Assessing fidelity: balancing methodology and reality in jail interventions

Patricia J. Kelly, Amanda Emerson, Chelsea Fair, Megha Ramaswamy

While fidelity to research protocols is important to ensure generalizable outcomes, interventions in criminal justice settings present unique challenges to uniform implementation. The goal of this paper is to describe the fidelity methods and outcomes for a sexual health intervention implemented in three local county jails. As part of a longitudinal cohort study, four trained fidelity assessors observed 25 of the 230 sessions presented (including both intervention and comparison groups) at three separate times during the 29 months of the intervention. Assessment methods included the assessors' field notes, a nine-item facilitator quality scale and a content inclusion scale with 6-13 items specific for each of the five sessions. Facilitator quality score ranged from 87.6 to 99.2%. Content inclusion scores ranged from 77.3 to 88%. Specific challenges to fidelity were found in two areas: the jail environment and the participants' response to content. The realities of conducting research in jails and prisons must be addressed in real time by adjusting program content to fit both unexpected facility and participant situations. Skilled facilitators are essential to this effort.