2.1 Laboratory Data

Sentinel Surveillance of Bloodborne Virus Testing in England (SSBBV) is a unique surveillance system that captures all blood-borne virus tests regardless of the result for tests conducted at any one of 23 sentinel laboratories across England. Participating laboratories are estimated to cover approximately 40% of the English population for primary and reference HBV testing and are broadly representative of most laboratories providing blood borne virus testing. Data collection methods for SSBBV have been described previously [13]. In summary, demographic and testing data for all individuals tested for hepatitis B surface antigen (HBsAg) and/or immunoglobulin M antibody against core antigen (anti-HBc IgM), the latter an indication of acute infection, are extracted from participating laboratory information systems. Individuals are de-duplicated, and test results for each individual are linked over time using a combination of soundex code (a coding based on the person's surname), first initial, date of birth and NHS number.

Two files were extracted from SSBBV; results from all women aged 16–49 years old tested as part of routine antenatal screening between 2015 and 2021, and individuals ≥ 16 years identified as non-antenatal testing, with a HBsAg test during the same time period. The test request location and clinical details accompanying the test request were used to distinguish women assumed to be tested for HBsAg as part of routine antenatal screening from those tested in other settings and for other reasons. For those with multiple tests in the study period, the first test was used in analyses.

2.2 Assignment of Ethnicity

2.3 Exclusions

Individuals with an anti-HBc IgM positive result within 1 month of being HBsAg positive and/or had a negative HBsAg within 6 months were excluded as indicated an acute infection, which, in adults generally results in virus clearance rather than chronic infection. Acute infections were also identified by linking to the reports of an acute infection reported by local teams using NHS number. Those tested through sources investigating sub-groups of the population also were excluded (including occupational health and renal services). Finally, individuals who had conflicting ethnic groups through linkage were excluded, as the true ethnicity of the individual could not be established.

2.4 Population Estimates

Population estimates were based on the 2021 census published by the Office for National Statistics. The dataset used was broken down by age, sex and ethnic group for all residents [14].

2.5 Data Processing and Analysis

SQL was used for data storage and data linkage. Statistical analyses were performed using STATA version 15 (Stata Corp., College Station, Tx, USA).

2.6 Prevalence Estimates

Using population estimates of age, sex and ethnicity for England, we extrapolated the number of people with chronic hepatitis B using the estimated prevalence estimates.

2.7 Ethics

UK Health Security Agency has approval from the Secretary of State for Health and Social Care under ‘section 251’ of the National Health Service Act 2006 and the associated Health Service (Control of Patient Information) Regulations 2002, to collect and process confidential patient information without consent for diagnosing, recognising trends, controlling and preventing, and monitoring and managing communicable diseases and other risks to public health. Therefore, no research or ethics review is required.

3.1 HBV Positivity in Females Aged 16–49 Years by Ethnicity

Among the 793,463 women aged 16–49 years who were tested as part of antenatal care between 2015 and 2021 and reported by laboratories participating in SSBBV, 4560 (0.6%) were HBsAg positive. Data on ethnicity were available on 714,287 (90.0%) and 4174 (91.5%) of women antenatally tested and those who were HBsAg-positive, respectively. Table 1 shows HBsAg positivity among those tested through antenatal screening. The majority of those who were tested were of White British (59.5%), followed by White other (including White Irish) (14.4%) and Asian (12.8%) ethnic groups. Among those who tested HBsAg positive in antenatal test settings, 94% were in ethnic groups other than White British, with 30% in Black ethnic groups and 23% in White other ethnic groups. HBV positivity was highest in people of other ethnic groups (3.0%), followed by those of Black ethnic groups (2.5%) (Table 1).

3.2 HBV Positivity in Individuals Tested Through Non-antenatal Settings

Table 2 shows non-antenatal testing overall and by ethnic group, age and sex. Between 2015 and 2021, 2,207,126 individuals 16 years and over were tested in non-antenatal settings, of which 38,195 (1.7%) were HBsAg-positive (Table 2).

