Alcohol and Alcoholism Advance Access originally published online on May 17, 2007
Alcohol and Alcoholism 2007 42(4):308-316; doi:10.1093/alcalc/agm013
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Detection and management of alcohol use disorders in German primary care influenced by non-clinical factors
1 University Medical Center Freiburg, Department of Psychiatry and Psychotherapy, Hauptstr. 5, D-79104 Freiburg, Germany
2 University Medical Center Tübingen, Hospital for Psychiatry and Psychotherapy, Osianderstr. 22, D-72076 Tübingen, Germany
3 University Medical Center Tübingen, Department of General Medicine, Keplerstr. 15, D-72074 Tübingen, Germany
* Author to whom correspondence should be addressed at: Dr. Michael M. Berner, Department of Psychiatry and Psychotherapy, University Medical Center Freiburg, Hauptstr. 5, D-79104 Freiburg, Germany. Tel: +49 761/270-6974; Fax: +49 761/270-6989; E-mail: michael.berner{at}uniklinik-freiburg.de
Received 13 November 2006; in revised form 14 December 2006; in revised form 15 December 2006; accepted 26 February 2007
 |
ABSTRACT |
|---|
 TOP ABSTRACT IntroductionMethodsResultsDiscussionConflict of InterestReferences |
|---|
Aims: The primary objective was to assess the proportion of detected and correctly referred patients in German primary care. The secondary objective was to identify patient and practitioner characteristics that predict detection and correct referral. \AbsHI{Methods} In this clustered cross-sectional survey in German primary care, 3003 patients were consecutively invited to participate, and were asked to fill in a standardized health questionnaire. They were then screened for problematic alcohol consumption using the Alcohol Use Disorders Identification Test. The physicians recorded their assessment of the presence of any alcohol use disorder and documented the treatment course of all identified patients for 3 months. Results Correctly identified problem drinkers were 38.6% in a per-protocol analysis and 33.6% using a worst-case scenario. Referral behaviour of physicians was in conformity with current practice guidelines in 64.6% of the documented cases and 27.0% in a worst-case scenario. Several patient (e.g. sex, age) and practitioner characteristics (e.g. age), which influence the diagnosis and referral of patients, could be identified. Conclusions There is a clear need to increase the special diagnostic and therapeutic skills of general practitioners so that they may be able to indicate and perform secondary prevention. Further research should focus on the likely effects of the implementation of these diagnostic and management tools.
|
Introduction |
|---|
|
TOPABSTRACTIntroductionMethodsResultsDiscussionConflict of InterestReferences |
|---|
There is a general on going debate about the effectiveness of screening and brief intervention techniques in the management of at-risk alcohol consumption in primary health care. Several trials and reviews demonstrate the efficacy, effectiveness and cost-effectiveness of such strategies (Wallace et al., 1988
; Wilk et al., 1997; Fleming et al., 1997; Fleminget al., 2000; Moyer et al., 2002; Fleming et al., 2002; Whitlock et al., 2004), while other findings suggest problems with their long-term effectiveness, efficiency, acceptance and implementation (Kaner et al., 1999; Andreasson et al., 2000; Aalto et al., 2000; Aalto et al., 2001; Beich et al., 2002; Beich et al., 2003). Several studies explored physician-related barriers to implementation (Bradley et al., 1995; Kaner et al., 1999; Kaner et al., 2001a; Kaner et al., 2001b; Anderson et al., 2003; Anderson et al., 2004; Berner et al., 2006b). In these studies, the training, attitudes, role security, commitment, belief in effectiveness, and environmental factors were found to influence the administration and quality of treatment. Two recent studies provide evidence that practitioner, as well as clinical and non-clinical patient characteristics also affect the likelihood of receiving appropriate care for excessive drinking and mental disorders (Raine et al., 2000); (Kaner et al.,2001a). The primary objective of this study was to assess the proportion of detected and correctly referred patients in a German primary care guideline implementation trial. The secondary objective was to identify patient and practitioner characteristics that predict detection and correct referral in this population. |
Methods |
|---|
|
TOPABSTRACTIntroductionMethodsResultsDiscussionConflict of InterestReferences |
|---|
Setting
This study was performed in the context of a randomized controlled trial, which evaluated the implementation of a comprehensive quality management approach for alcohol-related disorders in German outpatient care (Mänz et al., 2003
). This quality management system was derived from a systematic review of existing guidelines (Berner et al., 2004), and consisted of screening, guideline-oriented management, and documentation of patients. In the randomized controlled trial, it was examined whether implementing this system by training of the GPs improves the quality of provided care. In January 2003, all 299 GPs in the Ortenau region, and all 894 GPs in the Tübingen and Stuttgart regions of Baden-Württemberg county in southwest Germany were invited to participate in the study. Altogether, 58 GPs (4.9%), 29 from each region, agreed to participate. Patients were recruited in two stages (before and after GP training) from the GPs' offices. The first stage (baseline) was conducted from January to April 2003, the second one (follow-up) from July to November 2004, following the implementation of the quality management system in the intervention group. The present study is based on data from the baseline survey before the implementation of the guideline in the intervention group.
