TY - JOUR PY - 2016// TI - Automatic identification of firefighters with post-traumatic stress disorder based on demographic characteristics and self-reported alcohol consumption JO - Journal of Uncertain Systems A1 - Magoc, D. A1 - Magoc, T. A1 - Tomaka, J. A1 - Morales-Monks, S. SP - 198 EP - 203 VL - 10 IS - 3 N2 - Post-Traumatic Stress Disorder (PTSD) is an anxiety disorder that involves a specific set of symptoms that develop after experiencing, witnessing, or confronting an event comprising actual or threatened serious injury or death, threat or death to the physical integrity of self or others, or actual or threatened sexual violence. Firefighters participate in a wide array of stressful and traumatic events including threat of injury to self and others, death of and injuries to other people, exposure to gruesome accidents, body handling, multiple casualties, and suicides. Repeated exposure to such traumatic events places firefighters at risk for developing post-traumatic-stress (PTS) symptoms and related problems such as alcohol misuse, especially if they use alcohol as a means of coping with stress. 740 municipal firefighters completed assessments of PTS symptoms, alcohol consumption, alcohol problems, drinking motives, and coping with stress as part of a larger study. We used demographic data, data on PTS symptoms and data on alcohol related outcomes to build an automated predictor of the presence of PTSD in firefighters based on all attributes provided except those related to PTSD questionnaire. The results of the PTSD questionnaire were used to train and test the machine learning algorithms, including Neural Network (NN), Naïve Bayes Method (NB), and Decision Tree (DT), to build and validate the automated predictor for PTSD in municipal firefighters. The results of this study indicated that the automatic predictors can successfully predict PTSD with the accuracy of 88.65% using NB and 91.76% using both NN and DT. Collecting additional data points that contain more at risk individuals will improve the machine learning algorithms' prediction of at-risk individuals. © 2016 World Academic Press, UK. All rights reserved.

Language: en

LA - en SN - 1752-8909 UR - http://dx.doi.org/ ID - ref1 ER -