1 Health and Social Security Department, Program on Substance Abuse, Autonomous Government of Catalonia,
2 Alcohol Unit, Hospital Clinic, Barcelona, Spain and
3 Centre for Alcohol and Drug Studies, University of Northumbria at Newcastle upon Tyne, UK
Received 19 September 2001; in revised form 1 February 2002; accepted 27 May 2002
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Concern about those drinking over recommended levels has led to the concept of risky drinking (Higgins-Biddle and Babor, 1996), usually referring to men drinking >280 g (168 g for women) of alcohol per week. Although not defined precisely, the concept of risky drinking is often used to include both hazardous and harmful drinking.
Several studies have shown that drinking above World Health Organization recommendations (Dawson and Archer, 1993; Anderson, 1996
; Prada et al., 1996
; Portella et al., 1998
) raises the risk of alcohol-related physical and psychosocial problems (100% for liver cirrhosis, 2030% for cancer of the oral cavity, pharynx and larynx, 10% for cancer of the oesophagus, 14% for cancer of the liver, 1020% for cancer of the female breast and possibly 20% for stroke). Overall, the World Health Organization estimates that, in developed countries, alcohol accounts for 1011% of all illnesses and deaths each year (Murray and Lopez, 1996
). Risky drinking and alcohol misuse or dependence are common in primary health care (PHC) patients. In European countries, the prevalence of risky drinking in PHC settings ranges from 2.1 to 41% among men and from 0.8 to 21% among women (World Health Organization, 2001
). This high prevalence of risky drinking has led to the development of screening tools and brief intervention packages (Heather et al., 1987
; Gomel et al., 1994
; University of Sydney, 1994
; National Institute on Alcohol Abuse and Alcoholism, 1995
; Òrgan Tècnic de Drogodependencies, 1996
), which aim to help general practitioners (GPs) and other health care professionals to identify, assess and advise risky drinkers. The fact that brief interventions have proven efficacy (Wallace et al., 1988
; Babor and Grant, 1992
; Nuffield Institute for Health, 1993
; Altisent et al., 1997
; Fernández et al., 1997
) and cost-effectiveness (Fleming et al., 2000
) clearly increases the need to develop reliable, valid and user-friendly screening tools.
Fiellin et al. (2000) concluded in their systematic review that the literature supports screening for lifetime and current misuse or dependence disorders by means of the CAGE questions and, for less severe alcohol problems, such as at-risk, harmful, and hazardous drinking, by means of the Alcohol Use Disorders Identification Test (AUDIT).
The AUDIT (Saunders and Aasland, 1987; Saunders et al., 1993
) was developed as part of the World Health Organization Collaborative Project on the Detection and Management of Alcohol-related Problems in Primary Health Care, to identify hazardous and harmful alcohol use. AUDIT is a 10-item questionnaire that enquires about alcohol consumption and frequency of drinking, the presence of alcohol-related problems and alcohol-dependence symptoms. Several studies have shown its validity and reliability in the detection of risky drinking, alcohol misuse and alcohol dependence (Bohn et al., 1995
; Martínez, 1996
; Piccinelli et al., 1997
; Volk et al., 1997
; Rubio et al., 1998
; Contel et al., 1999
). Depending on the cut-off and the criterion standards used, studies have reported sensitivities between 51 and 97% and specificities between 78 and 96% (Fiellin et al., 2000
).
Seppä et al. (1998) developed the Five-Shot Questionnaire for detecting risky drinking, by combining two items from AUDIT asking about drinking amounts and three items from CAGE that correspond to the three different types of question in the AUDIT (hazardous alcohol consumption, dependence symptoms and harmful alcohol consumption). This instrument was tested in a middle-aged male population and, although it performed better than the CAGE, its usefulness among other age groups, among women and in PHC settings has not been demonstrated.
