Quality improvement for vibration controlled transient elastography using a novel analysis algorithm.

BACKGROUND: The correlation of liver stiffness (LS) score with histopathology is limited by diagnostic equipoise and variability by fibrosis etiology. This study devised a novel probabilistic reporting algorithm for VCTE output, stratified by etiology of liver disease, and obtained feedback from providers about this new framework. METHODS: We devised a novel probability algorithm by re-analyzing and converting optimal sensitivity and specificity cut-offs from the Refining the Baveno VI Elastography Criteria Study (i.e., <7 kPa, 10 to 15 kPa, and >15 kPa) into probabilities of F0-F4. We subsequently conducted an email-based, 9-item survey to hepatology providers at the Cleveland Clinic, assessing patient and provider experience. Responses to each question ranged from 1 to 5 (strongly disagree to strongly agree); each question was aggregated to create 3 categories of response: favorable, neutral, and not favorable. RESULTS: The statistical conversion yielded probabilities of LS on VCTE for patients with cACLD stratified by MASLD, Hepatitis B, Hepatitis C, and Alcoholic Liver Disease and was implemented in VCTE output in March 2023 at the Cleveland Clinic. Survey of provider experience showed it overall more favorable than the conventional algorithm with respect to facilitating understanding of fibrosis, communicating with referring physicians and surveying patients. The new algorithm was most beneficial for patient experience when biopsy and VCTE results are divergent. CONCLUSION: Our novel probabilistic algorithm acknowledging the intrinsic limitations of correlating LSM with histologic fibrosis stage, particularly at intermediate values. Overall, providers found this reporting output improved their experience in caring for patients and helpful for the patient experience when biopsy and VCTE results are discordant. Refinements to this reporting output are needed to enhance patients' understanding of test results.