To the editor

We have read with great interest the article by Stopa et al1 comparing the effectiveness of unattended automatic office blood pressure measurement (UAOBPM) with conventional office readings in predicting hypertension-mediated organ damage, especially considering 2 excellent accompanying editorials.

The conclusions drawn in the original paper are intriguing. We expected that, similarly to ambulatory methods, such as home blood pressure monitoring (HBPM) and 24-hour ambulatory blood pressure monitoring (ABPM), UAOBPM would prove superior to office blood pressure measurement (OBPM) in predicting hypertensive organ damage.2,3

Why then were different conclusions drawn?

The value of BP measurements lies in predicting hypertensive complications that usually develop over a long period of time. In our opinion, the cross-sectional methodology of this study does not allow for verification of the research hypothesis relating to the predictive value of BP measurements, and such a goal can only be achieved using retrospective or prospective observation.

Secondly, BP is not the only parameter influencing the risk for hypertensive complications. During modeling, cofactors such as age, sex, nutritional status, number of antihypertensive medications, smoking, and comorbidities should be considered, and only then models containing BP values measured with different methods should be compared. This thesis is supported by the fact that in most of the receiver operating characteristic curve analyses presented in the article, the confidence interval of the area under the curve almost always included the value of 0.5. In light of these findings, it seems that the value of diastolic BP is no better than that of a coin toss for predicting any of the hypertensive complications mentioned in the article, and that systolic BP is not useful for predicting left ventricular hypertrophy, relative wall thickness greater than 0.43, systolic and diastolic dysfunction, presence of atherosclerotic plaque, or albumin-creatinine ratio. Considering our existing knowledge about BP measurement, such results generate questions.

Ultimately, the conclusion that UAOBPM is not superior to OBPM seems to be questionable, because the study power analysis was based on the assumption of differences in BP values between the 2 methods, not the predictive value of the created models.

Therefore, we urge not to dismiss UAOBPM based on a single article. Indeed, the only prospective study demonstrating the relationship between UAOBPM and cardiovascular events is the Systolic Blood Pressure Intervention Trial.4 However, we have also published a post-hoc analysis presenting target diastolic BP values using this measurement method.5 Cuffless BP measurement methods combined with wearable technologies are emerging, but we still have more data on the prospective relationship between cardiovascular events and UAOBPM-based BP values than for BP values measured using these new technologies. Also, HBPM cannot replace OBPM or ABPM, simply because of limited accuracy (or confidence in the accuracy) of HBPM. With OBPM, UAOBPM, and ABPM, we can make every effort to ensure the measurement is done correctly. Moreover, UAOBPM should primarily replace ABPM, which is expensive and requires 4 visits before a therapeutic decision is made (ordering of ABPM, device fitting, device removal, and the actual medical consultation). In our opinion, broader implementation of UAOBPM could reduce the use of ABPM and redirect efforts toward better achieving therapeutic goals, which, as emphasized in one of the editorials accompanying the discussed article, is the main issue in treating hypertension.