Ethnic group data were available for 1,447,467 (65.6%) individuals, with the majority categorised as White British (60.4%), followed by White other (including White Irish) (12.4%) and Asian ethnic groups (11.8%). Among those tested in non-antenatal services where reported, 52.4% were female, 44.5% were aged 30–49 years, and 58.3% were tested in primary care services. HBsAg test positivity was higher in males than females (2.0% vs. 1.4%), higher in those aged 30–49 years (2.3%), higher among those testing in secondary care compared with primary care (2.2 vs. 1.4), and among those tested in London compared to outside London (2.3% vs. 1.4%). In non-antenatal test settings, 87.1% of positive HBsAg tests were in ethnic groups other than White British, with 30.3% positive HBsAg tests in people in Black ethnic groups and approximately 17%–18% in people of White other, Asian, and other ethnic groups (Table 2). Test positivity was 1.9% (27,628) for those where ethnicity was reported. HBsAg positivity rates were highest among those of any other ethnicity (10.0%), followed by those of black ethnicity (5.9%), whereas individuals of White British ethnicity had the lowest HBsAg positivity rate at 0.4%.

People of White British, Black, Asian or other/mixed ethnicity regardless of sex, aged 30–49 years and ≥ 50 years were more likely to test positive for HBsAg compared with those aged 16–29 years (Table 3). Among people of White other ethnicity, compared to those aged 16–29 years, individuals aged 30–49 years were more likely to test positive, whereas those aged ≥ 50 years were less likely. We ran sensitivity analysis removing setting of test from the model and running separate models for primary and secondary care and identified little variation between the models.

3.3 HBV Prevalence Estimates

Table 4 shows HBV estimates aggregated by sex, age, ethnicity, and region (London, outside London) in England. Using population estimates from the 2021 census broken down by age, sex and ethnicity for England, we estimate the number of people with chronic hepatitis B in England at 268,767 (227,896–314,044), equivalent to a prevalence estimate of 0.58% (95% CI: 0.50–0.68) (Table 4). Chronic hepatitis B prevalence was higher in London 1.52% (95% CI 1.17–1.97) than outside London 0.39% (95% CI 0.32–0.48).

Among people with chronic hepatitis B, approximately two-thirds (64%) were male and one-third female (36%), with 68% aged under 50. We estimate that 83% of HBV infections are in people from ethnic groups other than White British, with 23% in people from Black ethnic groups, 21% from other White ethnic groups, 19% in Asian ethnic groups and 16% in other ethnic groups. HBV prevalence for men was highest among Black ethnic groups aged 30–49 years at 6.70% (95% CI 5.84–7.58) followed by those of other ethnic groups aged 30–49 years at 6.36% (95% CI 5.36–7.41). Among women, HBV prevalence was highest among by those of other ethnic groups aged ≥ 50 years at 5.61% (95% CI 4.71–6.69) followed by other ethnic groups aged 30–49 years 3.01% (95% CI 2.80–3.27). For both men and women, the lowest prevalence was among those of white British ethnicity regardless of age.

4.1 Main Findings

Based on public health surveillance data (between years 2015–2021) from antenatal and non-antenatal testing sources we estimate that prevalence of chronic hepatitis B was 0.58% (95% CI 0.50–0.68) or 268,767 individuals (228,000–314,000). Over 85% of positive HBsAg tests in both antenatal and non-antenatal sources and an estimated 83% of the total number of people were from ethnic groups other than White British. We also estimated that two-thirds of the people with chronic HBV were male, one-third female, and approximately 70% were aged under 50 years. Prevalence of chronic HBV in London was nearly four times higher than outside London (1.52% vs. 0.39%).

4.2 Other Evidence

The baseline of the model uses data on women tested antenatally whose testing results were also reported by laboratories participating in SSBBV. These data are from a subset of all women screened for HBV in the NHS infectious Diseases in Pregnancy Screening (IDPS) programme. Compared to all women screened for HBV in IDPS, test positivity in testing was slightly higher: 1.14 for London and 0.39 for outside London in SSBBV for 2020, versus 0.65% in London and 0.26% outside London in IDPS programme data for the financial year 2020/21 [17]. These prevalence estimates are similar to those reported by Hahne et al. of 0.45% in England and Wales [7], and higher than previous estimates using surveillance data [18] due to changes to the census data related to the distribution of the underlying populations. However, our estimates are lower than the Global Burden of Disease estimates of approximately 0.7% [19]. Further work is needed to refine the prevalence model through statistical models using a similar method to that used for hepatitis C and HIV [20-22]. Gaps in data needed for statistical modelling should be identified and addressed using cross-sectional data such as seroprevalence studies, with efforts to ensure the data are representative of the heterogeneous populations affected by HBV in England.

