Gender differences in the performance of a computerized version of the alcohol use disorders identification test in subcritically injured patients who are admitted to the emergency department.
ABSTRACT The Alcohol Use Disorder Identification Test (AUDIT) has been recommended as a screening tool to detect patients who are appropriate candidates for brief, preventive alcohol interventions. Lower AUDIT cutoff scores have been proposed for women; however, the appropriate value remains unknown. The primary purpose of this study was to determine the optimal AUDIT cutpoint for detecting alcohol problems in subcritically injured male and female patients who are treated in the emergency department (ED). An additional purpose of the study was to determine whether computerized screening for alcohol problems is feasible in this setting.
The study was performed in the ED of a large, urban university teaching hospital. During an 8-month period, 1205 male and 722 female injured patients were screened using an interactive computerized lifestyle assessment that included the AUDIT as an embedded component. World Health Organization criteria were used to define alcohol dependence and harmful drinking. World Health Organization criteria for excessive consumption were used to define high-risk drinking. The ability of the AUDIT to classify appropriately male and female patients as having one of these three conditions was the primary outcome measure.
Criteria for any alcohol use disorder were present in 17.5% of men and 6.8% of women. The overall accuracy of the AUDIT was good to excellent. At a specificity >0.80, sensitivity was 0.75 for men using a cutoff of 8 points and 0.84 for women using a cutoff of 5 points. Eighty-five percent of patients completed computerized screening without the need for additional help.
Different AUDIT scoring thresholds for men and women are required to achieve comparable sensitivity and specificity when using the AUDIT to screen injured patients in the ED. Computerized AUDIT administration is feasible and may help to overcome time limitations that may compromise screening in this busy clinical environment.
- SourceAvailable from: Dennis Donovan[Show abstract] [Hide abstract]
ABSTRACT: In 2005 the American College of Surgeons passed a mandate requiring that Level I trauma centers have mechanisms to identify and intervene with problem drinkers. The aim of this investigation was to determine if a multilevel trauma center intervention targeting both providers and patients would lead to higher quality alcohol screening and brief intervention (SBI) when compared with trauma center mandate compliance without implementation enhancements. Cluster randomized trial in which intervention site (site n =10, patient n =409) providers received 1-day workshop training on evidence-based motivational interviewing (MI) alcohol interventions and four 30-minute feedback and coaching sessions; control sites (site n =10, patient n =469) implemented the mandate without study team training enhancements. Trauma centers in the United States of America. 878 blood alcohol positive inpatients with and without traumatic brain injury (TBI). MI skills of providers were assessed with fidelity coded standardized patient interviews. All patients were interviewed at baseline, and 6- and 12-months post-injury with the Alcohol Use Disorders Identification Test (AUDIT). Intervention site providers consistently demonstrated enhanced MI skills compared with control providers. Intervention patients demonstrated an 8% reduction in AUDIT hazardous drinking relative to controls over the course of the year after injury (RR =0.88, 95%, CI =0.79, 0.98). Intervention patients were more likely to demonstrate improvements in alcohol use problems in the absence of TBI (p =0.002). Trauma center providers can be trained to deliver higher quality alcohol screening and brief intervention than untrained providers, which is associated with modest reductions in alcohol use problems, particularly among patients without traumatic brain injury. Key Words: Alcohol, Screening and Brief Intervention, Traumatic Injury, American College of Surgeons, Policy Mandate, Motivational Interviewing, Dissemination and Implementation Research.Addiction 01/2014; · 4.58 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Introduction Alcohol use is an important contributor to injuries. Simple bedside tools to identify trauma patients with potentially harmful drinking may assist in brief intervention efforts in clinical practice. The objective of the study was to determine and compare the accuracy of Alcohol Use Disorders Identification Test (AUDIT) and an abbreviated version of this test, in the detection of hazardous drinking. Methods A cross sectional study of a convenience sample of admitted trauma patients at a single Australian major trauma centre. Eligible patients completed the AUDIT. AUDIT survey responses were scored in two ways, using the full form scale and secondly an abbreviated (AUDIT C) scale which uses only the first 3 questions. AUDIT and AUDIT-C scores were then evaluated with respect to the primary study measure; the detection of hazardous alcohol consumption based on a full alcohol consumption history. Sensitivities for each relevant score were calculated and Receiver Operator Characteristic Curve (ROC) analysis was used to determine test accuracy. Results During the study period, 523 trauma admissions were identified and of these 146 (28%) were screened. The optimum cut off scores for AUDIT and AUDIT-C were 8 and 5 respectively corresponding to sensitivities of 88% and 91% and both tests had excellent overall accuracy for the detection of hazardous alcohol consumption. There was no significant difference between AUDIT-C and AUDIT performance (p = 0.395) (AUDIT-C AUROC 0.96 95%CI 0.93, 0.99). Conclusion AUDIT-C appears to be a potentially useful screening tool for use trauma centres, but that further research with larger samples is required.Injury 01/2014; · 2.46 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: The time required to conduct drug and alcohol screening has been a major barrier to its implementation in mainstream healthcare settings. Because patient self-administered tools are potentially more efficient, we translated the Alcohol, Smoking and Substance Involvement Screening Test (ASSIST) into an audio guided computer assisted self interview (ACASI) format. This study reports on the test-retest reliability of the ACASI ASSIST in an adult primary care population. Adult primary care patients completed the ACASI ASSIST, in English or Spanish, twice within a 1–4 week period. Among the 101 participants, there were no significant differences between test administrations in detecting moderate to high risk use for tobacco, alcohol, or any other drug class. Substance risk scores from the two administrations had excellent concordance (90-98%) and high correlation (ICC 0.90-0.97) for tobacco, alcohol, and drugs. The ACASI ASSIST has good test-retest reliability, and warrants additional study to evaluate its validity for detecting unhealthy substance use.Journal of substance abuse treatment 01/2014; · 2.90 Impact Factor
Gender Differences in the Performance of a
Computerized Version of the Alcohol Use Disorders
Identification Test in Subcritically Injured Patients
Who Are Admitted to the Emergency Department
Tim Neumann, Bruno Neuner, Larry M. Gentilello, Edith Weiss-Gerlach, Henriette Mentz, Jordan S. Rettig, Torsten Schröder,
Helmar Wauer, Christian Müller, Michael Schütz, Karl Mann, Gerda Siebert, Michael Dettling, Joachim M. Müller,
Wolfgang J. Kox, and Claudia D. Spies
Objective: The Alcohol Use Disorder Identification Test (AUDIT) has been recommended as a screening tool
scores have been proposed for women; however, the appropriate value remains unknown. The primary purpose of
this study was to determine the optimal AUDIT cutpoint for detecting alcohol problems in subcritically injured
male and female patients who are treated in the emergency department (ED). An additional purpose of the study
was to determine whether computerized screening for alcohol problems is feasible in this setting.
