Pre-Injury Driving Records of Trauma Center Patients Relative to Psychoactive Substance Use Disorders, Mode of Injury, and BAC Status

Abstract

The goal of this study was to assess pre-injury driving convictions (speeding, reckless driving, impaired driving, major license violations) among trauma center patients with respect to psychoactive substance use disorder (PSUD) diagnoses, mode of injury, and blood alcohol concentration (BAC) status at the time of admission. Driving records of 778 patients were linked to PSUDs and BAC. Twenty-nine percent of patients had a conviction in the 3 years prior to injury. Types of violation were not related to mode of injury. Although there was a positive association between prior impaired driving convictions, current alcohol dependence, and BAC+ status, a consistent pattern relative to other pre-injury convictions, PSUDs, and BAC status was not apparent.

IN HER EXTENSIVE REVIEW, Vingilis (1983) asked if drinking drivers and alcoholics are from the same population. She concluded that “…alcoholics as a group, when compared to controls, seem to be ‘high-risk’ drivers. Not only are they involved in more alcohol-related violations and collisions but, with few exceptions, many studies also find that alcoholics have more non-alcohol-related violations and collisions than the general population.”

The prevalence of alcohol use diagnoses among injured patients treated in trauma centers and emergency departments (EDs) has been documented. Using a screening test, Rivara and associates (1993) found that 45% of almost 2,500 patients treated in a Seattle trauma center were positive for possible alcoholism. More recent studies used standardized criteria to diagnose alcoholism, i.e., alcohol dependence. Among 1,161 vehicular crash victims treated in two Michigan EDs, one of which was a component of a trauma center, Maio and colleagues (1977) documented that the prevalence of current alcohol abuse or dependence (they were not differentiated) was 15% among those who were treated and released and 30% among those admitted to the hospital. Also using standardized criteria, we documented a 24% prevalence of current alcohol dependence among 1,118 trauma center patients [Soderstrom, Smith, Dischinger, et al, 1997a]. In addition, 18% of patients had current dependence on a drug other than alcohol at the time of injury. (Throughout this report, the term “drug” indicates a psychoactive substance other than alcohol.)

A number of studies have provided data relative to pre-injury driving records of patients admitted to trauma centers. In a study involving 56 highly intoxicated injured drivers (BAC ≥ 150 mg/dl), Maull and associates (1984) found that 16% had at least one previous conviction for driving under the influence and 22% had previously completed an Alcohol Safety Action Program. Further, 33% had one or more prior reckless driving convictions and 72% had at least one violation. Finally, 22% were injured while driving with a suspended license and 36% had a prior license suspension.

Lillis and others (1995) found that 15% of impaired (BAC ≥ 80 mg/dl) non-fatally injured drivers who required hospitalization in upstate New York had received a driving while intoxicated or impaired conviction in the preceding 2 years, compared with 4% of non-impaired drivers.

A Canadian regional trauma unit study [Vingilis, Stoduto, Macartney-Filgate, et al, 1994] assessed pre-injury driving records of 76 drivers who tested BAC− at the time of admission and of 22 drivers who tested BAC+. Similar percentages of BAC− and BAC+ drivers had no prior traffic violations (58.1% vs 59.1%). However, the percentage of BAC+ victims who had ≥2 violations was double that of BAC− victims, i.e., 27.3% vs 13.5%.

Finally, among motorcycle drivers treated in our center, we found that alcohol-impaired drivers had significantly higher rates of pre-injury impaired driving convictions (28.9% vs 7.2%) as well as reckless driving (68.4% vs 43.5%) and speeding convictions (73.7% vs 58.0%) than did unimpaired drivers [Soderstrom, Dischinger, Ho, et al, 1993].

The above cited studies did not link alcohol use diagnoses and pre-injury driving records. In this study of a large cohort of trauma center patients, those factors were linked. Our goal was to test Vingilis’ (1983) observation that alcoholic drivers treated in a trauma center have higher rates of impaired driving and other driving-related convictions than do non-alcoholic drivers. This hypothesis was also tested for patients with drug diagnoses.

STUDY DESIGN

CLINICAL STUDY SITE

The R Adams Cowley Shock Trauma Center, located in Baltimore City, is a regional trauma center for the most populated counties of central Maryland and a quadrant of the city. About 80% of patients are admitted from the injury scene. With the use of helicopter transport from rural/suburban settings and ambulance transport of patients injured in the city, times from injury to admission average about 1 hour.

