AJCC
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

American Journal of Critical Care. 2002;11: 106-114
Copyright © 2002 by the American Association of Critical-Care Nurses.
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Respond to This Article
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sommers, M. S.
Right arrow Articles by Fleming, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sommers, M. S.
Right arrow Articles by Fleming, M.

"Nurse, I Only Had a Couple of Beers": Validity of Self-reported Drinking Before Serious Vehicular Injury

By Marilyn Sawyer Sommers, RN, PhD, CCRN, Janice M. Dyehouse, RN, PhD, Steven R. Howe, PhD, Kathryn Wekselman, RN, PhD and Michael Fleming, MD, MPH. From the College of Nursing (MSS, JMD) and Department of Psychology (SRH), University of Cincinnati, and Procter and Gamble (KW), Cincinnati, Ohio, and Department of Family Medicine, University of Madison-Wisconsin, Madison, Wis (MF).


   Abstract
Top
 Abstract
Background
 Research Questions
 Methods
 Results
 Discussion
 Recommendations for Clinical...
 Summary
 References
 
• Background Of the more than 40 000 Americans killed each year in vehicular crashes, 40% are involved in alcohol-related collisions. Although self-reported alcohol use has become an anchor for alcohol intervention after traffic crashes, clinicians are often skeptical about the truthfulness of self-reporting.

• Objective To determine the validity of self-reported alcohol consumption of vehicular occupants hospitalized for a serious, alcohol-related injury.

• Methods Non-alcohol–dependent subjects 18 years and older who were injured in motor vehicle crashes were interviewed. The self-reported number of standard drinks, time that drinking commenced, sex, and weight were used to calculate estimated blood alcohol concentration. This value was compared with the blood alcohol concentration measured at admission.

• Results Of the 181 subjects, 60% provided sufficient data to calculate the estimated concentration. Seven men with admission concentrations of 10 mg/dL or more denied drinking. Among the 113 subjects with estimated concentrations who acknowledged drinking (excluding the 7 who denied drinking), the mean concentration at admission was 158.67 mg/dL, and mean estimated concentration was 83.81 mg/dL. According to multiple regression analyses, weight and number of drinks accounted for 3% of the variance in alcohol concentration at admission for women (R =0.174, F2,40 = 0.623, P = .54) and for 29% of the variance in men (R=0.543, F2,128 =26.71, P< .001).

• Conclusions Most persons who drink before vehicular injury acknowledge drinking. Self-reported data from men generally reflect the overall trend of alcohol consumption but with systematic underreporting. Reports from women are less predictable.

Clinicians in healthcare settings and police in arrest situations are familiar with a stereotypical response when an alcohol-impaired driver or passenger is asked about drinking: "I just had a couple of beers, nurse" (or doctor, or officer).1 Babor et al2 wrote of the controversy between clinicians and research scientists about the validity of self-reported alcohol use. Persons involved in alcohol research generally think that verbal reporting procedures are valid and reliable under most circumstances. On the other hand, many clinicians who deal with alcohol-dependent or high-risk drinkers in clinical settings think that self-reports of drinking are untrustworthy.2


   Background
Top
 Abstract
 Background
Research Questions
 Methods
 Results
 Discussion
 Recommendations for Clinical...
 Summary
 References
 
Role of Self-reported Alcohol Use After Injury
Although self-reported alcohol use is a necessary component of alcohol screening, several obstacles must be overcome before such screening becomes routine after serious alcohol-related vehicular injury. Studies3,4 indicate that the majority of physicians who care for trauma patients would support implementation of an intervention program on the physicians’ services if screening and intervention were effective in reducing drinking. However, only 76% of trauma surgeons were familiar with common alcohol screening instruments, and 83% reported no training in alcohol screening.3 Another obstacle to alcohol screening is concern about denial of insurance coverage if screening information appears in a patient’s record. In a study5 of the insurance commissioners in all 50 states, investigators at the Harborview Injury Prevention and Research Center in Seattle found that 38 states had statutes that allowed denial of coverage for injuries sustained during impairment due to alcohol. Finally, the concern of nurses and physicians that patients will not provide a truthful self-report may deter the implementation of intervention programs.1,2

Despite these obstacles, an increasing number of scientists are using self-reported alcohol consumption to tailor brief interventions for trauma patients.6–10 Brief interventions are counseling sessions that include the following: assessment of and direct feedback on drinking, contracting and goal setting to reduce the risk of problem drinking, behavioral modification techniques, and the use of written material such as a self-help manual.11 These interventions are based on the patient’s self-reported drinking patterns. In short, self-reported alcohol consumption is used to tailor interventions directed at changing problem drinking and reducing risk for further injury. A valid self-report, therefore, may be the linchpin for a successful, brief intervention.

