2
Research Design and Methodologies
Research often ignores sex and gender, underrepresenting women in preclinical and clinical studies, and rarely analyzing outcomes by their sex or gendered dimensions. Integrating intentional design and novel, ethical methodologies can close knowledge gaps, personalize care, and reduce disparities in women’s health.
Despite evidence that sex and gender influence health outcomes, researchers have not consistently considered these influences in the design, data collection, outcome measurement, analysis, and reporting of research, leading to an incomplete understanding of how sex and gender influence health.
Preclinical research informs clinical trial design, which in turn informs individuals’ treatment. Therefore, female representation in all research stages is crucial to ensure an understanding of potential sex influences on health processes and outcomes. However, both basic research (e.g., animal models) and clinical research use males as the default. For example, a review of dermatological studies found that 60 percent of papers did not disclose the sex of studied cells or animals, and of papers that disclosed sex, 70 percent of the studied cells were male. In a review of phase I clinical trials for frequently prescribed drugs, only 22 percent of participants were female. While female participation in phase III trials has improved in recent years due to policy and regulatory changes, a review of over 56,000 articles and clinical trial records found that females were substantially underrepresented relative to the prevalence of key health conditions, particularly in studies of HIV/AIDS, chronic kidney diseases, and cardiovascular diseases. Female inclusion alone is insufficient: researchers should also meaningfully analyze outcomes by sex and gender, but such analyses remain limited. Moreover, sub-groups such as pregnant and lactating women are often excluded from trials, and gender minorities are often made invisible by research that adopts a gender binary or uses biological sex as a proxy for gender. As a result, the scientific community’s understanding of the benefits and risks of therapeutic options for these populations is limited.
In addition to increasing female inclusion in study design, researchers should leverage existing and novel methodologies to understand the underpinnings of conditions and diseases in women more comprehensively.
For example, novel technologies—from translational models like “organs on a chip” to machine learning and advanced data analytics—offer tremendous potential to deliver personalized care for women across the life course. Further developing these methodologies—and ensuring their ethical and unbiased application—can enhance risk identification, prevention measures, and treatment options fit for diverse populations
of women.
Integrating a sex and gender lens across all stages of research design (the structure of studies) and methodologies (the processes and tools used to collect, analyze, and apply data) will enhance the rigor and generalizability of research findings for women while filling critical knowledge gaps and enabling innovation tailored to women’s unique health needs. Sex- and gender-intentional design and methods are feasible and precedented. Their adoption will reduce disparities in diagnoses, treatment plans, and outcomes for women.
Overview Research Design and Methodologies
2.1 Sex/gender-intentional research design
Despite evidence that sex and gender matter to health outcomes, researchers insufficiently incorporate of sex and gender variables across research stages. Inconsistent use of sex and gender variables in data and analysis plans leads to statistically weak or non-existent methodologies to assess inter-group differences. Furthermore, methods to ensure representative and intersectional inclusion of different genders and sexes are often absent. These trends persist across cellular research, animal models, and medical and health research, including clinical trials
Progress Assessment
Progress made against Opportunities, from the 2024 Progress Report
Status Moderate Progress
0 % Achievement
Solution Strategies
-
Support academic and industry bodies to develop sex as a biological variable (SABV) and gender analysis policies at the national level, leveraging regional regulatory bodies and ethics review committees to establish policies consistent across regions and countries.
-
Advocate for research funders to establish SABV and gender analysis requirements for the research they fund—drawing on best practice examples that already exist—and to develop incentives for researchers to incorporate a sex- and gender-lens in their research (e.g., grants to offset the additional costs required to design studies that adequately capture sex differences, which may involve more time and resources). In tandem, advocate for journals and publishers to ensure stronger uptake of standardized SABV reviewing and reporting policies for manuscripts and published research, such as the Sex and Gender Equity in Research (SAGER) guidelines for scholarly literature and the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines for animal research.
-
Develop cross-functional teams (among academic research groups, pharmaceutical companies, etc.) that implement a community-engaged, adaptive design approach. This involves engaging women and relevant research stakeholders from high-income countries (HICs) and low- and middle-income countries (LMICs) in consultations, collaboration, and, ideally, shared leadership roles to understand key needs within specific disease areas and co-design solutions to meet those needs. An adaptive design approach would allow both soliciting women’s voices and needs while also following emerging insights from data, including big data.
-
Identify incentives for and pathways to cultivate champions of sex- and gender-intentional research design and analysis across sectors, especially in the pharmaceutical industry
2.2 Resource-sharing on research landscape in LMICs
Infrastructure and capacity to conduct preclinical and clinical studies vary among LMICs, and several challenges impede their development. Technical support for using and maintaining advanced equipment may insufficient, or facilities may lack the necessary equipment to conduct studies. Procuring reagents or kits can be unaffordable, especially with fluctuating exchange rates. As a result, LMICs often rely on shipping samples to HICs where further research and analysis is conducted. Basic research can be disconnected from clinical research; as a result, clinical trials remain disproportionately concentrated in HICs. Standardizing and certifying laboratories that can participate in preclinical and clinical studies will help democratize opportunities and globalize medical knowledge of different populations.
