ORIGINAL ARTICLE
Helmet use is associated with safer bicycling behaviors and reduced hospital resource use following injury Rachel Webman, MD, Linda A. Dultz, MD, MPH, Ronald J. Simon, MD, S. Rob Todd, MD, Dekeya Slaughter, BSc (Hons), Sally Jacko, RN, MPH, Omar Bholat, MD, Stephen Wall, MD, MSc, MAEd, Chad Wilson, MD, Deborah A. Levine, MD, Matthew Roe, MS, H. Leon Pachter, MD, and Spiros G. Frangos, MD, MPH, New York, New York
While the efficacy of helmet use in the prevention of head injury is well described, helmet use as it relates to bicyclists’ behaviors and hospital resource use following injury is less defined. The objective of this study was to compare the demographics, behaviors, hospital workups, and outcomes of bicyclists based on helmet use. METHODS: This study was a subset analysis of a 2.5-year prospective cohort study of vulnerable roadway users conducted at Bellevue Hospital Center, a New York City Level 1 trauma center. All bicyclists with known helmet status were included. Demographics, insurance type, traffic law compliance, alcohol use, Glasgow Coma Scale (GCS) score, initial imaging studies, Abbreviated Injury Scale (AIS) score, Injury Severity Score (ISS), admission status, length of stay, disposition, and mortality were assessed. Information was obtained primarily from patients; witnesses and first responders provided additional information. RESULTS: Of 374 patients, 113 (30.2%) were wearing helmets. White bicyclists were more likely to wear helmets; black bicyclists were less likely ( p = 0.037). Patients with private insurance were more likely to wear helmets, those with Medicaid or no insurance were less likely ( p = 0.027). Helmeted bicyclists were more likely to ride with the flow of traffic (97.2%) and within bike lanes (83.7%) ( p G 0.001 and p = 0.013, respectively). Nonhelmeted bicyclists were more likely to ride against traffic flow ( p = 0.003). There were no statistically significant differences in mean GCS score, AIS score, and mean ISS for helmeted versus nonhelmeted bicyclists. Nonhelmeted patients were more likely to have head computed tomographic scans ( p = 0.049) and to be admitted ( p = 0.030). CONCLUSION: Helmet use is an indicator of safe riding practices, although most injured bicyclists do not wear them. In this study, helmet use was associated with lower likelihood of head CTs and admission, leading to less hospital resource use. Injured riders failing to wear helmets should be targeted for educational programs. (J Trauma Acute Care Surg. 2013;75: 877Y881. Copyright * 2013 by Lippincott Williams & Wilkins) LEVEL OF EVIDENCE: Epidemiologic study, level III. KEY WORDS: Bicyclist; helmet; behavior; injury. BACKGROUND:
I
n 2010, 52,000 bicyclists were injured, and 618 were killed in motor vehicle collisions in the United States.1 Bicyclists were involved in 1.9% of all US traffic crashes.1 In New York City (NYC), bicyclists sustained traumatic head injuries in 77% of fatal crashes; helmets were used in only 3% of these incidents.2 The number of bicyclists in NYC is rising, and the local government is contributing toward making streets safer by implementing new bike lanes and vehicle-protected bike paths as well as by revising transportation laws, which include requiring employers to provide helmets for commercial bicyclists.3,4
Submitted: May 16, 2013, Revised: July 22, 2013, Accepted: July 23, 2013. From the Departments of Surgery (R.W., L.A.D., R.J.S., S.R.T., D.S., S.J., O.B., C.W., H.L.P., S.G.F.), Emergency Medicine (S.W., D.A.L.), and Pediatrics (D.A.L.), Bellevue Hospital Center, New York University School of Medicine; and Office of Research, Implementation and Safety (M.R.), NYC Department of Transportation, New York, New York. The sponsor did not participate in the design and conduct of the study; in the collection, management, analysis, and interpretation of the data; or in the preparation, review, or approval of the manuscript. Address for reprints: Spiros G. Frangos, MD, MPH, NYU School of Medicine, 550 First Ave, NBV 15 S7, New York, NY 10016-9196; email: Spiros.frangos@ nyumc.org. DOI: 10.1097/TA.0b013e3182a85f97
Morbidity and mortality remain high with 6 bicyclist deaths and 1,144 bicyclist injuries in New York County (i.e., Manhattan) in 2011.5 The efficacy of helmets in preventing or minimizing head injuries among bicyclists has been shown.6Y12 However, helmet use as it relates to bicyclists’ behaviors and hospital resource use following injury is less defined. Assessing the demographics and behaviors of bicyclists who wear helmets and comparing these variables to those who do not may provide insight into injury prevention strategies and targeted outreach. The objective of this study was to compare the demographics, behaviors, initial hospital evaluation, and outcomes of bicyclists based on their use of helmets. Our hypothesis was that helmeted bicyclists engage in safer riding practices.
PATIENTS AND METHODS This study was a subset analysis of a prospective cohort study, which evaluated vulnerable roadway users injured by motor vehicles in NYC.13 Data collection was performed at Bellevue Hospital Center, a NYC Level 1 trauma center, between December 2008 and June 2011. Bellevue Hospital Center’s catchment area includes midtown and lower Manhattan and
J Trauma Acute Care Surg Volume 75, Number 5
Copyright © 2013 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
877
J Trauma Acute Care Surg Volume 75, Number 5
Webman et al.
western Brooklyn, and its emergency department (ED) evaluates more than 100,000 patients annually. Prospective data had been collected on all bicyclists and pedestrians who presented within 24 hours of collisions involving motor vehicles. Patients requiring hospitalization and those discharged from the ED were included; both walk-ins and transfers were also included. All pedestrian data were excluded for this subset analysis. Pertinent bicyclist variables captured in the database included patient demographics, helmet use at time of injury, behaviors (e.g., riding with or against the flow of traffic), Glasgow Coma Scale (GCS) score at time of presentation, alcohol use before incident, initial computed tomography (CT) imaging studies, Abbreviated Injury Scale (AIS) score, Injury Severity Score (ISS), admission status, hospital length of stay (LOS), disposition, and mortality. Unknown or indeterminate helmet status was a criterion for exclusion. ED staff were informed of the study before it began. Informational fliers were hung throughout the ED. A pager number dedicated to the study was provided to ED staff, and the pager was activated whenever a patient meeting criteria presented. Notifications were accepted 24 hours a day and 365 days per year. Data collection was performed primarily by a dedicated study coordinator, a trauma coordinator, and attending physicians in trauma and emergency medicine. Verbal informed consent was obtained from all patients before study inclusion. Most data were obtained from patients/families and transcribed onto a six-page questionnaire by the interviewer. Emergency medical technicians, New York Police Department officers, and scene witnesses provided additional scene data when possible. Ambulance call reports were used to corroborate or verify data for most patients. Data collected on patients unable to grant verbal consent were used only if consent could be obtained at a later time. Findings from imaging studies were reviewed, and hospital LOS and disposition data were added soon after discharge. AIS score and ISS were calculated for each patient by a single trauma surgeon after attending radiology evaluations were finalized. A blood alcohol concentration (BAC) greater than 0.01 g/dL was used to determine whether patients had consumed alcohol before injury. BAC level measurements were performed as part of a routine workup in most cases. If objective laboratory data were not available, alcohol use was determined by patient self-report. If BAC results conflicted with the patient self-report, BAC took precedence. Data were analyzed using SPSS version 18 software (IBM). Pearson’s W2 and independent two-tailed t tests were performed, with p G 0.05 representing statistical significance. A univariable analysis was used to assess efficacy of helmet use with respect to the prevention of significant head injury. This study was approved by both the New York University School of Medicine and the Bellevue Hospital Center institutional review boards. This study was funded by a Highway Safety Grant from the State of New York Governor’s Traffic Safety Committee.