The AUDIT-C (Bush et al., 1998; Aertgeerts et al., 2001
; Gordon et al., 2001
) includes only the three AUDIT alcohol consumption questions and its performance as a screening test has been evaluated in three different studies. Bush et al. (1998)
evaluated the AUDIT-C for active alcohol misuse or dependence and/or risky drinking in a male population. Although the AUDIT-C performed better than the full AUDIT and the CAGE in identifying risky drinkers, this study was restricted to men, performed at three Veterans Affairs (VA) general medical clinics and the interviews were conducted by telephone. Recent data (Kraus and Augustin, 2001
) suggest that telephone interviews can produce a significant bias in results. Gordon et al. (2001)
used the AUDIT-C to identify hazardous drinkers in a large PHC sample. The AUDIT-C proved to be as effective as the AUDIT, even though criteria for hazardous drinking were not established on the basis of clinical judgement, but using quantityfrequency measures obtained from a self-administered questionnaire. In general, the AUDIT-C has shown a sensitivity of 54 to 98% and a specificity of 57 to 93% for various definitions of heavy drinking (Fiellin et al., 2000
).
In Europe, a large study of alcohol screening questionnaires in PHC carried out in Belgium (Aertgeerts et al., 2001) compared the full AUDIT with two shorter forms (Bush et al., 1998
; Gordon et al., 2001
) and the 5-shot questionnaire (Seppä et al., 1998
). The AUDIT-C performed significantly less well than the full AUDIT among female patients, but compared well with other questionnaires. It should also be noted that this study focused on alcohol misuse and dependence, not on risky drinking. Nevertheless, the authors suggested that the simplicity of the AUDIT-C supports its routine use by GPs.
Throughout Phase III of the World Health Organization Collaborative Study (Implementing and Supporting Early Intervention Strategies in Primary Health Care) the opinions of all GPs who participated were collected by means of systematic focus groups. Most GPs complained about the fact that it was too long to be used as a systematic screening tool. They also stated that the alcohol consumption questions (13) were well accepted by patients, but questions dealing with alcohol dependence and alcohol-related problems (410) tended to arouse defensiveness. These observations, together with the encouraging results previously obtained with short forms of the AUDIT (Piccinelli et al., 1997; Bush et al., 1998
; Aertgeerts et al., 2001
; Gordon et al., 2001
) led us to design the present study, in order to identify suitable short versions of AUDIT and, if appropriate, to test their effectiveness as screening tests for risky drinking among men and women in PHC settings. Also, this study provided an opportunity to examine possible differences between a population with a Mediterranean drinking pattern and other populations previously studied.
![]() |
SUBJECTS AND METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Alcohol screening measures
Before the interview, patients were asked to complete the 10-item AUDIT. GPs interviewed patients about their drinking habits using a Systematic Interview of Alcohol Consumption that included three questions exploring frequency and amount of consumption: If you ever drink alcoholic beverages (wine, beer, etc.), how many beverages a day? (measured in standard drinks); How often? (number of days in a week); and On weekends (or workdays) do your drinking habits change?. Detailed information about normal and exceptional drinking patterns was obtained with this recently validated instrument (Gual et al., 2001). The clinicians diagnosis of risky drinking was used as a gold standard. GPs identified as risky drinkers all patients whose weekly alcohol consumption was above the World Health Organization recommendations (280 g for men and 168 g for women) and/or who fulfilled criteria for hazardous or harmful drinking (World Health Organization, 2000
).
Statistical analysis
The Statistical Package for the Social Sciences (SPSS 9.0) was used for data analysis. Logistic regression analysis was carried out to identify those items that minimized the probability of misclassification between subjects with and without risky drinking. Following the Piccinelli et al. (1997) study, a stepwise selection of items was adopted by using the likelihood ratio statistic as a test for removal and a P-value of 0.10 to remove an item. Cross-tabulations and t-tests were used for group comparisons. Relationships between the three forms of AUDIT and reported mean weekly alcohol consumption were examined by regression analysis. Sensitivity, specificity, positive predictive and overall accuracy (OA) values were calculated for the full, 3-item and 4-item AUDIT forms in relation to the gold standard (diagnosis of risky drinking). Receiver operating characteristics (ROC) curves and their areas under the curve (AUROCs) were inspected to choose the optimal screening test and the best cut-off scores according to sensitivity and specificity levels. To compare AUROCs, we used the Z statistic defined following the method described by Hanley and McNeil (1983)
which takes into account the correlation between the areas that is induced by the paired nature of the data.