Investigators have studied the validity of self-reports in patients with a wide range of drinking patterns and in a variety of settings.12–15 Little is known, however, about self-reports of alcohol consumption obtained after hospitalization for serious vehicular injury.

Alcohol, Motor Vehicle Crashes, and the Need for Intervention After Serious Injury
Motor vehicle crashes are the leading cause of death and disability due to unintentional injury in the United States.16–18 The cost in lives lost and dollars expended is extraordinarily high. In 1998, more than 41 000 persons in the United States died in traffic crashes, and 2.2 million received disabling injuries. Approximately 40% of all traffic fatalities involve an intoxicated or alcohol-impaired driver or nonmotorist. In addition, alcohol is involved in 7% of all fatal and nonfatal traffic collisions.17,18 The National Safety Council estimated that in 1998 the cost of alcohol-related motor vehicle crashes was $26.9 billion.17

Persons with alcohol-related vehicular injury, like all persons with unintentional injury, have a higher risk for repeated injury than do members of the general population. In one study,19 the likelihood of readmission for trauma patients who required hospitalization was 2.5-fold greater for patients who were intoxicated than for patients who were not. A serious injury, therefore, offers a window of opportunity or a "teachable moment" for healthcare providers and patients to discuss strategies for injury prevention.

Validity of Self-reported Alcohol Use
Most research supports self-reported data as a valid basis for measuring alcohol intake. Sobell et al20 noted that self-reports in persons who misuse alcohol are generally truthful so long as the interviews are conducted in a confidential clinical research setting and the person being interviewed is alcohol-free at the time of the interview. Specific features in the interview that can enhance validity and reliability include the use of (1) detailed but separate questions about quantity, frequency, and variability of drinking; (2) questions of an incremental nature so that answers to each part of the questions are summed to arrive at a final calculation; (3) a day (24-hour) period of reference; and (4) the concept of a "standard drink."21 Other studies also support the validity and reliability of self-reports. Brown et al22 found a 97% agreement between self-reported alcohol use and laboratory data among subjects with alcohol dependence. In a group of adolescents, Smith et al23 found a correlation of 0.62 for the total amount of alcohol consumed when the adolescents’ self-reports were compared with the reports of collateral informants such as friends or family members.

Several studies by Cherpitel identified the accuracy of self-reported alcohol use among injured patients in the emergency department. In a 1994 review of emergency department studies, she noted that a larger proportion of patients acknowledged drinking within the hours prior to the event than actually had alcohol detected by laboratory analysis of blood or breath.12 She also found that self-reports of drinking before injury were accurate when compared with the results of breath analysis. Specifically, Cherpitel found that only 3.6% of all patients with positive breath analyzer readings reported no drinking before admission to the emergency department, whereas the remainder admitted alcohol use.24

Patients who are hospitalized after a vehicular injury present clinicians with a unique set of issues during assessment and intervention for high-risk drinking. The patients are usually anxious about their injury and are often in pain. If they were driving at the time of the crash, they may also be concerned about prosecution for drinking and driving. The literature supports the validity of self-reports of alcohol use in both injured and noninjured patients, but little is known about the truthfulness of hospitalized patients in acknowledging the drinking that occurred before a vehicular crash.


   Research Questions
Top
 Abstract
 Background
 Research Questions
Methods
 Results
 Discussion
 Recommendations for Clinical...
 Summary
 References
 
The purpose of this study was to determine the validity of self-reported alcohol consumption after a serious injury in a motor vehicle crash in subjects with a blood alcohol concentration (BAC) of 10 mg/dL or more at the time of hospital admission. Therefore, we asked the following question with regard to subjects after serious, alcohol-related vehicular crashes: What is the relationship between estimated BAC (EBAC) based on self-reported alcohol consumption and the BAC determined by laboratory tests of blood samples obtained at the time of admission?


