Predictors of hypertension detection in English general practices: a cross sectional study.

Background Worldwide, high systolic blood pressure is the leading risk factor for deaths and disability-adjusted life-years but has been

historically under-detected. This study aimed to quantify differences between estimated and practice-detected prevalences of hypertension

across English general practices, and to determine how variations in detected prevalence could be explained by population-level and

service-level factors.

Methods Descriptive statistics, pair wise correlations between the independent variables and a multivariable regression analysis were

undertaken. In the regression model, the outcome was detected hypertension prevalence, adjusted for estimated prevalence, person-related

and disease-related determinants of illness and characteristics of general practices.

Results Detected prevalence was substantially lower than estimated prevalence (mean difference 16.23%; standard deviation 2.88%). Higher

detected prevalence was associated with increased deprivation, increased non-white ethnicity and urban location. Lower detected prevalence

was associated with larger list sizes, more general practitioners and being located in the South outside London. The final multivariable model’s

adjusted R2 value was 0.75.

Conclusions Substantial under-detection of hypertension is widespread across England. Independent of estimated prevalence, factors

associated with greater morbidity and population density predicted higher detected prevalence. Identifying patients with undetected

hypertension and coordinating care for these patients will require further resources and logistical support in community settings