- Email: [email protected]
Motorcycle-related hospitalizations in the United States, 2001
Research Articles
Motorcycle-Related Hospitalizations in the United States, 2001 Jeffrey H. Coben, MD, Claudia A. Steiner, MD, MPH, Pamela Owens, PhD Objectives:
To estimate the prevalence of motorcycle-related hospitalization in the United States in 2001 and to describe the demographic, clinical, hospital, and financial characteristics associated with these injuries.
Methods:
Cross-sectional analysis of the 2001 Nationwide Inpatient Sample of the Healthcare Cost and Utilization Project was conducted in 2003.
Results:
There were an estimated 30,505 (confidence interval⫽26,566 –34,445) motorcycle-related hospital discharges in 2001. Approximately 62% of cases were aged ⱖ30 years, and males accounted for 89% of cases. The most common principal diagnoses were fractures of the lower limb (29.4%), fractures of the upper limb (13.1%), and intracranial injuries (12.3%). The mean length of stay was 5 days, the median hospital charge was $15,404, and the total estimated hospital charges were ⬎$841 million. The majority of patients (56.5%) were admitted to large urban teaching hospitals, and these hospitals accounted for nearly 70% of all hospital charges. Approximately 26% of cases were self-pay or listed public insurance as the expected payer.
Conclusions: These findings shed light on the substantial morbidity and financial impact of motorcyclerelated injuries. Renewed and strengthened prevention efforts are warranted. (Am J Prev Med 2004;27(5):355–362) © 2004 American Journal of Preventive Medicine
Introduction
M
otorcycle crashes are an increasing public health problem in the United States. Approximately 4.9 million motorcycles were registered in the United States in 2001.1 While these represent only 2% of all registered vehicles, motorcycles account for 7% to 12% of all motor vehicle–related fatalities.2,3 Per mile traveled, motorcyclists are 16 times more likely than passenger car occupants to die in a traffic crash and four times as likely to be injured.4 While only 20% of car crashes result in injury or death, that number increases to 80% for motorcycle crashes.5 After falling to a historic low in 1997, the number of motorcycle-related deaths and injuries has increased substantially. In 2001, there were 3181 motorcycle occupants killed in the United States. This represents a 10% increase from the number of fatalities reported in 2000, and a 50% increase from the number reported in From the Center for Delivery, Organization, and Markets, Agency for Healthcare Research and Quality, U.S. Department of Health and Human Services (Coben, Steiner, Owens), Washington DC; and Departments of Emergency Medicine and Community Medicine, Injury Control Research Center, West Virginia University (Coben), Morgantown, West Virginia Address correspondence and reprint requests to: Jeffrey H. Coben, MD, Injury Control Research Center, West Virginia University, P.O. Box 9151, Morgantown, WV 26506-9151. E-mail: [email protected] The full text of this article is available via AJPM Online at www.ajpm-online.net.
1997.2 This increasing trend is temporally associated with an increasing number of states repealing or modifying motorcycle helmet-use laws and a decreasing helmet-use rate among observed motorcyclists.2 Despite the considerable mortality and morbidity associated with motorcycle crashes, there are relatively few population-based studies examining hospitalizations associated with motorcycle trauma. A recent comprehensive review of this subject identified five studies utilizing statewide hospital discharge data.6 –11 Several other studies have relied upon single institution experience and/or trauma center experience. The Crash Outcome Data Evaluation System (CODES) project, involving seven states, used linked data from multiple sources to compile detail on individual crashes and their victims.12 The National Highway Traffic Safety Administration reported an estimated 60,000 nonfatal injuries among motorcyclists in 2001, and the National Electronic Injury Surveillance System (NEISS) reported an estimated 25,070 hospitalizations attributed to motorcycle crashes in 2001.2,13 However, NEISS provides insufficient detail on the clinical, hospital, and financial characteristics associated with these injuries. Research suggests that there has been no nationally representative study of the prevalence and impact of hospitalized motorcycle-related injuries. The primary objective of this study is to estimate the prevalence of motorcycle-related hospitalization in the
Am J Prev Med 2004;27(5) © 2004 American Journal of Preventive Medicine • Published by Elsevier Inc.
0749-3797/04/$–see front matter doi:10.1016/j.ampre.2004.08.002
355
United States in 2001, and to describe the demographic, clinical, hospital, and financial characteristics associated with these injuries. Recent studies have also highlighted concerns over off-road use of and injuries associated with all-terrain vehicles.14 Since motorcycles are also used for off-road recreational activities, it was further sought to identify the proportion of hospitalized cases resulting from off-road motorcycle crashes and determine if these cases present differing characteristics from in-traffic–related cases.
Methods Data were obtained from the Healthcare Cost and Utilization Project (HCUP), maintained by the Agency for Healthcare Research and Quality (AHRQ), and analyzed in 2003. The data in HCUP are derived from hospital discharge summaries and abstracts, which are created by hospitals primarily for billing and payment purposes. The hospital discharge summary contains the patient’s demographic information, conditions, procedures received, and other features about the hospital stay. Hospitals in many states provide discharge summaries to the state government, a hospital association, or other organizations. The HCUP is built through a partnership between the state-level data organizations and the AHRQ. Over time, beginning in 1988,15 the number of states contributing to HCUP has grown. The state data organizations provide their unique statewide database to HCUP. The data are then subjected to internal consistency and edit checks. Data elements that are similar across states are recoded into a uniform coding scheme and data elements unique to individual states are retained if they are useful for research purposes. These uniformly formatted data sets are the core of the HCUP databases.16 In 2001, 33 states provided data to HCUP. The HCUP State Inpatient Database (SID) contains the universe of each participating state’s community hospital inpatient discharge records. The 33 states contributing data to the SID captured approximately 80% of all hospital discharges in the United States. The year 2001 was chosen because it was the most recent year with complete data at the time of the study. The Nationwide Inpatient Sample (NIS) is the major derivative database created using the SID data. The NIS is a stratified probability sample of hospitals included in the SID, and is designed to approximate a 20% sample of all community hospitals in the United States. Hospitals are selected on the basis of a sampling frame that uses five strata: rural/urban location, number of beds, region, teaching status, and ownership. All discharges are retained for each sampled hospital.16 The 2001 NIS includes information on 7.4 million discharges from 986 nonfederal community hospitals located in the 33 states, which when weighted, provide estimates representing the total number of inpatient hospital discharges in the United States. The NIS provides a research database for conducting national and regional studies of inpatient care delivered in the United States. Motorcycle-related injuries were identified using E codes. Recent studies indicate that E coding within the NIS is relatively complete, and the data possess several strengths for conducting surveillance of hospitalized injuries.17–19 Since, by
356
