ORIGINAL ARTICLE Ten Years of All-Terrain Vehicle Injury, Mortality, and Healthcare Costs Emily Marie Krauss, MSc, MD, Dlanne M, Dyer, BN, MN, Kevin B. Laupland, MSc, MD, and Richard Buckley, MD Background: All-terrain vehicles (ATVs) are increasing in popularity worldwide. The province of Alberta accounts for 25% of Canadian ATV sales. This study describes UK epidemiology, outcome;, and associated healthcare costs for a decade of ATV traumatic injury incidents. Methods This is a retrospective population based cohort study using wo provincial databases; the Alberta Trauma Registry and the Officeof the Chief Medical Examiner of Alberta. Data for individuals agud IS years or older with Injury Severity Score 12or deaths benveen April 1,1998, and March 31, 2003 were included. Healthcare costs were extrapolated using figures from a Level I trauma center. Results ATV incidents resulted in 459 serious trauma cases, 395 trauma center admissions (atotal of4,117days), and a 17%morlality rate. Postdisdiarge earn was required for nearly 30% of patients. Mala patients aged IS years to ISycars had the highest incidence (6.5 of 100,000 people). Head, neck, and cervical spine injuries were most common (59%) and predictive of mortality {relative risk[RR], 2.19; interquartile range [IQR],1.35D.54;p 0.001). Vehicle rollovers (RR, 2.75; 1QR, 1.13CC.70; p 0.01), vehicle ejection (RR, 4.1S;IQR, 1.70010.32; p 0.000), alcohol intake (RR, 2.33; 1QR, 1.52n,5(i; p 0.000), helmet use (RR. 1.82; IQR, 1.1103.02) p 0.01), and incident location were predictive of mortality. Conduac/ia Increasing rates of ATV-related serious trauma and death arc described in young males riding without helmets after consuming alcohol. Serious injuries contributed to healthcare costs in exciss of S6.5 million USD. Prodi cro rs of mortality include rider behaviors and mechanical factors, Prevention should include rider education and industry measures to improve ATV stability. Key Words All-terrain vehicle. Off-road highway vehicle, Trauma. Wounds and injuries, Mortality. (J Trauma. 20IO;69; I338D1343) A ll-terrain vehicles (ATVs), defined as four-wheeled offhighway motor vehicles, have increased in popularity in many countries worldwide in the recent years. Associated Submitted forpubltcationAiigustlO,2010, Accepted for pub It cation September 14,2010, Copyright(C) 2010 by Lippincott Williams & Wdkins From the Faculty of Medians [E.M.K.J.Um'versify of Calgary, Calgary, Alberta, Canada; Rockyvlcw General Hospital (D.M.D.),Oilg[iry, Alberta, Canada; Department of Critical CarcMulicine (K.B.L.)and Di vision of Orthopedics (R.B.), Foothills Medical Centre, University of Calgary, Calgary, Alberta, Canada. Presented at the Annual Scientific Musing of the Trauma Association of Canada. May 6D7, 2010, Halifax, Nova Scoria, Canada, and as an abstract and presentation at the Twenty-EighthAnnual Calgary Surgeons' Day, June 25, 2010. Address for reprints: Richard Buckley, MD, FRCSC, Divisionof Orthopedics, Foothills Medical Centre, 1403-29th Street NW, Calgary, Alberta, Canada T2N4NI; entail: buckclin@ucalgary.ca. D01:!0.1097/rA.ObOia3181fc5c7b 1338 with their increased use has been several reports of severe injury.^ Many reports have illustrated differing pictures of the ATV trauma population and their injuries, including data on rider populations on a closed-circuit recreational course with an enforced helmet rule1 to rural Australian farmers using ATVs primarily for work.4 These findings are in contrast to the extensive literature, primarily from the United States, illustrating a young male population with injuries sustained during recreational use; suffering commonly head and neck injuries with poor rates of helmet use. The literature from the United States has also emphasized high rates of alcohol intoxication and its contribution to ATV crashes.