Regresar

Forecasting of cardiovascular disease using time series modeling in patients diagnosed with hypertension

Abstract:

Introduction: Time series models use patterns and trends observed in historical data to estimate and predict future values. Objective: To predict cardiovascular disease using a time series model in patients diagnosed with arterial hypertension in a health center in the province of Tungurahua, Ecuador. Methods: The study population consisted of patients diagnosed with arterial hypertension at a health center in Tungurahua, Ecuador. The endogenous variable was the cases of cardiovascular disease and the exogenous variables related to time and the parameters of the time series model (alpha, gamma and delta) influenced the predictions. A time series model was created and its validation used the Ljung-Box test and other goodness-of-fit tests to assess the quality of the model and the independence of the residuals. Results: The Root Mean Square Error (RMSE) had a value of 2.802 and the Ljung Box Q of 11.541 which showed that the model did not present a significant lack of independence in the residuals. A t-value of 2.998 and a P-value of 0.007 for the alpha parameter indicated that it was statistically significant, which meant that the level component was relevant in the model. Conclusions: By using time series models, it was possible to make accurate predictions of the incidence of cardiovascular disease in patients diagnosed with arterial hypertension, which provided valuable information for the planning of preventive interventions and the clinical management of this population.