Methods: The study was performed in the ED of a large, urban university teaching hospital. During an 8-month
dependence and harmful drinking. World Health Organization criteria for excessive consumption were used to define
high-risk drinking. The ability of the AUDIT to classify appropriately male and female patients as having one of these
three conditions was the primary outcome measure.
Results: Criteria for any alcohol use disorder were present in 17.5% of men and 6.8% of women. The overall
accuracy of the AUDIT was good to excellent. At a specificity ?0.80, sensitivity was 0.75 for men using a cutoff of
8 points and 0.84 for women using a cutoff of 5 points. Eighty-five percent of patients completed computerized
screening without the need for additional help.
Conclusions: Different AUDIT scoring thresholds for men and women are required to achieve comparable
sensitivity and specificity when using the AUDIT to screen injured patients in the ED. Computerized AUDIT
administration is feasible and may help to overcome time limitations that may compromise screening in this busy
Key words: Trauma, Alcohol Use Disorder Identification Test, Gender, Computer, Alcohol.
man, 1998; D’Onofrio et al., 1998; Gentilello et al., 1995,
HERE IS A strong association between alcohol use
disorders (AUDs) and injuries (Brismar and Berg-
1999, Spies et al., 1996). Brief, preventive interventions
delivered in the acute care setting may reduce posttrau-
matic morbidity and the risk of recurrent trauma (Dinh-
Zarr et al., 2000; D’Onofrio et al., 1998; Gentilello et al.,
1999). Gentilello et al. (1999) demonstrated that moti-
vational interviewing in severely injured patients reduced
the risk of repeat injury requiring emergency department
(ED) treatment from 10 to 5% during 1 year of follow-
up. Given the high prevalence of alcohol problems in
EDs and accumulating evidence of brief intervention ef-
ficacy, a variety of professional, expert, and consensus panels
currently recommend that EDs and trauma centers expand
their focus to include not just treatment of the patient’s pre-
senting injury but also the provision of screening and brief
interventions as a routine component of care. For optimizing
detection rates, screening questionnaires have been recom-
mended as supplements to the history, physical examination,
and blood alcohol test result (D’Onofrio et al., 1998;
D’Onofrio and Degutis, 2002; Runge et al., 2001; Spies et al.,
From the Department of Anesthesiology and Intensive Care Medicine, Campus
Mitte (TN, BN, EW-G, HM, JSR, TS, HW, WJK, CDS), Department of Clinical
Chemistry and Biochemistry (CM), Department of Trauma and Reconstructive
Surgery, Campus Virchow-Klinikum (MS), Department of Abdominal Surgery,
and Department of Medical Statistics (GS), Charite ´ University Medicine Berlin,
Germany (KM); and Division of Burns Trauma, Critical Care, Parkland Medical
Center, University of Texas Southwestern Medical School, Dallas, Texas, USA
Received for publication September 8, 2003; accepted September 3, 2004.
Supported by the German Health Ministry (BMG 217-43794-5/5).
Reprint requests: Tim Neumann, MD, Department of Anesthesiology and
Intensive Care Medicine, Charite ´ University Medicine Berlin, Campus Mitte,
Schumannstrasse 20/21, 10117 Berlin, Germany; Fax: ?49-30-450-531-900;
Copyright © 2004 by the Research Society on Alcoholism.
ALCOHOLISM: CLINICAL AND EXPERIMENTAL RESEARCH
Vol. 28, No. 11
Alcohol Clin Exp Res, Vol 28, No 11, 2004: pp 1693–1701
The 10-item Alcohol Use Disorders Identification Test
(AUDIT) (Babor et al., 1992; Saunders et al., 1993b) was
developed by the World Health Organization to detect
alcohol problems and was designed to be useful across a
broad spectrum of problem severity, ranging from depen-
dence to harmful or hazardous use (Allen et al., 1997;
Bradley et al., 1998b; Reinert and Allen, 2002). However,
use of the original cutoff of 8 points (Babor et al., 1992;
Saunders et al., 1993b) has been questioned, particularly
when screening women, as sensitivity is lower when the
same cutoffs are used for both genders (Aertgeerts et al.,
2001; Allen et al., 1997; Bergman and Källmen, 2002; Brad-
ley et al., 1998b; Cherpitel, 1995a,b, 1999; Cherpitel and
Borges, 2000; Conigrave et al., 1995; Hermansson et al.,
2000; Reinert and Allen, 2002; Rumpf et al., 2002; Stein-
bauer et al., 1998).