^ Helmkarap ct al.? reported that ATV trauma in the United States resulted in national hospitalizarion costs exceeding Sl.l billion from 2000 to 2004, a significant burden of injury. Most studies on ATV trauma have included injured persons who presented to a major trauma centcrscu or who survived injury.15 By restricting studies to patients presented to trauma center, the most severely injured people were excluded and significant predictors of severe Injury or mortality may be missed. Only one study outlined adult ATA' mortality including deaths at the scene but did not characterize multisystem injury, concentrating only on the single injury most likely to cause death.5 A study including trauma center admissions and deaths at the scene may provide incident and injury features that are predictors of mortality. In Canada, the province of Alberta leads the country in ATV sales and use, with an increase in yearly sales of 50% over the last 8 years, accounting for 25% of all ATVs sold nationally.'' Alberta Traffic Collision Statistics illustrate a nearly threefold increase in ATV collisions since 2001." In recent years, a larger number of patients with severe injuries from ATV use have been admitted to the trauma service at a Level I trauma center in southern Alberta. Despite the increase in popularity of ATVs and theincrease in incidence of ATV collisions, information in the trauma literature on Canadian ATV-related injury including mortality is limited to pediatric populations. 15 - 19 The limitations of the worldwide and Canadian literature mean that despite increasing uscofATVs, a clear picture of the extent of adult ATV trauma and mortality, its contributing factors, and associated healthcare costs is not available to Inform public policy-makers. Therefore, the objective of this study was to define the adult incidence and mortality rates, describe the injuries and costs associated with major ATV injuries, and assess changes over a 10-year period in the province of Alberta, Canada. TTieJxjrna/ofTRAUMA(R) lijury, Infect/en, and CriticsJ Care o Volume 69, Number 6, Dumber 2010 Copyright (C) Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. Ihe Journal of TRAUMA 8 Injury, Infection, and Critical Care o Volume 6'), Number 6, December 2010 ATV Injury bbrta!ify and Healthcare Costs MATERIALS AND METHODS This study used a population-based retrospective cohort design. Ethics approval was obtained from the Conjoint Health Research Ethics Board at University of Calgary, All adults (aged 18 years) suffering major trauma or death directly caused by a four-wheeled ATV in the Province of Alberta from April 1, 1998, to March 31, 200S, were included. Study subjects were identified, and data were obtained using two independent provincial databases: the Alberta Trauma Registry and the Office of the Chief Medical Examiner (ME) database. The Alberta Trauma Registry database registers all patients aged 18 years who suffer trauma with an Injury Severity Score (ISS) 12 and present to one of the three major Level I and Level II trauma centers in Alberta. The database of the Office of the Chief ME of Alberta was used for data on all individuals aged 18 years who died before presenting to a major trauma center. After identification of cases through the source databases, specific-data were abstracted. This abstraction included detailed demographic data (age and gender), injury geographic location (urban or rural defined as areas with 50,000 population), terrain or environment descriptors including streets orfields, mechanism of collision, ATV use for work or recreation, injury demographics (number of fatalities and number of injured persons), injury type (blunt or penetrating trauma), and injury date and time. Whether the patient was a driver or passenger on the ATV, alcohol levels if measured, and helmet use were also collected. Injury severity among hospitalized patients was assessed using the ISS. Injuries were grouped into six body systems (head and neck, face, chest, abdominal and pelvic, extremities, and external) according to the Maximum Abbreviated Injury' Scale classification system for ease of analysis.20 The Alberta Trauma Registry included the ISS coding; however, the data from the Office of the Chief ME could not be coded by ISS because of the limitations of the data descriptors. The ME data were grouped into body system categories. Hospital stay information included duration of hospital stay, intensive care unit (ICU) admissions and duration of stay, number of operating room procedures, and discharge destination (home, hospital, or rehabilitation center). All statistical analyses were performed using Stata statistical software version 11 (Stata, College Station, TX). For description, continuous variables were expressed as median with interquartile ranges (IQRs) when nonnormally distributed and were compared using the Mann-Whitney U test. Categorical data were compared using the Fisher's exact t test. Incidence rates were calculated using population data from the Alberta Health Registry. Overall all-cause casefatality rates were calculated by dividing the number of deaths by