VirA+EmiC project: Evaluating real‐world effectiveness and sustainability of integrated routine opportunistic hepatitis B and C testing in a large urban emergency department

Abstract Innovative testing approaches and care pathways are required to meet global hepatitis B virus (HBV) and hepatitis C virus (HCV) elimination goals. Routine blood‐borne virus (BBV) testing in emergency departments (EDs) in high‐prevalence areas is suggested by the European Centre for Disease Prevention and Control (ECDC) but there is limited evidence for this. Universal HIV testing in our ED according to UK guidance has been operational since 2015. We conducted a real‐world service evaluation of a modified electronic patient record (EPR) system to include opportunistic opt‐out HBV/reflex‐HCV tests for any routine blood test orders for ED attendees aged ≥16 years. Reactive laboratory results were communicated directly to specialist clinical teams. Our model for contacting patients requiring linkage to care (new diagnoses/known but disengaged) evolved from initially primarily hospital‐led to collaborating with regional health and community service networks. Over 11 months, 81,088 patients attended the ED; 36,865 (45.5%) had a blood test. Overall uptake for both HBV and HCV testing was 75%. Seroprevalence was 0.9% for hepatitis B surface antigen (HBsAg) and 0.9% for HCV antigen (HCV‐Ag). 79% of 140 successfully contacted HBsAg+patients required linkage to care, of which 87% engaged. 76% of 130 contactable HCV‐Ag+patients required linkage, 52% engaged. Our results demonstrate effectiveness and sustainability of universal ED EPR opt‐out HBV/HCV testing combined with comprehensive linkage to care pathways, allowing care provision particularly for marginalized at‐risk groups with limited healthcare access. The findings support the ECDC BBV testing guidance and may inform future UK hepatitis testing guidance.


| INTRODUC TI ON
There is now broad access to efficacious treatments that reduce both mortality and morbidity for both viral infections, 3,4 yet key challenges remain. 5 This is primarily because many infections remain undiagnosed: in 2018, around two-thirds of chronic HCV infections were estimated to be undiagnosed. 6 81% of HBV infections in the UK were estimated to be undiagnosed in 2016. 2 The 2016 World Health Organization (WHO) Global Health Sector Strategy on Viral Hepatitis prioritized the need for innovative testing strategies and efficient linkage to care pathways with the aim of eliminating viral hepatitis as a major public health threat by 2030. 5 Routine blood-borne virus (BBV) testing in emergency departments (EDs) in high-prevalence areas is suggested by the European Centre for Disease Prevention and Control (ECDC) but with concession that evidence for its effectiveness is lacking. 7 In England, there were >24.8 million attendances at EDs in 2018-2019. 8 For often marginalized populations at high risk of BBVs, the ED is often a key healthcare access point. 9,10 Findings from pilot ED opt-out viral testing studies and recent multicentre UK ED seroprevalence surveys 11,12 suggested a consistent, high level of active undiagnosed BBV infections, much higher than at general population level. 13,14 UK guidance from the National Institute for Health and Care Excellence (NICE) recommends that, in high-prevalence areas, human immunodeficiency virus (HIV) testing is universally offered to adult ED attendees who are undergoing blood tests for another healthcare reason. 15 However, in contrast, no similar consideration exists in current UK hepatitis testing guidelines. 16 In 2014, a 1-week opt-out ED testing campaign in nine UK EDs showed that this type of testing was feasible and effective. 17 Subsequently, pilot initiatives in London (UK), 18,19 and Dublin (Ireland) 20 have shown success using opt-out ED blood screening for HBV and HCV in identifying undiagnosed patients or those lost to follow-up and linking them to care. Previously we evaluated a 6-week pilot study in our centre, which demonstrated success of our model in a large inner-city ED serving a diverse population with high levels of social deprivation. 21 This current paper describes an extension of the model to evaluate its realworld sustainability. In particular, we discuss the challenges that were faced when integrating sustainable effective linkage to care into clinical routine, and methods that we employed to overcome these challenges.

| Study design, setting and participants
We conducted a clinical service development evaluation at the ED of Guy's and St Thomas' NHS Foundation Trust in London, UK. The project was conducted in two phases: phase 1 lasted 5 months (October 2016-February 2017) and phase 2 lasted 6 months (December 2017-May 2018). In contrast to traditional healthcare professionalinitiated testing models, the electronic patient record (EPR) system was modified to pre-select HBV and HCV tests for any blood test orders in the ED for patients ≥16 years on an opt-out basis, unless a positive viral hepatitis test result within the previous 6 months was available on the EPR system. Multilingual information leaflets were handed out to patients at registration and provided by the clinician, posters were also available throughout the ED. These leaflets included disease awareness, information on testing policy and process, and how to access care for either infection. Additionally, all eligible patients were verbally informed before blood draw that viral hepatitis tests would be carried out unless they declined (opted-out). Staff in the ED department were regularly trained by senior clinicians.
This project was a service development evaluation (http://www.hradecis ionto ols.org.uk/resea rch/) and did not require further ethical

