Prediction Model to Estimate the Probability of Ischemic Cerebral Vascular Event in Essential Hypertensive Patients

Abstract

Introduction: It is common to observe the development of an ischemic cerebrovascular event in the evolution of patients with arterial hypertension, which is rarely diagnosed, only until compatible clinical signs appear.

Objective: To validate a risk prediction model for ischemic cerebral vascular events in essential hypertensive patients.

Methods: A cohort study was conducted to develop a prediction model using multiple linear regression algorithms. One thousand two hundred consecutive patients treated in the outpatient medicine clinic at Dr. Luis Díaz Soto Hospital were taken into account from January to December 2012. The sample was made up of 250 essential hypertensive patients who met the intentionality criteria (risk of incidence of ischemic cerebrovascular disease), and according to the exposure of the risk factors, one hundred twenty-five cases (cohort exposed) and 125 cases (non-exposed cohort) were taken. Follow-up was 10 years from inclusion.

Results: The variables that were included in the model after the multiple logistic regression analysis were severe arterial hypertension, diabetes mellitus, dyslipidemia and smoking. The value of the coefficient of determination R2 explained 82.9% of the dependent variable. The adjusted R2 value (0.819) expressed good fit between the real data and the modeled prediction data. The model revealed that 41% of the variance of the dependent variable was significant.

Conclusions: The model demonstrated good fit and internal validity, in addition, it was good predictor of ischemic cerebrovascular disease in hypertensive patients.