Original Research

Alcohol misuse in patients attending a defence force general medical clinic

Bekker D, MBChB, MFamMed Van Velden DP, MBChB, MPraxMed Department of Family Medicine, University of Stellenbosch

Correspondence: Dr D Bekker, 3 Grobler Street, Aurora , 7550 Tel: 021-5902635, e-mail: iti00729@.co.za Keywords: alcohol misuse; Alcohol Use Disorders Identification Test (AUDIT); general medical practice; defence force; prevalence

Abstract

Aim.: To determine the prevalence of alcohol misuse in adult patients attending a defence force general practice clinic. Design: Descriptive cross-sectional survey. Methods: The setting is a military sickbay situated in the metropolitan area, South Africa. Participants included all adult patients (³ð18 years) attending the clinic during a three week period midyear. The Alcohol Use Disorders Identification Test (AUDIT) questionnaire was completed by patients. Two direct alcohol consumption questions were added determining the total drinks per week as well as the maximum number of drinks per occasion. Results: Of the 658 patients approached, 40 (6%) declined to fill in the questionnaire. 42% of the 618 patients surveyed were female and 58% were male with an average age of 41 years. The group consisted of 44% current defence force members, 25% retired defence force members, 22% medical dependents and 7% civilians. The overall rate of alcohol misuse according to the AUDIT-questionnaire was 13,27%. Among male patients, 20,6% scored as misusers of alcohol, compared with 2,7% among females. A much higher level of misuse was also found under younger age groups. 38,1% of male apprentices at the technical college and 40,0% of male current defence force members staying on the base scored 8 or more on the AUDIT. In the group scoring less than 8 on the AUDIT-questionnaire a further 9 patients reported drinking on average more than 14 drinks per week for males and more than 9 drinks per week for females. In this group (AUDIT-score <8) 76 patients reported maximum number of drinks per occasion of more than 4 for men and more than 3 for women. If these two questions are included the overall rate of alcohol misuse found was 26,05%. Conclusion: The high rate of alcohol misuse found in this group of patients is need for concern. The overall rate of alcohol misuse was not found to be higher than what was found in other studies in South Africa as well as other countries. This study supports the implementation of screening and intervention strategies in general medical clinics in the defence force.

SA Fam Pract 2003;45(2):___

○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○

INTRODUCTION dependence in the adult population.2 the only study found pertaining to However, relatively few studies have military personnel looked at the pattern Alcohol misuse may come to be been published on drinking patterns in of drug-taking of 188 drug-dependent recognized as one of the most significant South Africa. With regards to adults, national servicemen seen in the public health concerns facing South epidemiological studies in recent years department of psychiatry of 1 Military Africa over the next few years. This is have also focussed almost entirely on Hospital in Pretoria in the year 1971. a direct result of alcohol’s impact on the the black and coloured population while Alcohol were used by only 18% of the health services, the economy, and the they are not the only groups who have subjects.3 In our study practice we are South African society as a whole.1 high levels of risky drinking.1 not aware of many alcohol dependent Alcohol misuse is common world- There has always been a perception patients. According to the social work wide and in South Africa there is an that a high prevalence of alcohol misuse department, they have detected a high estimated prevalence of 10% of alcohol exists amidst defence force patients, but prevalence of alcohol misuse especially

