Acronyms and abbreviations

ANC

antenatal care

BD

birth dose

DALY

disability-adjusted life year

DC

decompensated cirrhosis

ESLD

end-stage liver disease

GBD

Global Burden of Disease

GDG

Guidelines Development Group

GDP

gross domestic product

GHSS

Global Health Sector Strategy

HBeAg

hepatitis B e-antigen

HBIG

hepatitis B immune globulin

HBsAg

hepatitis B surface antigen

HBV

hepatitis B virus

HCC

hepatocellular carcinoma

HIV

human immunodeficiency virus

ICER

incremental cost–effectiveness ratio

PMTCT

prevention of mother-to-child transmission

PPT

peripartum antiviral therapy

SQ

status quo

WHO

World Health Organization

Background

Hepatitis B virus (HBV) infection contributes to a high global burden of disease and the 2016 World Health Organization (WHO) elimination targets of the Global Health Sector Strategy (GHSS) on viral hepatitis called for a reduction in incidence of new HBV infections by 90% by 2030. The only current WHO recommendation for prevention of mother-to-child transmission (PMTCT) of HBV is birth dose (BD) vaccination within 24 hours of birth. However, in light of the accumulating evidence on the benefit of adding peripartum antiviral therapy in reducing HBV MTCT further, a WHO Guidelines Development Group (GDG) is currently reviewing the evidence for the addition of such a strategy. As part of this process, an evaluation of the impact, costs and cost–effectiveness of a population-level testing and antiviral treatment strategy for pregnant women as an HBV PMTCT intervention was undertaken to help inform these guidelines and was presented to the GDG in September 2019.

Methods

A previously published dynamic age-, sex- and region-structured simulation model of the global HBV epidemic was adapted for the purposes of this analysis, to evaluate the regional impact and cost–effectiveness of scaling up peripartum antiviral treatment for pregnant mothers. The model has been described in full elsewhere.¹ In brief, the model is composed of 21 Global Burden of Disease (GBD) world regions, and is fit to data on hepatitis B surface antigen (HBsAg)² and hepatitis B e antigen (HBeAg)³ prevalence and liver cancer deaths⁴ for each region independently. The model incorporates national-level demographic data on fertility, mortality and population structure (UN Population Division) and national intervention coverage levels of infant vaccination and BD vaccination.⁵ Transmission and natural history parameters were taken from the literature. In the model, transmission occurs from mother to child, from child to child, and across the whole population. The relative strengths of each mode of transmission are inferred through the calibration procedure.

Strategies evaluated

Table 1 summarizes the main strategies evaluated.

In strategy 1, infant vaccination is scaled up to 90% (or continues at higher levels if over 90% coverage has already been achieved) – this is considered to be the “status quo” (SQ) scenario in this report.

In strategy 2, both infant vaccination and timely BD vaccination is scaled up to 90% (or higher if already achieved).

Strategies 3 and 4 consider the scale up of antiviral therapy for pregnant women (referred to as peripartum antiviral therapy or PPT). In both PPT strategies, all pregnant women are screened for chronic HBV infection using an HBsAg test. For strategy 3, all pregnant women who test positive for HBsAg subsequently have an HBV viral load test. Women with a high viral load (the cut-off threshold was assumed to be 200 000 IU, as per existing international guidelines) are treated with tenofovir. For strategy 4, all pregnant women who test positive for HBsAg have an HBeAg test (instead of an HBV viral load). Those found to be HBeAg positive are treated with tenofovir. The PPT interventions are scaled up by 2021 and are assumed to include up to 4 months of antiviral tenofovir and monitoring.

An important assumption used for the baseline analysis is that the PPT intervention package that was modelled was incremental to a dose of BD vaccination only, without the use of hepatitis B immune globulin (HBIG), which has associated cost and logistical issues and is not currently a universal WHO recommendation. However, the efficacy of an HBIG-free strategy with BD and PPT alone is currently unknown as all existing studies evaluating tenofovir are incremental to BD and HBIG (Shimakawa et al., systematic review). Therefore, for the purpose of this analysis, it is assumed that a BD and PPT strategy would have the same efficacy as a “triple intervention strategy”, which includes BD, HBIG and PPT, of 1% residual transmission in women with a high viral load (Shimakawa et al., systematic review).

