Evaluation of a screening test for alcohol-related problems (CAGE) among employees of the Campus of the University of São Paulo

Abstracts

O uso do álcool pode ser responsável por acidentes, atrasos e faltas no trabalho. Sua detecção é limitada pelas dificuldades de pacientes e médicos quanto ao assunto. O questionário CAGE pode ser uma alternativa fácil, rápida e pouco intimidativa na detecção dos problemas relacionados ao uso de álcool (PRA). OBJETIVOS: Avaliar os indicadores de validade do CAGE - sensibilidade (S), especificidade (E), valor preditivo positivo, VPP e a área sob a curva ROC, ASC - entre funcionários da Prefeitura da Cidade Universitária, utilizando a Entrevista Clínica Estruturada para o DSM-IV, a SCID 2.0. MÉTODOS: Foram selecionados aleatoriamente 203 funcionários para entrevista com um questionário sociodemográfico seguido do CAGE e da SCID 2.0 Os indicadores de validade do CAGE foram analisados através dos resultados da SCID 2.0 para abuso e dependência do álcool e os dados sociodemográficos pelo cálculo do qui-quadrado. RESULTADOS: Entre os 192 funcionários entrevistados, a prevalência do CAGE positivo foi 19,8%, com os seguintes indicadores de validade para a detecção de PRA: S=84,4%, E=93,1%, VPP=71,1% e ASC=0,88 (p<0,01) e para dependência do álcool: S=91,3%, E=89,9%, VPP=55,3% e ASC=0,90 (p<0,01). Houve associações significativas entre o CAGE positivo e as seguintes características: 1) faixa etária de 61 anos ou mais (p=0,04); 2) sexo masculino (p=0,01); e 3) função operacional (p=0,02). CONCLUSÕES: O CAGE teve menos falsos positivos na detecção de PRA (abuso e dependência) que na detecção de dependência do álcool isoladamente. Os resultados para faixa etária e sexo podem refletir limitações do CAGE. Funções operacionais podem indicar um risco para PRA. O CAGE foi considerado válido na detecção de PRA no local de trabalho.

Transtornos relacionados ao uso de álcool; Escalas; Alcoolismo; Local de trabalho

ORIGINAL ARTICLE

Evaluation of a screening test for alcohol-related problems (CAGE) among employees of the Campus of the University of São Paulo

Ricardo Abrantes do Amaral^I; André Malbergier^II

^IInterdisciplinary Study Group on Alcohol and Drugs (GREA) of the Institute of Psychiatry of the Medical School of the University of São Paulo IPq-HCFMUSP

^IIDepartment of Psychiatry of the Medical School of the University of São Paulo

^{Correspondence}

ABSTRACT

Alcohol intake may play a significant role in absenteeism, delays and accidents at the workplace. However, its detection is limited by difficulties of both patients and physicians regarding the subject. The CAGE questionnaire may be an easy, fast and non intimidative alternative to detect alcohol-related problems (ARP).

OBJECTIVES: To evaluate the validity coefficients of the CAGE (sensitivity, Sen; specificity, Spec; positive predictive value, PPV, and the area under the ROC curve - AUC) among employees of the Campus of the University of São Paulo using the Structured Clinical Interview for DSM-IV, SCID 2.0.

METHODS: A random 203-worker sample was selected to be surveyed with a socio-demographic questionnaire followed by the CAGE questions and the SCID 2.0. CAGE validity coefficients were analyzed according to the SCID 2.0 results for alcohol abuse and dependence, while sociodemographic data were analyzed by the chi-square test.

RESULTS: Among 192 interviewed workers, the prevalence of the CAGE-positive was 19,8%. Validity coefficients results were for ARP: Sen=84,4%, Spec=93,1%, PPV=71,1% and AUC=0,88 (p<0,01); and for alcohol dependence: Sen=91,3%, Spec=89,9%, PPV=55,3% and AUC=0,90 (p<0,01). There were significant associations between CAGE-positive and the following characteristics: 1) age (61 years or more, p=0,04), 2) male gender (p=0,01) and 3) blue color occupations (p=0,02).

