key: cord-0746089-2i1d5z6y
authors: Clark, R. A.; Mukandavire, C.; Portnoy, A.; Weerasuriya, C. K.; Deol, A.; Scarponi, D.; Iskauskas, A.; Bakker, R.; Quaife, M.; Malhotra, S.; Gebreselassie, N.; Zignol, M.; Hutubessy, R. C. W.; Giersing, B.; Jit, M.; Harris, R. C.; Menzies, N. A.; White, R. G.
title: The impact of alternative delivery strategies for novel tuberculosis vaccines in low- and middle-income countries: a modelling study
date: 2022-04-17
journal: nan
DOI: 10.1101/2022.04.16.22273762
sha: 754a453cdfe3e98477b2f567e5f829fa993fa98d
doc_id: 746089
cord_uid: 2i1d5z6y

Background Tuberculosis is a leading infectious cause of death worldwide. Novel vaccines will be required to reach global targets and reverse setbacks from the COVID-19 pandemic. We estimated the impact of novel tuberculosis vaccines in low- and middle-income countries (LMICs), under alternative delivery scenarios. Methods We calibrated a tuberculosis model to 105 LMICs (93% of global incidence). Vaccine scenarios were implemented as Basecase: routine vaccination of 9-year-olds and a one-time vaccination campaign for ages [≥]10 with country-specific introduction between 2028[ndash]2047 and 5-year scale-up to target coverage; Accelerated Scale-up: as Basecase, but all countries introducing in 2025 with instant scale-up; and Routine Only: as Basecase, but routine vaccination only. Vaccines protected against disease for 10-years, with 50% efficacy. Findings The Basecase scenario reduced tuberculosis incidence (19.5% [95% uncertainty range=18.3-21.6%]) and mortality (20.6% [19.2-23.4%]) rates in 2050 and prevented 3.6 (3.3-3.9) million deaths before 2050, including 1.6 million in the WHO South-East Asian region. The Accelerated Scale-up scenario reduced tuberculosis incidence (25.2% [23.9-27.5%]), mortality (26.7% [25.2-29.9%]), and prevented 7.9 (7.3-8.5) million deaths. The Routine Only scenario reduced tuberculosis incidence (9.9% [9.0-11.6%]), mortality (9.9% [8.9-12.3%]), and prevented 1.1 (0.9-1.2) million deaths. Interpretations Novel tuberculosis vaccines could have substantial impact, which will vary depending on delivery strategy. Including a campaign will be crucial for rapid impact. Accelerated introduction similar to the pace of COVID-19 vaccines could approximately double the lives saved before 2050. Investment is required to support vaccine development, manufacturing, prompt introduction and scale-up. Funding WHO (2020/985800-0)

Tuberculosis is one of the leading causes of infectious disease death worldwide, second only to COVID-19 in 2020. 1 The negative impact of COVID-19 on tuberculosis-related health services, such as delays in diagnosis, treatment, and neonatal vaccination has paused and reversed previous slowly declining trends in mortality. 1, 2 The WHO established the "End TB Strategy" in 2015, with the goal of reducing disease incidence, deaths, and costs worldwide from tuberculosis. 3 Targets for 2025 and 2035 include reductions in the absolute number of tuberculosis deaths by 75% and 95% and the tuberculosis incidence rate by 50% and 90%, respectively, compared to 2015 levels. 3 However, the majority of countries are not on track to achieve these targets. 1, 4 The 2035 End TB targets explicitly assumed the introduction of new tools, including a novel tuberculosis vaccine, in 2025. 3 The WHO has proposed Preferred Product Characteristics for New Tuberculosis Vaccines (WHO PPCs) developed through a highly consultative process, including regulators and policy makers from high burden countries. 5 While progress has been made, it is unlikely that the 2025 target for novel tuberculosis vaccine introduction will be achieved.

