EXPLAINING INCONSISTENCIES BETWEEN PATIENTS AND COLLATERALS - VALIDITY IN OUTCOME STUDIES OF COERCIVE ALCOHOLISM-TREATMENT

This study aims to assess the consistency of replies to questionnaires mailed to patients and two kinds of collaterals, i.e., social workers and significant others, at a public treatment center with socially un stable and compulsory committed patients. It compares the quantity and kind of discrepancies between replies by patients and collaterals on outcome data concerning social situation and drinking habits. It aims to measure the amount of systematic bias among factors that may explai n inconsistencies between reports, especially the factors compulsory c ommitment, worse outcome, frequency of contact and type of collateral. The responders generally agreed. Variables in which there was less ag reement were explored in logistic regressions using ten explanatory va riables. Significant relations did not exceed those expected by chance . Discrepancies were not systematic in size and kind. On a six-rank or dinal scale of alcohol use or abuse, however, patients tended to under estimate the extent of their abuse. Inconsistencies here concerned the degree but not the presence of abuse. No difference in consistency du e to type of collateral was found. In conclusion, the consistency of t he questionnaires was high and independent of the social situation of the patient, of compulsory commitment and of other background or treat ment factors, as well as of treatment outcome and type of collateral.