Reliability of Identification of Behavior Change Techniques in Intervention Descriptions

The aim of this paper is to assess the frequency of identification as well as the inter-coder and test-retest reliability of identification of behavior change techniques (BCTs) in written intervention descriptions. Forty trained coders applied the "Behavior Change Technique Taxonomy version 1" (BCTTv1) to 40 intervention descriptions published in protocols and repeated this 1 month later. Eighty of 93 defined BCTs were identified by at least one trained coder, and 22 BCTs were identified in 16 (40 %) or more of 40 descriptions. Good inter-coder reliability was observed across 80 BCTs identified in the protocols: 66 (80 %) achieved mean prevalence and bias-adjusted kappa (PABAK) scores of 0.70 or greater, and 59 (74 %) achieved mean scores of 0.80 or greater. There was good within-coder agreement between baseline and 1 month, demonstrating good test-retest reliability. BCTTv1 can be used by trained coders to identify BCTs in intervention descriptions reliably. However, some frequently occurring BCT definitions require further clarification.