DIAGNOSTIC PERFORMANCE OF COMMUNITY HEALTH WORKER-LED ATTENDED AND UNATTENDED OFFICE BLOOD PRESSURE MEASUREMENTS IN RURAL SUB-SAHARAN AFRICA

Abstract

Objective: The burden of hypertension in rural sub-Saharan Africa (SSA) has been on the rise. In rural SSA, screening, diagnosis and treatment monitoring for hypertension is typically through attended office blood pressure (OBP) conducted by a trained health worker. A community-centered approach where community health workers (CHWs) conduct out-of-office blood pressure (BP) measurements at the homes of community members may increase the proportion of patients diagnosed early and improve monitoring during treatment. However, the diagnostic performance of these BP measurements when conducted by CHWs remains unclear. Design and method: We are conducting a cross-sectional study to determine the diagnostic performance of OBP performed by CHWs in identifying individuals with hypertension in rural Kenya and The Gambia. A random age-stratified random sample of 1250 participants aged 30 years and above from the Kilifi Health and Demographic Surveillance System and the Kiang West Demographic Surveillance System are being enrolled into the study. Participants undergo randomly assigned attended and unattended OBP measurement and 24-hour ambulatory BP monitoring (reference standard) conducted by trained CHWs at their homes. Results: Recruitment is ongoing. We have enrolled and completed BP measurements for 166 participants, all from the Kenya study site (13% of the total sample). Of these, 114 (69%) are female with an average age of 57 years (standard deviation 15 years). Based on ambulatory BP, 71 (43%) of included participants are hypertensive. The preliminary sensitivity of attended and unattended OBP is 0.50 (0.38, 0.62) and 0.47 (0.35, 0.59) respectively with both measurement methods having a high specificity of 0.93 (0.85, 0.97). The diagnostic odds ratio (DOR) for attended OBP (DOR 13.6; likelihood ratio (LR) positive 7.33 (3.26, 16.47)) is higher than that of unattended OBP (DOR 12.1; LR positive 6.89 (3.05, 15.54)). Diagnostic performance does not appear to differ by age or sex. Further analyses will be conducted when recruitment is complete. Conclusions: Attended OBP appears to have better diagnostic performance than unattended OBP when conducted at participants’ homes by CHWs, but the sensitivity of both methods for detecting hypertension is low.