the total number of ATV trauma cases. Estimated provincial hospital costs for ATV-relatcd trauma patients were calculated using figures from one of the study's Level I trauma centers, Foothills Medical Centre, for the years 2002 to 2006. Foothills Medical Centre patients admitted for ATV-related injuries between 2002 and 2006 were considered a representative sample of the provincial ATV trauma population. Individual patient costs were ob(C) 2010 UpprcoK Wll'iarrs & Wlh'ns tained by matching their trauma number to the former Calgary Health Region corporate database. The average daily cost of inhospital care was then calculated and applied to the overall study cohort. For all comparisons, a p value of 0.05 was determined to represent statistical significance. Costs included both direct and indirect costs of nursing, laboratory, diagnostic, surgical, surgical supplies,medications, and support staff.21 The costs associated with repeat hospitaladmissionsor the provision ofhealthcare services after discharge from a Level I trauma center acute care settingwere not assessed. RESULTS Population During the 10-year studyperiod, atotal of459cases of severe trauma involving ATVs were identified: 395 (86%) through the Alberta Trauma Registry and 64 by the Office of the Chief ME of Alberta. Eighty-seven percent (n 401) were male, and the median age for the entire study population was 36 years (IQR, 24D50 years). The median ISS (n 395) was 19 (IQR, 16E5). Of the459 cases, there were 79 deaths either prehospital or before discharge for an all cause casefatality rate of 17%. Incidence The overall annual incidence of ATV associated severe trauma was 2.0 per 100,000 population. Overall incidence of severe ATV trauma increased yearly during the study period, and this was attributable to dramatic increases in the number of male victims as shown in Figure 1. Overall, male subjects were at a much higher risk than females (3.5 vs. 0.5 per 100,000; incidence rate ratio, 7.1; 95% confidence interval, 5.3C9.5; p 0.0001) with this risk differential between the genders most pronounced in the more recent years of the study (Fig. 1). The incidence ofsevere ATV trauma incidents decreased with advancing age (Fig. 2). Event Description Events resulting in severe injury or mortality involved ATV drivers (94%, 427 of 459), riding for recreational purposes (92%, 423 of 459) in rural areas (87%, 363 of 417). D Ft(TM)Hi DToul Rgure 1. Yearly incidence of mq'or ATV injury in the province of /yberta, Canada 1339 Copyright (C) Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. Krauss d a!. The Jwnal (/TRAUMA(R) Injury, Infection, and Critical Care o Volume 69, Number 6, December 2010 aoSO - o Ualii aFenuMi DToBI 18-19 10-34 J5.2B JIMS 1D-49 SO-S9 60-68 70-TB BO- *SI orooD <>>H Rgure 2. ?ge- and gender-related trajms in Aberta, Canada incidence of 3^ere ATV These Incidents commonly occurred on weekends (54%, 249 of 459) in the summer months of June through August (41%, 187 of 459). Helmet use was reported in only 35% (144 of 408) ofthosc where information was recorded. The most common mechanism of injury was blunt trauma (98%, 448 of 459) sustained in a vehicle rollover (7S%, 307 of 395) or when ejected from the vehicle (67%, 261 of 391). The most common locations for an ATV injury incident were the street or highway (18%, 81 of 459) and defined recreational areas (12%, 57 of 459). ATV Injury incidents frequently occurred on farms (9%, 39 of459). Positive blood alcohol levels were reported in 45% (n 132) of 292 available cases. Although, in the majorilyof cases, alcohol levels were not tested or available, positive alcohol levels were significamlyassodated with ATV incidcmson the street or highway (39 of 132, 30%, vs. 27 of 160, 17%; p 0.01) and with reduced helmet use (23 of 122, 19%, vs. 62 of 142,44%; p 0.000). Posltiveblood alcohol levels were also associated with recreational use of ATVs compared with ATV use for work (129 of 132,98%, vs. 146of 160,91%; p 0.02). Because of the high incidence of ATV trauma in persons younger than 30 years, we chose to compare event factors between 18-year and 30-year olds and those older than 30 years (Table 1). Both groups showed comparable rates of weekend recreational use in rural areas and had similar rates of helmet use. People younger than 30 years were more likely to have been Injured as passengers on an ATV. Collisions with stationary objects were more common among riders younger than 30 years and may suggest a lack of