In this analysis, it is assumed that only those HBsAg-positive women whose children get a BD will be able to benefit from a PPT intervention. Further theoretical strategies whose feasibility and efficacy have yet to be established, including the administration of PPT to those who do not receive a timely BD, were also explored but are not presented here.

We assumed that all pregnant women would receive one lifetime antenatal screening test for HBsAg. In contrast with human immunodeficiency virus (HIV), acquisition of new chronic HBV infection during adulthood is rare, therefore a one-off test was assumed likely to be sufficient in the presence of a previous negative test. This was calculated by adjusting the antenatal care (ANC) screening cost by the regional total fertility rate (UNPOP 2010–2015). Such a strategy should be accompanied by a process to ensure that each woman keeps a record of her HBV status (whether positive or negative) once screened throughout her childbearing years. The impact of repeated antenatal screening for HBsAg at each pregnancy was explored in the sensitivity analysis (see below).

Results

Impact

Compared to SQ, the scale up of BD vaccination has the largest incremental impact in terms of new infections and DALYs averted, in all world regions apart from regions where BD vaccination coverage is already over 90%. Globally, this strategy will avert 14 million new neonatal HBV infections and 38 500 DALYs over the next 10 years. From a longer-term health perspective, BD scale up will avert 40 million new infections and 122 million DALYS (to 2100). The three world regions where BD scale up will have the highest impact are South Asia (12.6 million cases and 8.8 million DALYs averted to 2100), West Africa (7.4 million cases and 4.6 million DALYs averted to 2100) and East Africa (4.3 million cases and 3.1 million DALYs averted to 2100). In the following regions BD scale up will also have a significant impact and avert over 1 million new cases to 2100; Southern Africa (3 million), South-East Asia (1.5 million) and Central Africa (1.4 million).

Compared to the scale up of BD vaccination, the subsequent addition of HBsAg testing and antiviral treatment of pregnant women would avert an additional 2.9–3 million neonatal infections over the next 10 years and, over the longer term (to 2100), would avert a further 6–7 million new neonatal infections or 22–25 million DALYs. However, the incremental impact of such a strategy is highly heterogeneous, depending on the world region. The regions where a PPT strategy is estimated to have the highest impact (incremental to BD scale up) are South Asia (1.6–1.8 million infections averted and 1.3–1.7 million DALYS averted to 2100) and West Africa (1.3–1.5 million infections averted and 790 000–920 000 DALYs averted to 2100)

Table 4(A)Summary of global impact – short term (to 2030)

View in own window

GLOBAL IMPACT (2020 - 2030, undiscounted)	Compared to SQ			Compared to BD
	New chronic infections Averted	New Neonatal infections Averted	DALYs Averted	New chronic infections Averted	New Neonatal infections Averted	DALYS averted
Infant Vacc + BD 90% by 2020	18,145,612	13,766,271	38,529	-	-	-
PPT guided by High VL (2021)	22,247,643	17,069,313	40,757	4,102,032	3,303,042	2,227
PPT guided by HBeAg (2021)	21,766,185	16,682,556	40,518	3,620,573	2,916,285	1,989

Table 4(B)Summary of global impact – long term (to 2100)

View in own window

GLOBAL IMPACT (2020 - 2100, undiscounted)	Compared to SQ			Compared to BD
	New chronic infections Averted	New Neonatal infections Averted	DALYs Averted	New chronic infections Averted	New Neonatal infections Averted	DALYS averted
Infant Vacc + BD 90% by 2020	56,835,741	40,582,235	122,171,349	-	-	-
PPT guided by High VL (2021)	66,601,188	47,598,754	147,213,941	9,765,446	7,016,519	25,042,592
PPT guided by HBeAg (2021)	65,246,027	46,631,438	143,886,134	8,410,286	6,049,203	21,714,785

Table 4(C)Main results summary by world region. ICERs are presented in US$ per DALY averted.*