   Methods
Top
 Abstract
 Background
 Research Questions
 Methods
Results
 Discussion
 Recommendations for Clinical...
 Summary
 References
 
Subjects
We recruited subjects from 2 level I trauma centers in southwestern Ohio who were admitted to the hospital after a motor vehicle crash. We collected self-reported drinking and laboratory data as part of a baseline assessment for a randomized clinical trial funded by the Centers for Disease Control and Prevention. The trial tested the effectiveness of brief intervention strategies after a motor vehicle crash in reducing alcohol consumption and adverse events in adults 18 to 45 years old. We obtained informed consent for the initial interview and all follow-up interviews for the clinical trial. Informed consent procedures were consistent with those outlined by our institutional review board for enrollment and randomization of subjects into the clinical trial. Eligible patients were admitted within 24 hours of injury and had a BAC of 10 mg/dL or greater at the time of admission. Other inclusion criteria were age 18 to 45 years, ability to speak English, and intact cognition as judged by nurse clinicians upon physical assessment and chart review.

Because the randomized clinical trial focused on non-alcohol–dependent patients, we screened potential subjects for alcohol dependence and referred the subjects for further assessment of dependence if they met one or more of the following criteria: (1) score of 2 or higher on any or all of items 4, 5, and 6 on the Alcohol Use Disorders Identification Test25,26; (2) participation in an alcohol treatment program within the preceding year; (3) evidence of signs or symptoms of alcohol withdrawal upon physical examination and chart review; and (4) history of daily alcohol use exceeding 150 g of alcohol per day. We did not enroll subjects into the clinical trial protocol who were referred for further assessment for alcohol dependence. We did, however, enroll subjects who were on pain medication for their injuries if they had intact cognition, were able to participate in informed consent procedures, and were willing to complete the interview.

Blood Alcohol Analysis
Emergency department nursing staff obtained blood samples on hospital admission and sent them to the laboratory for analysis by gas chromatography at one site and by the automated alcohol dehydrogenase enzymatic method at a second site. To test intra-site validity of alcohol measurements, we sent 7 specimens with known values across a range of levels (0–500 mg/dL) to each hospital laboratory. Correlations between the known and measured BAC values were r greater than 0.98 at both institutions with each round of testing.

Study Protocol
The study was approved by the institutional review boards of both institutions. Nurse clinicians were trained by a social psychologist experienced in collecting data on self-reported alcohol consumption. The training sessions focused on the feedback, responsibility, advice, menu, empathy, self-efficacy, and technique of interviewing.27 To ensure consistency across interviewers and time, we videotaped nurse clinicians every 6 months, and experts critiqued the interview technique to maintain rigor in data collection procedures. After the 6-month review, each nurse clinician had a retraining session with the social psychologist to prevent drift of the intervention. No nurse clinician had to resign from the study team because of lack of consistency in the intervention technique based on the recommendation of the expert review panel.

Nurse clinicians identified eligible subjects by daily review of the admitting logs, requested consent for enrollment from eligible patients, and interviewed subjects at the subjects’ bedsides during hospital admission. To calculate when potential subjects were alcohol-free, clinicians divided the admission BAC by an alcohol metabolism rate of 20 mg/dL per hour. This calculation was used to predict the approximate number of hours needed for the subject to be alcohol-free28,29 and enabled the interviews to be done only with subjects who were not under the influence of alcohol at the time of the interview. Nurse clinicians obtained each subject’s weight and explained the concept of a standard drink across different preferred drinks. One standard drink was defined as 360 mL (12 oz) of beer, 150 mL (5 oz) of wine, or 45 mL (1.5 oz) of distilled spirits.30 If subjects reported the use of alcoholic preparations that varied from the alcohol content of the standard drink, such as wine coolers or foreign beers, nurse clinicians converted the alcohol content of those preparations into the equivalent of standard drinks. During a health-assessment interview, the nurse clinicians asked subjects to report the time drinking began in the 24 hours before the injury and the number of standard drinks consumed. Subjects were also asked to report the use of other mind-altering drugs during the same 24 hours.

Blood Alcohol Estimator
The Blood Alcohol Content Estimator is a computer software package developed by the National Highway Traffic Safety Administration.31 The program enables users to calculate an estimate of BAC based on a person’s weight, sex, number of drinks consumed, and period of consumption. The most controversial element of the software program is the elimination rate.13,28,29 Alcohol is eliminated from the body by metabolism and through excretion in the breath and through body fluids such as urine and saliva. Higher elimination rates appear to occur at BACs greater than 300 mg/dL and in alcohol-dependent drinkers. The computer program has 3 elimination rates: "above average" (decrease in BAC equals 20 mg/dL per hour), "average" (decrease in BAC equals 17 mg/dL per hour), and "below average" (decrease in BAC equals 12 mg/dL per hour).

