
2005, BOXALL et al	76.7	67 − 85	26/20	NR	NR	NR	NR	NR	NR	NR	Severe – very severe	NR	NR	NR	Multidisciplinary education on COPD and its management UL training program: from level 1 (6 repetitions of each weightless exercise) to 18 (3 sets of 6 repetitions of each exercise, 1.5 kg weights in both hands) Walking program from level 1 (walking on level ground for 1-minute, maximum rest, 2 minutes and walking again for 1 min) to 10 (walking for 15 min, rest and then walk for 15 min	standard medical treatment and orientations
2005, MURPHY; BELL; COSTELLO	66	54 − 76	17/9	Y:9 N: 19	NR	NR	NR	NR	NR	NR	Moderate – severe	NR	NR	NR	Training program: follow-up for 6 consecutive weeks, with 12 exercise sessions lasting 30 to 40 min, including rest. Aerobic exercises: climbing up and down a ladder and sitting up in a chair. UL training: Thera-Band. Exercise for at least 15 min on other days according to received diary	standard medical treatment
2006, KOPPERS et al	55.7	38–73	17/19	NR	B G1: 150 (17) G3: 144 (17) P G1:135 (16) G3:142 (17)	NR	NR	B G1: 1.52 (0.41) G3: 1,52 (0.42) P G1: 1.42 (0.41) G3: 1.50 (0.40)	NR	NR	Moderate – grave	NR	NR	NR	Respiratory muscle endurance training (RMET) by means of normocapnic hyperpnea. trained twice daily for 15 min, 7 days per week, for 5 weeks	Standard medical treatment sham training
2007, RESQUETI et al	68	61 − 75	35/3	NR	NR	NR	NR	NR	NR	NR	Severe – very severe	NR	NR	NR	Training for 9 weeks. • 1st week: 1 hour of patient education and 30 min of physical therapy • 2nd week: 3 sessions at the clinic to learn how to do the exercises that they should continue doing at home. 3rd to 9th week: program at home 5x per week for a period of 1.5 hours performing exercises for lower limbs, upper limbs and inspiratory	Training for 9 weeks. • 1st week: 1 hour of patient education and 30 min of physical therapy 2nd week: perform respiratory physiotherapy exercises at home and walk, without any supervision
2007, SHEA; TAYLOR; PARATZ	67.65	59 − 78	NR	NR	NR	NR	NR	NR	NR	NR	Moderate- severe	NR	NR	NR	Training program: 6 exercises with progressive load for 3x per week for 12 weeks, being 1x per week at the clinic and 2x at home. Exercises: standing hip abduction, sitting and standing, with 3 sets of 8 to 12 repetitions, with Theraband.	Did not receive intervention
2008, MOULIN et al	67	63 − 77	14/6	Y:2 N: 18	NR	NR	NR	NR	NR	NR	Moderate	NR	NR	NR	Training program: walking distance equivalent to 125% of your last 6MWT, 3x / day for 15 min. Daily and telephone follow-up every 4 weeks	Perform your daily life activities
2008, MALTAIS et al	66	NR	140/100	Y: 126 N: 114	NR	NR	NR	NR	NR	NR	Moderate – severe	NR	NR	NR	The self-management educational program “Living Well with COPD” consisted of an educational flipchart and 6 skill-oriented, self-help, patient workbook modules. The program was provided in the hospital on an outpatient basis. A qualified exercise trainer presented the exercise module.	The training program combined aerobic and strength exercises. Briefly, the aerobic training consisted of stationary leg cycling for 25 to 30 minutes in each session. The target training intensity was 80% of peak work capacity during incremental exercise.
2009, FERNANDEZ et al	66	51–79	41/0	NR	NR	NR	NR	NR	NR	NR	Moderate- severe – very severe	NR	NR	NR	Training program: 2 one-hour sessions in the Hospital: respiratory reeducation, taught the exercises at home with educational materials. Muscular training of the UL, initially through isotonic exercises without weights and subsequently a progressive weight increases 30 min. Muscular training of the LL, initially through isotonic exercises and subsequently resistance exercises	standard medical treatment and orientations
2010, OLIVEIRA et al	69.5	56 − 79	65/20	NR	NR	NR	NR	NR	NR	NR	Moderate – severe – very severe	NR	NR	NR	Exercises to strengthen the UL and LL. Aerobic exercise was performed by walking on flat ground, from 60 to 80% of the maximum heart rate achieved on the 6MWT.	Control group: standard medical treatment; Outpatient group: received the same training program as home rehabilitation
2010, GHANEM et al	56.69	45 − 68	NR	NR	NR	NR	NR	NR	NR	NR	severe – very severe	NR	NR	NR	Education of Patient: 4 days for 1 hour Exercise program: Respiratory muscle training, aerobic training (walking and cycling), UL strength training and hamstring, quadriceps, calf, shoulder, neck and lower back stretching	standard medical treatment
2012, AKINCI, OLGUM	71.8	55 − 79	NR	NR	NR	NR	B: G1: 93.5 (3.7) G3: 94.8 (3.4) P: G1: 94.8 (3.4) G3: 94.3 (3.3)	NR	NR	NR	severe - very severe	NR	NR	NR	Patient education: 2 to 3x at home for 2 to 3 hours. Training program: mmii aerobic exercise (walking) and UL and breathing exercises (lips and diaphragmatic breathing)	standard medical treatment
