Pragmatic fidelity measurement in youth service settings.

Background: Fidelity measurement is critical for developing, evaluating, and implementing evidence-based treatments (EBTs). However, traditional fidelity measurement tools are often not feasible for community-based settings. We developed a short fidelity rating form for the Collaborative Problem Solving (CPS) approach from an existing manualized coding system that requires extensive training. We examined the reliability and accuracy of this short form when completed by trained observers, untrained observers, and self-reporting providers to evaluate multiple options for reducing barriers to fidelity measurement in community-based settings.

Methods: Community-based treatment providers submitted recordings of youth service sessions in which they did, or did not, use CPS. For 60 recordings, we compared short-form fidelity ratings assigned by trained observers and untrained observers to those provided by trained observers on the manualized coding system. For 141 recordings, we compared providers' self-reported fidelity on the short form to ratings provided by trained observers on the manualized coding system and examined providers' accuracy as a function of their global fidelity.

Results & conclusions: The short form was reliable and accurate for trained observers. An assigned global integrity score and a calculated average of component scores on the short form, but not component scores themselves, were reliable and accurate for observers who had CPS expertise but no specific training on rating CPS fidelity. When providers self-reported fidelity on the short form, their global integrity score was a reliable estimate of their CPS integrity; however, providers with better CPS fidelity were most accurate in their self-reports. We discuss the costs and benefits of these more pragmatic fidelity measurement options in community-based settings.