Impact of Helmet Use on Injury and Financial Burden of Motorcycle and Moped Crashes in Hawai‘i: Analysis of a Linked Statewide Database (2024)

Emergency Medical Service Reports

This study is based on patient care reports (PCRs) from Hawai‘i emergency medical service (EMS) providers from 2007 to 2009. All Hawai‘i EMS providers complete PCRs electronically with standardized input screens. Discrete fields describe vehicle type (motorcycle or moped), helmet use, demographic variables, patient disposition, and a free-text space to record a narrative of the incident. All 2,923 EMS records for injured motorcyclists or moped riders over the 3-year period were reviewed. A total of 287 records were excluded from further analysis because they were related to off-road or non-traffic crashes (n = 150), riders of all-terrain vehicles (n = 77), or documented a patient transfer to another EMS unit (n = 60). Another 83 records were excluded due to unknown helmet status, resulting in records from 2,553 riders, 75% of whom (1,913) were transported to hospitals by EMS for further medical attention.

Linkage to Hospital Records

The Hawai‘i Health Information Corporation (HHIC) linked EMS records to hospital records for both emergency department (ED) visits and hospitalizations. HHIC is a non-profit organization which maintains a repository of billing data abstracts for the medical records of all acute care hospitals in the state, with the exception of the ED records of one facility on the island of O‘ahu.¹⁵ Linkage of EMS records to HHIC records was accomplished deterministically, using patient name, date of birth, date of crash, and location of crash. Since HHIC maintains a master patient identifier, the definitive patient disposition and cumulative medical charges were available for patients who were transferred between hospitals in the state. Charges include those for room and board, pharmacy, laboratory, X-ray and hospital-based physician charges. Some analyses utilize a threshold of $25,000 in medical charges, to indicate treatments resulting in relatively high medical charges; $25,000 represented the 77th percentile of the distribution of medical charges. The principal source of payment was categorized into groups of interest. Payment from automobile insurance plans was considered “private insurance.”

Hospital records were linked to 1,783 (93.2%) of the 1,913 transported by EMS to hospitals, leaving 130 records that could not be linked to HHIC data. These 130 patients were significantly younger than the 1,783 linked to HHIC data (33.3 versus 36.2 years, respectively, P = .026), and there was a higher prevalence of helmet use among motorcyclists (67.4% versus 50.0%, respectively, P = .002. Pre-hospital condition, as characterized by EMS personnel, was generally better among the 130 unlinked riders. Only 1.5% were in “critical” condition (indicating unstable vital signs and level of consciousness), compared to 7.8% of the other riders (P = .01). Patient condition was more likely to be “serious” (stable vital signs but impending deterioration) among the latter (60.9%), compared to the 130 unlinked riders (41.5%, P < .001). Data from these 130 riders were excluded from further analyses. Hospital records were available for another 182 riders who refused transport by EMS, but made their own way to hospitals, resulting in a final study sample size of 1,965 records.

HHIC data was used to define injury categories, using the principal diagnosis by ICD9-CM code: fractures (ICD9-CM code series 800–829), skull fractures (800–804), dislocations (830–839), sprains and strains (840–848), internal injuries (850–869), intracranial injury without skull fracture (850–854), open wounds (870–897), contusions and superficial injuries (910-924), as well as other injuries and non-injuries (ICD9-CM code less than 800). Traumatic brain injury (TBI) and spinal cord injuries were identified using the code series recommended by the CDC.¹⁶ A category of “head injury” was also constructed, using the Barell Matrix.¹⁷ HHIC data identified 62 patients who eventually died from their injuries, and all deaths were confirmed through linkage with the Hawai‘i Department of Health's death certificate database.

Statistical Analysis

All statistical analyses for this report were performed using JMP software, version 5.¹⁸ Differences in the distributions of categorical variables were tested using differences in proportions, while differences in the means of continuous variables were assessed by 2-tailed t-tests. Differences described as “significant” had an associated P value of less than .05. For multivariate logistic regression analyses, rider age was categorized into 4 groups: those 25 years and younger (referent group), 26 to 35 years, 36 to 54 years, and 55 years and older. Helmet status was dichotomized, with helmeted riders as the referent group. Our previous study¹ showed the associations between helmet use and injury outcomes varied considerably between motorcycle and moped riders, so all of the analyses included stratification by vehicle type. Study procedures were approved by the Hawai‘i Department of Health's Institutional Review Board.

