PEDIATRICS Vol. 107 No. 4 April 2001, pp. 632-637

A Population-Based Study of Crashes Involving 16- and 17-Year-Old Drivers: The Potential Benefit of Graduated Driver Licensing Restrictions

Natalie Z. Cvijanovich, MD, Lawrence J. Cook, Mstat, N. Clay Mann, PhD, MS, and J. Michael Dean, MD, MBA

From the Intermountain Injury Control Research Center, Department of Pediatrics, University of Utah School of Medicine, Salt Lake City, Utah.

	ABSTRACT

Top Abstract Methods Results Discussion Conclusion References

Objective.  To evaluate the potential effectiveness of graduated driver licensing programs using population-based linked data for motor vehicle crashes (MVCs) that involved teenaged drivers (TDs).

Methods.  Utah crash, inpatient hospital discharge, and emergency department databases were analyzed and probabilistically linked. We computed hospital charges and compared violations, contributing factors, seatbelt use, and passengers for TDs (16-17 years old) relative to adult drivers (18-59 years old).

Results.  TDs comprised 5.8% of the study population, but were involved in 19.0% of MVCs. TD crashes resulted in $11 million in inpatient hospital charges and 158 fatalities. TD crashes were 1.70 times (95% confidence interval [CI]: 1.34, 2.04) less likely to result in fatal injury to drivers than were crashes that involved adult drivers, but TDs were 2.20 times (95% CI: 1.96, 2.47) more likely to receive citations. The following were findings of the study: 1) 11% of all TD crashes but 19% of fatal TD crashes occurred between 2200 and 0600 hours; 2) TDs used seatbelts less often than did adult drivers (79.1% vs 84.4%) and less often with passengers present (81.9% vs 75.0%; 3) TDs were 1.72 times (95% CI: 1.38, 2.14) more likely to be involved in crashes that resulted in seriously or fatally injured occupants when driving with passengers than when driving alone.

Conclusions.  TDs are overrepresented in MVCs. TD crashes have a higher fatality rate at night, and TDs wear seatbelts less often than do adult drivers. Passengers affect TD crash characteristics. Graduated driver licensing programs that target state-specific characteristics of TDs may decrease morbidity and mortality.  Key words:  teenaged drivers, graduated driver licensing, provisional licensing, probabilistic linkage.

In recent years, the public health risk associated with novice drivers has received an increasing amount of attention. Teenaged drivers (TDs) in particular are known to be overrepresented in motor vehicle crashes (MVCs).^1-7 Factors that contribute to the increased crash rate among TDs have been studied. These include not only inexperience, which is independent of age,⁸ but also age-dependent factors that are unique to teenagers. For example, the TD may not yet have fully developed decision-making abilities and judgment.¹ TDs perceive less risk in most situations than do older drivers, overestimate their own skills, and consider themselves less vulnerable in the event of a crash.^8-16 Furthermore, the tendency to take risks may be increased by peer pressure, emotional lability, and other stresses. Finally, teenagers drive more frequently under higher risk conditions (ie, at night and/or without seatbelts).¹

The American Academy of Pediatrics, the National Highway Traffic Safety Administration, and other governmental, private, and professional organizations have recommended legislation to institute a program of graduated driver licensing (GDL). These programs remove driving restrictions in stages as the TD gains experience and maturity, until full licensure is granted a minimum of 18 months after the learner's permit is obtained.¹⁷ This type of program has been shown to decrease the crash rate of TDs in New Zealand, Canada, and several states in the United States.^18-22 Each of these programs incorporates different features of the National Highway Traffic Safety Administration recommendations, yet it has not been well established which components are the most effective. In the present study, we determined the population-based rate of crashes that involved TDs in a single state, hospital charges associated with all people involved in those crashes, and the characteristics of these crashes to determine the potential effectiveness of various components of GDL programs. We hypothesized that TDs are overrepresented in crashes, especially at night, that TDs wear seatbelts less often, and that the presence of passengers significantly increases the risk of a crash that results in fatal or nonfatal injury.

	METHODS

Top Abstract Methods Results Discussion Conclusion References

Data Sources

Four data sets were used in these analyses: 1) Utah MVC records; 2) the Utah hospital discharge database; 3) the Utah Emergency Department database; and 4) the Utah Division of Motor Vehicles driver license file. The 1992 to 1996 MVC records were obtained from the Utah Department of Transportation, Division of Traffic and Safety. This crash database contains information on all reportable crashes (defined as occurring on a public road and resulting in at least 1 injury or at least $1000 in property damage). These data are collected on reports that are filled out by police officers at the scene of the crash. Data points include detailed information on the time, location, type of crash, vehicles, drivers, and any other people involved (eg, passengers, pedestrians).

