Florida Department of Transportation, 605 Suwanee St, MS 36, Tallahassee, FL 32399, USA.
Accid Anal Prev. 2016 Mar;88:105-16. doi: 10.1016/j.aap.2015.12.012. Epub 2015 Dec 30.
The United States of America and other nations are grappling with the incidence of wrong-way driving (WWD). The issue is as important today (NTSB, 2012) as it was a half-century ago (Hulbert and Beers, 1966). In the absence of a comprehensive analysis, any effort to implement WWD countermeasures can be counterproductive. Hence, this effort began with the express intent to identify the factors that cause WWD crashes and fatalities. This work is sizeable in that it evaluated one million complete crash records from Florida. The methodology comprised (a) administering a survey on the perceptions about WWD; (b) developing binomial logistic models for computing the odds of WWD crashes, and of fatal crashes within the WWD space; (c) analyzing the contributing variables; and (d) comparing perceptions with crash analysis results. The study parameters included driver's age, gender, licensing state, physical defect, blood alcohol concentration, vehicle use, seatbelt compliance, day and time of crash, roadway lighting, facility type, weather conditions, road geometrics, and traffic volumes. Individual variable analysis of 23 parameters and the model development process included the determination of odds ratios and statistical tests for the predictive power and goodness-of-fit. The results of this work are generally consistent with expectation, yet surprising at times. This work concludes with decision-making inputs to the scientist, policy-maker and practitioner on the need for effectively engineering the roads, actively educating people about wrong-way driving, and strictly enforcing traffic laws, rules and regulations.
美国和其他国家正在努力应对逆向行驶(WWD)的发生率。这个问题在今天(NTSB,2012)和半个世纪前(Hulbert 和 Beers,1966)一样重要。如果没有全面的分析,任何实施 WWD 对策的努力都可能适得其反。因此,这项工作从明确导致 WWD 碰撞和死亡的因素开始。这项工作规模庞大,评估了佛罗里达州的一百万份完整碰撞记录。该方法包括:(a) 对 WWD 的看法进行调查;(b) 建立二项逻辑回归模型,以计算 WWD 碰撞的几率,以及 WWD 空间内的致命碰撞几率;(c) 分析相关变量;(d) 将看法与碰撞分析结果进行比较。研究参数包括驾驶员年龄、性别、驾驶执照状态、身体缺陷、血液酒精浓度、车辆使用、安全带合规、碰撞发生的日期和时间、道路照明、设施类型、天气条件、道路几何形状和交通量。对 23 个参数的个体变量分析和模型开发过程包括确定优势比和统计检验,以确定预测能力和拟合优度。这项工作的结果通常与预期一致,但有时也令人惊讶。这项工作的结论是为科学家、政策制定者和从业者提供决策投入,需要有效地对道路进行工程设计,积极教育人们关于错误驾驶的知识,并严格执行交通法规、规则和条例。