Authors’ reply
Thank you for the opportunity to respond to the letter1 regarding our recent publication.2 We appreciate the readers’ engagement and would like to address the points raised.
Similarly to the authors of the letter, we also anticipated that unattended automated office blood pressure measurement (UAOBPM) would prove more effective than office blood pressure measurement (OBPM) in predicting hypertension-mediated organ damage (HMOD) in our study. The results may be surprising; however, a thorough review of the literature on this topic, including the studies cited in our publication, indicates that both methods have shown similar effectiveness in most analyses.3
We agree with the authors of the letter that a cross-sectional design is not optimal for conducting such an analysis. A longer, prospective study would indeed be more valuable. At the same time, it is worth noting that many studies comparing these BP measurement methods also used a cross-sectional methodology. Interestingly, even one of the papers cited by the authors of the letter, comparing 24-hour central aortic and brachial cuff BP, was a cross-sectional study.4 This suggests that while not optimal, this approach is useful for such analyses.
We also agree with the authors that incorporating cofactors into the modeling would enhance the analysis. This would undoubtedly add value to the results.
Regarding the areas under the receiver operating curves (AUCs), their low values result from the low absolute number of HMOD cases in the study population. As noted in the Limitations section, BP control among our study participants was very good, resulting in low incidence of HMOD.2 We are confident that in a larger study group or in a treatment-naïve population, the AUC values would reach greater significance. Nevertheless, our results and AUC values are similar to those reported by Salvetti et al3 in their study performed in the Italian population.
It is acknowledged that our power analysis was based on the assumption of differences in BP values between the methods under investigation. The primary objective of our study was to directly compare these measurement techniques in terms of raw BP values. This approach is standard in studies aiming to establish the fundamental equivalence or noninferiority of measurement techniques. We recognize that predictive value and model accuracy are also crucial metrics in clinical settings. However, the scope of our study was deliberately narrow to ensure methodological rigor in assessing measurement equivalence.
Future research could build upon our findings by incorporating predictive modeling and evaluating the clinical outcomes associated with each measurement method. Such studies would benefit from distinct power analyses tailored to those specific end points. We appreciate the suggestion and agree that further investigations on predictive values and clinical implications are warranted and would enrich understanding of UAOBPM’s utility in diverse health care settings.
Similarly to the authors of the letter, we believe that UAOBPM has significant potential and may have applications in clinical practice, which is why we conducted this study.
However, we cannot agree with the statement that UAOBPM should primarily replace ambulatory blood pressure monitoring (ABPM). The latter method has the unique potential to determine BP during physical activity and home- or work-related psychological stress as well as to measure BP during sleep. While semiautomatic home BP monitoring (HBPM) devices are now available to measure BP during sleep at fixed intervals, OBPM, including UAOBPM, lacks this capability. Thus, ABPM, HBPM, and UAOBPM should be regarded as complementary rather than absolute alternatives.
Moreover, from a practical point of view, more attention should be given to the standardized conditions recommended for a valid office BP measurement. Unfortunately, BP measurement is often suboptimally performed in clinical practice, and this inaccuracy has persisted despite extensive education.5 Adherence to these recommendations would allow for better risk prediction and better BP control, which—and here we fully agree with the authors of the letter—is the main issue in treating hypertension.
While awaiting more data based on hard end points, we believe that these letters to the editor will allow the readers to better and more consciously interpret the results of our study.
Agnieszka Olszanecka, MD, PhD, First Department of Cardiology, Interventional Electrocardiology and Hypertension, Institute of Cardiology, Jagiellonian University Medical College, ul. Jakubowskiego 2, 30-688 Kraków, Poland, phone: +48 12 400 21 50, email: agnieszka.olszanecka@uj.edu.pl
July 23, 2024.
July 26, 2024.
August 8, 2024.
None declared.
Stopa M, Zięba K, Tofilska A, et al. Is unattended automated office blood pressure measurement truly not superior to office blood pressure measurement in predicting home blood pressure monitoring? Authors’ reply. Pol Arch Intern Med. 2024; 134: 16824. doi:10.20452/pamw.16824
- 1.
- Sobieraj P, Siński M, Lewandowski J. Is unattended automated office blood pressure measurement truly not superior to office blood pressure measurement in predicting hypertension-mediated organ damage? Pol Arch Intern Med. 2024; 134: 16823.Crossref
- 2.
- Stopa M, Zięba K, Tofilska A, et al. Unattended automatic blood pressure measurements vs conventional office readings in predicting hypertension-mediated organ damage. Pol Arch Intern Med. 2024; 134: 16699.Crossref
- 3.
- Salvetti M, Paini A, Aggiusti C, et al. Unattended versus attended blood pressure measurement. Hypertension. 2019; 73: 736-742.Crossref
- 4.
- Hu Y, Zhao J, Wang Q, et al. Disparate associations of 24-h central aortic and brachial cuff blood pressure with hypertension-mediated organ damage and cardiovascular risk. Front Cardiovasc Med. 2022; 9: 795509. Crossref
- 5.
- Sharman JE, O’Brien E, Alpert B, et al. Lancet Commission on Hypertension group position statement on the global improvement of accuracy standards for devices that measure blood pressure. J Hypertens. 2020; 38: 21-29.Crossref