the Department of Pediatrics, University of Arkansas for Medical Sciences College of Medicine and Arkansas Children's Hospital, Little Rock, Arkansas
ABSTRACT
Objective. All-terrain vehicle (ATV) injuries among children represent a significant and growing problem. Although state-level analyses have characterized some aspects of pediatric ATV-related injuries, little information on the national impact on hospitalization is available. This study was designed to characterize more fully the patterns of injury, hospital length of stay, and hospital charges associated with ATV-related injuries, with a nationally representative sample.
Methods. Analyses were based on the 1997 and 2000 Healthcare Cost and Utilization Project Kids' Inpatient Database (KID). The KID is the only national, all-payer database of hospital discharges for children. KID data were weighted to represent all discharges from general hospitals in the United States. Discharges with external cause-of-injury codes consistent with off-road ATV-related injuries were selected, and the affected population was described. Nationally representative rates of ATV-related injuries were calculated, and changes between 1997 and 2000 were documented.
Results. An estimated 5292 children were hospitalized because of ATV-related injuries during the 2-year period, and hospitalizations increased 79.1% between 1997 and 2000. Rates of ATV-related hospitalization were highest among adolescent white male subjects, consistent with previous studies. Most patients had hospital lengths of stay of <4 days (68%), but 10% had stays of >8 days. Injury severity varied considerably, with more than one third of patients sustaining moderate to severe injuries. Approximately 1% of hospitalizations resulted in in-hospital deaths. Total hospital charges for this injury mechanism were $74367677 for the 2-year study period. Most of these charges were paid by private insurers.
Conclusions. This study provides evidence supporting recent substantial increases in childhood ATV-related injuries. The hospitalization impact of ATV-related injuries among children is considerable. Our data support the need for ongoing creative attempts to identify effective strategies to decrease ATV injuries among children.
Key Words: pediatrics injury all-terrain vehicle
Abbreviations: ISS, injury severity score ATV, all-terrain vehicle CPSC, Consumer Product Safety Commission KID, Kids' Inpatient Database ICD-9-CM, International Classification of Diseases, Ninth Revision, Clinical Modification E-code, external cause-of-injury code
Pediatric all-terrain vehicle (ATV)-related injuries represent a significant and growing problem. ATVs have been sold in the United States since 1972 and have experienced a popularity surge in recent years.1 Accompanying this increase in popularity is concern regarding the associated risks for children who ride on or drive ATVs. Current discussions have centered on appropriate vehicle sizes for children and the potential for legislative and educational measures to decrease the ATV injury burden.
ATVs are 3- or 4-wheeled machines designed with motorcycle-style handlebars, a high center of gravity, and large, low-pressure tires intended for off-road use. ATVs typically range in engine size from 50 cm3 to >500 cm3 of displacement and can weigh as much as 600 pounds. Nearly all ATVs are designed for 1 rider, because precise, quickly executed, weight-distribution adjustments must be performed to maintain balance and control of the vehicle in turns or in rough-terrain situations.
ATVs have proved useful for a wide range of occupational and recreational purposes but are also associated with significant risks for fatalities and injuries. In 1986, emergency departments admitted an estimated 106000 patients with ATV-related injuries, representing a 230% increase from 1983 estimates. In addition, there was an 18% increase in estimated ATV-related deaths from 1985 to 1986.2
Recognizing the growing problem of ATV-related injuries, public health organizations have attempted to prevent injuries through policy recommendations. The US Consumer Product Safety Commission (CPSC) developed a consent decree with ATV manufacturers and distributors, which began in 1988. The decree banned the sale of new 3-wheeled ATVs, established a nationwide ATV training program, improved warning materials such as labels to inform the public about the hazards associated with ATV use, and required age recommendations based on the size of the ATV. Specifically, ATVs with engines of >90 cm3 were not recommended for children <16 years of age.3 In addition, the American Academy of Pediatrics issued a policy statement discouraging ATV use for children <16 years of age4 and produced a model ATV safety bill that prohibits children <16 years of age from operating ATVs, mandates license, insurance, and registration requirements for all ATV operators, and bans operation of ATVs on public roads or with passengers. The bill also endorses the use of motorcycle-style helmets, eye protection, and safety clothing, as well as vehicle modifications including seatbelts, roll bars, and automatic headlights.4,5
Despite these efforts, the numbers of pediatric injuries continue to increase.6–9 The CPSC conducted ATV injury and exposure studies and reported that overall exposure to ATVs increased by 36% (number of drivers) to 50% (number of driving hours) from 1997 to 2001.10 The same report showed a disproportionate increase of 104% in ATV-related injuries. The risk of death for children in these studies was 4.5 to 12 times greater than that for adult comparison groups.11,12
Because of this increasing problem, effective methods to prevent ATV-related injuries, particularly among children, are needed. Recent literature has described the benefits of helmet requirements and safety legislation.1,4,6,12–15 The CPSC continues to have an active interest in this problem and held a public hearing in June 2003,16 during which ATV owners and dealers, state legislators, CPSC officials, doctors, researchers, and injury patients and their families discussed ATV injury prevention.