INCLUSION/EXCLUSION CRITERIA

The center’s computerized trauma registry [Dunham, Cowley, Gens, et al, 1989] was used to identify patients who were ≥18 years age, were admitted from the injury scene, and had lengths of stay ≥2 days. Patients who met initial criteria and who had intact cognition were eligible subjects.

To obtain a representative sample of women and minorities, every other white male and all other subjects were recruited initially. When it became obvious that sufficient numbers of women and minority subjects were available, all potentially eligible patients were targeted for study. Hence, this study includes subjects from the “startup” period, who were not included in our report on the prevalence of substance use disorders, which emanated from this effort [Soderstrom, et al, 1997a].

Patients were not candidates while in intensive care units (ICUs). Patients who were initially in ICUs and those with impaired cognition were followed until they became eligible or were discharged. Patients were also not approached if their physician thought an interview would have a negative impact on their clinical course.

DIAGNOSIS OF SUBSTANCE USE DISORDERS

The psychoactive substance use disorder (PSUD) section of the Structured Clinical Interview for the DSM-III-R (SCID) [Spitzer, Williams, Gibbons, 1987] was used to make alcohol and drug diagnoses. These criterion standard diagnoses are classified as either abuse or dependence and as lifetime or current. Patients with a dependence diagnosis that was not in full or partial remission for 6 months prior to injury were considered to have current alcohol and/or drug dependence.

BAC TESTING

The admitting blood alcohol concentration (BAC) is determined to assist in clinical management. BAC results were extracted from a computerized clinical toxicology database [Soderstrom, Kufera, Dischinger, et al, 1997b].

DEMOGRAPHIC AND INJURY DATA

Demographic and mechanism of injury information were obtained from the trauma registry. Mechanisms of injury were grouped into three categories: vehicular (includes vehicular occupants and pedestrians), non-intentional (non-vehicular victims injured as the result of falls and of work related and other mishaps) and violence (shootings, stabbings, assaults).

DRIVING RECORDS

Driving records were obtained from the Driving Records Section of the Maryland Motor Vehicular Administration (Glen Burnie, MD). Maryland driving records--which are available to the general public--contain a list of convictions for the preceding 3-year period. Since records were requested in batches three times per year, they contained a list of convictions for each patient for approximately 2.75 years prior to injury. Violations were grouped into four categories: 1) speeding, 2) reckless driving (failure to stop/yield, improper passing, tailgating, etc.), 3) impaired driving (driving under the influence [DUI] of alcohol/drugs, driving while intoxicated [DWI], refusal to take a chemical test), and 4) other major license violations (driving with a revoked or suspended license, driving with a suspended license, failure to comply with Medical Advisory Board sanction/restriction, fleeing police, hit and run). In Maryland, a BAC of 70–99 mg/dl defines DUI and ≥100 mg/dl defines DWI. Despite repeated requests, not all records were obtained; however, there was no systematic bias with regard to the representativeness of available records.

INTERVIEWERS

Interviewers were a social worker, a nurse, and two psychologists, who were trained by an addiction psychiatrist and a psychologist.

CONSENT AND INTERVIEW SESSION

After intact cognition was ascertained, written consent to participate in the study was obtained. The consent form was approved by the Institutional Review Board of the University of Maryland. Candidates were informed that refusal to participate would not affect their care.

Interviews began with “ice-breaking” questions concerning previous injury and risk-taking behavior. They were followed by three alcoholism screening tests and concluded with administration of the SCID. Sessions lasted 20 to 45 minutes. Interviewers were blinded to BAC and other toxicology test results.

ANALYSIS OF RESULTS

As noted, SCID results provided alcohol and other drug diagnoses. The primary analyses involved correlating SCID diagnoses and mechanism of injury data to history of driving records. Data were analyzed using chi-square tests and t-tests.

RESULTS

STUDY PERIOD AND CONSENT

The study period was from May 1994 through November 1996. During that period, 1338 subjects were approached for study, of whom 1216 (90.9%) consented. One patient was excluded at the request of the attending physician. All of those who consented to study provided complete interview data. No significant differences were noted among consenting, non-consenting, and not approached eligible subjects relative to sex, age, race, BAC test results, Injury Severity Scores (ISS) [AAAM, 1990], and length of stay (LOS). For those interviewed, the mean ISS was 16.0 and the mean LOS was 10 days.

PATIENT PROFILE

Of the 1216 patients who completed interviews, 778 were eligible for the study, i.e., complete driving records were available. Figure 1 illustrates how the candidates for study were derived. Table 1 presents a profile of the patients relative to sex, age, race, area of residence, mode of injury, and BAC status. It also compares of the population that was included in the study and the group that was excluded.