For this study, we used the average elimination rate because our subjects were not alcohol-dependent drinkers, and we had no reason to suspect that elimination would occur at the below-average rate. We did not use the above-average rate because only a few subjects (5 of 181; 3% of the study population) had a BAC of 300 mg/dL or greater. The choice of the average elimination rate, however, was a study limitation because it did not allow for variations in liver function across all subjects. It also did not allow for a higher rate of alcohol metabolism in the 5 men with very high BAC (>300 mg/dL). The Blood Alcohol Content Estimator program does, however, control for body water and metabolic differences between the sexes.13,31

Data Management and Statistical Analysis
We calculated subjects’ EBAC by using data on the self-reported quantity and frequency of drinking in the 24 hours before the vehicular crash. Study staff then entered the number of standard drinks and length of the drinking episode into the Blood Alcohol Content Estimator program to determine EBAC. Nurse clinicians extracted and recorded laboratory BAC values at the time of hospital admission from each subject’s medical record. After data collection, we used nonparametric correlations to determine if the correlations between BAC values and EBAC values were significantly different from zero (P≤.05). We did these tests separately for men and for women and defined overreporting as EBAC greater than BAC and underreporting as BAC greater than EBAC.

For all subjects who acknowledged drinking, we used regression analyses to determine how well body weight and the number of drinks consumed were predictive of laboratory BAC. Because drinking and driving have legal ramif ications, we recognized the possibility that drivers altered their self-reports more than nondrivers did. To rule out the possibility that men might more often have been a driver, and that this factor might partially account for any sex-related differences in accuracy, we used analysis of variance to examine accuracy as a function of sex, driver status, and the interaction of sex and driver status.

To explore the role of injury severity in the accuracy of self-reporting, we obtained the Injury Severity Score from the staff of the trauma centers involved with data collection. The Injury Severity Score was devised by Baker et al32 and is used to describe the severity of injury and probability of survival in patients with multiple injuries.32,33


   Results
Top
 Abstract
 Background
 Research Questions
 Methods
 Results
Discussion
 Recommendations for Clinical...
 Summary
 References
 
Subjects
During the 40 months of data collection, all potential subjects were screened at both trauma centers. We entered 43 women and 138 men into the study. Of the 181 subjects, 130 were drivers of vehicles; 45 were passengers; and 6 were injured in vehicles, but their position in the vehicle was unknown. Of the 181 subjects, 7 men (4%) with BAC of 10 mg/dL or greater on admission denied drinking in the 24 hours before the injury. The remainder of the analyses was completed on the 174 subjects who acknowledged drinking and had a BAC of 10 mg/dL or greater. Means by sex for age, Injury Severity Score, weight, number of drinks before the injury, and BACs are reported in Table 1Go for these 174 subjects.


View this table:
[in this window]
[in a new window]
  		Table 1 Demographic data for subjects with self-reports of drinking verified by blood alcohol concentration*

 
Although all subjects provided self-reported data on quantity and frequency of drinking, some subjects could not recall the time they began drinking, and for others, the nurse clinician did not elicit information about the time drinking began. When data on the 7 men who denied drinking at the time of admission were excluded, we had enough information (sex, weight, number of drinks, and hours from start of drinking) on the remaining 86 men and 27 women to calculate EBAC by using the National Highway Traffic Safety Administration program.31 We used a 1-way analysis of variance to compare subjects for whom we had EBACs with subjects for whom we did not. The results indicated no significant differences in injury severity, age, weight, or admitting BAC (Table 2Go).


View this table:
[in this window]
[in a new window]
  		Table 2 Group comparison of subjects with and without data on estimated blood alcohol concentration*
 
Statistical Findings
The Pearson product moment correlations between laboratory BAC and EBAC derived from self-reports were r = 0.264 (P = .005) for all subjects (n = 113), r = 0.17 (P = .40) for women (n = 27), and (r = 0.30 (P = .005) for men (n = 86). Because the scat-terplots were not bivariate normal, the Spearman rank correlation was used to test the correlation between EBAC and BAC. Spearman r values between BAC and EBAC were 0.254 (P = .007) for all subjects (n = 113), 0.020 (P = .93) for women, and 0.324 (P = .002) for men. Among the nonintoxicated subjects, 0% overreported their drinking, and 39% underreported their drinking. Among the intoxicated subjects, 72% underreported their drinking, and 14% overreported it.