Additional information about the extent and character of ATV-related injuries could inform the debate about these vehicles and might strengthen the efforts of those working to prevent injuries through education or other activities. This study was designed to determine the impact of ATV-related injuries on the hospital care of children and adolescents. National data were used to assess the numbers of children and adolescents admitted to US hospitals for treatment of ATV-related injuries in 1997 and 2000. Additional analyses determined the distributions of hospitalizations across gender, ethnicity, income, and age categories and the severity of the injuries sustained.
METHODS
Study Design and Data
This study was a retrospective analysis of 1997 and 2000 hospitalizations based on the Kids' Inpatient Database (KID).17,18 The KID is a pediatric hospital discharge database developed by the Agency for Healthcare Research and Quality as part of the Healthcare Cost and Utilization Project. Data from 22 states in 1997 and 27 states in 2000 are included in the database. A 10% sample of normal newborn discharges and an 80% sample of all other pediatric discharges from each US community hospital (short-term, nonfederal, general and specialty hospitals) located in the participating states are included in the KID.19
The KID is designed to permit the analysis of a broad range of pediatric conditions, such as hospitalizations for treatment of specific injury mechanisms. Information on patient demographic factors, hospital characteristics, diagnosis and procedure codes, and injury severity is included for each hospitalization. Stratification weights are included, which allow the generation of population-based national estimates. This study used the strengths of the KID to develop a national sample of ATV-related hospitalizations.
Children 18 years of age with an International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) external cause-of-injury code (E-code) for an off-road motor vehicle accident were selected for analysis.20 The 821 category is described in the ICD-9-CM codebook as a "nontraffic accident involving other off-road motor vehicles."20 This category specifically excludes off-road vehicle accidents that occur on public highways. E-codes used in this study for case selection included 821.0, 821.1, 821.8, and 821.9, corresponding to injuries of the driver, a passenger, or another person injured in an off-road motor vehicle accident. We excluded codes 821.3 through 821.7, because these refer specifically to passengers, occupants, or riders of motorcycles, streetcars, animals, or pedal cycles, as well as pedestrians.
Modifications were made to the dataset to address coding inconsistencies. A small number of records were coded for both an off-road motor vehicle accident and an in-hospital birth, and these cases (n = 8) were excluded from analysis. The KID is based on hospitalizations and does not use unique identifiers that would enable tracking of individuals over several hospitalizations. To avoid overestimating the ATV-related injury incidence because of double-counting, hospitalizations (n = 44) that resulted in transfer to another short-term hospital were excluded from analysis.21 Short-term hospital transfers were retained for analysis of charges.
Statistical Analyses
We analyzed ATV-related hospitalizations according to patient demographic variables including age, gender, race, median household income according to zip code, and expected primary payer. Hospital characteristics, including hospital region, hospital type, and hospital bed size, were also considered in our analysis. KID variables used to examine the severity of injury included length of stay, patient disposition at discharge, and total hospital charges. In addition, ICDMAP-90 software was used to calculate injury severity scores (ISSs).22,23
National estimates were calculated with a stratified sampling design and were generated with Stata statistical analysis software.19,24,25 Because a sample size of 70 is required for national estimates, small categories were combined to produce cells large enough for inclusion in weighted analyses.
