Comparing theory-driven and intuition-based approaches to inform implementation strategies in practice: an...

To explicitly assess the impact of theory in implementation strategies, it is essential to isolate theoretical components while controlling other variables.

This exploratory two-arm cluster-randomized implementation trial compared theory-driven versus intuition-based approaches to the development of strategies to improve risk-appropriate tumor testing and referral to genetics services for Lynch syndrome (LS), a hereditary cancer predisposition condition.

Seven major Australian hospitals received an identical, seven-step implementation approach, with the only difference being either the use of theory-driven or intuition-based implementation strategy development, which was randomly assigned (theory n = 4; intuition n = 3). The theory-driven approach was guided by the Theoretical Domains Framework (step 4) and matched Behavior Change Techniques (step 5), with all other steps identical between both trial arms. The primary outcome was the rate of risk-appropriate LS tumor testing and referral within 2 months of colorectal cancer (CRC) resection, using data from January 1, 2017 to September 30, 2021 (n = 3,321 patients). Secondary outcomes included the proportion of risk-appropriate patients with a genetic referral.

The theory-driven arm showed a modest but non-significant improvement in the primary outcome of risk-appropriate LS pathway completion (adjusted risk ratio [aRR]=1.22, 95% CI: 0.92- 1.61, p  = 0.1713); however, this was driven by one large hospital, with null effect when excluding that hospital (aRR = 1.00, 95% CI: 0.86-1.18, p  = 0.9639) and heterogeneous results across different hospitals. For patients with high LS risk tumor test results, the theory-driven arm showed a significantly lower proportion without a genetics referral (secondary outcome: aRR = 0.50, 95% CI: 0.41-0.61, p  < 0.0001).

Theory-driven implementation led to stronger improvements in a secondary outcome. Primary outcome evidence was inconclusive. While theory-driven approaches led to some improvements in a secondary outcome, there was no conclusive difference in the primary outcome. A published process evaluation reports on the contextual factors influencing results at each site, as well as theoretical alignment of intuitive strategies and their proposed mechanisms of action. This trial design offers a template for isolating and evaluating the contribution of theory to implementation strategies.

Australian New Zealand Clinical Trials Registry, ACTRN12618001072202. Registered June 27, 2018 https://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=375348&isReview=true.

http://links.lww.com/IJEBH/A491.