BMJ team develops QUALITY-EDI to measure inclusion across healthcare research
Healthcare research has long depended on study populations that do not always reflect the patients who ultimately receive care. The result is familiar and clinically consequential: evidence can be difficult to apply to groups who carry disproportionate disease burden yet remain under-represented in trials, observational studies, and qualitative work. A new quality assessment tool, QUALITY-EDI, is designed to put structure around that problem by scoring how well equality, diversity, and inclusion are built into a project from planning through dissemination.
The tool was created as a concise framework for researchers, editors, reviewers, funders, lay members, and other interest holders to assess and monitor the quality of EDI implementation in healthcare research. Its goal is not abstract advocacy. It is to make inclusion visible, measurable, and harder to ignore.
Eight domains now define EDI quality across the research cycle
QUALITY-EDI was refined by a multidisciplinary working group of 22 members drawn from five countries, including clinicians, methodologists, statisticians, qualitative researchers, epidemiologists, clinical research staff, journal editorial personnel, and public representatives. Two public representatives were embedded in the executive group and took part in co-design work, with reimbursement aligned to National Institute for Health and Care Research guidance. The breadth of that group matters because inclusion failures rarely arise from a single point in the research process; they accumulate across recruitment, design, analysis, and publication.
The final tool contains eight domains: research context; study design; data collection; data analysis and interpretation; knowledge mobilisation; research team diversity; patient and public involvement; and budgeting for research inclusion. Each domain has two subdomains. In practice, the structure moves EDI from a broad ethical principle into a sequence of observable decisions. A trial can claim inclusivity, but this tool asks where the evidence actually sits.
Why inclusion remains a methodological issue, not only a moral one
The development rationale is rooted in persistent healthcare inequalities. Minoritised groups are often at higher risk of certain conditions, less likely to access medical care, and less likely to be represented in research. When those gaps persist, the evidence base becomes narrower than the patient population it is meant to serve. Generalisability weakens, and bias can enter through exclusion that is deliberate in some settings and inadvertent in others.
Research populations with greater heterogeneity permit a more credible assessment of how a diagnostic test or intervention performs across subgroups and settings. In biomarker studies, device evaluations, interventional trials, and implementation research alike, inclusion changes the external validity of the evidence. If underserved groups are missing, a promising result may still fail at the bedside because the setting, the participants, or the barriers to access were never represented in the study design.
A single design choice can distort the usefulness of an entire evidence package.
The NIHR toolkit provided the starting point for the scoring framework
The project began with the National Institute for Health and Care Research Research Support Service EDI toolkit, originally developed in 2022 and built around eight topics: historical context and structural inequality; selection of participants, sites, and samples; data collection; data analysis and presentation; dissemination, implementation, and impact; research team; public involvement; and budgeting for inclusion. That toolkit was intended as guidance rather than an evaluative instrument, but it provided a practical scaffold for the new assessment tool.
No comparable EDI quality assessment tools for healthcare research were identified at the time of development. That absence left a real gap. Researchers could describe inclusive intentions, but there was no standardised method to judge whether those intentions were carried through the work itself. QUALITY-EDI was built to address that gap by translating best-practice concepts into a more direct assessment structure.
Patient and public input shaped the tool rather than decorating it
A half-day workshop with patient and public representatives was held at Hammersmith Hospital Campus, Imperial College London, to review and co-design elements of the proposed tool. Participants were recruited through charities, community partnerships, online platforms, and existing patient and public networks. This is where the project moved beyond expert consensus alone. Public involvement was not treated as a ceremonial addition; it was used to refine the tool’s content and wording.
The workshop fed into later refinement by the working group, which adjusted the framework before settling on the final eight-domain structure. The process aligns with a straightforward principle of applied research: inclusion cannot be evaluated only from the investigator side of the table. Patients and public contributors can identify where access barriers, stigma, communication failures, or assumptions about participation are likely to enter the process. Those lived experiences are central when the outcome of interest is not just scientific rigor but whether research is usable by the communities most affected by disease.
A tool for editors, funders, and reviewers could alter what gets rewarded
QUALITY-EDI is aimed at multiple users, not only study teams. Journal editors and peer reviewers can use it to appraise whether inclusion has been built into a manuscript rather than mentioned in passing. Funders can use it when deciding which proposals deserve support. Lay members can use it to ask clearer questions about who was invited, who was missing, and whether budgets actually allowed meaningful participation.
Budgeting for research inclusion deserves special attention. Inclusion costs time, staff effort, access support, engagement work, and often extra communication. If those costs are not anticipated, exclusion becomes the path of least resistance. By making budgeting a formal domain, the tool recognizes that EDI is not merely a values statement; it is an operational commitment that has to be financed.
The same logic applies to research team diversity and knowledge mobilisation. Teams that lack diversity may miss barriers that are obvious to affected communities. Dissemination plans that do not reach those communities can leave the evidence stranded in journals, unshared where it might change care.
Application to gastrointestinal cancer research shows how the framework can be used
The project also applied QUALITY-EDI to a contemporary set of articles focused on early detection of gastrointestinal cancers. The article body available here does not provide the detailed scoring output, but the choice of topic is instructive. Cancer pathways often involve screening, diagnostic performance, referral patterns, and access to specialist care, all of which can be distorted if underserved groups are not adequately represented in the underlying research. A quality tool that inspects inclusion at each stage is especially relevant in that setting, where minor differences in access can affect whether disease is caught early or late.
By placing EDI assessment alongside study design and analysis, QUALITY-EDI creates a framework for asking whether research is merely representative on paper or genuinely built to support equitable benefit. The distinction influences how confidently clinicians can apply findings to their own patient populations.
What the tool can and cannot solve
QUALITY-EDI does not solve healthcare inequality on its own. It does not replace better recruitment, community trust-building, or structural changes in access to care. It also does not guarantee that every study will become fully inclusive. What it does offer is a standardised way to detect omissions, compare studies, and hold research teams to a clearer account.
The most useful effect may be indirect but durable: if researchers know their EDI methods will be appraised with the same seriousness as their outcomes, more of them will plan inclusion from the outset. That pressure could influence protocols, grant applications, peer review, and reporting norms. The tool’s value will ultimately depend on whether those users adopt it consistently and whether future studies show that it changes research conduct rather than merely documenting it.
For now, the central limitation is straightforward. QUALITY-EDI has been developed and refined, but broader validation across more research settings will be needed before its impact on actual inclusion practices can be measured.