Predicting New Cases of Hypertension in Primary Care With a Machine Learning Tool

Objective: Many individuals with hypertension remain undiagnosed and therefore untreated. Consequently, there is a pressing need to pinpoint patients at risk of developing hypertension to avoid cardiometabolic morbidity and mortality. We aimed to develop a predictive model for hypertension using diagnostic codes from prevailing electronic medical records in primary health care. Design and method: This sex- and age-matched case-control (1:5) study included patients aged 30–65 years living in Stockholm County, Sweden with a newly recorded diagnosis of hypertension during 2010–19 (cases) and individuals without a recorded hypertension diagnosis during 2010–19 (controls), in total 507,618 individuals. Patients with diagnoses of cardiovascular diseases or diabetes were excluded. A stochastic gradient boosting machine learning model was constructed using the 1,309 most registered ICD-10 codes from primary health care during the three years preceding the diagnosis of hypertension. Results: The model showed an area under the curve (95% confidence interval) of 0.748 (0.742–0.753) for females and 0.745 (0.740-0.751) for males for predicting newly diagnosed hypertension (Figures 1 and 2). The sensitivity was 63% and 68%, and the specificity 76% and 73%, for females and males, respectively (Table 1). The 25 diagnoses that contributed the most to the model for females and males all exhibited a normalized relative influence > 1%, as detailed in Tables 2a and 2b. The diagnoses contributing most to the model, all with an odds ratio of marginal effects > 1 for both sexes were dyslipidemia and obesity, along with people encountering health services in other circumstances, painful conditions, and anxiety. Conclusions: This machine learning model, developed with artificial intelligence based on earlier diagnoses within primary health care may contribute to identification of patients at risk of developing hypertension.