Figure 1.

Derivation of Study Subjects. 1216 patients completed interviews of whom 778 (64%) were candidates for study.

Table 1.

Comparison of, Included and Excluded Subjects

	Included in study (N=778)	Excluded from study (N=438)	Chi- square P-value
Sex
Men	571 (73.4%)	304 (69.4%)	0.137
Women	207 (26.6%)	134 (30.6%)
Mean age (yr) (SD)	37.8 (15.2)	36.2 (16.4)	0.0941
Race
White	509 (65.4%)	175 (40.0%)	0.001
Non-white	269 (34.6%)	263 (60.1%)
Area of residence2
Baltimore	355 (45.7%)	265 (60.9%)	0.001
Rural/suburban	422 (54.3%)	170 (39.1%)
Mechanism of injury
Vehicular	440 (56.6%)	172 (39.3%)	0.001
Other non-intentional	146 (18.8%)	79 (18.0%)
Intentional/violence	192 (24.7%)	187 (42.7%)
PSUD3 diagnosis
No diagnosis	369 (47.4%)	186 (42.5%)	0.014
Curr. alcohol depend. only	125 (16.1%)	67 (15.3%)
Curr. drug depend. only	59 ( 7.6%)	59 (13.5%)
Curr. both	59 ( 7.6%)	39 ( 8.9%)
Alcohol &/or drug lifetime	166 (21.3%)	87 (19.9%)
BAC status4
Positive	244 (30.4%)	110 (26.1%)	0.363
Negative	524 (69.6%)	312 (73.9%)

¹T-test P-value (continuous variable)

²Number missing=4

³PSUD=Psychoactive Substance Use Disorder

⁴BAC testing rate=96.2%; 30 (3.9%) of included subjects and 16 (3.7%) of excluded subjects not tested

There were no differences relative to sex, age, or BAC status on admission. Exclusion from study was associated with being non-white, being a resident of Baltimore City, and being injured as a result of violence. Among Maryland residents for whom license status was known, 50% (108/214) of the Baltimore City residents and 20% (16/82) of the non-Baltimore residents who were excluded from the study never had a license.

PREVALENCE OF PSYCHOACTIVE SUBSTANCE USE DISORDERS

Four hundred nine patients (52.6%) had a lifetime PSUD. Further, 243 (59.4%) of those 409 patients had a current alcohol and/or drug dependence diagnosis at the time of injury, of whom 184 (75.7%) had an alcohol dependence diagnosis alone or concurrent with a drug dependence diagnosis.

DRIVING RECORD CONVICTIONS

The types and number of convictions received are presented in Table 2. Two hundred twenty-six subjects (29%) had at least one conviction in one or more of the violation categories (Table 3). The number of convictions total 297. For each category of violation, the number of convictions was limited to only one conviction in that category for each subject studied.

Table 2.

Distribution of Convictions of Patients in Study (N=778)

Type of Violation	Number of Convictions1
Negligent driving (reckless)	90 (11.6%)
Speeding	139 (17.9%)
Impaired driving	29 (3.7%)
License/other2	39 (5.1%)

¹Number of subjects with one or more convictions for each violation category. The violation categories are not mutually exclusive.

²31 driving with suspended or revoked license or failing to comply with Medical Advisory Board sanction, 8 fleeing an officer or hit and run

Table 3.

Comparison of Subjects With and Without Convictions in Regard to Mode of Injury, PSUD Diagnosis, and BAC Status

	At Least One Conviction (n=226)	No Convictions (n=552)	Chi- Square P-value
Mechanism of injury
Vehicular	141 (62.4%)	299 (54.2%)	0.105
Other non-intentional	38 (16.8%)	108 (19.6%)
Intentional/violence	47 (20.8%)	145 (26.3%)
PSUD diagnosis
No diagnosis	100 (44.3%)	269 (48.7%)	0.017
Curr. alcohol depend. only	43 (19.0%)	82 (14.9%)
Curr. drug depend. only	10 (4.4%)	49 (8.9%)
Curr. both	13 (5.8%)	46 (8.3%)
Alcohol &/or drug lifetime	60 (26.6%)	106 (19.2%)
BAC status1
Positive	79 (35.9%)	145 (27.5%)	0.038
Negative	141 (64.1%)	383 (72.5%)

¹6 (2.7%) of those with convictions and 24 (4.4%) of those without convictions were not BAC tested

The percentages of subjects with at least one conviction in one or more of the violation categories relative to mode of injury, PSUD diagnosis, and BAC status at the time of admission are presented in Table 3. There is no association between mechanism of injury and the rate of conviction. On the other hand, there is a significantly different distribution among those receiving convictions relative to PSUD diagnoses. The percentage of patients without any PSUD history was slightly lower in the group of patients having at least one conviction compared with the group with no convictions (44.3% vs 48.7%). Of the patients with at least one conviction, 10.2% had a current drug dependence diagnosis at the time of admission, whereas 17.2% of those without any convictions were currently drug dependent. Finally, a significantly higher percentage of patients with at least one conviction was BAC+ on admission compared with those who had no convictions (35.9% vs. 27.5%).

