HARSHBARGER ET AL. IMPLEMENTATION, ADAPTATION, AND TAILORING An Empirical Assessment of Implementation, Adaptation, and Tailoring: The Evaluation of CDC’s National Diffusion of VOICES/VOCES Camilla Harshbarger, Gretchen Simmons, Helen Coelho, Kira Sloop, and Charles Collins The Centers for Disease Control and Prevention (CDC), through its Diffusion of Ef- fective Behavioral Interventions (DEBI) program, trained over 260 agencies on VOICES/VOCES between August 2003 and April 2005. ORC Macro conducted in- terviews with agency staff 3 months after receiving VOICES/VOCES training. This article discusses the diffusion of VOICES/VOCES; agencies’ adoption, adaptation, and implementation of this intervention; and needs for ongoing proactive technical assistance (TA) for agencies to successfully integrate behavioral interventions into their programs. The vast majority of agencies implemented VOICES/VOCES with fi- delity to the core elements, and agencies successfully adapted the intervention to make it more appealing to target populations. TA is needed for interventions to be successfully adapted and implemented with fidelity to the core elements, and to en- sure program sustainability. More effective interventions of short duration and mini- mum complexity to easily match with existing resources and conditions of agency capacity among HIV prevention providers in the community are needed. BACKGROUND Despite the availability of resources for implementing interventions shown to be effective with certain populations, there is insufficient literature addressing systematic processes for translating, adapting, implementing, or evaluating behavioral HIV prevention interven- tions in community settings (Gandelman & Rietmeijer, 2004; Universitywide AIDS Re- search Project, 2004). Once an agency adopts an intervention, its actual impact depends on how it is implemented in the field. Replication is based on the premise that a successful research–based HIV prevention intervention conducted with fidelity in the community should produce outcomes similar to those in the intervention trial (Kelly, Somlai, et al., 2000). Therefore, the entire intervention protocol, including the number and sequence of sessions; the content, procedures, and activities in each of the sessions; and the settings and characteristics of the participants, should mirror those used in the research trial. 184 AIDS Education and Prevention, 18, Supplement A, 184–197, 2006 © 2006 The Guilford Press Camilla Harshbarger and Charles Collins are with the Centers for Disease Control and Prevention. Gretchen Simmons, Helen Coelho, and Kira Sloop are with ORC Macro. The authors would like to ac- knowledge the following former ORC Macro staff for instrument design, piloting, and data collection con- tributions to this project: Jyothi Thrivikramen, Erica Amato, and Gingi Pica. The findings and conclusions in this article are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention. Address correspondence to Camilla Harshbarger, Ph.D., Behavioral Scientist, Centers for Disease Control and Prevention, Division of HIV/AIDS Prevention, 1600 Clifton Road, NE, Mailstop E-40, Atlanta, GA 30333; e-mail: UZZ9@cdc.gov