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Reported weekly alcohol consumption was translated into Spanish standard drinks (Gual et al., 1999) (one drink = 10 g of pure ethanol). Alcohol consumption in men ranged from 0 to 105 standard drinks and for women from 0 to 32 (Table 1
). AUDIT-10 scores ranged in men from 0 to 34 (mean ± SD: 6.76 ± 5.64) and in women from 0 to 10 (2.75 ± 1.91).
|
|
According to clinician diagnosis, 41.7% of men and 8.6% of women were classified as risky drinkers. According to the AUDIT cut-off scores established in its Spanish validation (9 for men and
6 for women), 31.5% of men and 6.3% of women were classified as risky drinkers.
ROC curves comparing the three AUDIT forms with the gold standard are presented with their AUROCs and 95% confidence intervals for men and women in Figs 1 and 2. Sensitivities and specificities at different cut-off scores are detailed for both genders in Table 3
. Among men, AUROCs (Fig. 1
) were 0.913 for AUDIT-3, 0.924 for AUDIT-4 and 0.920 for AUDIT-10 and no significant differences were found in comparisons among them (AUDIT-3 vs AUDIT-10, P = 0.646; AUDIT-4 vs AUDIT-10, P = 0.751). When comparing all AUROCs (0.957, 0.945 and 0.871 respectively) among women (Fig. 2
), no significant differences were found (AUDIT-3 vs AUDIT-10, P = 0.889; AUDIT-4 vs AUDIT-10, P = 0.90).
|
|
|
Risky drinker prevalences according to the best cut-off scores for AUDIT-3, -4 and -10 are also presented in Table 4.
|
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
As a comparison standard, we used the clinical diagnosis of risky drinking made by study physicians after interviewing patients attending the PHC centre. Clinicians made 10% more risky drinking diagnoses among men (2% more among women) than those made by AUDIT on the basis of Spanish cut-off scores established in previous validation studies (Rubio et al., 1998
; Contel et al., 1999
). Furthermore, The National Household Survey implemented in Spain in 1999 (Dirección General para el Plan Nacional de Drogas, 2001
) showed a heavy drinking prevalence of
9.5% (12.1% for males and 6.9% for females). This difference may be explained by the fact that the criterion standard used in our study was more rigorous and included criteria for hazardous and harmful drinking. The study clinicians had also been deeply trained in screening and brief intervention strategies in conjunction with the Drink-Less Program. In addition, the percentages found in our study (Table 4
) are higher than those of harmful and hazardous drinking (from 1 to 13%) obtained in previous studies. These discrepancies may be explained by the variation in risky alcohol consumption levels etablished. Since alcohol consumption in women has been increasing in Mediterranean countries (Dirección General para el Plan Nacional de Drogas, 2001
), the low prevalence obtained in females should be a cause of concern. This phenomenon may be explained partly by the tendency to hide alcohol consumption, which has been recorded as higher in females.
Correlations between the scores of the three AUDIT forms and alcohol consumption (in standard drinks) were positive and highly significant. The three AUDIT forms performed similarly and had equivalent AUROCs for detecting risky drinking among men and women attending PHC centres. No statistically significant differences were found when comparing the three AUROCs in either gender. Both short screening forms showed acceptable sensitivity and specificity levels for the detection of risky drinkers.
Although the standard criteria applied, the population screened and settings have varied in all studies previously published, the present results are partially in agreement with those studies (Bush et al., 1998; Aertgeerts et al., 2001
; Gordon et al., 2001
). The AUDIT cut-off scores which emerged in our study were higher (
7 among males,
5 among females) than those previously proposed (
5 among males and females), but were still lower than those recommended in the validation studies (
9 among males,
6 among females). For the shorter forms of AUDIT, the cut-off scores were similarly reduced.
The results for male patients found with the AUDIT-3 (or AUDIT-C), including the cut-off point (5), fit well with previous studies. For female patients, AUDIT-4 performed exactly as the full AUDIT at the same cut-off points. The results reported here have led us to recommend that clinicians use the AUDIT-3 questionnaire in general health screening interventions. If a score of
5 among men and
4 among women is observed, a more in-depth assessment of drinking pattern and alcohol-related problems should be carried out. If including question 10 increases the total score to 7 among men and 5 among women, this should definitely lead to a diagnosis of risky drinking. More detailed questioning, including the full AUDIT if necessary, could be used to make decisions about the need for referral to specialist help for alcohol dependence. Screening devices, such as the AUDIT in all of its self-administered forms, facilitate the GPs daily work, and, as our results confirm, are as reliable as a systematic GPs interview on alcohol consumption.