2013, DIAS et al	65.25	58 − 72	8/15	NR	NR	NR	NR	NR	NR	NR	Mild - Moderate - severe	NR	NR	NR	Education program Training program: stretching, respiratory reeducation exercises and aerobic exercises for the UL (with individual load determined at 50% of the load obtained after the evaluation)	Instructions for performing breathing and stretching exercises
2014, PINTO et al	70.2	61 − 78	39/2	NR	NR	NR	NR	NR	NR	NR	Severe - very severe	NR	NR	NR	Weekly visits and monitoring by phone Training program: breathing exercises, resistance (UL and LL) resistance (walking for 30 minutes, climbing stairs, cycling and walking on a treadmill or stationary bicycle from 5 to 30 min) and stretching. Exercises coordinated with breathing patterns and rest time. Patients on long-term oxygen treatment were instructed to exercise using flow to maintain saturation above 90%	standard medical treatment
2014, TABAK et al	63.45	55 − 73	12/12	Y:8 N: 16	NR	NR	NR	NR	NR	NR	mild -moderate – severe- very severe	31	77	61	4 modules: 1st: training in real time using the activities present in the activity diary 2nd: web exercise program for home exercises 3rd: self-management of exacerbations of COPD through a screening diary 4th: teleconsultation	standard medical treatment
2016, HOLLAND et al	69	56 − 82	99/67	Y: 28 N: 138	NR	R	NR	NR	NR	NR	Moderate – severe	NR	NR	NR	Training program: 30 min aerobic activity, resistance training (functional activities). Motivational follow-up 1x per week.	Outpatient group: received the same training program as home rehabilitation.
2016, TSAI et al	74 (8)	66 − 82	NR	Y: 2 N: 34	NR	NR	NR	NR	NR	NR	mild -moderate – severe	NR	NR	NR	Training program: 3x a week, for 8 weeks, with LL ergometry, walking and strengthening exercises Remote monitoring by videoconference	standard medical treatment
2017, HORTON et al	68 (8.86)	59 − 77	100/187	Y: 73 N: 214	NR	NR	G1: 94.54 G2: 93.91	NR	NR	NR	Moderate – severe	NR	NR	NR	Training program: walking every day of the week and resistance exercises 3x a week. Motivational follow-up	Outpatient group: supervised exercise and twice a week for 7 weeks
2017, VASILOPOULOU et al	66.9 (9.6)	57 − 76	119/28	Y: 14 N:147	NR	NR	NR	NR	NR	NR	Moderate – severe – very severe	NR	NR	NR	12-Month training program: • Individualized action plan • Educational session on self-management of the disease; • Physical exercise sessions • Access to the call center; remote monitoring	Outpatient group: visit to the hospital twice a week for 12 months where he received the same home rehabilitation training program
2018, COULTAS et al	70.3 (9.5)	60–79	151/154	Y: 282 N: 23	NR	NR	NR	NR	NR	NR	Moderate – severe	NR	NR	NR	Training program: 30 min of moderate intensity physical activity Motivational follow-up	Education advice on self-management and usual care
2019, LAHHAM et al	68 (8)	59–77	23/22	Y:7 N: 45	NR	NR	NR	NR	NR	NR	Moderate – severe	NR	NR	NR	Training program: resistance and strengthening exercises for at least 30 min a day, starting at an intensity of 80% of the 6MWT Motivational follow-up	Outpatient group: received the same training program as home rehabilitation
2019, BURGE et al	70	60 − 80	63/96	NR	NR	NR	NR	NR	NR	NR	Moderate – severe – very severe		NR	NR	Training program: aerobic exercises for 30 min, exercises to strengthen the lower limbs and upper limbs; Patient education on self-management; Motivational follow-up	Outpatient group: received the same training program as home rehabilitation
2020, LAHHAM et al	68 (9)	57–77	34/24	NR	NR	NR	NR	NR	NR	NR	Mild	NR	NR	NR	endurance training programme calculated at 80% of initial walking speed from the baseline (6MWT) Resistance training comprised upper and lower limb exercises using equipment available at home (home stairs for step ups and sealed water bottles as weights)	advised to keep active and to follow their medication
2020, JIANG et al	71.83 (7.60)	64–79	87/19	Y: 12 N: 94	NR	NR	NR	NR	NR	NR	Moderate – severe – very severe	NR	NR	NR	Exercise training included upper extremity training, lower extremity training, and balance training; Respiratory training included lip contraction breathing and abdominal breathing	other activities, but in a clinic
2020, HANSEN et al	68.3 (9)	NR	60/74	Y: 30 N: 104	NR	NR	NR	NR	NR	NR	Moderate – severe	NR	NR	NR	It was a group-based, supervised and standardized programme performed by the patients in their homes three times weekly for 10 weeks via a videoconference software system installed on a single touch screen. The exercise sessions lasted 35 min with incorporated warm-up and high repetitive time-based muscle endurance training followed by 5 min’ rest before beginning a patient education session of 20 min.	The exercise sessions lasted 60 min and incorporated warm-up, endurance and resistance training and a cool-down period
2021, KO et al	75 (6.70)	68–82	132/0	NR	NR	NR	NR	NR	NR	NR	Moderate – severe	NR	NR	NR	Our case manager provided support and reinforcement via telephone to the subject every fortnight on continuing exercise at home for1 year	received neither physiotherapy training nor phone calls from the case manager for reinforcement of home exercise