The current model is constructed by GBD world region; however, in this table the GBD regions are also approximated to the nearest WHO region (see Appendix B for mapping of GBD regions to WHO regions). The regions with the largest discrepancy regarding Member States are the GBD South-East Asian region which has countries in both WHO WPRO and SEARO (and are therefore combined) and the GBD South Asia region, which has countries in both the WHO South-East Asia and Eastern Mediterranean regions. Therefore, this approximation should be applied with caution. The last two columns represent the cost–effectiveness results of the low and high diagnostic cost scenarios, highlighted in blue if it is strategy 3 (viral load-guided) or green for strategy 4 (HBeAg-guided)

View in own window

GBD region	Approximated WHO region*	ICER BD (compared to SQ)	ICER PPT guided by High VL (compared to BD)	ICER PPT guided by HBeAg (compared to BD)	Is VL or HBeAg guided more cost effective?	ICER difference between PPT strategies (USD)	ICER difference between PPT strategies (%age)	ICER with low diagnostic costs	ICER with high diagnostic costs
South Africa	AFRO	388	1491	1493	VL	2	0.1%	481	6653
West Africa	AFRO	242	1066	992	eAg	74	7.5%	357	4946
Central Africa	AFRO	285	1106	1037	eAg	69	6.7%	365	4950
East Africa	AFRO	349	1250	1220	eAg	30	2.5%	405	5461
North Africa & Middle East	EMRO	548	1798	2003	VL	205	11.4%	534	7550
Eastern Europe	EURO	166	2028	2110	VL	82	4.0%	590	8009
Western Europe	EURO	842	7355	7973	VL	618	8.4%	1973	28221
Central Europe	EURO	333	1069	1068	eAg	1	0.1%	971	13903
Central Asia	EURO	-	2168	2334	VL	166	7.7%	622	8479
Southern LA	PAHO	133	1271	1237	eAg	34	2.7%	401	5316
Tropical LA	PAHO	952	7320	8320	VL	1000	13.7%	1916	29392
Andean LA	PAHO	198	1849	1932	VL	83	4.5%	548	7505
Caribbean	PAHO	275	1908	2026	VL	118	6.2%	557	7650
Central LA	PAHO	791	2031	2071	VL	40	2.0%	1595	24133
North America	PAHO	854	6743	7389	VL	646	9.6%	1824	26236
South Asia	WPRO/SEARO	286	2227	2297	VL	70	3.1%	647	8733
SE Asia	WPRO/SEARO	314	2214	2319	VL	105	4.7%	639	8687
Asia Pacific High-Income	WPRO/SEARO	397	3194	3300	VL	106	3.3%	938	12678
Australasia	WPRO/SEARO	397	2885	2964	VL	79	2.7%	866	11734
East Asia	WPRO/SEARO	-	890	835	eAg	55	6.6%	293	3895
Oceania	WPRO/SEARO	223	1823	1852	VL	29	1.6%	549	7391

Costs and cost drivers

The global cost of scaling BD to 90% is estimated to be US$ 1.6 billion for 2020–2030. The additional costs of antenatal screening of pregnant women for HBsAg and providing antiviral treatment for those at high risk of HBV MTCT transmission would be an extra US$ 2.2–2.7 billion over 10 years, depending on the exact strategy adopted, with large interregional heterogeneity.

Fig. 1 demonstrates that for the PPT antiviral strategies (strategies 3 and 4), the cost of ANC screening with HBsAg contributes to the largest proportion of incremental costs. The costs of antiviral drug and further diagnostic tests (HBV viral load or HBeAg) account for a smaller percentage of total costs. In the low-cost scenario (see Fig. 1(B)), where the cost of HBsAg testing has been reduced to US$ 0.4 per test, the contribution of ANC screening to the total costs is significantly reduced. However, if the cost of HBsAg screening was higher and women were screened more than once during their childbearing lifetime, the ANC screening costs would significantly increase and impact on the overall total costs. Although this figure represents the results of the South-East Asian region, similar patterns of results are seen in other world regions (results not shown).

