Integration and evaluation of implementation strategies to improve guideline-concordant bladder cancer surveillance:...

Despite guideline recommendations, our prior work revealed more than half of low-risk bladder cancer patients within the Department of Veterans Affairs (VA) undergo too many surveillance procedures and about a third of high-risk patients do not undergo enough procedures. Thus, we developed and integrated implementation strategies to improve risk-aligned bladder cancer surveillance for the VA.

Prior work used Implementation Mapping to develop nine implementation strategies: change record systems, educational meetings, champions, tailoring, preparing patients to be active participants, external facilitation, remind clinicians, audit & feedback, and a blueprint. We integrated these strategies as improvement approaches across four VA urology clinics. Primary implementation outcomes were qualitatively measured via coding of semi-structured interviews with clinicians and co-occurrence of codes. Implementation outcomes included: appropriateness, acceptability, and feasibility. Exploratory quantitative outcomes included clinicians' recommendations for guideline-concordant bladder cancer surveillance intervals and sustainability.

Eleven urologists were interviewed. Co-occurrence analysis of codes across strategies indicated that urologists most commonly reported on the acceptability and appropriateness of changing the record system, preparing patients to be active participants ("surveillance grid"), reminders (i.e., cheat sheet), and educational sessions. We confirmed feasibility of all implementation strategies. Urologists indicated that changing the record system had a high impact, reduced documentation time, and guided resident physicians. Preparing patients to be active participants using the "surveillance grid" was seen as an effective but time-consuming tool. Educational sessions were seen as critical to support implementation. In quantitative analyses, clinicians recommended guideline-concordant surveillance about 65% of the time at baseline for low-risk patients, and this improved to 70% during evaluation. Across all risk levels, the largest improvement was observed at site 2 while site 3 did not improve. All sites sustained use of the changed record system, while sustainability of other strategies was variable.

Based on summative interpretation of results, the most appropriate, acceptable, and feasible strategies include changing record systems via a template and educational meetings focused on guideline-concordant surveillance. Future work should assess the impact of the improvement approaches on clinical care processes, particularly on reducing overuse of surveillance procedures among low-risk patients.

The implementation strategies were not considered a healthcare intervention on human participants by the governing funding agency and IRB. Rather, they were seen as quality improvement interventions. Thus, this study did not meet criteria for a clinical trial and was not registered as such.