A phase 2b trial of the M72/AS01E candidate vaccine demonstrated an efficacy of 49.7% (95% confidence interval: 2.1-74.2) for preventing disease in adults positive by interferon-gamma release assay (IGRA+) from South Africa, Zambia, and Kenya after three years follow-up, 6 and a trial of BCG-revaccination appeared efficacious at preventing sustained infection in a cohort of IGRA negative (IGRA-) adolescents in South Africa with an efficacy of 45.4% (6.4-68.1). 7 Unfortunately, the phase 3 trial of M72/AS01E has not started, and therefore the realistic licensure date, should a positive result be found, may not be for a number of years. Policy changes on BCGrevaccination in adolescents could happen sooner in settings such as South Africa, as the trial is likely to be completed in 2024/2025, but BCG-revaccination has not been tested in infected individuals-a population shown previously to be epidemiologically important for rapid population-level vaccine impact. 8 This raises critical questions for global and country decision-makers, including: How many lives will be lost if we fail to roll out a novel tuberculosis vaccine by 2025? What is the potential impact if instead vaccines are introduced and rolled out following more traditional timelines, and how do we best prepare for that? And how would these impacts vary by WHO region, income level, and TB burden?

We estimated the potential impact of vaccines meeting the technical specifications of the WHO Preferred Product Characteristics for New Tuberculosis Vaccines in low-and middle-income countries (LMICs), in a range of introduction and scale-up scenarios.

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) 

To estimate the impact of novel tuberculosis vaccines, we developed a compartmental agestratified dynamic Mycobacterium tuberculosis (Mtb) transmission model (Figure 1 ), by adapting features of earlier models. 8, 9 In our model, tuberculosis natural history is represented using eight compartments, allowing for Mtb infection along a spectrum from uninfected to active clinical disease. 10, 11 A detailed description of the structure can be found in Supplementary Material section 1, with parameterisation in Supplementary Material section 2.

We incorporated an access-to-care structure to represent the systematic differences in tuberculosis burden and healthcare access by income. 12 The access-to-care structure contains high-access-tocare, representing the top three income quintiles (60% of the population) and low-access-to-care, representing the bottom two income quintiles (40% of the population). We assumed no transition between the high-and low-access-to-care classes, and random mixing between them.

To account for the influences of human immunodeficiency virus (HIV) and antiretroviral therapy (ART) on the risk of infection and progression to disease, 13,14 we included an HIV structure for countries if the proportion of tuberculosis cases among people living with HIV (PLHIV) was at least 15%, and the HIV prevalence was greater than 1% (countries listed in the Supplementary Material Table S5 .3). The HIV structure included HIV uninfected, HIV infected and not on ART, and HIV infected and on ART. The tuberculosis mortality rate and risk of progression is increased in both HIV compartments, with greater increases in those not on ART compared to on ART. We calibrated the model to epidemiologic data in each country separately using history matching with emulation through the hmer R package, 15 generating at least 1000 fitted parameter sets per country. We used the distribution of results produced by these parameter sets to quantify estimation uncertainty. 16 The model for each country was fit to nine calibration targets in 2019: the tuberculosis incidence rate (overall and by age), tuberculosis case notification rate (overall and by age), tuberculosis mortality rate (overall), the fraction of subclinical tuberculosis among active tuberculosis, and the risk ratio of active tuberculosis in the low-access-to-care group relative to . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 17, 2022. ; https://doi.org/10.1101/2022.04.16.22273762 doi: medRxiv preprint high-access-to-care. Models for countries with the HIV structure were fit to four additional all age HIV targets in 2019: HIV prevalence, ART coverage, tuberculosis incidence rate in PLHIV, and tuberculosis mortality rate in PLHIV.

For each country, separately, a primary baseline with no novel vaccine introduction was simulated, assuming that non-vaccine tuberculosis interventions continue at current levels ('Status Quo No-New-Vaccine' baseline). As reported country-level data includes the high coverage of neonatal BCG vaccination, 17 this was not explicitly modelled. We assumed that BCG vaccination would not be discontinued over the model time horizon.

Aligning with the product characteristics described in the WHO PPC, we evaluated a novel adolescent/adult and a novel infant vaccine. 5 Vaccines were assumed to prevent progression to disease and confer 10-years duration of protection on average, with exponential waning. We assumed the adolescent/adult vaccine would "take" in individuals in any infection state at the time of vaccination aside from active tuberculosis disease (i.e., a "pre-and post-infection" vaccine), with 50% vaccine efficacy. We assumed the infant vaccine would "take" in individuals who were uninfected at the time of vaccination (i.e., a "pre-infection" vaccine), with 80% efficacy. Further details are in Supplementary Material section 7.1.

