By Michael H. Crawford, MD, Editor

SYNOPSIS: A small, community-based study to detect hypertension revealed one week of twice-daily home blood pressure (BP) measurements are more reliable and more accurately predict increased left ventricular mass than clinic or 24-hour ambulatory BP monitoring.

SOURCE: Schwartz JE, Muntner P, Kronish IM, et al. Reliability of office, home and ambulatory blood pressure measurements and correlation with left ventricular mass. J Am Coll Cardiol 2020;76:2911-2922.

Controversy exists about which method of diagnosing and monitoring high blood pressure (BP) is best for guiding therapy to prevent end-organ damage. Schwartz et al investigated the reliability and predictive accuracy of office BP (OBP) measurements vs. home BP (HBP) vs. ambulatory BP (ABP) for predicting left ventricular mass index (LVMi) measured by echocardiography.

The study population included participants from the Detection of Hypertension (IDH) study, a community-based sample of 400 subjects without known cardiovascular disease recruited between 2011 and 2013 in New York City.1 Exclusion criteria included BP > 160/105 mmHg, evidence of secondary hypertension, pregnancy, or on medications that would affect BP. The study consisted of five visits over one month with OBP, 24-hour ABP twice, HBP over three weeks, and an echo on the last visit. OBPs were attended and performed three times with three different devices (mercury, clinic oscillometric, and home oscillometric). Complete ABP studies were obtained in 91% of subjects. HBP was conducted twice one minute apart in the morning and evening, and 96% recorded at least 12 out of 28 expected readings in a week for all three weeks. All measurements were corrected for regression dilution bias, and adjustments were made for clinical features of the subjects for estimating associations with LVMi. Mean age was 41 years, 60% were women, 26% were Black, and 64% were Hispanic.

The mean awake systolic (SBP) by ABP was 8-9 mmHg higher than the mean OBP and HBP SBPs, which were almost identical. The reliability of the first visit OBP with the subsequent visits was 0.89 and 0.85 for SBP and DBP, respectively. For the first week of HBP vs. the second week, it was 0.94/0.92. For the first 24-hour ABP vs. the second, it was 0.85/0.84. The corrected correlations between the three measurement techniques ranged from 0.74 to 0.89. The multivariate adjusted correlation between SBP by HBP and LVMi of 0.50 was significantly higher than those of ABP and OBP (0.43 for both; P < 0.001). Correlations between DBP and LVMi all were weaker, but the HBP value of 0.33 was significantly higher than the awake ABP value of 0.26 (P < 0.05). The three OBP equipment types produced similar adjusted correlations between SBP and LVMi (0.32 mercury, 0.39 office oscillometric, and 0.43 home oscillometric). The authors concluded HBP measurements were more reliable and more strongly associated with LVMi than OBP or ABP measurements and suggested that one week of HBP monitoring may be the best way to diagnose hypertension.


These results are somewhat surprising since prior studies have shown ABP is superior to OBP for predicting cardiovascular events and all-cause mortality. However, those studies used one OBP measurement as the comparator. Also, HBP has not been well studied, especially in comparison to other methods. In addition, ABP does not account for resting awake BP and usually is performed for only one 24-hour period. In the Schwartz et al study, HBP involved twice-daily measurements, averaged for a week. This would tend to average out day-to-day variation and arguably would be the best estimate of basal resting BP in the subject’s natural environment. OBP is well known to be problematic for several reasons. It occurs in an artificial environment. White coat hypertension is an issue. The SPRINT2 authors worked around some of these office issues by designating a mandatory five-minute unattended (in most cases) delay before an automatic oscillometric machine measured BP. These BP measures were lower than random home measures. However, the realities of clinic practice make such measurements impractical, and this technique has not gained widespread acceptance. The availability of ABP is limited, it is poorly reimbursed, and is used rarely. Thus, the one-week HBP approach is attractive, but there are barriers to widespread use of HBP. The daily schedule will not work for everyone, although in the Schwartz et al study, 96% of subjects recorded at least 12 of the expected 28 BP readings in a week. The cost of the device is not trivial, and there is no provider reimbursement for the data review. Also, there is no infrastructure to handle the data and provide the five minutes of training required to ensure accurate data.

There were weaknesses to this study. The population studied was small, young, and had few comorbidities. Whether HBP would work as well in an older population with more comorbidities and on medications is unknown. Also, although LVMi is a valid target organ abnormality to validate the technique’s utility, Schwartz et al did not pursue long-term clinical outcomes. Current guidelines recommend two or three OBP measurements to diagnose hypertension. Surely, HBP would be more cost-effective than two or three office visits. At this point, I believe HBP is a good technique for diagnosing hypertension, but more data are needed before using it exclusively for the management of patients with known hypertension and other comorbidities.


  1. Anstey DE, Muntner P, Bello NA, et al. Diagnosing masked hypertension using ambulatory blood pressure monitoring, home blood pressure monitoring, or both? Hypertension 2018;72:1200-1207.
  2. SPRINT Research Group; Wright JT Jr, Williamson JD, Whelton PK, et al. A randomized trial of intensive versus standard blood-pressure control. N Engl J Med 2015;373:2103-2116.