PREVENTION OF MOTHER-TO-CHILD TRANSMISSION OF HEPATITIS B VIRUS: Guidelines on antiviral prophylaxis in pregnancy July 2020

i

PREVENTION OF MOTHER-TO-CHILD TRANSMISSION OF HEPATITIS B

VIRUS: Guidelines on antiviral prophylaxis in pregnancy

July 2020

Web Annex C: Impact and cost–effectiveness of HBV peripartum antiviral therapy

Shevanthi Nayagam, Imperial College, London Timothy Hallett, Imperial College, London Yusuke Shimakawa; Pasteur Institute, Paris Nora Schmit, Imperial College, London Maud Lemoine, Imperial College, London Mark Thursz, Imperial College, London

ii

Prevention of mother-to-child transmission of hepatitis B virus (HBV): guidelines on antiviral prophylaxis in pregnancy. Web Annex C. Impact and cost–effectiveness of HBV peripartum antiviral therapy. Nayagam S et al.

ISBN 978-92-4-000864-9 (electronic version)

© World Health Organization 2020

Some rights reserved. This work is available under the Creative Commons Attribution-NonCommercial- ShareAlike 3.0 IGO licence (CC BY-NC-SA 3.0 IGO; https://creativecommons.org/licenses/by-nc-sa/3.0/igo).

Under the terms of this licence, you may copy, redistribute and adapt the work for non-commercial purposes, provided the work is appropriately cited, as indicated below. In any use of this work, there should be no suggestion that WHO endorses any specific organization, products or services. The use of the WHO logo is not permitted. If you adapt the work, then you must license your work under the same or equivalent Creative Commons licence. If you create a translation of this work, you should add the following disclaimer along with the suggested citation: “This translation was not created by the World Health Organization (WHO). WHO is not responsible for the content or accuracy of this translation. The original English edition shall be the binding and authentic edition”.

Any mediation relating to disputes arising under the licence shall be conducted in accordance with the mediation rules of the World Intellectual Property Organization (http://www.wipo.int/amc/en/mediation/rules/).

Suggested citation. Web Annex C. Nayagam S, Hallet T, Schmit N, Shimakawa Y, Lemoine M, Thursz M.

Impact and cost–effectiveness of HBV peripartum antiviral therapy. In: Prevention of mother-to-child transmission of hepatitis B virus (HBV): guidelines on antiviral prophylaxis in pregnancy. Geneva: World Health Organization; 2020. Licence: CC BY-NC-SA 3.0 IGO.

Cataloguing-in-Publication (CIP) data. CIP data are available at http://apps.who.int/iris.

Sales, rights and licensing. To purchase WHO publications, see http://apps.who.int/bookorders. To submit requests for commercial use and queries on rights and licensing, see http://www.who.int/about/licensing.

Third-party materials. If you wish to reuse material from this work that is attributed to a third party, such as tables, figures or images, it is your responsibility to determine whether permission is needed for that reuse and to obtain permission from the copyright holder. The risk of claims resulting from infringement of any third- party-owned component in the work rests solely with the user.

General disclaimers. The designations employed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of WHO concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted and dashed lines on maps represent approximate border lines for which there may not yet be full agreement.

The mention of specific companies or of certain manufacturers’ products does not imply that they are endorsed or recommended by WHO in preference to others of a similar nature that are not mentioned. Errors and omissions excepted, the names of proprietary products are distinguished by initial capital letters.

All reasonable precautions have been taken by WHO to verify the information contained in this publication.

However, the published material is being distributed without warranty of any kind, either expressed or implied.

The responsibility for the interpretation and use of the material lies with the reader. In no event shall WHO be liable for damages arising from its use.

The named authors alone are responsible for the views expressed in this publication. This publication forms part of the WHO guideline entitledPrevention of mother-to-child transmission of hepatitis B virus (HBV):

guidelines on antiviral prophylaxis in pregnancy. It is being made publicly available for transparency purposes and information, in accordance with the WHO handbook for guideline development, 2nd edition (2014).

