Am Fam Physician. 2004;70(8):1558-1561
Clinical blood pressure measurement with a mercury sphygmomanometer is subject to interobserver variability. In addition, blood pressure measurement is only intermittent, and white-coat hypertension might confound results. Ambulatory blood pressure monitoring has been shown to have better prognostic value than office monitoring, but home blood pressure monitoring is cheaper and more readily accepted by patients. However, only one prognostic study of cardiovascular morbidity and mortality comparing home and office blood pressure measurements has been conducted. Bobrie and colleagues investigated the prognostic value of home and office blood pressure measurements in a population of European patients with hypertension.
The Self-Measurement of Blood Pressure at Home in the Elderly: Assessment and Follow-up (SHEAF) study was a prospective cohort study with a two-week assessment phase and a three-year follow-up phase. The primary end point was cardiovascular mortality, and the secondary end point was all-cause mortality and the combination of all cardiovascular mortality. In the initial two-week phase, multiple office blood pressures were recorded at two visits, and home blood pressure was measured serially over four days. In the analysis, patients were divided into four groups: patients with “controlled” hypertension (office blood pressure below 140/90 mm Hg and home blood pressure below 135/85 mm Hg); patients with “uncontrolled” blood pressure over those limits; patients whose home blood pressure was normal but whose office blood pressure was high; and patients whose office blood pressure was normal and whose home blood pressure was high.
The prognostic value of home blood pressure measurements was analyzed at the first composite end point that occurred during follow-up. Only valid measurements, defined by specific criteria, were included in the analysis. The mean of six measurements defined office blood pressure, and the mean of 27 measurements defined home blood pressure.
At baseline, 13.9 percent of patients had controlled blood pressure by both measurements, 13.3 percent had elevated office blood pressure only, 9.4 percent had elevated home blood pressure but not office blood pressure, and 63.4 percent had uncontrolled hypertension by both measurements. Cardiovascular mortality and morbidity were known for 4,928 patients (99.8 percent) at the end of the mean three-year follow-up. There were 205 deaths, of which 85 were cardiovascular-related.
After adjustment for demographic and risk-factor variables, home blood pressure was predictive of the occurrence of cardiovascular events, with systolic blood pressure linked to prognosis in both sexes, and diastolic blood pressure linked to prognosis in men and trending to significance in women, whereas neither systolic nor diastolic office blood pressure was linked to prognosis. With home measurement, for each 10 mm Hg increase in systolic blood pressure, the cardiovascular event risk increased 17.2 percent, and for each 5 mm Hg increase, the cardiovascular event risk increased 11.7 percent. With office blood pressure measurement, there was no increase in event risk with either measurement level. Neither home nor office blood pressure measurement was significantly related to the secondary end point, all-cause mortality.
This study, which demonstrates the prognostic value of home blood pressure measurement, identified a subgroup of patients with poor home control of hypertension that appeared to be controlled in the physician’s office (“masked” hypertension). It also showed that the prognosis in patients with elevated office blood pressure and normal home blood pressure was similar to that in patients who had controlled hypertension in both places.
There was no prognostic value for the combined end point of all-cause mortality because of the small number of events. The results of larger studies have shown correlation between office measurement and cardiovascular morbidity and mortality, contradicting this study, but the larger studies had much greater statistical power.
The authors suggest that blood pressure monitoring should include home measurement to identify patients who have white-coat hypertension and those with masked hypertension. However, further studies are needed to determine whether, in identifying these patients, adjustment in treatment improves outcomes.