In some patients it is difficult to accurately assess blood pressure because of a “white-coat effect” that results in an elevated measurement when blood pressure is tested by medical personnel. In these cases, self-monitoring by patients or family members is recommended to obtain more representative blood pressure measurements. However, a few patients report even higher blood pressures when self-monitoring than when they are monitored by health professionals. This “inverse white-coat” phenomenon could be an intrinsic response to the act of measuring blood pressure or a result of poor measurement technique. Turnbull and colleagues investigated the inverse white-coat phenomenon in 173 patients in an English general practice.

Participants in the study were consecutive patients who were at least 20 years of age, regardless of treatment status for hypertension. The only exclusion factor was cardiac arrhythmia. Each participant was instructed by a nurse in measuring blood pressure using an oscillometric device. The nurse measured blood pressure on each patient with the oscillometer and with a mercury sphygmomanometer at baseline (phase 1), one month (phase 2), and two months (phase 3). Patients and nurses were not aware of each other’s results.

At baseline, patient recordings were significantly higher than those made by the nurses. Only 59.5 percent of systolic readings and 75 percent of diastolic readings taken by patients were within 10 mm Hg of the measurements made by nurses. Patients also recorded significantly higher systolic and diastolic readings in phases 2 and 3. The mean differences in systolic measurements between patient and nurse recordings were 4.7 mm Hg to 6.9 mm Hg across the three phases. The mean differences in diastolic measurements between the two groups were lower (2.7 mm Hg to 3.2 mm Hg) but still statistically significant.

The authors conclude that only one half of patients got blood pressure measurements within an acceptable range of the readings obtained by trained nurses. Patient readings were consistently higher than those taken by nurses. They suggest that further studies investigate if the effect is caused by faulty technique, patient anxiety, or other factors.

editor’s note: The results of this study cause concern since physicians frequently make treatment decisions based on blood pressure readings taken by patients. I had always assumed that errors in those readings would be falsely low. If the opposite is true and the errors are of the size reported by this study, we risk over-treating our patients. The results could be disastrous, particularly in elderly patients. The findings of this study should be examined in a larger group of patients who consistently take blood pressures at home.—a.d.w.