CONCLUSION: The PPV of the CAGE suggested its better performance in screening ARP (abuse and dependence) than only alcohol dependence. The association with age and sex could be related to limitations of the questionnaire. Blue color occupations could indicate a risk factor for ARP. The CAGE was considered valid for the screening of ARP at the workplace.

Keywords: Alcohol-related disorders;. Scales; Alcoholism; Working environment

Introduction

According to data from the World Health Organization,¹ problems related to alcohol use are responsible for 54% of work accidents with absence and for 40% of accidents followed by death. The International Labor Organization² adds that workers with alcohol-related problems use to be absent at the job 5- to 7-fold (26 days per year in average) than workers without these problems. In Brazil,¹ alcoholism is the third cause of absenteeism and the eighth for receiving the sickness absence in the welfare system; patients with alcohol-related problems use three-fold health services and are subject to delays 13- to 14-fold at work. Harwood³ has estimated in U$ 185 billion the costs related to alcohol abuse in the US in 1998, and more than 70% of these costs were attributed to a decrease in the productivity. Morawski et al⁴ and Blum et al⁵ observed that moderate alcohol use has a higher impact than heavy use in labor-related problems, such as absences and delays. Ames et al⁶ have studied the repercussion on the work not only of heavy drinking, but also of the consequences after the consumption, demonstrating that workers who came to work and reported having a 'hang-over', complained significantly more about criticism of supervisors, personal clashes with colleagues, about somnolence and feeling ill than workers who had not reported a 'hangover'.

The diagnosis of alcohol-related problems is hampered by several aspects: the defensive attitude of patients, denying the problem, and that of physicians who do not provide credibility to the possibility of treatment, using pre-conceived moral concepts, which frustrate the possibility of transforming the clinical interview in a process of sensitivization and motivation.⁷ The scarce available time of health professionals and the lack of training and specialized personnel in this kind of approach has been the target of studies and has inspired trainings.^8-9

The diagnosis of Alcohol Dependence Syndrome (ADS) through a fast and easily-applied method, with proven sensitivity and specificity, may be performed using the CAGE questionnaire, as proposed by Ewing & Rouse.¹⁰ Translated and validated into Portuguese by Masur & Monteiro,¹¹ with 88% of sensitivity (percentage of alcoholists correctly identified) and 83% of specificity (percentage of non-alcoholics correctly identified), the CAGE was successfully applied as a detecting instrument for alcoholism in hospitals,¹² in the general population^13-15 and in different working environments.^16-18

Some studies performed in the last two decades^13,19-20 have called the researchers' attention for the sensitivity of the CAGE regarding the detection of alcohol-related problems and not only for ADS, even when considering the traditional questionnaire's cut-off punctuation (two positive answers).

This study is justified for its being the first to use the diagnostic criteria of the DSM-IV, according to the Structured Clinical Interview for the DSM-IV (SCID-2.0),²¹ to assess the results of the CAGE regarding alcohol abuse and dependence, called in this study alcohol-related problems (ARP).

Objectives

The objective of this study is to assess the CAGE questionnaire regarding its sensitivity, specificity, positive predictive value and area under the ROC (Receiver Operating Characteristics) curve for ARP among civil servants of the Prefecture of the Campus of the University City (PCO) using the SCID-2.0.

Material and methods

The PCO is responsible for the administration and maintenance of the campus of the University City of the University of São Paulo (USP) and has 515 employees in the administrative and operational sectors. From 1995 onwards, the Program for the Prevention and Treatment of Drug Use at USP (PRODUSP) of the Interdisciplinary Study Group on Alcohol and Drugs (GREA) of the Department and Institute of Psychiatry of the Clinical Hospital of the Medical School of the University of São Paulo has started several preventive activities at USP, sponsored by the Rectorate and the University Hospital. This study is included among the activities being developed at PCO.