Fig. 1Costs and cost drivers of strategies (results from the South-East Asia region)

Cost–effectiveness

Compared to SQ, scaling up BD vaccination is the most cost-effective option that delivers the most health benefit for the lowest cost in all but two regions where BD vaccination coverage is already very high. The ICERs for this strategy vary between US$ 133 and US$ 952 per DALY averted depending on the world region. Eight world regions have ICERs of <US$ 300 per DALY averted.

The ICERs for an HBV DNA-guided antiviral screening and treatment strategy (strategy 3), compared to BD, are highly heterogeneous between world regions and vary between US$ 890 and US$ 7355 per DALY averted. The regions with the lowest ICERs for PPT scale up are East Asia, West Africa and Central Europe, Central Africa and East Africa with ICERs of US$ 890, US$ 1066, US$ 1069, US$ 1106 and US$ 1250 per DALY averted, respectively. The regions with the highest ICERs are western Europe, tropical Latin America and North America with ICERs of US$ 7355, US$ 7320 and US$ 6743 per DALY averted, respectively. At the baseline cost of diagnostics used in this analysis, the ICERs of an HBV viral load-(strategy 3) or HBeAg-guided strategy (strategy 4) are largely similar; in most regions, the ICER difference between the two strategies is less than US$ 200 (representing less than a 10% difference in ICER in most regions). The relative cost–effectiveness of each strategy is largely dependent on the relative costs of the two diagnostic modalities (Fig. 2 shows an example from South-East Asia).

Fig. 2Figure demonstrating how the relative costs of HBV viral load and HBeAg determine which strategy is more likely to be cost effective

Note that this figure represents only the strategy that dominates, rather than a comparison to a threshold. Therefore, a PPT intervention will not necessarily be cost effective at all costs presented in the figure. The red triangle represents the baseline cost combination used in this analysis (US$ 15 for viral load, US$ 7.5 for HBeAg) and the black triangles represent upper and lower cost bounds used. (Example figure for South-East Asia)

South-East Asia

The example of the South-East Asian Region will now be used to describe some of the results in further details. Table 5 and Fig. 3 summarize the main impacts, cost and cost–effectiveness results for this region, compared to both SQ and BD. In this region, compared to BD, the two PPT strategies (scenario 3: HBV viral load-guided or scenario 4: HBeAg-guided) have ICERs of US$ 2214 and US$ 2319 per DALY averted, respectively (range US$ 639–US$ 8687) or US$ 1654 and US$ 1737 per infection averted (range US$ 477–US$ 6490). Using midrange cost scenarios, the ICER (per DALY averted) is consistent with the estimated health opportunity costs in only three out of the 11 countries in the South-East Asia Region (Table 6), suggesting that a PPT intervention is likely to be cost effective in these countries. In the other seven countries in the Region, the estimated health opportunity costs are lower than the ICER, suggesting that a PPT intervention maybe not be cost effective at those costs. However, a PPT intervention is likely to be cost effective in all countries in this Region if there was access to lower costs of diagnostics as the ICERs of a PPT strategy would be reduced to US$ 639–US$ 672 per DALY averted. A PPT strategy guided by HBV viral load is slightly more cost effective than an HBeAg-guided strategy in this Region, but there is only a US$ 105 difference between the cost–effectiveness ratios. It should be noted that the superiority of the impact of an HBV viral load-guided strategy in this analysis is a reflection of our understanding of HBV MTCT, which assumes that MTCT is guided by viral load levels. Fig. 2 shows the combination of relative diagnostic costs that determine whether an HBV viral load-guided strategy or an HBeAg-guided strategy would be more cost effective.

The current base case analysis assumes once per lifetime screening test for HBsAg. Screening pregnant women for HBsAg at each pregnancy would decrease the cost–effectiveness of such a strategy and increase the ICERs to US$ 4298 (strategy 3) or US$ 4688 (strategy 4) per DALY averted.

