Related Experiment Video
Updated: Aug 11, 2026
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
Alcoholism screening questionnaires: are they valid in elderly medical outpatients?
T V Jones1, B A Lindsey, P Yount
1Department of Medicine, University of Nebraska Medical Center, Omaha 68198-5620.
Objective:
To assess the validity of the CAGE (cut down, annoyed, guilty feelings, eye-opener) questionnaire and the Michigan Alcoholism Screening Test (MAST) in distinguishing between elderly patients with and without alcohol abuse or dependence disorders.
Design:
A cross-sectional study, in which patients were interviewed with a "gold standard," the alcohol module of the Revised Diagnostic Interview Schedule (DIS-III-R), and two screening questionnaires: the CAGE and the MAST.
Setting:
The study was conducted in the outpatient medical practice of a university teaching hospital.
Patients:
All English-speaking continuity patients 65 years of age or older able to participate were eligible; complete data were available for 154 (91%) of the 170 people who agreed to participate.
Results:
Sixty-seven patients (44%) were active drinkers, whereas 87 (56%) reported abstinence. Twenty-five patients (16%) met Diagnostic and Statistical Manual of Mental Disorders-III-Revised (DSM-III-R) criteria for alcohol abuse or dependence. A CAGE score of 2, the conventional cutoff point, had a sensitivity and a specificity of 48% and 99%, respectively. A MAST score of 5, the originally recommended cutoff point, had a sensitivity and a specificity of 52% and 91%, respectively. The areas under the receiver operating characteristic (ROC) curves were 0.91 for the CAGE and 0.61 for the MAST.
Conclusion:
The CAGE and the MAST were both characterized by low sensitivities at conventional cutoff points, but the CAGE was significantly more effective than the MAST in discriminating between elderly medical outpatients with and without alcohol abuse or dependence.