Considering a study with a sectional design over a population of 515 employees at PCO, the sample should have 109 employees, according to the software Epi Info Version 6.04b, taking into account the expected 10% frequency of alcohol dependence and accepting an alpha error of 5%, with a 10% confidence interval. However, the small number of possible employees identified using this sample could compromise the statistical analysis of the socio-demographic data. Therefore, considering this limitation and that this study aims to assess conditions of abuse and dependence, we have chosen a sample with approximately half of the studied population. In order to obtain a random sample, inclusion criteria were: pertaining to the personnel of the PCO in the period of the interviews and having an even job employee id number (n=243). Thirty-four employees who were transferred to other units were excluded from the sample, as well as 11 who were licensed from work at the moment of the interviews, 3 who had recently died and 3 who refused to participate after the reading of the informed consent. Refusal cases occurred at the reading of the informed consent, as employees were in doubt about the use of the information. In these cases, license and death, the occupational records were consulted for possible ARP. Therefore, the sample initially composed by 243 employees was reduced to 192 employees of the PCO. Chi-square analysis has not revealed significant differences between the total population of employees and the sample regarding the main socio-demographic data (age, marital status, gender and function).

Selected employees received and signed the informed consent in order to participate, beginning then to answer the sociodemographic questionnaire, followed by the CAGE, which was created by Ewing & Rouse¹⁰ as a detecting instrument for alcoholism, being composed by four questions with yes-or-no answers:

C – Have you, sometime, feel that you should reduce the amount of drinking or stop it?

A – Do people upset you because they criticize your way of drinking?

G – Do you feel guilty for the way you use to drink?

E – Do you use to drink in the morning in order to reduce nervousness or hang-over?

According to their authors, a positive answer would justify a more detailed further assessment. Since the creation of the CAGE, some studies^11,22-23 have analyzed the different figures of positive answers (from one to four) to the questionnaire and have defined its cut-off point, according to sensitivity and specificity criteria, at two positive answers. Steinweg and Worth²⁴ have suggested that the CAGE be preceded by an open question ('Let's talk about your drinking habit?'), what would make the questionnaire less intimidative, and would increase its sensitivity. In the current study we have chosen to apply the CAGE to detect ARP (abuse and/or dependence) according to the cut-off point of two or more positive answers and with the open question as an introduction to the questions. When the interviewee reported having stopped drinking, the questions were referred to one year before that.

After applying the socio-demographic questionnaire and the CAGE, the employees were interviewed according to the Structured Clinical Interview for the DSM-IV (SCID-2.0) which starts with a general screening, as a script for a non-structured clinical interview, and, afterwards, is divided into modules which correspond to the major diagnostic categories. In the interviews criteria for Alcohol Abuse and Dependence of the SCID-2.0 were used to assess the CAGE results for the different ARP. When the employee had positively punctuated for alcohol abuse and dependence according to the SCID-2.0, only the second criterion was considered. A psychiatrist at a PCO consultation office has conducted the interviews for a two-month period.

The statistical analysis was initially based on the CAGE assessment as a detecting test for ARP, calculating its sensitivity (S=Number of true positive/total of cases, multiplied by 100), specificity (E=Number of true negatives/total of healthy subjects, multiplied by 100), positive predictive value (PPV=Number of true positives/total of positives in the test, multiplied by 100) and area under the ROC curve (AUC), which is the measure of the instrument's capability to estimate the diagnosis, according to the results for alcohol abuse and dependence defined by the SCID-2.0. AUC was deemed valid for values higher than 0.5 and significant for p values equal to or lower than 0.5.

Next, socio-demographic data of PCO employees were analyzed according to their statistical association regarding ARP, alcohol abuse and dependence and the CAGE results using the chi-square test and Fisher exact test when the expected frequency was lower than 0.06, considering p values equal to or lower than 0.05 as statistically significant.

It was necessary to redivide some categorical variables due to the insufficient data for the statistical analysis in some of them. Therefore, marital status (single, married/living together, separated and widower) was recategorized as married/living together and without partner); the wage ranges 1 minimum wage (MW), 2 to 5 MW, 6 to 9 MW, 10 to 15 MW, 16 to 19 MW, 20 to 30 MW and more than 31 MW became 3 (1 to 5 MW, 6 to 9 MW and 10 MW or more); the time at work, initially divided in 1 to 5 years, 6 to 11, 12 to 17, 18 to 23 and more than 24 years, was modified into the following ranges: from 1 to 11 years, from 12 to 17 years and 18 years or more. Satisfaction at work (yes, no, more or less) was transformed into yes and some dissatisfaction.

The questionnaires were always applied sequentially, preventing thus sample losses.

