info:eu-repo/semantics/bachelorThesis
Análisis Estadístico mediante Modelos de Supervivencia en pacientes con COVID-19 en el Hospital Básico de Sucúa 2020-2021
Fecha
2022-03-10Registro en:
Villalta Brito, Álvaro José. (2022). Análisis Estadístico mediante Modelos de Supervivencia en pacientes con COVID-19 en el Hospital Básico de Sucúa 2020-2021. Escuela Superior Politécnica de Chimborazo. Riobamba.
Autor
Villalta Brito, Álvaro José
Resumen
The objective of the research was to perform a statistical analysis using survival models for patients with COVID-19 in the Basic Hospital of Sucúa, province of Morona Santiago, 2020-2021. The information matrix was provided by the Health District 14D03, recording the dates of admission and dates of final diagnosis of the patients. The methodology used in the study was Kaplan-Meier Curve Comparison and Cox Regression. An exploratory analysis of the relevant variables was carried out and the results showed that the majority of patients suffering from this disease are women 65%, the type of comorbidity with the majority is overweight 32%. Out of the 251 patients, 44 had some comorbidity, the most frequent being hypertension 12%. The technique used determined that there were 161 patients who died from COVID-19 for Sex, Ethnicity, Type of Obesity and Comorbidity with different probabilities of death. After the survival analysis using the stratifying variable Sex and Comorbidity, men have a probability of death of 92.10% and women 72.2% with a median follow-up time of 3 and 10 days respectively, for patients with comorbidity we obtained a probability of death of 93.20% and those who did not present any comorbidity, 76.33% with a median time of 27 and 4 days, respectively. The global curve comparison model indicated that there were significant differences between sex and comorbidity with a confidence level of 95%. The Cox regression model determined that the risk factors leading to death in patients with COVID-19 are Sex and Comorbidity. It is recommended the creation of other survival models that help in decision making to reduce the number of deaths.