This systematic review will be conducted according to the Cochrane Handbook for Systematic Reviews of Interventions [19] and reported following the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines [20]. The protocol was registered in PROSPERO (CRD42023396234).

The following electronic databases will be searched in June 2023: MEDLINE (1946-Present), Emcare, Embase, CINAHL, PsycINFO, and Scopus. The search strategy was created in consultation with a health sciences librarian at McMaster University and will involve database-specific MeSH terms in combination with keywords (keywords will be consistent across databases). The reference list of the selected studies will be manually searched to identify further relevant articles. The search strategy was validated in MEDLINE, Emcare, and Embase databases by testing whether it could identify 3 relevant articles to be included, previously sought through Google Scholar [16-18].

EndNote reference manager (Clarivate) will be used to filter duplicate studies. Studies will subsequently be imported into Covidence (Veritas Health Innovation), which will be used for further removal of duplicates, screening, and selecting studies. Excel (Microsoft Corp) will be used for data extraction and management.

Abstracts and full-text screening will be pilot-tested by all reviewers with 12 articles to calibrate the selection process and ensure interrater reliability. An 80% or higher agreement will be sought before proceeding with screening [26]. Six reviewers will independently screen titles and abstracts, with each study requiring consensus between 2 reviewers before a final decision to include or exclude the study for the full-text screening is made [26]. Two authors will independently perform the full-text screening process. At any stage of the screening process, if articles do not meet the inclusion criteria or meet any of the exclusion criteria, they will be excluded and reasons for exclusion will be recorded [26]. Disagreements will be resolved through consensus, and if no consensus is reached, a third reviewer will be consulted.

Two authors will independently extract the data to a predefined data extraction table, including title, authors, year and country of publication, study design, study setting, sample size per group, population (age, gender, chronic respiratory disease, disease severity, and comorbidities) per group, intervention (type, frequency, components, and duration) per group, outcome measures used, the timing of data collection, follow-up times if applicable, adverse events, and results of the interventions or comparator groups (including the mean, SD, and sample size).

Any disagreements on the extracted data will be resolved by discussion, and a third reviewer will be consulted if no consensus is reached. In the case there are missing data, study authors will be contacted by email to request these data. A follow-up request will be sent after 2 weeks, and if there continues to be no response after 2 more weeks, there will be no further attempts to retrieve the data. If the data are reported only in figures and the authors did not provide the requested data, whenever possible WebPlotDigitizer will be used [27].

Risk of bias assessment will be performed independently by 2 authors at a study level using the Cochrane Risk of Bias 2 tool for RCTs (randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of the reported result) or the Risk of Bias in Non-randomized Studies of Interventions tool for non-RCTs (bias due to confounding, bias of selection of participants into the study, bias in classification of interventions, bias due to deviations from intended interventions, bias due to missing data, bias in measurement of outcomes, and bias in selection of the reported result) [28,29]. Disagreements will be solved through discussion, and if no consensus is reached, a third reviewer will be consulted.

Tables will be used to describe the various types of studies, patient populations, characteristics of cognitive interventions and comparators, and outcomes. Meta-analysis will be performed using the Review Manager (version 5; Cochrane) software. If at least 2 studies reported the same outcome, that outcome will be included in the meta-analysis. Results will be reported as mean difference if the studies used the same outcome measures or standardized mean difference if studies assessed the same outcome using different measures. Depending on the level of heterogeneity between applicable studies, a fixed effects or a random effects model will be used for meta-analysis (ie, a high level of heterogeneity between applicable studies would warrant the use of a random effects model). If a meta-analysis cannot be performed, data will be synthesized narratively, explaining the main findings as they relate to the impact of cognitive interventions for people with CRDs.

For the quality of evidence assessment, the GRADE (Grading of Recommendations, Assessment, Development, and Evaluation) system will be used to upgrade or downgrade the evidence [30]. The GRADE assessment will be performed independently by 2 authors and disagreements will be solved through consensus or consulting a third reviewer if a consensus cannot be reached. The domains to be looked at are as follows: risk of bias, imprecision, inconsistency, indirectness, and publication bias. Based on these domains, the quality of evidence will be determined to be high, moderate, low, or very low. The outcome of the GRADE assessment will be presented in a summary of the findings table.

The beneficial effects of cognitive interventions for healthy individuals and those with various cognitive impairments have been discussed in several studies [14-17]. However, its effects on people with CRDs have never been systematically documented [4,16-18]. To our knowledge, this is the first systematic review focusing on the effects of cognitive intervention in individuals with CRDs. The nature of this systematic review is a strength, as it provides an objective perspective on this topic while minimizing bias, by following established procedures and reporting guidelines. Other strengths include its robust search strategy as demonstrated by searching 6 databases, reporting on the percentage of agreement among reviewers and including a variety of health domains in the secondary outcomes.

One of the predicted limitations of this study is that it will collect data from various study types, not limited to RCTs. This may affect the quality of the literature included in the paper. However, due to the novelty of this rehabilitation topic, there may be a limited number of primary articles (RCTs) available. Despite this limitation, it is still essential to synthesize the current literature on the effects of cognitive interventions for people with CRDs to inform evidence-based clinical decisions and plan future research. Another limitation of the study is the inclusion of articles written only in English, Portuguese, and French. This may exclude relevant data published in other languages. To capture a broader range of articles, future research on this topic should include studies written in other languages. Finally, to ensure a focused and manageable analysis, we have made the decision to include only the most commonly reported forms of cognitive interventions in the literature. This selective approach allows us to streamline our review process and maintain clarity in our findings. Although this decision may exclude certain less frequently reported interventions, it enables us to provide a comprehensive evaluation of the widely recognized and established cognitive intervention techniques.

The systematic review's findings will assist in displaying the effectiveness of cognitive interventions as a potential treatment approach in reducing cognitive deficits in those with CRDs. This innovative field of clinical research is relevant to all health care disciplines as it impacts decisions related to screening and interventions and informs clinical practice for all health care professionals.