We use a set of models representing the HBV epidemic in each of 110 LMICs, which together represent all six WHO regions and include all current GAVI-eligible countries (Fig. S1). Each country model was calibrated to data on HBsAg (Table S1) and HBeAg seroprevalence, HBV-related deaths due to cirrhosis and HCC, and national data on coverage of vaccinations and treatment (Figs. 1a–d). We apply a set of scenarios for the future coverage of the TBD in each country (Table 1) and quantify impact with reference to the infections and deaths that would occur in a scenario in which the coverage of TBD remains at the level recorded for each country in 2019. These results were aggregated across country-level simulations to generate results for the six WHO regions (Fig. S1) and global results. Figs. 1e–f show the global coverage of infant and TBD vaccination (aggregated over country-level coverage values, weighted by population size in 2025) under a selection of these scenarios (Fig. S2 gives the breakdown by WHO region for status quo).
The Impact of Scaling Up the Timely Birth dose
Fig. 2 shows the modelled impact of scaling up TBD vaccination to ≥90% by 2030 (scenario 2) on HBV disease burden globally. The scale-up results in immediate reductions in chronic HBsAg incidence and the prevalence of HBsAg among five-year-olds relative to the status quo scenario (scenario 1), greatly accelerating the gradual reductions that would be expected otherwise (Figs. 2a and 2b). The WHO elimination target (less than 0.1% HBsAg prevalence in five-year-olds by 2030) cannot be reached globally in the countries modelled (before 2100) without the scale-up of the TBD. However, with the modelled fast and substantial scale-up of TBD, that target can be reached in 2053 (2051 to 2056; Fig. 2b). Scaling up TBD vaccination to ≥90% by 2030 would avoid 43,000,000 (37,000,000 to 57,000,000) chronic infections relative to status quo between 2020 and 2100. This scale-up results in 770,000 (600,000 to 1,060,000) fewer deaths among those born between 2020 to 2030 globally (Table S3). However, this would not be recorded as a drop in deaths (Fig 2c) or DALYs (Fig 2d) in calendar-time until ~2050. This is because HBV usually takes decades to progress from infection to death. The expected trend in deaths and DALYs is an increase until 2030-2040, followed by a fall due to the competing effects of population ageing and growth (and so more people in older cohorts reaching ages when they are at risk of death caused by HBV) and the effect of the infant HBV vaccine series (reducing infections in younger cohorts).
Fig. 3 identifies the populations in which the effect of TBD scale-up in reducing deaths is most strongly concentrated. For cohorts born before 2030, the number of deaths averted rises in line with the assumed increases in coverage of the TBD. Later cohorts benefit somewhat less, as the risk of infection to them is lower as a result of the declining HBV prevalence among women of child-bearing age. Most (>70%) of the deaths averted in the 2020 to 2030 birth cohorts are in the WHO Africa region (AFRO), where HBV prevalence is very high in many countries (Fig. 1a) and current TBD coverage is the lowest of all the WHO regions (~7%: see Fig. S2). In fact, 26% of all deaths averted globally in the 2020 to 2030 birth cohorts would be in a single country—Nigeria—which has a large and rapidly growing population with very low TBD coverage currently (0% in 2019). Large numbers of deaths averted would also be expected in the Middle East (WHO’s EMRO region) and South-East Asia (WHO’s SEARO region), which also have relatively low TBD coverages (~34% and ~55%, respectively; Fig. S2), but lower HBV prevalence than AFRO (Fig. 1a). Reflecting their large sizes, 9% and 8% of deaths averted globally in the 2020 to 2030 birth cohorts would be in India and Pakistan, respectively (Fig. 3b). Results for each region and country individually are presented for selected scenarios in Tables S3 to S6.
Achieving HBV Elimination
Fig. 4a shows the year by which the WHO elimination goal would be achieved for the combined populations within each region under status quo (scenario 1) versus if the TBD is scaled up to ≥90% by 2030 (scenario 2). Without any TBD scale-up, elimination would be reached first in the Europe (WHO’s EURO region) in 2038, followed by the Western Pacific (WHO’S WPRO region) in 2042, in both of which there is already high TBD coverage and low prevalence of HBV. In contrast, in the WHO regions of EMRO and AFRO, elimination would not occur before 2100. The TBD scale-up would bring the date of elimination earlier in all regions (with the exception of EURO, where TBD is already at high scale) and make it possible in EMRO (in 2051) and AFRO (in 2059). With the TBD scale-up, by 2040, the regions EURO, PAHO and WPRO would all have reached the elimination target; and by 2050, all regions except AFRO would have reached the elimination target. Results for each country are given in Table S4.