10 SA Fam Pract 2003;45(2) Original Research

under the younger uniformed members The Alcohol Use Disorders Identifi- compared to the Structured Clinical (age 18-25) who are busy with their cation Test (AUDIT) questionnaire29-34 Interview for DSM-III-R (used as the apprenticeship at the military technical was given for self-completion to patients “gold standard” for alcohol abuse or college, but no effort has been made in during normal clinic hours (07h30 - dependence).9 our practice to screen or treat for alcohol 16h00). Questions additional to those Results of the AUDIT have also been misuse as part of an organised health posed in the questionnaire were added associated with more distal indicators promotion plan. with regards to personal information as of problematic drinking. It was found Several studies have shown that well as alcohol consumption to deter- to be a good predictor of both alcohol- doctors fail to detect alcohol misuse in mine the prevalence of at-risk drinking. related social and medical problems.31,35-36 the majority of their patients. They are It has been demonstrated that direct The AUDIT questionnaire is currently particularly unlikely to identify misuse questions about average daily the only validated instrument designed in its earlier stages and therefore miss consumption or frequency of heavy to detect hazardous drinking. major opportunities for intervening.4-17 consumption are less sensitive than Research has established the effective- standard self-report screening question- Hazardous drinking ness of even brief interventions by naires in detecting alcohol misuse. Hazardous drinking is defined as a primary care physicians on patients with Biochemical markers, especially serum quantity or pattern of alcohol consump- hazardous and harmful alcohol gamma glutamyl transferase can assist tion that places individuals at risk for consumption.18-25 in identifying problem drinking but are adverse health events and is recognised New evidence continues to appear also less valid than self-report measures. by the World Health Organisation appertaining to levels and patterns of In detecting alcohol dependence or (WHO) as a distinct disorder. The alcohol use in relation to health effects, abuse, the most widely used instruments quantity or pattern of alcohol consump- social well-being, and economic costs. are the CAGE29-34 with 4 items and the tion that constitutes hazardous drinking Data from several recent large prospec- Michigan Alcoholism Screening Test is typically specified by setting tive studies suggest that alcohol (MAST)29-34 with 25 questions. They threshold values for an individual’s consumption in quantities consistent have proven to be useful screening average number of drinks consumed per with hazardous and harmful drinking instruments and have revealed a higher week or per occasion.26 may increase the risk of adverse health validity compared to laboratory data. events, such as haemorrhagic stroke and They are however insensitive for Heavy drinking breast cancer.26-28 hazardous or heavy drinking. The Heavy drinking is defined as a quantity Every effort should thus be made in Alcohol Use Disorder Identification of alcohol consumption that exceeds an primary care settings to detect and treat Test (AUDIT) was developed by the established threshold value.26 The alcohol misuse in its earlier stages World Health Organisation in an Addiction Research Foundation of (heavy, hazardous and harmful drinking) international multi-site trial29 in an Ontario and Canadian Centre on because it is more common and pro- attempt to also detect earlier stages of Substance Abuse adopted threshold bably more responsive to treatment than drinking which place the patient at risk levels of more than 14 drinks per week alcohol abuse or dependence. for suffering eventual severe alcohol for men and more than 9 drinks per week By looking at the prevalence and related problems. The authors gave for women based on current scientific pattern of alcohol misuse in patients particular attention to selecting items evidence.27 Individuals whose drinking attending our clinic we aimed to generalisable across culture, gender and exceeds these guidelines are thought to determine the size and distribution of age. Most of the AUDIT questions be at increased risk for adverse health the problem to enable us to motivate for inquire about the previous year rather events. The definitions of hazardous and do proper planning towards inter- than ever in the patient’s lifetime, and heavy drinking are thus similar. vention as part of health promotion. thereby decreasing errors of mislabeling individuals who have already resolved However, the threshold levels set for METHODS earlier problems with alcohol. hazardous drinking in developing the Questions on the AUDIT covers the AUDIT were an average daily consump- A descriptive, cross-sectional survey three domains of consumption, alcohol tion of 60g per day (6 drinks) for men design was used. The study setting was dependence and adverse consequences and 40g (4 drinks) for women.29 If a Wingfield Military Sickbay situated at of drinking.29-34 male were to drink 3 units of alcohol Goodwood, one of the northern suburbs At a cut-off value of 8 of the possible per day for 7 days a week and have no of Cape Town. Participants in the study 40 points on the test, sensitivity and adverse consequences of drinking or included all adult patients (≥ 18 years) specificity coefficients are high. signs of dependence, he may, depending attending the general medical clinic for Saunders et al reported a 92% sensi- on the rest of his answers on the AUDIT a period of three weeks (28 June to 18 tivity and 94% specificity for hazardous questionnaire, score only 5 points. July 2000). We aimed at a minimum and harmful drinking and a 100% Although women suffer from adverse sample size of 340-350 for validity sensitivity for alcohol dependence.29 consequences of drinking at lower levels purposes. The patient population served Isaacson et al found sensitivity and of consumption than men, the AUDIT by the clinic was estimated at more than specificity of 96% in the detection of like other validated questionnaires, does 11000. alcohol abuse and dependence not use a separate scoring system based

