It’s a rare occasion when my worlds of biomedical informatics and serialized lesbian melodrama fandom collide.
But that’s exactly what happened earlier this summer when two of my favorite actresses appeared on a popular podcast. I was excited to hear them talk about their new book and their history of working together, so I was confused but delighted when their conversation took a turn toward my area of expertise — electronic health records.
One actress noted that on a recent trip to the optometrist, she was asked about her ethnicity. “And I was like, what difference does it make?” she said.
The host chimed in with her experience of being asked similarly personal questions before a mammogram.
“Like, it doesn’t matter if I’m married or not. It doesn’t matter if I’m white or Asian, you know?” she remarked.
Listening to the host and actresses question a process that, to me, seems straightforward and purposeful served as a stark reminder of the chasm that often exists between how researchers like me use patient data and a patient’s actual experience of clinical data collection.
For those of us who use demographic data collected during health care encounters to conduct research and design interventions, it does matter whether patients answer their doctor’s demographic questions. But as a patient myself, I can see how these questions might seem unnecessary and even invasive.
So it may help to understand why your doctors collect this data, how researchers use it and what medical discoveries might be possible when we know more about who patients are.
Why your data matters
When you answer the demographic questions your doctor logs in your electronic health record, you’re doing more than disclosing personal information. You’re adding one small piece to a giant puzzle of data that allows researchers like me to see a bigger picture.
Your health information can help us understand who gets sick and why. It might even be used to design real health interventions.
As a researcher focused on improving health and health care for moms and their babies, I consider myself lucky to live in Indiana, a state with one of the nation’s most comprehensive health information exchanges.
These exchanges are interconnected networks of hospital system electronic health record databases from all over the state that allow researchers like me to learn about how individuals and groups experience health and medical care.
For example, my colleagues and I in the Indiana University Better AI for a Strong Rural Maternal and Child Health Environment Lab use this data to train machine learning models that predict preeclampsia, a life-threatening condition of high blood pressure during pregnancy, before a mom gets really sick.
We could use only clinical data: diagnoses, labs and vital readings like blood pressure that contribute to the outcome of preeclampsia. But for conditions like preeclampsia, Black moms are diagnosed at higher rates than their white counterparts.
Research shows that race and racism can be major contributing factors to this disparity.