Fig. 4b shows the relationship between the annual rate of scale-up of the TBD and the year by which elimination is reached in the populations within each of the WHO regions. Faster scale-up results in earlier elimination—especially where TBD coverage is lowest, in the AFRO, EMRO and SEARO regions—but no amount of accelerated scale-up would feasibly result in elimination being reached in any region by 2030 (the WHO’s target). Indeed, even an infeasibly high rate of scale-up in Africa (9% per year) would not bring the date of elimination to before 2060. This is because many countries in the region have either a very low TBD coverage (e.g. Nigeria), a very high HBV prevalence (e.g. South Sudan (22% prevalence among all ages), Sierra Leone (19%) Liberia (15%); http://whohbsagdashboard.com/#hbv-country-profiles, accessed in January 2021), or both. The Americas (WHO’s PAHO region) has the best chance of reaching the 2030 target—here moderate TBD coverage is combined with low overall HBV, so that an accelerate scale-up of TBD could have an effect quickly.
The Indirect Effects of the COVID-19 Pandemic
Fig. 5a shows that a temporary drop in TBD coverage in 2020 could result in 17,000 (13,000 to 23,000) additional HBV-related deaths, concentrated in the SEARO and WPRO regions, which have large population sizes and normally relatively high TBD coverage levels (55% and 91% in SEARO and WPRO, respectively; Fig. S2). The additional deaths occur mostly among those born in 2020 when the disruptions occurred, but also, to a lesser extent, to unprotected children (unvaccinated or vaccination failed) born in earlier and later years, who become infected following contact with a child born in 2020 (Fig. S5). The effect of the temporary disruption is long-lived—the additional deaths occur mostly from 2050 onwards (Fig. S6). However, the impact does not affect the year by which elimination is achieved (Tables S4 and S6).
Fig. 5b shows the effect that delays in scale-up of TBD could have (scenarios 4a to 4c) compared to the scale-up of TBD (scenario 2). The nature of these disruptions could be large: at worst (in scenario 4c), 630,000 (490,000 to 850,000) additional deaths globally in the 2020 to 2030 birth cohorts, concentrated mostly in AFRO. Even if TBD is scaled up faster after a period of delay between 2020 and 2023 (as is the case in scenario 4a), there would remain a significant excess number of deaths in the cohorts that missed out on vaccination. Delays in the scale-up of TBD lead to additional deaths that would occur mostly from 2060 onwards (Fig. S7).
Sensitivity analyses
Our results are sensitive to the assumptions made about other aspects of the HBV programme—the infant HBV vaccine series and treatment. Firstly, in the foregoing analyses it was assumed that infant vaccine series coverage levels will be maintained at the levels recorded in each country for 2019 as our primary aim was to explore the impact that TBD scale-up alone could have. However, increasing infant vaccine series coverage levels to 100% in every country leads to fewer infections (reduction of 35% globally) and deaths (reduction of 20% globally) being averted between 2020 and 2100 as a result of scaling up TBD to ≥90% by 2030 (scenario 2) and an earlier year of elimination being reached overall (2048 versus 2053; Table S4). This is because a somewhat higher coverage of infant vaccine series would result in a lower risk of infection for those who did not receive the TBD. The effects are strongest in the countries for which infant vaccination is currently lower (Tables S5 and S6). For example, in Nigeria, which has an infant vaccine series coverage of 57% in 2019, there are 28% fewer deaths averted in the 2020 to 2030 birth cohorts in scenario 2 relative to scenario 1 if infant vaccine series coverage is scaled up to 100% compared to if infant vaccine series coverage is maintained at status quo levels.
Secondly, whilst in the foregoing analyses it was assumed that the proportion of persons living with HBV that receive treatment remains at the same levels as those recorded in 2016 for each country, if instead treatment coverage were to increase to 40% or 80% of those eligible for treatment by 2030, then the impact of scaling up of the TBD to ≥90% by 2030 on deaths is greatly reduced (globally by 42% or 77%, respectively). Similarly, the effect of drops in TBD coverage relative to status quo is reduce (globally the effect of a 20% drop in TBD is reduced by 32% or 70%, respectively), as are the effects of delays in TBD coverage relative to scaling up the TBD to ≥90% by 2030 (globally the effect of scaling up TBD to ≥90% between 2025 and 2040 is reduced by 42% or 77%, respectively). This is because, with treatment, the risk of death following infection is reduced substantially.