The infant vaccine was implemented in two scenarios, and, separately, the adolescent/adult vaccine was implemented in three scenarios, with assumptions confirmed through consultation with a range of global tuberculosis vaccine stakeholders. The Basecase and Accelerated Scale-up scenarios included routine neonatal vaccination for the infant vaccine (85% coverage), and routine vaccination of 9-year-olds (80% coverage) with a one-time vaccination campaign for ages ten and older (70% coverage) for the adolescent/adult vaccine. The Routine Only scenario (adolescent/adult vaccine only) was introduced through routine vaccination of 9-year-olds (i.e., no campaign).

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 17, 2022. ; https://doi.org/10.1101/2022.04.16.22273762 doi: medRxiv preprint

We evaluated vaccine delivery scenarios by varying the introduction year and scale-up trends between scenarios and countries (Table 1 ). In the Basecase and Routine Only scenarios, based on data from historical vaccine introduction, vaccines were introduced in country-specific years and linearly scaled-up to coverage targets over five years once introduced. To estimate introduction years, countries were divided into 'procuring with Gavi support' and 'self-procuring'. Factors influencing the likely timing of vaccine introduction were identified through expert consultation, and included disease burden, prior early adopter status, timelines for Gavi processes, capacity for immunisation, country-specific registration timelines, and commercial prioritisation. A scoring system was applied to each factor, and countries were assigned an aggregate score ranking their 

We calculated tuberculosis incidence and mortality rate reductions in 2050 for each vaccine scenario compared to the Status Quo No-New-Vaccine baseline and calculated the cumulative numbers of tuberculosis treatments, deaths, and cases averted between vaccine introduction and 2050. Incidence rates in 2035 for each vaccine scenario were also estimated to investigate the feasibility of meeting the 2035 End TB target. Results are presented as the median and 95% uncertainty range for all countries modelled, WHO region, World Bank income group, 19 and WHO tuberculosis burden level. 22

We conducted scenario analyses to evaluate alternative assumptions regarding the Status Quo No-New-Vaccine baseline, vaccine characteristics, and delivery. We simulated vaccine scenarios with lifelong duration of protection for both vaccines, as well as scenarios with adolescent/adult vaccine efficacy increased to 75%. For each scenario, low-and high-coverage targets for five years postintroduction were compared to the medium-coverage targets used for the main analyses. We also explored an alternative baseline: the 2025 End TB No-New-Vaccine baseline, which assumed . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 17, 2022. ; https://doi.org/10.1101/2022.04.16.22273762 doi: medRxiv preprint strengthening of non-vaccine tuberculosis interventions to meet the 2025 End TB incidence target, 3 providing an alternative estimate of impact assuming existing measures would be deployed more effectively (Supplementary Material section 6.2).

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 17, 2022. ; https://doi.org/10.1101/2022.04.16.22273762 doi: medRxiv preprint

Epidemiologic and demographic data were available to model 115 of 135 LMICs. We successfully Table S8 .1), with the earliest Gavi-supported country introducing following a two-year delay compared to the earliest self-procuring. Figure 2 demonstrates the cumulative number of countries introducing per year between 2028 and 2047, with 50% of countries introducing the vaccine by 2034.

Our findings suggest that introducing a 50% efficacy adolescent/adult vaccine with 10-years protection in the Basecase scenario would reduce tuberculosis incidence and mortality rates in 2050 by 19.5% (95% uncertainty range=18.3-21.6%) and 20.6% (19.2-23.4%), respectively, compared to the Status Quo No-New-Vaccine baseline ( Table 2 ). The incidence reduction ranged from 11.2% in the WHO Region of the Americas (AMR) to 21.1% in the African (AFR) and

Eastern-Mediterranean (EMR) regions (Table 2, Figure 3 ). The tuberculosis mortality rate was reduced by 12.4% in AMR to 21.9% in EMR. By income group, relative impact was slightly higher in low-income countries compared to lower middle-income countries and upper middle-income countries (Table 2, Figure 3 ). Figure   3 ). High numbers of cases could be averted in both AFR and the WHO South-East Asian (SEAR) regions, which contribute the highest number to the global total. By 2050, 3.6 (3.3-3.9) million deaths could be averted for all countries, including 1.6 million in SEAR, 1.5 million in AFR, and 2.7 million in lower middle-income countries (Table 2, Figure 3 ). By 2050, 17.5 (15.6-19.2) . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 17, 2022. ; https://doi.org/10.1101/2022.04.16.22273762 doi: medRxiv preprint million treatments could be averted, with 8.2 (7.1-9.5) million averted treatments in SEAR alone.