To investigate whether underreporting differed by sex and by driver-nondriver status, we did a 2-way analysis of variance. The means and SDs are shown in Table 3Go. The results indicated a significant effect due to driver-nondriver status (F1,103 = 12.01, P = .001); drivers (n = 78) underreported more often than did nondrivers (n = 29). For this analysis, we did not include those subjects whose position in the car was unknown (n = 6). Men underreported significantly more than did women (F1,103 = 7.95, P = .006). Among women, drivers underreported their drinking, and non-drivers overreported their drinking. We found no interaction between sex and driver-nondriver status (F1,103 = 2.01, P = .16).


View this table:
[in this window]
[in a new window]
  		Table 3 Differences in estimated and actual blood alcohol concentration by sex in drivers and nondrivers injured unintentionally*

 
For 174 subjects, excluding those 7 subjects who denied drinking but had BACs of 10 mg/dL or greater, data were available on the number of drinks, weight, and sex. These variables represented all but one of the variables used for the National Highway Traffic Safety Administration estimates. For men and women separately, multiple regression analyses were used to predict BAC on the basis of weight and the reported number of drinks. The multiple regression scatterplots were not bivariate normal because the distribution of number of drinks was positively skewed; as a result, the joint distribution was curvilinear. The data on number of drinks were transformed by using a log + 1 transformation. The log transformation allowed for the fact that the subjects with the higher number of drinks were almost certainly drinking for a longer period than lighter drinkers were; heavier drinkers, therefore, had more time than the lighter drinkers for alcohol metabolism to occur. With this transformed predictor, the scatterplots were bivariate normal. For women (n = 43), weight and number of drinks accounted for 3% of the variance in laboratory BAC (R = 0.174, F2,40 = 0.623, P = .54). In contrast, for men (n = 131), these same predictors accounted for 29% of the variance (R = 0.543, F2,128 = 26.71, P<.001). For men, the standardized regression coefficient for the number of drinks (b = 0.545) was significantly different from zero (P<.001).


   Discussion
Top
 Abstract
 Background
 Research Questions
 Methods
 Results
 Discussion
Recommendations for Clinical...
 Summary
 References
 
Although the results of the analysis did not show as strong a relationship as we expected between measured BAC and EBAC, they did indicate that most subjects acknowledged drinking before the vehicular crash when such drinking occurred. Only 7 (5 drivers and 2 passengers) of 181 subjects with a BAC of 10 mg/dL or greater at the time of admission denied drinking before the injury. This finding is additionally remarkable because many subjects (115/181 or 64%) were legally intoxicated (BAC ≥ 100 mg/dL) drivers and therefore at risk for prosecution for driving under the influence of alcohol. The role of intoxication cannot be minimized. Although the interviews took place during hospitalization when the subjects were alcohol-free, subjects intoxicated at the time of admission were more likely than their nonintoxicated counterparts to underreport drinking. Subjects in the intoxicated group may have underreported their drinking because (1) they could not remember their drinking patterns because of alcohol impairment or (2) they were aware that their drinking exceeded legal limits and placed them at risk for legal consequences.

Sex-Related Differences in Accuracy
We found that self-reports of men were more highly correlated with laboratory BACs than were the self-reports of women. The results of the simple correlation we used may raise the question that the sex-related differences might be merely artifacts of the large difference in the numbers of women and men (the f inding for women might be a type II error). However, the multiple regression analyses provide stronger evidence that after alcohol-related vehicular crashes, self-reports from men more closely reflected drinking patterns than did the self-reports of women.

Although women and men may self-report with different degrees of accuracy, the small number of women in our sample precludes firm conclusions. Clearly, men systematically underreported their drinking before injury, whereas women provided less predictable self-reports. This pattern is particularly noteworthy when the SDs of the EBAC in both sexes are examined (Table 3Go); females had more variance in reporting than men did. The reason for systematic underreporting by men is unclear. More men than women who were eligible for entry into the study declined to be enrolled because of a concern about prosecution for driving under the influence. These same concerns may have led to underreporting despite assurances of confidentiality. The more random self-reports of women are equally perplexing. If indeed sex-related differences do exist, they may stem from complex physiological, psychological, and social differences across the sexes.