definition, E codes are a secondary diagnosis, all secondary diagnosis fields were queried for cases that contained codes E810 through E825 (motor vehicle traffic and nontraffic accidents), and also contained a fourth digit of 0.2 (motorcyclist) or 0.3 (passenger on motorcycle). Traffic and nontraffic cases were further distinguished by grouping E810 – E819 and E820 –E825, respectively. For each identified case, the following variables were extracted: age, gender, median household income for patient’s ZIP code (per 1999 census data), principal diagnosis, secondary diagnoses, principal procedures, number of procedures, length of stay (LOS), in-hospital death, disposition of patient, expected primary payer, hospital charges, admission day of week, admission month, hospital region, hospital bed-size category, and location/teaching status of hospital. Information on hospital location, bed-size category, and teaching status were obtained from the American Hospital Association’s Annual Survey of Hospitals. Hospital census region was defined according to the U.S. Census Bureau. Within HCUP, a metropolitan statistical area is considered urban, and a nonmetropolitan statistical area is considered rural. Bed-size categories are based on the number of hospital beds, and are specific to the hospital’s location and teaching status. A hospital is considered to be a teaching hospital if it has an American Medical Association–approved residency program, and is a member of the Council of Teaching Hospitals or has a ratio of full-time-equivalent interns and residents to beds of ⱖ0.25.20 Analyses were conducted using the aggregate data for all cases identified, and by location of the event (traffic vs nontraffic). SUDAAN, version 8.0.0 (Research Triangle Institute, Research Triangle NC, 2001), was used to generate standard errors and 95% confidence intervals (CIs) around the national estimates. Cross-sectional analyses were performed for all demographic, clinical, and hospital characteristics for the entire group, and for traffic and nontraffic comparisons. Chi-square tests for categorical and unpaired t -tests for continuous variables were used to analyze differences. Unadjusted rates were calculated for hospital region using publicly available 2000 census data as population denominators. Unadjusted rates of admission by registered motorcycles were derived using publicly available vehicle registration data.1 Clinical diagnoses and procedures were grouped using Clinical Classifications Software (CCS), which aggregates individual ICD-9-CM codes into broad diagnosis and procedure groups.21
Results An estimated 30,505 (CI⫽26,566 –34,445) motorcyclerelated hospital discharges were identified nationwide in 2001. Males accounted for 89% of cases (CI⫽88.4 – 89.9). In 93% of cases, the injured patient was recorded as a motorcyclist, and 7% of cases were recorded as a passenger on a motorcycle. Approximately 62% of cases were aged ⱖ30 years; 84% were associated with traffic crashes, while 16% were classified as nontraffic crashes (Table 1). Forty-one percent of motorcycle-related cases were admitted during the weekend days, and nearly 46% of cases were admitted during the four
American Journal of Preventive Medicine, Volume 27, Number 5
Table 1. Patient characteristics and disposition for motorcycle-related hospitalizations Characteristics
n (95% CI)
Total estimated number of cases Event location Traffic Nontraffic Gender Male Female Age group (years) 0–14 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 65–69 ⱖ70 Median income of patient’s ZIP code (dollars) 0–24,999 25,000–34,999 35,000–44,999 ⱖ45,000 Disposition of patient Routine (home) Short-term hospitalization Other type of facility (includes long-term and rehab) Home health care Against medical advice Died
30,505 (26,566–34,445)
% (95% CI)
25,604 (22,128–29,080) 4,906 (4,132–5,680)
83.9 (82.0–85.9) 16.0 (14.1–18.0)
27,179 (23,649–30,708) 3,311 (2,832–3,790)
89.1 (88.4–89.9) 10.9 (10.1–11.6)
1,346 (977–1716) 2,479 (1,906–3,054) 4,302 (3,371–5,231) 3,497 (2,978–4,016) 3,830 (3,220–4,440) 3,856 (3,309–4,403) 3,461 (2,913–4,009) 2,887 (2,451–3,322) 2,291 (1,932–2,650) 1,122 (908–1,335) 723 (566–880) 351 (262–439) 361 (270–452)
4.4 (3.3–5.6) 8.1 (6.5–9.8) 14.1 (11.8–16.4) 11.5 (10.6–12.3) 12.6 (11.6–13.5) 12.6 (11.8–13.5) 11.4 (10.5–12.2) 9.5 (8.6–10.3) 7.5 (6.8–8.2) 3.7 (3.2–4.2) 2.4 (2.0–2.8) 1.2 (0.9–1.4) 1.2 (0.9–1.5)
1,115 (854–1,377) 5,736 (4,726–6,747) 8,091 (6,808–9,375) 14,912 (12,719–17,106)
3.7 (2.9–4.5) 19.2 (16.9–21.6) 27.1 (24.9–29.3) 49.9 (46.1–53.8)
23,792 (20,865–26,719) 1,043 (819–1,266) 2,839 (2,200–3,478)
78.0 (75.9–80.1) 3.4 (2.8–4.0) 9.3 (8.0–10.6)
1,978 (1,531–2,425) 236 (156–316) 606 (418–794)
6.5 (5.3–7.6) 0.8 (0.5–1.0) 2.0 (1.5–2.5)
CI, confidence interval.
months of June through September (data not shown). Two percent (CI⫽1.5–2.5) of all motorcycle-related cases (n ⫽606) died during the hospitalization. Seventy-eight percent (CI⫽75.9 – 80.1) were discharged routinely, whereas approximately 13% were discharged to another long-term or short-term healthcare facility (Table 1). The most common principal diagnosis for all cases, using CCS diagnosis categories, was fracture of the lower limb, which accounted for 29.4% (CI⫽28.1–30.8) of the diagnoses. Fractures of the upper limb accounted for 13.1% (CI⫽12.2–14.0), and intracranial injuries for 12.3% (CI⫽11.0 –13.6) of the principal diagnoses (Table 2). There were statistically significant differences in in-hospital deaths according to the principal diagnosis, with those sustaining intracranial injuries much more likely to die than those with other principal diagnoses (10.6% vs 0.8% died, p ⬍0.001). Compared with all other CCS categories, those with intracranial injury were also significantly more likely to be admitted through the emergency department, admitted on weekends, admitted to larger urban hospitals
in the Northeast, and have a higher proportion of females (p ⬍0.001). The most common principal procedure performed on motorcycle-related cases was treatment of lower limb fracture, 25% (CI⫽23.0 –26.2), and another 10% (CI⫽9.1–10.9) undergoing treatment of a fracture of the hip and femur (Table 3). An estimated 625 cases (CI⫽460 –789) had respiratory intubation and ventilation as the principal procedure, and another 3101 cases underwent intubation and ventilation as a secondary procedure (data not shown). The mean number of procedures per case was 2.5, with 79% of patients undergoing at least one procedure, and an aggregate total of 75,780 procedures associated with motorcyclerelated injuries (data not shown). The mean LOS for patients with motorcycle-related injuries was 5 days, with an estimated total of 161,094 hospital days for cases in 2001. The estimated total hospital charges for motorcycle-related injuries were ⬎$841 million, and the median hospital charge was $15,404. Hospital charges and LOS varied according to the principal diagnosis, and are detailed in Table 2. Am J Prev Med 2004;27(5)
357
Table 2. Length of stay and hospital charges Hospital charges ($) n (95% CI) Total estimated cases 30,505 (26,566–34,445) Top ten principal diagnoses Fracture of lower limb 8,980 (7,892–10,068) Fracture of upper limb 3,994 (3,398–4,590) Intracranial injury 3,753 (3,027–4,479) Other fractures 3,489 (2,908–4,070) Crushing/internal injury 3,313 (2,801–3,824) Open wounds of extremities 1,103 (897–1,308) Skull and face fractures 849 (656–1,043) Other injuries and conditions 818 (594–1,043) Superficial injury, contusion 688 (509–868) Joint disorders and 581 (455–707) dislocations Hospital location Rural Urban nonteaching Urban teaching
% (95% CI)
29.4 (28.1–30.8) 13.1 (12.2–14.0) 12.3 (11.0–13.6) 11.4 (10.5–12.4) 10.9 (9.9–11.8) 3.6 (3.1–4.1) 2.8 (2.3–3.2) 2.7 (2.1–3.3) 2.3 (1.7–2.8) 1.9 (1.5–2.3)
LOS (days)
Mean Median Total charges Mean Median 28,151 15,404
841,685,148
5.3
3.0
26,821 19,990 38,970 31,934 36,197 23,214 22,212 14,718 11,077 20,333
17,817 13,427 17,446 15,089 17,792 12,525 15,354 8,379 6,970 11,369
235,220,490 77,825,146 144,228,116 109,307,629 117,536,517 25,382,185 18,741,773 11,718,719 7,562,694 11,351,840
5.1 3.3 6.5 6.2 6.8 4.5 3.7 2.5 2.6 3.5
3.0 2.0 3.0 4.0 5.0 3.0 2.0 1.0 2.0 2.0
13,554 9,143 22,441 14,061 34,574 18,205
49,542,210 213,105,141 579,047,797
CI, confidence interval; LOS, length of stay.
Those with intracranial injury had significantly increased LOS, total hospital charges, and discharge to a long-term care facility, and were more likely to be self-pay (p ⬍0.001).
The majority of patients (56.5%, CI⫽50.6 – 62.4) were admitted to urban teaching hospitals, and most were admitted to large hospitals (Table 3). Twelve percent of cases (CI⫽9.5–14.5) were admitted to rural
Table 3. Facility characteristics, procedures, and expected payers Characteristics Hospital location/teaching status Rural Urban nonteaching Urban teaching Bed size of hospital Small Medium Large Top ten principal procedures Treatment, fracture/dislocation of lower extremity Treatment, fracture/dislocation of hip and femur Other fracture and dislocation procedure Debridement of wound Treatment, fracture, or dislocation of radius and ulna Suture of skin and subcutaneous tissue Incision of pleura, thoracentesis, chest drainage Respiratory intubation and ventilation Traction, splints, and other wound care Tracheostromy, temporary and permanent Primary expected payer Medicare Medicaid Private insurance/HMO Self pay No charge Other
n (95% CI)
% (95% CI)
3,660 (2,989–4332) 9,611 (8,110–11,113) 17,234 (13,654–20,813)
12.0 (9.5–14.5) 31.5 (26.5–36.6) 56.5 (50.6–62.4)
1,998 (1,523–2474) 7,853 (6,015–9692) 20,654 (17,202–24105)
6.5 (4.9–8.2) 25.7 (20.4–31.1) 67.7 (62.1–73.3)
5,889 (5,166–6,613)
24.6 (23.0–26.2)
2,385 (1,987–2,782)
10.0 (9.1–10.9)
1,761 (1,467–2,056) 1,554 (1,282–1,825) 1,549 (1,288–1,810)
7.4 (6.6–8.1) 6.5 (5.7–7.3) 6.5 (5.7–7.2)
1,194 (965–1,422) 1,011 (818–1,204)
5.0 (4.3–5.7) 4.2 (3.6–4.9)
625 (458–791) 582 (447–716) 452 (278–625)
2.6 (2.1–3.1) 2.4 (1.9–3.0) 1.9 (1.3–2.5)
852 (692–1,013) 2,210 (1,774–2,646) 20,366 (17,624–23,109) 4,941 (3,671–6,210) 345 (40–651) 1,595 (1,171–2,018)
CI, confidence interval; HMO, health maintenance organization.
