a b s t r a c t

Objectives: Emergency department (ED) Hepatitis C virus (HCV) screening programs are proliferating, and it is unknown whether EDs are more effective than traditional community screening at promoting HCV follow-up care. The objective of this study was to investigate whether patients screened HCV seropositive (HCV+) in the ED are linked to care and retained in treatment more successfully than patients screened HCV+ in the commu- nity.

Methods: A retrospective cohort study was performed including patients screened HCV+ at twelve screening fa- cilities in New Orleans, LA from March 1, 2015 to July 31, 2017. Treatment outcomes, including retention and time to follow-up care, were assessed using the HCV continuum of care model.

Results: ED patients (n = 3008) were significantly more likely to achieve RNA confirmation (aRR = 1.91, 95% CI

= 1.54-2.37), initiate HCV therapy (aRR = 2.23 [1.76-2.83]), complete HCV therapy (aRR = 1.77 [1.40-2.24]), and achieve HCV functional cure (aRR = 2.80 [1.09-7.23]) compared to community-screened patients (n = 322). ED screening was associated with decreased likelihood of fibrosis staging (aRR = 0.65 [0.51-0.82]) and no difference in linkage to specialty care (aRR = 1.03 [0.69-1.53]). In time to follow up, RNA confirmation oc- curred at faster rates in the ED (aHR = 2.26 [1.86-2.72]), although these patients completed fibrosis staging at slower rates (aHR = 0.49 [0.38-0.63]) than community patients.

Conclusions: Compared to community screening, HCV screening in the ED was associated with higher rates of dis- ease confirmation, treatment initiation/completion, and cure. Our findings provide new evidence that EDs may be the most effective setting to screen patients for HCV to promote follow-up care.

(C) 2022

1. Introduction

Hepatitis C virus (HCV) is the leading cause of mortality among Americans when compared to the next 60 highest-mortality, reportable infectious diseases combined (pre-COVID-19 pandemic) [1]. HCV is now a curable disease and diagnosis remains the primary barrier to combatting the HCV epidemic [2]. Community screening in ambulatory care clinics is the traditional model for HCV screening, yet nearly half of the HCV+ population (totaling 1.8 million Americans) remains un- aware that they are infected [3]. Recognizing the diagnostic gap,

* Corresponding author at: 777 Bannock Street, Mail Code 0108, Denver, CO 80204, USA.

E-mail address: austin.jones@denverem.org (A.T. Jones).

emergency department (ED) HCV screening programs are expanding to identify this population of asymptomatic carriers.

Emergency department HCV screening developed out of two de- cades of ED-based human immunodeficiency virus screening, a disease similar to HCV with high prevalence among the underserved population accessing the ED. [4] National Health and Nutrition Exami- nation Survey data demonstrate that uninsured HCV+ individuals use the ED more often than any other healthcare venue [5]. Similarly, dem- ographics with increased ED use – including non-white, Medicaid recip- ients, and uninsured individuals – are disproportionately affected by HCV [6]. Incorporating HCV screening in the ED utilizes existing health infrastructure to screen at-risk patients at a frequently accessed health venue.

https://doi.org/10.1016/j.ajem.2022.03.041

0735-6757/(C) 2022

EDs commonly adopt opt-out screening models, automatically test- ing all patients seeking care in the ED who meet pre-established param- eters. In 2013, the first opt-out screening program started at the University of Alabama at Birmingham. Targeting the birth cohort (those born between January 1, 1945 and December 31, 1965), HCV prevalence was 11.1% [7]. Highland Hospital Emergency Department in Oakland, CA followed in 2014, screening all patients in the birth co- hort or with a history of intravenous drug use (prevalence 10.3%) [8]. In 2015, University Medical Center in New Orleans, LA began similar risk-based HCV screening in the ED.

ED HCV screening programs are in their infancy and it is unknown whether they are more effective at linking HCV patients to care than the traditional community screening model. Community clinics report linkage to treatment rates of approximately 4% [9]. A multicenter evalu- ation of ED screening programs found rates of linkage to treatment equaling 8%, yet could not definitely prove greater success than the community model [10]. No study to date has conducted a direct com- parison of patient retention based on site of screening.