The 1992 to 1996 hospital discharge database was acquired from the Utah Health Data Committee/Office of Health Data Analysis, Utah Department of Health. The Utah Department of Health mandates that all Utah licensed hospitals report all inpatient admissions. The hospital discharge database includes demographic information and up to 9 International Classification of Diseases-Ninth Revision diagnosis codes and 5 procedure codes, up to 2 external-cause-of-injury codes (E codes), discharge status, and hospital charges.

Emergency department (ED) records for 1996, the first year for which they were complete, were acquired from the Utah Department of Health. The Department of Health mandates that all Utah licensed hospitals report information on ED patient encounters. The ED database includes demographic information, up to 9 International Classification of Diseases-Ninth Revision diagnosis codes and 5 procedure codes, up to 2 E codes, discharge status, and hospital charges. Patients in this database were by definition treated and released from the ED.

The driver license file includes all information required for a driver's license, as well as the original and current issue dates and expiration date. The Institutional Review Board of the University of Utah approved use of these databases for this study.

Probabilistic Linkage

The database that was used for analysis was created by probabilistically linking the crash and inpatient databases for 1992 to 1996. ED data were included in the 1996 linkage. Jaro²³ described probabilistic linkage. By comparing several common fields, probabilistic linkage uses an iterative approach to mathematically linking databases. The variables used for linkage in this study included date of incident, date of birth, hospital code, county code, city code, gender, and age. Successful linkage is affected by errors in the data sources as well as the effectiveness of the method used to achieve the linkage. Because the crash database includes noninjured persons, a substantial portion of the crash database is expected not to link with an inpatient or ED record.

To eliminate 16- or 17-year-old teenagers who were not yet licensed, as well as older unlicensed drivers, the driver license file was linked to the crash database and any unmatched drivers were eliminated from our analyses. Probabilistic record linkage was performed using Automatch 4.2 (Matchware Technologies, Inc, Boston, MA).

Measurements

To study the group of inexperienced drivers that would be most affected by interventions such as a GDL program, we compared 16- to 17-year-old drivers with 18- to 59-year-old drivers. Sixteen- to 17-year-old drivers are referred to as TDs, and 18- to 59-year-old drivers are referred to as adult drivers. A person who is "hospitalized" is one who is probabilistically linked to a hospital discharge record, and a person who is seen in the ED is one who is linked only to an ED record. Information from the hospital databases allowed us to compute hospital charges.

Fatalities were determined by a KABCO score, as assigned by the police officer who completed the accident report. "K" is killed, "A" is incapacitated, "B" is probably injured, "C" is possibly injured, and "O" is uninjured.

Statistical Analysis

Descriptive analysis of crashes and occupants and odds ratios with 95% confidence intervals (CIs) were computed. A ² test was used to test the association between study population and severity of crash, driver at fault, proportion of nighttime crashes, seatbelt use, and number of passengers. The number of passengers was determined by counting the actual number of occupants in each vehicle in each record in the crash file, rather than by using the number recorded by the investigating officer in the crash record, which may be inaccurate. We performed multiple logistic regression to evaluate the association of various environmental factors with crashes, comparing TDs with adult drivers. Factors included driving after dark, driving on dry road, driving on a weekend, single vehicle crash, seatbelt use, hospitalization of any occupant, and fatality of any occupant. Statistical analyses were performed using SAS (SAS Institute, Inc, Cary, NC).

	RESULTS

Top Abstract Methods Results Discussion Conclusion References

In Utah from 1992 to 1996, among licensed drivers from 16 to 59 years old, there were 217 024 crashes that involved 341 383 vehicles and 496 487 occupants. Based on licensing information, TDs make up 5.8% of the 16- to 59-year-old driving population but were involved in 19% (41 249) of all crashes in this population. Crashes that involved TDs resulted in 158 fatalities and $11 million in inpatient hospital charges. In 1996 alone, crashes that involved TDs resulted in $1.77 million in ED charges for patients who did not require hospitalization.

Of the 158 people who were killed in crashes that involved TDs, 63% (n = 99) were in a TD's car. Twenty-five percent (n = 40) of the fatalities were the TD. Of all teens (13- to 19-year-olds) who were killed in crashes during the study period, 85% (n = 84) were in cars of TDs.

To compare the crash conditions of adult drivers versus TDs, we performed multiple logistic regression on the crash file (Table 1). Interestingly, crashes that involved TDs were less likely to result in an occupant fatality than were those that involved adult drivers. However, TDs were more likely to be involved in single vehicle crashes and crashes that occurred at night. In addition, occupants of cars that were driven by teenagers were significantly less likely to be wearing seatbelts than were those with adult drivers.