Procedures
Within a consecutive 3- to 4-day period, the GPs sampled all patients who were over the age of 16 visiting them and who displayed sufficient German language skills, until they reached a minimum of 50 patients. In the waiting room, a member of the study team or a physician's assistant invited patients to participate in the study. Participating patients were asked to give informed consent. If patients were under 18, permission of parents was required. Before seeing the GP, participating patients were asked to fill out a four-page structured health questionnaire including questions on socio-demographic characteristics, context of the visit to the GP's office, and mental disorders. The questionnaire contained the Alcohol Use Disorders Identification Test (AUDIT), a screening instrument to detect at-risk and harmful alcohol consumption (Babor et al., 2001). Using the AUDIT scale, patients were assessed as being problem drinkers (a composite term to refer to either at-risk drinking, harmful consumption, or alcohol dependence) if they scored eight points or above (men) or five points or above (women) (Babor et al., 2001; Rumpf et al., 2002; Neumann et al., 2004; Dybek et al., 2006). The GPs were blinded to the results of the patient questionnaire. According to study protocol, immediately after discharging the patient, the GPs had to assess their patient with regard to alcohol use disorders according to their usual practice and criteria. They recorded their judgment in a standardized list, classifying the patients in one of four categories: no alcohol problem, at-risk drinking, harmful consumption or alcohol dependence.
After having seen the patient, the GPs were asked to include all patients assessed as having an alcohol problem in the longitudinal part of the study. The GPs were instructed to offer the patient their participation. This could be done at the end of the consultation or by contacting the patient afterwards, and was independent of any actions (treatment or referral) taken. After receiving informed consent, a second visit 3 months later was arranged to perform a follow-up assessment. At inclusion and at the follow-up, a document was filled out by the GP including diagnostic measurements, information on treatment and referral of patients, as well as assessing change in patient's status. Patients were counted as treated or referred correctly whenever they had been either treated successfully (with a significant reduction in alcohol consumption) or had been referred according to current practice guidelines (Berner et al., 2004). The detailed algorithm is presented in Table1.
|
Due to low rates of patient recruitment by the GPs, they were allowed to include additional patients who did not participate in the screening part of the study.
Statistical analysis
All analyses were of explorative nature, i.e. they aimed to generate rather than to confirm hypotheses. This approach was chosen because both theoretical concepts and empirical findings concerning the examined issues are missing. Furthermore, it must be kept in mind that the present study was conducted in the context of a randomized controlled trial, so it was not primarily designed to answer the questions that are investigated here.
A detailed patient flow-chart was created to examine detection and referral behaviour of GPs in German primary care. A descriptive analysis was performed by calculating detection and correct referral rates. Two scenarios were used for analysis. In an ‘as-treated’ scenario, only valid cases were examined. In an ‘intention-to-treat’ scenario, a worst-case assumption (no detection and incorrect referral) and a best-case assumption (detection and correct referral) were applied in case of missing data.
To explore the association between patient/practitioner characteristics and outcomes (detection and proper referral) both uni- and multivariable methods were used. To increase power, to make results comparable, and to ease interpretation, all independent variables were collapsed in two or three categories.
Univariable 
2-tests were performed separately for the following independent variables: (A) patient characteristics: (A1) sex, (A2) age, (A3) family status, (A4) first language, (A5) level of education, (A6) employment status, (A7) physical health status, (A8) severity of alcohol problem, (A9) frequency of visiting GP last year, and (A10) length of time being the present GP's patient; as well as (B) practitioner characteristics: (B1) sex, (B2) age, (B3) length of practice time, (B4) qualification in addiction medicine, (B5) population of local residents, (B6) practice type, and (B7) number of patients in a quarter year.
