This article describes two sets of field studies undertaken by the program evaluation unit of a community mental health center. These studies analyzed clients' utilization of service and assessed service impact in the process of testing procedural variations in service delivery. In the first set of studies, a procedure for ensuring verbal client-therapist contact prior to the first appointment was developed and tested. This procedure reduced the no-show rate for initial appointments from 22 to 12%. In the second set of studies, a brief pretherapy orientation nearly eliminated dropout during the first month of therapy. Orientation had both short- and long-range impact on the amount of services used by clients as well as on their outcomes. Therapist's global ratings of client functioning reflected more change for oriented clients, who reported greater short-term symptom reduction as well. Non-oriented clients were more likely to drop out early and to impress their therapists less favorably. The results of these studies suggest that a combination of pretherapy orientation and verbal client-therapist contact prior to the initial appointment might greatly reduce the failure to complete treatment.

A multi-attribute utility analysis employing ideal outcome measure criteria was applied to seven brief rating scales in order to identify the best performing instrument. A variety of judgmental data were collected from therapists working in mental health service agencies and from evaluation research experts to contrast the performance of the seven rating scales on criteria for selecting outcome measures developed by an NIMH task force. Transformations of the performance data were weighted in accordance with priorities assigned to the criteria by the task force. Comparing the sums of the weighted scores across scales, two rating scales emerged as preferred selections for monitoring the effectiveness of programs that serve the chronically mentally ill.

No abstract

No abstract

multidimensional scaling, INTERSCAL, incomplete data designs,

No abstract

No abstract

No abstract