Visitor restrictions are in place due to COVID-19. Review the latest information about the virus and how you can help by donating funds.

Notice of Privacy Incident. Learn More.

Science and Research

How Numbers Can Predict Hypertension

Hypertension can be a difficult condition to predict and prevent. But new technology may provide better treatment for the “silent killer,” deemed so due to the lack of warning signs preceding the strokes, coronary artery disease and congestive heart failure it can cause.

By applying algorithms to electronic health records, scientists from Northwestern Medicine successfully and accurately identified patients with undiagnosed high blood pressure. Of the patients identified, many were then formally diagnosed with hypertension while a high number were found to have related blood pressure conditions such as prehypertension, white-coat hypertension or elevated blood pressure.

Screening Past the White Coat Effect

Hypertension can go undiagnosed because of understandable human errors. If someone is seeing multiple physicians, one set of test results may be explained as the effect of a sick day, and another from the rush of setting a last-minute appointment. When the readings are left separate, physicians can miss a pattern of elevated results.

The white coat effect is a common obstacle to the successful diagnosis of hypertension, often used to account for falsely elevated blood pressure results associated with patients being near a health care provider.

In the Northwestern Medicine study, patients identified as at risk for undiagnosed hypertension participated in voluntary automated office blood pressure testing that gathered six readings after the medical assistant left the room, throwing out the first result to eliminate the white coat effect.

Through monitoring and follow-ups after the initial testing, the Northwestern Medicine scientists, led by principal investigator Michael K. Rakotz, MD, assistant professor of Clinical Family and Community Medicine at Northwestern University Feinberg School of Medicine, created a system that notifies medical staff and primary care providers of at-risk patients.

Given the algorithm’s success with hypertension, the scientists are hopeful that similar screenings of EHRs for other commonly undiagnosed chronic diseases may be possible.