AUDs in women have unique social and biological char-
acteristics, suggesting the need for different scoring thresh-
olds. Women develop a higher blood alcohol concentration
than men after consuming the same amount of alcohol,
which may affect the reliability of classifying patients on the
basis of the quantity of alcohol consumed. Women are also
more vulnerable to the toxic affects of alcohol, e.g., liver
cirrhosis or cognitive disorders (Bergman et Källmen, 2002;
Bradley et al., 1998a; Brienza and Stein, 2002; British Med-
ical Association, 1995; Graham et al., 1998; National Insti-
tute on Alcohol Abuse and Alcoholism, 1995), suggesting a
need for a lower safety threshold. Women may also differ in
their social and demographic characteristics and in their
approach to the health care system, which may also influ-
ence the performance of a screening test (Aertgeerts et al.,
2001; Bergman et Källmen, 2002; Bradley et al., 1998a;
Brienza and Stein, 2002; Cherpitel, 1995a,b, 1999; Cherpi-
tel and Borges, 2000; Gentilello et al., 2000; Rehm et al.,
2001; Rumpf et al., 2002; Steinbauer et al., 1998). The need
for a different AUDIT cutpoint may be particularly rele-
vant when the AUDIT is used in injured patients, as women
have a higher risk of injury than do men who have con-
sumed the same quantity of alcohol (Stockwell et al., 2002).
It is important to determine whether lower cutpoints for
women are appropriate so that individuals who may be
candidates for intervention or referral can be identified
(Cherpitel 1999; D’Onofrio et al., 1998; Gentilello et al.,
1999). However, no study on injured patients in the ED has
provided recommendations for gender-based AUDIT ad-
justment (Cherpitel, 1995a,b, 1999; Cherpitel and Borges,
2000; Soderstrom et al., 1998). One approach would be to
define an acceptable rate of false positives, or appropriate
specificity, and define AUDIT cutpoints that result in com-
parable sensitivities for men and women (Mundle et al.,
The primary aim of this study was to determine whether
gender-specific cutpoints are needed when using the
AUDIT in subcritically injured patients who present to
the ED. World Health Organization (WHO) criteria
were used to define alcohol dependence and harmful
drinking. In addition, the WHO uses the term “high-
risk” drinking to identify patients who may not currently
be experiencing harm but who are drinking at levels
associated with an increased risk for developing chronic
health problems or other consequences (WHO Depart-
ment of Mental Health and Substance Dependence,
2000). Such patients are appropriate candidates for pri-
mary prevention efforts. Therefore, the study was struc-
tured to determine the ability of the AUDIT to identify
patients with dependence, harmful drinking, or excessive
Implementation of alcohol screening and interventions
in EDs has been limited by lack of staff and time constraints
(D’Onofrio et al., 1998; Graham et al., 2000). Self-
administered computerized screening has been proposed as a
means of overcoming these barriers because it may reduce
demands placed on already overburdened ED personnel
(Maio et al., 2003). Computerized screening has been used
successfully and reliably for the detection of alcohol-related
problems, and computerized versions of the AUDIT have
been shown to produce scores similar to those obtained by
conventional means (Chan-Pensley, 1999; Cloud and Pea-
cock, 2001; Shakeshaft et al., 1999). Therefore, an additional
purpose of this study was to determine the feasibility of com-
puterized AUDIT screening in the ED.
MATERIALS AND METHODS
The study was approved by the Institutional Review Board of Univer-
sity Hospital Charité, (Berlin, Germany) and was conducted between
December 2001 and July 2002. All trauma patients ?18 years of age were
considered potential study candidates. Patients were excluded when they
were ?18 years of age, had altered mental status that precluded consent,
were medically unstable, were in significant pain (?3 points on a 10-point
visual analog scale), had significant psychiatric comorbidity, were intoxi-
cated by alcohol or drugs, were non–German-speaking, were in police
custody, were pregnant, were not residents of Germany, or were members
of the hospital staff. Patients who were readmitted for treatment of a
second injury during the study period were also not considered eligible.
A total of 5169 patients were admitted to the ED for treatment of an
injury during the study period. A total of 902 (17.4%) patients had a brief
ED stay and were discharged before contact. Of the remaining 4267
patients, 1060 were excluded for the following reasons: A severe psychi-
atric illness was present in 260 (6.1%) patients; 175 (4.1%) were critically
ill, unstable, or in significant pain; 317 (7.4%) did not speak German; 135
(3.2%) were acutely intoxicated; 68 (1.6%) were members of the hospital
staff; 49 (1.1%) were nonresidents; 41 (1%) were readmitted patients; and
15 (0.4%) were under arrest. The remaining 3207 patients were ap-
proached for study participation. Forty percent refused consent, and 1927
(60%) were enrolled. The resulting study sample included 1205 (62.6%)
men and 722 (37.4%) women.
A random cohort of 120 patients who refused consent was analyzed. Of
these, 71% were men and 29% were women. The gender distribution of
patients who refused consent contained a tendency toward a greater
proportion of men compared with enrolled patients, where 63% were
male and 37% were female (p ? 0.057). There were no significant differ-
ences in age or Injury Severity Score (Committee on Injury Scaling, 1990)
between patients who did and did not consent.
NEUMANN ET AL.
