Study type

All published randomised controlled trials (RCTs) with a parallel, cluster or cross-over design will be included.

Participants

Pneumoconiosis patients diagnosed by relevant standard (such as GBZ 70–2015 National occupational health standards)3 will be included. There will be no restrictions on age, sex, ethnicity, education or economic status. We will exclude studies including participants with complications, such as pulmonary heart disease, tuberculosis and chronic respiratory failure.

Interventions

The intervention we focus on in this review is a comprehensive PR, which is based on exercise training, including or not including health education, nutritional intervention and psychosocial support.

Comparisons investigated are:

• PR versus no treatments;

• PR adjunctive to other treatments versus other treatments alone.

Outcome measures

• Primary outcomes

  The primary outcome measures were functional capacity and HRQL, as measured by 6 min walk distance,16 19–21 St. George’s Respiratory Questionnaire.16 21 22

• Secondary outcomes

  The secondary outcomes measures were pulmonary function, symptoms, acute exacerbations and adverse events, as measured by forced vital capacity, forced expiratory volume in the 1 s,21 22 the modified Medical Research Council dyspnoea scale,23 frequency of acute exacerbations16 24 and incidence of adverse events.

Search methods

We will search PubMed, Embase, The Cochrane Library, Web of Science, Chinese Biomedical Literature Database, China National Knowledge Infrastructure, Chongqing VIP and Wanfang Data from their inception to 1 April 2019. We have developed detailed search strategies for each electronic database without language restrictions to attempt to identify all eligible studies. We will also review the reference lists of included studies or relevant systematic reviews to identify any potentially eligible studies. The search strategy for PubMed is shown in table 1.

Searching other resources

We will conduct a search on the website of ClinicalTrials.gov, WHO International Clinical Trials Registry platform and Chinese Clinical Trial Registry, Conference Proceedings Index (Web of Science Core Collection) to identify additional ongoing or unpublished studies.

Study selection

Two review authors (JW and XL) will independently examine titles and abstracts retrieved from the search and select all potentially eligible studies. Then these full-text articles will be obtained and the same review authors will review them independently according to the inclusion criteria. We will resolve all disagreements by consensus, and a third review author (HZ) will act as an arbiter when consensus cannot be reached. The review authors will record all full texts that do not meet the inclusion criteria and provide the rationale for their exclusion. We will count multiple publications reporting the same group of participants or their subsets as one single study. Details of the selection process are presented in figure 1.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

PRISMA flow diagram.

Data extraction and management

Two investigators (JW and XL) will independently extract data from included studies. A structured and standardised data extraction form will be used to extract the relevant information. We will complete a data extraction sheet for every study included in the review, involving information on details of authors, year of publication, study design, characteristics of participants, intervention, comparator and outcomes.

Included studies with greater than 20% attrition will be considered at high risk of attrition bias.25 When SDs of the change of included studies are missing, we will substitute for them the mean SD of other included studies.26 We will exclude from the analysis studies, in which only medians and percentiles are available and there are no means of calculating mean change scores.26

Quality assessment and analysis

Methodological quality will be independently assessed using the Cochrane Collaboration’s tool for assessing risk of bias in RCTs.27 The assessment details include: 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 will be assessed as ‘low risk’, ‘high risk’ or ‘unclear risk’ according to the description details of eligible studies. Two review authors (JW and XL) will complete the data extraction and score each study, with a third review author (HZ) acting as an arbiter when differences occur between JW and XL. We will summarise the risk of bias and settle differences in author interpretation of data through discussion.

Data analysis

We plan to pool data from the outcomes of each study to provide an overall measure of the effect. For dichotomous outcome data, we will present the effect using relative risks with 95% CI. For continuous data, we will present the effect using mean differences with 95% CI. We will convert the unit to the commonly used one, when different units of measurement are employed. For example, we will convert the ‘month’ or ‘year’ into ‘weeks’.

We plan to use a χ² test to estimate heterogeneity. Further analysis can be performed using the I² statistic. A random-effect model will be used to interpret the results if heterogeneity is statistically significant, whereas a fixed-effect model will be used if heterogeneity is not statistically significant. We will regard heterogeneity as substantial when I² is greater than 50% or a low p value (<0.01) is reported for the χ² test for heterogeneity.28

When 10 or more studies are included in the meta-analysis, we will investigate reporting biases (such as publication bias) by using funnel plots.27

We will undertake statistical analysis by using Review Manager software.29 As with cross-over trials, we will consider only the first phase and exclude from the analysis data obtained during the second phase (ie, up until the point of crossover).

Subgroup analysis

If data permit, we plan to conduct subgroup analyses for different intervention forms to assess whether the treatment effects are different in different subgroups. Different intervention forms refer to whether exercise training is combined with other rehabilitation measures, such as health education, nutritional intervention and psychosocial support.

Sensitivity analysis

We will conduct sensitivity analyses to explore the robustness of the findings regarding the study quality and sample size. We will exclude studies one by one and comparing the results in the analysis.30 Sensitivity analyses will be showed in a summary table.

Quality of evidence assessment

The quality assessment of the body of evidence is performed to determine the extent to which an estimate of effect is close to the true quantity/value, that is, it is not distorted by internal or external bias within and across studies. The assessment will be conducted by outcome of interest using the Grading of Recommendations Assessment, Development and Evaluation system.31 The quality of outcome measures will be categorised as high, moderate, low and very low.

Patient and public involvement

Patients were not involved in the development of the research question or the design of this study at this stage.

The procedure for this protocol will be conducted according to the guidance provided by the Preferred Reporting Item for Systematic Reviews and Meta-analyses protocols.32

Potential amendments

We do not envisage any further amendments to this protocol. However, in case of any changes, the amendment shall be detailed out in the final report.