Many GDL systems restrict nighttime driving among teenagers, usually between the hours of 2200 and 0600. To evaluate the effect of such a restriction in our study population, we studied the distribution of crashes throughout the day. Figure 1 shows the distribution of all crashes, both fatal and nonfatal, by hour of day among teenage and adult drivers. The pattern for TDs is similar to that for adult drivers, although slightly more crashes occur among TDs during after-school and evening hours. Figure 2 depicts the hourly fatality rate of crashes for each age group. In accordance with our regression model, adult drivers demonstrate a higher fatality rate than TDs in general, including between the hours of 2200 and 0600, when most nighttime driving restrictions would be in effect. However, the fatality rate among TDs is still highest during that time period, especially between 2400 and 0600 hours. Finally, Fig 3 depicts the distribution of all fatal crashes throughout the day. Although only 11% of all crashes that involved TDs occurred between 2200 and 0600, 19% (n = 23) of fatal crashes that involved TDs occurred during these hours.

View larger version (30K): [in this window] [in a new window]   Fig. 1.   Percentage of all crashes by hour of day.  Adult drivers.  TDs.

View larger version (24K): [in this window] [in a new window]   Fig. 2.   Fatality rate for crashes by hour of day. The y axis is the number of fatalities per 100 crashes for each driver age group at each hour.  Adult drivers.  TDs.

View larger version (33K): [in this window] [in a new window]   Fig. 3.   Distribution of fatal crashes by hour of day. The y axis is the percentage of all fatal crashes for each driver age group at each hour.  Adult drivers.  TDs.

In our regression model, TDs were less likely to wear seatbelts than were adult drivers. Self-reported seatbelt use among TDs was 79.1%, compared with 84.4% among adult drivers (P < .001). We also found that seatbelt use among TDs was related to the number of passengers in the vehicle (belted alone, 81.9%; with passengers, 75.0%; P < .001).

It has been shown that the presence of passengers has variable effects on crash risk depending on driver and passenger age.^9,24 To evaluate the effect of imposing restrictions on the number and/or age of passengers carried by TDs, we determined whether the risk of serious injury when passengers were present was higher for TDs than for adult drivers. Because the risk of injury is higher as a result of the greater number of people being subjected to the force of a crash when passengers are present, we compared odds ratios for hospitalization or fatality of the driver alone for TDs and adult drivers. The results are shown in Table 2. The presence of passengers increased the risk of hospitalization or fatality to the driver regardless of age group, but the increase was substantially higher among TDs. It is interesting to note that among adult drivers, there was a trend toward decreased risk with increasing number of passengers (up to 3), whereas among TDs, the risk steadily increased as the number of passengers increased.

To assess whether this increased crash risk was attributable to different driving behavior, as has been suggested in previous studies,^9,24 we used the police officer assignment of an at fault citation as an indicator of risky behavior. Overall, TDs were 2.20 times (95% CI: 1.96, 2.47) more likely to be at fault, defined as receiving a moving violation, than adult drivers. The presence of passengers in the vehicle increased the likelihood of receiving a citation among TDs, whereas it decreased the likelihood among adult drivers (Table 3). Because police officers may be biased against TDs and thus be more likely to consider them at fault, we also determined the likelihood of receiving a citation for speeding or reckless driving for TDs as compared with adult drivers as a more objective measure of responsibility for the crash (Tables 4 and 5). Overall, TDs were 1.43 times (95% CI: 1.35, 1.50) more likely than adult drivers to receive a speeding citation and 2.83 times (95% CI: 2.44, 3.27) more likely than adult drivers to be cited for reckless driving. The presence of passengers exacerbated this problem for TDs. The effect was most striking when there were 4 or more passengers present. Among adult drivers, there was little to no effect.

	DISCUSSION

Top Abstract Methods Results Discussion Conclusion References

In this population-based study, we showed that TDs are involved in a disproportionately high number of MVCs. Sixteen- and 17-year-old drivers accounted for <5% of the study's driving population but were involved in 19% of all MVCs. Crashes that involved TDs had a significant economic impact, directly resulting in $11 million in inpatient hospital charges over the 5-year study period.

TDs are more likely than adult drivers to be involved in a crash, but crashes that involve TDs generally are less severe than those that involve adult drivers. This may be because TDs more often have minor crashes than adult drivers. In addition, the risk of fatality increases simply as a factor of increasing age, so younger drivers (ie, TDs and their passengers, who tend to be young) can be expected to have a lower fatality rate.

