Emergency department hypertension and regression to the mean

To evaluate the importance of regression to the mean in the assessment of asymptomatic hypertension in the emergency department.

This was an historical cohort study of patients in the adult ED of a large urban teaching hospital. THe main outcome was changed in diastolic blood pressure (DBP). Subjects were 195 consecutive hypertensive patients with two sets of vital signs. Patients with specified acute conditions potentially associated with abnormal blood pressure were excluded, as were patients given vasoactive medications.

A statistical formula was used to predict the average blood pressure for hypertensive patients, using the observed mean and standard deviation of an all-patient sample. Given a threshold of 90 mm Hg, the expected mean DBP for hypertensive patients was 102.7 mm Hg, compared with an observed value of 104.5 mm Hg. Given an observed correlation coefficient of .73 between first and second measurements, a formula for regression to the mean predicted a spontaneous blood pressure decline of 7.2 mn Hg. A mean decline if 11.6 mm Hg was observed. The decline of 4.4 mm Hg more than expected among asymptomatic hypertensives was similar to the spontaneous decline of 3.7 mm Hg observed in the all-patient sample.

Patients who present with asymptomatic hypertension in the ED on average experience a spontaneous decline in blood pressure after they arrive. Most of this effect can be explained by regression to the mean. A small amount of this drop may represent attenuation of an initial alerting reaction.