NR: Not reported B: Baseline; P: Post-intervention; LL: lower limb; UL: upper limb; 6MWT: 6-minute walk test; SWT: Shuttle Walk Test; COPD: chronic obstructive pulmonary disease; Min: minutes; G1: home rehabilitation; G2: conventional outpatient rehabilitation; G3: no rehabilitation/ control group.

Table 3. Clinical characteristics of the population enrolled in included studies.

Year, author	Level of dyspneia	Forced expiratory volume in 1 s (FV1)	Forced vital capacity (FVC)	Inspiratory capacity (IC)	Residual volume (RV)	Inspiratory and expiratory maximal pressure (MIP and MEP)	Total lung capacity	Distance walked on six-minute walk test (6mwt)	Bode index	Health related quality of life	Impact of the disease (CAT)
1996, BAULDOFF; ZULLO; HOFFMAM	NR	G1: 27 (16) G2: 37 (18)	G1: 40 (20) G2: 55 (24)	NR	NR	NR	NR	NR	NR	NR	NR
1996, STRIJBOS; ALTENA; KOETER	NR	G1: 40.40 (19.60) G2: 45.50 (6.90) G3: 42.60 (8.80)	NR	NR	NR	NR	G1: 103.50 (19.60) G2: 99.20 (13.70) G3: 98.70 (17.80)	NR	NR	NR	NR
1996, WIJKSTRA et al	NR	G1: 44 (11) G3: 45 (9)	NR	NR	NR	NR	G1: 118 (14) G3: 114 (11)	NR	NR	NR	NR
2000, HERNANDEZ et al	MRC G1: 3.20 (0.90) G3: 3.40 (0.90)	G1: 41.70 (15.60) G3: 40 (16.40)	G1: 71.10 (18.90) G3: 74.70 (14.70)	NR	NR	NR	NR	NR	NR	CRQ G1: 82.60 (18.90) G3: 91.20 (21.40)	NR
2003, SINGH	CRDQ G1: 4.12 (0.88) G3: 3.58 (0.84)	G1: 29.10 (10.40) G3: 26.90 (9.20)	NR	NR	NR	NR	NR	G1: 315 (118) G3: 264 (157)	NR	CRDQ G1: Dyspnea: 4.12 (0.88) Fatigue: 3.70 (0.90) Emotion: 3.90 (1,10) Mastery: 3.78 (0.90) G3: Dyspnea: 3.58 (0.84) Fatigue: 3.04 (0.80) Emotion: 3.20 (0.90) Mastery: 3.15 (0.80)	NR
2003, OH	BORG G1: 4.67 (2.91) G3: 3.43 (1.13)	G1: 42.12 (15.07) G3: 44.91 (17.75)	NR	NR	NR	NR	NR	G1: 350.70 (59.42) G3: 360.10 (59.46)	NR	CRDQ G1: 82.00 (22.53) G3: 89.75 (10.38) G1: Dyspnea: 14.44 (8.14) Fatigue: 16.20 (5.28) Emotion: 29.93 (8.61) Mastery: 19.20 (3.35) G3: Dyspnea: 16.38 (7.65) Fatigue: 15.88 (3.91) Emotion: 34.00 (6.72) Mastery: 23.50 (3.89)	NR
2005, BOXALL et al	NR	G1: 40.5 (15.9) G3: 37.7 (15.0)	G1: 1.39 (0.61) G3: 1.61 (0.58)	NR	NR	NR	NR	G1: 163.0 (60.1) G3: 147.5 (62.3)	NR	CRDQ G1: 56.50 (11.0) G3: 61.0 (10.70)	NR
2005, MURPHY; BELL; COSTELLO	MRC G1: 3.10 (1.10) G3: 2.80 (1.13)	G1:38 (12) G3: 42 (12)	NR	NR	NR	NR	NR	NR	NR	SGRQ G1: 70.20 (14.20) G3: 65.30 (17)	NR
2006, KOPPERS et al	BORG G1: 8.40 (1.90) G3: 8.30 (1.70)	G1: 1.50 (0.40) G3: 1.70 (0.50)	NR	NR	NRN	NR	NR	G1: 512 (86) G3: 549 (75)	NR	CRQ G1: 78.70 (20.60) G2: 82.40 (14.70)	NR
2007, RESQUETI et al	MRC G1: 3.40 (0.80) G2: 3.60 (0.80)	G1: 27.50 (9) G2: 29.60 (8)	G1: 60.10 (15) G2: 62.20 (18)	NR	NR	NR	NR	NR	NR	CRQ G1: 15.40 G2: 16.90	NR
2007, SHEA; TAYLOR; PARATZ	CRDQ G1: 3.20 (0.80) G3: 3.70 (0.90)	G1: 49 (25) G3: 52 (22)	NR	NR	NR	NR	NR	G1: 376 (114) G3: 337 (83)	NR	CRDQ G1: 16.20 G3: 17.50	NR
2008, MOULIN et al	CRQ G1: 5.30 (4.50–6.10) G3: 4.80 (4–5.60)	G1: 58.60 (53.80–63.40) G3: 58.40 (50–66.90)	NR	NR	NR	NR	NR	G1: 511.20 (461.70–560.70) G3: 465.20 (415.70–514.70)	NR	CRQ G1: 5.40 (4.90–6) G3: 5 (4.40–5.50)	NR
2008, MALTAIS et al	NR	G1: 46 (13) G2: 43 (13)	G1: 2.55 (0.76) G2: 2.58 (0.84)	NR	NR	NR	NR	G1: 370 (89) G2: 368 (85)	NR	G1: 46 (16) G2: 46 (16)	NR
2009, FERNANDEZ et al	MRC G1: 2.60 (0.80) G3: 2.30 (0.80)	G1: 34 (10) G3: 38 (12)	G1: 56 (10) G3: 67 (19)	NR	B: G1: 131 (36) G3: 130 (20) P:	MIP B: G1: 59 (16) G3: 60 (20) MEP G1: 75 (18) G3: 86 (30) P: MIP G1: 66 (17) G3: 67 (16) MEP: G1: 77 (21) G3: 77 (25)	G1: 89 (14) G3: 98 (15)	G1: 313 (72) G3: 348 (85)	NR	SGRQ G1: 40.60 (13.80) G3: 50.00 (12.70) G1: Symptoms: 36.30 (20.40) Activity: 59.90 (13.90) Impact: 30.90 (16.30) G3: Symptoms: 36.60 (17.30) Activity: 70.00 (12.10) Impact: 42.90 (16.90)	NR