358
American Journal of Preventive Medicine, Volume 27, Number 5
2.8 (2.3–3.3) 7.3 (6.0–8.6) 67.2 (63.8–70.6) 16.3 (12.9–19.7) 1.1 (0.2–2.1) 5.3 (4.0–6.5)
Table 4. Hospitalizations and hospitalization rates
Characteristics
n (95% CI)
United States total Hospital region Northeast Midwest South West
30,505 (26,566–34,445) 7001 (5271–8730) 4986 (3911–6061) 9816 (7216–12417) 8703 (6555–10850)
% (95% CI)
22.9 (17.8–28.0) 16.3 (12.8–19.9) 32.2 (25.6–38.7) 28.5 (22.6–34.4)
Registered motorcycles
Registered motorcycles/ 100,000 population
Admissions/ 100,000 motorcycles
10.9
4,903,056
1742
623
13.1 7.7 9.8 13.8
781,589 1,541,096 1,283,722 1,295,207
1461 2397 1281 2053
896 324 765 672
Admissions/ 100,000 population
CI, confidence interval.
hospitals. Approximately 67% of cases had private insurance or health maintenance organization as the primary expected payer, while 26% were self-pay or had public insurance as the expected payer. Hospitalization rates per 100,000 population were highest among hospitals located in the Northeast and West. Hospitalizations per 100,000 motorcycles were highest in the Northeast and South (Table 4). Comparing traffic cases with nontraffic cases revealed a number of significant differences between the
two groups (Table 5). Cases occurring in traffic were found to be older and to include a relatively higher percentage of females. Traffic cases were more likely to have increased LOS, higher total hospital charges, greater number of procedures, increased prevalence of in-hospital deaths, and increased nonroutine disposition (including transfer to long-term care facilities), and were more likely to be self-pay (all p ⬍0.001). In contrast, cases resulting from nontraffic motorcycle use were more likely to be admitted on the weekend days,
Table 5. Motorcycle crash location Number of cases (95% CI)* Age (mean years)* Length of stay (mean days)* Total charges (mean dollars)* Number of procedures (mean)* Gender (% female)** Top five principal diagnoses (% of total) Fracture lower limb Fracture upper limb Intracranial injury Other fracture Crushing or internal injury Disposition (% of total)** Routine (home) Short-term hospitalization Other type of facility (includes long-term and rehab) Home health care Against medical advice Died Payer (% of total)** Medicare Medicaid Private Self-pay No charge Other Hospital location/teaching status (% of total)** Rural Urban, nonteaching Urban, teaching
Traffic
Nontraffic
25,604 (22,128⫺29,080) 36.5 5.6 30,086 2.6 11.7
4,906 (4,132⫺5,680) 27.7 3.5 18,018 1.9 6.4
28.8 12.7 12.8 11.4 10.4
32.7 15.3 9.7 11.6 13.3
75.9 3.4 10.6
88.9 3.3 2.5
7.0 0.8 2.2
4.0 0.5 0.8
3.0 7.2 66.2 17.1 1.3 5.2
1.9 7.7 72.3 12.4 0.4 5.3
11.0 30.4 58.5
17.1 37.1 45.8
*t-test: p ⬍ 0.0001 (bolded). **Chi-square test: p ⬍ 0.0001 (bolded), indicates a significant difference in response distributions between the two groups. CI, confidence interval.
Am J Prev Med 2004;27(5)
359
during the nonsummer months, to smaller hospitals, and to rural and nonteaching hospitals. Nontraffic cases had a lower prevalence of intracranial injuries as the principal diagnosis, and a corresponding higher prevalence of extremity fractures and crushing injuries (Table 5).
Discussion These findings confirm that motorcycle-related injuries are a significant cause of hospitalization, hospital bed utilization, disability, and healthcare costs in the United States. Paramount among the findings is the uneven distribution of these cases and their resultant financial burden. Of the $842 million in total hospital charges for motorcycle-related cases in 2001, nearly 70% was from urban teaching hospitals. While 10.1% of patients used public sources of insurance, another 16.3% were uninsured. This compares with an estimated 4.6% of U.S. hospitalizations for all causes being uninsured in 2001.22 Patients with intracranial injury were more likely to be admitted to large urban teaching hospitals and these patients had longer stays, with higher total charges. They were also more likely to be self-pay. These cases likely contribute a substantial economic burden to academic medical centers. Many of the findings are consistent with prior studies and expectations. The high proportion of male patients, age distribution, day of week and month of year admission patterns, and clinical diagnoses are similar to what has been previously reported.6,23,24 However, this study provides the first national picture of the magnitude of this problem. Comparing the 30,505 estimated motorcycle-related hospitalizations with a similar analysis performed on the estimated 35,810 firearm-related hospitalizations in 1997,18 reveals that while firearm cases were more likely to die in the hospital, motorcycle cases were more likely to be discharged to a long-term care facility and incur greater hospital charges. In addition to the economic consequences, these data reveal that on average, for each day in 2001, there were approximately 25 new lower extremity fractures, 10 new intracranial injuries, and one new spinal cord injury resulting from motorcycle crashes. The procedural data provided in Table 3 only demonstrate principal procedures. The majority of patients also had a number of secondary procedures performed during their hospitalization, and the overall mean number of procedures was 2.5 per patient. When considering both principal and secondary procedures, it was found that every day approximately 208 medical procedures are performed on victims of motorcycle-related injuries, including an average of ten intubations per day. The finding that 79% of patients undergo at least one procedure further underscores the severity of these cases, as this compares with 60% among all hospitalized patients.25 360
Hospitalization rates varied by region, and interesting differences emerged when different denominators were used to examine hospitalizations. As seen in Table 4, overall hospitalization rates per population were highest in the Northeast and West. However, these crude population-based rates provide little information on exposure. Motorcycle ownership provides a better measure of exposure, and motorcycle ownership varies by geographic region, with the Midwest and West having the highest number of registered motorcycles per population (Table 4). When examining hospitalization rates according to the number of registered motorcycles, it was found that rates were highest in the Northeast and South. Numerous factors, including the number of miles driven, weather, road conditions, population density, injury patterns, helmet legislation, and local treatment practices may account for this variation. A recent analysis revealed that weather and population density were important explanatory variables for the variance in motorcycle deaths across different states.26 There were also several interesting differences between cases occurring in traffic and cases involving nontraffic crashes. While those occurring in traffic demonstrate greater severity of injury (increased LOS, greater number of procedures, increased in-hospital death rate), nontraffic cases constitute a significant proportion of the total number of admitted patients. These patients tend to be admitted to smaller rural hospitals and they are a younger population of riders. In fact, 17% of those admitted from nontraffic motorcycle use were aged ⬍16 years. This population appears to be different from those experiencing traffic crashes and may warrant different prevention strategies. A significantly higher proportion of nontraffic cases were admitted during the weekend days and during the winter months. According to ICD-9-CM coding guidelines, a motor vehicle accident is considered a traffic accident if it “occurs on a public highway” and “a motor vehicle accident is assumed to have occurred on the highway unless another place is specified.” Examples of nontraffic accidents include a crash occurring in a driveway, parking lot, farm, or other off-road location. It is suspected, but cannot be confirmed, that the popular use of motorcycles on “dirt tracks” or other off-road locations may be responsible for many of these cases. There are a number of important limitations to these data. First, NIS data are derived from hospital discharge data, which contains no information on motorcycle helmet usage or other protective devices. Therefore, no conclusions can be drawn regarding the impact of helmet use or nonuse. There is, however, a large body of previous research demonstrating both the health and cost savings associated with helmet use.6 Second, the identification of motorcycle-related cases, and the distinction between traffic and nontraffic cases relied
American Journal of Preventive Medicine, Volume 27, Number 5
on the use of E codes, and the validity of E codes in national morbidity data systems has been questioned.27 Among other considerations, E code validity is dependent on the accuracy and amount of information supplied in the medical record, accurate capture of this information and appropriate coding by medical records personnel, transfer of those codes to the discharge abstract, and inclusion of those codes in the final administrative data set. While the coding that occurs at the hospital level could not be verified, recent studies indicate that E coding within the HCUP data is quite complete, with nearly 90% of all injury diagnoses contained in the NIS having a corresponding E code.17–19 The estimate of 30,505 hospitalizations in 2001 is consistent with estimates from the NEISS, which reported 25,070 cases of hospitalized motorcyclists in 2001. The consistency of the findings with other published data on motorcycle injuries and other national estimates provides further support for the validity of E codes within HCUP data. Nevertheless, it is possible that some motorcycle cases were either not counted or misclassified as a result of coding errors. Third, the financial data provided within the NIS are based on hospital charges, not actual costs. Hospital charges are generally more than actual costs.6 However, it should also be noted that hospital charge data do not include a number of important cost items including physician professional fees, emergency transportation costs, and subsequent rehabilitation costs. In general, the reported median hospital charge of $15,404 falls within the range of hospital charges for motorcycle injuries reported in previous studies.6 Despite these limitations, findings demonstrate the significant morbidity and economic consequences associated with motorcycle-related hospitalizations. The number of motorcyclist fatalities has risen dramatically in the last 7 years. During this time, the number of states repealing mandatory motorcycle helmet use legislation has increased, and national surveys have demonstrated a reduction in helmet use from 71% in the fall of 2000 to 58% in the summer of 2002.2 While the data provide only a snapshot of the acute care hospitalizations occurring in 2001, they further illustrate the growing burden and need for preventive interventions. Recent research has demonstrated that states with full helmet laws have lower motorcyclist death rates than states without full helmet laws.26 A similar crosssectional analysis examining the relationships among helmet laws, morbidity, and hospitalization costs is warranted, and will likely provide valuable information to policymakers. We would like to acknowledge the statewide data organizations that participate in the Healthcare Cost and Utilization Project Nationwide Inpatient Sample in 2001: Arizona Department of Health Services; California Office of Statewide Health Planning and Development; Colorado Health and
What This Study Adds . . . Motorcycle crashes are a significant and rising cause of injury mortality. Despite the mortality and morbidity associated with motorcycle crashes, there are relatively few population-based studies examining hospitalizations associated with motorcycle trauma. This study provides national estimates, derived from statewide hospital discharge data from 33 states, of the prevalence, morbidity, and costs associated with motorcycle-related hospitalizations. The study also identifies a unique subset of motorcycle injuries, associated with the off-road use of these vehicles.
