Epidemiology and patterns of musculoskeletal motorcycle injuries in the USA

Introduction: Motorcycles have become an increasingly popular mode of transportation despite their association with a greater risk for injury compared with automobiles. Whereas the recent incidence of annual passenger vehicle fatalities in the United States of America (USA) has progressively declined, motorcycle fatalities have steadily increased in the past 11 years. Although motorcycle injuries (MIs) have been studied, to the author’s knowledge there are no published reports on MIs in the USA during this 11-year period. Methods : Study data were derived from a prospectively collected Level I trauma center database. Data sampling included motorcycle crash injury evaluations for the 10-year period ending on 31 August 2008. This retrospective analysis included patient demographic and medical data, helmet use, Glasgow coma scale (GCS) score, injury severity score (ISS), length of hospital stay (LOS), specific injury diagnosis, and death. Data statistics were analyzed using the Spearman correlation coefficient, Kruskal-Wallis tests, and logistic regression. Results: The study identified 1252 motorcycle crash injuries. Helmets were worn by 40.7% of patients for which helmet data were available. The rates of the most common orthopedic injuries were tibia/fibula (19.01%), spine (16.21%), and forearm (10.14%) fractures. The most common non-orthopedic motorcycle crash injuries were concussions (21.09%), skull fractures (8.23%), face fractures (13.66%), and hemo- and pneumothorax (8.79%). There was a significant correlation between greater age and higher ISS (r=0.21, P<0.0001) and longer LOS (r=0.22, P<0.0001). Older patients were also less likely to wear a helmet (OR=0.99, 95% CI: 0.98, 0.997), associated with a significantly higher risk for death (after adjustment for helmet use OR=1.03, 95% CI: 1.00, 1.05). All patients without helmets had a significantly lower GCS score (P=0.0001) and a higher mortality rate (after adjustment for patient demographic data OR=2.28, 95% CI: 1.13, 4.58). Conclusion : Compared with historical reports, the prevalence of skull, face, spine, and pelvis fractures have increased in American motorcycle crashes. Compared to recent European studies, the incidence of USA skull and face fractures is much higher, while the incidence of USA spine and pelvis fractures is more comparable; however, this is not associated with increased in-hospital mortality.

In recent years, the annual incidence of automobile-related fatalities in the USA has progressively decreased, while the annual incidence of motorcycle crash fatalities has progressively increased 19 . However, despite the expanding use of motorcycles, there has been a paucity of orthopedic epidemiological studies related to motorcycle crash in recent years. Our institution has been the primary Level I trauma facility for Galveston County and the surrounding region. The area is one of the most popular motorcycle destinations in Texas, and Galveston is home to the state's largest motorcycle rally 1 . The purpose of this study was to review a single-institution experience with MIs in regard to the epidemiology and patterns of both musculoskeletal and non-musculoskeletal injuries. In addition, this study will compare our modern USA motorcycle injury data with those of earlier international reviews 2,27 .

Materials and methods
The study was conducted in compliance with the University of Texas Medical Branch policies and regulations regarding human subject research following study protocol review and approval by the Institutional Review Board (approval IRB #09-259). Study data were derived from a prospectively collected injury database at our Level I trauma center. The data sampling included 1,221 patients injured in motorcycle crashes for the 10-year period ending August 31, 2008. Each motorcycle trauma event was considered separately and evaluated independently; patients with more than one motorcycle trauma event evaluation during the study time period had each event considered as a separate incident.
This retrospective analysis included data on patient age, gender, race/ethnicity, documentation of helmet use, Glasgow Coma Scale (GCS) score at the time of presentation in the emergency department, injury severity score (ISS), length of hospital stay (LOS), specific injury diagnosis, and mortality.
Descriptive statistics were used to describe the sample characteristics (age, gender, race) and distribution of the outcome variables (GCS, ISS, LOS, type of injury, and death). Non-parametric methods (Spearman correlation coefficient and Kruskal-Wallis tests) were used to test the association between the general patient demographic data, helmet use, and the outcome variables ISS and LOS. Logistic regression analyses were used to test the association of the general patient demographic data and helmet use with patient GCS classification (mild vs. moderate/severe), vital status at discharge and type of injuries (with/without central nervous system injury, dislocation, fracture, and thorax injury). All tests were two-sided with alpha of 0.05 and were performed using SAS 9.3 (SAS Institute Inc, Cary, NC, USA).

