Modeling Time Series of Mortality Cases Due to Traffic and Non-Traffic Accidents in Northwestern Iran from 2013 to...

Globally, accidents rank as the third leading cause of mortality, following cardiovascular diseases and cancer. Accidents also account for 12% of the global disease burden. This study aims to examine the trends in mortality cases resulting from both traffic and non-traffic accidents.

This descriptive-analytical study utilized recorded data from the Health Deputy of West Azerbaijan Province, Iran, spanning from 2013 to 2022. A complete census was conducted across the entire province during the study period, yielding a total of 7716 fatalities due to traffic accidents and 7316 deaths due to non-traffic accidents. The Box-Jenkins AutoRegressive Integrated Moving Average (ARMA) models were employed for time series analysis. Descriptive analysis was performed using SPSS software, while modeling conducted using R Studio and SAS.

A total of 7716 deaths were recorded due to traffic accidents, and 7316 deaths were attributed to non-traffic accidents. Among traffic-related fatalities, 19.67% were female and 80.33% were male. In contrast, for non-traffic accidents, 29.65% of the victims were female and 70.35% were male. The ARIMA model was employed for modeling based on autocorrelation and partial autocorrelation plots, with the ARIMA (2, 0, 0) (2, 0, 0) model identified as the best fit for traffic accidents and the ARIMA (1, 0, 1) (2, 0, 0) model for non-traffic accidents in the context of monthly mortality time series.

The predictions for both traffic and non-traffic accidents indicated a decreasing trend in mortality. Furthermore, mortality trends and distributions for both categories of accidents exhibited similarities.