Inaccurate Self-reports and Fear of Prosecution
Although persons seriously injured in alcohol-related motor vehicle crashes may underreport their drinking because of the fear of legal consequences, our subjects were willing to acknowledge a mean of 7.48 standard drinks before the injury (Table 1Go). This finding is particularly remarkable because of the high percentage of the 181 subjects (83%; 115 drivers and 35 passengers) who were legally intoxicated at the time of the crash. We found that most subjects were not trying to conceal their drinking entirely but rather were either unable or unwilling to describe the exact quantity and frequency of alcohol consumption. Possibly, the subjects did not understand the concept of a standard drink. If they underestimated the number of standard drinks contained in their servings of alcohol, they would have underreported their drinking. In addition, some subjects may have underreported their drinking because they could not remember their drinking patterns because of confusion from the injury or because of alcohol intoxication itself.

Because of these factors, we do not think that the data support the hypothesis that underreporting was a function of fear of prosecution alone. Perrine et al1 identified 5 possible errors in processing that may contribute to underreporting in drinking drivers: inaccurate perception at the time of consumption, inaccurate registration of the perceptual information, differential decay of the stored information, inaccurate retrieval of the information, and conscious distortion of the verbal report. If our subjects were intent on distorting their report of drinking, it seems unlikely that they would have reported a mean of more than 7 drinks before the motor vehicle crash.

Other Explanations for Underreporting
One factor involved in the calculation of the EBAC was the time drinking commenced. If men and women consistently overreported the hours between the first drink and the injury, their EBACs would reflect levels lower than the BACs. Indeed, those subjects who drank at higher levels may have had more difficulty remembering when drinking commenced. This factor may explain why the mean BAC in those subjects for whom we did not have necessary data to calculate EBAC was higher than the mean BAC of those subjects who remembered the time drinking commenced (177.87 mg/dL vs 158.67 mg/dL; Table 2Go). These findings are interesting in light of the findings of Perrine et al1 that heavier drinkers retrospectively underreported alcohol consumption significantly more than did lighter drinkers. Because we did not do a laboratory screen for other drugs of abuse, self-reported drinking may have been inaccurate if the subject was under the influence of other drugs without our knowledge. Finally, inaccurate or imprecise laboratory techniques, despite our periodic reliability testing procedures, may have led to errors in BAC measurements.


   Recommendations for Clinical Practice: Understanding the Context for Collecting Self-reports
Top
 Abstract
 Background
 Research Questions
 Methods
 Results
 Discussion
 Recommendations for Clinical...
Summary
 References
 
The setting and context for collecting for self-reported data may have an effect on the accuracy of the data. By using specific strategies during an interview, healthcare practitioners can encourage patients to provide accurate self-reports. These data will then provide a useful assessment on which to base an intervention to change harmful drinking practices. O’Farrell and Maisto34 recommended that 9 conditions be met:

  1. The patient is alcohol-free at the time of assessment.
  2. The patient has no major signs or symptoms of alcohol withdrawal.
  3. A structured instrument is used to measure quantity and frequency of drinking.
  4. The patient is aware that the self-report will be checked with other sources such as collateral interviews or laboratory markers.
  5. Good rapport is established with the patient.
  6. The patient complies with other aspects of the treatment regimen.
  7. The patient has no motive to distort the self-report.
  8. The patient is assured of confidentiality.
  9. The setting and procedures promote honest reporting.

The results of our study indicate that women and men have different patterns of self-reporting drinking after vehicular injury. Men tended to underreport their drinking, whereas women were more random in their self-reports; both underreported and overreported their drinking. Of note, most persons who drink before vehicular injury, even when they drink to the level of legal intoxication, do not deny the drinking. When screening for problem drinking following a vehicular crash, clinicians can use the recommendations of O’Farrell and Maisto to enhance the efforts to obtain accurate self-reports of drinking patterns before injury. Self-report in combination with other assessment data can then be used without skepticism to tailor an intervention to reduce drinking.


   Summary
Top
 Abstract
 Background
 Research Questions
 Methods
 Results
 Discussion
 Recommendations for Clinical...
 Summary
References
 
Although obstacles to the standardization of alcohol screening and intervention in hospitals after vehicular injury remain, clinicians can have reasonable confidence that most patients will acknowledge drinking after alcohol-related motor vehicle crashes. According to our findings, self-reported data from men may reflect the overall trends of alcohol consumption but tend to indicate lower amounts than were actually ingested. Reports from women are less predictable. When providing care to injured patients, nurses and physicians do not use self-reported alcohol consumption to predict a specific BAC but rather to tailor interventions to reduce risky situations, drinking, and future risk for injury. Therefore, the self-reported data were, for the most part, a critical addition to the results of the history and physical examination.

