Di Bartolomeo S, Valent F, Marchetti R, Sbrojavacca R, Barbone F
Cattedra di Igiene ed Epidemiologia, Università degli Studi di Udine.
Ann Ig. 2010 Sep-Oct;22(5):401-18.
In case-crossover studies, useful for assessing the effect of acute transient exposures, each case acts as his/her own control, therefore, interpersonal confounding is controlled for by study. We used such design for studying the effect of the acute consumption of alcohol and meals, and of sleep and work hours on the risk of road traffic crashes. Subjects, enrolled at the Emergency Room (ER) of Udine from 12/3/2007 to 11/3/2008, were a sample of the drivers who arrived alive at the ER after a crash. They were interviewed by trained interviewers, who systematically covered predefined shifts, using a semi-structured questionnaire. The questionnaire collected information on the subjects, vehicles, and crashes, and contained a hourly diary of the exposure to driving, sleeping, working in the 48 hours before the crash and to alcohol and meals 24 hours before the crash. The statistical analysis was based on the matched pair interval approach. The exposures in the hours immediately before the crash (case window) were compared with those in previous hours (control window). Different window durations were chosen for different exposures. The relative risk (RR) of having a crash and 95% confidence intervals (95% CI) were estimated with conditional logistic regression, adjusting for potential confounders such as day of the week and time. We observed a statistically significant two-fold increase in the RR for drivers who had consumed alcohol (even small amounts) and a four-fold increase for those who had worked > 12 ore. The RR was increased by 10 times for drivers who had been awake for at least 16 hours. Meals were not associated with the risk of crash; the findings regarding sleep amount were controversial. In conclusion, the study confirms an increased risk of road crashes after consuming alcohol, even for amounts below the legal limit, and suggests that extended work hours and prolonged wakefulness may increase the risk of crashes.
在病例交叉研究中,这种研究方法对于评估急性短暂暴露的影响很有用,每个病例都作为其自身的对照,因此,人际混杂因素通过研究得以控制。我们采用这种设计来研究急性饮酒、用餐以及睡眠和工作时长对道路交通事故风险的影响。研究对象为2007年12月3日至2008年3月11日在乌迪内急诊室(ER)登记的样本,这些司机在发生车祸后活着抵达急诊室。由经过培训的访谈者对他们进行访谈,访谈者使用半结构化问卷系统地涵盖预先确定的时间段。问卷收集了有关研究对象、车辆和车祸的信息,并且包含了车祸前48小时内驾驶、睡眠、工作的每小时记录,以及车祸前24小时内饮酒和用餐的记录。统计分析基于配对区间法。将车祸发生前即刻的时间段(病例窗口)内的暴露情况与之前时间段(对照窗口)内的暴露情况进行比较。针对不同的暴露情况选择了不同的窗口时长。通过条件逻辑回归估计发生车祸的相对风险(RR)和95%置信区间(95%CI),并对潜在混杂因素如星期几和时间进行了调整。我们观察到,饮酒(即使少量)的司机RR有统计学意义的两倍增加,工作超过12小时的司机RR增加四倍。清醒至少16小时的司机RR增加了10倍。用餐与车祸风险无关;关于睡眠时间的研究结果存在争议。总之,该研究证实饮酒后道路交通事故风险增加,即使饮酒量低于法定限值,并表明延长工作时间和长时间清醒可能会增加车祸风险。