SA Fam Pract 2003;45(2) 11 Original Research

on gender.37 A woman who drinks 2 reasons stated were as follows: 2 did not scoring less than 8. Among male drinks per day would by definition and have time, 3 felt too ill, 4 could not write patients, 20,6% scored as misusers of current evidence on high risk drinking (due to stroke, Parkinson’s disease and alcohol, compared with 2,7% among be considered a hazardous or heavy hand injuries) and 11 left their reading females. Of the retired defence force drinker, but may score only 4 on the glasses at home. Of the other 20 people, members (average age of 58), 5.2% AUDIT. It has therefore been decided 3 stated that they did not drink, 1 felt scored positive compared to 24,2% of to add 2 questions on consumption that the questionnaire was not relevant the current defence force members which follows after the AUDIT, deter- to him, 1 wanted the questionnaire in (average age of 29). Higher levels of mining the total drinks per week as well Afrikaans, 1 wrote that he felt uncom- alcohol misuse was found in lower age as the highest total of drinks per fortable, 1 wrote that it was too personal groups as demonstrated in Table I. All occasion. Many studies define heavier and 14 gave no reasons. women scoring 8 or more on the drinking episodes as occasions on which Of the 618 patients surveyed, 42% AUDIT-questionnaire were below the more than 4 drinks were consumed. The were female and 58% were male. 58% age of 46 years. National Institute of Alcohol Abuse and Alcoholism (USA) sets this threshold at >4 drinks for men and >3 drinks for Table I: Age of males scoring 8 or more on the AUDIT women. However, there is no strong Age group 18-25 26-35 36-55 71-86 evidence favouring any specific amount as marking a discrete threshold of Frequency 30 29 14 1 risk.26,27 Percentage 34,5% 28,2% 14,1% 1,4% To ensure that confidentiality about the identity of the patients were pre- Total in age group 87 103 99 70 served, questionnaires were completed anonymously and no code were kept of indicated that they were white, 18% In the current defence force group, the identities of patients. After coloured, 7% black, 2% Indian and 14% 69,2% (9 out of only 13) Indians scored completion, the questionnaire were declined to reveal there race. Of this as misusers compared to 32,5% (13/40) deposited by the patient in a visibly 14%, 10 (1,6%) stated that their home blacks, 25,0% (19/76) coloureds, 17,7% locked wooden box. The scoring system language were African (Xhosa, Zulu, (17/96) whites and 16,7% (8/48) in the were not indicated on the questionnaire. Tswana, S/Sotho), 49 (7,9%) Afrikaans, unknown race group. Due to the diversity of home 26 (4,2%) English and 4 (0,6%) Afri- Male current defence force members languages in the defence force it was kaans and English. The majority of the had