Patients were approached after provision of initial care and after
resolution of any significant pain (visual analog score ?3). After enroll-
ment and written informed consent, a self-administered paper question-
naire was provided to obtain height, weight, body mass index, marital or
intimate partner status, educational level, and occupational status. Injury
Severity Score (range 0–75 points) (Committee on Injury Scaling, 1990)
was derived from the medical record.
Patients were asked to complete a lifestyle assessment questionnaire
using a portable laptop computer that contained questions regarding
nutrition, tobacco use, drug use, social status, exercise, risky behaviors,
and sexual problems. The 10 AUDIT questions were embedded within the
computerized questionnaire. The AUDIT has been published in German
and validated in general population samples (Rumpf et al., 2002). For
simplifying administration, a “mouse only” technique that did not require
typing was used. Each question was displayed alone on a gray screen. The
next question was not displayed until 1 sec after completion of each item,
which prevented question skipping. When the first AUDIT (frequency)
question was answered “never,” no further AUDIT questions were asked.
Most patients completed the paper-and-pencil questionnaire before the
computerized assessment. However, the specific order varied according to
ED contingencies and medical factors.
The presence of (1) current alcohol dependence, (2) current harmful
use (DIMDI-Deutsches Institut fu ¨r Medizinische Dokumentation und
Information, 2002), or (3) current high-risk drinking (men ?420 g, women
?280 g/week) (WHO Department of Mental Health and Substance De-
velopment, 2000) was used to define the presence of an AUD. Patients
who did not meet any of these criteria were considered control subjects.
Trained research fellows who were working under the supervision of a
trained psychiatrist (BN) or a psychologist (EWG) conducted face-to-face
diagnostic interviews to determine whether patients met criteria for an
AUD. Questions used in the paper-and-pencil questionnaire about weekly
alcohol intake used a modified quantity-frequency method that differen-
tiated between alcohol intake on an average weekday and on an average
weekend (Feunekes et al., 1999; Rehm, 1998; WHO Department of
Mental Health and Substance Dependence, 2000). They also distinguished
between different types of alcohol-containing beverages (beer, wine, dis-
tilled, other) and amounts consumed. For maintaining consistency with
WHO definitions for risky drinking, daily intake was recalculated and
expressed as weekly consumption.
As Feunekes et al. (1999) stated in a review of 33 methodological
papers published after 1984, methods that inquire about both the fre-
quency and the amount consumed for beer, wine, and liquor separately
minimize the risk of underreporting and yield the most reliable informa-
tion. If the context of the questioning is nonstigmatizing, is confidential, is
imbedded in other questionnaires, and asks specific questions about drink-
ing, then the results are also more likely to reflect intake accurately
(Feunekes et al., 1999; Rehm, 1998).
All statistical analyses were performed using SPSS 10.0 software. Basic
patient characteristics are reported as frequencies (nominal variables) or
as median and range of the 25th to 75th percentiles (metric variables).
Differences in frequencies across genders and criteria for dependence,
harmful, or risky drinking were compared using ?2. When the expected
frequency was smaller than 5, an exact test was used (Monte-Carlo
Method). Other ordinal or metric data were compared using the Kruskall-
Wallis test, or Mann-Whitney U Test. The Spearman rank test was used to
determine the correlation between the ordinal AUDIT and weekly alcohol
intake. A p ? 0.05 was considered statistically significant.
The sensitivity of the AUDIT was determined by calculating the rate of
individuals who were identified correctly as having an AUD. Specificity
was calculated as the rate of true negatives. By plotting the rate of true
positives versus the rate of false positives (1 ? specificity), receiver
operating characteristics were generated to allow determination of sensi-
tivity and specificity over a range of potential cutoffs. The area under the
curve (AUC) can be used to assess the validity of a test independent of any
particular cutoff. An AUC of 0.8 was defined a priori as reasonable
according to the following classification AUC: 0.90 to 1.00 ? excellent,
0.80 to 0.90 ? good, 0.70 to 0.80 ? fair, 0.60 to 0.070 ? poor, 0.50 to 0.60
? fail (Luna-Herrera et al., 2003). Overlapping 95% confidence intervals
of the AUC were considered not significantly different. Cronbach’s ? was
calculated to assess internal consistency.
When the ratio of sensitivity and specificity varies, desired specificity
may depend on the underlying prevalence of the condition, as well as the
risks and cost-effectiveness of erroneously classifying a patient as having a
positive screen. For purposes of this analysis, we considered a specificity of
?0.80 as a desirable performance level for a screening tool for alcohol
problems in the ED.
There were significant gender differences in baseline
characteristics and sociodemographics (Table 1).
Women were significantly less likely to be employed and
more likely to be retired than men. Weight and body mass
index were also significantly lower in women. Women were
less likely to report current dependence, harmful use, or high-
risk drinking and had a lower alcohol intake and lower total
AUDIT scores than men (Table 2). Women also had a dif-
ferent distribution of responses to each of the 10 AUDIT
questions, with each item scored significantly lower (p ?
0.001; Table 3).
As shown in Table 1, men and women with dependence,
harmful use, or high-risk use were less likely to be married
or living with an intimate partner than were control sub-
jects. Men with these alcohol use characteristics were more
likely to be unemployed, retired, or without a high school
degree. Women with these characteristics were younger
than women without an AUD.