RESULTS
The unweighted datasets contained a combined total of 2373 ATV-related hospitalizations for children 18 years of age across both study years, 1997 and 2000. After weighting, this yielded a national estimate of 5292 cases, with 1896 (SE: 136) hospitalizations in 1997 and 3396 (SE: 147) hospitalizations in 2000, representing a 79.1% increase from 1997 to 2000.
We calculated the percentages of ATV-related hospitalizations that resulted in an in-hospital death for both years. The numbers of deaths for 1997 and 2000 were too small to be used in national estimates of in-hospital mortality rates; therefore, the percent died category is based on unweighted data. The percent died results were equal for the 2 years, representing 0.9% of hospitalizations.
Hospitalization rates (per 100000 population) were calculated for age and gender variables based on 1997 and 2000 US census data.26 Rates increased strikingly with time, most notably among older children and male subjects. Hospitalization rates for children 0 to 9 years of age increased 30%, to 1.49 cases per 100000 population. Rates for children 10 to 14 years of age increased 67%, from 4.45 cases per 100000 population in 1997 to 7.41 cases per 100000 population in 2000. A 103% increase was observed among children 15 to 18 years of age (from 3.91 cases per 100000 population to 7.95 cases per 100000 population). Male hospitalization rates increased 74%, from 3.91 cases per 100000 population to 6.81 cases per 100000 population; female rates increased 63%, from 1.20 cases per 100000 population to 1.95 cases per 100000 population.
The distribution of patient income categories changed considerably from 1997 to 2000. The low-income category ($0–35000) represented 73.0% of ATV hospitalizations in 1997 but only 46.2% of hospitalizations in 2000. A corresponding increase was observed for the high-income category (>$35000), which represented 19.1% of hospitalizations in 1997 and 52.8% in 2000.
The proportions of hospitalizations covered by public insurance, including Medicaid and Medicare, remained relatively equivalent during the study period, ie, 16% in both 1997 and 2000. The private insurance category, including health maintenance organizations, represented a large majority of hospitalizations for both years, increasing from 70.1% in 1997 to 73.5% in 2000. The third insurance category contains all other KID expected-payer designations, including self-pay, no charge, and other. This category decreased from 12.8% of hospitalizations in 1997 to 10.0% in 2000.
The proportion of mild injuries (ISS values of 0–8) decreased slightly from 1997 to 2000, ie, from 62.7% to 58.9% of ATV-related hospitalizations. The proportions of moderate injuries (ISS values of 9–16) remained approximately equal, at 28% of ATV-related hospitalizations, for the 2 years. The most severe injury category (ISS values of >16) increased from 9.1% of ATV-related hospitalizations in 1997 to 12.0% of hospitalizations in 2000. In addition to examination of ISS values according to category, KID weight variables were used to create a national estimate of the mean ISS values for the 2 years. The estimated ISS mean increased slightly, from 7.50 (SE: 0.42) in 1997 to 8.29 (SE: 0.20) in 2000.
Hospital charges were calculated from a dataset that included transfers to a short-term hospital. KID weight variables were used to create a national estimate of charges for the 2 years. The mean charge per hospitalization remained stable across the years, with charges of $13336 (SE: $654) in 1997 and $13823 (SE: $430) in 2000. Estimated total charges associated with these hospitalizations were more than $70 million for the 2 years studied. The proportions of charges according to insurance type were largely unchanged from 1997 to 2000, with 70% and 20% of charges being paid by private insurance and public insurance, respectively.
Hospital location (urban versus rural) and teaching status are recorded in the KID. Rural hospitals received relatively equal percentages of hospitalizations (21%) in 1997 and 2000. The representation of urban nonteaching hospitals decreased considerably, from 42.1% of hospitalizations in 1997 to 28.1% of hospitalizations in 2000. A corresponding increase was seen in the representation of urban teaching hospitals, from 36.1% in 1997 to 49.8% in 2000.
Hospital size (measured as the number of available beds) is described in the KID as small, medium, or large. Small and medium-sized hospitals had decreased representation of ATV-related hospitalizations from 1997 to 2000, whereas that of larger hospitals increased from 51.3% to 59.4%.