familiarity with the riding environment or with the operation of the ATV. Rollover single-vehicle collisions were more commonly reported for people aged 30 or older as were ATV incidents on farms. Although only approaching significance, TABLE 1. Descriptive Favors of ATV &ents by Totd Study Population and Age Gregory Factor Total audy Population, n (%) AgaOOyr, Rate(%) Mem ISS (IQR) Malt Rural Urban Out of province Weekend Driver Passuigcr Other rider Recriouonal use Wisiring helmet Alcohol inlakc Rollover Ejtciid Stationary collision Moving collision Lociiion street Recrisitional arta Location industry Location mine Location farm Location home Winter {DeconbcrCFebrua ry) Spring (March CM ay) Summer (JuneD\t]gust) Autumn (SeptembiTQJovtnibLT) 19(16C25) 401/459(87.4) 363/417(87.1) 16/471 (3.4) 38/471(8.1) 249/459 (54.3) 427/450(93.6) 25/456 (5.5) 4/456 (0.9) 423/459 (92.2) 144/408 (35.3); 144/459 (31.4) 132/292 (45.2); 132/459 (28,8) 307/395 (77.7) 261/391 (66.S) 61/395(15.4) 26/395 (6.6) 81/459(17.7) 57/459(12.4) 9/459 (2.0) 4/459 (0.9) 39/459 (8.5) 9/459 (2.0) 35/459 (7.6) 112/459(24.4) 137/459(40.7) 125/459(27.2) 20(16129) 135/162 (S3 .3) 131/137(95.6) 6/137(4,4) 13/162(8.0) 87/162(53.7) 143/159(89.9) 14/159(8.8) 2/159(1.3) 152/162(93.3) 56/148(37.8) 55/105 (52.4) 94/134(70.1) 99/141(70.2) 29/134(21.6) 11/134(3.2) 36/162(22.2) 23/162(14.2) 2/162(1.2) 1/162(0.6) 8/162(4.9) 1/162(0.6) 11/162(6.8) 45/162(27.3) 66/162(40.7) 40/162(24.7) Age> 30 yr, Rate(%) 19(16D!5) 266/297 (89.6) 232/242 (95.9) 10/242 (4.1) 26/297 (8.8) 162/297(54.5) 284/297 (95.6) 1 1/297 [3.7) 2/297 (0.7) 271/297(91.2) SS/260 (33.8) 77/187 (41.2) 213/261 (81.6) 162/250(64,8) 32/261 (12.3) 15/261 (5.7) 45/297 (15.2) 34/297(11.4) 7/297 (2,4) 3/297 (1.0) 31/297(10.4) 8OT7 24/297 67/297 121/297 S5/297 (2.7) (S.I) (22.6) (40.7) (28.6) P 0.10 0.06 1.00 1.00 O.S6 0.92 0.03' 0.03* 0.61 0.37 0.45 0.07 0.01* 0.32 0.02* 0.39 0.07 0.46 0.50 1.00 0.050.17 0.72 0.21 1.00 0.33 Oihei rii-r [relinks h:hinJ ATV; Sta lioiuiy collision includes trees, fences, and touklcis, o Significant it 0.05 level. 13^0 (C) 2010 Llppncctt Wlliams & Wlh'ns Copyright (C) Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. The Journal o^TRAUMA* Injury, Infection, and Critical Care o Volume 69, Number 6, DaxiiiW 2010 ATV Injury Mortality and Healthcare Costs there is a trend toward more female riders, more incidents on streets or highways, and increased rates of confirmed alcohol consumption in the group younger than 30 years. Injury ATV trauma resulted in 4,117 total admission days. The median total length of stay per patient was 7 days (IQR, 4D13 days). Thirtypercent (120of395)ofpatientsrequired admission tothelCU for a median 6-day (IQR, 2D1 days) length of stay. Among the survivors to hospital discharge, nearly 30% (n 110, 28%) of patients required direct transfer to an alternate acute care facility (17%), a rehabilitation center (11%), or a chronic care facility (n 1) after the'acutc care admission. Injuries organized by body system are included in Table 2. Head, neck, and cervical spine (c-spine) injuries were the most common injuries among fatal ATV incidents and incidents resulting in major trauma center admission. These injuries were identified in nearly 60% of the study population and were frequently classified as serious, severe, and critical among hospitalized patients. Helmet use was associated with a significantly lower rate of head, neck, and c-spine injuries (63 of 144, 44%, vs. 184 of 264, 70%; p 0.000). Patients with these injuries were more likely to require ICU admission (91 of224, 40%, vs. 29 of 170, 17%; p 0.001), intubations (71 of224, 32%, vs. 10 of 170, 6%; p 0.001), or a surgical procedure (115 of 246,47%, vs. 66 of 170, 39%; p 0.02) and were less likely to be discharged home (136 of224, 61%, vs. 120 of 170, 71%; p 0.04). Patients with these injuries were often discharged to a rehabilitation center after their stay at the major trauma center (35 of 224, 16%, vs. 7 of 170, 4%; p 0.000). Injuries to the chesl and T-spinc occurred in nearly 60%of the study population (275 of 459). Although the chesl and T-spine injuries were commonly categorized as serious and severe, they were not associated with an increased likelihood of ICU admission, intubation, or requiring further care on discharge. These patients were less likely to require an operating room procedure (95 of 246, 39%, vs. S6 of 149, 58%; p 0.000). Mortality Of the 79 patients who died, 43 died on scene (54%), 15 