Background: Cardiovascular risk scales in hypertensive populations have limitations for clinical practice.
Aims: To develop and internally validate a predictive model to estimate one-year cardiovascular risk for hypertensive patients admitted to hospital.
Methods: Cohort study of 303 hypertensive patients admitted through the Emergency Department in a Spanish region in 2015-2017. The main variable was the onset of cardiovascular disease during follow-up. The secondary variables were: gender, age, educational level, family history of cardiovascular disease, Charlson score and its individual conditions, living alone, quality of life, smoking, blood pressure, physical activity and adherence to the Mediterranean diet. A Cox regression model was constructed to predict cardiovascular disease one year after admission. This was then adapted to a points system, externally validated by bootstrapping (discrimination and calibration) and implemented in a mobile application for Android.
Results: A total of 93 patients developed cardiovascular disease (30.7%) over a mean period of 1.68 years. The predictors in the points system were: gender, age, myocardial infarction, heart failure, peripheral arterial disease and daily activity (quality of life). The internal validation by bootstrapping was satisfactory.
Conclusion: A novel points system was developed to predict short-term cardiovascular disease in hypertensive patients after hospital admission. External validation studies are needed to corroborate the results obtained.
Keywords: Models; cardiovascular; hypertension; mobile applications.