a positive score of 28,9% (65/225). decided to present the questionnaire in sample’s home language was Afrikaans Wingfield being a naval base, mainly English seeing that it is the language (54,7%) followed by English (33,5%), current naval defence force members mastered by the majority of the patients. Afrikaans and English (3,4%), Xhosa were seen with 29,2% (52/178) males A research assistant was appointed (3,4%), Zulu (1,5%), Tswana (1,5%), S/ scoring as misusers of alcohol compared and trained to distribute the question- Sotho (1,4%), Tsonga (0,5%), Ndebele to 30,0% (9/30) male air force members. naires to the patients in the waiting (0,2%), N/Sotho (0,2%) and Tsonga/ The numbers for the army and medical room. This person was available full- Swazi (0,2%). services were too small for comparative time to assist with any problems that The group consisted of 44% current purposes. 38,1% (32/84) of male might arose in completing the defence force members, 25% retired apprentices studying for electrical or questionnaire. defence force members (85% white, 6% mechanical fitters at SAS Wingfield Before finalisation, the questionnaire coloured, 9% race unknown), 22% Naval College scored as misusers of was first piloted on 6 patients. medical dependents (95% female) and alcohol. Descriptive statistics were deter- 7% civilians. 40,0% (24/60) of male current mined. Within the samples comparisons defence force members staying on the were made with respect to groups AUDIT results defence force base (average age of 26 formed by gender, race, occupation etc. The overall rate of alcohol misuse years), scored as misusers of alcohol Contingency tables were used to according to the AUDIT questionnaire according to the AUDIT-questionnaire. investigate the influence of discrete was 13,27%. Of the 536 people scoring 274 of the 618 people (44,3%) that classificatory factors. Percentages out below 8, 70 left out one or more ques- took part in the study, stated that they of a total less than 20 were calculated tions. However, 38 of the 70 indicated never drink alcohol. only for comparative purposes, but the that they never drink. Of the other 32, Weekly alcohol consumption reader are advised that these 15 could have scored 8 or more if they compared to the AUDIT questionnaire percentages are highly variable. filled in all the questions. If that would In males scoring <8 on the AUDIT have been the case, it would have questionnaire, only 5 of the 266 that RESULTS brought the total up to 15,67%. answered the question (1,9%) said that The average age of the 13,27% who they consumed a number of units in Of the 658 people approached, 40 (6%) scored 8 and higher was 29 years, excess of the limits set for low-risk declined to fill in the questionnaire. The compared to 43 years in the group drinking (>14 units per week). In