Hospital Association; Connecticut—CHIME, Inc.; Florida Agency for Health Care Administration; Georgia—GHA: An Association of Hospitals and Health Systems; Hawaii Health Information Corporation; Illinois Health Care Cost Containment Council; Iowa Hospital Association; Kansas Hospital Association; Kentucky Department for Public Health; Maine Health Data Organization; Maryland Health Services Cost Review Commission; Massachusetts Division of Health Care Finance and Policy; Michigan Health and Hospital Association; Minnesota Hospital Association; Missouri Hospital Industry Data Institute; Nebraska Hospital Association; New Jersey Department of Health and Senior Services; New York State Department of Health; North Carolina Department of Health and Human Services; Oregon Association of Hospitals and Health Systems; Pennsylvania Health Care Cost Containment Council; Rhode Island Department of Health; South Carolina State Budget and Control Board; Tennessee Hospital Association; Texas Health Care Information Council; Utah Department of Health; Vermont Association of Hospitals and Health Systems; Virginia Health Information; Washington State Department of Health; West Virginia Health Care Authority; and Wisconsin Department of Health and Family Services. The views expressed are those of the authors and not necessarily those of the Agency for Healthcare Research and Quality.
References 1. Federal Highway Administration. Highway statistics 2001, U.S. Department of Transportation. Available at: www.fhwa.dot.gov/ohim/hs01/index.htm. Accessed September 14, 2003. 2. National Highway Traffic Safety Administration. Traffic safety facts 2001— motorcycles. U.S. Department of Transportation. Available at www-nrd. nhtsa.dot.gov/departments/nrd-30/ncsa/. Accessed September 14, 2003. 3. Sosin D, Sacks J, Holmgreen P. Head injury associated deaths from motorcycle crashes: relationship to helmet use laws. JAMA 1990;264:1395–9. 4. National Highway Traffic Safety Administration. Traffic safety facts 1996 — motorcycles. U.S. Department of Transportation. Available at www-nrd. nhtsa.dot.gov/departments/nrd-30/ncsa/. Accessed September 14, 2003. 5. National Highway Traffic Safety Administration. Fatality analysis reporting system data, 1996. U.S. Department of Transportation. Available at: www. nhtsa.dot.gov/people/injury/pedbimot/motorcycle/safebike/approach. html. Accessed September 14, 2003.
Am J Prev Med 2004;27(5)
361
6. Pacific Institute for Research and Evaluation. Costs of injuries resulting from motorcycle crashes: a literature review. Washington DC: U.S. Department of Transportation, 2002. 7. Braddock M, Schwartz R, Lapidus G, Banco L, Jacobs L. A population-based study of motorcycle injury and costs. Ann Emerg Med 1992;21:273– 8. 8. Karlson TA, Quade CA. Head injuries associated with motorcycle use— Wisconsin, 1991. MMWR Morb Mortal Wkly Rep 1994;43:423, 429 –31 9. Max W, Stark B, and Root S. Putting a lid on injury costs: the economic impact of the California motorcycle helmet law. J Trauma 1998;45:550 – 6. 10. Rowland J, Rivara F, Salzberg P, Soderberg R, Maier R, Koepsell T. Motorcycle helmet use and injury outcome and hospitalization costs from crashes in Washington State. Am J Public Health 1996;86:41–5. 11. Shankar BS, Ramzy AI, Soderstrom CA, Dischinger PC, Clark CC. Helmet use, patterns of injury, medical outcome, and costs among motorcycle drivers in Maryland. Accid Anal Prev 1992;24:385–96. 12. National Highway Traffic Safety Administration. Report to Congress: benefits of safety belts and motorcycle helmets. Washington DC: U.S. Department of Transportation, February 1996. 13. Office of Statistics and Programming, National Center for Injury Prevention and Control, Centers for Disease Control and Prevention. Home page. Available at: www.cdc.gov/ncipc/wisqars/. Accessed April 12, 2004. 14. Helmkamp JC. A comparison of state-specific all-terrain vehicle-related death rates, 1990 to 1999. Am J Public Health 2001;91:1792–5. 15. Agency for Healthcare Research and Quality. Healthcare cost and utilization project, 1988 –2001: a federal-state-industry partnership in health data. Available at: www.ahrq.gov/data/hcup/hcup-pkt.htm. Accessed March 19, 2004. 16. Steiner C, Elixhauser A, Schnaier J. The healthcare cost and utilization project: an overview. Eff Clin Pract 2002;5:143–51. 17. Schulman J, Provenzano G, Pierce B, Wolters C, Brustrom J, Miles E. Investigation of the utility of Healthcare Cost and Utilization Project
362
18. 19.
20.
21.
22.
23. 24.
25.
26. 27.
(HCUP) data for annual surveillance of hospitalized injuries. Final report from Batelle Centers for Public Health Research and Evaluation to the Centers for Disease Control and Prevention. Atlanta GA: Batelle Centers for Public Health Research and Evaluation, September 2003. Coben JH, Steiner CA. Hospitalization for firearm-related injuries in the United States, 1997. Am J Prev Med 2003;24:1– 8. Coben JH, Owens P. Evaluation and use of HCUP emergency department data. Proceedings of the 18th Annual Meeting of the National Association of Health Data Organizations, Baltimore MD, December 2003. Agency for Healthcare Research and Quality. HCUP data documentation. Available at: www.hcup-us.ahrq.gov/db/nation/nis/dbdoc/nisdde.jsp. Accessed May 27, 2004. Agency for Healthcare Research and Quality. HCUP tools and software. Available at: www.hcup-us.ahrq.gov/tools/ccs.jsp. Accessed March 29, 2004. Agency for Healthcare Research and Quality. HCUPnet 2001 national statistics. Available at: http://hcup.ahrq.gov/HCUPnet. Accessed May 25, 2004. Aare M, von Holst H. Injuries from motorcycle and moped crashes in Sweden from 1987 to 1999. Inj Control Saf Promot 2003;10:131– 8. Peek C, Braver E, Shen H, Kraus J. Lower extremity injuries from motorcycle crashes: a common cause of preventable injury. J Trauma 1994;37:358 – 64. Elixhauser A, Klemstine K, Steiner C, Bierman AS. Procedures in U.S. hospitals, 1997. HCUP fact book number 2. Available at: www.ahrq.gov/ data/hcup/factbk2/factbk2.htm. Accessed November 6, 2003. Branas CC, Knudson MM. Helmet laws and motorcycle rider death rates. Accid Anal Prev 2001;33:641– 8. Annest JL, Mercy JA. Use of national data systems for firearm-related injury surveillance. Am J Prev Med 1998;15(suppl 3):17–30.