RESULTS There were 382 bicyclists. Of these, eight had unknown or indeterminate helmet status and were excluded from the analysis. Of the remaining 374 patients, 113 (30.2%) were 878
wearing helmets at the time of injury (Table 1). Twenty-five percent of women and 31.1% of men wore helmets ( p = 0.420). White bicyclists were more likely to wear helmets, while black bicyclists were less likely ( p = 0.037). Patients with private insurance were more likely to wear helmets, while those with either Medicaid or no insurance were less likely ( p = 0.027). There was no difference in age, primary language, or working status between the helmeted and nonhelmeted groups. Among those who were riding as part of employment (e.g., delivery workers, bike messengers) (n = 163), 52 (31.9%) were wearing helmets; 60 (28.8%) of the nonworking bicyclists (n = 208) were wearing helmets when injured ( p = 0.110). Behaviors of bicyclists are depicted in Table 2. Helmeted bicyclists were more likely to ride with the flow of traffic (97.2%) and to be within bike lanes when available (83.7%) at the time of the collision, as compared with 83.6% and 61.3% of nonhelmeted bicyclists ( p G 0.001 and p = 0.013, respectively). Nonhelmeted bicyclists were also more likely to be riding against traffic ( p = 0.003). There was no statistical difference with regard to alcohol use. Of the seven helmeted alcohol users, five had BACs drawn (all were 90.08 g/dL) and two selfreported use. Of the 29 nonhelmeted alcohol users, 20 had BACs drawn (13 were 90.08 g/dL, while 7 were 0.01Y0.08 g/dL) and 9 self-reported use.
TABLE 1. Patient Demographics, Insurance, and Work Status Categories Sex Male, n (%) Age, Y Age, mean (SD) Ethnicity/race, n (%) White Black Latino East Asian South Asian Other Language,* n (%) English Spanish Chinese Russian Bengali Other Insurance,** n (%) Medicaid Medicare Private insurance Worker’s compensation Uninsured Other
Helmet (n = 113) No Helmet (n = 261)
p
100 (88.5)
222 (85.1)
0.420
32.73 (12.41)
31.84 (13.23)
0.541
49 (43.4) 11 (9.7) 41 (36.3) 7 (6.2) 5 (4.4) 0 (0)
87 (33.3) 43 (16.5) 100 (38.3) 22 (8.4) 3 (1.1) 6 (2.3)
0.037
64 (56.6) 34 (30.1) 5 (4.4) 1 (0.9) 4 (3.5) 5 (4.4)
150 (57.7) 77 (29.6) 14 (6.9) 1 (0.4) 1 (0.4) 13 (5.0)
0.214
8 (8.8) 0 (0) 41 (45.1) 0 (0) 37 (40.7) 5 (5.5)
27 (14.1) 1 (0.5) 48 (25.1) 2 (1.0) 102 (53.4) 11 (5.8)
0.027
*One unknown removed. **n = 292 because 92 were unknown.
* 2013 Lippincott Williams & Wilkins
Copyright © 2013 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 75, Number 5
Webman et al.
TABLE 2. Bicyclist Behaviors Categories
TABLE 4. Initial CT Imaging Studies
Helmet (n = 113) No Helmet (n = 261)
Riding patterns, n (%) With flow of traffic 106 (97.2)* Against flow of traffic 3 (2.8)* Crossing against signal 6 (5.5)† Riding in bike lane 36 (83.7)§ Alcohol involvement, n (%) Yes 7 (6.3)||
p
204 (83.6)** 30 (12.3)** 30 (12.6)‡ 46 (61.3)¶
G0.001 0.003 0.057 0.013
29 (11.1)
0.181
*Four unknowns removed. **Seventeen unknowns removed. †Four unknowns removed. ‡Twenty-two unknowns removed. §n = 43 (patients were excluded because bike lane was unavailable or bike lane availability or use unknown). ¶n = 75 (patients were excluded because bike lane was unavailable or bike lane availability or use unknown). ||Two unknowns removed.
TABLE 3. GCS Score, AIS Score, and ISS Categories GCS score* GCS score, mean (SD) GCS e 8, n (%) GCS 9 8, n (%) AIS categories, n (%) Head and neck AIS score 1Y6 Face AIS score 1Y6 Chest AIS score 1Y6 Abdomen AIS score 1Y6 Extremity AIS score 1Y6 External AIS score 1Y6 ISS* ISS, mean (SD) No injuries, n (%) 1Y8, n (%) 9Y15, n (%) 16Y24, n (%) Q25, n (%) Head AIS only,** n (%) AIS score 0 AIS score 1 AIS score 2 AIS score 3 AIS score 4 AIS score 5 AIS score 6 Craniotomy Yes, n (%)
Helmet (n = 113)
No Helmet (n = 261)
14.93 (0.499) 0 (0) 111 (100)
14.79 (1.293) 3 (1.2) 257 (98.8)
0.285 0.557
6 (5.3) 5 (4.4) 5 (4.4) 6 (5.3) 27 (23.9) 92 (81.4)
21 (8.0) 23 (8.8) 14 (5.4) 12 (4.6) 64 (24.5) 216 (82.8)
0.394 0.198 0.803 0.795 1.000 0.769
3.18 (7.205) 20 (17.7) 83 (73.5) 7 (6.2) 1 (0.9) 2 (1.8)
4.13 (7.478) 34 (13.0) 187 (71.6) 23 (8.8) 7 (2.7) 10 (3.8)
0.255 0.388
107 (94.7) 0 (0) 0 (0) 2 (1.8) 0 (0) 1 (0.9) 0 (0)
240 (92.0) 0 (0) 1 (0.4) 5 (1.9) 8 (3.1) 1 (0.4) 0 (0)
0.360
1 (0.4)†
1.000
0 (0)
*Three unknowns excluded. **Excludes those with neck injuries. †Evacuation of epidural hematoma.