2010, OLIVEIRA et al	MRC G1: 0 = 8 1 = 9 2 = 10 3 = 6 4 = 0 G2: 0 = 6 1 = 8 2 = 5 3 = 4 4 = 0 G3: 0 = 8 1 = 8 2 = 10 3 = 2 4 = 1	G1: 47.50 (23.30) G2: 51.50 (23.90) G3: 41.40 (18.40)	G1:68.70 (30.20) G2: 79.10 (30) G3: 69.90 (28)	NR	NR	NR	NR	NR	G1:4 (0-7) G2: 4 (0-8) G3: 4 (1-9)	NR	NR
2010, GHANEM et al	CRQ G1: 11.80 (5) G3: 12.40 (4.40)	G1:29.44 (13.14) G3: 23.21 (7.70)	G1: 36.24 (14.17) G3: 29.00 (10.91)	NR	NR	NR	NR	G1: 88.79 (19.14) G3: 83.79 (15.9)	NR	SF-36 Physical component: G1: 30.60 (14.20) G2: 40.60 (21.9) Mental component G1: 26.30 (14.60) G2: 30.4 (19.90)	NR
2012, AKINCI, OLGUM	BDI G1: 5.20 (1.60) G3:6.10 (2.10)	G1: 36.10 (10.80) G3: 39.20 (11)	NR	NR	NR	NR	NR	G1: 157.90 (64.50) G3: 176.30 (54.90)	NR	SGRQ G1:55 (16) G3: 45 (18)	NR
2013, DIAS et al	NR	G1: 55.14 (24.80) G3: 60 (20)	NR	NR	NR	B MIP G1: 65.10 (11) G3: 71.10 (16.70) MEP: G1: 77.90 (38.80) G3: 51.60 (21.70) P MIP G1: 74.2 (19.3) G3: 70.2 (13.9) MEP G1: 82.7 (38.3) G3: 53.7 (22.9)	NR	NR	NR	AQ-20 G1: 12 (3–18) G3: 10 (4–20)	NR
2014, PINTO et al	MRC G1: 3,.0 (0.70) G3: 3.50 (0.70)	G1: 33.50 (7.30) G3: 34.50 (9.50)	G1: 2.50 (0.80) G3: 2.30 (0.50)	NR	NR	NR	NR	G1: 197 (106) G3: 190 (120)	NR	SGRQ G1: 56.30 (13.10) G3: 62.30 (11.60)	NR
2014, TABAK et al	G1: 3.0 (2.0–3.30) G3: 4.0 (3.0–4.0)	G1: 50.0 (33.30–61.50) G3: 36.0 (26.0–53.50)	NR	NR	NR	NR	NR	G1: 409.5 (111.60) G3: 313.0 (79.40)	NR	NR	NR
2016, HOLLAND et al	MMRC G1: 0 = 2 1 = 33 2 = 22 3 = 21 4 = 2 G2: 0 = 0 1 = 36 2 = 28 3 = 19 4 = 3	G1: 52 (19) G2: 49 (19)	G1: 78 (17) G2: 79 (22)	NR	NR	NR	NR	NR	NR	NR	NR
2016, TSAI et al	MMRC G1: 2 G3:2	G1: 60 G3: 68	G1: 2.60 (0.70) G3: 2.60 (0.60)	NR	NR	NR	NR	G1: 363 (66) G3: 383 (93)	G1: 6 G3: 6	CRDQ G1: 90 (18) G3: 88 (23)	G1: 16 (7) G3: 15 (6)
2017, HORTON et al	MRC G1: 2: (n:28) 3: (n:52) 4: (n: 47) 5: (n: 18) G2: MRC 2: (n:20) 3: (n:57) 4: (n: 39) 5: (n: 26)	G1:47.89 (18.67) G2:48.79 (17.19)	G1: 2.73 (0.84) G2: 2.67 (0.90)	NR	NR	NR	NR	NR	NR	CRQ-RS G1: Dyspnea (2.58) Fatigue (3.42) Emotion (4.41) Mastery (4.50) G2: Dyspnea (2.42) Fatigue (3.36) Emotion (4.37) Mastery (4.36)	NR
2017, VASILOPOULOU et al	MRC G1:2.30 (1.00) G2:2.50 (1.00) G3:2.20 (1.10)	G1:49.60 (21.90) G2:51.80 (17.30) G3:51.70 (21.00)	G1: 3.07 (0.90) G2: 2.70 (0.65) G3: 2.77 (0.81)	G1:81.70 (33.00) G2:77.30 (30.00) G3:76.90 (31.00)	G1:184.6 (80.6) G2:180.2 (59.9) G3:182.0 (70.9)	NR	G1:118.8 (30.3) G2:120.7 (25.7) G3:19.9 (28.8)	G1: 389.10 (91.30) G2: 385.10 (80.30) G3: 384.80 (80.20)	G1:3.50 (2.70) G2: 3.20 (2.10) G3: 3.30 (2.30)	SGRQ G1:46.2(19.7) G2: 43.5 (16.7) G3: 44.1 (16.6)	G1:17.6 (8.1) G2:15.7 (5.6) G3: 15.8 (4.9)
2018, COULTAS et al	CRDQ G1:4.48 (1.30) G2:4.33 (1.35)	G1:45.50 G2:47.30	NR	NR	NR	NR	NR	G1:342.80 G2:337.50	B:G1:4.40(1.90) G2:4.40 (2.00) P: G1: 4.10(1.80) G2:4.40 (2.10)	NR	NR
2019, LAHHAM et al	NR	G1: 54 (19) G2: 52 (19)	G1: 2.70 (0.60) G2: 2.9 (0.60)	NR	NR	NR	NR	NR	NR	NR	NR
2019, BURGE et al	NR	G1: 53 (18) G2: 49 (20)	NR	NR	NR	NR	NR	NR	NR	G1: 0.671 (0.429 − 0.912) G2: 0.643 (0.405 − 0.882)	NR
2020, LAHHAM et al	CRDQ G1: 19 (7) G3: 19 (5)	G1: 90 (8) G3: 92 (7)	G1: 104 (8) G3: 104 (24)	NR	NR	NR	NR	G1: 504 (123) G3: 488 (88)	NR	CRDQ G1: 93 (21) G3: 95 (17)	NR
2020, JIANG et al	mMRC G1: 2.79 (0.66) G2: 2.75 (0.71)	NR	NR	NR	NR	NR	NR	NR	NR	SGRQ G1: 50.24 (20.95) G2: 49.57 (17.52)	G1: 21.79 (6.85) G2: 25.55 (6.48)
2020, HANSEN et al	NR	G1: 32.6 (10.3) G2: 33.7 (8.4)	NR	NR	NR	NR	NR	NR	NR	G1: 2.7 (0.9) G2: 2.9 (1.0)	G1: 19.8 (7.3) G2: 20.4 (6.6)
2021, KO et al	mMRC G1: 2.1 (0.7) G3: 2.1(0.5)	G1: 49 (17) G3: 46 (15)	NR	NR	NR	NR	NR	G1: 249.10 (96.07) G3: 246.12 (88.38)	NR	G1: 38.15 (20.64) G3: 36.03 (15.25)	G1:12.85 (6.88) G3: 13.99 (7.84)