1 Contents

ACRONYMS AND ABBREVIATIONS ...2

BACKGROUND ...3

METHODS ...3

STRATEGIES EVALUATED ...3

ECONOMIC ANALYSIS ... 4

OUTCOME MEASURES ... 5

COSTINGS ... 5

SENSITIVITY ANALYSES ... 5

RESULTS ...7

IMPACT ... 7

DISCUSSION ... 14

SELECTED REFERENCES... 17

APPENDICES... 18

APPENDIX A:REGIONAL SUMMARY RESULTS (SEE APPENDIX B FOR APPROXIMATE MAPPING OF GBD TO WHO REGIONS*) ... 18

APPENDIX B:APPROXIMATE MAPPING OF GBD REGIONS TO WHO REGIONS ... 29

GIVEN THE OVERLAP BETWEEN THE WESTERN PACIFIC AND SOUTH-EAST ASIAN REGIONS, THESE REGIONS ARE APPROXIMATED TOGETHER FOR THE RESULTS. ... 29

APPENDIX C:WHO ESTIMATES ON COSTS OF MANAGEMENT OF END-STAGE LIVER DISEASE (ESLD), SUMMARIZED BY REGION ... 30

2

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

3

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.

4

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.

Table 1. Main strategies modelled.*The coverage of PPT is only among those who get BD vaccination

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).

Economic analysis

The economic analysis was performed using a health provider perspective. All costs and health outcomes were discounted at 3% and a long-term time horizon was used to 2100, although this was varied in the sensitivity analysis. Estimated health opportunity costs are represented by half the gross domestic product (GDP) per capita of the country.⁶

The incremental cost–effectiveness ratios (ICERs) were calculated compared to two different baseline strategies;

1. Compared to the “status quo”: this represents 90% infant vaccination coverage (or higher if already achieved) and SQ levels of BD vaccination coverage. Comparison to this baseline scenario allows consideration of both strategies; BD scale up alone or BD and PPT scale up.

Appraisal of these options is useful to help inform decisions in countries/regions that are yet to scale up BD vaccination to high levels.

For all Pregnant women

For women who test HBsAg +ve

1 Infant Vacc to 90% by 2020 90 SQ - - -

2 Infant Vacc + BD to 90% by 2020 90 90 - - -

3 2 + PPT guided by High VL (2021) 90 90 HBsAg HBV VL High VL*

4 2 + PPT guided by HBeAg (2021) 90 90 HBsAg HBeAg Eag positive*

Eligibility for antiviral therapy Strategy

Number Description Infant Vacc

coverage

BD vacc coverage

Diagnostic Tests Used

5

2. Compared to “BD vaccination”: this represents 90% coverage of both infant vaccination (or higher if already achieved) and 90% BD vaccination coverage (or higher if already achieved).

Comparison to this baseline will allow countries to consider what to do next after scaling up BD vaccination and whether a PPT intervention would be cost effective.

Outcome measures

The primary outcome measures are new chronic infections averted, disability-adjusted life years (DALYs) averted and total costs. Cost–effectiveness results are presented as ICERs: (i. cost (US$) per DALY averted and (ii) cost (US$) per infection averted. Results are presented both relative to SQ and BD, as described above.

Costings

The costs (Table 2) in this analysis include the costs of the diagnostic tests (HBsAg test, HBeAg test and HBV viral load test), the cost of antiviral therapy for all those who fulfilled eligibility criteria and cost of treatment monitoring. Although it is assumed that rapid point-of-care tests would be used for HBsAg antenatal screening, the exact type of diagnostic test used is not specified but is expected to include one with high diagnostic performance and low cost, and suitable for use at a population level.

Peripartum antiviral treatment is assumed to include at least four months of antiviral tenofovir and monitoring. Where available, the costs are taken from recent work by WHO on costing elimination of hepatitis testing and treatment.⁷ Expert opinion was sought for other costs and ranges. The costs are the same for all world regions and the baseline costs were aimed at representing a price available if countries were to purchase the drug through optimal procurement. The cost of human resources, programme costs and overheads were excluded for the purpose of this analysis. All costs are presented in US$ (2019) and discounted at 3%.