The objective of this study was to investigate whether patients screened HCV seropositive (HCV+) in the ED are linked to HCV care and retained in treatment more successfully than patients screened HCV+ in the community. We hypothesized that ED HCV screening pro- grams are more effective than their community counterparts at linking and retaining patients in follow-up care. As HCV-infected individuals represent a fragile population experiencing high rates of competing health priorities, homelessness, and drug addiction, ED screening pro- grams confer the advantage of referring positive patients to outpatient specialty services within the same healthcare setting. By localizing follow-up care around the ED, we hypothesized that ED screening pro- grams would result in improved retention and timeliness of care, as compared to the traditional community screening approach.

1. Methods
   1. Design

A retrospective cohort study was performed. Participants included all patients screened HCV+ at 12 screening facilities (11 community, 1 ED) in New Orleans, LA. The study was approved by the Tulane Uni- versity and Louisiana State University New Orleans Health Sciences Center institutional review boards.

• 1. Setting and population

The exposed group contained participants screened in the University Medical Center (UMC) ED from March 1, 2015 to July 31, 2017. The start of the enrollment window coincided with when the ED initiated an opt- out screening of patients with lifetime history of intravenous drug use or membership in the birth cohort. Nurses screened patients during tri- age. Patients who met the testing criteria were notified that they would receive a confidential HCV screening test, at no cost to the patient, un- less they wished to decline. For those who did not decline, an HCV anti- body test with quantitative HCV ribonucleic acid (RNA) polymerase chain reaction reflex order was placed. HCV testing in the ED was per- formed using the Abbott(TM) ARCHITECT(TM).

The non-exposed group included patients screened in a neighboring system of 11 community clinics, operated by Acacia NOLA, in New Orleans, LA from March 1, 2015 to May 31, 2018. Patients from the com- munity were included for an additional 10 months to achieve the sample size. Community sites included facilities serving people experiencing homelessness (Ozanam Inn, New Orleans Mission, UNITY of Greater New Orleans), community health centers (Ruth Fertel Total Health Clinic), drug rehabilitation facilities (Bethel Colony South, GRACE House, Bridge House, Behavioral Health Group Methadone Clinic), and community gathering sites (St. Anna’s Episcopal Church, Street Medicine, and Broadmoor Food Pantry). Using the OraQuick(R)

HCV Rapid Antibody Test, community sites performed opt-in testing, free of cost to the patient. Community testing and counseling were per- formed by medical student volunteers who were trained and certified by the New Orleans Office of Public Health. Patients received their test result in 20 min. At all 12 study sites, seropositive patients were con- nected with care coordinators who guided all subsequent HCV follow- up care (RNA confirmation to SVR12) at a single health facility (UMC). Coordinators scheduled clinic follow-up appointments on behalf of pa- tients, reminded patients of follow-up times, and arranged transporta- tion to appointments.

• 1. Data collection

The inclusion criterion was testing HCV antibody positive at one of the participating sites. Eligible participants were retrospectively identi- fied via an automated query of the electronic medical record (EMR) for all HCV seropositive patients screened by the participating sites within the aforementioned time frames. Participants were restricted to their First visit in the enrollment window. In the event that a participant had multiple unique screens, only the first visit was included for analy- sis in order to produce the most conservative estimate of time-to-event outcomes. In these cases, the site of the first HCV screening test was used to determine exposure status. Participants were excluded if they were actively undergoing HCV care (underwent fibrosis staging and awaiting HCV specialist appointment, currently managed by an HCV specialist for their HCV infection, currently taking HCV therapy, or fin- ished HCV therapy and awaiting confirmation of functional cure). Cure was defined as achieving sustained virologic response at 12 weeks (SVR12). This design enabled the capture of both newly identified HCV+ patients, without history of HCV care, as well as patients who had a history of engaging in HCV care but had fallen out of the care con- tinuum prior to SVR12.

All outcome variables and covariates were manually collected via retrospective review of the EMR from February 9, 2019 through January 31, 2021. Collection was performed systematically using a standardized collection instrument in Research Electronic Data Capture (REDCap) [11]. Selected variables were those standardly included in all patient charts to avoid missing data. Abstractors were blinded to the hypothesis to minimize abstractor bias. All abstractors were medical students with contextual knowledge of the EMR and health system, and in addition, provided extensive study specific training. Abstractors collected across all study sites preventing differences in collection by site. Entries were reviewed by the primary author for accuracy.

