Why blood pressure can remain elevated even with medication

There is no single reason blood pressure may stay stubbornly high, and treating every case the same can overlook the real obstacle. Yale researchers have created a method for sorting patients according to the barriers that are keeping their blood pressure elevated.

The approach distinguishes between situations in which treatment was not intensified, the prescribed plan was not carried out, or blood pressure remained high even after treatment efforts. Each of those patterns may require a different type of support.

Published in Circulation: Cardiovascular Quality and Outcomes, the study analyzed clinicians’ notes in electronic health records. It helps lay the foundation for automated tools, including artificial intelligence, but it did not show that an AI system had already improved blood pressure outcomes.

The researchers identified 1,664 adults in the Yale New Haven Health System who had blood pressure readings of at least 160/100 at five or more straight outpatient visits. That definition focused the study on repeated, markedly elevated measurements rather than a one-time result.

They randomly chose 200 records for review. After examining 115 of them, they found that no new categories were emerging, which allowed them to define the main patterns consistently.

One category involved treatment not being stepped up even though high readings continued. In these cases, the care process did not lead to a stronger medication plan once the ongoing problem was recognized.

The second category reflected a gap between what was prescribed and what actually happened afterward. Trouble getting medicines, side effects, practical barriers, or other obstacles can keep a plan from becoming part of daily care.

The third category covered cases in which blood pressure stayed high despite documented attempts to increase treatment. These patients may need a different evaluation than someone whose main challenge is getting access to or following an existing plan.

The goal is to turn information already stored in medical notes into a clearer next step. A practical access issue, for example, calls for a different response than a condition that seems resistant to several medicines.

The researchers also pointed to ways this classification could be automated through electronic records. AI may help find relevant details in large volumes of clinical text, but any such system would still need testing to confirm that it classifies patients accurately and supports better care.

The original work was a qualitative analysis from one health system, not a trial measuring medical outcomes. The categories may also overlap, since the same person can face more than one barrier over time.

The study’s main contribution is making those barriers visible instead of treating every persistent reading as the same problem. That could help future care programs direct the right resources to the reason treatment is falling short.

If you are interested in this topic, please read articles about what to know about blood pressure medicines and their side effects, and how the DASH diet can help lower high blood pressure.

For more information, please see articles about salt and heart health: does less really mean more?, and how the Mediterranean diet may help lower blood pressure naturally.