Compared to people of White British ethnicity, HBV prevalence estimates were higher in all other ethnic groups. The highest prevalence estimate were in those of Black ethnicity and Other ethnicity, both above the 2% threshold at which the WHO testing guidance suggest scale-up of testing, diversifying approaches to reach those unaware of their infection [23]. Estimates for London are over three times higher than for outside London, with estimates similar to the 1.1% test positivity (including those with a known infection) observed in five London NHS trusts offering opt-out blood-borne virus (BBV) testing in emergency departments [24].

4.3 Strengths and Limitations

There are several limitations. First, the data used are derived from a sentinel network of laboratories, where, although covering approximately 40% of England's GP population, coverage varies geographically. This may introduce bias if sites missing in public health surveillance and more recent data have different ethnic groups, rates of testing and rates of HBV. Nonetheless, we believe the potential for biases from this limitation is small as the sample sizes are large, a prominent diagnostic laboratory in each area is included within SSBBV and although risk-based testing rather than screening may occur in other settings, antenatal screening has almost 100% uptake. Second, the antenatal data for the seven year period 2015–2021 used as the baseline is a subset of all antenatal screening tests (793,463 women/~4,550,000 pregnancies) and so may not be fully representative. To explore how representative this baseline was, age, geographical distribution and ethnicity of screening results in SSBBV were compared to all live births [25]. Live birth data was used as demographic data is not published for the screening programme but with > 99% uptake it will be comparable with published live birth data. The age distribution among the subset was comparable with all births, however, there was some variation by region, with the West Midlands and the East of England underrepresented in SSBBV, and the East Midlands overrepresented. Due to the regional variability in SSBBV we present the prevalence estimates for London and outside London only and further work will be required to present additional regional breakdowns. While mothers ethnicity is not available in the live births data, the babies ethnicity is [26] and using this as a proxy for the mothers ethnicity the distribution was comparable, apart from those of mixed ethnicity which was lower among mothers in SSBBV compared to all babies, however, mixed ethnicity is difficult to compare as the babies ethnicity will represent the ethnicity of both parents. We also know that the positivity in our sample is slightly higher than the overall positivity reported by IDPS, and work is underway to include all women antenatally screened in further prevalence models. Third, ideally we would also have information on country of birth—which is a better indicator of HBV risk and might provide a better indicator of HBV prevalence within the antenatal sample and multipliers from the non-antenatal sample. This information could be used to test whether there were any systematic differences in ethnic background in people tested through antenatal programmes and non-antenatal settings. Unfortunately, current record linkages cannot provide country of birth yet for all tests. Fourth, we assume that age and gender ratios within the non-antenatal samples are unbiased—which also requires additional evidence to test and justify this assumption. Finally, the method also assumes that the numerators and denominators match, that is, that ethnic group reported to census is equivalent to reporting of ethnic group to hospitals—which also requires additional evidence to test this assumption.

4.4 Implications

Our estimates confirm that overall England has a low endemic prevalence of chronic HBV infection, though the number of people in England with chronic HBV is still relatively common and may be larger than the number of people with HIV or HCV [27, 28]. The higher prevalence of HBV in people from ethnic groups other than White British also indicate unmet need and associated equity issues. Compared to UK-born populations migrants experience a disproportionate burden of disease, with poorer access and outcomes to healthcare [29-31]. Barriers for migrants accessing testing and treatment include language, stigma, sociocultural factors and limited understanding of HBV infection and its harms [32, 33].

Due to the asymptomatic nature of hepatitis B infection, individuals can remain undiagnosed for many years until they develop advanced liver disease. Early exploration of estimating the diagnosed proportion using laboratory-confirmed hepatitis B diagnoses and other established health care datasets, suggest that 38% living with chronic hepatitis B in England were diagnosed, 36% for London and 40% outside London [34]. These estimates and the finding that, within the first year of the BBV opt-out testing in ED programme, hepatitis B had the highest number of new diagnoses identified [24] supports the understanding that there are large numbers of people with undiagnosed HBV. Early diagnosis should ensure access to care and treatment, and better outcomes. Extensive work has already been undertaken to increase testing coverage for hepatitis C, with novel case-finding interventions developed and implemented in the community [27]. Case finding approaches in primary care for both HCV and HBV has been shown to be effective and cost-effective but not yet fully implemented [35]. Offer of BBV testing in a range of settings is a key part of elimination efforts [6].

Additional evidence is needed to test some of the assumptions of the method and validate our estimates. However, our England HBV prevalence estimates are crucial in providing a baseline denominator from which to measure the coverage of HBV diagnosis and care, monitor HBV elimination targets and evaluate the impact of prevention and control strategies on hepatitis B burden.