Table 5Summary results table for the South-East Asia Region (see Appendix B for summary tables of each of the other GBD regions)

View in own window

REGION: South East Asia	Outcome Measure	Infant Vacc + BD 90% by 2020	PPT guided by High VL (2021)	PPT guided by HBeAg (2021)
Impact compared to SQ	Cases Averted	1,464,167	1,861,897	1,813,965
	Cases Averted (%age)	31	39	38
	DALYs Averted	1,167,841	1,464,989	1,429,807
Impact compared to BD	Cases Averted	-	397,730	349,798
	Cases Averted (%age)	-	17	15
	DALYs Averted	-	297,148	261,966
Total Costs	Mid Cost Scenario	1,304,043,544	1,961,877,869	1,911,558,355
	Low Cost Scenario	-	1,493,955,050	1,480,094,603
	High Cost Scenario	-	3,885,476,062	3,487,740,778
ICER ($ per DALY averted) - compared to SQ	Mid Cost Scenario	314	699	681
	Low Cost Scenario	-	380	380
	High Cost Scenario	-	2,012	1,784
ICER ($ per case averted)- - compared to SQ	Mid Cost Scenario	250	550	537
	Low Cost Scenario	-	299	299
	High Cost Scenario	-	1,583	1,406
ICER ($ per DALY averted) - compared to BD	Mid Cost Scenario	-	2,214	2,319
	Low Cost Scenario	-	639	672
	High Cost Scenario	-	8,687	8,336
ICER ($ per case averted) - compared to BD	Mid Cost Scenario	-	1,654	1,737
	Low Cost Scenario	-	477	503
	High Cost Scenario	-	6,490	6,243

Fig. 3Incremental cost and impacts of strategies under consideration. Left panel shows results compared to SQ. Right panel shows results compared to BD scale up (results for South-East Asia)

Table 6Table of countries in the South-East Asian Region, with GDP per capita and estimate of health opportunity costs. The GDP per capita (World Bank, World Development Indicators 2019)

View in own window

SOUTHEAST ASIA	GDP per Capita (2018 USD)	Estimate of Health Opportunity Cost
Cambodia	1,512	756
Indonesia	3,894	1947
Lao PDR	2,568	1284
Malaysia	11,239	5619
Maldives	10,224	5112
Myanmar	1,326	663
Philippines	3,103	1551
Sri Lanka	4,102	2051
Thailand	7,274	3637
Timor-Leste	2,036	1018
Vietnam	2,564	1282

The sensitivity analysis performed with the costs averted of management of ESLD confirm that increasing the costs of the management of DC and HCC lower the ICER. For the South-East Asian Region, if annual costs of DC and HCC are US$ 500 then the ICER for strategy 3 reduces from US$ 2214 to US$ 2180 per DALY averted, and to US$ 2044 per DALY averted if it was US$ 2500 per year. The alternative method using provisional WHO values revealed ICERs for South-East Asia of US$ 2125, US$ 2191 or US$ 1801 when regional median costs (US$ 1080 DC, US$ 2048 HCC), lower-range costs (US$ 281 DC, US$ 545 HCC) and higher-range costs (US$ 5066 DC and US$ 9432 HCC) were applied, respectively. However, for example, in the Central African Region, if the costs of DC and HCC are as high as US$ 13 945 and US$ 2572, which WHO estimates suggest, then such an intervention could even be cost-saving (Table 7). However, this would appear to warrant further investigation as the cost of HCC is 19-fold higher than the median value for that region.

Table 7

Sensitivity analysis: impact on cost–effectiveness of including costs of management of end-stage liver disease (ESLD) using uniform costs and provisional WHO costs (results shown for South East Asia). The cost of ESLD refers to annual costs for (more...)

Table 8 summarizes the impact and cost–effectiveness results of varying the transmission parameters, for the two PPT strategies (strategy 3 and 4). As the correlation between HBeAg and viral load increases, the relative difference in cost–effectiveness of the two PPT strategies reduces. Having a higher overall percentage of all HBV-infected persons or HBeAg-negative persons with a high viral load who would benefit from antiviral therapy also improves the cost–effectiveness of a viral load-guided PPT strategy. If a BD + PPT strategy had a lower effectiveness, this would reduce the impact of a PPT strategy and increase the ICER to US$ 2801–US$ 2931 per DALY averted if residual transmission was 5% or to US$ 4196–US$ 4379 if residual transmission was 10%.