• 1. Outcomes

Outcomes were assessed using the HCV continuum of care, delineat- ing successive stages of care from screening to cure (Fig. 1) [10]. The

Antibody

RNA

Chronic Infection

Prior Spontaneous Clearance

No Follow-Up

Follow-Up

Testing

Cure

End Therapy

Start Therapy

Hepatitis Clinic

Fibrosis Staging

Fig. 1. The hepatitis C care continuum. The continuum serves as a Conceptual framework for how patients progress through the healthcare system from HCV screening to cure.

primary outcome was successful completion at each stage of HCV follow-up care, including determination of RNA positivity, fibrosis stat- ing, linkage to care, treatment initiation, treatment completion, and SVR12, each measured as a binary variable. The study was powered for retention at treatment initiation. RNA measurement was conducted using Cobas(R) AmpliPep/Cobas(R) TaqMan(R) HCV Test v2.0. Confirmation of chronic HCV infection occurred if the patient received a Blood draw for HCV RNA and the lab successfully processed the order. Fibrosis stag- ing occurred if the patient underwent FibroScan, FibroSure, hepatic ul- trasound, or liver biopsy with the indication of HCV. Linkage to care occurred if the patient attended a follow-up appointment with an infec- tious disease physician or gastroenterologist specifically for the man- agement of their HCV infection. Treatment initiation occurred if the physician prescribed HCV therapy and the patient was documented to have taken at least one dose of therapy. Treatment completion occurred with documentation that the full course of HCV had been completed. HCV cure was defined as achieving SVR12 following treatment comple- tion. A failure at this step indicated either failure to receive the confir- matory test or a result demonstrating persistent viremia.

For viral RNA confirmation, retention was a percentage of patients achieving RNA confirmation among those having screened HCV sero- positive. For subsequent steps in the care continuum, only patients with chronic HCV infection, measured by positive HCV RNA (>15 IU/ mL), were considered eligible to progress and be retained in care. There- fore, retention at linkage to care, treatment initiation, treatment com- pletion, and cure were a percentage of those found chronically infected by positive RNA.

The secondary outcomes were the time elapsed from HCV antibody screening to follow-up care. A time-to-event variable, this was a mea- sure of the time elapsed from the date of antibody screening to each subsequent stage of the HCV care continuum, including time to RNA confirmation, fibrosis staging, linkage to an HCV specialist, treatment initiation, treatment completion, and SVR12.

• 1. Analysis

The study was powered to retention at treatment initiation. Commu- nity clinics have reported linkage to treatment rates of approximately 4% [9], while a multicenter evaluation of ED HCV screening programs found rates of linkage to treatment equaling 8% [10]. To detect a dou- bling in linkage to care from 4 to 8%, n = 2942 participants were needed in a 10:1 ED:community allocation ratio totaling 2675 ED and 267 com- munity patients. The ratio was chosen to account for the differences in Patient throughput among the participating HCV screening programs. This sample size provided 80% power to detect a difference at ? =

0.05. The calculated sample size informed the window of time that pa-

tients would be enrolled retrospectively. To achieve this sample size, the first 29 months of patients (March 1, 2015 to July 31, 2017) from the UMC ED screening program were included and served as the ex- posed group. Patients found HCV+ in the community in the first 39 months of operation (March 1, 2015 to May 31, 2018) were included as the unexposed group.

Demographic variables collected were age, gender, race, insurance, and lifetime history of intravenous drug use. Variables detailing clinical history pre-HCV screen included history of prior HIV testing, currently taking highly active anti-retroviral therapy (HAART), and history of prior HCV testing. In bivariate analyses, ?² tests, t-tests, and Wilcoxon rank sum tests were used to compare baseline characteristics for cate- gorical variables, normally distributed continuous variables, and non- normally distributed continuous variables, respectively. Variables were eligible for inclusion in the model if differences were observed be- tween the two groups and found to be associated with the study out- comes, or were otherwise a known confounder.