Our data can be used to tailor interventions that incorporate the most effective restrictions on TDs. During the study period, Utah teens were able to obtain a learner's permit at a minimum age of 15 years 9 months. The permit is valid for 6 months, with required supervision by a licensed driver of any age and no minimum permit period. An unrestricted license could be obtained at age 16 with evidence of a previous learner's permit. During the study period, there were no nighttime restrictions, seatbelt requirements, or passenger restrictions. Our study demonstrates that a significant reduction in crashes and fatalities could be achieved by implementing these restrictions. A nighttime driving restriction from 2200 to 0600 hours, for example, may prevent 19% of fatal crashes and 11% of all crashes among TDs. Requiring seatbelt use among vehicle occupants would also have an impact. However, the single most effective restriction may be limiting the number of passengers, especially teenage passengers, in a car that is driven by a TD. The presence of any passengers in a car that is driven by a TD increases by 70% the risk of a crash that results in fatal or serious injury. These results are consistent with recently published results from several national databases.²⁵ In addition, TDs are 2.5 times more likely to be cited for reckless driving and wear seatbelts less often when passengers are present.

Only 2.8% of crashes that involved TDs occurred when there was a passenger who was >21 years old. The remaining 97.2% occurred when the drivers were alone or with passengers who were <21 years old. This may be because of very different exposure rates among teenagers who drive with passengers (ie, TDs, once they have their unrestricted licenses, rarely drive with adults as passengers). However, it also may be true that a passenger who is >21 years old positively influences the driving behavior of TDs.

Whereas previous studies have focused on societal costs of MVCs that involve TDs,^2,26-28 our study is unique in that we report hospital charges incurred as a result of these crashes. This is possible because of the linkage of crash files to hospital data, allowing a determination of actual hospital charges associated with each crash. This does not include physician charges or charges incurred after release from the hospital, such as home health services, physical therapy, rehabilitation services, and so forth. Nevertheless, crashes that involved TDs resulted in nearly $13 million in inpatient hospital charges alone. In addition, if ED charges from 1996 are extrapolated to the previous 4 years, nearly $10 million can be added to the total charges for crashes that involved TDs.

Several potential limitations may affect our results, and therefore warrant consideration. First, the higher rate of citations for speeding and reckless driving among TDs may be attributable, in part, to an inherent bias against TDs among police officers. Second, seatbelt use is self-reported; therefore, actual usage rates are probably significantly lower in nonfatal crashes. However, there is little reason to believe that the bias in self-reported seatbelt use is systematically different among TDs compared with adult drivers or between TDs when driving alone and driving with passengers. Finally, the actual crash rate may be underreported because crashes that occur on private property (eg, parking lots, driveways, sidewalks) are not included in the Utah crash data set.

	CONCLUSION

Top Abstract Methods Results Discussion Conclusion References

This population-based study of TDs shows that they are overrepresented in MVCs. Crashes that involve TDs have the highest fatality rate at night and are more often the result of driver error than are those that involve adult drivers. TDs are less likely than adult drivers to wear seatbelts. Furthermore, the presence of passengers increases the severity of crashes that involve TDs as well as the tendency of TDs to be cited for speeding or reckless driving. Finally, injuries sustained in crashes that involve TDs have a significant economic impact. The results of this study can be used to support interventions that are targeted specifically to TDs, such as programs of GDL, which can be tailored to the characteristics that are particular to the population studied.

	FOOTNOTES

Received for publication Nov 8, 1999; accepted Aug 25, 2000.

Reprint requests to (N.Z.C.) Intermountain Injury Control Research Center, 410 Chipeta Way, Suite 222, Salt Lake City, UT. E-mail: natalie.cvijanovich{at}hsc.utah.edu

	ABBREVIATIONS

TD, teenaged driver; MVC, motor vehicle crash; GDL, graduated driver licensing; ED, emergency department; CI, confidence interval.