p
Categories
Helmet (n = 113)
CT scan, n (%) CT head CT cervical spine CT chest CT abdomen/pelvis
35 38 20 23
No Helmet (n = 261)
p
110 (42.1) 96 (36.8) 44 (16.9) 66 (25.3)
0.049 0.639 0.881 0.355
(31.0) (33.6) (17.7) (20.4)
There were no significant differences in mean GCS score, severe head injury (i.e., GCS score e 8), AIS categories (including head and neck AIS), mean ISS, or severity for category-based ISS between groups (Table 3). Head and neck AIS was further broken down to ‘‘head AIS only’’ (Table 3). There was no difference in ‘‘head AIS only’’ or rate of craniotomy between helmeted and nonhelmeted patients. A majority in both groups (94.7% helmeted, 92% nonhelmeted) had no intracranial injuries. Head CTs were more likely to be performed on patients who were nonhelmeted (42.1%) at the time of injury as opposed to their helmeted counterparts (p = 0.049) (Table 4). CT scans of the cervical spine, chest, abdomen/pelvis were as likely to be performed in one group as the other. Nonhelmeted bicyclists were more likely to be admitted to the hospital (p = 0.030) (Table 5). Mean LOS, disposition, and mortality rates were not statistically different between groups. Three nonhelmeted bicyclists died, including two of multisystem organ failure and one of hemorrhagic shock at time of presentation. There were no deaths in the helmeted group.
DISCUSSION Evidence for a relationship between helmet use and bicyclist behaviors, including obeying traffic laws and bicycle lane use, is limited. Farris et al.14 looked at certain behaviors including hand signaling and adherence to stop signs; they found that bicyclists who wore helmets were more likely to engage in these safe biking practices. Wasserman et al.15 investigated seat belt use in motor vehicles and found that buckling up correlated with helmet use while bicycling. In both studies, helmet use was associated with safer behaviors. Our study supports the notion that observance of traffic laws correlates with helmet use. Specifically, bicyclists who wore
TABLE 5. Patient Outcomes and Disposition Admissions and Hospital LOS Admitted, n (%) LOS, mean (SD), d Patient disposition, n (%) Home Rehabilitation Other Mortality Died, n (%)
17 (15.0) 0.96 (4.819)
66 (25.3) 1.98 (9.254)
0.030 0.268
109 (96.5) 3 (2.7) 1 (0.9)
240 (92.0) 12 (4.6) 9 (3.4)
0.240
0 (0)
3 (1.1)
0.555
* 2013 Lippincott Williams & Wilkins
Copyright © 2013 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
879
J Trauma Acute Care Surg Volume 75, Number 5
Webman et al.
helmets were more likely to ride with the flow of traffic and to be in designated bike lanes at time of injury. A negative correlation between alcohol and helmet use is well documented.16Y18 This variable did not achieve statistical significance in our study. Our results suggest that the riskier behaviors of nonhelmeted bicyclists extend beyond the use and influence of alcohol. NYC requires working bicyclists and children younger than 13 years to wear helmets.3 Despite this legislation, our study found no difference in helmet use between working bicyclists and those biking for recreation or commuting purposes. While other studies have shown that instituting helmet laws increased helmet use in children,19,20 our small pediatric sample size precludes us from making definitive conclusions for this cohort. Previous studies have concluded that children of parents of higher socioeconomic status (SES) are more likely to wear helmets.20Y22 Rodgers23 looked at helmet use among bicyclists of all ages using level of education as a proxy for SES and found greater levels of helmet use to correlate with higher levels of education. Our study suggests that patients of lower SES, inferred from a lack of health insurance or need for Medicaid applications, were less likely to wear helmets. Dellinger and Kresnow22 reported racial and ethnic differences among pediatric bicyclists who wore helmets; this demographics and its association with helmets have not been well defined among adults. Consistent with their results,22 our study found that, among ethnic groups, white bicyclists were more likely while black bicyclists were less likely to wear helmets. Our study found that nonhelmeted bicyclists were more likely to undergo head CT scans and had higher rates of hospital admissions as compared with helmeted bicyclists. This difference was apparent despite injury scores that were not statistically different between these groups. CT scanning is an expensive imaging modality and has associated radiation exposure risks. Likewise, hospital admissions including a mere one-night stay may incur costs in thousands of dollars. Given that the nonhelmeted cohort in our patient population was more likely to have government insurance or no insurance, a substantial portion of these health care costs get distributed across the general public. Crocker et al.18 reported a similar correlation in lower SES bicyclists who were impaired by alcohol. Many studies have shown the efficacy of helmet use in preventing head and facial injuries and in reducing mortality.6Y12 Because our methodology involved inclusion of both admitted and nonadmitted patients, the low number of traumatic head injuries and the rarity of deaths likely precluded this study from being powered to capture the protective effects of helmet use. The Manhattan study area is in an extremely dense urban core with low-to-moderate traffic speeds and mostly low-speed interactions between bicyclists and motor vehicles. The low mean ISSs among bicyclists, while expected given the relatively slow motor vehicle speeds, make injury comparisons between bicycling cohorts more challenging. This study has several strengths. First, it was conducted in a prospective manner, helping to minimize missing data points. Second, nonadmitted patients were included in the sample and not merely hospital admissions as most previous studies have done; in evaluating at-risk behaviors for injury, 880
artificially limiting the sample by excluding nonadmitted patients introduces a selection bias, which our study was designed to limit. Finally, information for most incidents was obtained from multiple sources including first responders when available, thereby introducing an additional, objective viewpoint. There are also limitations to this study. Because the accuracy of data relied heavily on patient self-report, it is possible that individual patients may have underreported certain behaviors for personal, cultural, economic, and/or legal reasons. Second, a minority of patients had poor or no recall of the incident secondary to alcohol use or traumatic head injury. Third, the sample size was not powered to investigate injury and mortality differences, which could have been informative and added to the other analyses. The low mean ISS in both groups also denotes a population that sustained predominantly mild-to-moderate trauma, leading to less pronounced differences in injuries and outcomes between cohorts. Fourth, patients declared dead at the scene and those taken to other local hospitals were not included. Fifth, this study was performed in a single institution, so its results are representative of the unique catchment area of Bellevue Hospital Center and are not generalizable across NYC as a whole. Sixth, these behaviors reflect those of bicyclists involved in crashes and may not be representative of bicyclists who avoid them. Finally, this study was conducted in an urban environment, and therefore, whether these data are applicable to suburban or rural areas is unclear and requires further investigation. Although most injured bicyclists do not wear them, helmets are indicators of basic safe riding practices. In this study, helmet use was associated with reduced hospital resource use based on less head CT imaging and lower likelihood of admission. In light of our national focus on preventative health and reducing health care costs, bicyclists who opt not to wear helmets may benefit from safe bicycling educational programs, while injured nonhelmeted riders should be targeted; especially in urban core areas, bicycling without a helmet represents more than merely a vulnerability to head injury in the event of a crash. AUTHORSHIP R.J.S., D.S., S.J., O.B., D.A.L., and S.G.F. contributed to the study conception and design. R.W., L.A.D., R.J.S., S.R.T., D.S., S.J., O.B., S.W., C.W., D.A.L., and S.G.F. performed the acquisition of data. R.W., L.A.D., S.R.T., D.S., S.W., M.R., H.L.P., and S.G.F. performed the analysis and interpretation of data. R.W., L.A.D., and S.G.F. drafted the manuscript. R.W., L.A.D., R.J.S., S.R.T., D.S., S.J., S.W., O.B., C.W., D.A.L., M.R., H.L.P., and S.G.F. provided critical revision.