Demographics

The mean age of the 1,965 patients was 35.9 years (SD 14.3 years) (Table 1). Helmeted motorcyclists were younger than unhelmeted motorcyclists on average (34.8 vs 39.0 years, P < .001), while helmeted moped riders were older than unhelmeted moped riders (average 37.9 vs 34.2 years, P = .017). Motorcycle riders were more likely to be male (87.2%) compared to moped riders (70.0%, P < .001). Overall, helmet use was significantly less common among female riders (25.6%) compared to male riders (34.6%, P = .001). However, this association differed across vehicle types. In moped riders, helmet usage in females was significantly higher than in males (15.2% vs 9.8%, P = .027), but was comparable among motorcyclists (45.3% for females vs. 50.7% for males). About two-thirds of all riders were not wearing a helmet at the time of the crash. This proportion was significantly higher among moped riders (88.6%) than motorcyclists (50%, P < .001).

Motorcyclists were more likely to be Hawai‘i residents (89.6%) compared to moped riders (77.8%, P < .001). Hawai‘i residents were more likely to wear helmets (33.8%) compared to non-residents (27.4%, P = .027).

Insurance

Overall, the most common principal payer was private insurance (60.8% of patients), followed by self-pay (16.2%), income-qualifying insurance (such as Medicaid, QUEST) (10.7%), and Department of Defense (8.0%). The proportion of riders who did not utilize an insurance payer (ie, self-pay) was nearly double among unhelmeted riders (19.3%), compared to helmeted riders (9.8%, P < .001). The latter were also significantly less likely to have income qualifying insurance as their principle source of payment (5.0%), compared to unhelmeted riders (13.5%, P < .001), and more likely to have policies under the Department of Defense (17.7% versus 3.35%, respectively, P < .001). These differences in payer mix between helmeted and unhelmeted riders were also evident within the strata of motorcycle and moped riders.

Disposition and Pattern of Injury

Moped riders were more likely to be discharged from the ED, whereas motorcyclists were significantly more likely to be hospitalized (40.7% vs 28.4%, P < .001) or die (4.7% vs 1.3%, P < .001). There were no significant differences in the distribution of final medical disposition between helmeted and unhelmeted moped riders (Table 2). Among motorcyclists, however, unhelmeted riders were significantly more likely to be admitted to the hospital (44.1%), compared to helmeted motorcyclists (37.3%, P = .02), or to ultimately die (7.0% versus 2.4%, respectively, P = .001).

When principal diagnoses were reviewed, the incidence of fractures was statistically comparable between helmeted and unhelmeted riders of both vehicle types. However, the incidence of skull fractures was nearly 3 times greater for unhelmeted moped riders, and more than 6 times greater for unhelmeted motorcyclists.

When all diagnoses were considered, the risk for any type of head injury was significantly higher for unhelmeted riders of either type of vehicle. Approximately half of the unhelmeted motorcycle (53.5%) or moped riders (50.5%) sustained a head injury, compared to 35.0% of the helmeted moped riders and 23.4% of the helmeted motorcycle riders. The incidence of traumatic brain injury (TBI), which includes the spectrum from concussion to intracranial hemorrhage, was also significantly higher among unhelmeted riders of either type of vehicle. For moped riders, the incidence of TBI among unhelmeted riders was 60% higher (35.1% versus 22.0% for helmeted riders, P = .009). For motorcyclists, the incidence was nearly double (41.4% versus 21.1% for helmeted riders, P < .001).

Charges

Average medical charges were approximately 50% higher for unhelmeted riders of both vehicle types (Table 3). More than one-third (35.1%) of medical charges to unhelmeted motorcyclists were paid by income-qualifying plans or self-pay, compared to only 13.7% of the charges to helmeted motorcyclists. Medical charges among unhelmeted moped riders were also more likely to be paid by income-qualifying plans or self-pay, compared to helmeted riders (37.9% vs 15.2%, respectively).

