The Statistical Investigation of the Proportional Hazard Assumption in Mine Accidents

In recent years, the occupational accidents in our country again bring to the agenda the importance of occupational health and safety. In particular, the mining sector is one of the most dangerous business lines with many occupational accidents. In this study, the underground coal mine accident data was taken from Turkey Mining Engineers Chamber between the years 2009-2011. Median survival times were obtained by Kaplan-Meier method according to various factors for the analysis of the time until death. In the Cox regression model, the number of workers, the ventilation, the gas measurement condition and the amount of coal dust were found to be statistically significant. However the stratified Cox regression model was used with the control of proportional hazard assumption. It was determined that fatal accidents occurred in a longer period if appropriate ventilation and gas measurement were made for the mines where the number of workers was small