NR: Not reported; NA: Not available; † Data extracted from abstracts; MRC: Medical Research Council; MMRC: modified Medical Research Council; G1: home rehabilitation; G2: conventional outpatient rehabilitation; G3: no rehabilitation/control group; SGRQ: St. George’s Respiratory Questionnaire; CRQ/CRDQ: Chronic Respiratory Questionnaire/Chronic Respiratory Disease Questionnaire; AQ-20: Airways Questionnaire 20; SF-36: Short Form-36; MIP: inspiratory maximal pressure; MEP: expiratory maximal pressure.

Table 4. Characteristics of the intervention.

Year, Author	Type of activity HBPR	Type of activity CPR	Percentage of service (minimum 70%) (Y / NR)	Number of sessions per week
1996, BAULDOFF; ZULLO; HOFFMAM	Charge progression every week	NA	NR	5
1996, STRIJBOS; ALTENA; KOETER	After the third and fourth week, the workload was gradually increased up to 70% of stationary bicycles. When this exercise level was attained without any problems, exercise time was gradually increased	After the third and fourth week, the workload was gradually increased up to 70% of stationary bicycles. When this exercise level was attained without any problems, exercise time was gradually increased	Y	2
1996, WIJKSTRA et al	Charge progression every week	Charge progression every week	Y	2
2000, HERNANDEZ et al	The training intensity was set at 70% of the maximum speed reached in SWT	NA	Y	6
2003, SINGH et al	Continuous training	NA	Y	3
2003, OH	Load is gradually increased according to their condition	NA	Y	5
2005, BOXALL et al	Charge progression every week	Charge progression every week	NR	7
2005, MURPHY; BELL; COSTELLO	Continuous training	NA	Y	2
2006, KOPPERS et al	Continuous training	NA	NR	7
2007, RESQUETI et al	Load is gradually increased according to their condition	Load is gradually increased according to their condition	Y	5
2007, SHEA; TAYLOR; PARATZ	Progressive resistance exercises	NA	NR	3
2008, MOULIN et al	Continuous training	NA	Y	5
2008, MALTAIS et al	Continuous training	NA	Y	2
2009, FERNANDEZ et al	Continuous training	NA	NR	5
2010, OLIVEIRA et al	Load is gradually increased according to their condition	Load is gradually increased according to their condition	Y	5
2010, GHANEM et al	Continuous training	NA	Y	4
2012, AKINCI, OLGUM	Continuous training	NA	Y	3
2013, DIAS et al	Load is gradually increased according to their condition	NA	Y	5
2014, PINTO et al	Charge progression every week	NA	Y	7
2014, TABAK et al	Continuous training	NA	Y	7
2016, HOLLAND et al	Continuous training	Continuous training	Y	3
2016, TSAI et al	Load is gradually increased according to their condition	NA	Y	3
2017, HORTON et al	Continuous training	Continuous training	Y	2
2017, VASILOPOULOU et al	Load is gradually increased according to their condition	Load is gradually increased according to their condition	Y	5
2018, COULTAS et al	Continuous training	Continuous training	Y	7
2019, LAHHAM et al	Load is gradually increased according to their condition	Load is gradually increased according to their condition	NR	3
2019, BURGE et al	Load is gradually increased according to their condition	Load is gradually increased according to their condition	Y	5
2020, LAHHAM et al	Continuous training	NA	Y	7
2020, JIANG et al	Continuous training	NA	Y	3
2020, HANSEN et al	Continuous training	Continuous training	Y	3
2021, KO et al	Continuous training	NA	Y	2

NR: Not reported; NA: Not available; Y: yes; SWT: Shuttle Walking Test ^†Data extracted from abstracts.

Data extraction and meta-analysis

Dyspnea level

Seventeen studies (54.83%) evaluated dyspnea as an outcome [28–30, 32–36, 38, 39, 41, 44, 46, 47, 50–53]. The scales used to assess dyspnea were Borg scale [30, 33], Chronic Respiratory Questionnaire (CRQ/CRDQ) [29, 35, 36, 38, 44, 47, 51], Medical Research Council (MCRC) or the modified version (MMRC) [28, 32, 34, 41, 46, 52, 53], and Baseline Dyspnea Index (BDI) [39].

BORG scale scores

Two RCTs^, (6.45%) used the Borg scale to compare dyspnea level between HBPR and control group after two months [30] and five weeks [33] of rehabilitation from the baseline. In the experimental groups, the score on the Borg scale decreased, indicating a decrease in dyspnea levels, while there was an increase in dyspnea level in the control group [30]. There were no differences between the beginning and the end of the study [33].

CRQ/CRDQ scores

Seven studies (22.58%) used the CRQ/CRDQ scores to assess the level of dyspnea [29, 35, 36, 38, 44, 47]. Six studies (20%) [29, 35, 36, 38, 44, 51] compared HBPR and control group; however, only two of these studies were evaluated in the same follow-up period [38, 44]. Only one study (3.22%) [47], whose follow-up period were 18 months, compared HBPR with CPR, showing favorable results for the group treated with HBPR.

There was a significant decrease in dyspnea level among the patients classified as severe and very severe in the HBPR group when compared to the control group after two months of training (MD: 5.46; 95% CI: 1.97 to 8.96). One study assessed dyspnea levels after one month [29], while another study assessed after three and six months [35] from the baseline (Figure 2).

Figure 2. Mean difference (and 95% confidence interval) of dyspnea levels by CRQ/CRDQ between groups of HBPR (Home pulmonary rehabilitation) versus CG (Control group).

MCRC or a modified version of mMRC scores

Dyspnea was evaluated in five studies (16.12%) using the MCRC or MMRC scales comparing HBPR to patients of the control group [28, 32, 41, 46, 53]. One study also included a group of CPRs [46], while two studies (6.45%) compared the HBPR to CPR [34, 52].

There was no significant difference in dyspnea levels reported in the third month of follow-up between participants from HBPR and the control group (MD: −0.41; 95% CI: −0.82 to 0.01) (Figure 3).

Figure 3. Mean difference (and 95% confidence interval) of dyspnea levels by MRC/mMRC between groups of HBPR (Home pulmonary rehabilitation) versus CG (Control group).

BDI scores

Only one study [39] (3.22%) used the BDI scale for the assessment of dyspnea at the beginning of the study and after three months. The dyspnea level was lower in patients in the HBPR group than in the control group at the end of the follow-up period.

Lung function

Forced expiratory volume in 1 s

Three studies (9.67%) [27, 33, 53] presented the result of FEV1 differently from other studies, so they did not enter the meta-analyses. The results obtained in these studies demonstrate that there was no significant difference between the values ​​of this variable at the beginning and at the end of the programs between both groups evaluated (HBPR and control group). In addition, one study [36] did not present the standard deviation values, and for this reason, it did not enter the meta-analysis. This was done in an intervention period of 3 and 6 months, and in both periods, the HBPR obtained better results than the control group. Only one study that reported FEV1 results compared HBPR and CPR [47], and in 18 months, there was a small improvement in HBPR results, while in CPR, there was no change.