DISCLOSURE This study was funded by a Highway Safety Grant from the State of New York Governor’s Traffic Safety Committee (October 1, 2008, to September 30, 2011; Year 1, $122,242; Year 2, $129,748; Year 3, $130,670). D.S., S.J., S.G.F., and D.A.L. received salary support from this grant.
REFERENCES 1. US Department of Transportation, National Highway Safety Administration. Traffic Safety Facts. (2012) Available at: www-nrd.nhtsa.dot.gov/ Pubs/811606.pdf. Accessed January 28, 2013.
* 2013 Lippincott Williams & Wilkins
Copyright © 2013 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 75, Number 5
2. Nicaj L, Stayton C, Mandel-Ricci J, McCarthy P, Grasso K, Woloch D, Kerker B. Bicyclist fatalities in New York City: 1996Y2005. Traffic Inj Prev. 2009;10:157Y161. 3. New York City Administrative Code: N.Y.C. Admin. Code §§ 10Y157. (bicycles used for commercial purposes). 4. New York City. Department of Transportation. Bureau of Traffic Operations. NYCDOT-DOT Announces 35% Increase in Commuter Cycling from 2007 to 2008 and Calls on Cyclists to Use Lights to Be Seen & Safe. (2008) Available at: http://www.nyc.gov/html/dot//html/pr2008/pr08_047.shtml. Accessed January 28, 2013. 5. Governor’s Traffic Safety Committee, SafeNY. County Data Reports. (2011) Available at: www.safeny.ny.gov/hsdata.htm. Accessed January 28, 2013. 6. Amoros E, Chiron M, Martin JL, Thelot B, Laumon B. Bicycle helmet wearing and the risk of head, face, and neck injury: a French case-control study based on a road trauma registry. Inj Prev. 2012;18:27Y32. 7. Heng KWJ, Lee AHP, Zhu S, Tham KY, Seow E. Helmet use and bicyclerelated trauma in patients presenting to an acute hospital in Singapore. Singapore Med J. 2006;47:367Y372. 8. Persaud N, Coleman E, Zwolakowski D, Lauwers B, Cass D. Nonuse of bicycle helmets and risk of fatal head injury: a proportional mortality, casecontrol study. CMAJ. 2012;184:E921YE923. 9. Thompson RS, Rivara FP, Thompson DC. A case-control study of the effectiveness of bicycle safety helmets. N Engl J Med. 1989;320: 1361Y1367. 10. Thompson DC, Thompson RS, Rivara FP, Wolf ME. A case-control study of the effectiveness of bicycle safety helmets in preventing facial injury. Am J Public Health. 1990;80:1471Y1474. 11. Thompson DC, Rivara F, Thompson R. Helmets for preventing head and facial injuries in bicyclists. Cochrane Database Syst Rev. 1999;4: CD001855. 12. Wasserman RC, Buccini RV. Helmet protection from head injuries among recreational bicyclists. Am J Sports Med. 1990;18:96Y97.
Webman et al.
13. Dultz LA, Foltin G, Simon R, Wall S, Levine DA, Bholat O, SlaughterLarkem D, Jacko S, Marr M, Glass N, et al. Vulnerable roadway users struck by motor vehicles at the center of the safest, large US city. J Trauma Acute Care Surg. 2013;74(4):1138Y1145. 14. Farris C, Spaite DW, Criss EA, Valenzuela TD, Meislin HW. Observational evaluation of compliance with traffic regulations among helmeted and nonhelmeted bicyclists. Ann Emerg Med. 1997;29:625Y629. 15. Wasserman RC, Waller JA, Monty MJ, Emery AB, Robinson DR. Bicyclists, helmets and head injuries: a rider-bases study of helmet use and effectiveness. Am J Public Health. 1988;78:1220Y1221. 16. Li G, Baker SP, Smialek JF, Soderstrom CA. Use of alcohol as a risk factor for bicycling injury. JAMA. 2001;285:893Y896. 17. Spaite DW, Criss EA, Weist DJ, Valenzuela TD, Judkins D, Meislin HW. A prospective investigation of the impact of alcohol consumption on helmet use, injury severity, medical resource utilization, and healthcare costs in bicycle-related trauma. J Trauma. 1995;38:287Y290. 18. Crocker P, Zad O, Milling T, Lawson KA. Alcohol, bicycling, and head and brain injury: a study of impaired cyclists’ riding patterns R1. Am J Emerg Med. 2010;28:68Y72. 19. Macpherson AK, To TM, Parkin PC, Moldofsky B, Wright JG, Chipman ML, Macarthur C. Urban/rural variation in children’s bicycle-related injuries. Accid Anal Prev. 2004;36:649Y654. 20. Parkin PC, Khambalia A, Kmet L, Macarthur C. Influence of socioeconomic status on the effectiveness of bicycle helmet legislation for children: a prospective observational study. Pediatrics. 2003;112:e192Ye196. 21. Lang IA. Demographic, socioeconomic, and attitudinal associations with children’s cycle-helmet use in the absence of legislation. Inj Prev. 2007;13: 355Y358. 22. Dellinger AM, Kresnow M. Bicycle helmet use among children in the United States: the effects of legislation, personal and household factors. J Saf Res. 2010;41:375Y380. 23. Rodgers GB. Bicycle helmet use patterns in the United States: a description and analysis of national survey data. Accid Anal Prev. 1995;27:43Y56.
