Pulmonary Rehabilitation Can Improve the Functional Capacity and Quality of Life for Pneumoconiosis Patients: A Systematic Review and Meta-Analysis

This study evaluated the efficacy and safety of pulmonary rehabilitation (PR) for pneumoconiosis. We systematically searched PubMed, Embase, The Cochrane Library, Web of Science, SinoMed, CNKI, VIP databases and Wanfang Data from their inception to June 1, 2019. A systematic review and meta-analysis of randomized controlled trials (RCTs) of PR for pneumoconiosis was conducted and reported in compliance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA). Two reviewers independently screened literature, extracted data, and assessed bias risk. All statistical analyses were performed using the RevMan software. Sixteen RCTs with 1307 subjects were ultimately included for analysis. Compared with routine treatment, PR was able to improve the 6-minute walking distance (mean difference (MD) 69.10, 95% confidence interval (CI) 61.95–76.25); the 36-Item Short Form Health Survey total score (MD 17.60, 95% CI 13.59–21.61); physical function score (MD 15.45, 95% CI 3.20–27.69); role physical score (MD 17.87, 95% CI 12.06–23.69); body pain score (MD 14.34, 95% CI 10.33–18.36); general health score (MD 20.86, 95% CI 16.87–24.84); vitality score (MD 11.66, 95% CI 0.18–23.13); social function score (MD 9.67, 95% CI 1.27–18.08); mental health score (MD 20.60, 95% CI 13.61–27.59); forced vital capacity (FVC) (MD 0.20, 95% CI 0.12–0.29); forced expiratory volume in 1 s (FEV1) (MD 0.23, 95% CI 0.09–0.38); FEV1% (MD 5.19, 95% CI 1.48–8.90); maximal voluntary ventilation (MD 4.47, 95% CI 1.14–7.81); reduction in the St. George's Respiratory Questionnaire score (MD -9.60, 95% CI -16.40 to -2.80); and the modified Medical Research Council Scale score. Furthermore, PR did not increase the FEV1/FVC (MD 3.61, 95% CI -3.43 to 10.65), nor the emotional score (MD 6.18, 95% CI -23.01 to 35.38) compared with the control. We found no reports of adverse events associated with PR. Thus, to some extent, PR can improve functional capacity and quality of life in patients with pneumoconiosis. However, these results should be interpreted with caution because of high heterogeneity. This trial is registered with registration number CRD42018095266.


Introduction
Pneumoconiosis refers to a group of occupational lung diseases characterized by diffuse fibrosis of the lung tissue. It is caused mainly by long-term inhalation and deposition of mineral dust, with varying levels of pathogenicity, into the lungs during occupational activities [1][2][3][4][5][6]. There are 12 forms of occupational pneumoconiosis prescribed by the law in China, of which silicosis and coal worker pneumoconiosis are the most common. Patients with pneumoconiosis are typically affected by cough, expectoration, chest distress, and dyspnea [1]. Pneumoconiosis is a progressive and irreversible but preventable lung disease [7,8]. It compromises personal health and can cause serious harm among families [9,10].
According to the statistical bulletin on the development of health and health services in China in 2018, 19 468 cases of occupational pneumoconiosis were reported, accounting for 82.85% of a total of 23 497 cases of occupational diseases [11]. The prevalence of silicosis from 2002 to 2016 was 12.7% [12]. The number of new cases of pneumoconiosis in China showed an upward trend from 2000 to 2016 and a slight decline after 2016 [13,14]. At present, more than 870 000 cases of pneumoconiosis have been reported in China [15], with a mortality as high as 31.2% [16]. Particularly, longterm exposure to silica dust is associated with substantially increased mortality [17]. However, various reports indicate that the actual number of patients with pneumoconiosis in China has exceeded 6 million [18].
No specific drugs or other treatment methods currently exist for pneumoconiosis [19]. Therefore, the management of the condition should first strengthen a comprehensive approach to overall health and active interventions (including symptomatic treatment, complication/combination treatment, and rehabilitative treatment) to reduce pain, delay progression of the disease, and prolong the life of the patient [1]. Comprehensive intervention has been clearly demonstrated to alleviate dyspnea and improve exercise performance and the health-related quality of life (QoL) [20]. Comprehensive interventions, including exercise training, education and behavioral change, and pulmonary rehabilitation (PR), have been widely used in the treatment of respiratory diseases and can effectively yield sustained improvement in functional capacity and reduced need for clinical care [21][22][23][24][25][26].
Exercise training can improve physical ability and the quality of life in patients with nonmalignant respiratory diseases. However, generally, fewer patients with pneumoconiosis have been included in previous relevant studies [27]. Recently, several randomized controlled trials (RCTs) of PR in the treatment of pneumoconiosis have also been reported. To further evaluate the efficacy and safety of PR for pneumoconiosis, we included a comparatively greater number of patients with pneumoconiosis and conducted a meta-analysis.