In Table 4, the conviction rate relative to mechanism of injury, PSUD diagnosis, and BAC status is presented. There is no association between those variables and reckless driving convictions. On the other hand, speeding convictions are associated with PSUD diagnosis and mechanism of injury, with the lowest percentages noted among individuals with current drug dependency at the time of admission. Further, although impaired driving convictions were not correlated with mode of injury, there was a strong association with current alcohol dependence and a BAC+ status. Of the 29 individuals who had at least one impaired driving conviction, 18 (62.1%) were alcohol dependent on admission. Further, 26 (89.7%) of the 29 drivers who received an impaired driving conviction were BAC+ on admission. Finally, license and other violations were associated only with a BAC+ status.

Table 4.

Specific Convictions¹ versus Mechanism of Injury, Diagnosis, and BAC Status

	N N	Reckless	Speeding	Impaired	License/Other
Mechanism of injury
Vehicular	440	12.3%	20.2%	4.3%	4.8%
Other non-intentional	146	11.0%	18.5%	2.1%	2.7%
Intentional/violence	192	10.4%	12.0%	3.7%	6.3%
		p=0.7732	p=0.044	p=0.456	p=0.324
PSUD Diagnosis
No diagnosis	369	11.4%	17.9%	0.8%	3.8%
Curr. alcohol depend. only	125	11.2%	17.6%	13.6%	6.4%
Curr. drug depend. only	59	8.5%	8.5%	0.0%	1.7%
Curr. both	59	13.6%	5.1%	1.7%	5.1%
Alcohol &/or drug lifetime	166	12.7%	25.9%	4.8%	6.6%
		p=0.910	p=0.002	p=0.001	p=0.404
BAC status
Positive	224	14.3%	17.4%	11.6%	8.9%
Negative	524	10.7%	18.5%	0.6%	3.1%
Not tested	30	6.7%	10.0%	0.0%	3.3%
		p=0.257	p=0.485	p=0.001	p=0.002

¹The conviction categories are not mutually exclusive.

²Chi-square P-value

DISCUSSION

This is the first report of an examination of pre-injury driving convictions among a large cohort of trauma center patients, comparing convictions with psychoactive substance use disorder diagnoses. Further, conviction rates are assessed relative to mechanisms of injury and BAC status on admission.

The data from the current study indicate that Vingilis’ (1983) finding that, “with few exceptions, …alcoholics have more non-alcohol-related violations …than the general population” is not true for trauma center patients. The much higher 13.6% rate of prior impaired driving convictions among current alcohol-alone dependent patients compared with a 0.8% rate for patients with no history of PSUDs is no surprise. However, there was no difference in reckless driving or speeding convictions among current alcohol dependent patients and those with no PSUD history, being 11.2% vs 11.4% and 17.6% vs 17.9%, respectively.

Although 13.8% of current alcohol-alone dependent patients had a prior impaired driving conviction, none of the current drug-alone dependent patients had a prior impaired driving conviction. Among those with both current alcohol and drug dependence, the impaired driving conviction rate was only 4.8%. The speeding conviction rates for patients with current drug-alone dependence or drug and alcohol dependence (8.5% and 5.1%, respectively) were lower than for those with no PSUD history, those with lifetime alcohol and/or drug diagnoses, and those with current alcohol-alone dependence. The lowest rates of prior speeding convictions were found among subjects with current drug-alone or alcohol and drug dependencies. Hence, drug-addicted trauma center patients did not have higher rates of pre-injury driving convictions than did patients with no PSUD history or those with current alcohol-alone dependence.

Lillis and colleagues (1995) estimated that only 3% of New York state drivers in the general population received an alcohol-related conviction in a 5-year period compared with 28.3% of injured impaired drivers after discharge from the hospital. The investigators in this study are not aware of data sources that allow for similar comparisons of overall driving record conviction rates of injured patients compared with the general population. Because of different study designs, only a few imprecise comparisons can be made relative to previous trauma center studies.

