Measurement of blood pressure for the diagnosis and management of hypertension in different ethnic groups: one size fits all

BACKGROUND: Hypertension is a major risk factor for cardiovascular disease and prevalence varies by ethnic group. The diagnosis and management of blood pressure are informed by guidelines largely based on data from white populations. This study addressed whether accuracy of blood pressure measurement in terms of diagnosis of hypertension varies by ethnicity by comparing two measurement modalities (clinic blood pressure and home monitoring) with a reference standard of ambulatory BP monitoring in three ethnic groups. METHODS: Cross-sectional population study (June 2010 - December 2012) with patients (40–75 years) of white British, South Asian and African Caribbean background with and without a previous diagnosis of hypertension recruited from 28 primary care practices. The study compared the test performance of clinic BP (using various protocols) and home-monitoring (1 week) with a reference standard of mean daytime ambulatory measurements using a threshold of 140/90 mmHg for clinic and 135/85 mmHg for out of office measurement. RESULTS: A total of 551 participants had complete data of whom 246 were white British, 147 South Asian and 158 African Caribbean. No consistent difference in accuracy of methods of blood pressure measurement was observed between ethnic groups with or without a prior diagnosis of hypertension: for people without hypertension, clinic measurement using three different methodologies had high specificity (75–97%) but variable sensitivity (33–65%) whereas home monitoring had sensitivity of 68–88% and specificity of 64–80%. For people with hypertension, detection of a raised blood pressure using clinic measurements had sensitivities of 34–69% with specificity of 73–92% and home monitoring had sensitivity (81–88%) and specificity (55–65%). CONCLUSIONS: For people without hypertension, ABPM remains the choice for diagnosing hypertension compared to the other modes of BP measurement regardless of ethnicity. Differences in accuracy of home monitoring and clinic monitoring (higher sensitivity of the former; higher specificity of the latter) were also not affected by ethnicity. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1186/s12872-017-0491-8) contains supplementary material, which is available to authorized users.