A comparison of blood pressure measurements derived from primary care practice and research settings [abstract]
- View All
Background/Aims: Guidelines for blood pressure (BP) treatment thresholds and goals have been derived from measurements obtained in research clinics (RC), but few studies have compared BP measurements obtained in RC and routine primary care practice (PCP) settings, where decisions about BP treatment occur.
Aims: To compare BP obtained from RC and PCP settings using data from Hyperlink, a randomized controlled trial of home BP telemonitoring.
Methods: Eligible subjects included 1,138 subjects whose 2 most recent PCP BP in the last 12 months were >140/90, responded with interest to a mailed invitation, and attended a research clinic (RC) screening visit for the Hyperlink study. Of these subjects, 255 had an additional PCP BP measurement within 30 days before the RC visit and were included in this analysis. The systolic blood pressure (SBP) and diastolic blood pressure (DBP) used for RC was the mean of 3 readings obtained using an automated blood pressure device, and for PCP was the single measurement (with sphygmomanometer as usual procedure) recorded in the electronic medical record closest to the RC visit.
Results: The SBP mean (SD) in mm Hg for PCP was 136.9 (15.8) and for RC was 133.5 (17.6), a mean within-person difference of 3.3, p=.005. DBP mean (SD) for PCP was 77.9 (11.1) and for RC was 75.9 (12.2), difference of -2.0, p=.005. Correlation coefficients between PCP and RC readings were 0.38 for SBP and 0.55 for DBP. Of 103 subjects with PCP SBP >=140, 52 (50%) had RC SBP <140, and the overestimation was greater at higher levels of PCP SBP (PCP SBP >150, +17 mm Hg; PCP SBP 140-149, +8 mm Hg; PCP SBP 130-139, +0.3 mm Hg; PCP SBP <130, -6 mm Hg).
Conclusions: BP obtained in the RC setting was significantly lower than BP obtained in the PCP setting; only half of patients with uncontrolled PCP BP were confirmed in the RC. The differences could result in poorer outcomes on PCP BP performance measures and over-treatment of BP. The results support the use of better methods to more accurately assess and manage BP in the PCP setting.