Results

One hundred and ninety-two PCO employees were interviewed and answered the socio-demographic questionnaire and to questions of the CAGE and SCID-2.0. Eleven employees have not come to the interview and eight of them were not included in the study for being in sick leave, due to orthopedic (n=5), cardio-circulatory (n=2) and surgical (n=1) problems. Among the three employees who refused to participate after the reading of the informed consent, only one had previous history of ARP in his/her medical record. Among recent death cases, there were no references to ARP.

As shown in Table 1, the prevalence of ARP, according to the SCID 2.0, at the PCO in the year was 16.7%, that of abuse was 4.7% and of dependence, 12%. CAGE positive prevalence was 19.8%.

Table 1

Considering the age range, it was observed a higher frequency of ARP among employees aged 20 to 30 years, although there were high frequencies among age ranges from 41 to 50 years and from 61 years or more. Alcohol abuse was significantly associated with the youngest age range, decreasing relatively and absolutely with the increase in the age range, reaching 0 in the last age range. Among alcohol-dependent subjects, the highest concentrations were those of age ranges from 41 to 50 years and 61 years or more. No employee aged 20 to 30 years was identified as dependent. Positive CAGE was significantly associated with the age range of 61 years or more, representing almost half of the employees in this age range.

Considering gender, there was an association between males and ARP, what did not occur for the categories of alcohol abuse and dependence. Neither the CAGE nor the clinical interview identified any female employee. Positive results for the CAGE were also significantly associated with males.

The analysis of the marital status showed similar frequencies of ARP among married/living together subjects and those without partners; the same occurred regarding results for abuse and dependence, as well as for the CAGE.

Regarding the variable type of work there was higher concentration of employees with ARP in the operational sector. The abuse condition was similar between the different types of work, whereas, regarding dependence, there was a three-fold number of alcohol-dependent subjects, although this difference was not significant. When the CAGE results were analyzed, it was observed that almost all positives were from the operational sector, showing a statistically significant association between this function and positive CAGE.

According to the results shown in Table 2, the differences between wage ranges were not statistically significant. The lowest wage range had the highest frequency of ARP, dependence and CAGE-positive, although the intermediate wage range (6 to 9 MW) had results slightly lower. For the abuse condition, however, in absolute and relative figures, there was discrete predominance of the wage range between 6 and 9 MW. Neither the SCID 2.0 nor the CAGE identified any employee in the highest wage range.

Regarding time at work, differences were not significant, although alcohol dependence and CAGE- positive were nearly three times more frequent from 12 years at work onwards than in the range between 1 and 11 years of work at the PCO. Regarding alcohol abuse, among the range with the fewest working time, there was almost 2-fold frequency of positive results than among ranges with higher working time.

Considering satisfaction at work, there was no significant difference between groups who showed satisfaction and some dissatisfaction, with higher frequencies among all studied conditions among the former.

CAGE validity indicators are displayed in Table 3.

Thirty-eight employees (19.8%) answered positively to two or more questions of the CAGE. Considering the results of the SCID-2.0 for alcohol abuse and dependence as a whole (ARP), of each 10 identified subjects, more than 8 had positive results for the CAGE. Having a higher sensitivity for isolated dependence, the CAGE has identified more than 90% of the dependent subjects identified by the SCID 2.0. The results for CAGE specificity were higher for ARP (alcohol abuse and dependence) than for dependence alone.

Of each 10 subjects identified by the CAGE, 7 had probability of having any ARP and 5 may have alcohol dependence, what may be seen by the PPV results. Despite this difference, both results, alcohol abuse and dependence as a whole, (Figure 1) and alcohol dependence alone (Figure 2), when compared to the CAGE results according to the ROC curve, show areas significantly higher than 0.5.

The first CAGE question was positively and isolatedly answered by 60 employees. Of these, 5 were considered as false negatives according to the SCID-2.0 results (3 cases of abuse and 2 of dependence) and 21 received the diagnosis of alcohol abuse (n=4) or dependence (n=17) in lifetime. It is noteworthy that, of these 21 mentioned employees, 10 claimed having been abstinent for 10 years or more and 11, for more than 2 years. These employees were included as CAGE-negative in the study.