	REFERENCES

Top Abstract Methods Results Discussion Conclusion References

1. American Academy of Pediatrics, Committee on Injury and Poison Prevention and Committee on Adolescence The teenage driver. Pediatrics 1996; 98:987-990 [Abstract/Free Full Text]
2. Centers for Disease Control Involvement by young drivers in fatal motor-vehicle crashesUnited States, 1988-1995. MMWR Morb Mortal Wkly Rep 1996; 45:1049-1053 [Medline]
3. Ferguson SA, Leaf WA, Williams AF, Preusser DF Differences in young driver crash involvement in states with varying licensure practices. Accid Anal Prev 1996; 28:171-180 [CrossRef][Medline]
4. Liu JY, Mooney DP, Meyer MM, Shorter NA Teenage driving fatalities. J Pediatr Surg 1998; 33:1084-1088 [CrossRef][Medline]
5. Massie DL, Campbell KL, Williams AF Traffic accident involvement rates by driver age and gender. Accid Anal Prev 1995; 27:73-87 [CrossRef][Medline]
6. Phebo L, Dellinger AM Young driver involvement in fatal motor vehicle crashes and trends in risk behaviors, United States, 1988-95. Inj Prev. 1998; 4:284-287 [Abstract/Free Full Text]
7. Williams AF Overview of the young driver problem in the United States. Transportation Res Circ 1996; 458:6-8
8. Comsis C. Understanding Youthful Risk-Taking and Driving. Interim Report. Washington, DC: US Department of Transportation, National Highway Traffic Safety Administration; 1995. Report No. DOT HS 808 318
9. Doherty ST, Andrey JC, MacGregor C The situational risks of young drivers: the influence of passengers, time of day and day of week on accident rates. Accid Anal Prev 1998; 30:45-52 [CrossRef][Medline]
10. Farrow JA Young driver risk taking: a description of dangerous driving situations among 16- to 19-year-old drivers. Int J Addict 1987; 22:1255-1267 [Medline]
11. Centers for Disease Control Risky driving behaviors among teenagers-Gwinnett County, Georgia, 1993. MMWR Morb Mortal Wkly Rep 1994; 43:405-409 [Medline]
12. Gregersen NP, Bjurulf P Young novice drivers: toward a model of their accident involvement. Accid Anal Prev 1996; 28:229-241 [CrossRef][Medline]
13. Gregersen NP, Berg HY Lifestyle and accidents among young drivers. Accid Anal Prev 1994; 26:297-303 [CrossRef][Medline]
14. Jonah BA Accident risk and risk-taking behaviour among young drivers. Accid Anal Prev 1986; 18:255-271 [CrossRef][Medline]
15. Trankle U, Gelau C, Metker T Risk perception and age-specific accidents of young drivers. Accid Anal Prev 1990; 22:119-125 [CrossRef][Medline]
16. Vavrik J Personality and risk-taking: a brief report on adolescent male drivers. J Adolesc 1997; 20:461-465 [CrossRef][Medline]
17. National Highway Traffic Safety Administration. Graduated Driver Licensing System for Young Novice Drivers. Washington, DC: US Department of Transportation, National Highway Traffic Safety Administration; 1996. Report No. DOT HS 808 331
18. Foss RD, Evenson KR Effectiveness of graduated driver licensing in reducing motor vehicle crashes. Am J Prev Med 1999; 16:47-56 [CrossRef][Medline]
19. Begg DJ, Langley JD, Reeder AI, Chalmers DJ The New Zealand graduated driver licensing system: teenagers' attitudes toward and experiences with this car driver licensing system. Inj Prev 1995; 1:177-181 [Abstract/Free Full Text]
20. Boase P, Tasca L. Graduated Licensing System Evaluation. Interim Report. Ottawa, Ontario, Canada: Safety Policy Branch, Ministry of Transportation; 1998. Report No. 410 SPB-98-101
21. Langley JD, Wagenaar AC, Begg DJ An evaluation of the New Zealand graduated driver licensing system. Accid Anal Prev 1996; 28:139-146 [CrossRef][Medline]
22. National Highway Traffic Safety Administration. Saving teenage lives: the case for graduated driver licensing. http://www.nhtsa.dot.gov/people/injury/newdriver/SaveTeens/toc.html
23. Jaro MA Advances in record-linkage methodology as applied to matching the 1985 census of Tampa, Florida. J Am Stat Assoc. 1989; 84:414-420 [CrossRef]
24. Preusser DF, Ferguson SA, Williams AF The effect of teenage passengers on the fatal crash risk of teenage drivers. Accid Anal Prev 1998; 30:217-222 [CrossRef][Medline]
25. Chen LH, Baker SP, Braver ER, Li G Carrying passengers as a risk factor for crashes fatal to 16- and 17-year-old drivers. JAMA 2000; 283:1578-1582 [Abstract/Free Full Text]
26. Goldstein LA, Spurlock CW, Kidd PS Economic costs of motor vehicle crashes involving teenaged drivers in Kentucky, 1994. Inj Prev. 1997; 3:200-206 [Abstract/Free Full Text]
27. Miller TR, Lestina DC, Spicer RS Highway crash costs in the United States by driver age, blood alcohol level, victim age, and restraint use. Accid Anal Prev 1998; 30:137-150 [CrossRef][Medline]
28. Moore M Comparison of young and adult driver crashes in Alaska using linked traffic crash and hospital data. Alaska Med 1997; 39:95-102 [Medline]