Approximately one-quarter (23.3%) of the 1,965 riders had medical charges of $25,000 or more, although this proportion was significantly higher among motorcyclists (28.5%) than moped riders (16.9%, P < .001). Unhelmeted motorcyclists were almost 50%, and unhelmeted moped riders almost 40%, more likely to incur $25,000 or more in total charges compared to helmeted riders (Table 3). Unhelmeted riders were about 3 times as likely to have an income-qualifying plan as the principal source of payment for these higher charge totals. They were also more likely to self-pay. As a result, unhelmeted riders in both motorcycles and mopeds had more than double the incidence of accruing $25,000 or more in total charges, combined with an income-qualifying payer plan or self-pay status.

Adjusted Outcomes

After controlling for age and gender by multivariate logistic regression, unhelmeted motorcycle riders had significantly higher odds of hospital admission (1.54, P = .001), and nearly three times the odds of a fatal injury (2.85, P = .001), compared to helmeted riders (Table 4). There were no significant differences in either hospital admission or risk of fatal injury among moped riders.

Unhelmeted riders had significantly higher odds of TBI (2.49, P < .001), skull fracture (4.48, P < .001), or any head injury (3.18, P < .001), compared to helmeted riders, independent of age or gender. The disparity in outcomes was greater among motorcyclists, although unhelmeted moped riders also had significantly higher odds of TBI or head injury, compared to helmeted moped riders. Similar associations were found if only the principal diagnosis was considered.

The odds of incurring $25,000 or more in total charges were nearly 50% higher among unhelmeted motorcyclists, compared to helmeted riders. The former were also about 3 times as likely to have an income-qualifying plan as principal source of payment and twice as likely to self-pay. As a result, unhelmeted motorcyclists had more than twice the odds (2.82, P < .001) of accruing $25,000 or more in total charges, combined with an income-qualifying plan or self-pay as the principal source of payment.

Limitations

Limitations of this study are the inclusion of only injured riders who were attended to by EMS personnel, and the limited sample size of moped riders. Hospital records could not be located for approximately 7% of the injured riders who were transported to hospitals by EMS. However, the exclusion of these riders probably did not bias the results of multivariate logistic regression models, since these riders were more likely to have worn helmets and had a more favorable distribution of pre-hospital injury severity. Another limitation is the availability of only acute care charges in the initial hospitalization. Therefore, medical charges we report may underestimate the true financial burden of long-term injuries. It should be noted that we only had access to medical charges, and not the actual medical costs that were ultimately paid after insurance adjustments. Finally, diagnostic data may be less accurate than the original medical records because HHIC data systems is generated for billing reimbursem*nt.¹

Strengths

This study includes statewide data collection which was possible through uniform data collection for all EMS providers in Hawai‘i, and the existence of HHIC, a central repository of billing data from hospital medical records in the state. The capability of HHIC to track individuals transferred across medical facilities improved the ascertainment of final medical dispositions and reduced loss to follow-up. Finally, this study had a comprehensive comparison of demographics, injury, mortality, cost, and insurance information in four different groups: motorcycle riders with and without helmets, and moped riders with and without helmets. This is the first study with detailed analysis of helmet use between the two vehicle types.

Future Impact

Our study shows that unhelmeted riders are much more likely to suffer serious injury or death and to incur higher hospital charges. Furthermore, these higher hospital charges are more likely to be absorbed by public funds or the hospitals that provide them medical treatment. This is true for both motorcycle and moped riders, a distinction that was previously unstudied. These are important findings for those advocating for legislative policies requiring helmet use. Our study supports that implementing mandatory helmet laws for both motorcycle and moped riders would decrease severity of injury, medical charges, and cost to society.

1. Galanis DJ, Ly CL, Wong LL, Steinemann S, Rosen L. Helmet Use Among Motorcycle and Moped Riders Injured in Hawaii: Final Medical Dispositions from a Linked Database. J Trauma. 2014;77(5):743–748. [PubMed] [Google Scholar]

2. Brown CV, Hejl K, Bui E, Tips G, Coopwood B. Risk Factors for Riding and Crashing a Motorcycle Unhelmeted. J Emerg Med. 2011;41(4):441–446. [PubMed] [Google Scholar]

3. Heldt KA, Renner CH, Boarini DJ, Swegle JR. Costs Associated with Helmet Use in Motorcycle Crashes: The Cost of Not Wearing a Helmet. Traffic Inj Prev. 2012;13(2):144–149. [PubMed] [Google Scholar]

