Abstract-– Physicians are in a key position to diagnose and treat patients with alcohol-related problems. Early interventions before the onset of these problems may decrease the costly health care as well as the psychological and social burden of alcoholism on the patient as well as the society. At this stage, the need for physicians to screen alcohol users systematically with a simple, effective and accurate instrument is becoming more critical. Being an easy-to-administer, low-cost, sensitive and specific screening tool, CAGE Questionnaire meets these criteria and offers the promise of raising the identification rate of alcoholic patients substantially. However, CAGE has still been reported to miss nearly half of risk-drinkers because of the incorrect setting of the high likelihood criterion for the presence of alcoholism. Therefore, there is a need to determine a clinically significant cut-off point above which CAGE will be diagnostic. This article aims to identify these optimal work-points for three different clinical settings by employing a step-wise application of statistical indices such as the area under the ROC curve, leveling factor and Youden index. This method will enable health care providers to determine the optimal CAGE scores for different treatment settings and significantly decrease the number of unrecognised at-risk drinkers. Keywords – Alcoholism, CAGE Questionnaire, cut-off, leveling factor, Youden index I. INTRODUCTION Alcoholism is a chronic, progressive and potentially fatal disease characterized by continued use of alcohol resulting in emotional, social, physical, or legal problems. These problems take an enormous emotional toll on individuals as well as their families, and are a great financial expense to health care systems and society in purely economic terms (1). Early detection and identification of alcohol-related problems may alleviate ongoing medical and social problems due to drinking and reduce the future risks and costs from excessive alcohol use. This can only be possible by using a powerful screening test that coverts drinking problems and discriminates between at-risk and risk-negative alcohol users with high diagnostic accuracy. With proper screening for these conditions, physicians can identify individuals in a patient population who have begun to develop or who are at- risk for developing alcoholism. Once they are diagnosed, patients can be treated as outpatient or inpatient depending on the complication of their alcohol-related problems. Those patients with mild-to- moderate withdrawal symptoms for uncomplicated problems and psychological stability are usually treated as ‘outpatients’ and assigned to support groups, counseling, or both. On the otherhand, patients with a coexisting medical or psychiatric disorder and those who may harm themselves or others, who have not responded to conservative treatments, or who have a disruptive home environment receive an ‘inpatient’ treatment in a general or psychiatric hospital or in a center dedicated to treatment of alcohol abuse. Since inpatient treatment is expensive, it is generally reserved for severely alcohol- dependent patients. This paper undertakes a study on psychiatric and medical inpatients as well as elderly General Medicine outpatients. It investigates where the optimal cut-off point should be placed for these clinical settings so that physicians using the CAGE Questionnaire as their screening tool for alcoholism can achieve superior results in catching the at-risk drinkers. II. METHODOLOGY CAGE Questionnaire was first developed by Ewing and Rouse (2). It was initially validated by Mayfield in psychiatric inpatients (3). Then, Bush subsequently studied the CAGE using medical inpatients (4). Consequently, Buchsbaum applied the CAGE on elderly General Medicine outpatients (5). Up-to-date, CAGE has been administered to a variety of other patient groups such as college students and general populations for measuring the dimensions of their alcohol problems (6,7). CAGE is an acronym arising from key concepts contained in each of the following four questions of the CAGE Questionnaire: 1. Have you ever felt you should Cut down on your drinking? 2. Have people Annoyed you by criticizing your drinking? 3. Have you ever felt bad or Guilty about your drinking? 4. Have you ever had a drink first thing in the morning to steady your nerves or to get rid of a hangover (Eye-opener)? Each of the above question yields an answer in the form of binary responses as ‘yes’ or ‘no’. Each ‘yes’ answer accounts for 1 point whereas each ‘no’ answer is given 0 points. The resulting total point is called the CAGE score. The CAGE score enables the physician to stratify patients along a continuum of risk for alcoholism: the higher the CAGE score the greater the probability of alcoholism. Here, a dichotonomous model is used in the interpretation of CAGE scores such that all patients above an pre- determined optimal cut-off point will be assigned the same risk of positive alcoholism. This cut-off point will be defined in terms of a CAGE score such that it will alert the physician for further investigation to the high likelihood of the presence of alcoholism and portray a clinical significance. Thus, Analysis of Cut-off Points for the CAGE Questionnaire for Alcohol Abuse Mehmet Tolga Taner 1 , Albert Guvenis 1 1 Institute of Biomedical Engineering, Bogazici University, Istanbul, Turkey 0-7803-7211-5/01$10.00©2001 IEEE Proceedings – 23rd Annual Conference – IEEE/EMBS Oct.25-28, 2001, Istanbul, TURKEY