Demographic and outcome data
The study identified 1,252 motorcycle trauma event evaluations performed at our institution within a 10 year period. Of these trauma event evaluations, 31 were for a patient's second or third motorcycle injury during the study period. These multiple-event evaluations involved 28 patients and constituted 59 of all 1,252 evaluations. Twenty-five patients were evaluated twice and three patients were evaluated three times. The interval between subsequent trauma event evaluations among these 28 patients ranged from 2 days to 8.5 years.

Specific injury data
The incidence and specific types of motorcycle injuries are listed in Table II-Table V. The most common orthopedic motorcycle injuries were tibia/fibula, spine, and forearm fractures, which occurred in 238 (19%) evaluations, 203 (16.2%), and 127 (10.1%) respectively (Table II and Table III). The highest rates of non-orthopedic motorcycle injuries were for concussions (264 evaluations, 21.09%),   Logistic regression analysis demonstrated that greater age was significantly associated with more severe injury (Spearman correlation coefficient r=0.21, P<0.0001) and a longer hospital stay (Spearman correlation coefficient r=0.22, P<0.0001). Greater age was also associated with significantly higher risks for death following injury, for fracture, and for thorax injury.
Whites had significantly higher ISS and LOS and higher risk for fracture.
Among all patients who did not wear a helmet, the risk was significantly higher for a more severe GCS score (adjusted odds ratio 2.98) and for death (adjusted odds ratio 2.95). These patients were also more likely to have a fracture of the skull (P<0.001) or face (P<0.001).
Fracture results from our current study and from earlier studies from England, Sweden, and the USA are presented in Table VI. The rates of skull and face fractures are higher in our current study than in England or Sweden. The incidence of skull, spine, and rib fractures has nearly doubled in the USA over the last 30 years, and pelvis fractures are much more prevalent.

Discussion
Many epidemiological studies have addressed motorcycle crash injuries, in particular the effect of helmet use 3,5,6,9,25 . There has been a relative paucity, however, of new information regarding orthopedic motorcycle crash injuries in USA over the last 30 years. Unfortunately, during this period motorcycle crashes continued to be a major cause of accident-related morbidity and mortality. It has been reported that in 2007 motorcyclists were 37 times more likely to die in a motor vehicle traffic crash than passenger car occupants, and nine times more likely to be injured 20 . Whereas car and truck fatality rates have decreased for the past 6 and 3 years, respectively, 2008 marked the eleventh consecutive year in which the annual motorcycle fatality rate increased 19 . Yet, to our knowledge, the most recent epidemiological studies on orthopedic motorcycle crash injuries in the USA were by Bried et al. 6 in 1987 and Peek et al. 22 in 1994.
According to our data, the rates of skull, face, spine, rib, and pelvis fractures have increased in the USA motorcycle crashes compared with historical reports 9,11,28 . The rates of skull and face fractures in our study are substantially higher than in recent European studies (Table VI), whereas the rates of spine and pelvis fractures have remained similar. The higher incidence of some of the injuries documented in our study may be attributed to better injury detection. In the USA, whole-body computed tomography has become a routine trauma screening tool 24 . Moreover, European riders are more likely to wear a helmet 2 , and their lower incidence of skull and face fractures and death might be attributed to this protective device. The protective nature of helmets has been described in a number of studies 13, 16 .   The limitations of our study include its retrospective nature and the fact that it relies upon a trauma database. Although the data were collected prospectively, they were not specifically designed for this study. Orthopedic injuries were grouped by anatomic location as opposed to specific fracture type to control for small errors of injury classification (e.g., injury was classified as a radial shaft fracture as opposed to a Galeazzi or simple distal radius fracture). However, these injury classifications are similar to those employed in prior studies 2,6,8,9,11,22,27,28 , and many of those study data were also derived from existing databases 2,4,12 .
Although the motorcycle fatality rate continues to increase, we did not see a similar increase in in-hospital mortality rate, indicating that more of the involved participants are likely being declared dead at the scene. The increase in the incidence of skull, spine, and face injuries noted in the study may be attributed to better injury detection (routine CT and/or MRI). The higher incidence of USA skull and facial fractures, as compared to other countries, may also be attributed to differences in compulsory helmet wearing laws. Older MI patients are a higher risk for serious injury and mortality, and this is further potentiated by not wearing a helmet.

Data availability
Figshare: Raw data used for the analysis in the study on epidemiology and patterns of musculoskeletal injuries. Doi: 10.6084/ m9.figshare.1402186 29 Author contributions STB -data collection, data analysis, manuscript preparation.
ZG -study design, data analysis, manuscript preparation.
All authors read and approved the revised manuscript.

Competing interests
No competing interests were disclosed.

Grant information
The author(s) declared that no grants were involved in supporting this work.