* 2013 Lippincott Williams & Wilkins
Copyright © 2013 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
881
Helmet use is associated with safer bicycling behaviors and reduced hospital resource use following injury Rachel Webman, MD, Linda A. Dultz, MD, MPH, Ronald J. Simon, MD, S. Rob Todd, MD, Dekeya Slaughter, BSc (Hons), Sally Jacko, RN, MPH, Omar Bholat, MD, Stephen Wall, MD, MSc, MAEd, Chad Wilson, MD, Deborah A. Levine, MD, Matthew Roe, MS, H. Leon Pachter, MD, and Spiros G. Frangos, MD, MPH, New York, New York
While the efficacy of helmet use in the prevention of head injury is well described, helmet use as it relates to bicyclists’ behaviors and hospital resource use following injury is less defined. The objective of this study was to compare the demographics, behaviors, hospital workups, and outcomes of bicyclists based on helmet use. METHODS: This study was a subset analysis of a 2.5-year prospective cohort study of vulnerable roadway users conducted at Bellevue Hospital Center, a New York City Level 1 trauma center. All bicyclists with known helmet status were included. Demographics, insurance type, traffic law compliance, alcohol use, Glasgow Coma Scale (GCS) score, initial imaging studies, Abbreviated Injury Scale (AIS) score, Injury Severity Score (ISS), admission status, length of stay, disposition, and mortality were assessed. Information was obtained primarily from patients; witnesses and first responders provided additional information. RESULTS: Of 374 patients, 113 (30.2%) were wearing helmets. White bicyclists were more likely to wear helmets; black bicyclists were less likely ( p = 0.037). Patients with private insurance were more likely to wear helmets, those with Medicaid or no insurance were less likely ( p = 0.027). Helmeted bicyclists were more likely to ride with the flow of traffic (97.2%) and within bike lanes (83.7%) ( p G 0.001 and p = 0.013, respectively). Nonhelmeted bicyclists were more likely to ride against traffic flow ( p = 0.003). There were no statistically significant differences in mean GCS score, AIS score, and mean ISS for helmeted versus nonhelmeted bicyclists. Nonhelmeted patients were more likely to have head computed tomographic scans ( p = 0.049) and to be admitted ( p = 0.030). CONCLUSION: Helmet use is an indicator of safe riding practices, although most injured bicyclists do not wear them. In this study, helmet use was associated with lower likelihood of head CTs and admission, leading to less hospital resource use. Injured riders failing to wear helmets should be targeted for educational programs. (J Trauma Acute Care Surg. 2013;75: 877Y881. Copyright * 2013 by Lippincott Williams & Wilkins) LEVEL OF EVIDENCE: Epidemiologic study, level III. KEY WORDS: Bicyclist; helmet; behavior; injury. BACKGROUND:
I
n 2010, 52,000 bicyclists were injured, and 618 were killed in motor vehicle collisions in the United States.1 Bicyclists were involved in 1.9% of all US traffic crashes.1 In New York City (NYC), bicyclists sustained traumatic head injuries in 77% of fatal crashes; helmets were used in only 3% of these incidents.2 The number of bicyclists in NYC is rising, and the local government is contributing toward making streets safer by implementing new bike lanes and vehicle-protected bike paths as well as by revising transportation laws, which include requiring employers to provide helmets for commercial bicyclists.3,4
Submitted: May 16, 2013, Revised: July 22, 2013, Accepted: July 23, 2013. From the Departments of Surgery (R.W., L.A.D., R.J.S., S.R.T., D.S., S.J., O.B., C.W., H.L.P., S.G.F.), Emergency Medicine (S.W., D.A.L.), and Pediatrics (D.A.L.), Bellevue Hospital Center, New York University School of Medicine; and Office of Research, Implementation and Safety (M.R.), NYC Department of Transportation, New York, New York. The sponsor did not participate in the design and conduct of the study; in the collection, management, analysis, and interpretation of the data; or in the preparation, review, or approval of the manuscript. Address for reprints: Spiros G. Frangos, MD, MPH, NYU School of Medicine, 550 First Ave, NBV 15 S7, New York, NY 10016-9196; email: Spiros.frangos@ nyumc.org. DOI: 10.1097/TA.0b013e3182a85f97
Morbidity and mortality remain high with 6 bicyclist deaths and 1,144 bicyclist injuries in New York County (i.e., Manhattan) in 2011.5 The efficacy of helmets in preventing or minimizing head injuries among bicyclists has been shown.6Y12 However, helmet use as it relates to bicyclists’ behaviors and hospital resource use following injury is less defined. Assessing the demographics and behaviors of bicyclists who wear helmets and comparing these variables to those who do not may provide insight into injury prevention strategies and targeted outreach. The objective of this study was to compare the demographics, behaviors, initial hospital evaluation, and outcomes of bicyclists based on their use of helmets. Our hypothesis was that helmeted bicyclists engage in safer riding practices.
PATIENTS AND METHODS This study was a subset analysis of a prospective cohort study, which evaluated vulnerable roadway users injured by motor vehicles in NYC.13 Data collection was performed at Bellevue Hospital Center, a NYC Level 1 trauma center, between December 2008 and June 2011. Bellevue Hospital Center’s catchment area includes midtown and lower Manhattan and
J Trauma Acute Care Surg Volume 75, Number 5
Copyright © 2013 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
877
J Trauma Acute Care Surg Volume 75, Number 5
Webman et al.
western Brooklyn, and its emergency department (ED) evaluates more than 100,000 patients annually. Prospective data had been collected on all bicyclists and pedestrians who presented within 24 hours of collisions involving motor vehicles. Patients requiring hospitalization and those discharged from the ED were included; both walk-ins and transfers were also included. All pedestrian data were excluded for this subset analysis. Pertinent bicyclist variables captured in the database included patient demographics, helmet use at time of injury, behaviors (e.g., riding with or against the flow of traffic), Glasgow Coma Scale (GCS) score at time of presentation, alcohol use before incident, initial computed tomography (CT) imaging studies, Abbreviated Injury Scale (AIS) score, Injury Severity Score (ISS), admission status, hospital length of stay (LOS), disposition, and mortality. Unknown or indeterminate helmet status was a criterion for exclusion. ED staff were informed of the study before it began. Informational fliers were hung throughout the ED. A pager number dedicated to the study was provided to ED staff, and the pager was activated whenever a patient meeting criteria presented. Notifications were accepted 24 hours a day and 365 days per year. Data collection was performed primarily by a dedicated study coordinator, a trauma coordinator, and attending physicians in trauma and emergency medicine. Verbal informed consent was obtained from all patients before study inclusion. Most data were obtained from patients/families and transcribed onto a six-page questionnaire by the interviewer. Emergency medical technicians, New York Police Department officers, and scene witnesses provided additional scene data when possible. Ambulance call reports were used to corroborate or verify data for most patients. Data collected on patients unable to grant verbal consent were used only if consent could be obtained at a later time. Findings from imaging studies were reviewed, and hospital LOS and disposition data were added soon after discharge. AIS score and ISS were calculated for each patient by a single trauma surgeon after attending radiology evaluations were finalized. A blood alcohol concentration (BAC) greater than 0.01 g/dL was used to determine whether patients had consumed alcohol before injury. BAC level measurements were performed as part of a routine workup in most cases. If objective laboratory data were not available, alcohol use was determined by patient self-report. If BAC results conflicted with the patient self-report, BAC took precedence. Data were analyzed using SPSS version 18 software (IBM). Pearson’s W2 and independent two-tailed t tests were performed, with p G 0.05 representing statistical significance. A univariable analysis was used to assess efficacy of helmet use with respect to the prevention of significant head injury. This study was approved by both the New York University School of Medicine and the Bellevue Hospital Center institutional review boards. This study was funded by a Highway Safety Grant from the State of New York Governor’s Traffic Safety Committee.
