Home, automated office, and conventional office blood pressure as predictors of cardiovascular risk

摘要

Automated office blood pressure (AOBP) has recently been shown to closely predict cardiovascular (CV) events in the elderly. Home blood pressure (HBP) has also been accepted as a valuable method in the prediction of CV disease. This study aimed to compare conventional office BP (OBP), HBP, and AOBP in order to evaluate their value in predicting CV events and deaths in hypertensives. We assessed 236 initially treatment naive hypertensives, examined between 2009 and 2013. The end points were any CV and non-CV event including mortality, myocardial infarction, coronary heart disease, hospitalization for heart failure, severe arrhythmia, stroke, and intermittent claudication. We fitted proportional hazards models using the different modalities as predictors and evaluated their predictive performance using three metrics: time-dependent receiver operating characteristics curves, the Akaike's Information Criterion, and Harrell's C-index. After a mean follow-up of 7 years, 23 participants (39% women) had experienced >= 1 CV event. Conventional office systolic (hazard ratio [HR] per 1 mm Hg increase in BP, 1.028; 95% confidence interval [CI], 1.009-1.048), automated office systolic (HR per 1 mm Hg increase in BP, 1.031; 95% CI, 1.008-1.054), and home systolic (HR, 1.025; 95% CI, 1.003-1.047) were predictive of CV events. All systolic BP measurements were predictive after adjustment for other CV risk factors (P < .05). The predictive performance of the different modalities was similar. Conventional OBP was significantly higher than AOBP and average HBP. AOBP predicts equally well to OBP and HBP CV events. It appears to be comparable to HBP in the assessment of CV risk, and therefore, its introduction into guidelines and clinical practice as the reference method for assessing BP in the office seems reasonable after verification of these findings by randomized trials. Copyright (C) 2017 American Society of Hypertension. All rights reserved.