Retention at each point in the care continuum was measured as a bi- nary variable. Generalized estimating equation log-binomial models were used for retention outcomes, accounting for clustering by site of

HCV screening. Empirical estimates produced adjusted risk ratios (RRs). Time to follow-up served as a time-to-event variable. Cox regres- sion models, also clustering by screening site, were used to assess time- to-event differences between both groups. Analysis was conducted using SAS Studio Software version 3.8 (SAS Institute, Cary, NC).

1. Results
   1. Cohort construction

A total of 4150 patients (n = 3776 ED and n = 374 community) were identified as having screened HCV+ using automated searches of the participating facilities’ electronic health records (Fig. 2). After matching subjects’ records at the screening and the follow-up health fa- cilities, restricting to patients’ first visit, and applying study eligibility criteria, 3330 (n = 3008 ED and n = 322 community) unique HCV+ pa- tients, not currently engaged in the care continuum, were analyzed. Pa- tients screened in the community setting were included in the following distribution: Ozanam Inn, n = 91 (28.3%); New Orleans Mission, n = 56 (17.4%); Bethel Colony South, n = 56 (17.4%); St. Anna’s Episcopal Church, n = 1 (0.3%); Grace House, n = 45 (14.0%); Ruth Fertel Total Health Clinic, n = 1 (0.3%); UNITY of Greater New Orleans, n = 2 (0.6%); Bridge House, n = 51 (15.8%); Behavioral Health Group Meth- adone Clinic, n = 14 (4.4%); Broadmoor Food Pantry, n = 3 (0.9%); and Street Medicine, n = 2 (0.6%).

• 1. Characteristics of study subjects

Baseline characteristics are shown in Table 1. Compared to patients screened HCV+ in the community, those screened in the ED were more likely to possess health insurance (81.5% vs. 75.8%; p = 0.01), identify as Black/African American (56.3% vs. 26.1%; p < 0.01), have a history of HIV testing (82.5% vs. 72.0%; p < 0.001), and be older (53 years vs. 39 years; p < 0.001). Patients in the ED were less likely to have previous HCV testing (28.0% vs. 37.0%; p < 0.001) and be male (73.6% vs. 78.9%; p = 0.04), while ED patients were equally likely to have a history of IVDU (42.1% vs. 46.6%; p = 0.12) compared to patients in the community.

• 1. Retention in care

Retention outcomes are depicted in Table 2. Risk models were ad- justed for insurance (binary), race (white/non-white), IVDU (binary), and HCV testing history (positive/negative/never tested). Age, gender, HIV history were not found to be associated with the outcomes and therefore ineligible to be confounders requiring adjustment in the model. Confirmatory RNA testing was performed in 96.2% of ED and 49.1% of community patients. This included laboratory failure to process RNA samples in 44 ED patients (38.3% of failed RNA retention) and no community patients. Those testing HCV+ in the ED were significantly more likely to complete RNA confirmation of chronic infection (ad- justed RR [aRR] = 1.91, 95% confidence interval [CI] = 1.54-2.37). Among patients completing RNA, chronic infection was established in 2154 ED (74.5%) and 132 clinic (83.5%) patients (p = 0.01).

Retention at subsequent linkage and treatment outcomes was mea- sured using the denominator of those chronically infected, as they were the only patients eligible to progress in the care continuum and restric- tion on chronic infection was necessary given the significant difference in persistent RNA between the two groups. Fibrosis staging occurred in 36.7% of ED and 60.6% of community patients. Compared to patients screened in the community, those testing HCV+ in the ED had a signif- icantly lower likelihood of fibrosis staging (aRR = 0.65; 95% CI = 0.51-0.82; p < 0.001). Linkage to care with an HCV specialist was initi- ated in 25.1% of ED and 23.5% of community patients. Patients screened HCV+ in the ED were no more likely to achieve linkage to care than pa- tients screened in the community (aRR = 1.03; 95% CI = 0.69-1.53;

Restriction to Unique Patients

Community Clinics N=374

Emergency Department N=3,776

Unique HCV+ patients not engaged in care

n=3,330

Currently in HCV care continuum Finished fibrosis staging (n=5) Managed by HCV provider (n=10)

