Applying multinomial logistic regression to prediction effect of environmental and vehicle type factors on the severity of road accidents (Case study: Tabriz-Ahar Road)

Journal of Injury and Violence Research Pub Date : 2019-08-01 DOI:10.5249/jivr.v11i2.1367

Yousef Kazempour, S. Givehchi, H. Hoveidi

{"title":"Applying multinomial logistic regression to prediction effect of environmental and vehicle type factors on the severity of road accidents (Case study: Tabriz-Ahar Road)","authors":"Yousef Kazempour, S. Givehchi, H. Hoveidi","doi":"10.5249/jivr.v11i2.1367","DOIUrl":null,"url":null,"abstract":"Abstract: Background: Determining the factors that have an important role in increasing the severity of accidents is important to reduce or eliminate mortality and severity of injuries caused by accidents. According to the World Health Organization, the mortality rate in Iran is 32.1 per 100,000 population. Which is much higher than the population of 8.7 per 100,000 people, compared with other high-income countries. Methods: In this study road variables, environmental and climate variables, and vehicle types for four years (1394-1391) were selected accident data of Tabriz-Ahar road for review. The severity of crashes was classified into three levels include fatal, injured and PDO. In this study, multinomial logistic regression method was used to determine the probability and predict the severity of accidents. Results: The results showed that in the case of injured crashes compared with PDO, rainy weather (OR = 0.028), intersections (OR = 0.044) increase the severity of accidents. Also, in the case of fatal accidents compare with PDO, Driving during the night (OR = 0.005), intersections (OR = 2.24) and increase in heavy vehicles on the road (OR = 4.31) increase the severity of fatal accidents. Conclusions: The results of this study show that the multinomial logistic regression model offers a promising approach to predict the severity of accidents in future studies. According to the results, all the significant factors mentioned above should be improved. Therefore, it is necessary to reduce the severity of accidents by taking measures such as development Grade separation intersections at black spots and also increasing the number of lane and separating the slow lane from overtaking lane on this road. Keywords: Crash severity, Logistic Regression, Road Safety, Crash modelling","PeriodicalId":32422,"journal":{"name":"Journal of Injury and Violence Research","volume":" ","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2019-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Injury and Violence Research","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.5249/jivr.v11i2.1367","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 0

Abstract

Abstract: Background: Determining the factors that have an important role in increasing the severity of accidents is important to reduce or eliminate mortality and severity of injuries caused by accidents. According to the World Health Organization, the mortality rate in Iran is 32.1 per 100,000 population. Which is much higher than the population of 8.7 per 100,000 people, compared with other high-income countries. Methods: In this study road variables, environmental and climate variables, and vehicle types for four years (1394-1391) were selected accident data of Tabriz-Ahar road for review. The severity of crashes was classified into three levels include fatal, injured and PDO. In this study, multinomial logistic regression method was used to determine the probability and predict the severity of accidents. Results: The results showed that in the case of injured crashes compared with PDO, rainy weather (OR = 0.028), intersections (OR = 0.044) increase the severity of accidents. Also, in the case of fatal accidents compare with PDO, Driving during the night (OR = 0.005), intersections (OR = 2.24) and increase in heavy vehicles on the road (OR = 4.31) increase the severity of fatal accidents. Conclusions: The results of this study show that the multinomial logistic regression model offers a promising approach to predict the severity of accidents in future studies. According to the results, all the significant factors mentioned above should be improved. Therefore, it is necessary to reduce the severity of accidents by taking measures such as development Grade separation intersections at black spots and also increasing the number of lane and separating the slow lane from overtaking lane on this road. Keywords: Crash severity, Logistic Regression, Road Safety, Crash modelling

查看原文本刊更多论文

环境与车型因素对道路交通事故严重程度的多元logistic回归预测(以大不里士-哈哈公路为例)

摘要：背景：确定在增加事故严重程度方面起重要作用的因素对于降低或消除事故造成的死亡率和伤害严重程度很重要。根据世界卫生组织的数据，伊朗的死亡率为每10万人口32.1人。与其他高收入国家相比，这远远高于每10万人8.7人的人口。方法：本研究选取大不里士-阿哈尔公路四年（1394-1391）的道路变量、环境和气候变量以及车辆类型的事故数据进行回顾。撞车事故的严重程度分为三级，包括致命事故、受伤事故和PDO事故。本研究采用多项逻辑回归方法来确定事故发生的概率和预测事故的严重程度。结果：结果表明，与PDO相比，在受伤车祸的情况下，雨天（OR=0.028）、十字路口（OR=0.044）增加了事故的严重程度。此外，在致命事故的情况下，与PDO相比，夜间驾驶（OR=0.005）、十字路口（OR=2.24）和道路上重型车辆的增加（OR=4.31）会增加致命事故的严重程度。结论：本研究的结果表明，多项式逻辑回归模型在未来的研究中为预测事故的严重程度提供了一种很有前途的方法。根据结果，上述所有重要因素都应该得到改善。因此，有必要采取措施，如在黑点处开发立体交叉口，并增加车道数量，将该道路上的慢车道与超车道分开，以降低事故的严重程度。关键词：碰撞严重程度，逻辑回归，道路安全，碰撞建模

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Journal of Injury and Violence Research

自引率

0.00%

发文量

审稿时长

24 weeks

期刊介绍： The Journal of Injury and Violence Research (JIVR) is a peer-reviewed open-access medical journal covering all aspects of traumatology includes quantitative and qualitative studies in the field of clinical and basic sciences about trauma, burns, drowning, falls, occupational/road/ sport safety, youth violence, child/elder abuse, child/elder injuries, intimate partner abuse/sexual violence, self-harm, suicide, patient safety, safe communities, consumer safety, disaster management, terrorism, surveillance/burden of injury and all other intentional and unintentional injuries.