Materials and Methods
This systematic review and meta-analysis was conducted in accordance with the Cochrane Handbook for Systematic Reviews of Interventions [28], and reported in compliance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [29] (Supplementary data: Table S1). The study protocol was prospectively registered in the International Prospective Register of Systematic Reviews (PROSPERO) registry (CRD42018095266) and published [30].
2.1. Eligibility Criteria 2.1.1. Study Type and Participants. All published RCTs with a parallel, cluster, or crossover design of PR for patients with pneumoconiosis were included. Only studies in which the authors specifically noted a diagnosis of pneumoconiosis were considered for inclusion.

Intervention.
The intervention on which we were focused was PR, based on exercise training. Furthermore, the intervention may or may not have included health education, nutritional intervention, and/or psychosocial support.

Outcomes
(1) Primary Outcomes. The primary outcome measures were functional capacity and health-related QoL, as measured by the 6-minute walking distance (6MWD) and St. George's Respiratory Questionnaire (SGRQ).

Trial Selection.
After records were imported into the EndNote reference management software (Clarivate Analytics), duplicate records were removed. Two reviewers (J-JW and X-LL) independently examined titles and abstracts and selected all potentially eligible studies. Full-text articles were then obtained and reviewed independently, based on the inclusion criteria. We resolved all disagreements by consensus, and a third reviewer (H-LZ) acted as an arbiter when consensus could not have been achieved. Details of the selection process are presented in Figure 1.

Data Extraction.
Two investigators (J-JW and X-LL) independently extracted data from the included studies. Disagreements were solved by consultation with the third reviewer (H-LZ). Structured and standardized data extraction sheets were completed for every study included in the review. Those sheets included details of the authors, year of publication, study design, characteristics of participants, intervention, comparator, and outcomes. 3 BioMed Research International tool for assessing the risk of bias [31][32][33]. The assessment details included sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessors, incomplete outcome data, selective reporting, and other sources of bias. Each domain was assessed as "low risk," "high risk," or "unclear risk," according to the description details of the eligible studies. We summarized the risk of bias and solved differences in author interpretation of the data through discussion.
2.6. Quality of Evidence Assessment. We also evaluated the quality of evidence for the outcome measures using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) system [34]. The GRADE system divides the quality of evidence into four levels: high, medium, low, and very low. A summary table was prepared using the GRADE profiler (GRADEpro, version 3.6).

Data Analysis.
For continuous data, we calculated the mean differences (MD) with 95% confidence intervals (CI). When different units of measurement were employed, each unit was converted to the most common. For example, we converted the "month" or "year" into "weeks".
The χ 2 test for heterogeneity was performed. We used the random effects model across all analyses, to address clinical heterogeneity of the included population, as well as variations in treatment duration. Publication bias was assessed via a funnel plot.
We also performed subgroup analyses according to different intervention forms (exercise training and health education or exercise training and health education combined with other measures) and durations of PR (<3 months or ≥3 months). We excluded studies in succession and compared the results through sensitivity analysis. All statistical analyses were performed using the Review Manager (RevMan) software [35]. Publication bias was assessed by Begg's test and Egger's test with the STATA software version 12.0 [36].

