MMPI) or revised Minnesota Multiphasic Personality Inventory (MMPI-2) scales, purported to be predictive of treatment outcome, were related to length of stay (LOS) in a domiciliary and to discharge status from a domiciliary for 335 subjects. The revised MacAndrew Alcoholism Scale (MAC-R), Addiction Acknowledgement Scale (AAS), and Addiction Potential Scale (APS), scales indicating a propensity to use illicit substances, and the Psychopathic Deviate Scale (Pd), indicating acting-out, all correlate negatively with length of stay. AAS also correlates negatively with regular/irregular discharge. Other scales expected to predict treatment outcome do not significantly correlate with either length of stay or type of discharge. 0 1996 John Wiley & Sons, Inc.* Dahlstrom, Welch, & Dahlstrom (1975, pp. 171-172) list 159 different studies that focus upon Minnesota Multiphasic Personality (MMPI) scales used to predict treatment outcome. They categorized these studies according to need for treatment, general prognosis for treatment, predicting outcome of hospitalization, predicting response to shock therapy, predicting response to rehabilitation, predicting suitability for group therapy, predicting persistence in treatment regimen, role of therapist characteristics, and patient-therapist dyad studies.
Scarcity of resources in the health and mental health arenas has demonstrated a need for prognostic indicators. The revisors (Butcher, Dahlstrom, Graham, Tellegen, & Kaemmer, 1989) of the Minnesota Multiphasic Personality Inventory (MMPI-2) noted this need by developing yet another prognostic treatment scale (TRT).
The reliability of MMPI predictor scales may be questionable. At times, various MMPI scales have differentiated well and would seem to have potential as predictors, and, at other times, the same scales have performed poorly (Levitt, 1989). Mauger (1980) noted that the performance of outcome scales was often distorted by pathology, chronicity, the nature of the treatment, and even when the MMPI was administered during the course of treatment.
Most of these scales have not been cross-validated. The purpose of this study is to evaluate a number of scales expected to predict outcome of treatment for their suitability for predicting success of residents enrolled in a domiciliary, an extended care treatment facility for veterans with medical and/or psychiatric problems.
METHOD
Subjects
The subjects were a random sample of veterans in a Department of Veterans Affairs domiciliary. The MMPI-2 was administered after admission to all domiciliary residents who We gratefully acknowledge the assistance of Linda A.