In multivariable analysis, (A1) patient sex, (A2) patient age, (B1) practitioner sex, (B2) practitioner age, their interactions (A1x B1 and A2x B2) and all variables reaching a significance level of P 
0.10 in univariable analyses predicting at least one outcome were included simultaneously in a logistic regression model. Interaction effects were displayed graphically, regardless of significance level, to allow visual examination of results.
Characteristics and outcomes of patients who visited the same GP are likely to be more similar than those of patients treated by different GPs. This clustered structure of the data (with the GPs being of higher and the patients of lower level) was accounted for by an adjustment for the clustering effect (Wears, 2002). For this, the variance inflation factor (VIF, also called design effect) was calcuated. 2-values in case of univariable tests, Wald-statistics in case of logistic regression, and confidence intervals for adjusted odds-ratio estimates were corrected using this factor (Wears, 2002; Reed, 2004). The intraclass correlation coefficients (ICCs) needed for this adjustment were estimated by the analysis of variance (ANOVA) method (Fleiss, 1986; Wears, 2002).
Results with a probability of a type I (alpha) error of 0.05 were considered as statistically significant. Findings with a corresponding error level of 0.10 were termed as ‘trend’ and were also reported. Due to the explorative character of the study, a correction of the alpha error inflation arising from multiple testing was not performed.
All statistical analyses were carried out using SPSS 13.0 and Microsoft Excel 2002.
Ethical approval
The study protocol of the project ‘Out-Patient Quality Management of Alcohol-Related Disorders in Primary Care’, in which context this survey was conducted, was approved by the Ethics Committee of the Freiburg University Medical Centre, Germany.
|
Results |
|---|
|
TOPABSTRACTIntroductionMethodsResultsDiscussionConflict of InterestReferences |
|---|
GP population
Fifty six GPs included (screened or documented) at least one patient. Approximately three-quarters of the participating GPs were men (76.8%; N = 43/56). Over half of them were above the age of 50 (59.2%; N = 29/49) as well as practicing for a minimum of 15 years (53.1%; N = 26/49). Approximately one-quarter carried an additional degree in basic addiction medicine (26.5%; N = 13/49). The majority of the GPs had their offices in rural areas, in towns with less than 25 000 citizens (71.2%; N = 37/52). Sixty percent worked in a single practice (N = 30/50), and 62.5% saw more than 1000 patients per quarter of a year (medium to large practice size) (N = 30/48). Two GPs had to be excluded because they did not include any patients during the study. One of them was a 34-year-old male GP in a newly established rural team practice. The other excluded GP was a 59-year-old female practitioner with 18 years of practice experience in a rural region.
Patient population
Three thousand and three patients were recorded by the participating GPs (with a median of 52, a minimum of 24, and a maximum of 106 patients per practice). Characteristics of the total sample and of the patients that were included in further analyses are presented in Table 2. In 2940 patients, the screening procedure was performed. However, 378 of them (12.9%) could not be included in the analysis because either the AUDIT was filled out incompletely (42.9%; N = 162/378), the GP's assessment was not available (54.0%; N = 204/378), or the sex of the patient was not reported (3.2%; N = 12/378). More than half of the patients were women (56.3%). Approximately half of them were middle-aged (53.4% between 30 and 59 years old) and more than two-thirds were married or cohabiting with a partner (69.1%). Three of four patients had at least a medium level of education (76.0%) and one out of two patients was employed (54.6%). In the last year, 40.9% of the patients visited their GP more than five times. Most patients visited the same GP for more than a year (87.6%), with a notable proportion of patients attending the same GP for at least 10 years (42.8%).
|
Detection and referral of patients
Patient flow and management of alcohol use disorders in this sample are displayed in Fig. 1. Of the 2562 patients that were included in the analysis of diagnostic processes, 332 (13.0%) scored above the threshold for problem drinking. Of them, 128 (38.6%) were assessed by the GP as being problem drinkers. When assuming that the prevalence of problem drinking in the excluded patients is the same as in the included ones, and that none of the excluded patients was detected (worst-case scenario), the detection rate falls to 33.6% (N = 128/381). The best-case assumption (all excluded patients detected) increases the detection rate to 46.5% (N = 177/381).
|
There were 344 patients who consumed alcohol at-risk, drank harmfully, or had an alcohol dependence as judged by the GP (13.4% of analysed sample). A complete documentation was performed in the case of 144 of these patients (41.9%). Referral behaviour of GPs was in conformity with current practice guidelines in 93 cases (64.6%). The worst-case assumption that incompletely documented patients were referred incorrectly yields a correct referral rate of 27.0% (N = 93/344). The best-case assumption of all incompletely documented patients referred correctly results in a correct referral rate of 85.2% (N = 293/344).
