Clinical Implementation of DIGEST as an Evidence-Based Practice Tool for Videofluoroscopy in Oncology: A Six-Year Single Institution Implementation Evaluation.

Abstract

Clinical implementation of evidence-based practice (EBP) tools is a healthcare priority. The Dynamic Grade of Swallowing Toxicity (DIGEST) is an EBP tool developed in 2016 for videofluoroscopy in head and neck (H&N) oncology with clinical implementation as a goal. We sought to examine: (1) feasibility of clinical implementation of DIGEST in a national comprehensive cancer center, and (2) fidelity of DIGEST adoption in real-world practice. A retrospective implementation evaluation was conducted in accordance with the STARI framework. Electronic health record (EHR) databases were queried for all consecutive modified barium swallow (MBS) studies conducted at MD Anderson Cancer Center from 2016 to 2021. Implementation outcomes included: feasibility as measured by DIGEST reporting in EHR (as a marker of clinical use) and fidelity as measured by accuracy of DIGEST reporting relative to the decision-tree logic (penetration-aspiration scale [PAS], residue, and Safety [S] and Efficiency [E] grades). Contextual factors examined included year, setting, cancer type, MBS indication, and provider. 13,055 MBS were conducted by 29 providers in 7,842 unique patients across the lifespan in diverse oncology populations (69% M; age 1-96 years; 58% H&N cancer; 10% inpatient, 90% outpatient). DIGEST was reported in 12,137/13,088 exams over the 6-year implementation period representing 93% (95% CI: 93-94%) adoption in all exams and 99% (95% CI: 98-99%) of exams excluding the total laryngectomy population (n = 730). DIGEST reporting varied modestly by year, cancer type, and setting/provider (> 91% in all subgroups, p < 0.001). Accuracy of DIGEST reporting was high for overall DIGEST (incorrect SE profile 1.6%, 200/12,137), DIGEST-safety (incorrect PAS 0.4% 51/12,137) and DIGEST-efficiency (incorrect residue 1.2%, 148/12,137). Clinical implementation of DIGEST was feasible with high fidelity in a busy oncology practice across a large number of providers. Adoption of the tool across the lifespan in diverse cancer diagnoses may motivate validation beyond H&N oncology.