How to reliably diagnose arterial hypertension: lessons from 24 h blood pressure monitoring

BACKGROUND: Hypertension is a common condition in modern society. As blood pressure fluctuates with time, a single blood pressure measurement is useless to diagnose hypertension. Nevertheless, no well-defined number of measurements is often used for this purpose. Diagnosis and therapeutic control of hypertension are therefore suboptimal. OBJECTIVE: To determine the number and timing of measurements needed to give a trustworthy approximation of an individual's average blood pressure. METHODS: In this observational study 306 clinically indicated 24h ABPM datasets were analysed. Hypertension was defined as a daytime blood pressure mean exceeding 135/85 mm Hg. Kappa coefficients determined the best time of day for measuring blood pressure. The optimal number of measurements was estimated using canonical correlation. RESULTS: 162 (53%) patients were diagnosed with hypertension. Kappa statistics indicated that measuring during the afternoon gave the best agreement with the 24h blood pressure mean (κ = 0.78). According to canonical correlation, about 8-10 blood pressure readings give enough information for hypertension diagnosis. CONCLUSIONS: Eight to ten blood pressure measurements between 01:00 and 05:00 p.m. are sufficient to give a clinically useful approximation of the daytime mean blood pressure and therefore for diagnosing hypertension accurately. Future research should determine the ideal dispersion of measurements and include patient characteristics which could influence the required number and timing of measurements. These results may increase the future importance of telemonitoring in diagnosing hypertension.