The combination of these findings results in an ‘as-treated’ estimate of the proportion of altogether correctly managed patients with alcohol use disorders of 24.9%, in a worst case ‘intention-to-treat’ estimate of 9.1%, and in a best case ‘intention-to-treat’ estimate of 39.6%.
Association of patient and practitioner characteristics with the management of alcohol-related disorders
The ICCs were 0.183 (VIF = 2.30) for detection and 0.174 (VIF = 2.39) for referral behaviour of GPs, showing that almost a fifth of variance in these outcomes results from differences between the individual practitioners.
The univariable analyses of predicting detection of problem drinkers by the GP from patient and practitioner characteristics are summarized in Tables 3 and 4. Male patients, a severe drinking problem, and frequent GP visits were significantly associated with an increased probability of detection (P 0.05). Some trend was found that middle-aged (between 30 and 59 years old) patients and patients with a rather poor physical health status are more likely to be detected. The GP's qualification in addiction medicine tends also to be associated with a probable detection of problem drinkers (P 0.10). In univariable analyses, correctness of referrals proved to be significantly interrelated to the number of local residents at the location of the GP's office, with an increased accuracy in urban practices (P 0.05; see Table3. A trend was found that female patients were more correctly referred than males (P 0.10).
|
|
In multivariable analyses, including patient and practitioner characteristics as well as their interactions, several factors proved to be significantly (P
0.05) associated with detection of problem drinkers by the GP (see Table 5. Being a female patient decreased the odds of detection by a factor of 3.6 (OR = 0.28). Visiting the GP more than five times within the last year increased the odds of being detected by a factor above 3 (OR = 3.15). Female GPs proved to detect problem drinkers clearly better than their male colleagues (OR = 7.83). The odds of detection were more than 4 times higher by GPs who were at least 50 years old than by their younger counterparts (OR = 4.32). Furthermore, a significant interaction was found between the patients' and the GPs' age (OR = 0.06 by the interaction patient 
60 years x practitioner 50 years). In multivariable analyses, neither a significant association nor a trend could be identified between the patient or practitioner characteristics and correctness of referrals.
|
Interaction effects between the patients' and GPs' sex/age in case of predicting detection and correct referral are displayed in Fig. 2. While female patients were generally poorly detected (less than 25%), male patients tended to be detected better by female GPs (72.1% vs 42.0% by male GPs). However, these results were not statistically significant. It can also be seen that older GPs detect young and middle-aged patients almost twice as well as their younger colleagues, and that this tendency is inverse in the case of older patients with a statistically significant interaction effect. In case of analysing referrals, the most proper GP behaviour could be identified in female patients and female GPs (90.9%), with the least guideline-conforming rate in male patients and female GPs (35.0%; results are statistically not significant). Minor interaction effects can be observed between the patients' and GPs' age, with younger GPs referring young patients more incorrectly than older GPs (33.3% vs 83.3%; results are statistically not significant).
|
|
Discussion |
|---|
|
TOPABSTRACTIntroductionMethodsResultsDiscussionConflict of InterestReferences |
|---|
In the present study, a likely rate of correct diagnosis and referral of patients with problem drinking could be established. Furthermore, several non-clinical factors that influence the management of alcohol use disorders in German primary care could be identified.