3.2.
Characteristics of the Included Studies. This review included 16 RCTs published from 2009 to 2019, with a minimum sample size of 10 and a maximum of 200. The total numbers of patients in the PR group and control group were 655 and 652, respectively. The duration of treatment ranged from 1.5 to 12 months. The characteristics of the included studies are presented in Table 2.

Assessment of Risk of Bias.
Of the 16 included studies, 14 were judged to show unclear risk and 2, high risk. Four studies described randomization methods, of which two used a random table of numbers, one used a lottery, and one used a computer-generated randomization program. One study used allocation concealment via a lottery. Participants were not blinded to the specific intervention in any of the studies. In one study, the assessor was blinded. Three studies reported loss to follow-up and dropouts. Two studies did not report the primary outcomes. The results of risk of bias assessment are summarized in Figures 2 and 3. The statements for risk of bias for all 16 included studies are shown in Table 3.

Primary Outcomes
Twelve studies were included in the metaanalysis. The between-study heterogeneity was low for the 6MWD. The results showed that PR could significantly increase the 6MWD (MD 69.10, 95% CI 61.95 to 76.25, P < 0:001, I 2 = 4%) compared with the control (Figure 4).

SGRQ.
Four studies were included in the meta-analysis. The between-study heterogeneity was high for the SGRQ. The results showed that PR could reduce the SGRQ score (MD -9.60, 95% CI -16.40 to -2.80, P = 0:006, I 2 = 88%) compared with the control ( Figure 5).

Forced Vital Capacity (FVC).
Three studies were included in the meta-analysis. The between-study heterogeneity was low for FVC. The results showed that PR could increase the FVC (MD 0.20, 95% CI 0.12 to 0.29, P < 0:001, I 2 = 14%) compared with the control (Figure 6).

Forced Expiratory Volume in 1 s (FEV1).
Five studies were included in the meta-analysis. The between-study heterogeneity was high for the FEV1. The results showed that PR could significantly increase the FEV1 (MD 0.23, 95% CI 0.09 to 0.38, P = 0:002, I 2 = 77%) compared with the control (Figure 7).

FEV1%.
Six studies were included in the meta-analysis. The between-study heterogeneity was high for the FEV1%.

Maximal Voluntary Ventilation (MVV).
Four studies were included in the meta-analysis. The between-study heterogeneity was high for the MVV. The results showed that PR could significantly increase the MVV (MD 4.47, 95% CI 1.14 to 7.81, P = 0:009, I 2 = 77%) compared with the control (Figure 10).

BioMed Research International
3.5.6. mMRC. One study was included in the meta-analysis. The results showed that PR could reduce the mMRC compared with the control (Figure 11).

Subgroup Analysis.
We conducted subgroup analysis of the intervention and duration of PR. No statistical differences were noted in the improvement of the 6MWD among different subgroups. However, statistical differences were noted in the improvement of FEV1% between exercise training plus health education and exercise training plus health education combined with other measures. Compared with exercise training plus health education, exercise training plus health education combined with other measures showed more favorable results in improving FEV1% (Table 4). Statistical differences were noted in the improvement of SGRQ and MVV between the PR duration of <3 months and the duration of ≥3 months. Compared with the PR duration of <3 months, the duration of ≥3 months showed more favorable results in improving SGRQ scores and MVV ( Table 4). Results of all subgroup analyses in the meta-analysis are presented in the Supplementary materials: Figure S1-S13.  (Table 5).
3.8. Quality of Evidence. The overall quality of evidence, according to the primary outcome measures, was moderate or low. The GRADE evidence profiles of the primary outcomes are shown in Table 6.

Limitations.
We had hoped to also evaluate the effects of PR on acute exacerbations of pneumoconiosis and the occurrence of adverse events in patients with PR. However, the 16 studies included did not report acute exacerbations or adverse events. Therefore, the present analysis could not report them. This is a major limitation of this systematic review.