Table 2. Main intervention costs

Sensitivity analyses

The following sensitivity analyses were performed to demonstrate the impact of uncertainty in underlying parameters on the results.

i. Costs

Costs of intervention

COSTS (USD) Baseline

analysis Low High

HBsAg Test 1.6 0.4 2.8

HBeAg Test 7.5 3 40

HBV Viral Load 15 5 100

Peripartum treatment: Drug 10 - -

Peripartum treatment: monitoring 10 5 40

Birth Dose Vaccination 1 - -

Infant Vaccination 1 - .-

6

Three cost scenarios were evaluated, referred to in this analysis as mid-, low- and high-cost scenarios (Table 2 outlines the ranges used). For the high-cost scenario, we also assumed that some women would have more than one HBsAg screening test per childbearing lifetime. This would, for example, take into account repeat HBsAg testing due to health provider preference or where for administrative reasons a record of the previous testing result was not available at the time of re-attendance at ANC.

We also evaluated which combination of HBeAg and HBV viral load diagnostic costs would determine which PPT strategy – HBV viral load-guided (strategy 3) or HBeAg-guided (strategy 4) – would be more cost effective.

Costs averted of management of end-stage liver disease

The costs to the health system of managing end-stage HBV-related liver disease (ESLD) (defined as decompensated cirrhosis [DC] and hepatocellular carcinoma [HCC]), are currently highly uncertain. Therefore, for the baseline analysis, a conservative estimate of cost–effectiveness was taken by excluding these averted costs.

However, we also explored the impact of including these costs in the sensitivity analysis as, given the long-term time perspective, it is useful to consider the costs averted. In light of limited empirical data on the costs of management of ESLD, and between-regional patterns that are hard to determine and likely obscure significant within-country variation, we have chosen to primarily present results with respect to a wide uniform uncertainty on the costs of ESLD. We did this by considering two scenarios where the annual costs were either US$ 500 or US$ 2500 per year, assuming that the costs of DC were equal to those of HCC and applied the same values for each region. However, WHO has recently produced provisional estimates on the cost of ESLD for 166 countries (Tordrup et al. unpublished); therefore, we have also used these in a supplemental analysis. In this case we took the WHO-estimated median, lower and upper range of costs for each GBD region (see Appendix C).

ii. Transmission parameters

The current understanding of the epidemiology of HBV MTCT is that the rate of HBV MTCT is related to HBV viral load.⁸ Combinations of the fraction of pregnant women who were HBeAg positive and HBeAg negative with a high HBV viral load were varied, as shown in Table 3. These ranges take into account the global averages from the recent systematic review for PICO2 (Shimakawa et al., systematic review 2019). Currently the efficacy and effectiveness of a BD and PPT strategy without HBIG is unknown. For this analysis, it is assumed that a BD and PPT strategy would have the same efficacy as a “triple PMTCT strategy”, which includes BD, HBIG and PPT, of 1% residual transmission in women with a high viral load (Shimakawa et al., systematic review 2019). However, given than the effectiveness of a BD and PPT strategy (without HBIG) is unknown, we varied this parameter over a large range in our sensitivity analysis; increasing it to 5%

(assumption set 5) and 10% (assumption set 6). The high value of 10% would allow us to account for the impact of both lower efficacy and effectiveness, including the impact of low adherence to treatment.

Table 3. Assumption sets used for sensitivity analysis on transmission parameters

7

In the sensitivity analysis, we also considered a scenario where PPT was given as a package with HBIG and compared this to a baseline of BD only. We considered HBIG costs of US$ 50 and US$

100 (WHO GDG document).

iv. Discount rates and time horizon

In keeping with guidelines on cost–effectiveness analysis, sensitivity analyses were performed on the discount rate used by using five combinations of discount rate between 0% and 6%. The impact of a shorter time horizon was also evaluated.

