Seeking a deeper understanding of ‘distributed health literacy’: A systematic review

Abstract Background Previous research suggests that it would be useful to view health literacy as a set of ‘distributed competencies’, which can be found dispersed through the individual's social network, rather than an exclusively individual attribute. However, to date there is no focused exploration of how distributed health literacy has been defined, conceptualized or assessed in the peer‐reviewed literature. Aims This systematic review aimed to explore: (1) definitions and conceptual models of distributed health literacy that are available from the peer‐reviewed literature; and (2) how distributed health literacy has been measured in empirical research. Methods We searched MEDLINE, Embase, CINAHL, PsycInfo, Scopus, ERIC and Web of Science using truncated versions of the keywords ‘literacy’ and ‘distributed’ (within five words' distance). We collated the definitions and conceptual models of distributed health literacy, and report on how health literacy has been measured in empirical research studies. Findings related to distributed health literacy from included manuscripts were synthesized using thematic synthesis. Results Of the 642 studies screened, 10 were included in this systematic review. The majority were empirical manuscripts reporting on qualitative research in one of five countries, with two reviews, one conceptual analysis and one quantitative study. Edwards' definition of distributed health literacy, which emphasizes the health literacy abilities, skills and practices of others that contribute to an individual's level of health literacy was widely applied in a variety of clinical and geographical settings. However, we did not identify any quantitative instruments which directly measured distributed health literacy. There was significant variability in questions used to explore the concept qualitatively, and discrepancies across studies in regard to (a) what constitutes distributed health literacy and what does not (e.g., general social support), and (b) the relationship between distributed health literacy and other constructs (e.g., public health literacy). Conclusion Although there is a widely applied definition of distributed health literacy, our review revealed that the research space would benefit from the development of the concept, both theoretically for example via conceptual distinctions between distributed health literacy and other types of social support, and empirically for example through the development of a quantitative measurement instrument. Patient or Public Contribution This paper is a systematic review and did not involve patients or the public.


| INTRODUCTION
Health literacy has historically been defined as an observable set of individual skills that inform health actions. This individual focus is evident, for example, in the definition adopted by the World Health Organization in 1998-'the personal, cognitive and social skills which determine the ability of individuals to gain access to, understand and use information to promote and maintain good health', 1 and in newer definitions such as from the International Union of Health Promotion, where health literacy is defined as the combination of personal competencies and situational resources needed for people to access, understand, appraise and use information and services to make decisions about health. It includes the capacity to communicate, assert and act upon these decisions. 2 It is also clear from health literacy measurement instruments, which typically assess individual skills, such as skills in interpreting nutrition labels (e.g., the Newest Vital Sign), 3 recognizing medical terms (e.g., the Rapid Estimatre of Adult Literacy in Medicine), 4 and health-related reading and numeracy skills (e.g., Test of Functional Health Literacy in Adults). 5 However, more recently, the literature has begun to draw attention to the intersection between health literacy and the social context, acknowledging that other individuals, families and communities also play a role in one's health information acquisition, comprehension and decision-making. [6][7][8][9] Most often, social context is modelled as a construct that impacts health but is distinct from health literacy. In a recent systematic review, for example, 23 of 34 identified studies represented social context in this way, measuring an association between health literacy and a social context variable. 6 This included, for instance, measuring whether people with lower health literacy had more or less social support, social capital and social engagement compared to those with higher health literacy.
Alternatively, a smaller number of studies (n = 6) positioned social skills (i.e., the ability to interact with and draw upon others for support) as a specific type of individual health literacy. 6 This is also illustrated in the 'Social support for health' subscale of the Health Literacy Questionnaire. 10 Other work still has recognized that an individual's health literacy skills are supplemented by those of others (including family, carers and health professionals), together contributing to an improvement in individual or collective health outcomes. Edwards et al. 7 'distributed health literacy' model, for example, argues that while individual health literacy may vary within a group, individuals can overcome personal deficits in health literacy skills by combining their efforts. In this way, distributed health literacy is a resource that may buffer the adverse impacts of low health literacy. 11 Although previous research has provided a broad overview of the intersection between health literacy and the social context, 6 to date there is no focused exploration of how distributed health literacy has been defined, conceptualized or assessed in the peer-reviewed literature, and no attempt to synthesize the existing body of research. To progress this field, this study aims to explore: (1)