In our investigation, self-reports of alcohol use immediately before injury provided assessment data that generally reflected the patterns of drinking in most subjects. In particular, when men acknowledged drinking, they provided useful data upon which to tailor an intervention. Further research on self-reported alcohol consumption in women that includes a larger sample size is needed to provide information about the accuracy of women’s self-reports after vehicular crashes.


   ACKNOWLEDGMENTS
 
We acknowledge the assistance of the trauma services at University of Cincinnati Medical Center, Cincinnati, Ohio, Kenneth Davis, MD, director, and at Miami Valley Hospital, Dayton, Ohio, Mary McCarthy, MD, director. This study was supported by grant CCR-510153 from the Centers for Disease Control and Prevention, Grants for Unintentional Injury Prevention and Control Research.

To purchase reprints, contact the InnoVision Group, 101 Columbia, Aliso Viejo, CA 92656. Phone, (800) 809-2273 or (949) 362-2050 (ext 532); fax, (949) 362-2049; e-mail, reprints{at}aacn.org.


   REFERENCES
Top
 Abstract
 Background
 Research Questions
 Methods
 Results
 Discussion
 Recommendations for Clinical...
 Summary
 References
 

  1. Perrine MW, Mundt JC, Searles JS, Walter D. I only had a couple of beers: validation of drivers’ self-reported drinking in bars. In: Mercier-Guyon C, ed. Proceedings of the 14th International Conference on Alcohol, Drugs, and Traffic Safety. Annecy, France: Centre d’Etudes et de Recherches en Médecine du Trafic; 1997:769–776.
  2. Babor TF, Stephens RS, Marlatt GA. Verbal report methods in clinical research on alcoholism: response bias and its minimization. J Stud Alcohol. 1987;48:410–424.[Medline]
  3. Danielsson PE, Rivara FP, Gentilello L, Maier RV. Reasons why trauma surgeons fail to screen for alcohol problems. Arch Surg. 1999;135:564–568.
  4. Graham DM, Maio RF, Blow FC. Emergency physician attitudes concerning intervention for alcohol abuse/dependence delivered in the emergency department: a brief report. J Addict Dis. 2000;19:45–53.[Medline]
  5. Rivara FP, Tollefson S, Tesh E, Gentilello LM. Screening trauma patients for alcohol problems: are insurance companies barriers? J Trauma. 2000;48:115–118.[Medline]
  6. Gentilello LM, Rivara FP, Donovan DM, et al. Alcohol interventions in a trauma center as a means of reducing the risk of injury recurrence. Ann Surg. 1999;230:473–483.[Medline]
  7. Dyehouse JM, Sommers MS. Brief intervention as an advanced practice strategy for seriously injured victims of multiple trauma. AACN Clin Issues. 1995;6:53–62.[Medline]
  8. Gentilello LM, Donovan DM, Dunn CW, Rivara FP. Alcohol interventions in trauma centers. JAMA. 1995;274:1043–1048.[Abstract/Free Full Text]
  9. Sommers MS, Dyehouse JM, Howe SR, Lemmink J. Preventing alcohol-related motor vehicle injury with brief intervention strategies: preliminary findings with 60 subjects. Assoc Adv Automotive Med. 1998;42:426–427.
  10. Monti PM, Colby SM, Barnett NP, et al. Brief intervention for harm reduction with alcohol-positive older adolescents in a hospital emergency department. J Consult Clin Psychol. 1999;67:989–994.[Medline]
  11. Fleming MF, Barry KL, Manwell LB, Johnson K, London R. Brief physician advice for problem alcohol drinkers: a randomized controlled trial in community-based primary care practices. JAMA. 1997;277:1039–1045.[Abstract/Free Full Text]
  12. Cherpitel CJ. Alcohol and injuries resulting from violence: a review of emergency room studies. Addiction. 1994;89:157–165.[Medline]
  13. Midanik L. The validity of self-reported alcohol consumption and alcohol problems: a literature review. Br J Addict. 1982;77:357–382.[Medline]
  14. Sobell LC, Sobell MB. Self-report issues in alcohol abuse: state of the art and future directions. Behav Assess. 1990;12:77–90.