Three (9.67%) RCTs [29, 30, 39] compared FEV1 values between the HBPR and control groups but with different follow-up periods. The study with the intervention of one month [29] had an increase of 29.10% (SD: 10.40) for the HBPR group and 26.90% (SD: 9.20) for the control group. The two-month [30] intervention study had an increase of 41.91% (SD: 17.90) for the HBPR group and 53.89% (SD: 24.53) for the control group. The three-month [39] intervention study had an increase of 40.70% (SD: 13.20) for the HBPR group and 39.30% (SD: 12.10) for the control group.

Forced vital capacity

Two studies (6.45%) evaluated FVC, with a follow-up of 12 months [37] and 2 months [38]. In the first study, there was an increase of 57% (SD: 17) for the BPH group and a reduction of 66% (SD: 24) for the control group. In the second study, there was an increase of 40.40% (SD: 16.16) for the BPH group and of 26.57% (SD: 7.13) for the control group. The study included patients classified as moderate, severe and very severe COPD.

Residual volume

Only one (3.22%) study evaluated residual volume [37] pre and post-intervention, with an increase of 145% (SD: 34) for HBPR group and 148% (SD: 24) for the control group. The study included patients classified as moderate, severe and very severe COPD.

Total lung capacity

Only one (3.22%) study evaluated Total lung capacity [37] pre and post-intervention, with an increasing of 94% (SD: 15) for HBPR group and 108% (SD: 15) for control group. The study included patients classified as moderate, severe and very severe COPD.

Respiratory muscle strength

Inspiratory and expiratory maximal pressure

Two studies (6.45%) evaluated respiratory muscle strength, with an intervention period of 12 months [37] and two months [40]. There was an improvement in MIP and MEP in the HBPR group in comparison to the control group after the rehabilitation program. The population included in these studies obtained a classification of mild, moderate, severe and very severe COPD.

Oxygenation and BODE index

SpO2%

SpO2% showed no differences at three, six, 12 and 18 months from the baseline in comparison to HBPR with control group [26, 39]. Compiled analysis of the data was only possible for the three-month follow-up time (Figure 4).

Figure 4. Mean difference (and 95% confidence interval) regarding the results of SPO2 between groups of HBPR (Home pulmonary rehabilitation) versus CG (Control group).

iBODE

Only one study evaluated iBode [47] before and after the intervention, with a reduction in the score of 4.10 (SD: 1.8) for the HBPR group and 4.40 (SD: 2.10) for the control group. The study included patients classified as moderate and severe COPD.

Physical capacity

6MWT

The distance walked in a 6MWT was evaluated in 18 studies (58.06%) [29–31, 33, 35, 39, 41, 42, 44, 46–48, 50, 51, 53]. Of these, four (12.90%) [36, 48, 50, 51] presented only the value of mean distance obtained, and the results were individually better in the HBPR group than in the CPR group. In addition, eight studies (25.80%) compared HBPR to control group [29–31, 33, 35, 37–39, 41, 42, 44, 53]. A study (3.22%) with a 14-month intervention reported the distance covered in the three groups (HBPR, CPR and control group) [46].

The distance covered after two months of intervention [30, 38, 44, 46] was significantly greater for the HBPR group than for the control group (MD: 61.75 m; 95% CI: 42.94 m to 80.56 m). In a three-month intervention period [31, 35, 39, 41], the difference between the groups (participants with mild, moderate, severe, and very severe classification of COPD) remained significant, but the distance covered was shorter when compared to a two-month intervention period (MD: 40.24 m; 95% CI: 5.52 m to 74.97 m). After twelve months, the difference between groups was not significant (MD: 26.97 m; 95% CI: −12.02,52 m a 65.96 m) (Figure 5).

Figure 5. Mean difference (and 95% confidence interval) regarding the results of 6MWT between groups of HBPR (Home pulmonary rehabilitation) versus CG (Control group).

It was not possible to perform a meta-analysis between the HBPR and CPR groups because the studies [46, 47] presented different follow-up periods.

HRQoL

Twenty-two studies (70.96%) assessed HRQoL in COPD patients [28–41, 44–46, 48, 51–54]. There was no homogeneity in the time of follow-up between studies among patients in the HBPR and CRP groups; thus, a meta-analysis was performed only between HBPR and control group (Figure 6).

Figure 6. Mean difference (and 95% confidence interval) regarding the results of HRQoL by SGRQ between groups of HBPR (Home pulmonary rehabilitation) versus CG (Control group).

St. George’s respiratory questionnaire

Eight studies (25.80%) used the St. George’s Respiratory Questionnaire (SGRQ). Three studies (9.67%) including population classified as mild, moderate, severe, or very severe, were carried out in three months [31, 39, 41], with a significant improvement in HRQoL in the HBPR group compared to the control group (MD: −11.30; 95% CI: −19.81 to −2.79). In one study, therapists assessed the intervention group after six weeks [32] with no difference between groups (HBPR and CPR).

Two studies (6.45%) [37, 53] carried out the study in 12 months (MD: −08.03; 95% CI: −15.12 to −0.94) and showed a significant improvement in HRQoL in the HBPR group compared to the control group. One study [46] had the longest follow-up period, with the assessment of HRQoL after two and 14 months (Figure 6).

Short form 36

A single study (3.22%) [38] used SF-36 to assess HRQoL in individuals classified as severe and very severe; the total score and the domains were higher in the HBPR group than in the control group after two months from baseline.

Airways questionnaire 20

One study (3.22%) [40] used the AQ-20 with a population having COPD classification as mild, moderate, and severe. There was an increase in the score of 15 (0–18 minimum and maximum value) for the HBPR group and 11 (3–20 minimum and maximum value) for the control group. The study included patients classified as moderate, severe, and very severe COPD.

