Assessing Performance of a Hepatitis C Emergency Department (HepC-EnD) Screening Tool: IT Integration Process for Electronic Health Record System

Overview

The goal of this observational study is to develop, implement, and evaluate a machine learning algorithm-based Hepatitis C Emergency Department (HepC-EnD) screening tool for use in emergency departments (EDs) to identify patients at high risk of hepatitis C virus (HCV) infection. HepC-EnD will be integrated into the University of Florida Health electronic health record (EHR) system as a best practice alert (BPA) pop-up for ED providers, notifying them of patients at high risk for HCV infection and recommending both HCV and human immunodeficiency virus (HIV) screening. Investigators aim to enhance the screening and diagnosis of individuals who may otherwise remain undiagnosed and untreated. The implementation outcomes (e.g., usability) and effectiveness outcomes (e.g., HCV screening and diagnosis rates) of HepC-EnD targeted screening will be compared with universal screening (FOCUS) and conventional physician-initiated screening programs in EDs.

Detailed description

HCV infection has markedly increased in the United States, primarily resulting from injection drug use associated with the ongoing opioid epidemic. Despite the availability of highly effective direct-acting antiviral therapy, more than half of individuals with chronic HCV remain undiagnosed, leading to significant morbidity and mortality. EDs represent a critical setting for HCV and HIV screening, as they are currently the most common setting for missed diagnostic opportunities. However, universal ED-based screening programs are often costly and unsustainable. Moreover, existing targeted screening programs are limited, and have not been systematically developed or rigorously evaluated in clinical practice. Thus, there is a critical public health need to develop innovative, tailored, effective, and sustainable screening strategies to enhance HCV screening in EDs.

This study will accomplish three specific aims:

1. Develop and validate prediction algorithms using machine learning and natural language processing (NLP) to identify patients at high risk of HCV infection 2. Develop the HCV screening tool prototype HepC-EnD for implementation in EDs 3. Compare the usability, effectiveness, and cost-effectiveness of an automated HepC-EnD prompt for HCV (with HIV) testing versus universal and physician-initiated screening strategies

This study is guided by multiple implementation science frameworks, including the Exploration, Preparation, Implementation, Sustainment (EPIS) framework, Proctor's Implementation Outcomes, and Five Rights, which will greatly increase the tool's utility, sustainability, and generalizability.

The investigators will conduct a quasi-experimental study to compare HepC-EnD to two existing screening strategies across three UF Health EDs over 12 months (6 months pre-implementation and 6 months post-implementation). UF Jacksonville Downtown ED will transition from universal screening (FOCUS) to HepC-EnD. UF Jacksonville North ED will continue FOCUS throughout the study period to serve as a control. UF Gainesville ED will pilot HepC-EnD, as FOCUS has not previously been implemented at that site. The study will evaluate the effectiveness of HepC-EnD's within-site and between-site comparisons.

The investigator's central hypothesis is that the use of HepC-EnD will have lower HCV and HIV screening rates but higher diagnosis rates and will be more cost-effective than universal screening and physician-initiated screening strategies.

Interventions

• Other Physician-Initiated Screening (Conventional Screening)
  Screening for HCV and HIV in patients presenting to the ED occurs when an ED provider initiates screening based on symptoms or clinical judgement. Providers will manually order individual tests in the EHR.
• Other FOCUS (Universal Screening)
  During nurse triage, a FOCUS screening question will appear in the EHR and the patient will be asked to opt-in to HCV and HIV testing. For those who consented, if an ED provider enters a phlebotomy order for any reason in the EHR, a BPA will alert the providers to suggest HCV and HIV testing. The provider can decide to "order" or "do not order" for each test individually. Ordered tests automatically trigger the following in the EHR: HCV antibody with reflex to RNA and HIV 1/2 antigen/antibody wi
• Other HepC-EnD (Targeted Screening)
  HepC-EnD will run in real time once integrated into the hospital's Epic EHR system. When the patient comes to the ED waiting room, a risk score generated from HepC-EnD will be available and determine if the patient is at high risk of HCV infection (\> cutoff risk score). If the patient is determined to be at high risk, a HepC-EnD screening question will appear in the EHR during nurse triage and the patient will be asked will be asked to opt-in to HCV and HIV testing. For those who consented, a B

Primary outcome measures

• Proportion of new HCV or HIV diagnoses [Time frame: Time Frame: 6 months pre- and post-implementation]
• Absolute number of new HCV or HIV diagnoses [Time frame: 6 months pre- and post-implementation]

Secondary outcome measures (4)

• Proportion of BPA alerts among individuals presenting to EDs [Time frame: 6 months pre- and post-implementation]
• Proportion of HCV and HIV tests performed among BPA alerts [Time frame: 6 months pre- and post-implementation]
• Proportion of patients linked to care among those with positive HCV and HIV diagnoses [Time frame: 3 months after diagnosis]
• Composite HCV or HIV Diagnoses [Time frame: 6 months pre- and post-implementation]

Eligibility criteria

Inclusion criteria

• 18-79 years of age

Exclusion criteria

• < 18 years of age
• Medically unstable

Criteria are shown verbatim from the registry (in English). Final eligibility is always assessed by the study center.

Healthy volunteers: Yes

Study design

Observational model

Cohort

Study locations

United States · 3 centers

• UF Health Shands Emergency Room / Trauma Center — Gainesville
• UF Health Jacksonville Emergency Room — Jacksonville
• UF Health North Emergency Room — Jacksonville

Publications

• Jang SC, Lo-Ciganic WH, Hernandez-Con P, Jenjai C, Huang J, Stultz A, Yan S, Wilson DL, Norse A, Guirgis FW, Cook RL, Gage C, Nguyen KA, Hornes P, Wu Y, Nelson DR, Park H. Development and Validation of a Machine Learning-Based Screening Algorithm to Predict High-Risk Hepatitis C Infection. Open Forum Infect Dis. 2025 Aug 15;12(8):ofaf496. doi: 10.1093/ofid/ofaf496. eCollection 2025 Aug. PMID 40874186

Identifiers

NCT: NCT07406490 · IRB202500471 · R01DA057886 · IRB202600334

Primary sources (government registries)

• ClinicalTrials.gov (US NIH)

View this study on ClinicalTrials.gov ↗