| Information sources and search strategy
The article search for this review was completed on 4 February 2021.
The search strategy aimed to find published journal articles on the topic of distributed health literacy using a two-step search strategy: 1. An initial search of MEDLINE, Embase, CINAHL, PsycInfo, Scopus, ERIC and Web of Science was undertaken. Keywords used were truncated versions of 'literacy' and 'distributed' (within five words' distance). Where controlled vocabulary terms (e.g., MeSH) were available they were applied. The full search strategies for all databases, including any limits used, are included in Appendix SA.
2. The reference lists of all eligible articles were checked for any additional relevant studies.  extracted data from included papers using data extraction tools from the JBI. The data extracted included basic information about the study (author, year of publication, journal), study details (research questions, target groups, methods, settings, recruitment procedures, participant demographics, data analysis) and information related to distributed health literacy (definitions, conceptual models, instruments and approaches used to measure distributed health literacy, findings and authors' conclusions). For the purposes of this review, we did not seek to obtain or confirm data from study investigators.

| Synthesis of results
In line with our study aims, we collated the definitions and conceptual models of distributed health literacy that were referred to in the included studies, and report on how health literacy has been measured in empirical research studies (including quantitative measurement tools and qualitative interview guides).
In addition, we also synthesized the results from all included qualitative studies (including systematic reviews of qualitative studies) and theoretical manuscripts using thematic synthesis, as described by Harden and Thomas. 13 We took an inductive approach to find themes, which involved free line-by-line coding of the findings of primary studies and the organization of these 'free codes' into themes. Each stage was conducted by two authors (D. M. M. and D. G.), with discussion and input from the entire authorship team.

| Quality appraisal
Full texts selected for retrieval were assessed for quality using standardized critical appraisal instruments from the JBI (https:// joannabriggs.org/ebp/critical_appraisal_tools). Due to the heterogeneity of the study designs included, three JBI Critical Appraisal Tools were utilized to assess study quality. These appraisal tools included the Checklist for Qualitative Research 14 (n = 6; for example, 'Is there congruity between the research methodology and the methods used to collect data?'), the Checklist for Systematic Reviews and Research Synthesis 15 (n = 3; e.g., 'Were there methods to minimise errors in data extraction') and the Checklist for Quasi-Experimental Studies 16 (n = 1; e.g., 'Were outcomes measured in a reliable way').
Any disagreements between reviewers (D. G. and I. R. H) were resolved through discussion with a third member of the research team.

| Study selection
Of the 642 studies screened, 10 were deemed to meet the inclusion criteria for this systematic review after full-text screening (see Figure 1). The majority of included studies reported on qualitative research (n = 6). Other studies reported on quantitative research (secondary analysis of data; n = 1), or took the form of a systematic review (n = 1), conceptual analysis (n = 1) or 'perspective' article (n = 1). In addition to the primary studies, Gessler and colleagues 23 conducted a systematic review of qualitative studies that explored the process of decision-making and characterized how adolescents and young adults share healthcare information. Although the 14 eligible studies included in the review did not refer specifically to distributed health literacy, findings were synthesized using this conceptual frame. Bröder and colleagues conducted an iterative conceptual analysis of child and youth health literacy and offered a target-group-centred definition that embodied concepts of distributed health literacy. 24,25

| Quality appraisal
Quality appraisal scores can be found in Tables S1-S3. Qualitative studies were rated as high quality overall, while the quantitative study and two studies assessed using the checklist for systematic reviews and research synthesis had lower quality ratings, mainly because key details of methodological rigour were not elucidated in the text (e.g., details about data extraction; critical appraisal).
3.4 | Definitions and conceptual models of distributed health literacy available from the peer-reviewed literature

