Is home blood-pressure monitoring in hypertensive disorders of pregnancy consistent with clinic recordings?

OBJECTIVE: To assess the agreement between home blood-pressure monitoring (HBPM) and blood-pressure measurements in a clinic setting, in a cohort of pregnant women with hypertensive disorders of pregnancy (HDP). METHODS: This was a cohort study of 147 pregnant women with HDP conducted at St George's Hospital, University of London, London, UK, between 2016 and 2017. Inclusion criteria were chronic hypertension, gestational hypertension or high risk of developing pre-eclampsia, no significant proteinuria and no hematological or biochemical abnormalities. Each included patient was prescribed a personalized schedule of hospital visits and blood-pressure measurements, according to their individual risk as per UK National Institute for Health and Care Excellence guidelines. The blood-pressure measurement at the clinic and the HBPM reading obtained closest to that hospital visit were paired for analysis. Only one pair of measurements was used per patient. Differences between home and clinic blood-pressure measurements were tested using the Wilcoxon signed rank test or paired t-test, and were also assessed visually using Bland-Altman plots. Comparison of the binary outcomes was performed using McNemar's chi-square test. Subgroup analysis was performed in the following gestational-age windows: < 14 weeks, 15-22 weeks, 23-32 weeks and 33-42 weeks' gestation. RESULTS: A total of 294 blood-pressure measurements from 147 women were included in the analysis. Median systolic HBPM measurements were significantly lower than clinic measurements (132.0 (interquartile range (IQR), 123.0-140.0) mmHg vs 138.0 (IQR, 132.0-146.5) mmHg; P < 0.001). When stratified according to gestational age, systolic blood-pressure measurements obtained at home were significantly lower than those at clinic in all gestational-age periods except 23-32 weeks' gestation (P = 0.057). Median diastolic blood-pressure measurements at home were also significantly lower than those at clinic (85.0 (IQR, 77.0-90.0) mmHg vs 89.0 (IQR, 82.0-94.0) mmHg; P < 0.001). When stratified according to gestational age, diastolic HBPM measurements were significantly lower in the periods 5-14 weeks (P < 0.001), 15-22 weeks (P = 0.008) and 33-42 weeks (P < 0.001), compared with clinic measurements. The incidence of clinically significant systolic and diastolic hypertension based on clinic blood-pressure measurements was four to five times higher than that based on HBPM measurements (P < 0.001 and P = 0.005, respectively). CONCLUSIONS: Our study shows that, in women with HDP, blood pressure measured at home is lower than that measured in a clinic setting. This is consistent with observations in non-pregnant adults, in whom home and ambulatory monitoring of hypertensive patients is recommended. As such, HBPM has the potential to reduce the number of false-positive diagnoses of severe hypertension and unnecessary medical interventions in women with HDP. This must be carefully weighed against the risk of missing true-positive diagnoses. Prospective studies investigating the use of HBPM in pregnant women are urgently needed to determine the relevant blood-pressure thresholds for HBPM, and interval and frequency of monitoring. Copyright © 2018 ISUOG. Published by John Wiley & Sons Ltd