Discussion

The CAGE showed to be a hardly intimidating, easily- and quickly-applicable instrument (nearly 1 minute to be applied). Its validity indicators for ARP and alcohol dependence were similar, except for PPV, which was higher in the first case. The interpretation of a CAGE-positive as suggestive of any ARP would be the most indicated and could result in a less intimidating approach to the patient as the diagnostic of alcohol dependence is many times received as a recrimination. The diagnosis of alcohol abuse should not be interpreted as an indication of a future condition of dependence, and cannot be ruled out, as it is in itself a cause for health, social, family and occupational problems. Patients who were assessed regarding ARP and were oriented about the associated risks were receptive to the clinical orientations.^19-20 It should be emphasized that the questionnaire does not replace a detailed clinical interview.

The results for sensitivity and specificity found by Masur & Monteiro¹¹ are similar to those found in this study considering the criterion of alcohol dependence of the SCID-2.0. According to this interpretation, cases identified as false positives would be 17, although in 6 of them (15.7% of CAGE-positive ones) the SCID-2.0 has identified alcohol abuse patterns. Considering the CAGE only as a screening instrument for alcohol dependence, these 6 cases would be considered as negative ones although having risk patterns for alcohol consumption.

Utilizing the CAGE, Lima et al16 found 8.8% prevalence of 'problematic drinkers' among the workers of an oil refinery. Andrade et al¹⁷ also using the CAGE, found 8.51% of 'alcoholics' (the authors' expression) in a weaving industry. Furuno,¹⁸ studying workers from a car reseller with the CAGE, found 16.8% prevalence of possible cases of alcoholism, a result which is very close to that found among PCO employees for alcohol dependence (16.7%). Hemansson et al,²⁵ using the Alcohol Use Disorders Identification Test (AUDIT) and the dose of carbohydrate-deficient transferrine (CDT), found a 19.5% prevalence for alcohol-related problems among different working places. According to Webb et al,²⁶ using the Mortimer-Filkins test, there was a 5.7% prevalence of alcohol-related problems among blue-collar workers, whereas in other 15.2% of workers the results indicated the possibility of risk consumption patterns. The results reflect specific characteristics of the different work places studied and of the different types of occupations.

Prevalence rate of ARP and alcohol dependence at the PCO are related to the male gender. This is important when comparing these results with those of Carlini et al,²⁷ which found 11.2% prevalence of alcohol dependence (17.1% on males and 5.7% on females) among the interviewed.

The results for the variable age range (Table 1) show that 42.9% of employees aged above 61 years or more punctuated positively according to the CAGE, but, of them, only 3 had positive results for the SCID-2.0, that is, 3 identified cases would be false positives. In two of these cases, there was a positive response to questions one (decreasing or stopping) and two (being criticized) of the CAGE and, in the other case, positive answers in the first and third ones (feeling guilty). These responses, considering the negative results for abuse and/or dependence according to the SCID-2.0, may suggest a criticism to the consumption patterns, both by the subject and by others, due to the age and possibly chronic health problems. It would not be expected to find a statistically significant difference for this age range considering that, within the current population patterns, the increase in the ratio of people aged above 60 years in the general population would have led to a decrease in the average per capita of alcohol consumption.²⁸

Considering the variable gender (Table 1), we observed that no female employee had positive results in both questionnaires. Previous studies, using the CAGE in different population groups, showed a remarkable prevalence of alcohol-related problems among males. On the other hand, Monteiro et al,¹⁴ reapplying a methodology similar that of Capriglione et al,¹³ found a decrease from 3.6:1 in the first study to 2.1:1 in the second one in the male/female ratio for CAGE-positive cases, what may suggest a convergence of male-female behavior patterns regarding alcohol abuse. Moreover, according to the same study, the sensitivity of the CAGE in the detection of alcohol-related problems among females is lower than among males (46% and 75%, respectively). Female behavior of drinking less and having lower probability of drinking excessively as compared to males is a well-known behavior, even though the data should be described regarding the variable type of work to interpret adequately the result found. The PCO has 41 female employees, 24 of them in white-collar functions. Of the 20 female employees interviewed, 13 pertain to the administrative and 7 to the operational sector. The PCO is a predominantly male working environment, with men assigned to workshops or to external maintenance services. Among interviewed women, except for those working in the security sector (4 employees) all worked daily in the administrative sector, in administrative or maintenance functions (3 as general helpers) therefore in an environment which scarcely tolerates or hardly stimulates the consumption habits.

