Field Test of a Tool for Level-of-Care Decisions in Community Mental Health Systems

Abstract

Tools for supporting decisions about placing clients in different levels of mental health care can facilitate several important functions, such as structuring service allocation, providing standards for quality assurance and concurrent review, designing service and benefit packages, and planning for resource needs. This paper describes a level-of-care decision-support tool and tests its reliability and concurrent validity. Panels of clinical managers and administrators of mental health centers developed a decision-support tool with eight levels of care and a computerized decision-tree algorithm for level-of-care placement. A random sample of 1,034 adults from one county mental health system were assessed by their case managers using a package that included variables in the eight-level model, but not the algorithm itself. Other variables were included to assess the tool's concurrent validity. Level-of-care placements based on the decision-support tool showed strong interrater reliability. Concurrent validity was demonstrated by significant relationships in the expected direction between level of care and psychiatric hospitalizations, arrests, residential moves, homeless periods, residential independence, lack of work activity, medication noncompliance, and functioning as measured by the Global Assessment of Functioning Scale. Preliminary empirical evidence indicates that the level-of-care decision-support tool is reliable and valid. It could be further refined by incorporating the impact of social supports, collateral services, current mental health services, and motivation for services.