| Laboratory methods
Laboratory testing was carried out according to routine local protocol. Current HBV infection was diagnosed upon detection of hepatitis B surface antigen (HBsAg) (ARCHITECT; Abbott Laboratories).
If HBsAg was reactive on screening assay, specificity was confirmed with a neutralization step, and reflex testing of hepatitis B core immunoglobulin (Ig)G, IgM antibodies and e-markers. HCV testing comprised an initial antibody screening test (HCV-Ab) (ARCHITECT; Abbott Laboratories), followed by same-sample reflex-HCV antigen (HCV-Ag) (ARCHITECT; Abbott Laboratories) testing for positive or equivocal initial HCV-Ab results. For negative or equivocal HCV-Ag tests, a second HCV-Ab test (Bio-Rad Inc.) was performed to confirm the presence of HCV antibodies detected in the initial screening assay. Whenever the HCV-Ab or HBsAg result had been previously demonstrated using this algorithm, any subsequent sample from the same individual was only tested using the screening test and reported as a confirmed result. HIV testing as an opt-out test for the ED attendees had already integrated into routine ED testing practice in our hospital since 2015; this is reported separately 22 and, therefore, data regarding HIV testing outcome were not reported for this service evaluation.

| Follow-up/linkage to care
Patients testing positive for HBsAg and HCV-Ag had their EPR checked by the study investigators. If there was documentation indicating that the patient's result reflected a known diagnosis who were already engaged in care, no further action was taken.
Patients who were HCV-Ab positive but HCV-Ag negative (and for whom no prior local HCV RNA status was found) were asked by letter to attend their general practitioner (GP) for a follow-up HCV RNA sample to exclude current infection. As a linkage to care pathway already existed for patients with HIV infection, no further action was taken for patients with dual infection with HIV and either HBV or HCV.
Patients with a confirmed infection who were still alive at point of review were considered eligible for contact (see below). Among patients who were informed of their diagnosis (successfully contacted), those who were either newly diagnosed or known but disengaged with care, or those whose status had not been clarified (unknown) during the initial contact, were classed as 'requiring linkage to care'.
If the patient attended at least one appointment, the outcome of the linkage to care was recorded as 'engaged with care'. During the initial appointment, patients underwent assessment in accordance with the current UK clinical guidance. 4,23 Patients who were lost to follow-up at each stage of the linkage pathway were described.
Treatments offered for HCV and HBV were in accordance with the NICE and NHS England guidelines. 4,23

| Evolution of contacting and engaging patients during project
Linkage to care approaches were under constant review and evolved throughout the service evaluation. At the beginning of the project, the linkage to care pathway included a direct telephone call to the patient informing them of diagnosis, care status assessment and of-

| Data collection and management
A dataset that included all ED attendees aged ≥16 years, those with routine blood tests and those who were tested for HBsAg and/or HCV-Ab during the study period was extracted from the EPR system. This dataset included demographic information, including age, sex, ethnicity and residence type (including hostel, no fixed abode). For patients testing positive for HBsAg and/or HCV-Ag, a separate linkage to care dataset was also compiled, which included contact, diagnosis and linkage to care outcomes.
Each patient was assigned a unique patient identifier so that datasets could be collated and deduplicated. Data were handled in accordance with information governance policies to ensure patient confidentiality. Only staff with direct patient care responsibilities had access to patient-identifiable data. Equivocal test results for HBsAg or HCV-Ab with a negative HCV-Ag result were recorded as negative. Unsuitable samples or samples that were not received in the laboratory were recorded as 'not tested'. For patients who attended the ED multiple times during the study period, the earliest record was retained and updated with additional information from subsequent records, unless there was a later attendance linked with a positive test result. HIV status was recorded as negative, positive or unknown. Age was recorded as a categorical variable (16-29, 30-49, 50-69, ≥70 years) and ethnicity was classified into six categories, in line with UK census categories. ED arrival was categorized as day (08:00-19:59) or night (20:00-07:59) and also by arrival day (weekday: Monday to Friday, weekend: Saturday or Sunday).

| Statistical analysis
Continuous variables were described using medians with interquartile ranges (IQRs). Categorical variables were described as absolute and relative frequencies. Overall and stratified seroprevalence estimates with their 95% CI were calculated for the HCV-Ag and HBsAg tests. To investigate factors associated with requiring a HCV or HBV test and for being seropositive, we applied univariable and multivariable Poisson regression models to calculate prevalence ratios (PRs).
All tests were two sided with 95% CI; the level of significance was set at p < 0.05. All analyses were performed using STATA version 15 software (StataCorp LP).

| RE SULTS
During the 11-month service evaluation period, 81,088 patients attended the ED and, of those, 36,865 (45.5%) received a blood test.
There was an approximately equal distribution of female to male ED participants (51% female); 37% were of White British ethnicity, 19% White other and 18% Black/Black British. The median (IQR) age was 44 (30-61) years. Figure 1 shows the numbers and proportions of patients prospectively screened who received the HBsAg, HCV-Ab and HCV-Ag tests, and the numbers and proportions of patients testing positive for each.
Testing uptake was relatively constant during the two phases of the  Figure S1). Test uptake was consistently >72% (range 72.6-77.3%) across strata (sex, age, ethnicity, homeless status, ED arrival day and ED arrival time).
With univariable testing (Table S1) (Table S1), all of the above-mentioned factors associated with testing positive for HBsAg remained statistically significant. Seropositivity according to study period and age group are shown graphically in Figure S1.