DISCUSSION
The CPSC 1997 and 2001 ATV injury studies, with a random-digit-dialing survey design, have provided the most comprehensive, nationally representative data demonstrating the recent increase in ATV-related injuries among children. Those studies were limited to the reports of survey respondents and lacked clinically important information about the nature of the injuries sustained. Our study extends those findings by using a large, nationally representative, pediatric discharge database. Results provide evidence supporting the increasing number of ATV-related injuries experienced by children. The study period is significant; the study used KID data collected in the years surrounding the end of the initial CPSC 10-year consent decree in 1998, which allows us to substantiate the subsequent increase in ATV-related injuries suggested in the CPSC injury reports.
Use of the KID data enabled a more detailed look at the hospital impact of ATV-related injuries at a national level than had been possible previously. The estimated number of ATV-related hospitalizations in our sample increased nearly 80%, to >3300 hospitalizations in 2000, which was even more dramatic than the CPSC report of a 56.5% increase in emergency department-treated injuries during the same time period. This larger change in admissions may reflect increasing severity of injuries, necessitating hospitalization, or other factors not accounted for in the available data.
Estimated gender and race characteristics were similar to those in previous studies, with a large proportion of ATV-related injuries being experienced by white male subjects.7,10,12,14,27–30 Despite warnings from manufacturers and policy statements from medical groups regarding the use of ATVs by children, most of these pediatric injuries occurred among younger children; two thirds of the children were 14 years of age, and approximately one fifth were <9 years of age. This underscores the need to intensify efforts to educate parents about ATV risks for children and to identify new strategies that might prevent inappropriate ATV use among young children.
The proportion of ATV-related hospitalizations associated with a median income of more than $35000 increased substantially between 1997 and 2000, from 19.1% to 52.8%. In the full KID, this category represented 45.2% of hospitalizations in 1997 and 59.9% of hospitalizations in 2000. The reasons behind this shift to a greater proportion of higher-income hospitalizations are not clear but may reflect, in part, the increasing costs associated with purchase of larger, more powerful ATVs over time. The CPSC documented a 202.7% increase in larger ATVs, with engine sizes of 400 cm3, during the same time period.10
We were able to identify 3 other studies in which injury severity was formally assessed with the ISS, all based on data from trauma referral centers in single states. Our severity findings were similar to those found in Ohio when ATV riders were compared with bicyclists.12 Reports with smaller samples from trauma centers in Georgia and Pennsylvania demonstrated mean ISS values of 13 for children 16 years of age.7,8 Our national data, by comparison, demonstrated mean scores of 7.50 in 1997 and 8.29 in 2000, representing injuries treated in general as well as in trauma referral centers. Although more than one half of ATV-related injuries in our study were associated with ISS values of <9, it is notable that many of the patients (>10%) sustained very severe injuries, with scores of >16 and correspondingly prolonged hospital stays of >1 week in many cases. Although few deaths were reported in our study, these hospitalizations represent substantial injuries. We were unable to characterize the long-term morbidity associated with these injuries using data available in the KID.
The mean charges reported here (more than $13000 per hospitalization) were substantially higher than those noted in a Utah study ($4240 per hospitalization) in the early 1990s.6 The total charges associated with the hospitalizations over 2 years exceeded $70 million, demonstrating an expensive, avoidable, public health problem. The bulk of these hospital expenses were borne by private insurers.
For several reasons, we think that the total cost of ATV-related injuries is underestimated in this study. Physician charges are not included in the KID; therefore, charge estimates described in this study underestimate the cost of ATV-related hospitalizations. In addition, we have no follow-up data to characterize the long-term health care needs of the affected patients. Finally, the KID does not include ATV-related injuries treated in outpatient or emergency department settings.
The burden of ATV-related injuries is unevenly distributed throughout the country, with injuries being most common in the South and Midwest regions of the country, consistent with previous usage reports.27,31 Our results also indicate a shift to hospitalization in larger, urban, teaching hospitals from 1997 to 2000. Rural hospitals maintained 21% of ATV-related hospitalizations for the 2 years. The notable change in location/teaching status from 1997 to 2000 was the shift from urban nonteaching hospitals to urban teaching hospitals.
