A review of the research on this comprehensive community-based service delivery model for the seriously mentally ill has shown that positive client outcomes are achieved when the original model is followed with regard to organization, staffing, and practice patterns. The reviewer concludes that not implementing the program fully and not providing the necessary staff training will jeopardize the ability of the program to assist clients in becoming fully functioning members of their communities (Allness, 1997).

Several experts have recommended that researchers gain a deeper understanding of the treatment programs they hope to influence (Kavanaugh, 1995). Sobell adapts business techniques and encourages behavioral scientists to get "close to the customer" in developing and fostering close working relationships (Sobell, 1996). Brown suggests that, at a minimum, to develop effective technology transfer, the innovator must obtain input from potential adopters about the relevance, clarity, credibility, and adaptability of an intervention (Brown, 1995). Trust-building experiences can include site visits, jointly sponsored seminars and staff development activities, and short-term exchanges of staff.

The building of successful research-treatment partnerships, which recognize the contribution of both the research and treatment communities, is one way to build trust. Partnerships can be successfully organized with the community group as senior partner, the academic group as senior partner, or in a balanced partnership (Mittelmark, 1990). The committee heard from several administrators of community-based drug treatment programs who emphasized the need to work for a collaborative relationship. One pointed out that who takes the lead in a proposal depends on the funding agency: if it is SAMHSA, the CBO leads, if the funder is NIH, the leader is the university-based researcher.

These issues are not unique to drug abuse. In the area of cancer treatment, 80 percent of care is provided in the community and the quality of care can be quite variable. In an attempt to improve the quality of care provided in the local communities, various organizations have issued guidelines for effective treatment procedures, but like the Treatment Improvement Protocols in the drug abuse area, writing a guideline does not guarantee that providers will comply (Czaja et al., 1997; Ford et al., 1987; Klabunde et al., 1997).

The National Cancer Institute's Community Clinical Oncology Program (CCOP) provides a model for researchers and clinicians seeking to collaborate (Cobau, 1994; Kaluzny et al., 1993, 1996). To better integrate policy, research, and treatment and thereby assure access to improved care within local communities, CCOP involves primary care physicians and oncologists in the conduct and management of clinical trials, in cooperation with NCI-funded centers and clinical cooperative groups. CCOP has become a valuable resource to NCI for performance of a wide variety of