Douglas Krakower, MD, explains how predictive models can be used to identify patients that are at a higher risk for HIV and could benefit from learning about PrEP.
Segment Description: Douglas Krakower, MD, assistant professor of medicine, Harvard Medical School, explains how predictive models can be used to identify patients that are at a higher risk for HIV and could benefit from learning about PrEP.
Interview transcript: (modified slightly for readability):
“I think one of the things about the study that was most interesting was we used a predictive model to give everyone in the population at this health center an HIV risk score. And so, you could stratify the whole population in terms of their risk scores.
If you look closely at the distribution of the scores, it looks like the population, in general, most people have a relatively low risk score. But towards the right end of the curve, there is a steep increase; in people with the highest risk scores, you tend to see a steep take off.
This group where the risk scores are much higher than the general population, if they’re not already using pre-exposure prophylaxis (PrEP), these are candidates for a conversation about PrEP.
One of the nice things about the study is that it can help you be really efficient as a health care organization in terms of using your electronic health record (EHR) data to target your prevention efforts for a subset of your whole population. Instead of having to inform clinicians to use the same level of intensity of screening across the population, you can hopefully have them target a smaller population which is more efficient and maybe more effective.
You’d never want to abandon risk assessment for people who may be identified as low risk by the model, because the model is not perfect, but it’s a way to try to be efficient as you can.
In an era in which clinicians are asked to do so many things and there are so many competing interests, we hope this can move things forward in terms of impact.”