Data related to marital status at the PCO showed that the groups with or without partners had similar results for alcohol-related problems. Rego et al¹⁵ have also not observed significant differences between marital status and positive results in the CAGE. Other factors, besides marital status, such as relationship with children, spouse and with other relatives should be considered concerning the relation of the subject with alcohol.²⁹

Mandell et al,³⁰ analyzing alcoholism among occupations, have observed a higher risk among workers of the civil work, industrial transportation, and maintenance sectors, as well as carpenters, gardeners, rural workers, cleaning personnel, and mechanics. Rodríguez-Martos et al,³¹ have also found abusive drinkers in risk sectors such as civil work, transportation and mass media industries. Vasse et al³² identified a higher alcohol consumption mean among operational workers than among operational workers and that 12.5% of the studied sample drank excessively.

The results found in this study show a significant relation between CAGE-positive and the operational occupation (Table 1), what may represent a risk for work and reflect the pressure of colleagues and problems in the selection of personnel on the alcohol consumption habits among workers.³⁰ However, this association was not observed in the results of the SCID 2.0, what may suggest subclinical conditions or else a risk perception related to alcohol consumption among these employees.

The progressive clinical evolution of the ADS is manifested when we consider the age ranges with the highest prevalence (Table 1). The presence of alcohol abuse prevailed significantly among the youngest age ranges and the diagnosis of alcohol dependence was more frequent from 41 years of age onwards. It would be expected a decrease in the prevalence of ARP among the highest age ranges as a result of the reduction in the consumption or the abstinence of alcohol-dependent subjects, besides a possible selection bias, as the clinical repercussions become more important with the evolution of these problems,^29,33 hampering workers with chronic health problems to remain active.

As mentioned above, the PCO is mainly composed by operational workers, males, working predominantly in jobs with low complexity and directly exposed to weather conditions, factors which directly influence the consumption of alcohol. When considering the marital status and wage ranges (Table 2), we observe the existence of a standardization of the socioeconomic profile of employees, what can be reflected in the formation of both professional and personal informal links between the workers.

The high level of satisfaction at work among the employees (Table 2) allows interpretations such as real satisfaction or else as a certain intimidation to criticize their working conditions. The current unemployment level may have somehow influenced the data interpretation, what would justify the idea that, despite their unsatisfactory working conditions, the fact of being employed would be the most relevant factor.

As we have seen, the application of the CAGE in recruiting or periodical exams or in those for special activities (drivers, machine operators, etc.) should have to consider the existence of false positives, as well as the presence of subclinical conditions, and, else, of alcohol abuse conditions. Other important aspect is the fact that there are no studies related to the test-retest validity of the questionnaire, as its systematic application among subjects of a same population could lead to its recognition, leading to a lower reliability of the instrument.

One limitation regarding the methodology applied in this study is due to the application of the CAGE and the SCID 2.0 by the same interviewer, as a positive result in the questionnaire could lead the interviewer to seek traces confirming this result. This kind of bias can be reduced by means of structured clinical interviews, like the one used in this study. According to the questionnaire's authors, the application of the CAGE should not be conclusive, determining that a more detailed examination of the positive cases be performed for the diagnosis. Therefore, interviewers, be they clinicians or labor physicians, should reassess as suspicious the previously detected cases.

Conclusions

The CAGE has proven a scarcely intimidating, economical, quickly and easily applied instrument. The interviewed had no difficulty to understand the questions and its application has not required technical explanations, what may allow its inclusion among self-reported questionnaires or among those filled in by trained personnel but not necessarily of the medical area.

Its utilization to detect alcohol-related problems resulted in indicators of sensitivity and specificity within the literature patterns and in an increase in its positive predictive value, enabling the detection of alcohol abuse conditions among employees at the PCO.

Besides the abuse condition, the application of the CAGE at the workplace should consider the presence of other alcohol consumption conditions, deemed subclinical in this study.

The inclusion of the CAGE in the routine of occupational exams may be useful, considering its limitations as a screening instrument.