RESULTS There were 382 bicyclists. Of these, eight had unknown or indeterminate helmet status and were excluded from the analysis. Of the remaining 374 patients, 113 (30.2%) were 878
wearing helmets at the time of injury (Table 1). Twenty-five percent of women and 31.1% of men wore helmets ( p = 0.420). White bicyclists were more likely to wear helmets, while black bicyclists were less likely ( p = 0.037). Patients with private insurance were more likely to wear helmets, while those with either Medicaid or no insurance were less likely ( p = 0.027). There was no difference in age, primary language, or working status between the helmeted and nonhelmeted groups. Among those who were riding as part of employment (e.g., delivery workers, bike messengers) (n = 163), 52 (31.9%) were wearing helmets; 60 (28.8%) of the nonworking bicyclists (n = 208) were wearing helmets when injured ( p = 0.110). Behaviors of bicyclists are depicted in Table 2. Helmeted bicyclists were more likely to ride with the flow of traffic (97.2%) and to be within bike lanes when available (83.7%) at the time of the collision, as compared with 83.6% and 61.3% of nonhelmeted bicyclists ( p G 0.001 and p = 0.013, respectively). Nonhelmeted bicyclists were also more likely to be riding against traffic ( p = 0.003). There was no statistical difference with regard to alcohol use. Of the seven helmeted alcohol users, five had BACs drawn (all were 90.08 g/dL) and two selfreported use. Of the 29 nonhelmeted alcohol users, 20 had BACs drawn (13 were 90.08 g/dL, while 7 were 0.01Y0.08 g/dL) and 9 self-reported use.
TABLE 1. Patient Demographics, Insurance, and Work Status Categories Sex Male, n (%) Age, Y Age, mean (SD) Ethnicity/race, n (%) White Black Latino East Asian South Asian Other Language,* n (%) English Spanish Chinese Russian Bengali Other Insurance,** n (%) Medicaid Medicare Private insurance Worker’s compensation Uninsured Other
Helmet (n = 113) No Helmet (n = 261)
p
100 (88.5)
222 (85.1)
0.420
32.73 (12.41)
31.84 (13.23)
0.541
49 (43.4) 11 (9.7) 41 (36.3) 7 (6.2) 5 (4.4) 0 (0)
87 (33.3) 43 (16.5) 100 (38.3) 22 (8.4) 3 (1.1) 6 (2.3)
0.037
64 (56.6) 34 (30.1) 5 (4.4) 1 (0.9) 4 (3.5) 5 (4.4)
150 (57.7) 77 (29.6) 14 (6.9) 1 (0.4) 1 (0.4) 13 (5.0)
0.214
8 (8.8) 0 (0) 41 (45.1) 0 (0) 37 (40.7) 5 (5.5)
27 (14.1) 1 (0.5) 48 (25.1) 2 (1.0) 102 (53.4) 11 (5.8)
0.027
*One unknown removed. **n = 292 because 92 were unknown.
* 2013 Lippincott Williams & Wilkins
Copyright © 2013 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 75, Number 5
Webman et al.
TABLE 2. Bicyclist Behaviors Categories
TABLE 4. Initial CT Imaging Studies
Helmet (n = 113) No Helmet (n = 261)
Riding patterns, n (%) With flow of traffic 106 (97.2)* Against flow of traffic 3 (2.8)* Crossing against signal 6 (5.5)† Riding in bike lane 36 (83.7)§ Alcohol involvement, n (%) Yes 7 (6.3)||
p
204 (83.6)** 30 (12.3)** 30 (12.6)‡ 46 (61.3)¶
G0.001 0.003 0.057 0.013
29 (11.1)
0.181
*Four unknowns removed. **Seventeen unknowns removed. †Four unknowns removed. ‡Twenty-two unknowns removed. §n = 43 (patients were excluded because bike lane was unavailable or bike lane availability or use unknown). ¶n = 75 (patients were excluded because bike lane was unavailable or bike lane availability or use unknown). ||Two unknowns removed.
TABLE 3. GCS Score, AIS Score, and ISS Categories GCS score* GCS score, mean (SD) GCS e 8, n (%) GCS 9 8, n (%) AIS categories, n (%) Head and neck AIS score 1Y6 Face AIS score 1Y6 Chest AIS score 1Y6 Abdomen AIS score 1Y6 Extremity AIS score 1Y6 External AIS score 1Y6 ISS* ISS, mean (SD) No injuries, n (%) 1Y8, n (%) 9Y15, n (%) 16Y24, n (%) Q25, n (%) Head AIS only,** n (%) AIS score 0 AIS score 1 AIS score 2 AIS score 3 AIS score 4 AIS score 5 AIS score 6 Craniotomy Yes, n (%)
Helmet (n = 113)
No Helmet (n = 261)
14.93 (0.499) 0 (0) 111 (100)
14.79 (1.293) 3 (1.2) 257 (98.8)
0.285 0.557
6 (5.3) 5 (4.4) 5 (4.4) 6 (5.3) 27 (23.9) 92 (81.4)
21 (8.0) 23 (8.8) 14 (5.4) 12 (4.6) 64 (24.5) 216 (82.8)
0.394 0.198 0.803 0.795 1.000 0.769
3.18 (7.205) 20 (17.7) 83 (73.5) 7 (6.2) 1 (0.9) 2 (1.8)
4.13 (7.478) 34 (13.0) 187 (71.6) 23 (8.8) 7 (2.7) 10 (3.8)
0.255 0.388
107 (94.7) 0 (0) 0 (0) 2 (1.8) 0 (0) 1 (0.9) 0 (0)
240 (92.0) 0 (0) 1 (0.4) 5 (1.9) 8 (3.1) 1 (0.4) 0 (0)
0.360
1 (0.4)†
1.000
0 (0)
*Three unknowns excluded. **Excludes those with neck injuries. †Evacuation of epidural hematoma.
