Injury severity in relation to seatbelt use.

5 January 2015, Vol. 105, No. 1 Injury severity in relation to seatbelt use To the Editor: I read the article by Van Hoving et al.[1] with great interest. Two interesting aspects of the study are that it included patients in the prehospital setting, and involved both the Division of Emergency Medicine and the Department of Civil Engineering at their institution (Stellenbosch University, Western Cape, South Africa). The authors acknowledge that it is a pilot study with many limitations. Nevertheless, important lessons can be learned from it that may improve the design and performance of future studies. First, it is known that using seatbelts reduces injury severity and death.[2,3] However, owing to the complexity of road traffic injuries, it is important to include the maximum number of injured patients, depending on the inclusion/exclusion criteria, to strengthen the analysis and permit a proper multivariate analysis to answer the research question properly. It is not clear to me, for example, why patients involved in taxi accidents (n=94) were excluded from the present study.[1] This is an unusual reason for exclusion, especially when only 107 patients were included in the prehospital phase. Second, study of the biomechanics of road traffic collisions (RTCs) in more detail, including ejection from the vehicle, is essential. The authors cited our study reporting a significant effect of seatbelt use in reducing injury severity when we performed a univariate comparison between patients who had used seatbelts and those who had not.[4] However, once we included other important factors using a general linear model,[5] we found that the mechanism of the collision and vehicle speed were the most important factors affecting the severity of RTC injuries, while the effect of seatbelts became non-significant, indicating that the main effect of seatbelt use was to reduce ejection from the vehicle.[5] Injury severity among patients who were ejected was double that in cases in which the vehicle sustained lateral impact or rolled over.[5] I note that ejection was not included in Van Hoving et al.’s[1] study. Third, there were many missing hospital data in the study.[1] We have addressed this issue by appointing a full-time researcher to collect data prospectively from hospitals on a daily basis. This has increased the completeness of data on seatbelt use to 98% in our Road Traffic Collision Registry.[4] Finally, I congratulate the authors on their important study and hope that they will find these comments useful.