12 SA Fam Pract 2003;45(2) Original Research

However, it was not expected that this Table II: Derivation of sensitivity and specificity of weekly would lead to a unrepresentative peak alcohol consumption compared to the AUDIT in the results. The AUDIT-question- AUDIT-questionnaire naire also specifically refers to alcohol experiences in the past year. Weekly alcohol consumption 8 or more Less than 8 Total Six percent of patients approached Positive 15 9 24 declined to participate in the study. This compares well with the study of Negative 66 481 547 Isaacson et al.9 where 12% declined to Total 81 490 571 participate as well as the study of Sharkey et al.6 where 9% of outpatients Sensitivity = 15/81 = 18,5% declined. These were the only studies Specificity = 481/490 = 98,2% found where the AUDIT-questionnaire was used in similar settings. Only 20 females scoring <8 on the AUDIT In females scoring <8, only 22 out of of the 40 patients did not have a valid questionnaire, only 4 of the 224 that 233 who filled in the question (9,4%) reason for not participating and 3 of answered the question (1,8%) said that stated that the maximum number of these said that they did not drink. If all they consumed >9 units per week. drinks that they drank per occasion was of the 17 patients were regarded as Those whose AUDIT score was 8 or greater than 3. misusers of alcohol, the overall rate of more, had a less impressive showing in In males whose AUDIT-score was alcohol misuse would then be 15,6% comparison. All identified alcohol ³ð8, 63 out of 74 (85,1%) stated that (99/635) which is only slightly higher misusers answered this question, with their maximum number of drinks per than what was found. 13 of 74 males (17,3%) and 2 of 7 occasion was more than 4. All alcohol The questionnaire was handed out females declaring an intake above the misusers filled in this question (male for self-completion and therefore it can limit. and female). In females whose AUDIT- be expected that some questionnaires When compared to the AUDIT, self- score was ≥ 8, 7 out of 7 (100%) stated would not have been completed in full. reported alcohol consumption per week that their maximum number of drinks Fortunately most patients that left had a sensitivity of only 18,5% and a per occasion was more than 3. questions on the AUDIT unanswered, specificity of 98,2% (Table II). When When compared to the AUDIT, self- did so because of the fact that they were compared to self-reported alcohol reported maximum number of drinks per teetotallers and therefore felt that the consumption per week, the AUDIT occasion had a sensitivity of 86,4% and questions were not applicable to them. questionnaire had a sensitivity of 62,5% a specificity of 84,9% (Table III). When To try and prevent patients from not and a specificity of 87,9%. compared to self-reported maximum completing the questionnaire honestly, it was decided on an anonymous Maximum number of drinks per number of drinks per occasion, the questionnaire. The questions were occasion compared to the AUDIT AUDIT had a sensitivity of 47,9% and obviously of a very personal nature and questionnaire. a specificity of 97,5%. it is to be expected that some patients In males scoring <8 on the AUDIT, would be inclined to minimise their 54 out of 234 (23,1%) who filled in this DISCUSSION alcohol consumption habits, which question stated that the maximum could lead to an under-estimation of the number of drinks that they drank per Limitations of study prevalence. However, the aim of the occasion was greater than 4. Of these The data was obtained during 3 weeks study was explained on the question- 54 patients 7 indicated that their in the middle of the year which naire as well as by the research assistant, maximum was greater than 12 drinks. coincided with the school holidays. and the subjects were assured of complete confidentiality. Table III: Derivation of sensitivity and specificity of maximum Notable results number of drinks per occasion compared to the AUDIT The greatest influences on the rate of AUDIT-questionnaire alcohol misuse found were age and gender. There is by far a greater level Maximum number of 8 or more Less than 8 Total of alcohol misuse among males and drinks per occasion younger age groups. 28,9% of male Positive 70 76 146 current defence force members scored as misusers of alcohol on the AUDIT. Negative 11 427 438 Our suspicion was confirmed by finding Total 81 503 584 a high level of misuse (38,1%) under Sensitivity = 70/81 = 86,4% male apprentices studying at the technical college on the base. 40% of Specificity = 427/503 = 84,9% male defence force members staying on