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)

Assumption

Sets Description FractionofHBeAg +vewithHighVL

FractionofHBeAg- vewithHighVL

MTCTwithBD+PPT inthosewithHighVL

1 Baselineparameters 0.9 0.05 0.01

2 VaryingFractionHighVL:GlobalaverageforHBeAg+ve 0.83 0.13 0.01 3 VaryingFractionHighVL:GlobalaverageforHBeAg-ve 0.96 0.07 0.01 4

VaryingFractionHighVL:HighercorrelationbetweenVL

andHBeAg 1 0.001 0.01

5 LowerefficacyofBD+PPT 0.9 0.05 0.05

6 LowerefficacyofBD+PPT 0.9 0.05 0.1

8

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

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)

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

GLOBAL IMPACT (2020 - 2100, undiscounted)

Compared to SQ Compared to BD

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

Newchronic infections

Averted

NewNeonatal infections

Averted

DALYs Averted

Newchronic infections

Averted

New Neonatal infections Averted

DALYS averted

InfantVacc+BD90%by2020 18,145,612 13,766,271 38,529 - - - PPTguidedbyHighVL(2021) 22,247,643 17,069,313 40,757 4,102,032 3,303,042 2,227 PPTguidedbyHBeAg(2021) 21,766,185 16,682,556 40,518 3,620,573 2,916,285 1,989

GLOBALIMPACT (2020-2030,undiscounted)

ComparedtoSQ ComparedtoBD

9

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. 1. Costs 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

10

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. 2. Figure 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

A viral load-guided strategy is more more likely to be cost effective (strategy 3)

An HBeAg-guided strategy is more likely to be cost effective (strategy 4)

11

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

Fig. 3. Incremental 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)

REGION: South East Asia Outcome Measure

Infant Vacc + BD 90%

by 2020

PPT guided by High VL (2021)

PPT guided by HBeAg (2021)

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

Cases Averted - 397,730 349,798

Cases Averted (%age) - 17 15

DALYs Averted - 297,148 261,966

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

Mid Cost Scenario 314 699 681

Low Cost Scenario - 380 380

High Cost Scenario - 2,012 1,784

Mid Cost Scenario 250 550 537

Low Cost Scenario - 299 299

High Cost Scenario - 1,583 1,406

Mid Cost Scenario - 2,214 2,319

Low Cost Scenario - 639 672

High Cost Scenario - 8,687 8,336

Mid Cost Scenario - 1,654 1,737

Low Cost Scenario - 477 503

High Cost Scenario - 6,490 6,243

ICER ($ per case averted) - compared to BD Impact compared to SQ

Impact compared to BD

Total Costs

ICER ($ per DALY averted) - compared to SQ ICER ($ per case averted)- -

compared to SQ ICER ($ per DALY averted) -

compared to BD

0 1 2 3 4 5 6 7 8 9 10

Cost (USD) vs SQ 10⁸

0 2 4 6 8 10 12 14

DALYs Averted vs SQ

10⁵ Incremental Costs and Impacts compared to SQ (to 2100)

Infant Vacc + BD 90% by 2020 PPT guided by High VL (2021) PPT guided by HBeAg (2021) Southeast Asia

0 1 2 3 4 5 6

Cost (USD) vs BD ₁₀⁸

0 0.5 1 1.5 2 2.5

DALYs Averted vs BD

10⁵ Incremental Costs and Impacts compared to BD (to 2100) PPT guided by High VL (2021)

PPT guided by HBeAg (2021)

Southeast Asia

12

Table 6. Table 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)

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 DC HCC. ICER is presented in US$ per DALY averted.

(A) Method 1: uniform costs (results for the South-East Asia Region)

(B) Method 2: using provisional WHO costs for selected countries in the Region (results for the South-East Asia Region)

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

Estimate of Health Opportunity Cost GDP per Capita

(2018 USD) SOUTHEAST ASIA

Strategy Costs of ESLD = $500 Costs of ESLD = $2500 Costs of ESLD = $0 (ie basecase)

PPT guided by High VL (2021) 2,180 2,044 2,214

PPT guided by HBeAg (2021) 2,285 2,149 2,319

13

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 8. Results 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)

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).

Strategy Costs of ESLD median Costs of ESLD highest Costs of ESLD lowest

PPT guided by High VL (2021) 2,125 1,801 2,191

PPT guided by HBeAg (2021) 2,230 1,906 2,296

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

14

Fig. 4. Costs 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 9. Sensitivity analysis on various combinations of discount rate

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

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

Scenario Cost ($) per DALY averted

15

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

16

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.