Not HCV seropositive

HCV seronegative (n=136) HCV test did not result (n=23)

Exclusion

First visit (n=3,506)

Multiple visits Second test (n=478) Third test (n=111) Fourth test (n=28) Fifth test (n=12) Sixth test (n=5)

Assessed for eligibility (n=4,140)

Unmatched records (n=2)

Records Matched

Automated Search

Analysis

Fig. 2. Constructing a cohort of HCV seropositive patients.

p = 0.89). HCV therapy was started in 11.4% of ED patients and 4.6% of community patients. Compared to those screened in the community, patients testing HCV+ in the ED had a significantly higher likelihood of starting therapy (aRR = 2.23; 95% CI = 1.76-2.83; p < 0.001). HCV therapy was completed in 8.9% of ED patients and 4.6% of the commu- nity. Patients who tested HCV+ in the ED had significantly higher like- lihood of completing therapy compared to community patients (aRR = 1.77; 95% CI = 1.40-2.24; p < 0.001). Functional HCV cure was achieved in 7.2% of ED patients and 2.3% of clinic patients. Patients testing HCV+ in the ED also demonstrated a significantly higher likelihood of achiev- ing cure compared to community-screened individuals (aRR = 2.80, 95% CI = 1.09-7.23; p = 0.03).

• 1. Time to follow-up care

Secondary outcomes of time to follow-up are shown in Table 3. Fol- lowing the HCV screening test, RNA confirmation occurred, on average, the same day (median = 0 months) for ED patients and 0.9 months (IQR = 3.7 months) for community patients. Patients testing HCV+ in the ED achieved confirmation of their infection at a significantly faster rate than those screened in the community (adjusted hazard ratio [aHR] = 2.26; 95% CI = 1.86-2.72; p < 0.001). Time to fibrosis staging occurred 7.6 months following screening for ED patients and 2.0 months for community patients. Patients screened HCV+ in the ED had a signif- icantly slower rate of fibrosis staging than patients screened in the

Table 1

Characteristics of study participants (n = 3330)

finding have demonstrated that HCV+ patients screened at Federal Qualified Health Centers that provide both HCV screening and therapy

Variable, median (25th,75th percentiles)

Emergency Department (n = 3008)

Community Clinic

(n = 322)

P-value

were more likely to undergo a medical evaluation and receive fibrosis staging compared to patients who were screened at a location offering HCV screening alone [12]. ED screening programs refer positive patients

Age, years 53.0 (40.0, 59.0) 39.0 (31.0, 51.0) <0.001*

Variable n (%)

Male 2214 (73.6) 254 (78.9) 0.04*

Insurance <0.001*

Medicaid	1512 (50.3)	139 (43.2)
Medicare	511 (17.0)	31 (9.6)
Private insurance	212 (7.1)	62 (19.3)
Uninsured	555 (18.5)	78 (24.2)
Other	217 (7.2)	12 (3.7)
Race
Black or African American	1693 (56.3)	84 (26.1)	<0.001*
White	1073 (35.7)	206 (64.0)
Hispanic Latino	30 (1.0)	1 (0.3)
Asian	15 (0.5)	1 (0.3)
American Indian/Alaska Native	4 (0.1)	0 (0.0)
Native Hawaiian or Pacific Islander	5 (0.2)	0 (0.0)
More than one race	82 (2.7)	9 (2.8)
Not reported	105 (3.5)	21 (6.5)
History of intravenous drug use HIV status HIV-HCV coinfection	1267 (42.1) 85 (2.8)	150 (46.6) 2 (0.6)	0.12 <0.001*
HIV negative No previous HIV testing	2397 (79.7)	230 (71.4)
history HCV status Positive	525 (17.5) 790 (26.3)	90 (28.0) 94 (29.2)	<0.001*
Negative	52 (1.7)	25 (7.7)
Never tested	2165 (72.0)	203 (63.0)

community (aHR = 0.49; 95% CI = 0.38-0.63; p < 0.001). Following screening, rates of linkage to care with an HCV specialist, treatment ini- tiation, treatment completion, and cure were similar among patients screened in the ED and the community.