The most striking finding of the present study is the low proportion of patients with problem drinking who are detected by their GP. Not even half of them were identified. The worst case ‘intention-to-treat’ estimate suggests that only one of three patients with problematic alcohol consumption is identified by his/her general practitioner. In another German study by Rumpf et al. in a similar setting, 50.4% of the patients were detected (Rumpf et al., 2001). The difference between these findings could be explained by the fact that in the study by Rumpf et al., only patients with a diagnosis of alcohol abuse or dependence were considered, but in the present trial, at-risk drinkers also had to be identified. This assumption is supported by the finding that the problem severity is associated with an increasing probability of detection. Referral behaviour of the GPs was correct in two of three cases, when only properly documented patients were considered. This could mean that once the secondary prevention in the patient is initiated, a majority of patients are treated correctly. However, the worst-case ‘intention-to-treat’ estimate of one of four correctly referred patients is clearly lower. Best-case ‘intention-to-treat’ analyses suggest more optimistic, though by far not optimal, figures, with every second problem drinker detected and six of seven referred according to guidelines. Combining the diagnostic and management findings suggests that in a best-case scenario, around one in three patients with alcohol use disorders is managed correctly. The worst-case estimate presents an alarming scenario, in which only one of eleven patients is detected and referred properly.
Female patients are more likely to be overlooked. This is in concordance with the findings by Kaner and colleagues, who describe that male at-risk drinkers are more likely to receive brief intervention than females (Kaner et al., 2001a). Similarly, there is a trend that younger people and patients with good health status are overlooked, whereas in the male population, the ‘high utilizers’ of the general practice, and the severely affected are more subject to detection. This underlines the necessity for conducting early secondary prevention programmes to detect alcohol use disorders in early stages as well as in populations in which drinking problems are not expected by the GP. It is of special note that female GPs appear to be better in detecting problematic alcohol consumption in men than their male colleagues. Older GPs were more successful in identifying alcohol problems than their younger colleagues. However, the formal qualification did not improve diagnostic or referral performance.
Whereas the detection rates of the practitioners might reflect a true performance, the referral rates must be interpreted with caution. Patients with alcohol use disorders, especially in early stages of the disease, may not have enough insight into the problem to proceed to behavioural changes or accept treatment. German GPs report frequently about patient-related barriers of management of alcohol-related problems (Berner et al., 2006b). Similarly, basic system requirements, such as cooperation, detoxification, and counselling centres have to be available to allow adequate referrals. This is probably reflected in the finding that urban practices tend to refer their patients more properly than the rural ones.
However, the findings are limited by some methodological issues. First, a possible self-selection bias of the GPs that participated in the study cannot be ruled out. The central element of the randomized controlled trial, in which context the presented analysis was performed, is training the GPs in detection and referral of patients with alcohol use disorders. This could possibly have led to a participation of GPs with suboptimal skills concerning this topic. On the contrary, there is evidence that the need for continuing education in the field of alcohol use disorders increases with knowledge of the practitioners (Berner et al., 2006a). Thus a positive selection of GPs is also possible. In addition, the knowledge of participating in a study might have drawn the GPs’ focus on the subject of alcohol use disorders and therefore might have biased the findings. Second, as comparable population or practitioner data are not available, no certain judgment about the representativeness of the patient sample can be made, although the GPs were instructed to include all patients consecutively, and they reported that the proportion of patients declining participation was below 5%. Third, the use of the AUDIT as a reference standard for at-risk drinking might result in a lower estimate of detection rates than the use of formal diagnostic criteria (Rumpf et al., 2001). Nevertheless, as yet, there is no general consensus of how at-risk drinking should be ‘formally’ defined. Fourth, disregard of study instructions by the GPs in the form of not/incompletely documenting positively assessed patients in the longitudinal part of the study was unfortunately quite frequent. This might be explained by the fact that it is very uncommon for German general practitioners to participate in clinical or epidemiological trials. The practitioners may have found it difficult to comply with the documentation system. These facts were accounted for by calculating both ‘as-treated’ and ‘intention-to-treat’ estimates. Fifth, due to the large variation between the GPs' performances (as indicated by the intracluster correlation coefficient), the power of the performed analyses might be insufficient to point out small effects.
The most essential implication of the present findings is that there is a clear need to increase the special diagnostic and therapeutic skills of GPs enabling them to indicate and perform secondary prevention. This education should focus on the use of AUDIT (or comparable instruments) as a screening instrument in detecting the need (or indication) for early intervention (i.e. in young people without serious health consequences) and provide special attention for the detection of female patients. Furthermore, skills and/or instruments (such as screening questionnaires) could be introduced, which enable the physicians to detect at-risk drinking patients regularly: for instance, if they visit the practice for the first time. Further research should focus on the likely effects of the implementation of those diagnostic and management tools.