![]() |
ACKNOWLEDGEMENTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
FOOTNOTES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Altisent, R., Córdoba, R., Delgado, M. T., Pico, M. V., Melus, E., Aranguren, F., Alvira, U., Barberá, C., Moran, J. and Reixa, S. (1997) Estudio multicéntrico sobre la eficacia del consejo para la prevención del alcoholismo en atención primaria (EMPA). Medicina Clínica 109, 121124.[ISI][Medline]
Anderson, P. (1996) Alcohol and Primary Health Care. World Health Organization Regional Office, Copenhagen.
Babor, T. F. and Grant, M. (1992) WHO Collaborating investigators project on identification and management of alcohol-related problems. Combined analyses of outcome data: the cross-national generalizability of brief interventions. Report on phase II: a randomized clinical trial of brief interventions in primary health care. World Health Organization Regional Office, Copenhagen.
Bohn, M. J., Babor, T. F. and Kranzler, H. R. (1995) The Alcohol Use Disorders Identification Test (AUDIT): validation of a screening instrument for use in medical settings. Journal of Studies on Alcohol 56, 423432.[ISI][Medline]
Bush, K., Kivlahan, D. R., McDonell, M. S., Fihn, S. D. and Bradley, K. A. (1998) The AUDIT Alcohol Consumption Questions (AUDIT-C): an effective brief screening test for problem drinking. Archives of Internal Medicine 158, 17891795.
Contel, M., Gual, A. and Colom, J. (1999) Test para la identificación de transtornos por uso de alcohol (AUDIT): traducción y validación del AUDIT al catalán y castellano. Adicciones 11, 337347.
Dawson, D. A. and Archer, L. D. (1993) Relative frequency of heavy drinking and the risk of alcohol dependence. Addiction 88, 15091518.[ISI][Medline]
Dirección General para el Plan Nacional de Drogas (DGPND) (2001) Informe n° 4-Marzo del 2001 del Observatorio Español sobre Drogas. Ministerio del Interior, Spain.
Fernández, M. I., Bermejo, C. J., Alonso, M., Herreros, B., Nieto, M., Novoa, A. and Marcelo, M. T. (1997) Efectividad del consejo médico breve para reducir el consumo de alcohol en bebedores. Atención Primaria 19, 127132.[Medline]
Fiellin, D. A., Reid, M. C. and OConnor, P. G. (2000). Screening for alcohol problems in primary care: a systematic review. Archives of Internal Medicine 160, 19771989.
Fleming, M. F., Mundt, M. P., French, M. T., Manwell, L. B., Stauffacher, E. A. and Barry, K. L. (2000) Benefit-cost analysis of brief physician advice with problem drinkers in primary care settings. Medical Care 38, 718.[Medline]
Gomel, M. K., Saunders, J. B., Burns, L., Hardcastle, D. M. and Sumich, M. (1994) Dissemination of early intervention for harmful alcohol consumption in general practice. Health Promotion Journal of Australia 4, 6569.
Gordon, A. J., Maisto, S. A., McNeil, M., Kraemer, K. L., Conigliaro, R. L., Kelley, M. E. and Conigliaro, J. (2001) Three questions can detect hazardous drinkers. Journal of Family Practice 50, 313320.[ISI][Medline]
Gual, A., Martos, A. R., Lligona, A. and Llopis, J. J. (1999) Does the concept of a standard drink apply to viticultural societies? Alcohol and Alcoholism, 34, 15360.[Abstract]
Gual, A., Segura, L., Contel, M. and Colom, J. (2001) The ISCA (Systematic Interview of Alcohol Consumption), a new instru-ment to detect risky drinking. Medicina Clinica (Barcelona) 117, 685689.[ISI][Medline]
Hanley, J. A. and McNeil, B. J. (1983) A method of comparing the areas under receiver operating characteristic curves derived from the same cases. Radiology 148, 839843.[Abstract]
Heather, N., Campion, P. D., Neville, R. G. and Maccabe, D. (1987) Evaluation of a controlled drinking minimal intervention for problem drinkers in general practice (The DRAMS scheme). Journal of the Royal College of General Practitioners 37, 358363.[ISI][Medline]
Higgins-Biddle, J. C. and Babor, T. F. (1996) Reducing Risky Drinking: A Report on Early Identification and Management of Alcohol Problems Through Screening and Brief Intervention. (Report to Robert Wood Johnson Foundation.) Alcohol Research Center, University of Connecticut Health, Farmington, CT.