died at a major trauma center (19%), and 21 died at a TABLE 2, Injuries Susta'ned in ATV Trauma Dearibed Head ard Ne* Induing C-Sjane Total, n (%) ME cases MAIS severity code Minor Moderate Sorious Sevi-TU Critical Maximum 271 (50.04) 46 peripheral rural hospital (27%). Case-fatality rates are provided in Table 3 outlining the relative risk (RR) of death for various incident factors. A significantly increased risk of mortality was observed in ATV incidents of rollover or ejection from the vehicle (RR, 2.75 [IQR, 1.13C6.70]; p 0.01, and RR, 4.18 [IQR, 1.7D0.32]; p 0.000, respectively). ATV incidents occurring at designated recreational areas and on streets or highways hadasignificanlly increased RRof mortality (RR, 3.66 [IQR, 2.52E32]; p 0.000, and RR, 2.56 [IQR, 1.73CB.80]; p 0.000, respectively). Factors representing rider behaviors also showed a significantly increased risk ofmortalityin ATV trauma. These factors included the consumptionof alcohol with aRR of mortality of 233 (IQR, 1.52D.56;p 0.000).An individual'schoicenotto wear a helmet was significantly linked to an increased risk of death (1.S2 [IQR, 1.11D.02]; p 0.008). Use of ATVs for work, despite being a smaller proportion of the injured population.also had a significantlyincreased riskfor mortalityat RR of 2.10 (IQR, 1.26D3.51;p 0.008). Two Injury classes were associated with significantly increased predictive risk of mortality. These were head injuries (2.19 [IQR, 1.35B.54]; p 0,001) and face injuries (RR, 1.59 [IQR, 1.051239]; p 0.03). A large number of crush injuries to the head and face were observed during the chart review of files from the ME. Although not statistically significant, many additional factors may increase the risk of mortality. Injuries occurring at home had a RR of mortality o f l .97 (IQR, 0.77C5.09). The ME charts revealed that the majority of events at home occurred during the loading of an ATV for transport. The increased risk of mortality, although not statistically significant, emphasizes the importance of public safety training and protective measures not only during ATV operation but also in ATV storage and transport. The ME chart review identifieda common theme associated with a lack of knowledge and familiarity with the terrain chosen for travel on the ATV resuUingin death. Many incidents were related to sudden unexpected falls off cliffs, do wn embankments, into rivers resulting in drowning, or collisions with stationary objects (e.g., trees, rocks, and fences). Because of the speculative natureof the observations contained in thesecharts, quantification for further analysis was not possible. by Body System Abdomen and Pdvic Cortot Including L-Spne 118(25.71) Upper arrf Lcwer Ext rarities 200 (43.57) Extartal 118(29.9) Face US (25.71) Chest Inducting T-^ana 275(59.91) 20 10 51 32 1 2] 7 24 >>JB 9S 19 S I 56 30 17 6 17 4 103 73 2 1 * 116 2 57 60 51 56 1 MAIS, Max! nun AHuevmtcd InjufV Scale. o Exlcnul injuics ret reputed fa Office .of th: Cfiief MS cai<<. (C) 2010 Uppncott Wliiams & Wlklns 1341 Copyright (C) Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. Krauss d al. 77)9 Journal of TRAUMA* Injury, Infection, and Critical Cars o Volume 69, Number 6, Duconber 2010 TAELE 3. Case fixity fetes end the Fa alive Rd-cof Mort^ity by ATV E/ent Descriptive Fkiors Factor Ejected Recreational area Rural Rollover Location street PenetratinB injury Alcohol intake Head injury Work-related use Location home No helmet Face injury Location mine Winter (DcccmbcrLfebnm ry) Spring (March CM ay) Male Location farm Age 30 yr Moving collision Weekend Summer (JanuaryCAugust) Autumn (Stp limber CNovem b< Infgd. 2006;fi3:124QI32. 22. CANSIM (Canadian Socio-tcononiic Infonnadon Management System) [database online]. Table 326-0021--Consumer Price Index (CPI). Ottawa. Canada: Statistics Canada. Available at http^/cansim2.statcan.gc,ca. Accessed May 12.2010. 23. Monk JP, Buckle)' R, Dyer D, Molorcyde-relat ed trauma in Alberta; a sad and expulsive story. Can J Bjrg, 2009;52:E235CE240, 24. CANSIM (Canadian Socio-economic Information Management System) [datibasu online]. Table 405-0004--Road mnlor vehicle, registralions, annual, Ottawa, Canada: Statistics Canada. Accessed February 23, 2010, 25. Rodgas GB.F:Ktors associated witli die all-terrain vehicle mortality rate in the United States: an analysis of slate-level data. Acdd Anal Pre/. 200S;40;725D732. (C) 2010 Upprcdt Wiliams & Wlki'rts 1343 Copyright (C) Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.