  15. Sommers MS, Dyehouse JM, Howe SR, Lemmink J, Volz T, Manharth M. Validity of self-reported alcohol consumption in nondependent drinkers with unintentional injuries. Alcohol Clin Exp Res. 2000;24:1406–1413.[Medline]
  16. National drunk and drugged driving month-December 2001. MMWR Morb Mortal Wkly Rep. 2001;50:1063–1064.
  17. Injury Facts 1999. National Safety Council, Itasca, Ill: 1999.
  18. National Center for Injury Prevention and Control. December is drugged and drunk driving prevention month. December 11, 1998. Available at: http://www.cdc.gov/ncipc/duip/spotlite/3dfacts.htm. Accessed June 30, 2000.
  19. Rivara FP, Koepsell TD, Jurkovich GJ, Gurney JG, Soderberg R. The effects of alcohol abuse on readmission for trauma. JAMA. 1993;270:1962–1964.[Abstract/Free Full Text]
  20. Sobell LC, Toneatto T, Sobell MB, Leo GI, Johnson L. Alcohol abusers’ perceptions of the accuracy of their self-reports of drinking: implications for treatment. Addict Behav. 1992;17:507–511.[Medline]
  21. Embree BG, Whitehead PC. Validity and reliability of self-reported drinking behavior: dealing with the problem of response bias. J Stud Alcohol. 1993;54:334–344.[Medline]
  22. Brown J, Kranzler HR, Del Boca FK. Self-reports and drug abuse inpatients: factors affecting reliability and validity. Br J Addict. 1992;87:1013–1024.[Medline]
  23. Smith GT, McCarthy DM, Goldman MS. Self-reported drinking and alcohol-related problems among early adolescents: dimensionality and validity over 24 months. J Stud Alcohol. 1995;56:383–394.[Medline]
  24. Cherpitel CJ. Breath analysis and self-reports as measures of alcohol-related emergency room admissions. J Stud Alcohol. 1989;50:155–161.[Medline]
  25. Bohn MJ, Babor TF, Kranzler HR. The Alcohol Use Disorders Identification Test (AUDIT): validation of a screening instrument for use in medical settings. J Stud Alcohol. 1995;56:423–432.[Medline]
  26. Allen JP, Litten RZ, Fertig JB, Babor T. A review of research on the Alcohol Use Disorders Identification Test (AUDIT). Alcohol Clin Exp Res. 1999;21:613–619.
  27. Bien TH, Miller WR, Tonigan JS. Brief interventions for alcohol problems: a review. Addiction. 1993;88:315–335.[Medline]
  28. Midanik L. Over-reports of recent alcohol consumption in a clinical population: a validity study. Drug Alcohol Depend. 1982;9:101–110.[Medline]
  29. Midanik L. Validity of self-reported alcohol use: a literature review and assessment. Br J Addict. 1988;83:1019–1029.[Medline]
  30. The Physicians’ Guide to Helping Patients With Alcohol Problems. Bethesda, Md: National Institute on Alcohol Abuse and Alcoholism; 1995. NIH Publication 95–3769.
  31. National Highway Traffic Safety Administration. Blood Alcohol Content Estimator. Springfield, Va: US Dept of Commerce; 1994.
  32. Baker SP, O’Neill B, Haddon W, Long WB. The Injury Severity Score: a method for describing patients with multiple injuries and evaluating emergency care. J Trauma. 1974;14:187–196.[Medline]
  33. The Abbreviated Injury Scale. Des Plaines, Ill: Association for the Advancement of the Automotive Medicine; 1990.
  34. O’Farrell TJ, Maisto SA. The utility of self-report and biological measures of alcohol consumption in alcoholism treatment outcome studies. Adv Behav Res Ther. 1987;9:91–125.



This article has been cited by other articles:


Home page
 CirculationHome page
E. Ruttmann, L. J. Brant, H. Concin, G. Diem, K. Rapp, H. Ulmer, and the Vorarlberg Health Monitoring and Promotion Pro
{gamma}-Glutamyltransferase as a Risk Factor for Cardiovascular Disease Mortality: An Epidemiological Investigation in a Cohort of 163 944 Austrian Adults
Circulation, October 4, 2005; 112(14): 2130 - 2137.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Respond to This Article
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sommers, M. S.
Right arrow Articles by Fleming, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sommers, M. S.
Right arrow Articles by Fleming, M.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS