Research and analysis

Emergency department bloodborne virus opt-out testing: 12-month interim report 2023

Published 9 November 2023

Applies to England

Public health and implementation evaluation from April 2022 to March 2023.

Executive summary

In April 2022, implementation began of an NHS England (NHSE) funded programme of opt-out testing for bloodborne viruses (BBV) in emergency departments (EDs) in areas of very high diagnosed HIV prevalence (more than 5 per 1,000 people aged 15 to 59 years). According to NHSE programmatic data, across 33 EDs 857,117 HIV tests, 473,723 hepatitis C virus (HCV) tests and 366,722 hepatitis B virus (HBV) tests were done during the first year. The scale of the programme makes it a substantial contribution to all BBV testing in England, with the number of tests done equivalent to more than half of all BBV tests done in general practice (GP), sexual health, drug and prison services combined (reported through the UK Health Security Agency (UKHSA) sentinel surveillance of BBV testing).

In this report, UKHSA provides an interim public health evaluation of the first 12 months of the programme, and an interim implementation optimisation evaluation, conducted by the University of Bristol. Using linked data sets, analysis was undertaken of 16 of 33 sites to understand the characteristics and numbers of people newly and previously diagnosed, and the extent both groups were promptly linked to care. A deep dive of 5 of 33 sites was undertaken to investigate test uptake and positivity.

The ED population who had blood tests as part of their ED attendance and were therefore eligible for BBV testing in the programme, tended to be older compared to other settings where BBV tests are undertaken. Although approximately half of all eligible attendees were tested for BBVs, testing uptake varied considerably between the 5 sites included in the deep dive, ranging from 22% to 74%. There was little difference in test uptake in relation to ethnicity, sex and age group within sites. Achieving universally high rates of testing across the programme is needed to ensure that between-site variations do not cause inequalities in offer of and access to testing.

The highest proportion of new diagnoses and highest positivity was for HBV; however, pathways into care were more robust for HIV and HCV. Mapping and optimising care pathways for HBV is needed to ensure that the needs of those newly diagnosed with HBV through the programme are met. Linkage to care was sub-optimal for all 3 BBVs, and it will be important for ICBs, commissioners and service providers to reflect upon the different needs of individuals diagnosed in EDs.

The implementation optimisation evaluation found variation in sites standard operating procedures (SOPs) on how opt-out testing was being implemented. Continuing to support sites to implement good practice processes and procedures will be essential to the effectiveness of the programme and could enhance equity in offer and access to testing.

There were limitations in the coverage and completeness of surveillance data reported to UKHSA for this evaluation, which need to be addressed going forward. Nevertheless, the availability of both programmatic and surveillance data has provided a rare opportunity to triangulate data between sources and has been beneficial in providing a starting point to work towards filling the gaps in the available surveillance data for BBV tests and diagnoses.

Main messages

The Public health and implementation evaluation 12-month interim report has identified 13 main messages. These messages are explored in more detail in the main findings section of the report.

1. Over its first year, the NHS England (NHSE) funded opt-out bloodborne virus (BBV) testing in emergency departments (ED) programme (the programme) was rolled out across 33 sites selected based on being in areas of very high diagnosed prevalence for HIV. In programme reported data, there were 857,117 HIV tests, 473,723 HCV tests and 366,722 HBV tests undertaken by participating EDs during the first year, with roll out and ‘go live’ dates for each BBV varying by site. The programme makes a substantial contribution to all BBV testing in England. The number of tests done in the programme is equivalent to more than half of all BBV tests done in general practice (GP), sexual health, drug and prison services combined (reported to UKHSA sentinel surveillance of BBV testing (SSBBV).

2. The public health evaluation aims to evaluate the impact and effectiveness of the programme and assess any potential inequalities in uptake of BBV testing and linkage to care. This cannot be done with programmatic data alone, which does not collect all relevant person characteristics and outcome information. This interim evaluation report of the first 12 months of the programme therefore combines programmatic data with data from UKHSA SSBBV, a voluntary sub-national surveillance system representing 16 out of 33 ED sites, 15 of which are London based. A deep dive analysis is done for 5 out of 33 sites with complete surveillance data for both positive and negative tests (all London based) to estimate test uptake and positivity. Future evaluation reports will include increased site coverage of SSBBV.

3. Across all sites, the population eligible to be tested by the programme due to having a blood test as part of their ED attendance was older than the population tested for BBVs in other settings. The ethnic distribution of eligible attendees varied by geography and reflected local population demographics.

4. At the 5 sites selected for a deep dive analysis, half of eligible attendees had an HIV test and this was similar for HCV and HBV (taking into account ‘go live’ dates). Within sites, BBV testing was broadly equitable by age, sex, ethnicity and IMD but was lower in the youngest and oldest age groups. However, the proportion tested at the 5 deep dive sites varied considerably by site, between 22% to 74%.

5. The highest number and proportion of new diagnoses was for HBV, reflecting the higher prevalence of people living with undiagnosed HBV compared to HIV and HCV. There has been substantial effort and financial commitment to increase diagnosis and linkage to care for HIV and HCV in recent years, which needs to be replicated for HBV if the World Health Organization (WHO) elimination targets by 2030 are to be reached.

6. Across the 16 sites served by laboratories that report through the SSBBV system, there were a total of 646 new HBV diagnoses between April 2022 and March 2023. In the same time period, there were 175 new HCV diagnoses and 141 people previously diagnosed with HCV and not in care, and 78 new HIV diagnoses and 59 people previously diagnosed with HIV and not in care. It is noted that new HIV diagnoses, identified in the HIV and AIDS Reporting System (HARS) and HIV and AIDS new diagnoses database (HANDD), could be identified only up to the end of 2022.

7. At the 5 sites selected for a deep dive analysis (all within London), test positivity (including people with known infections) was 1.1% for HBV (HBsAg), 0.9% for HIV and 0.2% for HCV (ribonucleic acid (RNA)). For all 3 BBVs, new diagnoses were higher among men. New diagnoses for HIV and HBV were higher among people aged 35 to 64 years and aged 50 to 80 years for HCV. For HIV and HBV, the proportion of new diagnoses among those tested was highest among people of black African ethnicity, and for HCV they were highest among people of white other ethnicity.

8. At the 16 SSBBV sites, the rate of linkage to care for people newly diagnosed with HIV or HCV was lower than reported in national data, demonstrating the additional difficulties in linking to care from ED, which may include challenges in contacting individuals due to incomplete details, being homeless or not registered to a current local address. Linkage to care was lower for HCV and HBV than for HIV, which mirrors findings from other ED studies and national surveillance. For HIV, linkage to care was considerably lower among people previously diagnosed and not in care compared to people newly diagnosed.

9. It was feasible to produce indicators on testing and diagnosis from surveillance data for all 3 BBVs. However, linkage to care for HBV cannot currently be monitored because there is currently no large-scale HBV treatment data set available to monitor HBV care. Surveillance data for BBV tests and diagnoses is sub-national with only 16 out of 33 sites in the programme reporting to SSBBV. Due to data incompleteness for negative tests, only 5 out of 33 had complete data for both positive and negative tests to enable analysis of test uptake and positivity.

10. Whilst most responding sites had SOPs or guidance documents (28 out of 33 sites), there was variability in opt-out testing strategies at each site, and some procedures differed from good practice guidance. SOPs and guidance documents were primarily HIV focused, with HCV mentioned in 13 out of 28 sites and HBV mentioned in 7 out of 28.

11. All SOPs from sites suggested that testing was delivered using an opt-out model although there was some variation in how opt out testing was implemented. All sites used posters and banners to inform individuals about testing and how to opt out, with some sites additionally recommended that staff provide a verbal reminder or signposting individuals to the ED opt-out testing programme.

12. Most sites have an automated system for adding a BBV test to the ED blood set, although a small number of sites required staff to manually order tests. In the public health analysis, testing rates were higher in sites that had automated ordering of BBV tests.

13. Sites used a ‘no news is good news’ approach to results management. If the individual had not opted-out but no BBV test was performed, the SOPs had no mention of procedures to highlight this to individuals.

Interim recommendations

Delivery of testing

1. Develop and implement SOPs for opt-out testing for all BBVs if these are not yet in place (including opt-out approach, documentation, communication of results and linkage to care for positive results) and ensure ED staff are fully briefed (sites).

2. Adopt opt-out procedures recommended as good practice, using verbal prompts where appropriate (for example visually impaired, English not first language) (sites).

3. Continue to work with electronic patient record (EPR) teams in sites that do not yet have automated test ordering in place to replicate approaches taken by other sites with high uptake (sites, NHSE and Integrated Care Boards (ICBs)).

4. Develop procedures to contact individuals in the event of insufficient blood samples, and to inform individuals if no BBV test has been performed (sites).

5. Continue to work with sites with low test uptake to understand barriers to testing and to facilitate higher testing rates (NHSE and ICBs).

Linkage to care

1. Map and optimise care pathways for people newly diagnosed with HBV in ED as part of roll out of ED testing to new sites (ICBs and hepatology teams).

2. Continue to improve linkage to care from ED by identifying the needs of individuals diagnosed in ED and structural facilitators to linkage to care, including additional interventions such as community support (NHSE, ICBs, UKHSA and the University of Bristol).

3. Continue to share learning from different care pathways used within the programme (NHSE, UKHSA and the University of Bristol).

Evaluation and surveillance

1. Collaborate to increase recruitment of laboratories to SSBBV to improve representative coverage across sites, including those outside London (NHSE, UKHSA, ICBs and sites).

2. Work with laboratories to understand and address data incompleteness and recording of ED test setting in SSBBV surveillance data (UKHSA, NHSE and laboratories).

3. Investigate ways to identify confirmatory testing for HIV in SSBBV surveillance data (UKHSA).

4. Undertake a deep dive with selected sites to understand why some people with positive HIV results are not matching to HARS and are not categorised as ED test setting (UKHSA and sites).

5. Work with HIV and sexual health clinics to improve recording of first site of HIV test to better understand the extent of ED testing nationally (UKHSA and clinics).

6. Encourage sites to work with UKHSA on monitoring HBV linkage to care (NHSE, ICBs, sites and UKHSA).

Background

Public health relevance

The national HIV Action Plan, published in 2021 by the Department of Health and Social Care (DHSC), commits to ending HIV transmissions in England by 2030, and has an interim target of reducing HIV transmissions by 80% between 2019 and 2025 (1). To achieve this, HIV testing needs to be scaled up to reduce the number of people living with undiagnosed HIV infection, decrease the number of people diagnosed late and ensure those diagnosed are rapidly linked to care and treatment. This will reduce morbidity, mortality and onward transmission. In 2021, an estimated 4,400 people were living with undiagnosed HIV, and a further 7,600 (mid estimate) were living with diagnosed HIV and had transmissible levels of virus (1).

In 2016, the UK government signed up to the WHO global health sector strategy to eliminate viral hepatitis as a public health threat by 2030 (2). In 2022 WHO developed this into a shared strategy, to include HIV and sexually transmitted infections alongside hepatitis B (HBV) and hepatitis C (HCV) (3).

An estimated 70,649 people were living with chronic HCV in England in 2022 (4). This number has reduced by 45% since 2015 largely due to the roll out of direct-acting antiviral (DAA) treatment through the NHS England Hepatitis C Elimination Programme. People who inject drugs (PWID) are the largest risk group for HCV and this population has been successfully targeted for testing and treatment, with routine testing in place in drug services. Most people who remain undiagnosed either have a past injecting history, or no injecting history (5). Pilots are underway to increase testing for people who are not aware they are at risk, including in primary care, through web testing and in antenatal services.

An estimated 206,000 people were living with chronic HBV in England in 2021, and the proportion that is undiagnosed is unknown. Chronic HBV is most commonly acquired in infancy and childhood in countries that have a higher prevalence of HBV, and around 95% of new chronic HBV diagnoses in England are among people who were born in a higher prevalence country. HBV is routinely tested for in antenatal services, but there has been no other systematic or widespread testing for HBV, and the majority of people living with HBV remain undiagnosed and unlinked to care (6).

Evidence for opt-out testing in ED

There is good evidence that ED opt-out HIV testing is acceptable to individuals and is feasible and cost-effective (7). Based on this evidence, the National Institute for Health and Care Excellence (NICE) guidance recommends offering an HIV test to people who have a blood test when they attend an ED in areas of high (more than 2 per 1,000 people aged 15 to 59 years) or very high (more than 5 per 1,000 people aged 15 to 59 years) diagnosed HIV prevalence, and to all ED admissions in areas of very high diagnosed HIV prevalence (7). EDs are not currently recommended by NICE as a setting for opt-out viral hepatitis testing, but the available evidence shows that EDs are an important setting for HBV and HCV testing and has a key role in elimination efforts (8).

EDs offer open access healthcare to the entire population, with an estimated 24.4 million ED attendances annually in England, a rate of 28,537 attendances per 100,000 population (9). ED opt-out testing can therefore address inequalities in testing among people who may not identify themselves as being at risk of a BBV and do not seek testing, and or may not be offered testing in health care services. ED opt-out testing also offers an opportunity to reengage people previously diagnosed with BBV but who are not currently in care. The approach seeks to normalise the process of testing for BBVs and reach groups underserved by other forms of BBV testing, treatment or care (8, 10). Studies have shown that testing uptake is higher if implementing an opt-out approach compared to an opt-in approach (11). Opt-out testing in ED can be cost-effective even at low prevalence. However, the cost-effectiveness of such testing is largely dependent on people being referred to and receiving healthcare following diagnosis (8, 12, 13). Linkage to care from EDs presents additional challenges due to the nature of the ED attendance and the circumstances of the population who attends, which may include difficulties in contacting people due to being homeless, and incomplete or outdated contact information (8).

The ED opt-out testing programme

The HIV Action Plan committed £20 million in funding over 3 years to implement opt-out HIV testing in EDs serving areas of very high diagnosed HIV prevalence. Through a partnership with NHSE Hepatitis C Elimination Programme, HBV and HCV testing were included to form a combined BBV opt-out testing approach aligned with the WHO Shared Strategy on HIV, viral hepatitis and sexually transmitted infections (3).

The opt-out testing in ED programme (the programme) launched in April 2022 in EDs in areas of very high HIV diagnosed prevalence in London, Manchester, Blackpool, and Brighton, with London adopting a city-wide implementation approach to ED testing, including all Type 1 EDs (major EDs that provide a consultant-led 24-hour service with full facilities for resuscitating patients). HBV and HCV testing were added later than HIV testing at most sites, with roll out and ‘go live’ dates for each BBV varying by site. A table of participating sites and their ‘go live’ dates is provided in the Appendix Table A1. Implementation of the programme and reflections on design, delivery and early outcomes were published by NHS England after the first 100 days (15).

The programme provides testing for HIV, HBV and HCV for anyone aged 16 years and over (at some sites, aged 18 years and over) and having a routine blood test during their ED attendance, unless they opt out. Under the opt-out approach, people attending EDs are informed about testing using accessible and translated public-facing information including posters and banners. Public facing information informs people about testing, its benefits, and how they can opt out or seek further information.

NHSE good practice guidance for opt-out BBV testing in EDs was published in April 2022 and includes a set of key principles informed by national guidance and models of good practice (14). Although initially written for London sites, the guidance seeks to support implementation and consistency in delivery across all participating sites in England. Key principles include:

• the use of an opt-out approach
• dedicated staff training
• public-facing information that clearly informs about the testing, benefits, and how to opt out

Automated electronic patient record (EPR) requests of BBV tests are recommended so that ordering of BBV tests is automated for all eligible attendees, unless they opt out. Repeat BBV testing of an individual should be blocked by EPR for an agreed period, with 12 months recommended. Positive results are reported to and managed by the relevant specialist team (for example HIV, sexual health or hepatology), who notify the individual and facilitate linkage to care; ED staff do not need to manage non-negative results. A ‘no news is good news’ approach is advised for results management. Guidance is also provided around data collection, reporting and monitoring. Implementation varies by site and is discussed in the Implementation optimisation evaluation section of this report .

The programme provides a proof of concept for large scale roll out of ED opt-out testing and aims to assess the feasibility of the approach and its contribution to elimination goals.

NHSE has committed funding for 3 years from April 2022 for the programme. To understand the impact of the programme and inform its potential roll out to other sites, NHSE asked UKHSA to conduct an evaluation to be delivered during the lifespan of the programme (16). The final evaluation will consist of 3 evaluation workstreams:

• a public health evaluation to be delivered as an interim report at 12 months (the focus of this report), followed by a 24-month and then final evaluation in year 3
• a qualitative implementation optimisation evaluation
• a cost-effectiveness analysis

The second and third workstream are being conducted by the University of Bristol.

This report summarises the early data from the public health and implementation optimisation evaluations and provides early recommendations to optimise continued roll out.

Public health evaluation

Aims and objectives 

The primary aim of the 12-month evaluation is to assess the impact and effectiveness of the programme over its first 12 months. It uses healthcare activity and public health disease surveillance data, from the ‘go live’ start dates for testing for each BBV to the end of March 2023. 

The secondary aim of the 12-month evaluation is to use the operational data reported by the programme to assess the completeness of available surveillance data sources and the feasibility of linking between data sets. This is to assess the validity of the indicator definitions for the evaluation so that these can be refined if needed before the 24-month and final evaluations. 

The 12-month public health evaluation has the following objectives to address the primary aim: 

• describe the demographic and risk characteristics of those being tested and diagnosed in EDs and how these compare to testing in other settings 
• describe the numbers and proportions of eligible people having a BBV test, diagnoses made, and referral to care for those newly diagnosed, or previously diagnosed and not currently seen for BBV care 
• identify any potential inequalities in uptake of BBV testing 
• identify any potential inequalities in linkage to care for those newly or previously diagnosed

To address the secondary aim, the 12-month public health evaluation will assess the feasibility of reporting the proposed indicators, including the frequency with which indicators can be updated, data source representativeness, available demographic breakdowns and data quality and completeness. 

Evaluation method 

The public health evaluation uses healthcare activity and public health disease surveillance data to understand the activity, outcomes and impact of the programme. There are benefits to using routine surveillance data: 

• it enables multiple records from the same individual to be deduplicated to more accurately find the number of people that were tested rather than numbers of tests done, and the characteristics of people tested 
• it allows data to be linked between data sources to understand outcomes for an individual, from attendance at ED, to having a BBV test, its result, and diagnosis and linkage to care 
• by linking to past diagnoses, it is possible to categorise which diagnoses are new and which people had a previous diagnosis 

In this report we present surveillance data alongside data from the NHSE programme dashboard to triangulate between data sources and to provide overall numbers from programme-reported data where surveillance data does not have complete coverage or is not available. However, evaluating the impact and effectiveness of the programme and assessing any potential inequalities in uptake of BBV testing and linkage to care requires surveillance data for the reasons described above. 

The public health evaluation uses a theory of change model as the framework for the evaluation (16). This is diagrammatically represented by logic models which show the relationship between long-term outcomes, direct programme outputs and programme activities. Logic models for HBV, HCV, and HIV are shown in Appendix 2A. 

Outcome indicators 

There are 13 indicators for the public health evaluation, which represent public health outcomes. These aim to address a series of questions for the evaluation that are outlined in Appendix Table A2. Definitions of each indicator (including numerator and denominator) are provided in Appendix Table A3. The indicators were co-developed with stakeholders including NHSE colleagues and clinicians involved in the ED opt-out programme.  

Most indicators are reported separately for each BBV. Indicators for late diagnosis are not reported in the 12-month report but will be included in the 24-month report. The indicators follow a care cascade: 

1. Testing.
2. Diagnosis.
3. Linkage to specialist BBV care. 
4. Attendance at care for BBV. 

Data sources 

The public health evaluation uses data from the following sources: 

National Emergency Care Data Set (ECDS)  

ECDS is the national data set for urgent and emergency care, managed by NHS Digital (17). Primarily a commissioning data set, it provides information on activity in EDs for a range of secondary uses. ECDS is used for data on ED attendances, blood tests and demographic information about attendees. ECDS includes data on all attendances from all EDs in England and is available for all 33 participating sites. ECDS is available monthly. 

NHSE BBV opt-out testing in ED programme dashboard 

The programme dashboard reports aggregate data for programme metrics and is updated monthly. It includes numbers of attendances, attendances with blood tests, BBV tests, diagnoses and linkage to care outcomes at site and Trust level. It is used to report data on numbers of tests, diagnoses and linkage to care outcomes in order to fill the gaps where surveillance data is not available, and to triangulate with available surveillance data as part of developing indicator definitions. 

Sentinel surveillance of bloodborne virus testing in England (SSBBV) 

SSBBV is a sub-national surveillance system for BBV testing activity and results in England, managed by UKHSA. It includes data on all positive and negative BBV tests taken in community, primary and secondary care services that are processed by approximately 22 sentinel NHS laboratories. Laboratory participation in SSBBV is voluntary. Testing coverage in SSBBV laboratories represents around 40% of the GP registered population in England. Sixteen of the 33 participating ED sites are represented in SSBBV data for the latest year. Because SSBBV includes both positive and negative tests, it can be used to calculate test uptake and positivity rates as well as total numbers of new and previous diagnoses. However, triangulating SSBBV data against the NHSE programme dashboard identified incomplete data for negative tests in SSBBV for 11 of the 16 sites. Therefore, a subset of 5 sites that have complete data for both positive and negative tests was used to present data on testing rates and positivity in this report. These 5 sites are all in London. SSBBV is available quarterly.  

UKHSA Second Generation Surveillance System (SGSS) 

SGSS is the national laboratory reporting system used in England to capture laboratory reports of infectious diseases and antimicrobial resistance, managed by UKHSA. It is the main system through which NHS diagnostic laboratories submit mandatory reports of notifiable organisms (which include HBV and HCV) to UKHSA, in accordance with the Health Protection (Notification) Regulations 2010. SGSS is used for linkage to identify past diagnoses of HBV and HCV and is available monthly. 

HIV and AIDS reporting system (HARS) and HIV and AIDS New Diagnoses Database (HANDD) 

HARS contains pseudo-anonymised data on all HIV diagnoses and HIV care attendances in England, and HANDD collates reports of new HIV diagnoses, AIDS and deaths of people with HIV. Both are managed by UKHSA and data is available annually. HARS and HANDD are used to provide information on HIV diagnoses and treatment information. HIV diagnoses were obtained by linking positive HIV tests in SSBBV to records in HARS. This means that HIV diagnosis data is limited to the 16 sites that have SSBBV data available for the latest year. In addition, HARS data is only comprehensively available up to the end of 2022. This means for the analysis of new and previous HIV diagnoses, only diagnoses identified between April and December 2022 are included.  

Arden and GEM national hepatitis C patient registry and treatment outcome system 

Arden and GEM national hepatitis C patient registry and treatment outcome system is used for HCV treatment information and to identify previous HCV diagnoses; available quarterly. The registry contains information on people who have received an HCV diagnosis and have initiated or will be initiating treatment. 

GUMCAD STI Surveillance System 

GUMCAD is the surveillance system for sexually transmitted infections (STIs) in England (18). In this context, it is used to provide information on numbers of HIV tests done in sexual health services (SHS), and demographics of those tested.  

Hospital Episode Statistics (HES) 

Hospital Episode Statistics (HES) is a curated data product containing details about admissions, outpatient appointments and historical ED attendances at NHS hospitals in England (19). It is used for linkage to supplement data on ethnicity. 

Data analysis and linkage

To understand programme outcomes, a descriptive approach was taken to summarise BBV testing, positivity and referral to care for each virus, with counts and proportions provided. Data was linked between data sets so that the outcomes (testing, test results, diagnosis and linkage to care) were linked to an individual. Indicators were stratified by site and by individual demographics including gender, age, ethnic group, index of multiple deprivation (IMD) quintile and risk group information. Further details of the data linkage are in Appendix 2D. Logistic regression was used to understand factors associated with BBV test uptake among people who had a blood test, using the variables site, age group, gender, ethnicity and IMD.

Findings

The following section explores the 13 main messages outlined at the beginning of the report in more detail.

Overview of programme 

Main message 1 

Over its first year, the programme was rolled out across 33 ED sites selected due to serving areas of very high diagnosed prevalence for HIV. In programme reported data EDs undertook 857,117 HIV tests, 473,723 HCV tests and 366,722 HBV tests during the first year. The programme made a substantial contribution to all BBV testing in England. The number of tests done in the programme was equivalent to more than half of all BBV tests done in general practice (GP), sexual health, drug and prison services. 

Coverage of surveillance data 

Main message 2

Alongside Programme data, this report uses data from sentinel surveillance of bloodborne virus testing (SSBBV), a voluntary sub-national surveillance system which has patient identifiers and basic demographics, and covered 16 out of 33 sites, 15 of which are London based. A deep dive analysis was undertaken for 5 out of 33 sites (all London based) where there was more complete surveillance data for positive and negative tests in relation to test uptake and positivity. Future evaluation reports will include increased SSBBV coverage of ED sites. 

Details 

The ED site coverage for indicators varies by indicator and surveillance or Programme data as outlined in Table 1.

Table 1. ED site coverage of indicators 

Indicator	Coverage and data source
Diagnoses and linkage to care: 1) attendances and attendees 2) attendances and attendees with a blood test	All sites: ECDS and NHSE programme dashboard
Deep dive on BBV test coverage and BBV test positivity	5 SSBBV sites with good data quality and completeness: Homerton University Hospital, Newham General Hospital, St Thomas’ Hospital, The Royal London Hospital, Whipps Cross University Hospital
Total numbers diagnosed (newly or previously); Linkage to care	16 SSBBV sites (as below)

The 16 sites included for diagnosis and linkage to care analysis are:

Charing Cross Hospital, Chelsea and Westminster Hospital, Croydon University Hospital, Hillingdon Hospital, Homerton University Hospital, King’s College Hospital, Kingston Hospital, Newham General Hospital, Princess Royal University Hospital, Royal Sussex County Hospital, St George’s Hospital (Tooting), St Mary’s Hospital, St Thomas’ Hospital, The Royal London Hospital, West Middlesex University Hospital and Whipps Cross University Hospital. 

A summary of indicator definitions, coverage and values, alongside the data from programme reporting, is available in Appendix Table A3.  

ED attendees and attendees who had a blood test 

Main message 3

Across all sites, the population eligible to be tested by the programme due to having a blood test as part of their ED attendance was older than the population tested for BBVs in other settings. The ethnic distribution of eligible attendees varied by geography and reflected local population demographics. 

Coverage  

Between April 2022 and March 2023, across all 33 sites participating in the programme, there were 2,464,160 attendances and 1,613,699 attendees (adjusted for sites’ individual go live dates for HIV testing) in ECDS data (Figure 1). Of these, 1,320,878 attendances (54%) and 934,466 attendees (58%) had a blood test and so were eligible for the opt-out BBV test. These numbers are similar to, but slightly lower than the overall numbers in the programme reported data: 2,791,968 attendances, of which 1,477,482 (53%) had a blood test. The NHSE Programme dashboard does not report numbers of attendees.  

The 28 London sites represented most (82%, 2,243,201) of attendances and attendees (81%, 1,306,804). Sites outside London (3 in Manchester, one in Blackpool and one in Brighton) represented 19% of attendees during the programme’s first year. The 16 sites represented in SSBBV data were all in London apart from one and represented 59% (1,643,268) of all attendances and 65% (1,044,864) of attendees. The 5 SSBBV sites used for testing and positivity analysis were all in London and represented 18% (488,850) of all attendances and 21% (332,997) of all attendees for the ED opt-out testing programme during its first year.  

Figure 1. Coverage of the programme sites and surveillance data 

Demographics 

Gender 

Across all sites, just over half (54%, 865,036) of attendees were women and this was similar at London sites, Manchester, Blackpool and Brighton sites (grouped), all 16 SSBBV sites and the 5 deep dive sites. Women were slightly more likely to have a blood test (60%) than men (56%). Overall, a higher proportion of people eligible to be tested in the programme were women when compared to the people tested for HIV, HCV or HBV in other settings (outside of antenatal). 

Age 

Across all sites, the median age of attendees was 44 years (interquartile range (IQR): 30 to 62 years) and the most common age group was 35 to 49 years (23%, 376,893). This was similar in London sites, Manchester, Blackpool and Brighton sites (grouped) and in SSBBV sites. Attendees who had a blood test, and so were eligible for the opt-out BBV test, were older compared to all attendees (median age 50 years, IQR 33 to 69) and compared to other settings where testing takes place (such as general practice (GP), sexual health, drug and prison services). (Figure 2). In the 5 deep dive sites used for testing and positivity data, eligible attendees were younger than the overall sample, with a median age of 39 years and IQR of 28 to 56. 

Figure 2. Age distribution of all attendees and attendees that had a blood test at all participating sites compared to people tested for HIV in SHS and for HBV and HCV in other primary care settings (excluding antenatal) 

Ethnicity 

Across all 33 sites, ethnicity was available for 85% (1,373,866) of attendees, of which 41% (559,357) were of white British ethnicity. This was similar for the 16 SSBBV sites. London sites had a lower proportion of white British ethnicity (33%) with sites in Manchester, Blackpool and Brighton having a much higher proportion of white British ethnicity (71%). Compared to all sites participating in the programme, the 5 deep dive sites used for testing and positivity analysis had a lower proportion of eligible attendees of white British ethnicity (26%) and higher proportions of white other (20%) and Asian (Indian, Pakistani and Bangladeshi) ethnicity (18%) (see data in the Appendix tables).  

Across the programme as a whole, the proportion of people receiving a blood test varied by ethnicity, ranging from 52% to 53% in people of white other, mixed or multiple or other ethnicity to 59% to 62% in people of white British, black Caribbean and Asian (Indian, Pakistani, or Bangladeshi) ethnicity (see data in the Appendix tables).  

Compared to people testing for HIV in SHS nationally, the population eligible to be tested under the programme included a higher proportion of people from non-white British ethnic groups. However, this is largely due to most participating EDs being in London. The ethnic distribution of attendees across the whole programme is largely similar to the population who have HIV testing in SHS in London, but with a higher proportion of people of Asian (Indian, Pakistani or Bangladeshi) and other ethnicity, and a lower proportion of black Caribbean, black African, mixed or multiple and white other ethnicity.  

IMD  

Over all sites, a higher proportion of attendees lived in the most deprived areas than in the least deprived areas. London sites had a slightly higher proportion living in the most deprived areas, and sites in Manchester, Blackpool and Brighton had a higher proportion living in less deprived areas compared to sites overall, but the difference was small. The 5 deep dive sites used for testing and positivity data also had a slightly lower proportion of attendees from the most deprived areas and a higher proportion from the least deprived compared to sites overall, but again this difference was small (see data in the Appendix tables). The proportion of people receiving a blood test was very similar across IMD quintiles, ranging from 59% to 60%. 

Overall, the population eligible to be tested under the programme live in less deprived areas than people testing for HIV in SHS nationally and in London (Figure 3).  

Figure 3. IMD distribution of attendees and attendees with a blood test, compared to people testing for HIV in SHS nationally and in London 

Bloodborne virus testing rates 

Main message 4

At the 5 sites selected for a deep dive analysis, half of eligible attendees had an HIV test and this was similar for HCV and HBV. Within sites, BBV testing was broadly equitable by age, sex, ethnicity and IMD but was lower in the youngest and oldest age groups. However, the proportion tested varied considerably by site, between 22% to 74%. 

Details 

We used data from the 5 sites that had adequate data on both positive and negative tests in SSBBV: Homerton University Hospital, Newham General Hospital, St Thomas’ Hospital, The Royal London Hospital, and Whipps Cross University Hospital 

Overall, 49% of eligible attendees (people who had a blood test) had an HIV test, and the proportion was slightly lower for HCV (46%) and HBV (46%) after adjustment for their respective ‘go live’ dates. Testing rates among eligible attendees varied between sites, ranging from 22% to 74%, with similar patterns for all 3 BBVs (Figure 4). 

Overall in the 5 deep dive sites, BBV testing rates among eligible attendees were similar for men and women (both 49% for HIV, and 45% for men and 46% for women for both HCV and HBV). Testing rates were lower in the youngest and oldest age groups, with people aged 24 years and under, and 65 years and over less likely to have a BBV test compared to those aged 25 to 64 years. By IMD quintile, a higher proportion of eligible attendees living in the most deprived areas had a BBV test compared to those living in the least deprived areas (from 51% in the most deprived to 46% in the least deprived for HIV, with similar patterns observed for HBV and HCV). BBV testing rates among eligible attendees varied by ethnicity, ranging from 43% in Asian other ethnicity to 52% in mixed or multiple ethnicities for HIV, with similar patterns observed for HCV and HBV (see data in the Appendix tables).  

After adjustment, the specific site attended was the strongest predictor of whether a BBV test was received. Differences in testing by age, IMD and ethnicity were small and are likely to be associated with site-level variations in demographics and testing practices. Notably, HIV testing rates was slightly lower among females after adjusting for other variables, which was not the case for HCV or HBV.  

Analysis of all sites’ SOPs for delivering opt-out testing is presented in the implementation optimisation section of this report. For the 5 sites included in the deep dive analysis, 4 had an SOP that recommended reminding the individual about their BBV test at the point of blood being drawn. For the 3 sites that had lower rates of testing, staff-dependent test ordering was in place, whereas the 2 sites that had higher test coverage had automated EPR test requests.  

Figure 4. Testing among ED attendees with a blood test by test site and BBV, April 2022 to March 2023, adjusted for site and HBV go-live dates 

Bloodborne virus positive tests and new diagnoses 

Main message 5 

The highest number and proportion of new diagnoses identified through the programme was for HBV, reflecting the higher prevalence of people living with undiagnosed HBV compared to HIV and HCV. There has been substantial effort and financial commitment to increase diagnosis and linkage to care for HIV and HCV in recent years, which needs to be replicated for HBV if WHO elimination targets by 2030 are to be reached. 

Main message 6

Across the 16 sites served by laboratories that report through the SSBBV system, between April 2022 and March 2023, a total of 646 new HBV diagnoses, 175 new HCV diagnoses were made with an additional 141 people previously diagnosed with HCV and not in care. In addition, 78 new HIV diagnoses and 59 people previously diagnosed with HIV and not in care were found, noting that new HIV diagnoses, identified in the HIV and AIDS Reporting System (HARS) and HIV and AIDS New Diagnoses Database (HANDD), could only be identified up to the end of 2022. 

Details 

Across all 33 sites, programme reported data indicated there were 5,068 positive HIV tests, of which 4,400 were people already in care, 341 were new diagnoses, 208 were people previously diagnosed and not currently in care, and for 119 status was not yet known. There were 2,454 HBV (hepatitis B surface antigen (HBsAg, indicative of current infection) positive tests. Of these 1,143 were new diagnoses, 768 were already in care, 180 were people previously diagnosed and not currently in care and there were 363 whose status was not yet known. There were 799 HCV (RNA positive, indicative of current infection) positive tests, of which 499 were new diagnoses, 129 were people currently in care, 99 were people previously diagnosed and not currently in care, and there were 72 whose status was not yet known. 

At the 16 sites where SSBBV data was available, a total of 1,424 people had a positive HIV test and 1,192 received this between April and December 2022. Of these 78 were new diagnoses, 921 were previously diagnosed and in care, 59 were previously diagnosed and not currently attending HIV services and 134 were unmatched. The unmatched group are people whose positive HIV test in surveillance data did not match to a record in HARS and so it is not possible to confirm whether they were new diagnoses not yet linked to care, people who had received a negative confirmatory test result following a positive reactive screening result and therefore did not have HIV, or those where a match between existing records in SSBBV and HARS was not possible. Further data will help categorise these individuals. 

For HBV, 1,318 people tested HBsAg positive, and 646 were new HBV diagnoses.  

For HCV, 339 people tested HCV antigen and or RNA positive, 175 of these were new HCV diagnoses and 141 were people previously diagnosed and not currently in care. There were 1,066 people who tested HCV antibody positive (the marker of past or current HCV infection) and of these, 891 (84%) had a confirmatory RNA test for current infection, and 814 (91%) of these were done within 7 days, the proxy marker of reflex testing (a test done on the same sample).  

The total numbers of people testing positive and the number of new diagnoses across the 16 SSBBV sites, which covers an estimated 64% of eligible attendees, is lower compared to the programme data, and the largest difference is for HIV. There are several reasons likely to be contributing to the differences: 

• programme data reports numbers of positive tests, whereas surveillance data enables individuals tested to be calculated 
• new diagnoses are likely to be over reported in programme data as trusts do not have full access to data on past diagnoses and so may record a diagnosis as new if they are unable to find information about a previous diagnosis or an individual does not disclose a previous diagnosis 
• there is possible incomplete identification of positive tests in surveillance data, for example tests may not have been identified if they were not coded as being done in an ED 
• some new HIV diagnoses will not be identified as new where they were not possible to match 

Further work with NHSE and sites is needed to understand whether all relevant tests are identified in surveillance data, which will be continued as part of developing metrics for the 24-month and final evaluation reports.  

Bloodborne virus test positivity 

Main message 7

At the 5 sites selected for a deep dive analysis, positivity (including people with known infections) was 1.1% for HBV, 0.9% for HIV and 0.2% for HCV RNA. For all 3 BBVs, new diagnoses were higher among men. New diagnoses for HIV and HBV were higher among people aged 35 to 64 years, and aged 50 to 64 years for HCV. For HIV and HBV, the proportion of new diagnoses among those tested was highest among people of black African ethnicity, and for HCV they were highest among people of white other ethnicity. 

Details 

To understand test positivity a deep dive into the 5 sites was undertaken that had complete data for both positive and negative tests in SSBBV: Homerton University Hospital, Newham General Hospital, St Thomas’ Hospital, The Royal London Hospital, and Whipps Cross University Hospital 

At the 5 deep dive sites there were 742 (0.9%) positive HIV tests, of which 48 (0.07%) were new diagnoses and 25 (0.04%) were people previously diagnosed and not currently in care; 633 (1.1%) positive HBV tests, of which 269 (0.5%) were new diagnoses; there were 591 (1.0%) HCV antibody positive tests and 139 (0.2%) HCV RNA positive tests, of which 67 (0.1%) were new diagnoses.   

For HIV, overall positivity at the 5 sites was higher than the diagnosed prevalence in the local areas that the 5 EDs are in (0.7%) and the positivity from all ED testing in higher prevalence areas nationally (0.6%) (20, 21). The proportion newly diagnosed was 0.07%, which is within the low end of the range of 0.06% to 0.28% reported in recent ED opt-out testing studies in the UK (22 to 26). 

For HBV, overall prevalence among those who tested was 1.1%, which is higher than the general population estimate of 0.45%, and higher than the range of 0.5% to 0.9% reported in recent UK ED opt-out studies (6, 23, 25 to 27). The proportion newly diagnosed was 0.4%, which is within the range of 0.1% to 0.5% reported in these studies (23 to 27).  

For HCV, overall prevalence of current infection among those tested was 0.2%, which is slightly higher than the England general population estimate (0.15%) (4). New HCV RNA diagnoses were 0.1%, which is lower than the range of 0.3% to 0.7% reported in recent UK ED opt-out testing studies (23 to 27). 

The proportion of people tested who were newly diagnosed with HBV was higher among males, people aged 35 to 64 years and people of black African ethnicity (Figure 5). The proportion of people tested who were newly diagnosed with HCV RNA was higher among males, people aged 35 years and over and people of white other ethnicity (Figure 6). The proportion of people tested who were newly diagnosed with HIV was higher among males, people aged 35 to 64 years, people of black African ethnicity and people from the most deprived IMD quintiles (Figure 7).  

Figure 5. Percentage of people newly diagnosed with HBV (SBBV data from 5 site)

Figure 6. Percentage of people newly diagnosed with HCV (SSBBV data from 5 sites) 

Figure 7. Percentage of people newly diagnosed with HIV (SSBBV data from 5 sites) 

Characteristics of people newly diagnosed with a BBV 

To understand the impact of the programme on identifying new diagnoses in groups who may be less likely to be diagnosed in other settings, we looked at the proportion of people who had a positive test who were newly diagnosed by person demographics, using data from all 16 SSBBV sites. Because the denominator for this proportion is the total number of people who tested positive in a group, groups with lower positivity overall can still have high proportions of new diagnosis by this definition.  

HBV had the highest number of new diagnoses, and the proportion of new diagnoses among those who tested positive was also higher than for HIV and HCV. The number of new diagnoses was higher in men, and as a proportion of those who tested HBV positive, the proportion newly diagnosed was also higher for men than for women. The proportion of new diagnoses was higher in the 80 and over age group than in other age groups, but the overall number of people who tested positive in this age group was low. The proportion of new HBV diagnoses did not vary significantly by ethnicity, and the highest numbers of new HBV diagnoses were among people of white other and black African ethnicity.  

For HCV, the number of new diagnoses and the proportion of all diagnoses that were new were lower than for HBV. The number of new diagnoses was higher in men than women, but as a proportion of people who tested positive for HCV, the proportion who were newly diagnosed was similar between men and women. For HCV, the proportion of new diagnoses among those who tested positive was higher in people of Asian (Indian, Pakistani or Bangladeshi), white other and other ethnicity. By ethnicity, the highest number of new HCV diagnoses was among people of white other ethnicity.  

HIV had the lowest number and proportion of new diagnoses among all 3 BBVs. For HIV, the number of new diagnoses was higher among men, and in people of black African ethnicity. The proportion of new diagnoses among those who tested positive for HIV was higher among people aged 16 to 24 years, although the overall number who tested positive in this age group was low. The proportion of new diagnoses was higher among people who probably acquired their infection through sex between men and women than those whose probable route of acquisition was sex between men.  

Of all 3 BBVs, both the proportion of new diagnoses and number of new diagnoses were highest for HBV, suggesting a need to increase opportunities for testing and diagnosis for HBV in other settings. 

Linkage to care 

Main message 8

At the 16 SSBBV sites, linkage to care for people newly diagnosed with HIV or HCV was lower than reported in national data, demonstrating the additional challenges in linking to care from ED, which may include challenges in contacting people due to incomplete details, being homeless or not registered to a current local address. Linkage to care was lower for HCV and HBV than for HIV, which mirrors findings from other ED studies and national surveillance. For HIV, linkage to care was considerably lower among people previously diagnosed and not in care than for people newly diagnosed. 

Details 

British HIV Association (BHIVA) guidelines recommend that people diagnosed with HIV should be linked to care within 14 days (28). In 2021, 79% of people first diagnosed with HIV in England in any setting were linked to care within 14 days, and 90% were linked within a month (22).  

In programme reported data, 79% of people newly diagnosed with HIV were linked to care overall, and 35% of people previously diagnosed with HIV and not currently in care were reengaged. These proportions were lower in SSBBV and HARS data from the 16 sites; among the 78 people who were newly diagnosed with HIV at their ED attendance, 31 (40%) were linked within 14 days of their positive test in ED, and 45 (58%) were linked to care within 30 days (Figure 8). Among the 59 people who were previously diagnosed with HIV and not currently receiving care in SSBBV and HARS data from the 16 sites, 8 (13%) were linked to care within 30 days of their positive test in ED, and 16 (27%) were linked to care within 90 days. 

Figure 8. HIV linkage to care for people newly diagnosed, and previously diagnosed and not currently in care 

For HCV, a Commissioning for Quality and Innovation (CQUIN) target of linkage to care within 28 days of diagnosis was introduced in April 2023 and fully implemented in August 2023. 

There is not yet any national data on timeliness of linkage to care for HCV. Overall, between 2015 and 2020, 55% of people who tested HCV RNA positive had evidence of being linked to treatment, using the Arden and GEM treatment registry (5). 

In programme reported data on people who tested HCV RNA positive, among the 604 people either newly diagnosed, diagnosed with a reinfection or previously diagnosed and not in care, 52% engaged with initial clinical contact. In SSBBV data from the 16 SSBBV sites, among all 339 people who tested HCV RNA positive, 155 (46%) had evidence of being linked to care, 25 (7%) of these within 28 days. There were 175 new diagnoses, and among these, 79 (45%) were linked to care overall of which 8 (5%) were linked within 28 days of diagnosis (Figure 9). There were 141 people who had previously been diagnosed with HCV and were not currently in care, and of these 62 (44%) were linked to care and 16 (11%) were linked within 28 days. The median time between HCV diagnosis and linkage to care was 58 days (IQR 33 to 110 days). There were 6 people who had HCV reinfections and 4 (67%) of these were linked to treatment. 

Figure 9. HCV linkage to care for people newly diagnosed, and previously diagnosed and not currently in care 

There are currently no national guidelines specifying timeliness of HBV linkage to care, and no national surveillance of HBV linkage to care. In programme reported data for people who tested HBsAg positive, the marker for current HBV infection, of the 1,323 people who were either newly diagnosed or previously diagnosed and not in care, 98% were notified and offered follow up and 38% attended a hepatology clinic.  

Feasibility of indicator definitions and limitations 

Main message 9 

It was feasible to produce indicators on testing and diagnosis from surveillance data for all 3 BBV, however data on linkage to care for HBV cannot currently be identified in surveillance data, because there is currently no large-scale HBV treatment data set available to monitor HBV care. Surveillance data for BBV tests and diagnoses is sub-national and only covered some of the sites in the programme, with 16 out of 33 sites having diagnosis data. Due to data incompleteness for positive and negative tests, only 5 out of 33 sites had complete data on both positive and negative tests to enable analysis of test uptake and positivity. 

Details 

Feasibility of indicator definitions has been demonstrated for the indicators that we have reported on for this report. The development of indicators identified incompleteness in some source data sets which was discovered by comparing data between sources (such as the NHSE Programme Dashboard and SSBBV). This represents a rare opportunity to triangulate data between sources and has been beneficial in providing a starting point to work towards filling in the gaps in this data. 

There are several challenges with using the SSBBV surveillance system for this evaluation that need to be resolved before the 24-month and final evaluation reports. 

Firstly, SSBBV does not cover all sites participating in the programme, and it is under representative of sites outside London. Increasing coverage of SSBBV overall is needed. 

Secondly, for sites that are represented by SSBBV there were data gaps that meant that some data could not be used. Numbers of negative tests reported were lower than expected for some sites, and identifying tests as having been done at an ED was not possible for some sites, as they did not code data to identify ED tests, meaning that these sites, although included in SSBBV, could not be included in the evaluation.   

Both of these issues need to be resolved before the 24-month and final evaluations, and are being taken forward as areas for action, with UKHSA working alongside laboratories, ICBs and the NHSE Programme group to increase the numbers of laboratories participating in SSBBV, and to understand and resolve data gaps within the data set.   

In this report we have not presented data on CD4 cell counts or disease stage at diagnosis to understand late diagnoses because this was not possible within the timeframe of this report. Presenting this data will provide valuable insight into the role of the programme in diagnosing people who are diagnosed late and will be included in the 24-month report. 

It has not been possible to report on linkage to care for HBV because there is currently no large-scale HBV treatment data set available to monitor HBV care. Methods to identify HBV attendance to care using proxy measures from the SSBBV data set were investigated but it was not possible to develop these for this report. Establishing a national treatment monitoring system for HBV is a priority area for development for UKHSA (6).

Implementation optimisation

Aims and objectives

The implementation optimisation evaluation primary aim is to evaluate the implementation of ED opt-out testing to provide recommendations to guide optimisation, standardisation and any future roll-out of ED testing in lower prevalence sites.

The objectives of the implementation optimisation are to:

• describe the different strategies and approaches of sites to implementing ED opt-out testing outlined in the sites’ SOPs or documentation (phase one)
• identify issues and barriers to ED opt-out testing implementation and uptake (phase two)
• develop recommendations for future implementation of ED opt-out testing, informed by (phase one) and (phase two)

The objectives of phase one work will seek to:

• gain an understanding of how ED opt-out testing is being delivered at each site and identify any variation in implementation that may influence the uptake or delivery of testing
• provide information to facilitate the optimisation and standardisation of any future roll-out of ED testing in lower prevalence sites
• inform areas to be explored in greater detail in phase two (qualitative interviews)

This report includes findings from phase one.

Methods

Implementation optimisation evaluation approach

Data collection approach

All 33 sites participating in the ED opt-out testing programme in year one were contacted by email to request standard operating procedures (SOPs) and any other documentation about the delivery of opt-out testing (for example posters, patient leaflets). The request was sent to staff involved in the delivery of opt-out testing, including ED clinical leads, hepatologists, HIV specialists and data managers. Sites were first contacted in February 2023, with follow-up requests to non-responding sites between June and September 2023.

If the site specified that they did not have an SOP but had documentation providing practical guidance on the delivery of opt-out ED testing, this was accepted and included in the final analysis.

A document analysis of SOPs and other guidance documentation was performed (29). A data extraction tool was produced, informed by discussions with key stakeholders and reviewing literature to identify key procedures associated with opt-out testing. The tool was applied to all SOPs and guidance documents to collate information on how sites were delivering opt-out testing. This allowed for variation and consistency in procedures to be explored.   

Verbatim quotes were extracted from SOPs to illustrate examples, and summary statistics were produced demonstrating the prevalence of any procedures or guidance used in the delivery of opt-out testing. Where possible, findings were contextualised alongside the NHS good practice guidance for opt-out testing in EDs (14).

Findings

Main message 10

Whilst most responding sites had SOPs or guidance documents, there was variability in opt-out testing strategies at each site, and some procedures differed from good practice guidance. SOPs and guidance documents were primarily HIV focused, with HCV mentioned in 13 out of 28 responding sites and HBV mentioned in 7 out of 28.

Summary of responses

From the 33 sites across England currently involved in the programme, all (100%) responded to an email request for a SOP or guidance documentation pertaining to opt-out testing. 17 (52%) of these sites provided a SOP, 11 (33%) provided information in the form of guidance or a flowchart, and 4 (12%) stated they had no SOP and did not provide any other documentation. One site (3%) was currently developing its SOP. Data were extracted and analysed from 28 out of 33 sites SOPs or documents (results summarised in Appendix Table A4).

All 5 deep dive sites selected for the testing and positivity analysis in the public health evaluation provided SOPs. Of these, 3 were from the same NHS trust and used the same joint HIV and HCV document. One site had a joint BBV SOP (including mentions of HIV, HBV and HCV) and one had an SOP that only mentioned HIV. These 5 sites’ SOPs are discussed with the findings from the public health evaluation.

Publicity and training

Public-facing materials

Good practice guidance: sites should display accessible information in prominent areas in ED about opt-out testing, including how to opt-out and receive results.

SOPs or guidance documents from all 28 responding sites suggested that they provided opt-out testing information in ED. Information obtained from each site’s SOP or guidance document advised that posters or leaflets should inform the individual of testing, provide information on why a test is being performed and advise on how results are communicated. Seven (25%) sites reported providing a leaflet on opt-out testing to each individual.  

Ten (36%) sites reported that they had translated information available, one of whom stated that non-English speaking individuals should have a translator present or accessible via telephone to inform the individual of testing. Six (21%) sites reported having a website or having material available online. No SOPs or guidance documents mentioned that they provided information for blind or partially sighted people.

Staff training

Good practice guidance: sites should provide staff with training to support the delivery of opt-out testing, and use BBV ‘testing champions’ to support staff with opt-out testing.

Of the 28 sites that responded, 14 (50%) reported providing opt-out testing training for staff. BBV testing champions were found to operate in 18 (64%) sites.

Where information was available, specific training included providing information on the importance of testing, how to conduct a test (such as ordering tests, use of bottles, process for consent) and procedures for results management. Sites provided training in both video and paper formats. In some cases, hepatology and HIV specialist teams supported ED teams with implementation. Some sites also recommended highlighting BBV testing at handovers, and one site promoted testing through an ‘SOP of the week’ award.

Opt-out testing processes and procedures

Main message 11

All SOPs from sites suggested that testing was delivered using an opt-out model although there was some variation in how opt-out testing was implemented. All sites used posters and banners to inform individuals about testing and how to opt-out with some sites additionally recommended that staff provide a verbal reminder or signposting people to the ED opt-out testing programme.

Opt-out process and eligibility procedures

Good practice guidance: all adults who are having blood tests should be routinely tested for BBVs unless they opt out, and verbal consent is not required. BBV testing for someone unconscious or lacking capacity is recommended if it is in their best interests.

All sites implemented testing using an opt-out model. Where it was possible to obtain information from SOPs, there were nuanced differences in how this model was implemented. For example, some sites recommended not to verbally remind individuals at the point of blood draw when posters were clearly displayed in cubicles:

HIV testing is OPT OUT, that is counselling and formal consent is NOT needed. There should be posters in each cubicle – if so, you DO NOT need to ask every time.

(Site 25)

Others advised that staff could provide verbal reminders at the point of blood draw if they wished, but this was not a requirement:

Patients may also be verbally informed by staff that routine BBV testing is being undertaken when blood tests are taken, but this is not a requirement.

(Site 1)

Finally, some site SOPs recommended that staff provide a verbal reminder or ask an individual if they are happy to receive a test:

All patients who require a blood test in this Emergency Department are routinely tested for HIV, are you happy to have the test?

(Site 23)

There were 6 sites that excluded individuals with a ‘temporary impairment or loss of capacity’ but suggested testing was suitable for individuals with a ‘permanent impairment’. Four sites excluded individuals who were unconscious but recommended testing if it was in the ‘best interest’ or testing was ‘necessary to save their life’. Three sites excluded individuals with a ‘lack of capacity’, and 4 sites stated that ‘all adults’ should be tested without mentioning temporary or permanent impairment or lack of capacity. Eleven sites did not specify exclusion criteria regarding capacity.

Main message 12

Most sites have an automated system for adding a BBV test to the ED blood set, although a small number of sites required staff to manually order tests. In the public health analysis, testing rates were higher in sites that had automated ordering of BBV tests.

Integration with EPR systems (including recording of opt-out testing)

Good practice guidance: BBV tests should be automatically added when any blood test is requested in ED, so that staff are not required to add on BBV testing manually. A patient’s decision to opt-out should be recorded on their medical record.

There were 22 (79%) sites that had an automated system for adding a BBV test to the ED blood set, and 15 sites also reported that they recorded reasons for opt-out through EPR.

There were 6 (21%) sites where the SOP reported that the clinician requesting blood tests had to add the BBV test manually alongside other bloods requested. 

At 4 of the 15 sites where an individual’s decision to opt-out was recorded, only 4 had an EPR system that allowed them to record why the individual opted-out. Some sites specified that individuals would be blocked from testing for a specific period if they opted-out, usually for 6 months.

Blocking

Good practice guidance: repeat BBV testing should be blocked by EPR; a timescale of 12 months is recommended.

Seventeen sites outlined their blocking timescales, ranging from 3 to 12 months.

Eight sites reported using a 12-month blocking period. A further 8 suggested a blocking period of 6 months, and a 3-month period was used at one site. Some sites justified blocking periods that were less than 12 months due to being in extremely high areas of diagnosed HIV prevalence:

For frequent visitors to EDs, NICE recommends at least one HIV test per 12 months. As [site] is an area of extremely high HIV prevalence, we will test patients every 6 months.

(Site 29)

Blood-taking procedure

Of the 20 sites that reported on their blood taking procedure for opt-out testing, all stated that an extra blood bottle was needed to obtain bloods for BBV testing in addition to ED routine bloods. One site tested for HIV from the same bottle as a full blood count, although HCV and HBV still required an extra bottle. 

Results management

Main message 13

Sites used a ‘no news is good news’ approach to results management. If the individual had not opted-out but no BBV test was performed, the SOPs had no mention of procedures to highlight this to individual. 

Management of negative results

Good practice guidance: use a ‘no news is good news’ approach for negative results.

A ‘no news is good news’ approach was advised by 24 (86%) sites. At one site, results were added to the individual’s discharge summary, and 3 sites did not specify their approach.

At 19 (68%) sites, provisions were in place for people to find out their result; 15 via a results phone line and four online or via an app. One of these sites specified that the individual must be registered with a GP to access their results either online or via an app. One other site also added a individuals result to their discharge summary whilst following a ‘no news’ approach.  

Management of non-negative results

Good practice guidance: for non-negative results, specialist clinical services (HIV, sexual health, hepatology) follow up on results and notify the individual. ED staff are not expected to contribute to this.

All sites were found to follow this procedure. At 3 sites there was a new role created within the HIV team for a ‘tracker’ who was responsible for the follow-up of BBV tests. 

Insufficient sample or unable to follow-up

Good practice guidance: develop appropriate cautious wording highlighting situations where it may not be possible to test or contact patients in the event of a non-negative results (for example in the event of tests not being processed due to technical or human error, or if a patient does not respond to attempt to contact them).

SOPs or guidance documents from 8 sites suggested contacting individuals via text if they were unable to test due to insufficient bloods, advising them to attend a local GUM clinic or GP to repeat the test. In the 3 sites that had a dedicated staff tracker role, individuals were recalled for a repeat test with the tracker.   

However, if the individual had not opted-out, but no BBV test was performed, the SOPs had no mention of procedures to highlight this to individuals who may have believed they had been tested.

Limitations

Although efforts were made to obtain all relevant SOPs and guidance documents, not all sites returned all relevant documents and of those that did, some contained ambiguous information. SOPs or guidance documents may have also been updated or produced in the period following our analysis. Although SOPs and guidance documents were analysed using established document analysis methods, there is always the possibility of procedures being implemented in different ways: for example, the absence of information relating to a specific procedure does not necessarily mean it was not implemented or adhered to. There may also be additional implementation issues that could not be obtained from the SOPs or guidance documents.

Conclusions

During its first year, the programme has demonstrated that opt-out ED BBV testing can be delivered at scale, and as far as can be evaluated, equitably within sites, although there were large differences between sites. Although approximately half of all eligible attendees were tested for BBVs, testing varied by site, ranging from 22% to 74% using deep dive analysis of 5 sites. Overall, this is relatively high compared to other ED opt-out testing studies and reflects the substantial work that has been done to implement effective opt-out testing processes across multiple sites. The programme has not yet reached its target of 95% of eligible people being tested at all sites, however these data represent the early stages of the programme and it takes time to implement and embed new processes. The 24-month and final evaluations will monitor how uptake develops over the second and third year of the programme. 

Achieving universally high rates of testing under the programme is needed to ensure that between-site variations do not cause inequalities in offer and access to testing. The SOP analysis identified variations in programme delivery that require standardisation to reduce inequalities, maximise test acceptability and coverage. Ensuring all sites implement opt-out testing procedures, including when to use verbal prompts and signposting, automated testing and blocking of unnecessary repeat testing will be key to the effectiveness of the programme, and can ensure equity in offer and access to testing (8, 30).

At sites where coverage remains low, the ‘no news is good news’ approach to results management could lead to individuals assuming they have been tested because they have not received a positive result, when in fact they have not been tested. It is therefore important that sites develop procedures to mitigate this risk, for example providing access to a results line where individuals can call to check the status of BBV test results, or text notifications that a BBV test was not done and signposting to alternative ways to access a BBV test.

The programme was effective in identifying new diagnoses for all 3 BBVs, with the highest number being for HBV, reflective of the greater proportion of people living with undiagnosed HBV compared to HCV and HIV. The high number of new HBV diagnoses has implications of how to meet the increased need for HBV care when considering expansion. Overall, the lower numbers of new diagnoses for HIV and HCV compared to HBV are reflective of the substantial effort and financial commitment there has been to increase diagnosis and linkage to care for HIV and HCV, which needs to be replicated for HBV if elimination targets are to be reached.

Linkage to care was sub-optimal for all 3 BBVs, and more so for HBV and HCV than for HIV. It will be important for ICBs, commissioners and service providers to reflect upon the different needs of individuals diagnosed in EDs. It is important to take on learning from the implementation of the programme to optimise care pathways and implement facilitators to linkage to care for both newly diagnosed and those previously diagnosed to ensure the opportunities offered by the programme are not wasted. This includes interventions such as community support.

There were limitations in the coverage and completeness of surveillance data for this evaluation. The availability of both programmatic and surveillance data allowed triangulation between data sources and has been useful in providing a starting point to work towards filling the gaps in the surveillance data for BBV tests and diagnoses. It will be important to address these gaps before the 24-month and final evaluations.

References

1. DHSC. ‘Towards zero - an action plan towards ending HIV transmission, AIDS and HIV-related deaths in England - 2022 to 2025’. 21 December 2021

2. WHO. ‘Global health sector strategy on viral hepatitis 2016 to 2021: towards ending viral hepatitis’. June 2016

3. WHO. ‘Global health sector strategies on, respectively, HIV, viral hepatitis and sexually transmitted infections for the period 2022 to 2030’. 18 July 2022

4. UKHSA. World Hepatitis Day 2023: infographic. 2023

5. UKHSA. Hepatitis C in England: 2022. 2022

6. UKHSA. Hepatitis B in England 2023. 2023

7. National Institute for Health and Care Excellence (NICE). ‘HIV testing: increasing uptake among people who may have undiagnosed HIV. NICE guideline [NG60]. 2016

8. Simmons R, Plunkett J, Cieply L and others. ‘Blood-borne virus testing in emergency departments - a systematic review of seroprevalence, feasibility, acceptability and linkage to care’. HIV Medicine. Volume 24 issue 1, page 6 to 26. 2023

9. NHS Digital. Hospital accident and emergency activity 2021 to 22. 15 September 2022

10. Cullen L, Grenfell P, Rodger A and others. ‘‘Just another vial…’: a qualitative study to explore the acceptability and feasibility of routine blood-borne virus testing in an emergency department setting in the UK’. BMJ Open. Volume 9 issue 4\, e024085. 2019

11. White DA, Sadoun T, Tran T and others. ‘Increased acceptance rates of HIV screening using opt-out consent methods in an urban emergency department’. Journal of Acquired Immune Deficiency Syndrome. Volume 58 issue 3, page 277 to 282. 1 November 2011

12. Williams J, Vickerman P, Douthwaite S and others. ‘An economic evaluation of the cost-effectiveness of opt-out hepatitis B and hepatitis C testing in an emergency department setting in the United Kingdom’. Value Health: The Journal of the international society for Pharmacological Outcomes and Research. Volume 23 issue 8, page 1003 to 10011. 23 August 2020

13. Prabhu VS, Farnham PG, Hutchinson AB and others. ‘Cost-effectiveness of HIV screening in STD clinics, emergency departments, and inpatient units: a model-based analysis’. PLOS ONE. Volume 6 issue 5, e19936. 20 May 2011 

14. NHS. ‘Blood-borne viruses opt-out testing in emergency departments in London: good practice guidance’. April 2022

15. NHS England. ‘Emergency department opt out testing for HIV, hepatitis B and hepatitis C: the first 100 days’. 29 November 2022

16. Breuer E, Lee L, De Silva M and others. ‘Using theory of change to design and evaluate public health interventions: a systematic review’. Implementation Science. Volume 11 issue 1, page 63. 6 May 2016

17. NHS Digital. ‘Emergency Care Data Set (ECDS): user guidance’. Version 4.2. 4 July 2023

18. UKHSA. ‘GUMCAD STI Surveillance System’. 2023

19. NHS Digital. ‘Hospital Episode Statistics (HES)’. 2023

20. UKHSA. ‘HIV monitoring and evaluation framework’. 2022

21. Office for Health Improvement and Disparities. ‘Public health profiles’. 2023 

22. Hempling MC, Zielicka-Hardy A, Ellis JP and others. ‘Routine HIV testing in the emergency department: feasible and acceptable?’. International Journal of STDs and AIDS. Sage Journals. Volume 27 issue 14, page 1267 to 1274. 11 July 2016

23. Parry S, Bundle N, Ullah S and others. ‘Implementing routine blood-borne virus testing for HCV, HBV and HIV at a London emergency department - uncovering the iceberg?’. Epidemiology and Infection. Volume 146 issue 8, page 1026 to 1035. 2018

24. Bath R, Ahmad K, Orkin C. ‘Routine HIV testing within the emergency department of a major trauma centre: a pilot study’. HIV Medicine. Volume 16 issue 5, page 326 to 328. May 2015

25. Bradshaw D, Rae C, Rayment M and others. ‘HIV, HCV, HBV testing in the emergency department: a feasibility and seroprevalence study’. HIV Medicine. Volume 19 Supplement 1, pages 52 to 57. February 2018

26. Orkin C, Flanagan S, Wallis E, and others. ‘Incorporating HIV, hepatitis B virus, hepatitis C virus combined testing into routine blood tests in nine UK emergency departments: the “going viral” campaign’. HIV Medicine. Volume 17 issue 3, pages 222 to 230. 17 March 2016

27. Evans H, Balasegaram S, Douthwaite S and others. ‘An innovative approach to increase viral hepatitis diagnoses and linkage to care using opt-out testing and an integrated care pathway in a London emergency department’. PLOS ONE. Volume 13 issue 7, e0198520. 25 July 2018

28. BHIVA British HIV Association. ‘Standards of care for people living with HIV 2018. 3a access to and retention in care’. 2018

29. Dalglish SL, Khalid H, McMahon SA. ‘Document analysis in health policy research: the READ approach’. Health Policy Plan. Volume 35 issue 10, page 1424 to 1431. December 2020

30. Smout E, Phyu K, Hughes GJ and others. ‘Real-world clinical effectiveness and sustainability of universal bloodborne virus testing in an urban emergency department in the UK’. Scientific Reports. Scientific reports Volume 12 issue 1, article number 19257. 10 November 2022

Acknowledgements

Report produced by the Public Health and Implementation Evaluation Group of the ED Opt-out BBV Testing Programme.

Report contributors: Olaide Adebayo-Clement, Siobhán Allison, Alison Brown, Erna Buitendam, India Clancy, Monica Desai, Jessica Edney, Rachel Hill-Tout, Stephen Hindle, Susan Hopkins, Jeremy Horwood, Clare Humphreys, James Lester, Kathy Lowndes, Sema Mandal, Veronique Martin, Tom May, Hannah Moore, Debbie Mou, Sarah Murdoch, Amber Newbigging-Lister, Rachel Roche, Ruth Simmons, Ruby Tabor, Georgia Threadgold.

The project was supported by National Institute for Health and Care Research Health Protection Research Unit (NIHR HPRU) in Behavioural science and Evaluation at University of Bristol, with a partnership between UKHSA and the University of Bristol. The views expressed are those of the author and not necessarily those of NIHR, DHSC, or UKHSA.