A prospective multicentre study of in-patient treatment for alcoholics: 18- and 48-month follow-up (Munich Evaluation for Alcoholism Treatment, MEAT).

In a prospective multicentre study of 1410 alcoholics (73% men) data were collected at five different times: admission, discharge, 6, 18 and 48 months after discharge. The details from the 21 treatment centres involved were acquired from staff-members; follow-up data on patients were collected by personal interviews and/or mailed questionnaires. In all, 85%, 84% and 81% of the patients could be traced, respectively. Additionally, data of patients' sick leave days and in-patient treatment were provided by the health insurance and pension insurance institutions, respectively. Outcome criteria were drinking behaviour, working and partner situation, and subjective complaints. The drinking behaviour was divided into three rough categories: abstinent, improved and unimproved. For 18 months 53% of the patients remained abstinent, 8.5% improved and 38% did not improve. For 48 months 46% remained abstinent, 12% improved and 42% did not improve. During the last 6 months prior to the 48-month data collection 66% were abstinent, 4% improved and 30% did not improve. Only 3% of patients succeeded in maintaining controlled drinking. The percentage of days of sick leave was reduced by 64% and of in-patient treatment from 34.7% to 14.6% during the 18-month period after discharge (in comparison with the 18-month period prior to treatment). Only 21% of the patients regularly attended self-help groups. Out of the patients' variables, ten for men and five for women could be identified as prognostically relevant. In the 48-month follow-up these factors were reconsidered. In men almost all, in women only three of the five factors were confirmed. The treatment variables were evaluated according to the prognosis factors (positive vs negative group). In the 48-month follow-up the treatment variables relevant in the 18-month follow-up were also reassessed. In the positive prognosis group five variables were confirmed, in the negative prognosis group only one. In addition, differentiated indication variables for the three treatment lengths were developed and applied to a model. The following appeared to be clues regarding the length of desirable treatment. For an unfavourable prognosis in both men and women no short-term treatment should be given; medium- or long-term treatment is to be preferred. For a medium prognosis men do better with short-term treatment; for women medium-treatment is preferred. For a favourable prognosis for men medium-term treatment should be avoided; long-term is preferred; for women short-term treatment may also be preferred.