Considering the whole range of potential cutoffs and
comparing their respective AUC, there were no sig-
nificant gender differences in accuracy of the AUDIT by
gender: accuracy was good for men and was good to
excellent in women. For detection of dependence or
harmful or risky drinking, at a specificity of ?0.80, op-
timal AUDIT cutoffs were 5 points for women (sensitiv-
ity 0.84) and 8 points for men (sensitivity 0.75). These
results are displayed in Fig. 1. The positive predictive
value using an AUDIT cutpoint of 5 was 33% for men
and 27% for women and at 8 points was 49% and 46%,
respectively. The corresponding negative predictive val-
ues were ?94%. The correlation between weekly intake
and AUDIT score was statistically significant (Spear-
man-Rho: men ? 0.69, women ? 0.76; p ? 0.001). The
AUDIT demonstrated acceptable reliability (Cronbach’s
? ? 0.83 for men and 0.79 for women).
Thirteen percent of men and 8% of women who scored
zero on the AUDIT acknowledged some alcohol intake,
AUDIT IN TRAUMA PATIENTS
whereas 9 (3%) of 276 patients who denied using alcohol
had an AUDIT score ?5 points. Five of the nine were
alcohol-dependent patients in recovery, and no further
information was available regarding the remaining four.
However, alcohol consumption reflected “average” in-
take, whereas the AUDIT contains questions about past-
year experience on some items, which may also have
contributed to this discrepancy.
More than 83% of both genders were able to complete
the computerized screen without assistance. Men who met
criteria for an AUD performed slightly more poorly, with
77% completing screening on their own (p ? 0.001). Pa-
tients who required assistance were also significantly older
(mean 48 vs. 31 years; p ? 0.001) and more likely to have
sustained a head (21% vs. 12%; p ? 0.001) or upper
extremity injury (36% vs. 28%; p ? 0.001). A requirement
for assistance was also associated with educational status
(high school degree 35% vs. 49%) and a lower rate of
employment (53% vs. 66%; all p ? 0.01). There was no
significant difference by gender (p ? 0.44).
This is the first prospective study performed to de-
velop gender specific AUDIT cutoff points for ED
screening of injured patients. The analysis demonstrated
that an AUDIT cutpoint of ?8 is appropriate for men,
but to achieve comparable sensitivity and specificity in
women, the definition of a positive screen should be
reduced to ?5 points. The accuracy of the AUDIT to
Table 1. Patient Characteristics and Sociodemographic Data
Patients with AUDa
Male 32 0.19 32330.57
Body weight (kg)
?0.01 7978 0.86
Body mass index (kg/m2)
Male23.7 0.04 23.723.7 0.96
Injury Severity Scoreb
Male1 0.1611 0.15
Computer use without assistance
High school degree
Marital or intimate partner
Frequencies (%) and median (25th to 75th percentiles).
aPresence of alcohol dependence, harmful use, or high-risk drinking (men ? 420 g, women ? 280 g/week). All other patients were considered control subjects.
bRange of Injury Severity Score: 1–10 points.
NEUMANN ET AL.
detect current dependence, harmful use, or high-risk use
was good in men and good to excellent in women. The high
prevalence of AUDs and frequent presence of adverse
consequences as reflected by answers to AUDIT questions
support the need for screening injured ED patients. The
different sensitivities for men and women that would occur
Table 2. Alcohol-Related Data of 1205 Men and 722 Women
High risk use
High risk use
Alcohol intake (g/week)
Frequencies and median (25th to 75th percentiles).
aPresence of alcohol dependence, harmful use, or high-risk drinking (men ? 420 g, women ? 280 g/week; patients may meet more than one criterion). All other
patients were considered control subjects.
Table 3. Distribution of AUDIT Response by Gender
Item 9 and 10
Yes, not in the
? 10 drinks
Yes, in the
1. Frequency of drinking?
2. No. of drinks per occasion?
3. ?6 drinks per occasion?
4. Could not stop drinking once started?
5. Failed to do what was expected?
6. Morning drink?
7. Guilty feeling?
8. Unable to remember night before?
9. Injury under influence?
10. Concern expressed by others?
The distribution of responses was significantly different by gender for all 10 AUDIT items (?2p ? 0.01).
AUDIT IN TRAUMA PATIENTS
when using an unmodified AUDIT cutoff of 8 points for
both genders confirms the need for gender-specific cut-
The use of cutpoints that correspond to 80% specificity
seems reasonable when applied to this patient population.
The resulting sensitivities of ?75% for both genders were
associated with positive predictive values of 27 to 33% and
negative predictive values of 94%. Further increases in
sensitivity would reduce the positive predictive value, which
may raise concerns in the resource-limited environment of
Our findings are consistent with previous analyses of the
screening properties of the AUDIT in primary care and ED
settings. When equal cutpoints were used for both genders,
lower sensitivity and higher specificity were noted in
women (Aertgeerts et al., 2001; Cherpitel, 1995a, 1999;
Cherpitel and Borges, 2000; Steinbauer et al., 1998). This
study screened not only for dependence and harmful use
but also for high-risk drinking. At-risk drinking, notably
bingeing—and related injuries—is one of the major health
problems in young adults and a target for secondary pre-
vention: In settings where opportunistic interventions are
provided, it may be important to identify not only patients
who have experienced harmful consequences but also those
who are drinking at levels that are associated with high
future risk (Barnett et al., 2003; Rehm et al., 2001; WHO
Department of Mental Health and Substance Dependence,
2000). In the context of the ED, this information might be
used to identify patients who may benefit from receiving
educational materials regarding safe limits, brief advice, or
a motivational interview to reduce their level of risk.