1. Discussion

Compared to patients diagnosed with HCV in the community, those diagnosed HCV+ in the emergency department were significantly more likely to receive RNA confirmation of HCV chronic infection, initiate HCV therapy, complete HCV therapy, and achieve HCV cure. ED screening was associated with a decreased likelihood of fibrosis staging and no dif- ference in linkage to HCV specialty care. In time to follow-up, RNA con- firmation occurred at faster rates in the ED, although these patients completed fibrosis staging at slower rates than their community coun- terparts.

We believe the success of ED screening can be attributed, in part, to co-localization of HCV care services in proximity to the ED. Previous

to outpatient specialty services within the same hospital. This forms a “one-stop shop” of HCV care, a network of care likely responsible for higher retention at follow-up appointments. We believe patients screened at community clinics have lower rates of retention as they re- quire an underserved population with limited healthcare utilization to navigate among multiple providers across distinct healthcare institu- tions. This imposes additional barriers on an underserved population, likely contributing to lack of follow-up. The geographic proximity of ser- vices around an ED screening program is likely leading to higher reten- tion of patients in care compared to their community-screened counterparts.

Furthermore, HCV screening in the ED benefits from reflex confirma- tory testing. Our findings demonstrate the reflex ordering of viral RNA in the ED significantly improved completion and timeliness of establish- ing HCV chronic infection. We believe decreased delays in confirmation may also be contributing to higher care retention following ED screen- ing. As the first stage of the HCV care continuum, same-day confirma- tion likely increased the momentum of ED patients and aided in improving retention at downstream stages of the care continuum. When patients had a confirmatory diagnosis and were thrust expedi- tiously into HCV evaluation, this inertia likely propelled them through the care continuum to achieve HCV cure. Expanding reflex ordering in the ED, beyond RNA confirmation, may produce additional benefits in retaining patients in HCV care.

This study has numerous strengths. While it has been demonstrated that ED screening programs detect high prevalence of seropositive pa- tients, limited data had been collected measuring the effectiveness of the emergency department at linking and retaining patients in follow- up care. In following over 3000 unique patients screened HCV+ in an ED over 29 months, this study serves as the largest known evaluation of retention in the HCV care continuum to date among all ED HCV screening programs. The study leveraged a unique healthcare delivery system where twelve screening facilities funneled into one care contin- uum. The study took place in the first 29 months of ED HCV screening, maximizing a newly captured population and minimizing patients with previous referrals to care. While differential loss to follow-up is a primary limitation to most cohort studies, retention in HCV care is the study outcome. This design, therefore, removed loss to follow-up as a threat to the study’s validity. The majority of studies to date evaluating ED HCV screening did not assess outcomes beyond linkage to care. Our study followed the entire HCV care continuum and included fibrosis staging, treatment start, treatment end, and cure. For these distant out- comes, the study’s multi-year observation period enabled the measure- ment of rare outcomes with long induction periods. Our rates of retention for ED and community programs are similar to those previ- ously described in the literature [10]; therefore, we believe our findings

Table 2

Retention in HCV follow-up care for patients screened in emergency departments and community clinics. Completion of RNA was measured among all HCV seropositive patients while RNA processed was measured among those with a successful RNA draw. Completion of all subsequent steps in care was measured among those identified as chronically infected with HCV. All models were adjusted for white race (binary), lifetime history of IVDU (binary), insurance (binary), and previous HCV testing history (multilevel: positive, negative, never tested)

Care Continuum	Emergency Department	Community Clinic	Adjusted Risk Ratio	P-value
	N = 3008	N = 322	[95% Confidence Interval]
	n (%)	n (%)
RNA drawn	2936 (97.6)	158 (49.1)	1.94 [1.56-2.40]	<0.001*
RNA processed	2892 (98.5)	158 (100)	–	–
Chronic infection	2153 (74.4)	132 (83.5)	–	–
Fibrosis staging	791 (36.7)	80 (60.6)	0.64 [0.51-0.81]	<0.001*
Linkage to care	544 (25.3)	31 (23.5)	1.03 [0.69-1.53]	0.90
Treatment initiation	233 (10.8)	5 (3.8)	2.52 [1.79-3.56]	<0.001*
Treatment completion	185 (8.6)	5 (3.8)	2.06 [1.46-2.90]	<0.001*
SVR12	150 (7.0)	3 (2.3)	2.73 [1.06-7.02]	0.037*

