Incorporating sex, gender and vulnerable populations in a large multisite health research programme: The Ontario Pharmacy Evidence Network as a case study

Over the past three decades, several important developments in health and social science literatures have affected the way that differences between men and women are addressed across fields of health research and policy. These changes are not independent, but reflect a more general shift toward consideration of the social as well as the biological dimensions of these differences and their causes, and toward including constructionist perspectives on social inequalities and differences alongside positivist explanations. Together, these changes have led research funders in Canada and elsewhere to require that research teams incorporate these perspectives in their research plans. This is the challenge that this paper hopes to help address.

An important change that has taken place across health and biomedical sciences has been the recognition that differences between males and females have been systematically ignored in health research. This has included women being under-represented in research designs, or absent entirely, resulting in results being inappropriately generalised from male-only samples to the entire population [1]. Evidence remains of continuing under-representation of women in research from the United States of America and other countries. This includes a lack of reporting of male/female differences, despite the United States National Institutes of Health requiring, since 1993, that federally supported Phase III clinical trials include female participants [2, 3].

Over the past several decades, social causes have attracted greater attention as explanations for observed differences between males and females for a range of health-related outcomes and processes. Since the 1970s, researchers have increasingly referred to the role of ‘gender’, or the socially derived differences between women and men, boys and girls, along with the physiological differences associated with ‘sex-linked biology’ [4]. Although the biological and social factors may be difficult to conceptually and methodologically delineate, this has led to greater attention to the importance of aspects of culture, norms and social roles, and their implications for a wide range of patient diagnoses, treatments and health outcomes [5‐7]. Researchers concerned with understanding gender differences in health have also called for attention to the role of unequal social power in their creation and maintenance. Drawing from critical feminist studies, this includes consideration of how the epistemology and practice of health and medical research might themselves be shaped by gender [8].

Other insights from critical social science have shaped the study of gender and health. Developed in the 1980s and 1990s, ‘intersectional’ approaches see social dimensions of inequality, such as gender, race/ethnicity, sexuality and social class, not as separate but as mutually constitutive and interlocking, whereby the implications of one’s positon on one dimension cannot be fully understood without reference to others [9]. The field of queer studies has contributed the ideas that both sexuality and gender identities are fluid and multiple, rather than fixed and binary. This has led to a new focus on the health of transgendered people, as well as lesbian, gay, bisexual and other sexual minorities, and of the importance of including these dimensions in health research and data [10, 11].

In response to the recognition of the complexities of sex-linked biology and gender, and their critical importance for health services research and policy, there have been calls for assessments of the scientific, social and economic benefits of research to be made specifically with an eye to reducing gender bias [1]. To this end, funding agencies have required researchers to incorporate both sex and gender into their research. Approaches to considering both the social and biologically-linked dimensions throughout research projects are sometimes referred to as ‘sex- and gender-based analyses’ or SGBA [12]. SGBA has been made either a requirement or a priority by public research funders [13‐15]. Similar to the National Institutes of Health [16], Health Canada implemented a Sex and Gender-Based Analysis Policy in 2000 [12], and applicants to any of the Canadian Institutes of Health Research (CIHR) funding programmes are now expected to incorporate sex and/or gender in research designs or to justify why they are not relevant [15, 17]. As that federal health research funding agency has put it, “if our research designs do not take sex and gender into account, the evidence we generate may be incomplete or simply incorrect; we risk not only doing harm (such as extrapolating findings based on male samples to females), but also missing critical opportunities to improve health (for example, not detecting the benefits of an intervention in a subgroup of men)” [18].

The proportion of CIHR-funded research projects incorporating sex and gender rose from 28% in 2010 to 48% in 2011, a result partly attributable to this new emphasis [19]. However, barriers to the integration of sex and gender into health research projects remain. Day et al. [20] identified a lack of consistent terminology, challenges with applying the concepts of sex and gender and using them to interpret results, and a lack of appropriate indicators on datasets as key challenges. Despite the priority given to integrating sex and gender in health research, there is also evidence that funding agencies have not given a great deal of practical guidance about how these dimensions should be integrated in research, instead leaving researchers the freedom and responsibility for developing their own plans and models. Several agencies and research groups have provided case studies of SGBA approaches within specific health research and healthcare settings [21‐28]. Examples include What a Difference Sex and Gender Make: A Gender, Sex and Health Research Casebook [18] and Rising to the Challenge: Sex- and Gender-Based Analysis for Health Planning, Policy and Research in Canada [29]. Researchers have also developed sex- and gender-specific tools in order to advance research and methods in this area. Doull et al. [30] have developed the Sex and Gender Appraisal Tool to assess systematic reviews with reference to sex and gender issues, and Briefing Notes [31] to aid researchers to implement sex/gender analysis in systematic reviews. Oertelt-Prigione et al. [6] have established GenderMedDB, an archive of medical literature containing sex- and/or gender-specific analyses.

Most of these examples focus on application of SGBA to specific research questions, within discrete research projects. In such projects, typically having one main source of data and a single research team, one might expect that researchers could easily determine how sex and gender might be important, and how to include these dimensions in the research design. For large and multi-centre research programmes, the intentional and systemic incorporation of sex and gender poses a greater challenge. There may be multiple sub-projects, and research questions and approaches might be emergent over the course of a programme. The extant literature provides few examples relevant to integrating a SGBA model in these larger organisations. This paper aims to begin to fill this gap by describing how considerations of sex, gender and vulnerable populations (GVPs) were promoted within a large multisite health research programme, reporting the results of an evaluation of the model and the lessons learned in its implementation.