This study was confined to injured patients. Most studies
of ED screening included both injured and noninjured
patients. These populations have different demographic
Fig. 1. Receiver operated characteristics with AUC and 95% confidence interval (CI) and sensitivity/specificity for the AUDIT to presence of (1) alcohol dependence,
(2) harmful use, or (3) high-risk drinking (men ?420 g, women ?280 g/week). Specificities ?0.80 are indicated in bold.
NEUMANN ET AL.
characteristics. Injured patients are more likely to be male,
are younger, have a higher level of education, and are more
likely to be current drinkers than noninjured patients.
Previous injury under while under the influence (AUDIT
question 9) occurred in 31.4% of men and 9.7% of women
with a positive AUDIT. More than half of these events
occurred within the past year. Concerns by others were
reported by 32.4% of men and 14.3% of women. Binge
drinking was reported in 93.8% of men and 82.5% of
women (question 3). Our findings suggest that binge drink-
ing is the most common drinking pattern found in this
patient population. Results of the U.S. National Alcohol
Survey (Rehm et al., 2001) suggest that heavy drinking
occasions are associated with an increased risk for mortal-
ity, underlining the utility of the AUDIT, which is sensitive
to heavy drinking episodes. This supports the need to use
screening criteria that accommodate maximum daily in
addition to total weekly consumption.
The correlation between consumption as reported using
the paper-and-pencil questionnaire and the computerized
AUDIT was satisfactory (Spearman-Rho: men 0.69;
women 0.76) and comparable to a previous report (Seppa
et al., 1995). Reliability (Cronbach’s ? was 0.83 for men and
0.79 for women) was in accordance with previously re-
ported ? between 0.77 and 0.94 (Allen et al., 1997; Her-
mansson et al., 2001; Reinert and Allen, 2002).
Although the incidence of dependence (men 5.7%;
women 1.0%) was approximately the same as in the adult
German population as a whole, the incidence of high-risk
drinking was greater than that reported in a recent
population-based survey (Kraus et al., 2001). In this study,
a positive screen was most frequently characterized by a
risky drinking pattern, rather than by harm or dependence.
In contrast, a previous study conducted by our group of
older, more seriously injured patients who required admis-
sion to the hospital found that 68% of patients met criteria
for dependence (Spies et al., 1996). This suggests that
opportunistic interventions with younger, less seriously in-
jured patients who present to the ED may enable detection
of problems at an earlier stage.
The proportion of patients with a positive screen and the
severity of drinking problems encountered are likely to vary
from center to center. However, the AUDIT was developed
for international use, and it has been validated for use in
different countries, cultures, and ethnic groups. Therefore,
although screening results may vary, our findings with re-
spect to gender differences are likely to be robust. The
feasibility of computerized screening is more likely to be
affected by age and educational factors than by culture or
There was a nonstatistically significant trend toward a
gender difference in consent rates between men and
women, with women being more likely to participate. Pa-
tients with more severe problems may be less likely to
consent, resulting in exclusion of more severely impaired
patients. Gender-related selection bias, with exclusion of
more severe cases, has the potential to reduce specificity, or
the rate of true positives, in men. It is less likely to have
affected sensitivity and thus gender-adjusted cutoffs in
women, although specificity may not actually be compara-
ble after adjustment.
Currently, screening is not performed in most EDs
(Danielsson et al., 1999; D’Onofrio et al., 1998; Graham et
al., 2000). The high incidence of previous injury while
under the influence confirms previous work demonstrating
that patients who are injured while under the influence
have a high rate of trauma recidivism (Gentilello et al.,
1999, 2000; Kaufmann et al., 1998; Kelly et al., 2002; Rivara
et al., 1993).
Although both injuries and alcohol problems were less
common in women, previous studies have shown that fe-
male trauma patients who do screen positive are just as
severely impaired, have at least as many adverse conse-
quences of alcohol use as men, and are more likely to have
experienced alcohol-related physical and psychological
harm than men (Gentilello et al., 2000). It therefore is
important to use appropriately sensitive screening tools for
women, as well as for men.
We found that computerized screening was feasible in
this patient population, with most patients able to complete
the process without requiring additional assistance. Pa-
tients who present to the ED are typically younger than
patients who present with other, more chronic medical
conditions, and many were already familiar with use of a
mouse and a digital screen. The use of a mouse-only tech-
nique further simplified this process. The computer auto-
matically scored the results, and programming that enables
gender-adjusted scoring can be implemented easily. Some
patients nevertheless may require assistance; therefore, it is
unlikely that computer technology will completely elimi-
nate the need for other forms of screening. Future feasi-
bility testing might include the use of portable digital assis-
tants and incorporation of other questions that enable
computerized generation of personalized feedback forms
that provide specific intervention components tailored to
the patient’s needs and specific situation.
In summary, we found that gender-specific AUDIT ad-
justments are necessary to optimize screening for AUDs in
subcritically injured patients in the ED. Alcohol problems
were common in this population, and early problems (risky
drinking) predominated, so the majority of positive screens
identify patients who are optimal candidates for brief in-
terventions. However, if traditional cutoffs are used for
women, then only those with significant abuse or depen-
dence are likely to be detected and patients with less severe
symptoms are likely to be missed.
Finally, the ED is a unique clinical environment character-
ized by frequent disruptions; overcrowding; and time, staffing,
and financial constraints. The use of a computerized version
of the AUDIT offers a potential means of offering this service
with minimal disruption of ED routines.