American Journal of Preventive Medicine, Volume 27, Number 5
Motorcycle-Related Hospitalizations in the United States, 2001 Jeffrey H. Coben, MD, Claudia A. Steiner, MD, MPH, Pamela Owens, PhD Objectives:
To estimate the prevalence of motorcycle-related hospitalization in the United States in 2001 and to describe the demographic, clinical, hospital, and financial characteristics associated with these injuries.
Methods:
Cross-sectional analysis of the 2001 Nationwide Inpatient Sample of the Healthcare Cost and Utilization Project was conducted in 2003.
Results:
There were an estimated 30,505 (confidence interval⫽26,566 –34,445) motorcycle-related hospital discharges in 2001. Approximately 62% of cases were aged ⱖ30 years, and males accounted for 89% of cases. The most common principal diagnoses were fractures of the lower limb (29.4%), fractures of the upper limb (13.1%), and intracranial injuries (12.3%). The mean length of stay was 5 days, the median hospital charge was $15,404, and the total estimated hospital charges were ⬎$841 million. The majority of patients (56.5%) were admitted to large urban teaching hospitals, and these hospitals accounted for nearly 70% of all hospital charges. Approximately 26% of cases were self-pay or listed public insurance as the expected payer.
Conclusions: These findings shed light on the substantial morbidity and financial impact of motorcyclerelated injuries. Renewed and strengthened prevention efforts are warranted. (Am J Prev Med 2004;27(5):355–362) © 2004 American Journal of Preventive Medicine
Introduction
M
otorcycle crashes are an increasing public health problem in the United States. Approximately 4.9 million motorcycles were registered in the United States in 2001.1 While these represent only 2% of all registered vehicles, motorcycles account for 7% to 12% of all motor vehicle–related fatalities.2,3 Per mile traveled, motorcyclists are 16 times more likely than passenger car occupants to die in a traffic crash and four times as likely to be injured.4 While only 20% of car crashes result in injury or death, that number increases to 80% for motorcycle crashes.5 After falling to a historic low in 1997, the number of motorcycle-related deaths and injuries has increased substantially. In 2001, there were 3181 motorcycle occupants killed in the United States. This represents a 10% increase from the number of fatalities reported in 2000, and a 50% increase from the number reported in From the Center for Delivery, Organization, and Markets, Agency for Healthcare Research and Quality, U.S. Department of Health and Human Services (Coben, Steiner, Owens), Washington DC; and Departments of Emergency Medicine and Community Medicine, Injury Control Research Center, West Virginia University (Coben), Morgantown, West Virginia Address correspondence and reprint requests to: Jeffrey H. Coben, MD, Injury Control Research Center, West Virginia University, P.O. Box 9151, Morgantown, WV 26506-9151. E-mail: [email protected] The full text of this article is available via AJPM Online at www.ajpm-online.net.
1997.2 This increasing trend is temporally associated with an increasing number of states repealing or modifying motorcycle helmet-use laws and a decreasing helmet-use rate among observed motorcyclists.2 Despite the considerable mortality and morbidity associated with motorcycle crashes, there are relatively few population-based studies examining hospitalizations associated with motorcycle trauma. A recent comprehensive review of this subject identified five studies utilizing statewide hospital discharge data.6 –11 Several other studies have relied upon single institution experience and/or trauma center experience. The Crash Outcome Data Evaluation System (CODES) project, involving seven states, used linked data from multiple sources to compile detail on individual crashes and their victims.12 The National Highway Traffic Safety Administration reported an estimated 60,000 nonfatal injuries among motorcyclists in 2001, and the National Electronic Injury Surveillance System (NEISS) reported an estimated 25,070 hospitalizations attributed to motorcycle crashes in 2001.2,13 However, NEISS provides insufficient detail on the clinical, hospital, and financial characteristics associated with these injuries. Research suggests that there has been no nationally representative study of the prevalence and impact of hospitalized motorcycle-related injuries. The primary objective of this study is to estimate the prevalence of motorcycle-related hospitalization in the
Am J Prev Med 2004;27(5) © 2004 American Journal of Preventive Medicine • Published by Elsevier Inc.
0749-3797/04/$–see front matter doi:10.1016/j.ampre.2004.08.002
355
United States in 2001, and to describe the demographic, clinical, hospital, and financial characteristics associated with these injuries. Recent studies have also highlighted concerns over off-road use of and injuries associated with all-terrain vehicles.14 Since motorcycles are also used for off-road recreational activities, it was further sought to identify the proportion of hospitalized cases resulting from off-road motorcycle crashes and determine if these cases present differing characteristics from in-traffic–related cases.
Methods Data were obtained from the Healthcare Cost and Utilization Project (HCUP), maintained by the Agency for Healthcare Research and Quality (AHRQ), and analyzed in 2003. The data in HCUP are derived from hospital discharge summaries and abstracts, which are created by hospitals primarily for billing and payment purposes. The hospital discharge summary contains the patient’s demographic information, conditions, procedures received, and other features about the hospital stay. Hospitals in many states provide discharge summaries to the state government, a hospital association, or other organizations. The HCUP is built through a partnership between the state-level data organizations and the AHRQ. Over time, beginning in 1988,15 the number of states contributing to HCUP has grown. The state data organizations provide their unique statewide database to HCUP. The data are then subjected to internal consistency and edit checks. Data elements that are similar across states are recoded into a uniform coding scheme and data elements unique to individual states are retained if they are useful for research purposes. These uniformly formatted data sets are the core of the HCUP databases.16 In 2001, 33 states provided data to HCUP. The HCUP State Inpatient Database (SID) contains the universe of each participating state’s community hospital inpatient discharge records. The 33 states contributing data to the SID captured approximately 80% of all hospital discharges in the United States. The year 2001 was chosen because it was the most recent year with complete data at the time of the study. The Nationwide Inpatient Sample (NIS) is the major derivative database created using the SID data. The NIS is a stratified probability sample of hospitals included in the SID, and is designed to approximate a 20% sample of all community hospitals in the United States. Hospitals are selected on the basis of a sampling frame that uses five strata: rural/urban location, number of beds, region, teaching status, and ownership. All discharges are retained for each sampled hospital.16 The 2001 NIS includes information on 7.4 million discharges from 986 nonfederal community hospitals located in the 33 states, which when weighted, provide estimates representing the total number of inpatient hospital discharges in the United States. The NIS provides a research database for conducting national and regional studies of inpatient care delivered in the United States. Motorcycle-related injuries were identified using E codes. Recent studies indicate that E coding within the NIS is relatively complete, and the data possess several strengths for conducting surveillance of hospitalized injuries.17–19 Since, by
356