Table 3

Time to HCV follow-up care for patients screened in the emergency department and community clinics. All models were adjusted for white race (binary), lifetime history of IVDU (binary), insurance (binary), and previous HCV testing history (multilevel: positive, negative, never tested)

Care Continuum	Emergency Department Months Median (25th, 75th percentiles)	Community Clinic Months Median (25th, 75th percentiles)	Adjusted Hazard Ratio [95% Confidence Interval]	P-value
RNA	0 (0, 0)	0.87 (0.3, 3.9)	2.29 [1.91-2.76]	<0.001*
Fibrosis staging	7.6 (2.9, 22.6)	2.0 (1.2, 5.4)	0.49 [0.38-0.62]	<0.001*
Linkage to care	5.1 (3.0, 11.7)	6.2 (4.7, 9.5)	1.01 [0.69-1.48]	0.97
Treatment initiation	16.6 (8.8, 31.8)	17.5 (14.0, 29.1)	1.01 [0.41-2.48]	0.98
Treatment completion	19.3 (12.0, 33.2)	20.3 (15.9, 31.9)	0.99 [0.40-2.43]	0.98
SVR12	24.5 (16.3, 36.5)	23.8 (22.2, 32.6)	1.12 [0.35-3.54]	0.85

are likely generalizable to the growing number of ED screening pro- grams nationwide.

The study findings have high pragmatic significance. ED HCV pro- grams are new and very few studies have looked at care outcomes of pa- tients tested HCV+ in the ED upon being discharged. This study serves as the first direct comparison of outcomes between emergency depart- ment and community HCV screening programs. Our data provides new evidence that EDs may be the most effective setting to screen patients to promote follow-up care. This study also provides a proof of concept for the role of EDs in infectious disease surveillance and linkage to care. ED screening has been effective with HIV and evidence is building for its utility for HCV. With advances in point of care testing, this model can ex- pand to other diseases of public health import, including sexually trans- mitted infections as well as non-communicable diseases. This study further shows that EDs are an Ideal setting to engage and deliver ser- vices to patients who are hard to reach and who do not engage the healthcare system at more traditional venues. Additional interventional studies are needed to determine the most effective setting to screen for hepatitis C to maximize patient outcomes.

• 1. Limitations

Given the retrospective nature of this study, there are a few limita- tions. Data was collected by retrospective review of the EMR. There is a possibility that baseline information in the medical record may have changed from enrollment to time of data collection. As is common in the EMR, data may be subject to updating and overwriting. For example, insurance status at time of chart review may be different than insurance status at time of enrollment. In these instances, every effort was made to collect demographic data on, or as close to, the point of enrollment (HCV Antibody testing). An additional limitation to EMR data is that variables collected were limited to those documented by Healthcare personnel in the chart. As a result, there may be additional, unrecognized differences between the exposed and unexposed groups that were unable to be measured and subsequently adjusted for in analysis. The selected vari- ables for the study were those commonly documented in the EMR lead- ing to no missing data for the collected variables. Patients may have also sought care at other medical institutions. While we utilized EMR sharing agreements with partnering hospitals, if evidence of follow-up was not found in the EMR, it was documented in our study as failed retention. Low retention at treatment initiation, treatment completion, and cure may have yielded unstable estimates for treatment and cure. Given the high volume of EDs, those screened in the community were a smaller proportion of the study sample.

• 1. Conclusions

In conclusion, compared to community screening, ED-based HCV screening was associated with significantly higher confirmation of HCV infection, treatment initiation/completion, and HCV cure. Our find- ings provide new evidence supporting the role of emergency depart- ments in promoting retention in HCV follow-up care. Expanding ED-based HCV screening may serve as the most effective screening

strategy and a viable solution to close the profound hepatitis C treat- ment gap.

Prior presentations

Results were delivered by oral presentation at the American Acad- emy of Emergency Medicine Scientific Assembly in St. Louis, MO (June 5, 2021).