AUDIT IN TRAUMA PATIENTS
Aertgeerts B, Buntinx F, Ansoms S, Fevery J (2001) Screening properties
of questionnaires and laboratory tests for the detection of alcohol abuse
or dependence in a general practice population. Br J Gen Pract 51:206–
Allen JP, Litten RZ, Fertig JB, Babor T (1997) review of research on the
Alcohol Use Disorders Identification Test (AUDIT). Alcohol Clin Exp
Babor TF, de la Fuente JR, Saunders J, Grant M (1992) AUDIT, the
Alcohol Use Disorders Identification Test: Guidelines for Use in Pri-
mary Health Care. WHO, Geneva, Switzerland. Available at: http://
Barnett NP, Monti PM, Cherpitel C, Bendtsen P, Borges G, Colby SM,
Nordqvist C, Johansson K (2003) Identification and brief treatment of
alcohol problems with medical patients: An international perspective.
Alcohol Clin Exp Res 27:262–270.
Bergman H, Källmen H (2002) Alcohol use among Swedes and a psycho-
metric evaluation of the alcohol use disorders identification test. Alco-
hol Alcohol 37:245–251.
Bradley KA, Badrinath S, Bush K, Boyd-Wickizer J, Anawalt B (1998a)
Medical risks for women who drink alcohol. J Gen Intern Med 13:627–
Bradley KA, Boyd-Wickizer J, Powell SH, Burman ML (1998b) Alcohol
screening questionnaires in women: A critical review. JAMA 280:166–
Brienza RS, Stein MD (2002) Alcohol use disorders in primary care: Do
gender-specific differences exist? J Gen Intern Med 17:387–397.
Brismar B, Bergman B (1998) The significance of alcohol for violence and
accidents. Alcohol Clin Exp Res (Suppl) 22:299S–306S.
British Medical Association (1995) Guidelines on Sensible Drinking. British
Medical Association, London.
Chan-Pensley E (1999) Alcohol-Use Disorders Identification Test: A
comparison between paper and pencil and computerized versions. Al-
cohol Alcohol 34:882–885.
Cherpitel CJ (1995a) Analysis of cut points for screening instruments for
alcohol problems in the emergency room. J Stud Alcohol 56:695–700.
Cherpitel CJ (1995b) Screening for alcohol problems in the emergency
department. Ann Emerg Med 26:158–166.
Cherpitel CJ (1999) Gender, injury status and acculturation differences in
performance of screening instruments for alcohol problems among US
Hispanic emergency department patients. Drug Alcohol Depend 53:
Cherpitel CJ, Borges G (2000) Screening instruments for alcohol prob-
lems: A comparison of cut points between Mexican American and
Mexican patients in the emergency room. Subst Use Misuse 35:1419–
Cloud RN, Peacock PL (2001) Internet screening and interventions for
problem drinking: Results from the www.carebetter.com pilot study.
Alcohol Treat Q 19:23–44.
Committee on Injury Scaling (1990) The Abbreviated Injury Scale Revision.
Association for the Advancement of Automotive Medicine, Des
Conigrave KM, Hall WD, Saunders JB (1995) The AUDIT questionnaire:
Choosing a cut-off score. Alcohol Use Disorder Identification Test.
Danielsson PE, Rivara FP, Gentilello LM, Maier RV (1999) Reasons why
trauma surgeons fail to screen for alcohol problems. Arch Surg 134:
DIMDI-Deutsches Institut für Medizinische Dokumentation und Informa-
tion (2002) Internationale Statistische Klassifikation der Krankheiten und
verwandter Gesundheitsprobleme 10. Revision Version 2.0 Available at:
Dinh-Zarr T, DiGuiseppi C, Heitman E, Roberts I (2000) Interventions
for preventing injuries in problem drinkers. Cochrane Database Syst
D’Onofrio G, Bernstein E, Bernstein J, Woolard RH, Brewer PA, Craig
SA, Zink BJ (1998) Patients with alcohol problems in the emergency
department, part 1: Improving detection. SAEM Substance Abuse Task
Force. Society for Academic Emergency Medicine. Acad Emerg Med
D’Onofrio G, Degutis LC (2002) Preventive care in the emergency de-
partment: Screening and brief intervention for alcohol problems in the
emergency department: A systematic review. Acad Emerg Med 9:627–
Feunekes GI, van ’t Veer P, van Staveren WA, Kok FJ (1999) Alcohol
intake assessment: The sober facts. Am J Epidemiol 150:105–112.
Gentilello LM, Donovan DM, Dunn CW, Rivara FP (1995) Alcohol
interventions in trauma centers. Current practice and future directions.
Gentilello LM, Rivara FP, Donovan DM, Jurkovich GJ, Daranciang E,
Dunn CW, Villaveces A, Copass M, Ries RR (1999) Alcohol interven-
tions in a trauma center as a means of reducing the risk of injury
recurrence. Ann Surg 230:473–480.
Gentilello LM, Rivara FP, Donovan DM, Villaveces A, Daranciang E,
Dunn CW, Ries RR (2000) Alcohol problems in women admitted to a
level I trauma center: A gender-based comparison. J Trauma 48:108–
Graham DM, Maio RF, Blow FC, Hill EM (2000) Emergency physician
attitudes concerning intervention for alcohol abuse/dependence deliv-
ered in the emergency department: A brief report. J Addict Dis 9:45–53.
Graham K, Wilsnack R, Dawson D, Vogeltanz N (1998) Should alcohol
consumption measures be adjusted for gender differences? Addiction
Hermansson U, Helander A, Huss A, Brandt L, Ronnberg S (2000)
Alcohol Use Disorder Identification Test (AUDIT) and carbohydrate-
deficient transferrin (CDT) in a routine workplace health examination.
Alcohol Clin Exp Res 24:180–187.