Hollingshead, A. (1957) Two-factor Index of Social Position. New Haven, CT.
Kraus, L. and Augustin, R. (2001) Measuring alcohol consumption and alcohol-related problems: comparison of responses from self-administered questionnaires and telephone interviews. Addiction, 96 459471.[ISI][Medline]
Martínez, J. M. (1996) Validación de los cuestionarios breves: AUDIT, CAGE y CBA para la detección precoz del síndrome de dependencia de alcohol en Atención Primaria. Tesis Doctoral, Departamento de Neurociencias, Universidad de Cadiz.
Murray, C. M. and Lopez, A. (1996) The Global Burden of Disease: A Comprehensive Assessment of Mortality and Disability from Diseases, Injuries, and Risk Factors in 1990 and Projected to 2020. World Health Organization, Geneva.
National Institute on Alcohol Abuse and Alcoholism (1995) Screening and Brief Intervention Package. NIAAA, Rockville, MD.
Nuffield Institute for Health (1993) Brief interventions and alcohol use. Effective Health Care Bulletin 7, 114.
Òrgan Tècnic de Drogodependències (1996) Dossier Beveu Menys. Departament de Sanitat i Seguretat Social, Barcelona.
Piccinelli, M., Tessari, E., Bortolomasi, M., Piasere, O., Semenzin, M., Garzotto, N. and Tansella, M. (1997) Efficacy of the Alcohol Use Disorders Identification Test as a screening tool for hazardous alcohol intake and related disorders in primary care: a validity study. British Medical Journal 314, 420424.
Portella, E., Ridao, M., Carrillo, E., Ribas, E., Ribó, C. and Salvat, M. (1998) El alcohol y su abuso: impacto socioeconómico. Ed. Médica Panamericana, Madrid.
Prada, C., Del Río, M. C., Yañez, J. L. and Alvarez, F. J. (1996) Mortalidad relacionada con el consumo de alcohol. Gaceta Sanitaria 10, 161168.[Medline]
Rubio, G., Bermejo, J., Caballero, M. C. and Santo-Domingo, J. (1998) Validación de la Prueba para la Identificación de Transtornos por Uso de Alcohol (AUDIT) en Atención Primaria. Revista Clínica Española 198, 1114.[ISI][Medline]
Saunders, J. B. and Aasland, O. G. (1987) WHO Collaborative Project on the identification and treatment of persons with harmful alcohol consumption. Report on phase I: development of a screening instrument. World Health Organization, Geneva.
Saunders, J. B., Aasland, O. G., Babor, T. F., de la Fuente, J. R. and Grant, M. (1993) Development of the Alcohol Use Disorders Identification Test (AUDIT): WHO Collaborative Project on Early Detection of Persons with Harmful Alcohol Consumption II. Addiction 88, 791804.[ISI][Medline]
Seppä, K., Lepistö, J. and Sillanaukee, P. (1998) Five-shot questionnaire on heavy drinking. Alcoholism: Clinical and Experimental Research 22, 17881791.[ISI][Medline]
University of Sydney (1994) Drink-less. University of Sydney, Sydney, NSW.
Volk, R. J., Steinbauer, J. R., Cantor, S. B. and Holzer, C. E. (1997) The Alcohol Use Disorders Identification Test (AUDIT) as a screen for at-risk drinking in primary care patients of different racial/ethnic backgrounds. Addiction 92, 197206.[ISI][Medline]
Wallace, P., Cutler, S. and Haines, A. (1988) Randomized controlled trial of general practitioner intervention in patients with excessive alcohol consumption. British Medical Journal 297, 663668.[ISI][Medline]
World Health Organization (2000) International Guide for Monitoring Alcohol Consumption and Related Harm. Department of Mental Health and Substance Dependence. World Health Organization, Geneva.
World Health Organization (2001) Alcohol in the European Region Consumption, Harm and Policies. World Health Organization Regional Office, Copenhagen.