READINESS TO CHANGE QUESTIONNAIRE: RELIABILITY STUDY OF ITS SPANISH VERSION

• Published in: Alcohol and alcoholism (Oxford) 2000-05, Vol.35 (3), p.270-275
• Main Authors: RODRÍGUEZ-MARTOS, A., RUBIO, G., AUBÀ, J., SANTO-DOMINGO, J., TORRALBA, Ll, CAMPILLO, M.
• Format: Article
• Language: English
• Subjects: Addictive behaviors; Adult; Adult and adolescent clinical studies; Alcohol Drinking - epidemiology; Alcohol Drinking - prevention & control; Alcohol Drinking - psychology; Alcoholism; Analysis of Variance; Biological and medical sciences; Factor Analysis, Statistical; Female; Humans; Male; Medical sciences; Middle Aged; Problem drinking; Psychological tests; Psychology. Psychoanalysis. Psychiatry; Psychopathology. Psychiatry; Readiness to Change Questionnaire; Reliability; Spain - epidemiology; Spanish version; Surveys and Questionnaires; Translating
• ISSN: 0735-0414; 1464-3502; 1464-3502
• DOI: 10.1093/alcalc/35.3.270

— The present study explored the reliability and validity of a Spanish version of the Readiness to Change Questionnaire (RCQ) (12-item short form) as it might be used for opportunistic intervention. The test has three scales to allocate patients to a stage of change: pre-contemplation (P), contemplation (C) or action (A). The RCQ was translated and back-translated prior to pilot administration to 15 patients. From two settings (a general hospital ward and a primary health care centre), 201 patients were identified as excessive drinkers on the Alcohol Use Disorders Identification Test. Patients known to be alcohol-dependent and attending for alcohol-related reasons were excluded. Patients completed the RCQ. Test–retest reliability after 2 days was assessed in 35 patients. A components analysis was performed. Patients were classified on RCQ scores to a stage of change. Two experts separately interviewed the patients and made an allocation to stage of change, blind to the RCQ score. Test–retest reliability was good (P: r = 0.81; C: r = 0.87; A: r = 0.86). Within the three scales, RCQ items showed fair consistency in terms of Cronbach's alpha (P: 0.58, C: 0.75, A: 0.80). Component analysis showed that together the scales accounted for 57.4% of the variance. The experts agreed between themselves on patients' stage of change (weighted kappa 0.92) but much less with the stage of change according to RCQ (expert A, kappa = 0.44; expert B, kappa = 0.52). Omitting patients with low consumption did not improve internal reliability, and omitting those with low educational level who might have filled in the questionnaire wrongly did not improve internal reliability or agreement between RCQ and the experts. We conclude that the Spanish RCQ did not function efficiently in a population of opportunistically identified excessive drinkers.