definition, E codes are a secondary diagnosis, all secondary diagnosis fields were queried for cases that contained codes E810 through E825 (motor vehicle traffic and nontraffic accidents), and also contained a fourth digit of 0.2 (motorcyclist) or 0.3 (passenger on motorcycle). Traffic and nontraffic cases were further distinguished by grouping E810 – E819 and E820 –E825, respectively. For each identified case, the following variables were extracted: age, gender, median household income for patient’s ZIP code (per 1999 census data), principal diagnosis, secondary diagnoses, principal procedures, number of procedures, length of stay (LOS), in-hospital death, disposition of patient, expected primary payer, hospital charges, admission day of week, admission month, hospital region, hospital bed-size category, and location/teaching status of hospital. Information on hospital location, bed-size category, and teaching status were obtained from the American Hospital Association’s Annual Survey of Hospitals. Hospital census region was defined according to the U.S. Census Bureau. Within HCUP, a metropolitan statistical area is considered urban, and a nonmetropolitan statistical area is considered rural. Bed-size categories are based on the number of hospital beds, and are specific to the hospital’s location and teaching status. A hospital is considered to be a teaching hospital if it has an American Medical Association–approved residency program, and is a member of the Council of Teaching Hospitals or has a ratio of full-time-equivalent interns and residents to beds of ⱖ0.25.20 Analyses were conducted using the aggregate data for all cases identified, and by location of the event (traffic vs nontraffic). SUDAAN, version 8.0.0 (Research Triangle Institute, Research Triangle NC, 2001), was used to generate standard errors and 95% confidence intervals (CIs) around the national estimates. Cross-sectional analyses were performed for all demographic, clinical, and hospital characteristics for the entire group, and for traffic and nontraffic comparisons. Chi-square tests for categorical and unpaired t -tests for continuous variables were used to analyze differences. Unadjusted rates were calculated for hospital region using publicly available 2000 census data as population denominators. Unadjusted rates of admission by registered motorcycles were derived using publicly available vehicle registration data.1 Clinical diagnoses and procedures were grouped using Clinical Classifications Software (CCS), which aggregates individual ICD-9-CM codes into broad diagnosis and procedure groups.21
Results An estimated 30,505 (CI⫽26,566 –34,445) motorcyclerelated hospital discharges were identified nationwide in 2001. Males accounted for 89% of cases (CI⫽88.4 – 89.9). In 93% of cases, the injured patient was recorded as a motorcyclist, and 7% of cases were recorded as a passenger on a motorcycle. Approximately 62% of cases were aged ⱖ30 years; 84% were associated with traffic crashes, while 16% were classified as nontraffic crashes (Table 1). Forty-one percent of motorcycle-related cases were admitted during the weekend days, and nearly 46% of cases were admitted during the four
American Journal of Preventive Medicine, Volume 27, Number 5
Table 1. Patient characteristics and disposition for motorcycle-related hospitalizations Characteristics
n (95% CI)
Total estimated number of cases Event location Traffic Nontraffic Gender Male Female Age group (years) 0–14 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 65–69 ⱖ70 Median income of patient’s ZIP code (dollars) 0–24,999 25,000–34,999 35,000–44,999 ⱖ45,000 Disposition of patient Routine (home) Short-term hospitalization Other type of facility (includes long-term and rehab) Home health care Against medical advice Died
30,505 (26,566–34,445)
% (95% CI)
25,604 (22,128–29,080) 4,906 (4,132–5,680)
83.9 (82.0–85.9) 16.0 (14.1–18.0)
27,179 (23,649–30,708) 3,311 (2,832–3,790)
89.1 (88.4–89.9) 10.9 (10.1–11.6)
1,346 (977–1716) 2,479 (1,906–3,054) 4,302 (3,371–5,231) 3,497 (2,978–4,016) 3,830 (3,220–4,440) 3,856 (3,309–4,403) 3,461 (2,913–4,009) 2,887 (2,451–3,322) 2,291 (1,932–2,650) 1,122 (908–1,335) 723 (566–880) 351 (262–439) 361 (270–452)
4.4 (3.3–5.6) 8.1 (6.5–9.8) 14.1 (11.8–16.4) 11.5 (10.6–12.3) 12.6 (11.6–13.5) 12.6 (11.8–13.5) 11.4 (10.5–12.2) 9.5 (8.6–10.3) 7.5 (6.8–8.2) 3.7 (3.2–4.2) 2.4 (2.0–2.8) 1.2 (0.9–1.4) 1.2 (0.9–1.5)
1,115 (854–1,377) 5,736 (4,726–6,747) 8,091 (6,808–9,375) 14,912 (12,719–17,106)
3.7 (2.9–4.5) 19.2 (16.9–21.6) 27.1 (24.9–29.3) 49.9 (46.1–53.8)
23,792 (20,865–26,719) 1,043 (819–1,266) 2,839 (2,200–3,478)
78.0 (75.9–80.1) 3.4 (2.8–4.0) 9.3 (8.0–10.6)
1,978 (1,531–2,425) 236 (156–316) 606 (418–794)
6.5 (5.3–7.6) 0.8 (0.5–1.0) 2.0 (1.5–2.5)
CI, confidence interval.
months of June through September (data not shown). Two percent (CI⫽1.5–2.5) of all motorcycle-related cases (n ⫽606) died during the hospitalization. Seventy-eight percent (CI⫽75.9 – 80.1) were discharged routinely, whereas approximately 13% were discharged to another long-term or short-term healthcare facility (Table 1). The most common principal diagnosis for all cases, using CCS diagnosis categories, was fracture of the lower limb, which accounted for 29.4% (CI⫽28.1–30.8) of the diagnoses. Fractures of the upper limb accounted for 13.1% (CI⫽12.2–14.0), and intracranial injuries for 12.3% (CI⫽11.0 –13.6) of the principal diagnoses (Table 2). There were statistically significant differences in in-hospital deaths according to the principal diagnosis, with those sustaining intracranial injuries much more likely to die than those with other principal diagnoses (10.6% vs 0.8% died, p ⬍0.001). Compared with all other CCS categories, those with intracranial injury were also significantly more likely to be admitted through the emergency department, admitted on weekends, admitted to larger urban hospitals
in the Northeast, and have a higher proportion of females (p ⬍0.001). The most common principal procedure performed on motorcycle-related cases was treatment of lower limb fracture, 25% (CI⫽23.0 –26.2), and another 10% (CI⫽9.1–10.9) undergoing treatment of a fracture of the hip and femur (Table 3). An estimated 625 cases (CI⫽460 –789) had respiratory intubation and ventilation as the principal procedure, and another 3101 cases underwent intubation and ventilation as a secondary procedure (data not shown). The mean number of procedures per case was 2.5, with 79% of patients undergoing at least one procedure, and an aggregate total of 75,780 procedures associated with motorcyclerelated injuries (data not shown). The mean LOS for patients with motorcycle-related injuries was 5 days, with an estimated total of 161,094 hospital days for cases in 2001. The estimated total hospital charges for motorcycle-related injuries were ⬎$841 million, and the median hospital charge was $15,404. Hospital charges and LOS varied according to the principal diagnosis, and are detailed in Table 2. Am J Prev Med 2004;27(5)
357
Table 2. Length of stay and hospital charges Hospital charges ($) n (95% CI) Total estimated cases 30,505 (26,566–34,445) Top ten principal diagnoses Fracture of lower limb 8,980 (7,892–10,068) Fracture of upper limb 3,994 (3,398–4,590) Intracranial injury 3,753 (3,027–4,479) Other fractures 3,489 (2,908–4,070) Crushing/internal injury 3,313 (2,801–3,824) Open wounds of extremities 1,103 (897–1,308) Skull and face fractures 849 (656–1,043) Other injuries and conditions 818 (594–1,043) Superficial injury, contusion 688 (509–868) Joint disorders and 581 (455–707) dislocations Hospital location Rural Urban nonteaching Urban teaching
% (95% CI)
29.4 (28.1–30.8) 13.1 (12.2–14.0) 12.3 (11.0–13.6) 11.4 (10.5–12.4) 10.9 (9.9–11.8) 3.6 (3.1–4.1) 2.8 (2.3–3.2) 2.7 (2.1–3.3) 2.3 (1.7–2.8) 1.9 (1.5–2.3)
LOS (days)
Mean Median Total charges Mean Median 28,151 15,404
841,685,148
5.3
3.0
26,821 19,990 38,970 31,934 36,197 23,214 22,212 14,718 11,077 20,333
17,817 13,427 17,446 15,089 17,792 12,525 15,354 8,379 6,970 11,369
235,220,490 77,825,146 144,228,116 109,307,629 117,536,517 25,382,185 18,741,773 11,718,719 7,562,694 11,351,840
5.1 3.3 6.5 6.2 6.8 4.5 3.7 2.5 2.6 3.5
3.0 2.0 3.0 4.0 5.0 3.0 2.0 1.0 2.0 2.0
13,554 9,143 22,441 14,061 34,574 18,205
49,542,210 213,105,141 579,047,797
CI, confidence interval; LOS, length of stay.
Those with intracranial injury had significantly increased LOS, total hospital charges, and discharge to a long-term care facility, and were more likely to be self-pay (p ⬍0.001).
The majority of patients (56.5%, CI⫽50.6 – 62.4) were admitted to urban teaching hospitals, and most were admitted to large hospitals (Table 3). Twelve percent of cases (CI⫽9.5–14.5) were admitted to rural
Table 3. Facility characteristics, procedures, and expected payers Characteristics Hospital location/teaching status Rural Urban nonteaching Urban teaching Bed size of hospital Small Medium Large Top ten principal procedures Treatment, fracture/dislocation of lower extremity Treatment, fracture/dislocation of hip and femur Other fracture and dislocation procedure Debridement of wound Treatment, fracture, or dislocation of radius and ulna Suture of skin and subcutaneous tissue Incision of pleura, thoracentesis, chest drainage Respiratory intubation and ventilation Traction, splints, and other wound care Tracheostromy, temporary and permanent Primary expected payer Medicare Medicaid Private insurance/HMO Self pay No charge Other
n (95% CI)
% (95% CI)
3,660 (2,989–4332) 9,611 (8,110–11,113) 17,234 (13,654–20,813)
12.0 (9.5–14.5) 31.5 (26.5–36.6) 56.5 (50.6–62.4)
1,998 (1,523–2474) 7,853 (6,015–9692) 20,654 (17,202–24105)
6.5 (4.9–8.2) 25.7 (20.4–31.1) 67.7 (62.1–73.3)
5,889 (5,166–6,613)
24.6 (23.0–26.2)
2,385 (1,987–2,782)
10.0 (9.1–10.9)
1,761 (1,467–2,056) 1,554 (1,282–1,825) 1,549 (1,288–1,810)
7.4 (6.6–8.1) 6.5 (5.7–7.3) 6.5 (5.7–7.2)
1,194 (965–1,422) 1,011 (818–1,204)
5.0 (4.3–5.7) 4.2 (3.6–4.9)
625 (458–791) 582 (447–716) 452 (278–625)
2.6 (2.1–3.1) 2.4 (1.9–3.0) 1.9 (1.3–2.5)
852 (692–1,013) 2,210 (1,774–2,646) 20,366 (17,624–23,109) 4,941 (3,671–6,210) 345 (40–651) 1,595 (1,171–2,018)
CI, confidence interval; HMO, health maintenance organization.