The results show, furthermore, that the studied workplace shows a prevalence of alcohol-related problems which deserve attention, but that are within the patterns of other services. This information is relevant for the self-esteem at the institution's workplace in order to develop a preventive work.

References

Sponsoring: The State of São Paulo Research Foundation (Fundação de Amparo à Pesquisa do Estado de São Paulo) – FAPESP. Grant 01/07463-9.

Received in 01.27.2003

Accepted in 11.05.2003

Original version accepted in Portuguese

Institute and Department of Psychiatry of the Clinical Hospital of the Medical School of the University of São Paulo – IPq-HCFMUSP

• 1. Odo SA, Araújo AC, Santos AF, Toledo FCP, Yonamine M, Silva OA, Leite MC. Indicações e limites das análises toxicológicas para substâncias psicoativas. Rev Psiquiatr Clin. 2000;27(1):50-6.
• ²
  Organização Internacional do Trabalho - International Labour Office – Guiding principles adopted by the ILO Inter-regional Tripartite Experts Meeting on Drug and Alcohol Testing in the Workplace. Geneve; 1993.
• 3. Harwood H. Updating Estimates of the economic costs of alcohol abuse in the United States: Estimates, Update Methods and Data. Report prepared by the Lewin Group for the National Institute on Alcohol Abuse and Alcoholism; 2000.
• 4. Morawski J, Moskalewicz J, Wald I. Economic costs of alcohol abuse, with special emphasis on productivity. In: Aasland OG, editor. The Negative Social Consequences of Alcohol Use, Oslo, Norway: Norwegian Ministry of Health and Social Affairs; 1991. p. 95-128.
• 5. Blum TC, Roman PM, Martin JK. Alcohol consumption and work performance. J Stud Alcohol. 1993;54(1):61-70.
• 6. Ames GM, Grube JW, Moore RS. The relationship of drinking and hangovers to workplace problems: an empirical study. J Stud Alcohol. 1997;58(1):37-47.
• 7. Rueff B. Lês maladies de l'alcool. Paris: John Libbey Eurotext; 1990. p. 203.
• 8. Rowland N, Maynard A, Beveridge A, Kennedy P, Wintersgill W, Stone W. Doctors have no time for alcohol screening. Br Med J (Clin Res Ed). 1987;295(6590):95-6.
• 9. Johnson JG, Sptizer RL, Williams JB, Kroenke K, Linzer M, Brody D, et al. Psychiatric comorbidity, health status, and functional impairment associated with alcohol abuse and dependence in primary care patients: findings of the PRIME MD-1000 study. J Consult Clin Psychol. 1995;63(1):133-40.
• 10. Ewing JA, Rouse BA. Identifying the hidden alcoholic. In: Program and abstracts of the 29^th Internacional Congress on Alcohol and Drug Dependence. Sidney, Australia; 1970.
• 11. Masur J, Monteiro MG. Validation of the "Cage" alcoholism screening test in a brazilian psychiatric inpatient hospital setting. Braz J Med Biol Res. 1983;16(3):215-8.
• 12. Bush B, Shaw S, Cleary P, Delbanco TL, Aronson MD. Screening for alcohol abuse using the CAGE questionnaire. Am J Med. 1987;82(2):231-5.
• 13. Capriglione MJ, Monteiro MG, Masur J. Aplicação do questionário CAGE para detecção da síndrome de dependência do álcool em 700 adultos na cidade de São Paulo. Rev Assoc Bras Psiquiatr. 1985;7(25):50-3.
• 14. Monteiro MG, Monteiro MA, Santos BR. Detecção do alcoolismo na população geral através do questionário CAGE: o que mudou em cinco anos? Rev ABP-APAL. 1991;13(2):45-8.
• 15. Rego RA, Oliveira ZMA, Berardo FAN, Oliveira MB, Ramos LR. Epidemiologia do alcoolismo: prevalência de positividade do teste CAGE em inquérito domiciliar no município de São Paulo. Rev ABP-APAL. 1991;13(2):75-80.