It seems that GPs tend to detect problem drinking more likely in patients who are different from themselves: female practitioners identify male problem drinkers and young practitioners, the older patients more often. This fact may be based on a type of cognitive dissonance bias (e.g. assuming that similar people behave similarly and therefore rejecting possible problem behaviour of a similar person) and deserves further research interest (Festinger, 1962).
Considering the adverse somatic, psychological, social, and economic consequences of increased alcohol consumption, it is necessary that the management of alcohol-related disorders in primary care rests on clinical facts rather than on non-clinical factors.
|
Acknowledgement |
|---|
Acknowledgements–This study was conducted as part of the project ‘Out-Patient Quality Management of Alcohol-Related Disorders in Primary Care’ (AQAH) of the Addiction Research Network Baden-Württemberg, sponsored by the German Federal Ministry for Education and Research (grant number 01EB0111).
We would like to thank all participating general practitioners and patients. We are grateful for the help of the assistants of the AQAH Study Group: Michael Bentele, Inga Gendo, Sandra Habbig, Constantin Mänz, Tanja Peters and Sandra Schäfer. We would also like to thank Lili Mundle for the revision of the manuscript.
|
Conflict of Interest |
|---|
|
TOPABSTRACTIntroductionMethodsResultsDiscussionConflict of InterestReferences |
|---|
We declare that there is no competing interest of the study authors.
|
References |
|---|
|
TOPABSTRACTIntroductionMethodsResultsDiscussionConflict of InterestReferences |
|---|
Aalto M., Saksanen R., Laine P., et al. Brief intervention for female heavy drinkers in routine general practice: a 3-year randomized, controlled study. Alcoholism: Clinical and Experimental Research (2000) 24:1680–1686.[CrossRef][ISI][Medline]
Aalto M., Seppa K., Mattila P., et al. Brief intervention for male heavy drinkers in routine general practice: a three-year randomized controlled study. Alcohol and Alcoholism (2001) 36:224–230.
Anderson P., Kaner E., Wutzke S. World Health Organization Brief Intervention Study Group. Attitudes and management of alcohol problems in general practice: descriptive analysis based on findings of a world health organization international collaborative survey. Alcohol and Alcoholism (2003) 38:597–601.
Anderson P., Kaner E., Wutzke S. WHO Brief Intervention Study Group. Attitudes and managing alcohol problems in general practice: an interaction analysis based on findings from a WHO collaborative study. Alcohol and Alcoholism (2004) 39:351–356.
Andreasson S., Hjalmarsson K., Rehnman C. Implementation and dissemination of methods for prevention of alcohol problems in primary health care: a feasibility study. Alcohol and Alcoholism (2000) 35:525–530.
Babor T. F., Higgins-Biddle J. C., Saunders J. B., et al. AUDIT. The Alcohol Use Disorders Identification Test. Guidelines for use in Primary Health Care (2001) 2nd edn. World Health Organisation, Geneva.
Beich A., Gannik D., Malterud K. Screening and brief intervention for excessive alcohol use: qualitative interview study of the experiences of general practitioners. BMJ (2002) 325:870–874.
Beich A., Thorsen T., Rollnick S. Screening in brief intervention trials targeting excessive drinkers in general practice: systematic review and meta-analysis. BMJ (2003) 327:536–542.
Berner M. M., Habbig S., Härter M. Quality of guidelines for the treatment of alcohol related disorders. A systematic review and content analysis. [German]. Fortschritte der Neurologie and Psychiatrie (2004) 72:696–704.[CrossRef]
Berner M. M., Härter M., Zeidler C., et al. German general practitioners' perceived role in the management of alcohol use disorders: responsible but undertrained. Primary Care and Community Psychiatry (2006a) 11:29–35.[CrossRef]
Berner M. M., Zeidler C., Kriston L., et al. Diagnostic and treatment approaches to alcohol-related disorders. [German]. Fortschritte der Neurologie and Psychiatrie (2006b) 74:157–164.
