Incidence of overestimation and underestimation of hypertension in a large sample of Canadians with mild-to-moderate hypertension

OBJECTIVE: To determine the prevalence of overestimated and underestimated hypertension in a large sample of Canadians with mild-to-moderate hypertension before and after antihypertensive treatment. DESIGN: Data were retrieved from files of 1019 patients in whom clinic and ambulatory blood pressures had been measured both at the start and at the end of several studies involving various pharmacological antihypertensive agents. METHODS: Overestimation and underestimation of hypertension were calculated from the difference between office and awake ambulatory blood pressure both for systolic and for diastolic values (standard criteria) and also from the difference between observed ambulatory blood pressure and the predicted level estimated from individual office blood pressure by means of a regression equation derived for the group (criteria of Perloff et al.). RESULTS: Approximately 20% of the patients satisfied both sets of criteria for overestimation of systolic and diastolic hypertension before treatment. A significantly greater proportion of women showed overestimation of diastolic hypertension compared with men (approximately 27 and 17%, respectively, P < 0.05) on the basis of either set of criteria. Treatment did not affect the prevalence of overestimation of hypertension for the group as a whole, but the proportion of women in whom diastolic hypertension was overestimated decreased to above 17% after treatment (P < 0.001). The same proportion of patients satisfied the criteria of Perloff et al. for underestimation of hypertension, but only about 10% satisfied the standard criteria for this condition, both before and after treatment. The reproducibility of overestimation or underestimation of hypertension was 30-40%, independently from the criteria used. CONCLUSIONS: About one-fifth of patients display overestimated hypertension and, according to the criteria of Perloff et al., as great a proportion show underestimated hypertension. Thus, as many hypertensive patients may be overmedicated and undermedicated on the basis of clinic blood pressure compared with ambulatory blood pressure measurement.