Table 8Results of sensitivity analysis on varying transmission parameters. 3% discounting of costs and health outcomes, long-term time horizon (results for the South-East Asia Region)

View in own window

Scenarios	Cases Averted	DALYs averted	ICER ($ per DALY averted)	ICER ($ per case averted)
Assumption Set 1
Fraction High VL: HBeAg pos (0.9), HBeAg neg (0.05)
PPT guided by High VL (in 2021)	397,730	297,148	2,214	1,654
PPT guided by HBeAg (in 2021)	349,798	261,966	2,319	1,737
Assumption Set 2
Fraction High VL: HBeAg pos (0.83), HBeAg neg (0.13)
PPT guided by High VL (in 2021)	483,636	359,226	1,851	1,375
PPT guided by HBeAg (in 2021)	354,891	264,769	2,296	1,713
Assumption Set 3
Fraction High VL: HBeAg pos (0.96), HBeAg neg (0.07)
PPT guided by High VL (in 2021)	416,469	310,740	2,129	1,589
PPT guided by HBeAg (in 2021)	351,759	263,223	2,308	1,727
Assumption Set 4
Fraction High VL: HBeAg pos (1.0), HBeAg neg (0.001)
PPT guided by High VL (in 2021)	359,153	269,326	2,442	1,831
PPT guided by HBeAg (in 2021)	350,220	262,759	2,311	1,734
Assumption Set 5
Lower BD + PPT efficacy (5% residual transmission)
PPT guided by High VL (in 2021)	315,363	235,052	2,801	2,088
PPT guided by HBeAg (in 2021)	277,470	207,386	2,931	2,191
Assumption Set 6
Lower BD + PPT efficacy (10% residual transmission)
PPT guided by High VL (in 2021)	211,391	157,087	4,196	3,118
PPT guided by HBeAg (in 2021)	186,302	138,896	4,379	3,265

The current baseline analysis assumed that a tenofovir disoproxil fumarate (TDF) intervention had the same efficacy as a BD, HBIG and TDF intervention. However, if HBIG was used in combination with TDF, the ICER of strategy 3 (compared to BD only) would increase by over US$ 1000 to US$ 3325 per DALY averted (if HBIG was US$ 50 per dose) or over US$ 2000 to US$ 4436 if HBIG was US$ 100 per dose (Fig. 4).

Fig. 4Costs and cost drivers of strategies, including cost of HBIG (results for the South-East Asia Region)

The choice of discount rate has a large impact on ICER (Table 9). If costs and health benefits are undiscounted, the ICER reduces to US$ 1070 per DALY averted. A shorter 10-year time horizon (undiscounted) reveals a lower cost per infection averted of US$ 828–US$ 870.

Table 9Sensitivity analysis on various combinations of discount rate

View in own window

Scenario	Cost ($) per DALY averted
	0%, 0%	3%, 3%	6%, 6%	6%, 0%	0%, 6%
PPT guided by High VL (in 2021)	1,070	2,214	5,626	203	29,690
PPT guided by HBeAg (in 2021)	1,169	2,319	5,686	206	32,255

Discussion

Scaling up BD vaccination will have the largest impact for the lowest cost and would therefore be the most cost-effective strategy compared to SQ in most world regions. This is consistent with previous studies showing the impact and cost–effectiveness of BD vaccination,^1,9,10 and supports existing WHO HBV PMTCT recommendations for a universal BD vaccination policy.

Our study shows that a PPT intervention will have an impact on new cases averted, including neonatal infections averted and DALYs averted. However, there is inter- and intraregional heterogeneity as to whether such a strategy would be considered cost effective using the middle range of diagnostic costs. The cost–effectiveness of a PPT strategy is largely influenced by the costs of ANC screening and the number of times a woman is screened during her lifetime, the efficacy of antiviral therapy in addition to BD (i.e. HBIG-free strategy), whether HBIG is used and its associated cost and the costs of management of ESLD. Further health economic evaluation characteristics, including choice of discount rate and time horizon taken, also affect the ICER.