Bradley K. A., Curry S. J., Koepsell T. D., et al. Primary and secondary prevention of alcohol problems: U.S. internist attitudes and practices. Journal of General Internal Medicine (1995) 10:67–72.[ISI][Medline]
Dybek I., Bischof G., Grothues J., et al. The reliability and validity of the Alcohol Use Disorders Identification Test (AUDIT) in a German general practice population sample. Journal of Studies on Alcohol (2006) 67:473–481.[ISI][Medline]
Festinger L. Cognitive dissonance. Scientific American (1962) 207:93–102.[ISI][Medline]
Fleiss J. L. The Design and Analysis of Clinical Experiments (1986) New York: John Wiley and Sons. 1–32. Reliability of measurement.
Fleming M. F., Barry K. L., Manwell L. B., et al. Brief physician advice for problem alcohol drinkers. A randomized controlled trial in community-based primary care practices. JAMA (1997) 277:1039–1045.[Abstract]
Fleming M. F., Mundt M. P., French M. T., et al. Benefit-cost analysis of brief physician advice with problem drinkers in primary care settings. Medical Care (2000) 38:7–18.[CrossRef][ISI][Medline]
Fleming M. F., Mundt M. P., French M. T., et al. Brief physician advice for problem drinkers: long-term efficacy and benefit-cost analysis. Alcoholism: Clinical and Experimental Research (2002) 26:36–43.[CrossRef][ISI][Medline]
Kaner E. F., Heather N., Brodie J., et al. Patient and practitioner characteristics predict brief alcohol intervention in primary care. British Journal of General Practice (2001a) 51:822–827.[ISI][Medline]
Kaner E. F., Wutzke S., Saunders J. B., et al, WHO Brief Intervention Study Group. Impact of alcohol education and training on general practitioners' diagnostic and management skills: findings from a World Health Organization collaborative study. Journal of Studies on Alcohol (2001b) 62:621–627.[ISI][Medline]
Kaner E. F., Heather N., McAvoy B. R., et al. Intervention for excessive alcohol consumption in primary health care: attitudes and practices of English general practitioners. Alcohol and Alcoholism (1999) 34:559–566.
Mänz C., Berner M. M., Lorenz G., et al. Project AQAH: comprehensive quality management in outpatient care. [German]. Suchtmedizin (2003) 5:41–44.
Moyer A., Finney J. W., Swearingen C. E., et al. Brief interventions for alcohol problems: a meta-analytic review of controlled investigations in treatment-seeking and non-treatment-seeking populations. Addiction (2002) 97:279–292.[CrossRef][ISI][Medline]
Neumann T., Neuner B., Gentilello L. M., et al. Gender differences in the performance of a computerized version of the Alcohol Use Disorders Identification Test in subscritically injured patients who are admitted to the emergency department. Alcoholism: Clinical and Experimental Research (2004) 28:1693–1701.[CrossRef][ISI][Medline]
Raine R., Lewis L., Sensky T., et al. Patient determinants of mental health interventions in primary care. British Journal of General Practice (2000) 50:620–625.[ISI][Medline]
Reed J. F. III. Adjusted chi-square statistics: application to clustered binary data in primary care. Annals of Family Medicine (2004) 2:201–203.
Rumpf H. J., Bohlmann J., Hill A., et al. Physicians' low detection rates of alcohol dependence or abuse: a matter of methodological shortcomings? General Hospital Psychiatry (2001) 23:133–137.[CrossRef][ISI][Medline]
Rumpf H. J., Hapke U., Meyer C., et al. Screening for alcohol use disorders and at-risk drinking in the general population: psychometric performance of three questionnaires. Alcohol and Alcoholism (2002) 37:261–268.
Wallace P., Cutler S., Haines A. Randomised controlled trial of general practitioner intervention in patients with excessive alcohol consumption. BMJ (1988) 297:663–668.[ISI][Medline]
Wears R. L. Advanced statistics: statistical methods for analyzing cluster and cluster-randomized data. Academic Emergency Medicine (2002) 9:330–341.[CrossRef][ISI][Medline]
Whitlock E. P., Polen M. R., Green C. A., et al, U.S. Preventive Services Task Force. Behavioral counseling interventions in primary care to reduce risky/harmful alcohol use by adults: a summary of the evidence for the U.S. Preventive Services Task Force. Annals of Internal Medicine (2004) 140:557–568.
Wilk A. I., Jensen N. M., Havighurst T. C. Meta-analysis of randomized control trials addressing brief interventions in heavy alcohol drinkers. Journal of General Internal Medicine (1997) 12:274–288.[CrossRef][ISI][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||