Including disability in administrative health data research to improve health equity
Over a quarter of the Canadian population experiences a physical, sensory, intellectual or developmental disability. But the failure to track disability in population level research contributes to health inequities for people with disabilities, according to Dr. Hilary K. Brown, Associate Professor at the University of Toronto. “Disability is very common, it’s considered to be the most common minority population. Yet it is also often excluded from research conversations about health equity and population health. This is why the inclusion and measurement of disability is critical in population based research.”
Dr. Brown and her colleague Dr. Yona Lunsky are adjunct scientists at HDRN Canada member organization ICES. They discussed their experience applying disability-related algorithms in research using administrative health data at a recent Big IDEAs about Health Data webinar. Dr. Brown’s research focuses on maternal and child health and mental health for populations with disabilities and chronic disease. Dr. Lunsky, who is the Scientific Director of the Azrieli Adult Neurodevelopmental Centre, has a focus on the mental health needs of individuals with intellectual and developmental disabilities and their families.
Disability is very common, it’s considered to be the most common minority population. Yet it is also often excluded from research conversations about health equity and population health. ~ Dr. Hilary K. Brown
Administrative health data, which are generated through routine encounters with health care programs and services, are a valuable resource for population based research. Yet there are a number of challenges when it comes to including disabilities in administrative health data research. “Because administrative data are often collected for purposes other than research, we are limited in terms of what we have access to in the data sets available,” says Dr. Brown. “When we think about the algorithms that are available to look at disability in administrative data, we are somewhat constrained by the data fields.”
While there are algorithms that use diagnostic codes to identify certain disabilities, Dr. Brown explains they come with a number of limitations. Diagnostic codes, which provide a standardized method for describing various medical conditions, can be inaccurate or incomplete. There are also certain disabilities, like chronic pain, that don’t have a diagnosis and can’t be captured in this type of data.
Another issue with diagnostic algorithms is that most people’s interactions with the health care system are focused on immediate or urgent problems, so disabilities that affect people over a lifetime tend to be under-reported in health administrative data. Dr. Lunsky conducted a comparison of the prevalence of intellectual and developmental disabilities in adults using health administrative data versus social services data and found that health administrative data only captured two-thirds of individuals with those disabilities.
“To improve the measurement of disability in administrative data, we need to move beyond just being able to capture diagnoses,” said Dr. Brown. “There are ways of measuring self-reported disability that have been used in routine reporting in home care and rehabilitation services in several countries that could be useful,” she added. “Including self-reported measures of disability as an equity-related variable, like race or gender identity, in health records will improve our ability to measure disability with administrative data.” This health equity information can be used to plan health care policies that are inclusive to people with disabilities.
Beyond the question of how disability can be measured in health research, Dr. Lunsky emphasized the need to include the disability community in health research throughout the study process. “Accessibility should be a key consideration for researchers from how the study is planned to how the data are interpreted and findings are shared. There’s a really important mantra that comes from the disability community: ‘Nothing about us without us.”
In reflecting on their experiences conducting community-engaged research with populations with disabilities, both researchers recommend taking time to build relationships, making room for flexibility in grant budgets to ensure advisors are adequately compensated and being responsive to community input in order to ensure research is meaningful and impactful. “When you’re doing work that matters to people about their health,” said Dr. Lunsky, “there is an obligation to build trusting relationships with the people for whom the research will make a difference.”
View the recording of Dr. Brown’s and Dr. Lunsky’s talk, Administrative Data & Working with Disability Communities.