Search strategy and selection of studies

Six electronic databases (Medline, Embase, Global Health, PsychInfo, Scopus and LILACS) were searched in May 2022. Subject headings and keywords were related to the Caribbean, mental health, and stigma (full list of search terms shown in S1 Appendix). Help-seeking was not included as a subject heading with relevant keywords to not limit the amount of potentially relevant studies that may be generated from this initial search. Experts in the field were contacted to provide further publication recommendations, and reference lists of included papers were hand-searched by the main author (J-BG) for eligible papers not detected in the original search.

The inclusion criteria were full-text, data-based, peer-reviewed articles, of any study design (quantitative, qualitative or mixed methods), published up to May 2022 in English or Spanish (see Table 1). Studies were eligible for inclusion if they explored the relationship between mental health-related stigma and mental health-related help-seeking in individuals living in the Caribbean. All types of stigma were included in this review, and the definitions used are guided by the Lancet Commission’s report on stigma and discrimination in mental health [16]. This includes: self-stigma, stigma by association, public stigma, and structural discrimination. This review includes all areas of help-seeking, including attitudes, intentions, and behaviours.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Inclusion and exclusion criteria.

https://doi.org/10.1371/journal.pone.0291307.t001

All results from this search were imported into Endnote20 (Clarivate Analytics), where deduplication was applied. The remaining studies were uploaded to Rayyan [17] for title and abstract screening. All titles and abstracts were screened by author J-BG. Authors KGD and CG independently screened a randomly selected 20% of the sample (10% each) for consistency, and to help ensure all relevant studies were included in the review. Discrepancies were resolved by discussion and arbitration, where any further clarification over the inclusion criteria was discussed and amendments appropriately made.

Full-text papers were obtained and assessed by author J-BG against the inclusion and exclusion criteria. Authors KGD and CG independently screened a randomly selected 10% sample each. Discrepancies were resolved by discussion, and where necessary arbitration with another author (PCG). There was uncertainty over one paper, where author PCG made the decision as to whether it should be included. Author KGD assessed full papers published in Spanish for eligibility solely due to the linguistic abilities of the team.

Data extraction

Author J-BG extracted data from all included studies using a review-specific data extraction form developed using Microsoft Excel. Data extracted from all study types included: author, year, title, aim, country, study design, sample characteristics, type of mental health condition addressed, aspect of stigma explored, aspect of help-seeking explored and key findings relating to the impact of mental health-related stigma on help-seeking–extracted from both original data and author’s reflections from the results section only, unless results and discussion sections were combined. For quantitative studies, the measures of stigma and help-seeking used were also extracted. For qualitative studies, the method of data collection and analysis, and relevant data extracts and authors comments were recorded. Author KGD independently extracted data from two randomly selected papers of each design type for consistency. The independently extracted results were compared, and any discrepancies were discussed and resolved. It was established that consistency in extraction had been achieved, following which data extraction for the remaining papers were conducted by author J-BG only.

Quality appraisal

Author J-BG conducted a quality assessment of all included studies using the Mixed Methods Appraisal Tool (MMAT) [18]. Five quality criteria are listed for each study design (e.g., Is the sample representative of the target population? Are the findings adequately derived from the data?), where responses are ‘Yes’, ‘No’ and ‘Can’t tell’. Following the guidance from Hong et al. [18], rather than calculating an overall score for each paper, a more detailed presentation of the ratings of each criterion was constructed to appraise the quality. Authors J-BG and KGD independently assessed the same two quantitative and qualitative papers to determine consistency, after which the remaining papers were appraised by author J-BG alone.

Data synthesis

Data synthesis was conducted in three stages. Following guidance from the Evidence for Policy and Practice Information and Co-ordinating Centre (EPPI-Centre; [19], quantitative and qualitative research evidence were considered independently before merging.

First, a narrative synthesis [20] was conducted on findings from quantitative evidence. A preliminary synthesis was developed, providing an initial description of the results of the included studies. The relationship between mental health-related stigma and help-seeking within and between studies was explored, with continuous reference to the extracted data, to help identify factors that influence this relationship. Lastly, the robustness of the synthesis was assessed. This was achieved by appraising the methodological quality of the included studies using the MMAT, and how the evidence was synthesised [21]. A textual narrative is provided in the results.

The second stage involved a thematic analysis [22] conducted on findings from qualitative evidence. Extracted data was input into the qualitative analysis software NVivo (Release 1.7.1). Initial codes were generated to develop a codebook by author J-BG, which continued to be revised when appropriate after the coding of each paper. Codes were then clustered into groups to explore meanings, interconnections, and patterns. This led to the development of initial themes, refinement of main themes and subthemes, which were reviewed by authors PCG and TTS, to then be structured into a thematic map.

Lastly, the data underwent a process of triangulation through a thematic meta-synthesis, where the findings from the quantitative and qualitative syntheses were integrated. This involved two stages. First, the quantitative-based synthesis papers were re-reviewed to see which factors were also present in the qualitative evidence. Next, it was determined whether any relevant factors were identified in the quantitative papers that did not arise from the qualitative papers, and vice versa. The thematic map developed from the qualitative synthesis was extended to reflect findings from the quantitative synthesis.

