Mis-implementation of evidence-based behavioural health practices in primary care: lessons from randomised trials in Federally Qualified Health Centers

Abstract

Background: Implementing evidence-based practices (EBPs) within service systems is critical to population-level health improvements, but also challenging, especially for complex behavioural health interventions in low-resource settings. ‘Mis-implementation’ refers to poor outcomes from an EBP implementation effort; mis-implementation outcomes are an important, but largely untapped, source of information about how to improve knowledge exchange. Aims and objectives: We present mis-implementation cases from three pragmatic trials of behavioural health EBPs in US Federally Qualified Health Centers (FQHCs). Methods: We adapted the Consolidated Framework for Implementation Research and its Outcomes Addendum into a framework for mis-implementation and used it to structure the case summaries with information about the EBP and trial, mis-implementation outcomes, and associated determinants (barriers and facilitators). We compared the three cases to identify shared and unique mis-implementation factors. Findings: Across cases, there was limited adoption and fidelity to the interventions, which led to eventual discontinuation. Barriers contributing to mis-implementation included intervention complexity, low buy-in from overburdened providers, lack of alignment between providers and leadership, and COVID-19-related stressors. Mis-implementation occurred earlier in cases that experienced both patient- and provider-level barriers, and that were conducted during the COVID-19 pandemic. Discussion and conclusion: Multilevel determinants contributed to EBP mis-implementation in FQHCs, limiting the ability of these health systems to benefit from knowledge exchange. To minimise mis-implementation, knowledge exchange strategies should be designed around common, core barriers but also flexible enough to address a variety of site-specific contextual factors, and should be tailored to relevant audiences such as providers, patients, and/or leadership.