Chronic respiratory questionnaire (CRQ/CRDQ)

Nine studies (29.03%) evaluated HRQoL between the HBPR and control group using CRQ or CRDQ scales [28–30, 33–36, 44, 45]. The meta-analysis was not performed due to the variation in the follow-up period. Three studies verified the comparison between HBPR and CPR. In the study [33], the score in the five-week intervention period was higher in the HBPR, and in the others [34, 45], the score was higher in CPR, in an interval period of seven weeks and six months, and nine weeks and six months, respectively.

CAT

Four studies (12.90%) performed assessments related to the impact of COPD using the CAT score among participants in the HBPR and control groups [44, 46]. The results of the meta-analysis (Figure 7) showed that the patients included in the groups treated with HBPR reported a decrease in impact of disease. The studies were addressed in a population with moderate, severe and very severe COPD - after two months of intervention (MD: −4.71; 95% CI: −7.95 to −1.47) when compared to the participants included in the control group. Only one study [46] evaluated the impact of the disease after 14 months from baseline, showing a reduction of the impact in this population.

Figure 7. Mean difference (and 95% confidence interval) regarding the results of impact of the disease by CAT between groups of HBPR (Home pulmonary rehabilitation) versus CG (Control group).

The RCTs evaluating HBPR and CPR groups [46, 48, 54] had different follow-up periods, and meta-analyses were not performed.

Other findings

Wijkstra PJ et al., [27], performed an intervention with a 12-month follow-up comparing HBPR and CG. This RCT was the only one that performed an evaluation and reevaluation of the participants in the following variables: heart rate, breathing frequency, VO2, VCO2, and blood lactate concentration. Only one study evaluated the heart rate at the beginning and at the end of the study. In this case the heart rate in the intervention group had a small increase at the end of the study when compared to the initial values; however, this increase was not significant: baseline: 120 (SD:13) beats·min-1 and post intervention: 122 (SD:17) beats·min-1. Although the post-intervention values of the conventional rehabilitation group were reduced, this result was also not significant: baseline: 126 (SD: 15) beats·min-1 and post-intervention: 121 (SD: 16) beats·min-1. Both groups performed the same activities only the settings were different (home-based or clinic/ambulatory). For breathing frequency there was a reduction of 29 (SD: 4) breaths·min-1 in the HBPR group and reduction of 28 (SD: 3) breaths·min-1 in the control group. There was a reduction in blood lactate of 3.10 (SD: 1.30) mEq·L-1 in the HBPR group and a reduction of 3.4 (SD: 1.7) mEq·L-1 in the control group. In addition, for VO2 there was an increase of 1.1 (SD: 0.40) L·min-1 in the HBPR group and reduction of 1.0 (SD: 0.30) L·min-1 in the control group. For VCO2, was an increase of 1 (SD: 0.40) in the HBPR group and reduction of 0.90 (SD: 0.30) in the control group.

Quality assessment

The quality assessment of the included studies is presented in Appendice 2 and 3, Supplementary Material. Twenty-two articles (70.96%) showed a high risk of bias in ‘blinding of outcome assessment’. The domain in which the studies had the lowest risk of methodological bias was ‘generation of random sequence’, for which all had low risk of bias.

Reliability of cumulative evidence

The evaluation of the quality of the evidence through GRADE showed very low certainty of evidence for the dyspnea levels measured by CRQ/CRDQ; low certainty for the dyspnea levels measured by MRC/mMRC, as well for the outcomes FEV1 and SpO2% and moderate certainty of evidence for 6MWT and HRQoL. The main reasons for the decrease in the certainty of the evidence of the evaluated outcomes were the possibility of risk of bias, inconsistency, and imprecision (Appendix 4, Supplementary Material).

Discussion

COPD is a health condition characterized by persistent limitation of expiratory airflow [55]. Individuals diagnosed with COPD may have peripheral muscle deconditioning, dyspnea, decreased exercise tolerance and reduced HRQoL [56, 57]. In this systematic review, 31 RCTs were conducted to evaluate a HBPR program as an alternative treatment choice for COPD patients. Our results showed that HBPR programs reduce dyspnea levels, improve physical capacity (increasing the distance covered in the 6MWT), improving HRQoL and reduce the impact of the disease compared to participants who received only education and standard medical care.

The number of weekly sessions in the studies ranged from 2 to 7 days a week, with a mean of 4.29 (SD:1.76). All included studies counted only patients who were able to attend at least 70% of the treatment sessions, as availability to undergo treatment was one of the main inclusion criteria. Although 6 (19.35%) articles [25, 31, 33, 35, 37, 49] not reporting this information, it was clear that, in general, the sessions were held, as in most cases there was the supervision of a therapist performing stimuli and motivation.

Furthermore, in some studies [29, 32, 33, 36, 38, 39, 42, 43, 45, 47–49, 52–54] the protocol continuous exercises were used throughout the study, that is, the load, number of repetitions and sets were defined after the first assessment and followed the same values ​​until the end of the study. Differently, the other studies [12, 25–28, 30, 31, 34, 35, 40, 41, 44, 46, 49, 50] included chose to establish protocols with gradual increases in workloads according to the results obtained in the first assessment. In these studies, the load is gradually increased according to the condition of the participants.

The HBPR programs use several methods to achieve the goals, such as breathing exercises, resistance exercises of the upper and lower limbs and aerobic exercises. In addition, it is recommended to carry out education, self-management and follow-up activities through visits or telephone calls to monitor and motivate patients [58, 59]. When these methods are combined, they can result in better lung functions, symptoms, muscle respiratory strength, exercise capacity, dyspnea and HRQoL in patients with COPD [51, 60, 61].

The use of upper limbs is essential in activities of daily living, such as eating, combing hair and brushing teeth, and therefore, it is related to functionality and quality of life [62]. Previous studies report that the muscular oxidative capacity of the upper limbs, in people with COPD, has lower levels when compared to individuals with normal spirometry [63]. Thus, in order to mitigate the effects of reduced oxidative capacity, one of the approaches used within home programs is the use of resistance and aerobic exercises for the arms. These exercise modalities contribute to reduced perception of dyspnea and arm fatigue during exercises, increased peripheral muscle strength, improved performance in daily activities [64, 65], directly affecting quality improvement of life [60].

