Determining the prevalence of autoimmune diseases in the U.S.
Aaron Abend, M.B.A.Executive Director, Autoimmune Registry; Managing Director, Prognosis Data Corp; Member, All of Us Advisory Panel
Emily HolladayMasters in Public Health (MPH) student, The University of Alabama at Birmingham
For years, clinical observations have suggested that cases of autoimmune disease are rising, but few studies have used systematic data to test this hypothesis. Current estimates of how many people in the United States live with autoimmune diseases are based on informed guesses from knowledgeable clinicians. To more accurately calculate autoimmune disease prevalence and assess the extent of its impact on public health, a scientifically sound methodology is needed. The search for this methodology is what brought technology entrepreneur Aaron Abend and his colleagues at the Autoimmune Registry (ARI)—a hub for research, statistics, and patient data on autoimmune diseases—to the All of Us dataset.
Together with a team of young investigators, including early career researchers and students like Emily Holladay, Aaron is using the data available in the Researcher Workbench to more accurately calculate the likelihood of having an autoimmune disease.
The dataset’s diversity and accessibility make it an ideal place to conduct this type of research. Still, like any researcher working with a large dataset, the team must correct for variables in their analysis.
For example, type 2 diabetics will sometimes be incorrectly coded in the electronic health record (EHR) with the autoimmune type 1 diabetes, which could lead to overcounting people with autoimmune diseases. In addition, due to its focus on building a diverse dataset, the All of Us cohort is not demographically representative of the U.S. population. For example, Black women are more likely to develop an autoimmune disease and represent nearly 12% of the All of Us participant cohort while accounting for 6.6% of the U.S. population.
While calculating the prevalence of autoimmune diseases in the country, Aaron and his colleagues are working to account for this kind of selection bias. These calculations will enable the ARI team to more accurately extend their findings within the All of Us participant cohort to the U.S. population. They hope their research will also help identify individuals who are at risk or who are just beginning to develop symptoms of autoimmune diseases and are not currently in the care of a medical professional. For students like Emily, access to a dataset of this size and scope can be transformative.
By focusing on the determinants, distributions, and diagnoses of autoimmune diseases, we hope to better define prevalence. It typically takes around 4 to 5 years to diagnose an autoimmune disease, but some spend 10-15 years trying to get the correct diagnosis. As both a patient and a researcher, it is encouraging to know I can contribute to improving the understanding and awareness of these diseases.
– Emily Holladay
In addition to the All of Us dataset, ARI partners with researchers at the Keck School of Medicine at the University of Southern California and the Weill Cornell Medical College to validate their work. Disease prevalence data from All of Us in combination with lab tests and the patient’s signs and symptoms from ARI partner, can speed up the journey to a correct diagnosis. The team is excited to begin working with the All of Us Research Program’s genomic data, newly released in March of 2022, to develop statistical models that will predict autoimmune disease and patient outcomes.
Key benefits of All of Us data and resources
“Most health data I have worked with is from the doctor’s point of view or the patient’s point of view. The All of Us database provides both, plus rich geographic and socioeconomic data, that allows the kind of research that we hope will help doctors better understand patients, and patients to better understand these complex diseases. And the diversity of the data ensures our work will be relevant to all Americans, including members of groups that have historically had limited access to the health system.”
– Aaron Abend
As someone who has spent his career navigating big datasets, Aaron immediately recognized how unique it was to have access to protected EHR data that can be combined with other data types, such as survey responses, to gain a more complete picture of an individuals’ health.
Aaron also frequently directs students and ARI collaborators to the training and support tools available through the All of Us Research Hub and the All of Us Researcher Workbench as a way to get their questions answered clearly and quickly. Between the richness and longitudinality of the dataset and the support tools available to researchers, the ARI team views All of Us as a key component in unlocking discoveries about autoimmune disease prevalence in the United States.