SA Fam Pract 2003;45(2) 13 Original Research

the base also scored as misusers Detroit, USA. The rate of alcohol (1 out of only 8) black females scored according to the AUDIT-questionnaire. misuse was higher, with 21% (26/124) as misusers of alcohol. 24,7% of Even though low threshold levels identified. The average age of the coloured males and 2,8% of coloured were taken, self-reported alcohol sample was 45 years with 52% women. females scored as misusers according consumption per week had a sensitivity However, the patients were of low to the AUDIT-questionnaire. of only 18,5% compared to the AUDIT- socio-economic status and predomi- On average, alcohol consumption questionnaire confirming what we nantly of African American origin. rates in South Africa appear to be 15- already know namely that direct One other comparable study was 20% higher for men than for women in questions about alcohol consumption found which looked at the pattern of all ethnic groups.1 This study also found are less sensitive than screening alcohol consumption of a general a substantial difference in rates of questionnaires in detecting alcohol hospital population in north Belfast, alcohol misuse according to the misuse. It is however interesting to see Ireland (Sharkey et al.,1996)6. 15% of AUDIT-questionnaire between men that self-reported maximum number of outpatients scored as misusers of (20,6%) and women (2,7%). drinks per occasion had a much higher alcohol according to the AUDIT Two studies over 10 years ago sensitivity, namely that of 86,4%, questionnaire of which 76%, however (Rocha-Silva 1989,1991) compared compared to the AUDIT even though were female. Among male patients, 27% differences among race groups in the these thresholds were arbitrarily chosen scored as misusers, compared with 10% general population of South Africa and based on tradition rather than empirical among females which is higher than the found the highest level of risky drinking study. 20,6% and the 2,7% found in males and among blacks, followed by coloureds, When compared to self-reported females in our study group. The average Indians and whites. A fairly substantial alcohol consumption per week, the age was 45 which is slightly higher than 13% difference between white and black AUDIT-questionnaire had a sensitivity the average age of 41 in this study males was noted in the 1985 study of 62,5% and thus missed 9 patients who group. (Rocha-Silva 1989).1 The same trend had an increased risk for adverse health In South Africa high levels of alcohol was found in this study according to the events according to current evidence on dependence according to the CAGE AUDIT except for the Indian group at-risk drinking. questionnaire were found in very which consisted of only 15 patients in When compared to self-reported selected groups: 87% in coloured farm total and the high prevalence of alcohol maximum number of drinks per occa- workers in the Koue Bokkeveld38, 56% misuse found here (60,0%, 9/15), may sion, the AUDIT-questionnaire had a in a coloured rural community in Fraser- thus be due to random error. A sensitivity of 47,9% and so missed 76 burg 500 km from Cape Town39, 62% difference of 15% was found between patients who misuses alcohol according in hospitalised coloured TB patients40, black and white current defence force to threshold levels set by the National 31% in hospitalised black TB patients40 members where the average age were Institute of Alcohol Abuse and and 32% in black miners in Welkom41. comparable. Alcoholism (USA). 2 studies were conducted that targeted If both these numbers were included, the adult population in the Cape CONCLUSION the overall rate of alcohol misuse Peninsula, but again only the coloured detected by the AUDIT-questionnaire and black groups. Here, Gilles et al. In this group of patients attending a combined with these two direct (1973) found a prevalence of 19,6% defence force general medical clinic the consumption questions is then 26,05% (male) and 1,2% (female) of alcohol overall rate of alcohol misuse according (161/618). misuse under the general coloured to the AUDIT-questionnaire was not population of the and found to be higher than what was found Comparative studies Lombard and Steyn (1991) found a in other studies in South Africa as well In the USA, about 11% to 20% of prevalence of 26,7% (male) and 5% as in other countries. However, a 13% patients presenting to general medical (female) of alcohol misuse (³ð5 drinks rate of alcohol misuse according to the clinics are diagnosed as suffering from per day or communally on weekends) AUDIT-questionnaire is still need for alcohol abuse or dependence.30 This is under the black population of the Cape concern and with much higher levels of comparable to the 13,27% of alcohol Peninsula.1 With the new national misuse found in certain subgroups, it is misuse according to the AUDIT- defence force being more representative even more distressing. questionnaire that was found in this of all the race groups and the retired With current threshold levels of group, and the 26% through the defence force group consisting mainly consumption for hazardous drinking set combined methodology. of white patients, the average age of lower than what was originally defined Only one study was found that has black and coloured patients in this study with the development of the AUDIT, it used the AUDIT questionnaire to were much younger (26 years and 31 is recommended that the two direct determine the extent of alcohol misuse years respectively). Therefore you questions on consumption determining in those attending a general medical would expect to find a higher rate of average number of drinks per week as clinic. Isaacson et al. used the AUDIT- alcohol misuse in this study compared well as maximum number of drinks per questionnaire among those attending a to the general population as was the occasion be added to the AUDIT- general medical clinic in inner city case: 32,4% of black males and 12,5% questionnaire for routine screening for