Discussion
Sixteen RCTs with 1307 subjects were ultimately included in the analysis. The main findings of the present study indicated that PR can be effective in patients with pneumoconiosis, as evidenced by the changes in the 6MWD, SGRQ, mMRC, SF-36, and pulmonary function. No adverse events were reported in any of the included studies.   The mean improvement in the 6MWD following PR was 69.10 m, which exceeds the mean improvement of 44.34 m observed in people with COPD who have undergone PR [23] and the minimal clinically important difference (24-45 m) for idiopathic pulmonary fibrosis [53]. In the present study, we also found that PR could reduce the SGRQ score compared with the control. The mean reduction in the SGRQ score following PR was 9.60, which exceeds the minimal clinically important difference for COPD [54,55].
We also found that PR reduced the mMRC score and increased the SF-36 score, compared with the control. Regarding pulmonary function, PR was able to improve the     At present, pneumoconiosis is still considered an incurable and irreversible disease, and the quality of life among patients with pneumoconiosis is low [56,57]. As the treatment goal for pneumoconiosis is to alleviate symptoms or prevent deterioration, we propose that the current guidelines should include PR as a routine treatment for pneumoconiosis. We found that PR in patients with pneumoconiosis was        BioMed Research International generally safe and well tolerated. Thus, the present evidence, at least to some extent, supports the fact that PR can be recommended for patients with pneumoconiosis.
Specific weaknesses and shortcomings of the present study must be acknowledged, which may reduce the credibility of the results. First, almost all trials were conducted in    BioMed Research International

BioMed Research International
China, which may limit generalizability. Second, some weaknesses in methodology may have had considerable effects on the results. For example, only four trials clearly reported the randomization methods used. Only three trials reported attrition. Only one clinical trial was formally registered. Furthermore, we only evaluated effects after rehabilitation and found that patients may have derived more benefits by prolonging the duration of PR. However, we were unable to determine the best course of PR or evaluate the long-term effects of PR. Similar to the problems encountered in most systematic reviews, our analyses were greatly affected by heterogeneity. This is a major issue, on which we need to focus in the next steps of our research.
There are currently no effective drugs or measures to treat fibrosis due to pneumoconiosis. Our findings have important implications worldwide, for both policymakers and clinicians. We should apply PR as a routine treatment for pneumoconiosis.

Conclusions
Our results show that PR can improve patients' quality of life and functional capacity. However, the quality of the research included in the present study was low, which limits the strength of the inferences that can be drawn. This issue needs further investigation through a well-designed RCT. There is also a need for follow-up data, to demonstrate the extent to  12 BioMed Research International which the effects of PR are maintained over time. Future studies should also pay attention to the specific types of patients with pneumoconiosis who can derive the most benefit from PR. We should also try to determine which form of PR is the most effective and the best course of treatment. Multicenter RCTs of PR for pneumoconiosis are needed worldwide, in order to generalize the results. Long-term exposure to mineral dust can severely compromise clearance and defense mechanisms of the respiratory system. A chronic, progressive course of disease can lead to reduced patient immunity, which can eventually lead to a variety of complications. Complications can have considerable impact on the treatment, progression, and prognosis of pneumoconiosis and can be a direct cause of patient deaths. Timely diagnosis and treatment of various complications are crucial to improving the condition of the patients, prolonging their lives, and improving their quality of life. Therefore, in our future research, we would focus on the efficacy of PR for patients with comorbidities.
The results of this study, at least to an extent, support the use of PR to improve respiratory function and the quality of life in patients with pneumoconiosis; however, we should treat the results cautiously because of the high heterogeneity among studies in the present analysis.

Disclosure
The funders had no role in the study design, data collection and analysis, or interpretation of the data.  540 (4 studies) ⊕⊝⊝⊝ very low 1,2 * The basis for the assumed risk (e.g., the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; GRADE Working Group grades of evidence. High quality: further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate. 1 No blind method and assignment concealment. 2 High heterogeneity (I 2 = 88%) was found.