Progress Assessment
Progress made against Opportunities, from the 2024 Progress Report
Status Moderate Progress
0 % Achievement
Solution Strategies
-
Develop clinical trial methods that increase the representation of traditionally underrepresented demographics of women—including remote trial participation options, satellite sites, and wearables—and support the development of infrastructure and capacities to enable these methods and improve the clinical trial landscape (e.g., around challenges like limited insurance coverage in LMICs for clinical trials).
-
Establish regional hubs or centers of excellence to supply LMICs with the biomedical research supplies needed to conduct research and coordinate resource-sharing, technical support, validation, and intellectual property. Amplify these hubs by engaging local non-research community resources where available to proactively support women’s participation in research.
-
Establish a platform to negotiate affordable prices in LMICs for research supplies, similar to the African Medical Supplies Platform that was launched to leverage Africa’s bulk purchasing power to secure medical supplies during COVID-19.
-
Attract large biotech companies to establish presence in LMICs and nurture local biotech companies in LMICs to improve their capacities.
2.3 Computational & bioinformatics modeling and machine/deep learning
Machine learning, deep learning, and advanced data analytics are emerging tools that offer opportunities for novel and exciting applications of data. Advances in computing power enable researchers to improve data mining capacity and more efficiently combine and integrate information to generate new hypotheses. These approaches can enable a more complete understanding of complex biological pathways and diseases; this knowledge has the potential to accelerate innovation of better treatments and prevention strategies. However, researchers should take steps to mitigate potential biases in existing data sets—including biases due to race, gender, and socio-economic status. An assessment of the adequacy of existing data can determine if more comprehensive prospective data with reduced biases is needed. Investing in emerging data analytics methods can enable shared learning and more efficient research on key drivers of women’s health conditions.
Progress Assessment
Progress made against Opportunities, from the 2024 Progress Report
Status Unchanged Progress
0 % Achievement
Solution Strategies
-
Strengthen existing multi-country databases (including genomic and other omics data, and physical parameters) that can be leveraged via AI/ML methods to understand genetic linkages with women’s health conditions—including collaborating with or expanding the US National Human Genome Research Institute, UK Biobank, Human Heredity and Health Africa (H3Africa), and others, and establishing new efforts as needed.
-
Invest in a few illustrative use cases of developing computational models for women’s health conditions—including with collaborators and data from underserved settings—to generate actionable learnings and replicable practices for the field.
-
Strengthen the infrastructure and capacities required to undertake machine and deep learning applications in LMICs, including through mechanisms like grant opportunities, open-source software and resources, online training resources accessible to LMIC scientists, and more.
-
Strengthen and promote consistent implementation of data coding and quality standards—including the WHO’s Family of International Classifications, such as the International Classification of Diseases—to facilitate machine learning applications. Uniform standards across clinical records, surveillance data, research studies, and other data sources will enable vast amounts of data to be aggregated across multiple databases and platforms. Potential biases and inaccuracies in data sets can be identified by implementing quality standards, including checks for accuracy, completeness, validity, consistency, uniqueness, timeliness, fitness for purpose, and identification of miscoded data, missing data, and outliers. Such standards should be accessible across geographic regions and translated to multiple languages.
2.4 In-vitro translational models
Within the past decade, micro-physiological systems such as organ-on-a-chip and organoid technologies have emerged as promising alternatives to animal models for biopharmaceutical applications for women’s health. These technologies allow researchers to engineer living tissues and organ units within a controlled environment to mimic the complex biological activity of human organs better than conventional cell culture models while avoiding some of the limitations of animal models. For example, work is underway for organ chips and organoids models for the vagina, cervix, fallopian tube, placenta, and endometrium. Researchers are also investigating the use of organ-on-a-chip systems for human pregnancy, which is difficult to study in vivo. The potential benefits are enormous, as reasonable ethical restrictions limit research that would affect mother and fetus and few animal models can capture unique aspects of human pregnancy. These systems can allow for multiple cell types in three-dimensional culture using artificial extracellular matrix materials based on hydrogels to investigate cellular crosstalk including modeling maternal-fetal interactions. While these models are still in their infancy, they hold great potential to advance discovery and development for key women’s health applications. Given recent regulatory pathway openings such as the Food and Drug Administration (FDA) Modernization Act 2.0 (2022)—which newly allows for micro-physiological systems and computational models as alternatives to animal testing before clinical trials for drug development in the US—advancements in these areas are likely to accelerate in the coming years.
Progress Assessment
Progress made against Opportunities, from the 2024 Progress Report
Status Moderate Progress
0 % Achievement
Solution Strategies
-
Establish a bank of primary cells from diverse populations worldwide to inform more representative in vitro translational models.
-
Support public-private sector partnerships to overcome commercial hurdles for advancing and scaling in vitro translational and computational modeling. Especially with the passing of the FDA Modernization Act 2.0 in the US, many small companies and startups will be active in this space working on commercial solutions. A critical first step will be ensuring quality control early in device development prior to scaling.
-
Prioritize the development of affordable micro-physiological systems for use in under-resourced settings, including in LMICs. While paper-based point-of-care devices are available in under-resourced settings, extending the use of micro-physiological tools for research studies using simpler, more affordable approaches (including open, non-proprietary technologies) will extend their benefits to even broader populations.
-
Support technology transfer and training for these technologies to be made more widely available and adapted across geographic regions, including through educational and technology transfer research grants, development of open-source technologies, etc.