4. Brandt MM, Ahrns KS, Corpron CA, Franklin GA, Wahl WL. Hospital Cost Is Reduced by Motorcycle Helmet Use. J Trauma. 2002;53(3):469–471. [PubMed] [Google Scholar]

5. O'Keefe T, Dearwater SR, Gentilello LM, Cohen TM, Wilkinson JD, McKenney MM. Increased fatalities after motorcycle helmet law repeal: is it all because of lack of helmets? J Trauma. 2007;63:1006–1009. [PubMed] [Google Scholar]

6. Croce MA, Zarzaur BL, Magnotti LJ, Fabian TC. Impact of motorcycle helmets and state laws on a society's burden: a national study. Ann Surg. 2009;250:390–394. [PubMed] [Google Scholar]

7. Abbas AK, Hefny AF, Abu-Zidan FM. Does wearing helmets reduce motorcycle-related death? A global evaluation. Accid Anal Prev. 2012;49:249–252. [PubMed] [Google Scholar]

8. MacLeod JB, DiGiacomo CJ, Tinkoff G. An evidence-based review: helmet efficacy to reduce head injury and mortality in motorcycle crashes: EAST practice management guidelines. J Trauma. 2010;69:1101–1111. [PubMed] [Google Scholar]

9. McSwain NE, Jr, Belles A. Motocycle Helmets-Medical Costs and the Law. J Trauma. 1990;30(10):1197–1199. [PubMed] [Google Scholar]

11. Christmas AB, Brintzenhoff RA, Schmeizer TM, Head KE, Sing RF. MOPEDS-Motorized Objects Propelling Ethanol Drinking Subjects. Am Surg. 2011;77(3):304–306. [PubMed] [Google Scholar]

12. Brintzenhoff RA, Christmas AB, Braxton VG, Janulis KE, Huynh TT, Sing RF. Mopeds: The Legal Loophole for Repeat Driving While Intoxicated Offenders. Am Surg. 2011;202(6):697–700. [PubMed] [Google Scholar]

13. Blackman RA, Haworth NL. Comparison of Moped, Scooter, and Motorcycle Crash Risk and Crash Severity. Accid Anal Prev. 2013;57:1–9. [PubMed] [Google Scholar]

14. Matzsch T. Injuries in Moped and Motorcycle Accidents, a Five-Year Series. Laekartidningen. 1983;80(24):2514–2517. [PubMed] [Google Scholar]

15. Hawaii Health Information Corporation (HHIC), author About. [March 9, 2015]. Available at: http://hhic.org/

17. Centers for Disease Control and Prevention (CDC), author The Barell Injury Diagnosis Matrix, Classification by Body Region and Nature of the Injury. [January 20, 2014]. Available at: http://www.cdc.gov/nchs/data/ice/final_matrix_post_ice.pdf.

19. Matzsch T, Karlsson B. Moped and Motorcycle Accidents-Similarities and Discrepancies. J Trauma. 1986;26(6):538–543. [PubMed] [Google Scholar]

20. Miggins M, Lottenberg L, Liu H, Moldawer L, Efron P, Ang D. Mopeds and Scooters: Crash Outcomes in a High Traffic State. J Trauma. 2011;71(1):217–222. [PMC free article] [PubMed] [Google Scholar]

21. Braddock M, Schwartz R, Lapidus G, Banco L, Jacobs L. A Population-Based Study of Motorcycle Injury and Costs. Annals of Emergency Medicine. 1992;21(3):273–278. [PubMed] [Google Scholar]

22. Orsaw E, Holden JA, Williams J, Lumpkin JR. Motorcycle Trauma in the State of Illinois: Analysis of the Illinois Department of Public Health Trauma Registry. Ann Emerg Med. 1995;26(4):455–460. [PubMed] [Google Scholar]

23. Max W, Stark B, Root S. Putting a Lid on Injury Costs: The Economic Impact of the California Motorcycle Helmet Law. J Trauma. 1998;45(3):550–556. [PubMed] [Google Scholar]

24. Shankar BS, Ramzy AI, Soderstrom CA, Dischinger PC, Clark CC. Helmet Use, Patterns of Injury, Medical Outcome, and Costs Among Motorcycle Drivers in Maryland. Accident Analysis & Prevention. 1992;24(4):385–396. [PubMed] [Google Scholar]