Health Services Research & Development

Veterans Crisis Line Badge
Go to the ORD website
Go to the QUERI website

IIR 16-024 – HSR&D Study

New | Current | Completed | DRA | DRE | Portfolios/Projects | Centers | Career Development Projects

IIR 16-024
Advanced Prediction Models to Optimize Treatment and Access for Veterans with Hepatitis C
Akbar K Waljee MD MSc
VA Ann Arbor Healthcare System, Ann Arbor, MI
Ann Arbor, MI
Funding Period: April 2017 - March 2021

BACKGROUND/RATIONALE:
Chronic hepatitis C (CHC) continues to present challenges in the treatment of Veterans. Direct acting antiviral agents (DAAs) for the treatment of CHC are highly effective, have a favorable safety profile, and are well tolerated, eliciting significant consumer demand. However, DAAs are costly, making it difficult for some healthcare systems to meet this growing demand from patients.

OBJECTIVE(S):
The purpose of this study is to lay the groundwork for risk-based treatment of CHC among non-cirrhotic Veterans in the Veterans Health Administration (VHA) by: (Aim 1) developing accurate, clinically relevant, and implementable risk prediction models; (Aim 2) engaging Veterans to develop consensus on how to implement risk-based treatment; and (Aim 3) evaluating the clinical and economic effects of risk-based treatment.

METHODS:
Preliminary work demonstrated the feasibility of using a machine-learning (ML) risk prediction model to identify patients at high risk and low risk for disease progression in a clinical trial cohort. In this 4-year study, we will use VA electronic data from 2004-2014 to adapt, validate and refine this model among Veterans with CHC. We will also engage Veterans with and without CHC, eliciting their preferences and values regarding risk-based treatment of CHC by applying consensus techniques (e.g., deliberative democracy). Finally, to estimate the incremental benefit of risk-based treatment over current treatment, we will use simulation modeling.

FINDINGS/RESULTS:
Aim 1: Of the 150K Veterans in our cohort, an initial evaluation of the clinical trial cohort model was conducted on a subset of 15K Veterans and found an AuROC curve of 0.6 to predict cirrhosis. Given this relatively poor performance, we developed a new model on our 150K cohort using a 70% training, 30% testing cohort. We used a cox model to build a longitudinal model with a concordance of 0.76. We will continue to refine this model to predict both cirrhosis and fibrosis using both APRI and FIB4 and in addition, test other machine learning and deep learning models.

Aim 2: Out of 15 Veterans consented, 11 interviews were completed to assess knowledge of hepatitis C and provide insight into policy decisions surrounding treatment of HCV. It was found that Veteran knowledge is variable in terms of how you contract HCV and its symptoms. Additionally, most Veterans agreed that those with HCV that were the sickest should be treated first. The interviews provided insight that helped develop materials for the Deliberative Democracy session (Aim 2) that will take place in 2018.

IMPACT:
Despite highly effective treatment for CHC, we cannot treat all infected patients immediately. A risk-based, systematic approach will maximize treatment benefit while limiting clinical and economic harms. This research will produce a strategy to systematically identify high-risk patients for early treatment and lay the groundwork for future implementation of this strategy. The approach used in this study can also serve as a template for future efforts to implement risk-based management strategies in other clinical contexts. The DD session will allow Veterans to enhance our understanding of what Veterans believe is important when policy makers establish guidelines surrounding such treatments for high-cost medications like those for HCV.

PUBLICATIONS:
None at this time.


DRA: Infectious Diseases
DRE: Technology Development and Assessment, Treatment - Implementation, TRL - Applied/Translational
Keywords: Care Management Tools, Decision Support, Healthcare Algorithms
MeSH Terms: none