Epidemiology and SARIMA model of death cases in a tertiary comprehensive hospital in Hangzhou from 2015 to 2022

Abstract

Background The statistical analysis of death cases has important clinical research value. Our study aims to describe the epidemiology of death cases in a tertiary comprehensive hospital in Hangzhou from 2015 to 2022 and predict the number of future deaths, providing a reference basis for hospitals to formulate relevant strategies and measures. Methods Death data of inpatients and non-inpatients in the hospital from 2015 to 2022 were obtained through the CDC-DSP system. The data of death cases were described and analyzed by retrospective study, and the single factor difference was statistically inferred by χ2 tests. P < 0.05 was considered statistically significant. According to International Classification of Diseases 10th revision (ICD-10), the main causes of death of patients were obtained. SARIMA model was established by R 4.3.0 (forecast, aTSA, tseries) software for time series analysis. Results A total of 1938 death cases from 2015 to 2022, including 287 inpatients and 1651 non- inpatients. Among them, the highest was in 2022 (262, 13.52%), and the lowest was in 2019 (223, 11.51%). The gender ratio is 2.22:1, and there are differences (P < 0.05) between different genders in the age, marital status, educational level, and distribution of place of residence. The main cause of death were circulatory system diseases (32.66%), injury-poisoning (28.22%), tumors (14.76%), and respiratory system diseases (10.47%), with a cumulative proportion of 86.12%. Furthermore, the SARIMA (2,1,1)(1,1,1)12 model was ultimately determined to predict the number of deaths among patients, AIC = 380.23, BIC = 392.79, AICc = 381.81, MAPE = 14.99%. Conclusions The hospital should focus on improving the pre-hospital emergency treatment and the ability of multi-disciplinary cooperation in the hospital to reduce the number of deaths of hospital patients. the SARIMA model is suitable for predicting the number of death cases and provide reference value for the rational allocation of medical resources.