Objective: The aims of this study were to (1) develop an approach to assess the delivery fidelity of a complex intervention to simultaneously increase physical activity and reduce smoking and (2) use this approach to assess the variation of fidelity across the delivery of different intervention components of the intervention.
Method: Audio recorded and transcribed sessions (90 in total) involving 30 participants and 3 health trainers delivering a one-to-one intervention were purposively sampled across health trainer (HT) and stage of treatment. The Dreyfus system for skill acquisition informed a scoring system based on 12 intervention processes and applied by three experts in health behavior change. Scores ranged from 0 to 2 (poor quality), 3 to 4 (reasonable quality), and 5 to 6 (expert level quality). Scores were averaged across coders and presented in relation to fidelity of both HT and the intervention component.
Results: The methods were successfully applied with recommendations for future application. Average scores for each item by each coder differed by up to +0.7 to -0.9 points indicating reasonable agreement. Mean scores for the three HTs were 2.9, 2.2, and 2.4, across all 12 intervention processes. The delivery of all intervention components for physical activity was scored lower (<3) than their respective counterparts for smoking reduction (>3; p < .001).
Conclusions: Novel methods for assessing delivery fidelity were successfully applied and areas for improvement identified. Delivery fidelity was deemed to be of reasonable quality but was higher for smoking related intervention components over physical activity ones. (PsycINFO Database Record
(c) 2018 APA, all rights reserved).