Evaluation of occupational injuries with lost days among opencast coal mine workers through logistic regression models Seyhan Onder ⇑ Department of Mining Engineering, Eskisehir Osmangazi University, Eskisehir, Turkey article info Article history: Received 3 October 2012 Received in revised form 10 April 2013 Accepted 2 May 2013 Available online 28 May 2013 Keywords: Accident analysis Logistic regression models Non-fatal occupational injuries Opencast coal mining abstract Despite precautions, mining remains the most hazardous occupation, and coal mining is one of the most dangerous industries for non-fatal occupational accidents. Accidents are complicated events with many factors that affect their formation, and statistical evaluation of accident records can produce valuable information that may prevent such accidents. In this study, a logistic regression analysis method was applied to non-fatal occupational injuries from 1996 to 2009 in an opencast coal mine for Western Lignite Corporation (WLC) of Turkish Coal Enterprises (TKI). The accident records were categorized as occupa- tion, area, reason, age, part of body and lost days, and the SPSS package program was used for statistical analyses. Logistic regression analyses were used to predict the probability of accidents that resulted in greater or less than 3 lost workdays. It is found that the job group with the highest probability of expo- sure to accidents with greater than 3 lost workdays for non-fatal injuries was the maintenance personnel and workers. The employees were primarily exposed to accidents caused by a mining machine, and the lower and upper extremities have the highest probability of exposure to such risks. Finally, an equation for calculating the probability of exposure to accidents with greater or less than 3 lost workdays was derived. Then, the equation was used to determine the important accident risk factors. Ó 2013 Elsevier Ltd. All rights reserved. 1. Introduction Compared with other industries, the mining industry and re- lated energy resource industries are associated with high rates of occupational injuries and fatalities, and mining is one of the most hazardous work environments in many countries around the world (Sari et al., 2009; Groves et al., 2007; Bajpayee et al., 2004; Donog- hue, 2004). Mining is a hazardous profession and considered at war with the unpredictable forces of nature. As a result, the mining industry continues to be associated with a high level of accidents, injuries, and illness (Maiti et al., 2004). Despite the record of pro- gress in reducing mining fatalities and injuries, both the number and severity of mining accidents remain unacceptable (Kecojevic et al., 2007), and the incidence rates are high compared with other industries (Komljenovic et al., 2008). To identify the potential problem areas, it is necessary to investigate the causes of accidents and control exposure of such risks through quantitative analysis of accident data (Maiti et al., 2001). Human factors approaches to system safety have been used to provide greater insights into the causes of accidents and can be applied to the mining context (Lenné et al., 2012). These models of human error in organizational systems take a systems approach (Reason, 2000). Such models have supported the development of several methods of accident investigation and analysis that use error and latent condition classification schemes to provide an analysis of the types of failure involved in accidents. One of the more widely used approaches is the Human Factors Analysis and Classification System (HFACS) (Shappell and Wiegmann, 2000). HFACS describes four levels of failure: (1) Unsafe Acts, (2) Precon- ditions for Unsafe Acts, (3) Unsafe Supervision, and (4) Organiza- tional Influences (Shappell and Wiegmann, 2004). Reason proposed the ‘‘Swiss Cheese’’ model of human error where four levels of failure are described. Each level influences the next level as seen in Fig. 1 (Shappell and Wiegmann, 2000). Lost workdays in mining industries are valuable indicators for a number of aspects in job safety programs (Coleman and Kerkering, 2007). According to the European Statistics on Accidents at Work (ESAW), the definition of a non-fatal accident at work is ‘‘The def- inition of what constitutes a notifiable work accident ranges from any work accident, whether it results in an interruption of work or not, to a minimum absence of more than three days’’. Accidents with greater than 3 days’ absence from work are reported more than accidents with less than 3 days’ absence from work. Only accidents with greater than 3 days’ absence are considered in the ESAW methodology (EUROSTAT, 2001). In this study, based on the ESAW accident definition, a logistic regression method was used for categorical data analysis to predict 0925-7535/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ssci.2013.05.002 ⇑ Tel.: +90 2222393750; fax: +90 2222393613. E-mail address: sonder@ogu.edu.tr Safety Science 59 (2013) 86–92 Contents lists available at SciVerse ScienceDirect Safety Science journal homepage: www.elsevier.com/locate/ssci