Several limitations should be acknowledged. The use of ICD-9-CM E-codes as selection criteria excluded injuries that occurred on paved roads. ATV crashes occurring on paved roads have been associated with more severe injuries.9,11,31,32 Therefore, we think that our findings underestimate both the number and the severity of ATV-related injuries. However, because ATVs are not designed for use on paved roads, this analysis provides an opportunity to describe a population using ATVs in the setting for which they were designed. We attempted, through careful selection of the E-codes, to include only ATV-related hospitalizations in our data. It is possible that some cases of other off-road vehicles might have been included in our results, although the consistency of our data with previously published ATV injury studies suggests that the degree of misclassification was small.
Data from the KID are currently available only for the years 1997 and 2000, preventing a more detailed analysis of ATV-related hospitalizations over time. The anticipated release of KID data for 2003 will enhance our ability to describe hospitalization trends.
Some concerns might be raised that the increase in ATV-related hospitalizations found in the KID might be an artifact of its design. However, previous studies demonstrated stable KID hospitalization estimates for conditions for which no increase should be expected.33,34
CONCLUSIONS
This study provides additional evidence supporting recent substantial increases in the number of ATV-related injuries among children. The hospitalization impact of ATV-related injuries among children is considerable. Our data support the need for ongoing creative attempts to identify effective strategies to decrease ATV injuries among children.
ACKNOWLEDGMENTS
This work was supported by grants from the Health Resources and Services Administration, Maternal and Child Health Bureau (grant 1H33 MC 00088 01), and by the Robert Wood Johnson Foundation (Injury Free Coalition for Kids grant 047012).
We thank James M. Robbins, PhD, for helpful review of an earlier version of this manuscript and Suzanne Speaker for manuscript editing.
FOOTNOTES
Accepted Oct 6, 2004.
No conflict of interest declared.
REFERENCES
Scutchfield SB. All-terrain vehicles: injuries and prevention. Clin Orthop. 2003;(409) :61–72
Consumer Product Safety Commission. Annual Report of ATV Deaths and Injuries. Washington, DC: Consumer Product Safety Commission; 2000
US District Court for the District of Columbia. United States of America versus American Honda Motor Co., Inc., et al: Final Consent Decree: Civil Action 87-3525. March 14, 1988
American Academy of Pediatrics, Committee on Injury and Poison Prevention. All-terrain vehicle injury prevention: two-, three-, and four-wheeled unlicensed motor vehicles. Pediatrics. 2000;105 :1352 –1354
American Academy of Pediatrics. All-Terrain Vehicle Regulation Act [model bill]. In: American Academy of Pediatrics Policy Reference Guide. Elk Grove Village, IL: American Academy of Pediatrics; 1989
Cvijanovich NZ, Cook LJ, Mann NC, Dean JM. A population-based assessment of pediatric all-terrain vehicle injuries. Pediatrics. 2001;108 :631 –635
Lynch J, Gardner M, Worsey J. The continuing problem of all-terrain vehicle injuries in children. J Pediatr Surg. 1998;33 :329 –332
Ross R, Stuart L, Davis F. All-terrain vehicle injuries in children: industry-regulated failure. Am Surg. 1999;65 :870 –873
Bercher D, Staley K, Turner L, Aitken M. Pediatric injuries resulting from use of all-terrain vehicles. J Ark Med Soc. 2001;97 :351 –353
Consumer Product Safety Commission. All-Terrain Vehicle 2001 Injury and Exposure Studies. Washington, DC: Consumer Product Safety Commission; 2003
Rodgers GB. The effectiveness of helmets in reducing all-terrain vehicle injuries and deaths. Accid Anal Prev. 1990;22 :47 –58
Brown RL, Koepplinger ME, Mehlman CT, Gittelman M, Garcia VF. All-terrain vehicle and bicycle crashes in children: epidemiology and comparison of injury severity. J Pediatr Surg. 2002;37 :375 –380
Helmkamp JC. A comparison of state-specific all-terrain vehicle-related death rates, 1990–1999. Am J Public Health. 2001;91 :1792 –1795
Upperman JS, Shultz B, Gaines BA, et al. All-terrain vehicle rules and regulations: impact on pediatric mortality. J Pediatr Surg. 2003;38 :1284 –1286
Consumer Product Safety Commission. 35 Witnesses Set to Testify at CPSC Regional ATV Safety Hearing in West Virginia: ATV Injuries Double in 5-Year Period, Deaths Continue to Climb. Washington, DC: Consumer Product Safety Commission; 2003. Publication 03–136.
