ity of the research with the organization's mission and culture, and its financial stability. Thus, research roles may vary from relatively passive participation (completing surveys and submitting data to state databases) to involvement as a partner in the development of research questions, data collection, and data interpretation. However, incentives for all parties must be strategically aligned if real progress is to be made.
The committee identified barriers to closing the gap between treatment, research, and policy. These barriers range from organizational factors, stigma, and social policy to cultural differences and funding problems, all of which can be strong disincentives for the collaboration needed to advance the field.
A pervasive theme heard in our workshops was the need for communication, mutual respect, and trust. Values of researchers and providers often differ and these differences must be recognized and resolved. The conduct of community-based research is an intensely interpersonal enterprise and trust relationships must be cultivated, at different levels of the organization, with community residents, and often with members of other agencies connected to the CBO. These relationships often take years to build.
The trust necessary for long-term collaboration is generally based on a history of increasing involvement. Successful collaborative programs from other health fields include support for a permanent infrastructure that facilitates long-term development. The National Cancer Institute's Community Clinical Oncology Program (CCOP) uses this strategy to bring state-of-the-art oncology research to community-based cancer treatment programs. CCOP facilitates research collaborations and enhances the ability of treatment programs to apply research findings to the general patient population. Development of a similar mechanism for use in community-based drug and alcohol abuse treatment could catalyze research/practice collaborations and stimulate improvements in practice. CCOPs are not inexpensive and they present a significant managerial challenge. The infrastructure alone at each clinical site can exceed $200,000. However, the infrastructure recommendation that follows does not necessarily require a model with that complexity. It could begin as a demonstration project involving the funding of one full-time-equivalent staff person and some computer support to a small set of diverse treatment sites. This level of support could be the target, whichever of the various network collaboration models was implemented.
Based on these findings, the committee offers two recommendations and identifies certain key characteristics that will facilitate their successful implementation.