• 16. Lima CTS, Carvalho FM, Quadros CA, Gonçalves HR, Silva Junior JAS, Peres MFT, Bonfim MS. Hipertensão arterial e alcoolismo em trabalhadores de uma refinaria de petróleo. Rev Panam Salud Pública. 1999;6(3):185-91.
• 17. Andrade CL, Katz PD, Paulino APL, Bernik W. Diagnóstico de alcoolismo em empresa. Bol Centro Estud Pesqui Psiquiatr. 1986;4(2):68-74.
• 18. Furuno DF. Prevalência de alcoolismo em trabalhadores de uma revendedora de automóveis [tese]. São Paulo: Universidade de São Paulo, Faculdade de Saúde Pública, 1999.
• 19. Masur J. Diagnóstico precoce do alcoolismo através de questionários padronizados e testes biológicos. J Bras Psiquiatr. 1986;35(6):329-32.
• 20. Rosa AA, Gonçalves SC, Stefani SD, Martins SO, Rosa DD, Hunsche A, et al. Percepção e registro de abuso de álcool e de doenças relacionadas num hospital geral universitário. Rev Assoc Med Bras(1992). 1998;44(4):335-9.
• 21. Tavares M. Entrevista Clínica estruturada para diagnóstico dos transtornos do eixo I do DSM-IV. Ed. do Paciente, Versão 2.0. Tradução brasileira do Instituto de Psicologia da Universidade de Brasília; 1996.
• 22. Mayfield D, McLeod G, Hall P. The CAGE questionnaire: validation of a new alcoholism screening instrument. Am J Psychiatry. 1974;131(10):1121-3.
• 23. Buchsbaum DG, Buchanan RG, Centor RM, Schnoll SH, Lawton MJ. Screening for alcohol abuse using CAGE scores and likelihood ratios. Ann Intern Med. 1993;115(1):774-7.
• 24. Steinweg DL, Worth H. Alcoholism: the keys to the CAGE. Am J Med. 1993;94(5):520-3.
• 25. Hemansson U, Helander A, Brandt L, Huss A, Rönnberg S. The Alcohol Use disorders identification test and carbohydrate-deficient tranferrin in alcohol-related sickness absence. Alcohol Clin Exp Res. 2002 ;26(1):28-35.
• 26. Webb GR, Redman S, Hennrikus D, Rostas JA, Sanson-Fisher RW. The prevalence and sociodemographic correlates of high-risk and problem drinking at an industrial worksite. Br J Addict. 1990 ;85(4):495-507.
• 27. Carlini EA, Galduróz JCF, Noto AR, Nappo SA. Primeiro levantamento domiciliar sobre o uso de drogas psicotrópicas no Brasil: estudo envolvendo as 107 maiores cidades do país: 2001. São Paulo: CEBRID Centro Brasileiro de Informações sobre Drogas Psicotrópicas; UNIFESP Universidade Federal de São Paulo, 2002.
• ²⁸
  National Institute on Alcohol Abuse and Alcoholism. Alcohol and Health-Seventh Special Report to the US Congress, from the secretary of US Health and Human Services, Editorial Experts. Alexandria, Virginia; 1990. p. 13-38.
• 29. Lima MS. Epidemiologia do alcoolismo. In: Ramos SP, Bertolote JM. Alcoholismo hoje. 3Ş ed. Porto Alegre: Artes Médicas, 1997. p. 45-64.
• 30. Mandell W, Eaton WW, Anthony JC, Garrison R. Alcoholism and occupations: a review and analysis of 104 occupations. Alcohol Clin Exp Res. 1992;16(4):734-46. Review.
• 31. Rodríguez-Martos A. Problemas de Alcohol en el ámbito laboral. Madri, Espanha. Fundación de Ayuda contra la Drogadicción FAD; 1998. p. 16.
• 32. Vasse RM, Nijhuis FJ, Kok G. Associations between work stress, alcohol consumption and sickness absence. Addiction. 1998;93(2):231-41.
• 33. Robins LN, Regier DA. Psychiatric disorders in America: the epidemiologic catchment area study. New York; Toronto: Free Press: Collier Macmillan Canada, c1991.

Publication Dates

• Publication in this collection
  23 Feb 2005
• Date of issue
  Sept 2004

History

• Accepted
  05 Nov 2003
• Received
  27 Jan 2003