Quantitative synthesis

Six papers reported quantitative evidence, captured from 1,207 participants. All studies were cross-sectional self-completed surveys. The quality of the quantitative papers overall was average to below average, meeting two to three out of five criteria. The narrative synthesis of these data is presented in terms of: Associations between stigma and help-seeking and Stigma-related barriers to help-seeking.

Qualitative synthesis

Three articles reported qualitative evidence, consisting of 49 participants. One article (33%) considered the experiences of military veterans and their families [29]. Another paper (33%) explored individuals’ beliefs and use of the supernatural in their own lived psychiatric experiences [30]. The final article (33%) considered church leaders’ views [31]. The quality of the qualitative studies overall was good, with an average of four out of five criteria met.

Three themes were identified relating to the impact of mental health-related stigma and help-seeking in the Caribbean: (i) Making sense of mental health conditions; (ii) Anticipated/Experienced stigma-related experiences; and (iii) Individual characteristics. These themes are described below alongside select illustrative quotations. Further participant quotations are provided in S4 Appendix.

Overall meta-synthesis

The results from the quantitative and qualitative syntheses have been triangulated to produce an overall meta-synthesis. A conceptual model illustrating this is shown in Fig 2. Five of the six subthemes identified in the qualitative synthesis were also captured in the quantitative data, with the exception of the subtheme ‘lack of understanding’. One factor identified in the quantitative synthesis, but not in the qualitative synthesis, was the type of mental health condition and help-seeking source.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Conceptual model based on meta-synthesis of qualitative (n = 3) and quantitative (n = 6) results.

Key: Boxes with solid lines represent themes and subthemes. Subthemes that are underlined were identified in quantitative studies only. Subthemes that are asterisked were reported in qualitative studies only. Subthemes that are neither asterisked or underlined were reported in qualitative and quantitative studies. Dashed arrows indicate connections between the themes. Factors within a dotted circle were identified as characteristics of a theme.

https://doi.org/10.1371/journal.pone.0291307.g002

Future research

The review identified gaps in the existing evidence and highlights future research needs. There is a dearth of research into this important topic in the Caribbean and a need for more evidence. The majority of studies in this review only focused on public stigma. Previous research has demonstrated that different types of stigma, such as self-stigma have a differential impact on help-seeking [38], thus there is a need for this to be further explored in the Caribbean context. As most research has looked at mental health as a whole, we need to know more about how stigma for different types of mental health conditions impacts different types of help-seeking. MMR would also allow for a richer understanding of this topic. Research is needed that targets the heterogeneous racial and ethnic groups that exist in the Caribbean, as well as groups under-represented in the literature and likely to be particularly impacted by stigma on help-seeking, such as military veterans and LGBTQ+ individuals. Higher quality research is also required, which will assist in overcoming some of the current methodological issues identified from the quality appraisal, such as recruiting representative samples, and using appropriate methodological measures and data collection methods.

Implications for policy and practice

This review can be utilised in the development of policy and practice. There is a need for interventions that involve community-based mental health education and health promotion to contribute to public awareness on mental health, stigma and help-seeking. This may help to challenge and change existing labelling and negative language, as well as prevent social judgement and discrimination. When taking into account the factors and themes identified in the meta-synthesis, this indicates a need for particular attention to the needs of different groups of individuals, e.g., military personnel [29], when considering help-seeking preferences, needs and barriers, and providing support to address these. For practitioners providing treatment, there is an essential need for cultural understanding to be able to optimally engage and support an individual with a mental health condition. The results also suggest a need for self-coping strategies and teachings of ways to manage anticipated and/or experienced stigma to lessen the risk of this being a barrier to help-seeking.

Limitations and strengths

This review restricted its selection criteria to including published, peer-reviewed papers, which may have excluded relevant papers. However, this decision was made to be able to uphold the quality of the review. Some of the included studies had small samples, thereby caution may be needed when interpreting the findings of this review. Nonetheless, the findings from studies with lower participant numbers did not drive any single conclusions drawn in this research but rather corroborated findings from other studies, and by including all eligible studies regardless of sample size this review was able to analyse and synthesise important and interesting data from the limited existing literature to provide a thorough evaluation. Most of the included papers had many methodological issues, and there were no MMR studies, which could provide richer insights. The small number of included papers, limited Caribbean countries where the studies were set, and various target groups, restricts the generalisability of the results to other Caribbean settings. Whilst Caribbean countries share similarities, the Caribbean region is heterogenous. However, the findings from this review provide a comprehensive overview of the currently available evidence, for which research can be built upon in different Caribbean countries and contexts. The narrative synthesis, thematic analysis, and meta-synthesis were all conducted by author J-BG, which may raise issues of bias. However, analyses were reviewed by authors PCG and TTS to support the review’s validity.