Researchers harness EHR data to identify patients with undiagnosed hypertension
A new study successfully demonstrated how EHR algorithms can identify patients with previously undiagnosed hypertension with a high rate of accuracy.
Researchers from Chicago-based Northwestern Medicine developed algorithms, which reviewed records from 1,033 patients, according to the Annals of Family Medicine study. Of these, 361 were formally diagnosed with hypertension and 290 others were diagnosed with related blood pressure conditions such as prehypertension, white-coat hypertension or elevated blood pressure.
During the first half of the study, patients were identified as being at risk for undiagnosed hypertension using the three EHR algorithms and were invited to complete an automated office blood pressure (AOBP) test—which was created to eliminate falsely elevated blood pressure results associated with patients being in the presence of a healthcare provider.
In the second part of the study, additional patients at risk for unidentified hypertension were identified with the same three EHR algorithms, and were then observed over the course of two years. Follow-up appointments were arranged for such patients.
All of the primary care physicians that participated in the study also received monthly lists of their patients that the algorithms identified as being at risk of hypertension. Patients remained on the physicians' lists until an AOBP evaluation was completed or a diagnosis was entered into the chart that indicated the patient's at-risk status had been resolved.
"With this study, we created a surveillance system that notifies the medical staff and the primary care physician anytime a patient who is at risk arrives in the office," said principal investigator Michael K. Rakotz, MD, a family medicine physician at Northwestern Medicine Evanston, in a statement. "Once these patients are identified we proceed with an AOBP to more accurately measure their blood pressure and make a diagnosis. This surveillance system never stops running, so any patient that meets the EHR algorithm criteria for possible hypertension will automatically be flagged. In doing so we hope to put an end to undiagnosed hypertension."
The study's authors conclude that this method also could be applicable to other commonly undiagnosed chronic diseases.