Importantly, our analysis has revealed that the cost of diagnostics, particularly HBsAg tests, contribute more to the total cost of a PPT intervention than the cost of the antiviral drug. Therefore, further reduction in the cost of HBsAg tests from the currently available price of US$ 1.6 is needed in order to improve the cost–effectiveness of a PPT strategy. Our base-case scenario assumes that each pregnant woman has one ANC HBsAg screening test per childbearing lifetime, rather than a test repeated at each pregnancy. However, in practice, there might be a bias towards retesting at each pregnancy, for example, due to movement between clinics, lack of records being kept or health-care worker preference, which would make a screening and PPT strategy less cost effective, particularly in low-burden settings.

Under the current assumptions and given the available data, HBV viral load-guided or HBeAg-guided strategies have similar cost–effectiveness ratios and the choice of strategy would depend on the relative cost of diagnostics available and local-level considerations, including access to diagnostics and laboratory facilities. Further consideration may be appropriate about how different strategies might be adopted depending on urban or rural settings.

It is important to note that regional-level results provide some guide for policy-making, although they can give only a broad indication as to whether it may be more or less likely that a particular strategy would be cost effective under the simplifying assumptions adopted. Given the heterogeneity of epidemiological, cost and health opportunity costs of countries within regions, averaging across countries within a region obscures this variability. Therefore, regional analyses cannot directly inform decisions about the allocation of resources at the local level where the value is realized, and does not replace the need for careful analysis using local data in each country.

It should be noted that not all countries currently have access to diagnostics at the costs modelled. GeneXpert HBV viral load test kits are available for US$ 15 and are currently being validated in field settings. However, local consideration would need to be given as to the availability and capacity of the existing platforms to integrate HBV testing. Furthermore, HBeAg tests have, so far, been shown to have poor diagnostic performance.¹¹ In the absence of the development of a cheap and accurate HBeAg rapid test, if countries remain reliant on the use of laboratory-based HBeAg testing, this might limit the advantage of using an HBeAg-guided strategy over a viral load-guided one to guide PPT.

Although the current analysis aims to be as robust as possible, there are many data gaps that limit the study. Primarily there remains uncertainty about the epidemiology, transmission and efficacy of interventions, particularly in the sub-Saharan African region.¹² Importantly, there are currently no data on the efficacy of an HBIG-free (BD and PPT only) PMTCT intervention (Shimakawa et al., systematic review 2019). All studies that have demonstrated the efficacy of PPT have used PPT in addition to BD and HBIG. However, HBIG does not currently form part of WHO recommendations as the widespread use of HBIG is thought to be unfeasible in many settings due the need for a cold chain, problems with lack of availability, high cost and concerns around the use of blood products.¹³ The results of the ongoing study in Lao People’s Democratic Republic and Thailand evaluating the efficacy of a BD and PPT strategy will be useful to refine model projections (https://clinicaltrials.gov/ct2/show/NCT03343431), as our analysis has shown that the cost–effectiveness is sensitive to the effectiveness of such a strategy in reducing HBV MTCT. Furthermore, although there are data on the proportion of women with a high HBV viral load, by HBeAg status, the number of high-quality studies outside the WHO Western Pacific region is limited (Shimakawa et al., systematic review).

Currently accurate regional data on the costs of management of HBV-related decompensated cirrhosis and liver cancer are limited, and many persons in low–middle-income countries often present late when therapeutic options are limited. Our baseline analysis excludes the costs averted of the management of ESLD and therefore takes a conservative view of cost–effectiveness. However, the sensitivity analysis has revealed that country-specific information on the costs of management of these conditions would impact on whether such an intervention was cost effective or not. Therefore, further empirical data on resource utilization and the costs of managing DC and HCC will be useful to more accurately assess the cost–effectiveness of HBV interventions at a local level. Furthermore, we have not taken into account programme costs, management and overhead costs or that of human resources, thereby underestimating the true total costs of implementing a national HBsAg screening and treatment strategy. Whether ANC screening and PPT could be integrated into existing services, e.g. HIV services, needs further research. Additionally, this analysis has evaluated the cost–effectiveness of PMTCT strategies, assuming that there will not be a large scale up in antiviral therapy of chronic HBV carriers and in the absence of a cure, which would overestimate the cost–effectiveness.

