This study aimed to compare attended home blood pressure (BP) measurements (HBPM) with ambulatory BP monitor (ABPM) readings and examine if level of agreement between measurement modalities differs overall and by subgroup. This was a secondary analysis of data from a 2-year, multicenter observational study of children 11-19years (mean 15, SD = 2.7) with chronic kidney disease. Participants had 3 standardized resting oscillometric home BPs taken by staff followed by 24-h ABPM within 2weeks of home BP. BP indices (measured BP/95%ile BP) were calculated for mean triplicate attended HBPM and mean ABPM measurements. Paired HBPM and ABPM measurements taken during any of 5 study visits were compared using linear regression with robust standard errors. Generalized estimating equation-based logistic regression determined sensitivity, specificity, negative, and positive predictive values with ABPM as the gold standard. Analyses were conducted for the group overall and by subgroup. A total of 103 participants contributed 251 paired measurements. Indexed systolic BP did not differ between HBPM and daytime APBM (mean difference - 0.002; 95% CI: - 0.006, 0.003); the difference in indexed diastolic BP was minimal (mean difference - 0.033; 95% CI: - 0.040, - 0.025). Overall agreement between HBPM and 24-h ABPM in identifying abnormal BP was high (81.8%). HBPM had higher sensitivity (87.5%) than specificity (77.4%) and greater negative (89.8%) than positive (73.3%) predictive value, and findings were consistent in subgroups. Attended HBPM may be reasonable for monitoring BP when ABPM is unavailable. The greater accessibility and feasibility of attended HBPM may potentially help improve BP control among at-risk youth. A higher resolution version of the Graphical abstract is available as Supplementary information.
Read full abstract