358
American Journal of Preventive Medicine, Volume 27, Number 5
2.8 (2.3–3.3) 7.3 (6.0–8.6) 67.2 (63.8–70.6) 16.3 (12.9–19.7) 1.1 (0.2–2.1) 5.3 (4.0–6.5)
Table 4. Hospitalizations and hospitalization rates
Characteristics
n (95% CI)
United States total Hospital region Northeast Midwest South West
30,505 (26,566–34,445) 7001 (5271–8730) 4986 (3911–6061) 9816 (7216–12417) 8703 (6555–10850)
% (95% CI)
22.9 (17.8–28.0) 16.3 (12.8–19.9) 32.2 (25.6–38.7) 28.5 (22.6–34.4)
Registered motorcycles
Registered motorcycles/ 100,000 population
Admissions/ 100,000 motorcycles
10.9
4,903,056
1742
623
13.1 7.7 9.8 13.8
781,589 1,541,096 1,283,722 1,295,207
1461 2397 1281 2053
896 324 765 672
Admissions/ 100,000 population
CI, confidence interval.
hospitals. Approximately 67% of cases had private insurance or health maintenance organization as the primary expected payer, while 26% were self-pay or had public insurance as the expected payer. Hospitalization rates per 100,000 population were highest among hospitals located in the Northeast and West. Hospitalizations per 100,000 motorcycles were highest in the Northeast and South (Table 4). Comparing traffic cases with nontraffic cases revealed a number of significant differences between the
two groups (Table 5). Cases occurring in traffic were found to be older and to include a relatively higher percentage of females. Traffic cases were more likely to have increased LOS, higher total hospital charges, greater number of procedures, increased prevalence of in-hospital deaths, and increased nonroutine disposition (including transfer to long-term care facilities), and were more likely to be self-pay (all p ⬍0.001). In contrast, cases resulting from nontraffic motorcycle use were more likely to be admitted on the weekend days,
Table 5. Motorcycle crash location Number of cases (95% CI)* Age (mean years)* Length of stay (mean days)* Total charges (mean dollars)* Number of procedures (mean)* Gender (% female)** Top five principal diagnoses (% of total) Fracture lower limb Fracture upper limb Intracranial injury Other fracture Crushing or internal injury Disposition (% of total)** Routine (home) Short-term hospitalization Other type of facility (includes long-term and rehab) Home health care Against medical advice Died Payer (% of total)** Medicare Medicaid Private Self-pay No charge Other Hospital location/teaching status (% of total)** Rural Urban, nonteaching Urban, teaching
Traffic
Nontraffic
25,604 (22,128⫺29,080) 36.5 5.6 30,086 2.6 11.7
4,906 (4,132⫺5,680) 27.7 3.5 18,018 1.9 6.4
28.8 12.7 12.8 11.4 10.4
32.7 15.3 9.7 11.6 13.3
75.9 3.4 10.6
88.9 3.3 2.5
7.0 0.8 2.2
4.0 0.5 0.8
3.0 7.2 66.2 17.1 1.3 5.2
1.9 7.7 72.3 12.4 0.4 5.3
11.0 30.4 58.5
17.1 37.1 45.8
*t-test: p ⬍ 0.0001 (bolded). **Chi-square test: p ⬍ 0.0001 (bolded), indicates a significant difference in response distributions between the two groups. CI, confidence interval.
Am J Prev Med 2004;27(5)
359
during the nonsummer months, to smaller hospitals, and to rural and nonteaching hospitals. Nontraffic cases had a lower prevalence of intracranial injuries as the principal diagnosis, and a corresponding higher prevalence of extremity fractures and crushing injuries (Table 5).
Discussion These findings confirm that motorcycle-related injuries are a significant cause of hospitalization, hospital bed utilization, disability, and healthcare costs in the United States. Paramount among the findings is the uneven distribution of these cases and their resultant financial burden. Of the $842 million in total hospital charges for motorcycle-related cases in 2001, nearly 70% was from urban teaching hospitals. While 10.1% of patients used public sources of insurance, another 16.3% were uninsured. This compares with an estimated 4.6% of U.S. hospitalizations for all causes being uninsured in 2001.22 Patients with intracranial injury were more likely to be admitted to large urban teaching hospitals and these patients had longer stays, with higher total charges. They were also more likely to be self-pay. These cases likely contribute a substantial economic burden to academic medical centers. Many of the findings are consistent with prior studies and expectations. The high proportion of male patients, age distribution, day of week and month of year admission patterns, and clinical diagnoses are similar to what has been previously reported.6,23,24 However, this study provides the first national picture of the magnitude of this problem. Comparing the 30,505 estimated motorcycle-related hospitalizations with a similar analysis performed on the estimated 35,810 firearm-related hospitalizations in 1997,18 reveals that while firearm cases were more likely to die in the hospital, motorcycle cases were more likely to be discharged to a long-term care facility and incur greater hospital charges. In addition to the economic consequences, these data reveal that on average, for each day in 2001, there were approximately 25 new lower extremity fractures, 10 new intracranial injuries, and one new spinal cord injury resulting from motorcycle crashes. The procedural data provided in Table 3 only demonstrate principal procedures. The majority of patients also had a number of secondary procedures performed during their hospitalization, and the overall mean number of procedures was 2.5 per patient. When considering both principal and secondary procedures, it was found that every day approximately 208 medical procedures are performed on victims of motorcycle-related injuries, including an average of ten intubations per day. The finding that 79% of patients undergo at least one procedure further underscores the severity of these cases, as this compares with 60% among all hospitalized patients.25 360
Hospitalization rates varied by region, and interesting differences emerged when different denominators were used to examine hospitalizations. As seen in Table 4, overall hospitalization rates per population were highest in the Northeast and West. However, these crude population-based rates provide little information on exposure. Motorcycle ownership provides a better measure of exposure, and motorcycle ownership varies by geographic region, with the Midwest and West having the highest number of registered motorcycles per population (Table 4). When examining hospitalization rates according to the number of registered motorcycles, it was found that rates were highest in the Northeast and South. Numerous factors, including the number of miles driven, weather, road conditions, population density, injury patterns, helmet legislation, and local treatment practices may account for this variation. A recent analysis revealed that weather and population density were important explanatory variables for the variance in motorcycle deaths across different states.26 There were also several interesting differences between cases occurring in traffic and cases involving nontraffic crashes. While those occurring in traffic demonstrate greater severity of injury (increased LOS, greater number of procedures, increased in-hospital death rate), nontraffic cases constitute a significant proportion of the total number of admitted patients. These patients tend to be admitted to smaller rural hospitals and they are a younger population of riders. In fact, 17% of those admitted from nontraffic motorcycle use were aged ⬍16 years. This population appears to be different from those experiencing traffic crashes and may warrant different prevention strategies. A significantly higher proportion of nontraffic cases were admitted during the weekend days and during the winter months. According to ICD-9-CM coding guidelines, a motor vehicle accident is considered a traffic accident if it “occurs on a public highway” and “a motor vehicle accident is assumed to have occurred on the highway unless another place is specified.” Examples of nontraffic accidents include a crash occurring in a driveway, parking lot, farm, or other off-road location. It is suspected, but cannot be confirmed, that the popular use of motorcycles on “dirt tracks” or other off-road locations may be responsible for many of these cases. There are a number of important limitations to these data. First, NIS data are derived from hospital discharge data, which contains no information on motorcycle helmet usage or other protective devices. Therefore, no conclusions can be drawn regarding the impact of helmet use or nonuse. There is, however, a large body of previous research demonstrating both the health and cost savings associated with helmet use.6 Second, the identification of motorcycle-related cases, and the distinction between traffic and nontraffic cases relied
American Journal of Preventive Medicine, Volume 27, Number 5
on the use of E codes, and the validity of E codes in national morbidity data systems has been questioned.27 Among other considerations, E code validity is dependent on the accuracy and amount of information supplied in the medical record, accurate capture of this information and appropriate coding by medical records personnel, transfer of those codes to the discharge abstract, and inclusion of those codes in the final administrative data set. While the coding that occurs at the hospital level could not be verified, recent studies indicate that E coding within the HCUP data is quite complete, with nearly 90% of all injury diagnoses contained in the NIS having a corresponding E code.17–19 The estimate of 30,505 hospitalizations in 2001 is consistent with estimates from the NEISS, which reported 25,070 cases of hospitalized motorcyclists in 2001. The consistency of the findings with other published data on motorcycle injuries and other national estimates provides further support for the validity of E codes within HCUP data. Nevertheless, it is possible that some motorcycle cases were either not counted or misclassified as a result of coding errors. Third, the financial data provided within the NIS are based on hospital charges, not actual costs. Hospital charges are generally more than actual costs.6 However, it should also be noted that hospital charge data do not include a number of important cost items including physician professional fees, emergency transportation costs, and subsequent rehabilitation costs. In general, the reported median hospital charge of $15,404 falls within the range of hospital charges for motorcycle injuries reported in previous studies.6 Despite these limitations, findings demonstrate the significant morbidity and economic consequences associated with motorcycle-related hospitalizations. The number of motorcyclist fatalities has risen dramatically in the last 7 years. During this time, the number of states repealing mandatory motorcycle helmet use legislation has increased, and national surveys have demonstrated a reduction in helmet use from 71% in the fall of 2000 to 58% in the summer of 2002.2 While the data provide only a snapshot of the acute care hospitalizations occurring in 2001, they further illustrate the growing burden and need for preventive interventions. Recent research has demonstrated that states with full helmet laws have lower motorcyclist death rates than states without full helmet laws.26 A similar crosssectional analysis examining the relationships among helmet laws, morbidity, and hospitalization costs is warranted, and will likely provide valuable information to policymakers. We would like to acknowledge the statewide data organizations that participate in the Healthcare Cost and Utilization Project Nationwide Inpatient Sample in 2001: Arizona Department of Health Services; California Office of Statewide Health Planning and Development; Colorado Health and
What This Study Adds . . . Motorcycle crashes are a significant and rising cause of injury mortality. Despite the mortality and morbidity associated with motorcycle crashes, there are relatively few population-based studies examining hospitalizations associated with motorcycle trauma. This study provides national estimates, derived from statewide hospital discharge data from 33 states, of the prevalence, morbidity, and costs associated with motorcycle-related hospitalizations. The study also identifies a unique subset of motorcycle injuries, associated with the off-road use of these vehicles.