In summary, this study has shown that the most cost-effective PMTCT intervention is to scale up BD in all regions where it remains suboptimal. Incremental to birth scale up, a PPT strategy might be cost effective in some regions/countries but not others and careful local-level consideration needs to be given as to how such a strategy is implemented. Diagnostic costs, particularly HBsAg screening costs, the effectiveness of an “HBIG-free” PPT strategy and the costs of management of ESLD are large drivers of cost–effectiveness. Caution must be taken in interpreting these results as the data on the epidemiology and transmission of HBV MTCT are limited in some regions, particularly in sub-Saharan Africa, and research must be targeted to address these knowledge gaps.

Selected references

1.

Nayagam S, Thursz M, Sicuri E, et al. Requirements for global elimination of hepatitis B: a modelling study. Lancet Infect Dis. 2016;16(12):1399–408. [PubMed: 27638356]

2.

30. Ott JJ, Stevens GA, Groeger J, Wiersma ST. Global epidemiology of hepatitis B virus infection: new estimates of age-specific HBsAg seroprevalence and endemicity. Vaccine. 2012;(12):2212–9. [PubMed: 22273662]

3.

12 Ott JJ, Stevens GA, Wiersma ST. The risk of perinatal hepatitis B virus transmission: hepatitis B e antigen (HBeAg) prevalence estimates for all world regions. BMC Infect Dis. 2012;:131. [PMC free article: PMC3478174] [PubMed: 22682147]

4.

GLOBOCAN 2012: estimated cancer incidence, mortality and prevalence worldwide in 2012. Lyon: International Agency for Research on Cancer, World Health Organization; 2015.

5.

WHO/UNICEF Estimates of National Immunization Coverage (WUENIC). Geneva: World Health Organization (http://apps​.who.int/immunization​_monitoring​/globalsummary/timeseries​/tscoveragehepb_bd.html, accessed 14 February 2020).

6.

Woods B, Revill P, Sculpher M, Claxton K. Country-level cost-effectiveness thresholds: initial estimates and the need for further research. Value Health. 2016;19(8):929–35. [PMC free article: PMC5193154] [PubMed: 27987642]

7.

Tordrup D, Hutin Y, Stenberg K, et al. Additional resource needs for viral hepatitis elimination through universal health coverage: projections in 67 low-income and middle-income countries, 2016&#x2013;30. Lancet Glob Health. 2019;7(9):e1180–e8. [PubMed: 31353061]

8.

Wen W-H, Chang M-H, Zhao L-L, et al. Mother-to-infant transmission of hepatitis B virus infection: Significance of maternal viral load and strategies for intervention. J Hepatol. 2013;59(1):24–30. [PubMed: 23485519]

9.

Klingler C, Thoumi AI, Mrithinjayam VS. Cost-effectiveness analysis of an additional birth dose of Hepatitis B vaccine to prevent perinatal transmission in a medical setting in Mozambique. Vaccine. 2012;31(1):252–9. [PubMed: 22902676]

10.

Reardon JM, O’Connor SM, Njau JD, Lam EK, Staton CA, Cookson ST. Cost-effectiveness of birth-dose hepatitis B vaccination among refugee populations in the African region: a series of case studies. Conflict and Health. 2019;13:5. [PMC free article: PMC6390570] [PubMed: 30858875]

11.

Seck A, Ndiaye F, Maylin S, et al. Poor sensitivity of commercial rapid diagnostic tests for hepatitis B e antigen in Senegal, West Africa. Am J Trop Med Hyg. 2018;99(2):428–34. [PMC free article: PMC6090320] [PubMed: 29869595]

12.

Keane E, Funk AL, Shimakawa Y. Systematic review with meta-analysis: the risk of mother-to-child transmission of hepatitis B virus infection in sub-Saharan Africa. Aliment Pharmacol Ther. 2016;44(10):1005–17. [PubMed: 27630001]

13.

Spearman CW, Afihene M, Ally R, et al. Hepatitis B in sub-Saharan Africa: strategies to achieve the 2030 elimination targets. Lancet Gastroenterol Hepatol. 2017;2(12):900–9. [PubMed: 29132759]

Appendices