We can note that in recent years, measuring HRQOL in multicenter studies of patients with COPD has been a constant and common practice, having a very important meaning in clinical practice [66]. Most of the included studies in this review measured HRQoL with different instruments: AQ-20, SGRS, CRQ/CRDQ and SF-36. Among these, the most commonly used was the SGRS, which made it possible to carry out a meta-analysis. In the included studies, intervention periods of a maximum of two months proved not to be sufficient to achieve statistically significant changes. However, when the intervention programs are longer than two months, we notice statistically positive changes in HRQoL for patients with COPD.

To assess physical capacity in rehabilitation programs, several tests can be used, including the 6MWT, which is considered a cheap and easy test to apply, being used as a complement in the assessment of patients with pulmonary and cardiovascular diseases [67]. In this review, the patients included in the HBPR group demonstrated improvement in functional performance, since the distance covered after two and three months was greater in the HBPR group when compared to the control group. In two months of rehabilitation, the patients obtained an average of 61.75 m of in walking distance and in 3 months this average was 40.24 m. According to the literature, an increase in the distance walked of 30 m is considered a minimally important difference (MID) intervention in the analysis of adults with COPD [68]. In our study, the performance for the 6MWT after the HBPR exceeded the minimal important difference (MID). After five months of intervention, the HBPR did not show better results when compared to the control group. Two studies [43, 69], had shown that the HBPR was not effective in maintaining gains in the areas of functional testing over long periods.

The increase obtained in the distance covered in the 6MWT of the participants present in the HBPR group when compared to the control group can be explained by the type of activities that each of these groups was assigned to do during the study. Participants in the control group were instructed to maintain their daily activities, taking their medication properly; in addiction, the participants in the group HBPR entered the group that followed a structured format, with exercises that varied from standardized or with increased weekly workload, depending on the protocol established by the RCT; moreover, the home group received phone calls conducted by therapists trained and specialized in motivational interviewing techniques. Setting exercise goals was an essential component of all telephone consultations, and participants were also encouraged to set other health goals. In addition, participants were encouraged to exercise using equipment that was readily available in the home environment (e.g. resistance exercises with water bottles). Thus, it is clear that the best result obtained in the 6MWT may be closely related to the daily exercises and stimuli present in the HBPR group. Furthermore, when related to the HBPR group, one of the major difficulties found in the studies is the participants’ inability to attend treatment sessions, which can negatively influence the results of a rehabilitation program [70]. We noticed an increase in the distance walked in CG participants, although it was not greater than the distance walked by participants in the HBPR group. This fact can be explained by the method used to treat these patients, which was based on drug treatment. The correct use of medication contributes to the reduction of symptoms and exacerbations of COPD, which directly interferes with the physical capacity of the participants [6].

All studies included in this systematic review investigated the impact of COPD using the CAT scale. This is a reliable, valid, and good answer tool for assessing the health status of patients with COPD [71]. The HBPR program was effective in reducing the CAT scores after two months of follow-up. Our meta-analysis identified an average of −4.71 points reduced in this questionnaire after the HBPR. This result was superior to MCID, as it represents a reduction of 2 to 3 points [72, 73].

Dyspnea levels after two months of HBPR were also improved. The MRC/mMRC - are traditionally used because it is easy to apply and understand [74] and the CRQ/CRDQ - is the most responsive instrument among the existing ones [75]. Reducing of dyspnea symptoms leads to less restrictions on activities of daily living, triggered by shortness of breath, consequently, health-related quality of life improves [76].

In general, the included studies presented a high age range related to participants: the minimum age was 38 years [33], and the maximum was 85 years [31]. Current evidence suggests that early COPD is associated with ineffective clinical results and leads to early detection, diagnosis, and maintenance of COPD, combined with smoking cessation and exercise, to ensure symptom control, disease progress and COPD results [77]. Thus, although COPD is considered a disease that affects the elderly population, it is also present in younger individuals.

Even though another review on this topic was published in 2014, the present review highlighted some points and updated information about HBPR. In this systematic review it was included a total of 31 RCTs and the number of studies selected to carry out the meta-analysis was also higher. A total of 2235 COPD patients were included in the qualitative description and 898 in the meta-analysis. In addition, this new study obtained updated results from the measurement of dyspnea levels using the CRQ/CRDQ instrument and another scale that had not been used previously, the mMRC/MRC. Another issue addressed in this review that had not been addressed previously was the results on SPO2 levels. It was also possible to carry out new meta-analyses evaluating the physical capacity of the participants using the 6WMT. Due to the number of RCTs, analyses of the results obtained at two and three months of intervention were performed. Furthermore, this review presents the results of the meta-analysis on the impact of the disease on the patient using the COPD Assessment Test (CAT). Finally, GRADE system was used to assess the overall quality of the body of evidence associated with the main results.

Study limitations

There are some concerns to consider as limitations of the present review. Although the included studies had similar primary objectives, the methodology, follow-up periods, and assessment of outcomes were far different among them. Thus, few studies were included in our meta-analyses. In addition, no meta-analysis could be performed to compare HBPR and CPR groups. Thus, it is not possible to include patients with HBPR who have the same performance and improve the outcomes that patients receive in an outpatient setting. In this context, the scientific literature will benefit from new studies that compare these two approaches for COPD patients with protocols for evaluating primary and secondary outcomes considering the gold standard instruments and longer follow-up periods. In addition, although the PROSPERO protocol did not include secondary outcomes, the research team responsible for preparing the systematic review decided to include these other variables. Several of these variables were cited in the majority of studies included here.

Conclusion

There is a limited number of studies comparing patients with COPD performing HBPR vs standard treatment. However, in our systematic review we found a reduction in the levels of dyspnea, an increase in the distance covered on the six-minute walk test, improving HRQoL and decreasing the impact of the disease in COPD patients in home-based pulmonary rehabilitation. Nonetheless, these results are based on evidence of very low to moderate quality.

Ethics approval

The systematic review protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO) platform on 20 September 2019 (PROSPERO; CRD42020151669).

Acknowledgements

We offer our deepest thanks to the FAPEMIG for the scholarships and to the to the Programa de Pós-graduação em Reabilitação e Desempenho Funcional (PPGReab), Physiotherapy Department of the Universidade Federal dos Vales do Jequitinhonha e Mucuri, that provided technical support for the development and implementation of this study.

Disclosure statement

We declare that there are no conflicts of interest.

Additional information

Funding

This study was financed in part by the Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG) - APQ-00062-18.