Hepatitis B: linkage to care
Among the 235 patients who tested positive for HBsAg, 36 (15%) did not need any further follow-up as local care records showed that Seroprevalence of HCV-Ag was highest among homeless people (14.7%) and HIV-positive individuals (7.2%). In multivariable regression models (Table S2), factors significantly associated (p < 0.01)   Of the 251 patients who tested positive for HCV-Ag, 3

| NHS Find and Treat subset (these numbers are included in the above results)
The

| Overall summary
To our knowledge, this 11-month service evaluation of the realworld sustainability of ED EPR opt-out HCV and HBV testing is the largest intervention of its kind in the UK to date. It showed that optout ED testing is practical in a large urban hospital setting and allows identification and linkage to care of significant numbers of active HBV and HCV infected individuals from marginalized groups that have historically had poor access to healthcare. In the current study, seroprevalence for HBsAg was 0.9% and for HCV-Ag was 0.9%. The strata specific results mirror those seen in our previous 6-week pilot study, showing the high burden of disease in central London. 21 The demographic characteristics described here are in line with those reported previously in similar studies in UK. 25,26 It is noteworthy that prevalence of HCV-Ag among patients aged 50-69 years was significantly higher than among those aged 16-49 years. This highlights a potential benefit of ED opt-out testing: it captures the older patients who may no longer be attending services for people who inject drugs. Unfortunately history of drug use was not documented systematically in patients' notes and, therefore, formal analysis on this variable was not carried out.
Seroprevalence of both HCV-Ab and HCV-Ag appears to decrease over time ( Figure S2). This could be affected by several factors including seasonal variation in attendance patterns and/or be an artefact of EPR testing: the system avoided repeat testing within 6 months of a positive result. We, therefore, feel any interpretation would be speculative.
Interestingly, HBsAg prevalence in the current study was almost twice as high as in the pilot, reasons for this are unknown but might be chance-related due to the smaller sample size in the pilot phase.
In the current study, 141/235 patients (60%) who were positive for HBsAg were men which might suggest that ED testing could be of particular benefit in identifying men with HBV, whereas many HBV infections in women are detected in prenatal screening. 25,26 Importantly, opt-out ED testing provides an opportunity not only to identify new infections but also to re-engage patients previously diagnosed but not linked to care, the latter accounting for around a quarter of HBV and half of HCV patients requiring linkage to care in our sample.

| Testing in the ED
The testing uptake in the current study was 75% for both HBsAg and HCV-Ab over the 11-month evaluation period, demonstrating that EPR modification to create an automatic BBV order set and optout testing achieved consistently higher uptake compared with more traditional testing approaches. [17][18][19] HCV-Ag testing allows identification of current HCV infection and is standard of care throughout our hospital. HCV-Ag testing is cheaper than RNA-based testing, is less complex for the laboratory and has a more rapid turnaround time. 27

| Application in other settings
Routine, opt-out BBV testing has been shown to be feasible in multiple hospitals serving large UK cities. [17][18][19][20][21][22] A service of this kind is likely to be most useful in urban areas, which often have populations with high incidence and prevalence of BBVs. We believe that, while this kind of testing could be implemented in almost any ED with an EPR system amenable to modification to include automatic requests, the integration of a linkage to care pathway requires good coordination between ED, clinical treatment and community teams. EPR modification can be applied to in-patient and out-patient settings as well, but EDs have the advantage of being accessible to all and, therefore, provide an unfiltered population.

| Cost-effectiveness
A previous initial economic evaluation using pilot study data suggested that ED testing for HBV and HCV in the UK is likely to be cost-effective at a viral prevalence of 0.5% and above. 33 The prevalence for both infections observed in the current study are well above this threshold. Ongoing work, using data from this and other large-scale sustainability evaluations is looking at estimating the cost-effectiveness of ED testing under real-world conditions.

| CON CLUS ION
Our results demonstrate the effectiveness and sustainability of universal urban ED opt-out HBV and HCV testing combined with integrated linkage to care pathways embracing secondary, primary and community care. Utilization of an EPR supported testing approach achieved sustainable high test uptake and our comprehensive linkage network better connecting existing services allowed care provision to traditionally poorly served, at-risk groups that would not otherwise be possible without a large increase in health infrastructure.

AUTH O R CO NTR I B UTI O N S
All authors have made substantial contributions to conception and design, or acquisition of data, or analysis and interpretation of data; have been involved in drafting the manuscript or revising it critically for important intellectual content; and have given final approval of the version to be published. Each author has participated sufficiently in the work to take public responsibility for appropriate portions of the content and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available upon request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.