Prediction of Suicide Attempts Using the BoxJenkins Model

Objective: Suicide is a preventable social harm in the field of health. The present study aimed to design a prediction model for suicide incidence based on BoxJenkins model in Ilam province. Method: Using a retrospective method all completed and attempted suicide data were collected during 19932013. Then, using the autoregressive integrated moving average (ARIMA) model, the time series analysis of the BoxJenkins model was conducted to predict suicide status in the coming years (20142015). Results: According to the actual results obtained from the suicide data in 2014, a total of 1078 suicides were recorded and compared to the predicted results, according to the fitted model of the time series, which showed the selected model predicted 931 suicide cases, showing 86% accuracy of prediction. The series #x2019; prediction indicated 931 suicides in 2014 with a negative growth rate of 25.3% compared to 2013 and 969 suicide cases in 2015 with a positive growth rate of 3.93% compared to 2014. Conclusion: The results of this study showed the designed model provides a high diagnostic value to predict suicide rates. These types of models can help to predict suicide in future and plan to control and prevent suicide attempts.