Funding sources and disclosures

This work was supported by the National Center for Advancing Translational Sciences [5TL1TR001418 to A.T.J.] and the Emergency Medicine Foundation/Society of Academic Emergency Medicine Foun- dation Medical Student Research Grant [A.T.J.]. Gilead Frontlines of Communities in the United States (FOCUS) grants sponsored hepatitis C screening at all participating sites. The authors declare no competing interests.

CRediT authorship contribution statement Austin T. Jones: Writing – review & editing, Writing – original draft,

Validation, Project administration, Methodology, Funding acquisition, Formal analysis, Data curation, Conceptualization, Investigation. Lisa Moreno-Walton: Conceptualization, Funding acquisition, Investigation, Supervision, Writing – review & editing. Torrence Tran: Writing – review & editing, Project administration, Data curation. Christopher Briones: Data curation, Writing – review & editing. Kanayo Okeke- Eweni: Writing – review & editing, Validation, Supervision, Resources, Project administration. Rachael Stevens: Data curation, Writing – review & editing. Kacie Isaacson: Writing – review & editing, Data curation. Hua He: Formal analysis, Methodology, Writing – review & editing. Patricia J. Kissinger: Writing – review & editing, Supervision, Methodology, Investigation, Conceptualization.

Declaration of Competing Interest

None.

Acknowledgements

The authors would like to thank Alarica Dietzen, Alexander Jafari, Jenna Miller, Maya Mahendran, Mandy Majidian, Rachel Turner, and Scott Sabo for their efforts in data collection. The authors thank all per- sonnel responsible for carrying out HCV screening in the University Medical Center Emergency Department and Acacia NOLA.

References

1. Ly KN, Hughes EM, Jiles RB, Holmberg SD. Rising mortality associated with hepatitis C virus in the United States, 2003-2013. Clin Infect Dis. 2016;62:1287-8.
2. Terrault NA. Hepatitis C elimination: challenges with under-diagnosis and under- treatment. F1000Research. 2019;8:1-11.
3. Denniston MM, Klevens RM, McQuillan GM, Jiles RB. Awareness of infection, knowl- edge of hepatitis C, and medical follow-up among individuals testing positive for hepatitis C: National Health and nutrition examination survey 2001-2008. Hepatology. 2012;55:1652-61.
4. Galbraith JW. Hepatitis C virus screening: an important public health opportunity for United States emergency departments. Ann Emerg Med. 2016;67(129-130). https:// doi.org/10.1016/j.annemergmed.2015.08.002.
5. Stepanova M, Kanwal F, El-Serag HB, Younossi ZM. Insurance status and treatment candidacy of hepatitis C patients: analysis of population-based data from the United States. Hepatology. 2011;53:737-45.
6. Tsui JI, Maselli J, Gonzales R. Sociodemographic trends in national ambulatory care visits for hepatitis C virus infection. Dig Dis Sci. 2009;54:2694-8.
7. Galbraith JW, Franco RA, Donnelly JP, et al. Unrecognized chronic hepatitis C virus infection among Baby boomers in the emergency department. Hepatology. 2015; 61:776-82.
8. White DAE, Anderson ES, Pfeil SK, Trivedi TK, Alter HJ. Results of a rapid hepatitis C virus screening and diagnostic testing program in an urban emergency department.

Ann Emerg Med. 2016;67(119-128). https://doi.org/10.1016/j.annemergmed.2015. 06.023.

1. Norton BL, Southern WN, Steinman M, et al. No differences in achieving hepatitis C virus care milestones between patients identified by birth cohort or risk-based screening. Clin Gastroenterol Hepatol. 2016;14(1356-1360). https://doi.org/10. 1016/j.cgh.2016.04.017.
2. Anderson ES, Galbraith JW, Deering LJ, et al. Continuum of care for hepatitis C virus among patients diagnosed in the emergency department setting. Clin Infect Dis. 2017;64:1540-6.
3. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)-A metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42 (377-381). https://doi.org/10.1016/j.jbi.2008.08.010.
4. Coyle C, Moorman AC, Bartholomew T, et al. The hepatitis C virus care continuum: linkage to hepatitis C virus care and treatment among patients at an urban health network, Philadelphia. PA Hepatol. 2019;0:1-11.