Abstract

Background: AASLD and EASL guidelines recommend all people with cirrhosis undergo twice yearly screening for hepatocellular carcinoma (HCC) with hepatic imaging. However, patient, provider, and system level barriers impede ongoing surveillance efforts. This stepped-wedge hybrid effectiveness-implementation trial assessed the impacts of using a quality improvement playbook called Getting to Implementation (GTI) to support VA facilities to select, implement, and evaluate data-driven strategies to improve HCC surveillance.

Methods: This hybrid type III (implementation-effectiveness) stepped-wedge cluster randomized design was conducted at 12 VA sites between October 2020 and April 2023. We used a multi-faceted facilitation strategy consisting of manualized GTI during a 12-month active implementation and six-month sustainment period. The primary implementation outcome was GTI completion and strategy implementation. The secondary clinical outcome was receipt of guideline-concordant HCC surveillance at baseline, post-intervention, and sustainment. Analysis involved a three-level, generalized linear mixed model.

Results: Of 12 VA facilities, selected based on having low baseline HCC surveillance rates, 10 completed GTI with high fidelity. These 10 sites implemented a median of four implementation strategies while receiving an average of 19±5 facilitation hours. HCC surveillance improved from 21% at baseline to 30% during intervention and remained elevated at 32% during sustainment. Sites receiving more facilitation (r=0.59, p=0.048) and sites implementing a greater variety of strategies had higher HCC surveillance improvement. Generalized linear mixed models indicated significant changes in HCC surveillance during both implementation (aOR=1.306; 95% CI: [1.159, 1.472], p<0.0001) and sustainment (aOR versus control=1.511; 95% CI: [1.315, 1.73], p-value <0.0001). Sustainment, a challenge for implementation trials, was significantly associated with improvement in HCC surveillance compared with active implementation (aOR=1.168; 95% CI: [1.018, 1.340], p-value 0.0271).

Conclusion: Data-driven strategies with facilitated quality improvement sustainably improved HCC surveillance in Veterans with cirrhosis receiving care in the lowest-performing VA facilities. Further research is needed to understand the heterogenous effects across sites, which may have been driven by differences in site baseline characteristics and facilitation and strategy implementation nuances.