p
Categories
Helmet (n = 113)
CT scan, n (%) CT head CT cervical spine CT chest CT abdomen/pelvis
35 38 20 23
No Helmet (n = 261)
p
110 (42.1) 96 (36.8) 44 (16.9) 66 (25.3)
0.049 0.639 0.881 0.355
(31.0) (33.6) (17.7) (20.4)
There were no significant differences in mean GCS score, severe head injury (i.e., GCS score e 8), AIS categories (including head and neck AIS), mean ISS, or severity for category-based ISS between groups (Table 3). Head and neck AIS was further broken down to ‘‘head AIS only’’ (Table 3). There was no difference in ‘‘head AIS only’’ or rate of craniotomy between helmeted and nonhelmeted patients. A majority in both groups (94.7% helmeted, 92% nonhelmeted) had no intracranial injuries. Head CTs were more likely to be performed on patients who were nonhelmeted (42.1%) at the time of injury as opposed to their helmeted counterparts (p = 0.049) (Table 4). CT scans of the cervical spine, chest, abdomen/pelvis were as likely to be performed in one group as the other. Nonhelmeted bicyclists were more likely to be admitted to the hospital (p = 0.030) (Table 5). Mean LOS, disposition, and mortality rates were not statistically different between groups. Three nonhelmeted bicyclists died, including two of multisystem organ failure and one of hemorrhagic shock at time of presentation. There were no deaths in the helmeted group.
DISCUSSION Evidence for a relationship between helmet use and bicyclist behaviors, including obeying traffic laws and bicycle lane use, is limited. Farris et al.14 looked at certain behaviors including hand signaling and adherence to stop signs; they found that bicyclists who wore helmets were more likely to engage in these safe biking practices. Wasserman et al.15 investigated seat belt use in motor vehicles and found that buckling up correlated with helmet use while bicycling. In both studies, helmet use was associated with safer behaviors. Our study supports the notion that observance of traffic laws correlates with helmet use. Specifically, bicyclists who wore
TABLE 5. Patient Outcomes and Disposition Admissions and Hospital LOS Admitted, n (%) LOS, mean (SD), d Patient disposition, n (%) Home Rehabilitation Other Mortality Died, n (%)
17 (15.0) 0.96 (4.819)
66 (25.3) 1.98 (9.254)
0.030 0.268
109 (96.5) 3 (2.7) 1 (0.9)
240 (92.0) 12 (4.6) 9 (3.4)
0.240
0 (0)
3 (1.1)
0.555
* 2013 Lippincott Williams & Wilkins
Copyright © 2013 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
879
J Trauma Acute Care Surg Volume 75, Number 5
Webman et al.
helmets were more likely to ride with the flow of traffic and to be in designated bike lanes at time of injury. A negative correlation between alcohol and helmet use is well documented.16Y18 This variable did not achieve statistical significance in our study. Our results suggest that the riskier behaviors of nonhelmeted bicyclists extend beyond the use and influence of alcohol. NYC requires working bicyclists and children younger than 13 years to wear helmets.3 Despite this legislation, our study found no difference in helmet use between working bicyclists and those biking for recreation or commuting purposes. While other studies have shown that instituting helmet laws increased helmet use in children,19,20 our small pediatric sample size precludes us from making definitive conclusions for this cohort. Previous studies have concluded that children of parents of higher socioeconomic status (SES) are more likely to wear helmets.20Y22 Rodgers23 looked at helmet use among bicyclists of all ages using level of education as a proxy for SES and found greater levels of helmet use to correlate with higher levels of education. Our study suggests that patients of lower SES, inferred from a lack of health insurance or need for Medicaid applications, were less likely to wear helmets. Dellinger and Kresnow22 reported racial and ethnic differences among pediatric bicyclists who wore helmets; this demographics and its association with helmets have not been well defined among adults. Consistent with their results,22 our study found that, among ethnic groups, white bicyclists were more likely while black bicyclists were less likely to wear helmets. Our study found that nonhelmeted bicyclists were more likely to undergo head CT scans and had higher rates of hospital admissions as compared with helmeted bicyclists. This difference was apparent despite injury scores that were not statistically different between these groups. CT scanning is an expensive imaging modality and has associated radiation exposure risks. Likewise, hospital admissions including a mere one-night stay may incur costs in thousands of dollars. Given that the nonhelmeted cohort in our patient population was more likely to have government insurance or no insurance, a substantial portion of these health care costs get distributed across the general public. Crocker et al.18 reported a similar correlation in lower SES bicyclists who were impaired by alcohol. Many studies have shown the efficacy of helmet use in preventing head and facial injuries and in reducing mortality.6Y12 Because our methodology involved inclusion of both admitted and nonadmitted patients, the low number of traumatic head injuries and the rarity of deaths likely precluded this study from being powered to capture the protective effects of helmet use. The Manhattan study area is in an extremely dense urban core with low-to-moderate traffic speeds and mostly low-speed interactions between bicyclists and motor vehicles. The low mean ISSs among bicyclists, while expected given the relatively slow motor vehicle speeds, make injury comparisons between bicycling cohorts more challenging. This study has several strengths. First, it was conducted in a prospective manner, helping to minimize missing data points. Second, nonadmitted patients were included in the sample and not merely hospital admissions as most previous studies have done; in evaluating at-risk behaviors for injury, 880
artificially limiting the sample by excluding nonadmitted patients introduces a selection bias, which our study was designed to limit. Finally, information for most incidents was obtained from multiple sources including first responders when available, thereby introducing an additional, objective viewpoint. There are also limitations to this study. Because the accuracy of data relied heavily on patient self-report, it is possible that individual patients may have underreported certain behaviors for personal, cultural, economic, and/or legal reasons. Second, a minority of patients had poor or no recall of the incident secondary to alcohol use or traumatic head injury. Third, the sample size was not powered to investigate injury and mortality differences, which could have been informative and added to the other analyses. The low mean ISS in both groups also denotes a population that sustained predominantly mild-to-moderate trauma, leading to less pronounced differences in injuries and outcomes between cohorts. Fourth, patients declared dead at the scene and those taken to other local hospitals were not included. Fifth, this study was performed in a single institution, so its results are representative of the unique catchment area of Bellevue Hospital Center and are not generalizable across NYC as a whole. Sixth, these behaviors reflect those of bicyclists involved in crashes and may not be representative of bicyclists who avoid them. Finally, this study was conducted in an urban environment, and therefore, whether these data are applicable to suburban or rural areas is unclear and requires further investigation. Although most injured bicyclists do not wear them, helmets are indicators of basic safe riding practices. In this study, helmet use was associated with reduced hospital resource use based on less head CT imaging and lower likelihood of admission. In light of our national focus on preventative health and reducing health care costs, bicyclists who opt not to wear helmets may benefit from safe bicycling educational programs, while injured nonhelmeted riders should be targeted; especially in urban core areas, bicycling without a helmet represents more than merely a vulnerability to head injury in the event of a crash. AUTHORSHIP R.J.S., D.S., S.J., O.B., D.A.L., and S.G.F. contributed to the study conception and design. R.W., L.A.D., R.J.S., S.R.T., D.S., S.J., O.B., S.W., C.W., D.A.L., and S.G.F. performed the acquisition of data. R.W., L.A.D., S.R.T., D.S., S.W., M.R., H.L.P., and S.G.F. performed the analysis and interpretation of data. R.W., L.A.D., and S.G.F. drafted the manuscript. R.W., L.A.D., R.J.S., S.R.T., D.S., S.J., S.W., O.B., C.W., D.A.L., M.R., H.L.P., and S.G.F. provided critical revision.
