The Application of Non-Parametric Count Models for the Modeling of Female’s Accident Rates in Hamadan Province from 2009 to 2016

Abstract

Background: Accidents were just one of the general health problems. According to WHO forecasts (2013), deaths from road accidents will become the fifth-highest cause of death in the world by 2030. Therefore, we have attempted the application of non-parametric count models for modeling female’s accident rates. Methods: All accidents in Hamadan Province, western Iran are referred to as one of the emergency centers located in the hospitals. Data regarding the accidents were obtained from 21 emergency centers across Hamadan for the period 2009-2016. To assess the trend and pattern of the accidents, the Generalized Additive Model for the accident rate has been utilized. Results: The Mean±SD age of the females in study was 31.23±12.88 yr old. For each of the three kinds of road accidents (car accidents, motor accidents and pedestrian crashes), the accident rates in the “residential urban” areas are lesser than in the “non-residential” area (P=<0.001) and in "public and sports grounds" and "great roads, avenues and streets" are more than in "others". For the three kinds of accidents, the functional effect in the monthly trend of the accidents was signification (P=<0.001). Conclusion: The rates for all three kinds of accidents decreased. The increase in accident rates from the beginning of 2014 to Mar 2016 maybe due to the generalization of insurances in Iran and the increase in the number of accident victims being referred to the hospitals, which was the same with the results of other studies.