Professor Johns Hopkins University School of Medicine Baltimore, United States
Background: Resistant hypertension (RH) is defined as: (1) uncontrolled blood pressure (BP) despite the concurrent use of 3 different classes of antihypertensive agents, including a diuretic; or (2) concomitant use of 4 or more antihypertensive agents regardless of BP control. Patients with RH have an increased risk of morbidity and mortality. Identification of individuals at risk of RH is essential for early prevention.
Objectives: (1) To identify patients with RH using electronic medical records (EMR) from a large, multi-center health system; (2) to compare baseline characteristics of patients with incident hypertension (HTN) who did or did not develop RH; (3) to identify possible factors to be included in clinical prediction model for RH.
Methods: This was a retrospective cohort analysis using EMR data generated between 2013 and 2016. Cohort members were ≥ 18 years old with 2 or more outpatient visits with office BP measurements, without HTN during their baseline period, i.e. their first 12 months of observation. HTN is defined based on diagnosis codes, antihypertensive medication use, or elevated BPs. Cohort members were observed until the occurrence of RH, death, or administrative censoring. The algorithm for identifying RH was: (1) prescription orders for ≥ 3 antihypertensive agents overlapping for more than 21 days; and (2) for patients with 3 concurrent medications, an uncontrolled BP (> 140/90 mm Hg) was measured twice at least one week after the start of the third drug. Baseline characteristics were the last values of the covariates during the baseline period. Standardized mean difference (SMD) > 0.1 indicated important differences in baseline characteristics between groups.
Results: A total of 27998 patients were identified as having incident HTN with mean age 58 years; 44.3% were men and 38.2% were black patients. 613 developed RH with median follow-up of 4.7 months. Important characteristics that differed between those with and without RH were sex, race, BPs, employment status, alcohol use, tobacco use, body mass index (BMI), high-density lipoprotein cholesterol (HDL-C), creatinine, HbA1C, and atherosclerotic cardiovascular disease (ASCVD) risk. The population that did not develop RH was 44.1% men and 38.3% black; the population that developed RH was 51.4% men, and 56.9% black. The mean systolic blood pressure (SBP) and diastolic blood pressure (DBP) at baseline for those not developing RH were 144.0 (SD, 15.1) and 84.7 (SD, 11.5) mm Hg, while they were 152.8 (SD, 23.4) and 88.6 (SD, 16.7) mm Hg for those who did.
Conclusions: Difference in baseline characteristics between groups suggests that there may be an elevated risk of developing RH among individuals who are male, black, alcohol or tobacco user, or disabled with high baseline BPs, HDL-C, creatinine, HbA1C, or ASCVD risk. These variables are strong candidates for a prediction model targeting RH.
.jpg "Jodi B. Segal, MD, MPH (she/her/hers) photo")