DISCLOSURE This study was funded by a Highway Safety Grant from the State of New York Governor’s Traffic Safety Committee (October 1, 2008, to September 30, 2011; Year 1, $122,242; Year 2, $129,748; Year 3, $130,670). D.S., S.J., S.G.F., and D.A.L. received salary support from this grant.
REFERENCES 1. US Department of Transportation, National Highway Safety Administration. Traffic Safety Facts. (2012) Available at: www-nrd.nhtsa.dot.gov/ Pubs/811606.pdf. Accessed January 28, 2013.
* 2013 Lippincott Williams & Wilkins
Copyright © 2013 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 75, Number 5
2. Nicaj L, Stayton C, Mandel-Ricci J, McCarthy P, Grasso K, Woloch D, Kerker B. Bicyclist fatalities in New York City: 1996Y2005. Traffic Inj Prev. 2009;10:157Y161. 3. New York City Administrative Code: N.Y.C. Admin. Code §§ 10Y157. (bicycles used for commercial purposes). 4. New York City. Department of Transportation. Bureau of Traffic Operations. NYCDOT-DOT Announces 35% Increase in Commuter Cycling from 2007 to 2008 and Calls on Cyclists to Use Lights to Be Seen & Safe. (2008) Available at: http://www.nyc.gov/html/dot//html/pr2008/pr08_047.shtml. Accessed January 28, 2013. 5. Governor’s Traffic Safety Committee, SafeNY. County Data Reports. (2011) Available at: www.safeny.ny.gov/hsdata.htm. Accessed January 28, 2013. 6. Amoros E, Chiron M, Martin JL, Thelot B, Laumon B. Bicycle helmet wearing and the risk of head, face, and neck injury: a French case-control study based on a road trauma registry. Inj Prev. 2012;18:27Y32. 7. Heng KWJ, Lee AHP, Zhu S, Tham KY, Seow E. Helmet use and bicyclerelated trauma in patients presenting to an acute hospital in Singapore. Singapore Med J. 2006;47:367Y372. 8. Persaud N, Coleman E, Zwolakowski D, Lauwers B, Cass D. Nonuse of bicycle helmets and risk of fatal head injury: a proportional mortality, casecontrol study. CMAJ. 2012;184:E921YE923. 9. Thompson RS, Rivara FP, Thompson DC. A case-control study of the effectiveness of bicycle safety helmets. N Engl J Med. 1989;320: 1361Y1367. 10. Thompson DC, Thompson RS, Rivara FP, Wolf ME. A case-control study of the effectiveness of bicycle safety helmets in preventing facial injury. Am J Public Health. 1990;80:1471Y1474. 11. Thompson DC, Rivara F, Thompson R. Helmets for preventing head and facial injuries in bicyclists. Cochrane Database Syst Rev. 1999;4: CD001855. 12. Wasserman RC, Buccini RV. Helmet protection from head injuries among recreational bicyclists. Am J Sports Med. 1990;18:96Y97.
Webman et al.
13. Dultz LA, Foltin G, Simon R, Wall S, Levine DA, Bholat O, SlaughterLarkem D, Jacko S, Marr M, Glass N, et al. Vulnerable roadway users struck by motor vehicles at the center of the safest, large US city. J Trauma Acute Care Surg. 2013;74(4):1138Y1145. 14. Farris C, Spaite DW, Criss EA, Valenzuela TD, Meislin HW. Observational evaluation of compliance with traffic regulations among helmeted and nonhelmeted bicyclists. Ann Emerg Med. 1997;29:625Y629. 15. Wasserman RC, Waller JA, Monty MJ, Emery AB, Robinson DR. Bicyclists, helmets and head injuries: a rider-bases study of helmet use and effectiveness. Am J Public Health. 1988;78:1220Y1221. 16. Li G, Baker SP, Smialek JF, Soderstrom CA. Use of alcohol as a risk factor for bicycling injury. JAMA. 2001;285:893Y896. 17. Spaite DW, Criss EA, Weist DJ, Valenzuela TD, Judkins D, Meislin HW. A prospective investigation of the impact of alcohol consumption on helmet use, injury severity, medical resource utilization, and healthcare costs in bicycle-related trauma. J Trauma. 1995;38:287Y290. 18. Crocker P, Zad O, Milling T, Lawson KA. Alcohol, bicycling, and head and brain injury: a study of impaired cyclists’ riding patterns R1. Am J Emerg Med. 2010;28:68Y72. 19. Macpherson AK, To TM, Parkin PC, Moldofsky B, Wright JG, Chipman ML, Macarthur C. Urban/rural variation in children’s bicycle-related injuries. Accid Anal Prev. 2004;36:649Y654. 20. Parkin PC, Khambalia A, Kmet L, Macarthur C. Influence of socioeconomic status on the effectiveness of bicycle helmet legislation for children: a prospective observational study. Pediatrics. 2003;112:e192Ye196. 21. Lang IA. Demographic, socioeconomic, and attitudinal associations with children’s cycle-helmet use in the absence of legislation. Inj Prev. 2007;13: 355Y358. 22. Dellinger AM, Kresnow M. Bicycle helmet use among children in the United States: the effects of legislation, personal and household factors. J Saf Res. 2010;41:375Y380. 23. Rodgers GB. Bicycle helmet use patterns in the United States: a description and analysis of national survey data. Accid Anal Prev. 1995;27:43Y56.
* 2013 Lippincott Williams & Wilkins
Copyright © 2013 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
881
