Forecasting cardiovascular disease in ecuadorian patients with arterial hypertension using a time series model

Authors

  • Lina Espinoza Neri
  • Anahí Bonilla Rodríguez
  • Jenny Maribel Moya Arizaga

Keywords:

time series models, Root Mean Square Error, cardiovascular disease, arterial hypertension, stationary R-squared

Abstract

Introduction: Arterial hypertension is positioned as a fundamental risk factor in cardiovascular conditions.

Objective: To predict cardiovascular disease in patients diagnosed with hypertension in a hospital in Ecuador, using a time series model.

Methods: Individuals diagnosed with arterial hypertension in a hospital in Ecuador, were studied. The variable to be predicted was cases of cardiovascular disease and the variables that influenced the predictions were related to time and the parameters of the time series model. A time series model was created, whose validation used the exponential smoothing method and Winters' additive method as a criterion to calculate the coefficients that allowed the predictive model to be detailed: alpha (level), gamma (trend) and delta (season).

Results: The stationary R-squared and R-squared values indicated a good model fit, explaining about 70 % and 90 % of the variability in the data, respectively. The model contained no external predictors, only the original time series, presented a stationary R-squared of 0.699 and a Root Mean Square Error (RMSE) of 2.226. The alpha parameter was significant with a value of 0.803, implying that the data exhibited a stationary level. Gamma and delta were not significant, suggesting absence of trend and stationarity in the series.

Conclusions: The model parameters were consistent with a stationary process appropriate for modeling the occurrence of cardiovascular disease in the study population.

Downloads

Download data is not yet available.

Published

2024-04-08

How to Cite

1.
Espinoza Neri L, Bonilla Rodríguez A, Moya Arizaga JM. Forecasting cardiovascular disease in ecuadorian patients with arterial hypertension using a time series model. Rev Cubana Inv Bioméd [Internet]. 2024 Apr. 8 [cited 2025 Jul. 13];43. Available from: https://revibiomedica.sld.cu/index.php/ibi/article/view/3282