Initial search terms will be developed and piloted by the research team. Key search terms will include “men,” “male,” “man,” “female,” “women,” “woman,” “sex,” “gender,” “emotional perception,” “emotional processing,” “schizophrenia,” “schizophren,” “psychotic disorders,” “psychosis,” “psychoses,” “psychotic,” “schizoaffective,” “schizotypal personality disorder,” and “schizotyp.” Boolean operators “OR” and “AND” will combine these keywords. We will build a search terms and limits table to consider the keywords, terminology, and search limits. This table will also capture these keywords’ synonyms, abbreviations, alternative spellings, and collective nouns. We will use the MeSH (Medical Subject Headings) Browser to search for our topics and review the MeSH tree structures for each keyword to see whether there are any narrower terms to include in our search. We will then use this table to develop a preliminary search strategy with MEDLINE (PubMed; Multimedia Appendix 1). Peer review will be sought from a health services librarian with expertise in systematic reviews and meta-analyses. We will use the Polyglot search translator to adjust and translate our search strategy syntax to different databases [40]. The following electronic databases will be searched: MEDLINE, Embase, CENTRAL, CINAHL, and PsycINFO. Reference lists of included studies will also be manually scanned. Databases will continue to be searched for new publications until the end of June 2024. Study authors may be contacted to request full-text publications and clarify questions about study results. We will also search, at minimum, the first 3 pages of Google and Google Scholar to test our search strategy.

The literature search results will be uploaded to the web-based Covidence systematic review management software (Veritas Health Innovation) [41]. Abstracts that meet the search terms will be imported into Covidence, and titles and abstracts will be screened to determine whether they meet the inclusion criteria for a full-text review. Any duplicates will be removed during the screening process. Full-text papers will be uploaded for the formal reviewing process. If full-text versions are unavailable, the researchers will attempt to contact the study authors to request a copy of their paper via a maximum of 3 email attempts. Researchers (FD and GR) will independently screen titles and abstracts based on the study selection criteria. Before this screening process, screening questions will be developed to assist reviewers. Covidence software will flag any conflicts between reviewers at each of these stages. Reviewers will meet to discuss and resolve these conflicts. A third reviewer (HN) will be sought if a consensus cannot be reached. Full-text versions of studies that are considered relevant will then be formally reviewed by at least 3 researchers. Again, any conflicts during this process will be flagged, and reviewers will meet to seek consensus. Another reviewer will be contacted if an agreement cannot be reached. Eligible full-text publications will be stored in Endnote referencing software using smart groups. These selected papers will be checked for retractions and errata. Before data extraction, a search will be rerun to identify any new studies to be retrieved for inclusion. A PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flow diagram will record search results during the study selection process [42].

Two reviewers will independently critically appraise the quality of studies included in the systematic review for risk of bias using the relevant Joanna Briggs Institute checklist tools [43]. These tools provide checklists for a range of study designs, including randomized and nonrandomized controlled trials. Assessment will consider potential sources of bias for each study, such as in the selection of participants, reliability in the measurement of outcomes, appropriateness of the statistical analysis undertaken, and potential selective reporting of results. Reviewers will meet to resolve any discrepancies in these ratings. A third reviewer will be sought if required. No reviewers will be blind to journal titles, study authors, or institutions. We will report the results of this quality assessment in tabular format in the systematic review.

One researcher will extract the data from the selected full-text publications using the web-based Covidence systematic review management software [41]. This process will be reviewed by a second researcher. A data extraction template will be developed, piloted, and reviewed by the researchers before formal data extraction. If any data are unclear, the researchers will attempt to contact study authors to clarify their findings via a maximum of 3 email attempts over 28 days. We expect to extract the following data from the selected full-text publications as available: author and institution details, publication information (including study title, journal title, publication date, and study location), participant demographic and clinical information (including age, biological sex, primary and secondary mental health diagnoses, mean duration of onset of symptoms, premorbid functioning, current stage of illness, any other cognitive or physical impairments, and current medications), study methodology (including trial design, trial size, and emotional perception task), and all quantitative data on dynamic perception task performance including any dropouts or missing data.

The characteristics of the included studies will be presented in tabular format and summarized in narrative synthesis as appropriate. If sufficiently comparable data to permit meta-analysis emerges, this will be completed using the random-effects model. Both continuous (eg, standardized mean difference) and categorical outcomes (eg, odds ratio and relative risk) in the extracted studies will be assessed and compared. The heterogeneity of effect sizes of pooled studies will be measured via forest plots using chi-square and I² values. If appropriate, a funnel plot and Egger test will be used to report publication bias. No subgroup analyses are planned at this stage. The GRADE (Grading of Recommendations Assessment, Development, and Evaluation) system will be used to consider the strength of the evidence available [44]. Two reviewers will rate their level of confidence (ie, very low, low, moderate, or high) in the quality of evidence for outcomes. A third reviewer will be sought if consensus cannot be reached in these ratings. We will report this evaluation of the evidence in tabular format in the systematic review.

This study does not require ethics approval and consent to participate. No data used for analysis or reported in the systematic review will be identifiable.

This study will systematically review the impact of biological sex on emotional perception in adults with SSD. In comparison with the general population, deficits in emotional perception in schizophrenia are well established [13,14], but there is very little research into the potential impact sex has on performance. We did identify 1 systematic review and meta-analysis [34] that found only limited evidence of a female advantage for emotional perception. However, it is important to note that this construct was restricted to FEP tasks, and the study authors highlighted the need for future research into this area. We hypothesize that our review may provide support to these findings. In healthy adults, females tend to perform better at emotional perception tasks. Our review aims to understand whether this advantage extends to females with schizophrenia or whether the advantage is attenuated. The social cognitive profile of schizophrenia, more generally, is still largely heterogeneous and how it differs to other psychiatric disorders is still under investigation [9]. We hope to provide a greater understanding of how sex-related differences in emotional perception may explain some of this variation and, in turn, inform future treatment that targets functional outcomes in these individuals. The need for more research into sex-related differences in schizophrenia has been emphasized as a new but important field of research [33].

The main limitation anticipated in this review relates to the lack of consensus within the literature on how to measure emotional perception [12], compounded by the limited availability of well-validated and reliable tools. Often, attempts have been made to adapt measures from other participant samples to a schizophrenia population but with limited research on the psychometric properties of these measures in SSD [9]. A similar critique relates to existing social cognitive tasks’ lack of cross-cultural validity [9]. These restraints may limit the ability to draw conclusions and pool outcomes from comparable studies. If this is the case, we plan to provide a narrative synthesis and group studies by the type of emotional perception task, reporting on the reliability and validity of these measures. The study researchers must also acknowledge the cultural biases we bring to this review and be cautious in generalizing any conclusions drawn to other cultures, especially given the limit to studies available in English included in our study criteria. We also plan to measure the quality of evidence to assess the confidence in the available data and report these findings in the systematic review.

This study will review the evidence for sex differences in emotional perception within SSD. The results will provide a greater understanding of the social cognitive profile of schizophrenia and, in turn, inform future interventions tailored to improve long-term rehabilitation.