| Systematic reviews and conceptual analyses
The systematic reviews and conceptual analyses that we identified rarely included methods or search strategies specific to distributed health literacy. Rather, they adopted a broad approach to searching the health literacy literature, framing their results and/or forming conclusions and definitions of distributed health literacy from the analysis of their findings (see Table 2).  Social support: The remaining studies, however, mainly focused on family and social networks as sources of 'social support', rather than providing or supplementing a range of health literacy skills.
Uwamahoro et al., 18   Results acknowledged that identified studies addressed the health literacy of persons close to the child, such as caregivers, mothers, parents and teachers, and noted that 'researchers have proposed that child and adolescent health literacy should be regarded as the product of both individual health literacy skills and the skills or resources available in the proximal social context-namely, the adults, peers or institutions that young people trust. Among others, determinant of health and a coping strategy. Their qualitative analysis of distributed health literacy focused on social and community networks as assisting young people living with HIV to 'break the isolation by enabling them to meet others in the same situation'. While the authors did note that family was also a source of information as well as material and emotional support, this was offered as a single sentence in a much larger description of social support and self-stigma in the 'distributed health literacy' theme of their qualitative analysis.
Dayyani et al. 17   Provence, Vietnam. This is exemplified by the experience of a participant who mentioned that exclusive breastfeeding could be interrupted, and weaning commenced earlier than advised based on the preferences of their parents-in-law. Uwamahoro et al. 18 also noted that by providing special treatment to the HIV-positive client (and supporting distributed health literacy) at times exposed them to stigma in the process. In their conceptual analysis, Bröder et al. 24,25 also acknowledge that health literacy can be hindered by social structures, power relationships, societal demands and layers of autonomy. seek to capture the distribution of responsibilities among those involved in shared health literacy practices. This could entail looking at (a) the sum of health literacy resources available within one's proximal social context and community, and (b) how these resources are then used. In interpersonal psychotherapy, an 'interpersonal inventory' is used to identify people within a patient's network using concentric circles (closest on the inside). 32 We feel that something similar could also be applied in the measurement of distributed health literacy. Given the findings of this review, there may also be utility in trying to capture potentially nonsupportive networks, and specifically analysing distributed health literacy in the social context of children and adolescents.

| DISCUSSION
Finally, our review makes evident that much work in the area of distributed health literacy has occurred with children and young adults, in developing countries and/or with ethnic-minority groups.
Given the large body of research that exists regarding culture and individualist/collectivist orientations, 33 and the unique social context of children, it is unsurprising that researchers have chosen to explore social networks and distributed health literacy in these groups.
Moving forward, it would be useful to explore, compare and contrast the validity of these similar constructs among different groups.

| STRENGTHS AND LIMITATIONS
We adopted very focused search terms in this review, purposefully including variants of 'distributed' as a key term. In this way, we would have excluded a number of studies that explore the intersection between health literacy and the social context more generally without referring specifically to the concept of 'distributed health literacy', and biased results towards studies published after 2015. However, this matches our study aims and avoids duplication of research which has already been done. Our findings are also limited by the small number of identified studies, although they do represent five diverse countries across the globe and participants with various health conditions.
Blinding and independent assessments of articles for inclusion and risk of bias represent an important strength. We searched seven databases for relevant studies and checked reference lists to supplement our searching, although we acknowledge that not all potential databases were searched, which may have resulted in some missed articles.

| CONCLUSION
Although there is a rather widely applied definition of distributed health literacy, the research space would benefit from additional conceptual clarity. This includes the development of the concept, both theoretically for example via conceptual distinctions between distributed health literacy and other health literacy concepts, as well as other types of social support, and empirically for example, through the development of one or more quantitative measurement instruments.