Healthcare Cost and Utilization Project. Kids' Inpatient Database (KID), 1997 [CD-ROMs]. Rockville, MD: Agency for Healthcare Research and Quality; 2000
Healthcare Cost and Utilization Project. Kids' Inpatient Database (KID), 2000 [CD-ROMs]. Rockville, MD: Agency for Healthcare Research and Quality; 2002
Healthcare Cost and Utilization Project. 2000 Kids' Inpatient Database (KID) Technical Documentation. Rockville, MD: Agency for Healthcare Research and Quality; 2002
US Department of Health and Human Services. International Classification of Diseases, 9th Revision, Clinical Modification. Washington, DC: US Department of Health and Human Services, Public Health Service, Health Care Financing Administration; 1998
Westfall JM, McGloin J. Impact of double counting and transfer bias on estimated rates and outcomes of acute myocardial infarction. Med Care. 2001;39 :459 –468
ICDMAP-90 [computer program]. Baltimore, MD: Johns Hopkins University and Tri-Analytics, Inc; 1998
Baker SP, O'Neill B. The injury severity score: an update. J Trauma. 1976;16 :882 –885
Stata Corp. Stata statistical software: release 7.0. College Station, TX: Stata Corp; 2001
Agency for Healthcare Research and Quality. Nationwide Inpatient Sample Technical Documentation. Rockville, MD: Agency for Healthcare Research and Quality; 2004
US Bureau of the Census. Profile of General Demographic Characteristics: 2000. Washington, DC: US Bureau of the Census; 2002
Rodgers GB. The characteristics and use patterns of all-terrain vehicle drivers in the United States. Accid Anal Prev. 1999;31 :409 –419
Rodgers GB. Risk factors for all-terrain vehicle injuries: a national case-control study. Am J Epidemiol. 2001;153 :1112 –1118
Marciani R, Caldwell G, Levine H. Maxillofacial injuries associated with all-terrain vehicles. J Oral Maxillofac Surg. 1999;57 :119 –123
Lister D, Morgan J III, Denning D, Valentovic M, Trent B, Beaver B. Pediatric all-terrain vehicles trauma: a 5-year statewide experience. J Pediatr Surg. 1998;33 :1081 –1083
Consumer Product Safety Commission. All-Terrain Vehicle Exposure, Injury, Death, and Risk Studies. Washington, DC: Consumer Product Safety Commission; 1998
Hargarten SW. All-terrain vehicle mortality in Wisconsin: a case study in injury control. Am J Emerg Med. 1991;9 :149 –152
Holman R, Curns A, Belay E, Steiner C, Schonberger L. Kawasaki syndrome hospitalizations in the United States, 1997 and 2000. Pediatrics. 2003;112 :495 –501
Robbins JM, Tilford JM, Bird TM, et al. Have newborn hospitalizations for birth defects declined following fortification of foods with folic acid Presented at the Academy Health Annual Research Meeting; San Diego, CA; June 6–8, 2004
Intentional Injury Management and Prevention in Pediatric Practice: Results From 1998 and 2003 American Academy of Pediatrics Periodic Surveys
Pernio in Pediatrics
Part-time Residency in Pediatrics: Description of Current Practice
Safety and Immunogenicity of the American Academy of Pediatrics–Recommended Sequential Pneumococcal Conjugate and Polysaccharide Vaccine Schedule in Pediatric Solid Organ Transplant Recipients
Integrating Pediatrics and Public Health
The Essential Role of Research in Community Pediatrics
Physician Reimbursement Levels and Adherence to American Academy of Pediatrics Well-Visit and Immunization Recommendations
A Review of Evidence Supporting the American Academy of Pediatrics Recommendation for Prescribing Cephalosporin Antibiotics for Penicillin-Allergic Patients
The Pediatrician’s Role in Community Pediatrics
Community Pediatrics: The Rochester Story