14 SA Fam Pract 2003;45(2) Original Research

at-risk drinking. If findings from these problems in general practice. J Stud Alcohol 26. Bondy SJ, Rehm J, Ashley MJ, Walsh G, two consumption questions were 1992;53(3):197-202. Single E, Room R. Low-risk drinking 12. Buchsbaum DG, Buchanan RG, PosesRM, guidelines: the scientific evidence. Can J included in this study, the overall rate Schnoll SH, Lawton MJ. Pysician detection Public Health 1999;90(4):264-70. of alcohol misuse found was 26,05%. of drinking problems in patients attending 27. Parsons OA, Nixon SJ. Cognitive func- With brief intervention having been a general medicine practice. J Gen Intern tioning in sober social drinkers: a review of shown as a low cost, effective Med 1992;7(5):517-21. the research since 1986. J Stud Alcohol 13. Cleary PD, Miller M, Bush BT, Warburg 1998;59(2):180-90. preventive measure for heavy drinking MM, Delbanco TL, Aronson MD. Preva- 28. Saunders JB, Aasland OG, Babor TF, De la in primary care settings, the imple- lence and recognition of alcohol abuse in a Fuente JR, Grant M. Development of the mentation of intervention strategies primary care population. Am J Med Alcohol Use Disorders Identification Test should be considered in general medical 1988;85(4): 466-71. (AUDIT): WHO Collaborative Project on 14. Woodall HE. Alcoholics remaining anony- Early Detection of Persons with Alcohol clinics in the defence force and mous: resident diagnosis of alcoholism in a Consumption—II. Addiction 1993;88(6): elsewhere.r family practice center. J Fam Pract 791-804. 1988;26(3):293-6. 29. Allen J, Litten RZ, Lee A. What you need Acknowledgements: 15. Reid AL, Webb GR, Hennrikus D, Fahey PP, to know: detecting alcohol problems in Sanson-Fisher RW. Detection of patients general practice. Singapore Med J The author would like to thank the with high alcohol intake by general 1998;39(1):38-41. SANDF for allowing her to conduct and practitioners. Br Med J (Clin Res Ed) 1986; 30. Allen JP, Litten RZ, Fertig JB, Babor T. A publish this study as well as Theodata 293(6549):735-7. review of research on the Alcohol Use for statistical assistance. 16. Wallace P, Haines A. Use of a questionnaire Disorders Identification Test (AUDIT). in general practice to increase the Alcohol Clin Exp Res 1997;21(4):613-9. recognition of patients with excessive 31. Rumpf HJ, Hapke U, Hill A, John U. References alcohol consumption. Br Med J (Clin Res Development of a screening questionnaire 1. Parry CDH, Bennetts AL. Alcohol policy Ed) 1985;290(6486): 1949-53. for the general hospital and general and public health in South Africa. Cape 17. Ockene JK, Adams A, Hurley TG, Wheeler practices. Alcohol Clin Exp Res 1997;21(5): Town: Oxford University Press, 1998. EV, Hebert JR. Brief physician- and nurse 894-8. 2. Wilson D. Routine screening tests for alco- practitioner-delivered counselling for high- 32. Bohn MJ, Babor TF, Kranzler HR. The holism. S Afr Med J 1999;89(9):949-51. risk drinkers: does it work? Arch Intern Med Alcohol Use Disorders Identification Test 3. Levin A. The pattern of drug-taking among 1999;159(18):2189-205. (AUDIT): validation of a screening drug-dependant South African national 18. Fleming MF, Manwell LB, Barry KL, Adams instrument for use in medical settings. J servicemen. S Afr Med J 1972;46(44):1690- W, Stauffacher EA. Brief physician advice Stud Alcohol 1995;56(4):423-32. 4. for alcohol problems in older adults: a 33. Moore RD. Screening and assessment of 4. Sri EV, Raguram R, Srivastava M. Alcohol randomized community-based trial. J Fam alcohol and other drug abuse. Md Med J problems in a general hospital - a prevalence Pract 1999;48(5):378-84. 