Hospital Association; Connecticut—CHIME, Inc.; Florida Agency for Health Care Administration; Georgia—GHA: An Association of Hospitals and Health Systems; Hawaii Health Information Corporation; Illinois Health Care Cost Containment Council; Iowa Hospital Association; Kansas Hospital Association; Kentucky Department for Public Health; Maine Health Data Organization; Maryland Health Services Cost Review Commission; Massachusetts Division of Health Care Finance and Policy; Michigan Health and Hospital Association; Minnesota Hospital Association; Missouri Hospital Industry Data Institute; Nebraska Hospital Association; New Jersey Department of Health and Senior Services; New York State Department of Health; North Carolina Department of Health and Human Services; Oregon Association of Hospitals and Health Systems; Pennsylvania Health Care Cost Containment Council; Rhode Island Department of Health; South Carolina State Budget and Control Board; Tennessee Hospital Association; Texas Health Care Information Council; Utah Department of Health; Vermont Association of Hospitals and Health Systems; Virginia Health Information; Washington State Department of Health; West Virginia Health Care Authority; and Wisconsin Department of Health and Family Services. The views expressed are those of the authors and not necessarily those of the Agency for Healthcare Research and Quality.
References 1. Federal Highway Administration. Highway statistics 2001, U.S. Department of Transportation. Available at: www.fhwa.dot.gov/ohim/hs01/index.htm. Accessed September 14, 2003. 2. National Highway Traffic Safety Administration. Traffic safety facts 2001— motorcycles. U.S. Department of Transportation. Available at www-nrd. nhtsa.dot.gov/departments/nrd-30/ncsa/. Accessed September 14, 2003. 3. Sosin D, Sacks J, Holmgreen P. Head injury associated deaths from motorcycle crashes: relationship to helmet use laws. JAMA 1990;264:1395–9. 4. National Highway Traffic Safety Administration. Traffic safety facts 1996 — motorcycles. U.S. Department of Transportation. Available at www-nrd. nhtsa.dot.gov/departments/nrd-30/ncsa/. Accessed September 14, 2003. 5. National Highway Traffic Safety Administration. Fatality analysis reporting system data, 1996. U.S. Department of Transportation. Available at: www. nhtsa.dot.gov/people/injury/pedbimot/motorcycle/safebike/approach. html. Accessed September 14, 2003.
Am J Prev Med 2004;27(5)
361
6. Pacific Institute for Research and Evaluation. Costs of injuries resulting from motorcycle crashes: a literature review. Washington DC: U.S. Department of Transportation, 2002. 7. Braddock M, Schwartz R, Lapidus G, Banco L, Jacobs L. A population-based study of motorcycle injury and costs. Ann Emerg Med 1992;21:273– 8. 8. Karlson TA, Quade CA. Head injuries associated with motorcycle use— Wisconsin, 1991. MMWR Morb Mortal Wkly Rep 1994;43:423, 429 –31 9. Max W, Stark B, and Root S. Putting a lid on injury costs: the economic impact of the California motorcycle helmet law. J Trauma 1998;45:550 – 6. 10. Rowland J, Rivara F, Salzberg P, Soderberg R, Maier R, Koepsell T. Motorcycle helmet use and injury outcome and hospitalization costs from crashes in Washington State. Am J Public Health 1996;86:41–5. 11. Shankar BS, Ramzy AI, Soderstrom CA, Dischinger PC, Clark CC. Helmet use, patterns of injury, medical outcome, and costs among motorcycle drivers in Maryland. Accid Anal Prev 1992;24:385–96. 12. National Highway Traffic Safety Administration. Report to Congress: benefits of safety belts and motorcycle helmets. Washington DC: U.S. Department of Transportation, February 1996. 13. Office of Statistics and Programming, National Center for Injury Prevention and Control, Centers for Disease Control and Prevention. Home page. Available at: www.cdc.gov/ncipc/wisqars/. Accessed April 12, 2004. 14. Helmkamp JC. A comparison of state-specific all-terrain vehicle-related death rates, 1990 to 1999. Am J Public Health 2001;91:1792–5. 15. Agency for Healthcare Research and Quality. Healthcare cost and utilization project, 1988 –2001: a federal-state-industry partnership in health data. Available at: www.ahrq.gov/data/hcup/hcup-pkt.htm. Accessed March 19, 2004. 16. Steiner C, Elixhauser A, Schnaier J. The healthcare cost and utilization project: an overview. Eff Clin Pract 2002;5:143–51. 17. Schulman J, Provenzano G, Pierce B, Wolters C, Brustrom J, Miles E. Investigation of the utility of Healthcare Cost and Utilization Project
362
18. 19.
20.
21.
22.
23. 24.
25.
26. 27.
(HCUP) data for annual surveillance of hospitalized injuries. Final report from Batelle Centers for Public Health Research and Evaluation to the Centers for Disease Control and Prevention. Atlanta GA: Batelle Centers for Public Health Research and Evaluation, September 2003. Coben JH, Steiner CA. Hospitalization for firearm-related injuries in the United States, 1997. Am J Prev Med 2003;24:1– 8. Coben JH, Owens P. Evaluation and use of HCUP emergency department data. Proceedings of the 18th Annual Meeting of the National Association of Health Data Organizations, Baltimore MD, December 2003. Agency for Healthcare Research and Quality. HCUP data documentation. Available at: www.hcup-us.ahrq.gov/db/nation/nis/dbdoc/nisdde.jsp. Accessed May 27, 2004. Agency for Healthcare Research and Quality. HCUP tools and software. Available at: www.hcup-us.ahrq.gov/tools/ccs.jsp. Accessed March 29, 2004. Agency for Healthcare Research and Quality. HCUPnet 2001 national statistics. Available at: http://hcup.ahrq.gov/HCUPnet. Accessed May 25, 2004. Aare M, von Holst H. Injuries from motorcycle and moped crashes in Sweden from 1987 to 1999. Inj Control Saf Promot 2003;10:131– 8. Peek C, Braver E, Shen H, Kraus J. Lower extremity injuries from motorcycle crashes: a common cause of preventable injury. J Trauma 1994;37:358 – 64. Elixhauser A, Klemstine K, Steiner C, Bierman AS. Procedures in U.S. hospitals, 1997. HCUP fact book number 2. Available at: www.ahrq.gov/ data/hcup/factbk2/factbk2.htm. Accessed November 6, 2003. Branas CC, Knudson MM. Helmet laws and motorcycle rider death rates. Accid Anal Prev 2001;33:641– 8. Annest JL, Mercy JA. Use of national data systems for firearm-related injury surveillance. Am J Prev Med 1998;15(suppl 3):17–30.
American Journal of Preventive Medicine, Volume 27, Number 5