Similar to Vingilis and colleagues’ (1994) findings, the majority of trauma center patients in this study did not have prior convictions, being 71% and 58%, respectively. The 13% difference may be even less when one considers that in the Canadian study a history of convictions was queried for the previous 2 years, while the exposure of the subjects in the current study (in which 11.6% of patients had prior convictions for impaired driving) was approximately 3 years. Lillis and colleague (1995) found that 15% of injured hospitalized intoxicated drivers had impaired driving convictions in the preceding two years.

This study presents findings only for patients who sustained injuries serious enough to require trauma center admission, and who could be interviewed, i.e., those without serious brain injuries. Hence, the study involves a select group of patients. However, compared with patients with less serious injuries, this group of hospitalized patients generates a disproportionate burden of health care costs, and is potentially amenable to inpatient substance abuse interventions.

The impaired driving activity of the study’s patients is probably much higher than that reflected in their driving histories. This conclusion is predicated on the following studies and data:

1. Based on interview survey data, Liu and colleagues (1997) estimated that while 2.5% of adults admitted to alcohol-impaired driving in 1993, only 1 in 82 impaired driving episodes resulted in an arrest. This arrest rate is much lower than the 300 to 1,000 rate reported by Voas and Lacey (1989) based on roadside surveys.

2. During the period of study, the majority of drivers arrested for impaired driving were not convicted. Table 5 presents District Court data for the State of Maryland relative to arrests and legal outcome of drivers issued impaired driving citations during fiscal years 1994 and 1995. More than 98% of impaired driving arrests were alcohol-related. Of those arrested, only 32% received a conviction. If all of those who requested a jury trial were found guilty, the total found guilty would be only about 45%.

3. Studies of hospitalized and emergency department treated injured impaired drivers indicate that the vast majority are not arrested and convicted of impaired driving (range, 0 to 41%). [Maull et al, 1984; Evett, Finley, Nunez, et al, 1994; Runge, Pulliam, Carter, et al, 1996; Soderstrom et al, 1993; Soderstrom et al, 1990;Lillis et al, 1995; Barillo, 1993; Colquitt, Fielding, Cronan, 1987]. Among the patients in the current study, 51% of those with a history of alcohol dependence had at least one motor vehicular injury episode requiring emergency department treatment or hospitalization compared with 31% of those without such a history (p<0.001).

Table 5.

Impaired Driving Arrests¹ and Outcome State of Maryland - Fiscal Years 1994, 1995

	1994	1995
Number	29,826	30,648
Alcohol-related2	98.8%	98.4%
Outcome
Guilty	33.3%	29.9%
Probation before judgment	29.8%	27.0%
Nolle prosequi	18.7%	15.8%
Jury trial prayed	13.8%	13.1%
Other	4.0%	13.9%
Not guilty	0.4%	0.3%

¹Citations were for driving while intoxicated (DWI, BAC ≥ 100 mg/dl) or driving under the influence of alcohol (DUI, BAC 70–99 mg/dl), or driving while under the influence of a drug, drug and alcohol combined, or under the influence of a controlled substance.

²For FYs 1994 and 1995, 82.5% of alcohol-related arrests were for DWI.

Impaired drivers who sustained injury requiring hospital care rarely are arrested and convicted for the driving violation that caused the crash.

The importance of identifying drivers at risk of alcohol dependence is emphasized by Brewer and colleagues’ (1994) study of fatally injured drivers with BACs ≥20 mg/dl. They found that those drivers compared to fatally injured control drivers (BAC <20 mg/dl) were more likely (26% vs 3%) to have had one or more prior impaired driving arrests. In fact, of those with a prior arrest, 72% had only one arrest. Hence, a history of even one arrest, and even more so in light of the above discussion, a history of a prior conviction, should prompt an evaluation of an alcohol use diagnosis among injured impaired and unimpaired drivers.

SUMMARY

Except for a clear association between impaired driving convictions and current alcohol dependence and a BAC+ status, no consistent pattern was noted relative to types of convictions, BAC status, and substance use disorders diagnoses. All trauma center patients should be considered at risk of an alcohol or drug problem, particularly those who: 1) screen positive on substance use questionnaires [Soderstrom, Smith, Dischinger, et al., 1997c; Soderstrom, Gorelick, Carriero, et al., 1998], 2) screen positive for pre-injury alcohol or other drug use, or 3) have a history of impaired driving convictions. An individual meeting any of these criteria should have a formal evaluation for an underlying substance use disorder by a chemical dependency clinician.