1994;43(1):35-9. study. J Indian Med Assoc 1997;95(9):505- 19. Smith AJ, Shepherd JP, Hodgson RJ. Brief 34. Conigrave KM, Hall WD, Saunders JB. The 6. interventions for patients with alcohol- AUDIT questionnaire: choosing a cut-off 5. Stein MD, Wilkenson J, Berglas N, O’Sulli- related trauma. Br J Oral Maxillofac Surg score. Alcohol Use Disorder Identification van P. Prevalence and detection of illicit 1998;36(6):408-15. Test. Addiction 1995;90(10):1349-56. drug disorders among hospitalized patients. 20. Marlatt GA, Baer JS, Kivlahan DR et al. 35. Conigrave KM, Saunders JB, Reznik RB. Am J Drug Alcohol Abuse 1996;22(3):463- Screening and brief intervention for high- Predictive capacity of the AUDIT 71. risk college student drinkers: results from a questionnaire for alcohol-related harm. 5. Sharkey J, Brennan D, Curran P. The pattern 2-year follow-up assessment. J Consult Clin Addiction 1995;90(11):1479-85. of alcohol consumption of a general hospital Psychol 1998;66(4):604-15. 36. Bradley KA, Boyd-Wickizer J, Powell SH, population in north Belfast. Alcohol 21. Senft RA, Polen MR, Freeborn DK, Hollis Burman ML. Alcohol screening Alcohol 1996;31(3):279-85. JF. Brief intervention in a primary care questionnaires in women : a critical review. 6. Mehler PS, McClellan MD, Lezotte D, setting for hazardous drinkers. Am J Prev JAMA 1998;280(2):166-71. Casper E, Gabow PA. Improving identifi- Med 1997;13(6):464-70. 37. London L, Meyers J, Nell V, Taylor T, cation of and intervention for alcoholism. 22. Fleming MF, Barry KL, Manwell LB, Thompson ML, Mbuli S. An investigation West J Med 1995;163(4):335-40. Johnson K, London R. Brief physician into the neurological and neurobehavioural 7. Schmidt A, Barry KL, Fleming MF. advice for problem alcohol drinkers. A effects of long-term agrochemical exposure Detection of problem drinkers: the Alcohol randomised controlled trial in community- among deciduous fruit farm workers in the Use Disorders Identification Test (AUDIT). based primary care practices. JAMA Western Cape, South Africa. 1995. South Med J 1995;88(1):52-9. 1997;277(13):1039-45. : MD Thesis. 8. Isaacson JH, Butler R, Zacharek M, Tzelepis 23. Wilk AI, Jensen NM, Havighurst TC. Meta- 38. Claassen JN. The benefits of the CAGE as A. Screening with the Alcohol Use Disorders analysis of randomised control trials a screening tool for alcoholism in a closed Identification Test (AUDIT) in an inner-city addressing brief interventions in heavy rural South African community. S Afr Med population. J Gen Intern Med 1994; 9 (10): alcohol drinkers. J Gen Intern Med 1997; J 1999;89(9):976-9. 550-3. 12(5):274-83. 39. Schoeman JH, Parry CDH, Lombard CJ, 9. McMenamin JP. Screening for alcohol 24. WHO Brief Intervention Study Group. A Klopper HJ. Assessment of alcohol disorder in a general practice. N Z Med J cross-national trial of brief interventions screening instruments in tuberculosis 1994; 107(972):55-7. with heavy drinkers. Am J Public Health patients. Tuberculosis and Lung Disease 10. Moore RD. Screening and assessment of 1996;86(7):948-55. 1994;75:371-6. alcohol and other drug abuse. Md Med J 25. Reid MC, Fiellin DA, O’Connor PG. Hazar- 40. Kew G. A descriptive study of alcohol 1994; 43(1):35-9. dous and harmful alcohol consumption in consumption patterns on a South African 11. Rydon P, Redman S, Sanson-Fisher RW, primary care. Arch Intern Med gold mine. MRC Urbanisation and Health Reid AL. Detection of alcohol-related 1999;159(15):1681-9. Newsletter 1994;21:39-42.

SA Fam Pract 2003;45(2) 15