Kaufmann CR, Branas CC, Brawley ML (1998) A population-based study
of trauma recidivism. J Trauma 45:325–331; discussion 331–332.
Kelly TM, Donovan JE, Kinnane JM, Taylor DM (2002) A comparison of
alcohol screening instruments among under-aged drinkers treated in
emergency departments. Alcohol Alcohol 37:444–450.
Kraus L, Augustin R (2000) Population survey on the consumption of
psychoactive substances in the German adult population 2000. Sucht
Kraus L, Augustin R (2001) Population survey on the consumption of
psychoactive sustances in the German adult population 2000. Sucht
Luna-Herrera J, Martinez-Cabrera G, Parra-Maldonado R, Enciso-
Moreno JA, Torres-Lopez J, Quesada-Pascual F, Delgadillo-Polanco R,
Franzblau SG (2003) Use of receiver operating characteristic curves to
assess the performance of a microdilution assay for determination of
drug susceptibility of clinical isolates of Mycobacterium tuberculosis.
Eur J Clin Microbiol Infect Dis 22:21–27.
Maio RF, Shope JT, Blow FC, Gregor MA, Zakrajsek JS, Weber JE,
Nypaver MM (2003) Randomized controlled trial of an emergency
department-based interactive computer program to prevent alcohol
misuse among injured adolescents. Acad Emerg Med 10:514–515.
Mundle G, Munkes J, Ackermann K, Mann K (2000) Sex differences of
carbohydrate-deficient transferrin, gamma-glutamyltransferase, and
mean corpuscular volume in alcohol-dependent patients. Alcohol Clin
Exp Res 24:1400–1405.
National Institute on Alcohol Abuse and Alcoholism (1995) The Physi-
cian’s Guide to Helping Patients With Alcohol Problems. National Insti-
tutes of Health, Rockville, MD (NIH No. 95-3769).
Rehm J (1998) Measuring quantity, frequency, and volume of drinking.
Alcohol Clin Exp Res (Suppl) 22:4S–14S.
Rehm J, Greenfield TK, Rogers JD (2001) Average volume of alcohol
consumption, patterns of drinking, and all-cause mortality: Results from
the US National Alcohol Survey. Am J Epidemiol 153:64–71.
NEUMANN ET AL.
Reinert DF, Allen JP (2002) The Alcohol Use Disorders Identification
Test (AUDIT): A review of recent research. Alcohol Clin Exp Res
Rivara FP, Koepsell TD, Jurkovich GJ, Gurney JG, Soderberg R (1993)
The effects of alcohol abuse on readmission for trauma. JAMA 270:
Rumpf HJ, Hapke U, Meyer C, John U (2002) Screening for alcohol use
disorders and at-risk drinking in the general population: Psychometric
performance of three questionnaires. Alcohol Alcohol 37:261–268.
Runge JW, Hargarten S, Velianoff G, Brewer PA, D’Onofrio G, Soder-
strom CA, Gentilello LM, Flaherty L, Fiellin DA, Degutis LC, Pantalon
MV (2001) Recommended Best Practices of Emergency Medical Care
for the Alcohol-Impaired Patient: Screening and Brief Intervention for
Alcohol Problems in the Emergency Department. Available at: http://
Saunders JB, Aasland OG, Babor TF, de la Fuente JR, Grant M (1993)
Development of the Alcohol Use Disorders Identification Test (AU-
DIT): WHO Collaborative Project on Early Detection of Persons With
Harmful Alcohol Consumption–II. Addiction 88:791–804.
Seppa K, Makela R, Sillanaukee P (1995) Effectiveness of the Alcohol
Use Disorders Identification Test in occupational health screenings.
Alcohol Clin Exp Res 19:999–1003.
Shakeshaft AP, Bowman JA, Sanson-Fisher RW (1999) A comparison of
two retrospective measures of weekly alcohol consumption: Diary and
quantity/frequency index. Alcohol Alcohol 34:636–645.
Soderstrom CA, Dischinger PC, Kerns TJ, Kufera JA, McDuff DR,
Gorelick DA, Smith GS (1998) Screening trauma patients for alcohol-
ism according to NIAAA guidelines with alcohol use disorders identi-
fication test questions. Alcohol Clin Exp Res 22:1470–1475.
Spies CD, Neuner B, Neumann T, Blum S, Muller C, Rommelspacher H,
Rieger A, Sanft C, Specht M, Hannemann L, Striebel HW, Schaffartzik
W (1996) Intercurrent complications in chronic alcoholic men admitted
to the intensive care unit following trauma. Intensive Care Med 22:286–
Spies CD, Toennesen H, Andreasson S, Helander A, Conigrave K (2001)
Perioperative morbidity and mortality in chronic alcoholic patients.
Alcohol Clin Exp Res (Suppl) 25:164S–170S.
Steinbauer JR, Cantor SB, Holzer CE 3rd, Volk RJ (1998) Ethnic and sex
bias in primary care screening tests for alcohol use disorders. Ann
Intern Med 129:353–362.
Stockwell T, McLeod R, Stevens M, Phillips M, Webb M, Jelinek G (2002)
Alcohol consumption, setting, gender and activity as predictors of in-
jury: A population-based case-control study. J Stud Alcohol 63:372–379.
WHO Department of Mental Health and Substance Dependence (2000)
Noncommunicable Diseases and Mental Health Cluster, International
Guide for Monitoring Alcohol Consumption and Related Harm. World
Health Organization, Geneva. Available at: http://www.who.int/sub-
AUDIT IN TRAUMA PATIENTS