In the 27 countries categorised by WHO as high-TB-burden of the 105 modelled, 28.6 (24.4-33.4) million cases, 15.9 (14.1-17.5) million treatments, and 3.3 (3.0-3.6) million deaths could be averted by 2050; around ten times higher than those averted in all other countries combined (Table 2, Figure 3 ). Table 2) .

With the Accelerated Scale-up scenario, a 50% efficacy adolescent/adult vaccine could prevent 7.9 (7.3-8.5) million deaths-4.3 million more deaths than the Basecase delivery-and avert 65.7 

Impact results from scenarios with lifelong duration of protection, 75% efficacy, and low-and high-coverage targets are provided in the Supplementary Material section 10. Assuming lower . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 17, 2022. ; https://doi.org/10.1101/2022.04.16.22273762 doi: medRxiv preprint coverage targets or the 2025 End TB No-New-Vaccine baseline led to reduced health impact, and higher coverage targets, 75% efficacy, or lifelong duration of protection vaccines led to an increased health impact compared to the Status Quo No-New-Vaccine baseline, a mediumcoverage target, 50% efficacy vaccine, or vaccine with only ten-years protection.

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 17, 2022. ; https://doi.org/10.1101/2022.04.16.22273762 doi: medRxiv preprint

Our results suggest that novel tuberculosis vaccines could substantially reduce the tuberculosis burden in the coming decades. The Basecase scenario, in which a 50% efficacy adolescent/adult vaccine was introduced over 20 years, could reduce tuberculosis incidence (19.5%) and mortality (20.6%) rates in 2050, and prevent 3.6 million deaths before 2050, including 1.6 million in the WHO South-East Asian region and 1.5 million in the African region. The more ambitious Accelerated Scale-up scenario could reduce tuberculosis incidence (25.2%) and mortality (26.7%) rates in 2050 and prevent over double (7.9 million) the number of deaths before 2050. The less ambitious Routine Only scenario could reduce tuberculosis incidence (9.9%) and mortality (9.9%) rates in 2050 and prevent around a third (1.1 million) of the deaths before 2050.

Impact estimates for vaccine introduction varied by region. While incidence and mortality rate reductions achievable by 2050 were similar between high-TB-burden countries and all other countries, the number of cases, treatments, and deaths averted were around ten times higher than those averted in all other countries, emphasising the need to focus on high-burden countries to maximise health impact. Particularly large numbers of averted cases, treatments, and deaths were predicted in Africa and South-East Asia, and in lower-middle-income countries, arguably populations in the greatest need.

Vaccination campaigns will be important to expedite health gains from vaccination. The Basecase and Routine Only scenarios offer a direct comparison of implementing with and without a campaign. The Basecase scenario averted around four times as many cases, deaths, and treatments as the Routine Only scenario, supporting the need to include a campaign in any future delivery strategy to maximise health impact.

A new vaccine will be an important tool to accelerate progress towards 2035 End TB targets.

Conservatively assuming non-vaccine intervention coverage and quality does not improve in the future (Status Quo No-New-Vaccine baseline) and roll out from 2028 in line with previous vaccines (before the Basecase scenario suggests we could reach around a third of the 2035 global target. More optimistic assumptions, that assume 2025 End TB targets are met before . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Two recent systematic reviews have highlighted potential health impacts of novel tuberculosis vaccines. 20, 21 Our study expands on the findings of these reviews and addresses some identified gaps. We showed that the adolescent/adult vaccine would have greater and more rapid health impacts than an infant vaccine before 2050. The largest burden of pulmonary tuberculosis disease is most often found in adults, 1 showing that faster vaccine introduction in LMICs is possible with high political will and financial resources. 22 This is more similar to our Accelerated Scale-up scenario, which averted up to 4.3 million more deaths than our Basecase scenario, and while the benefits of rolling out a vaccine from 2028 at pre-COVID-19 pace are predicted to be large, the increase in deaths demonstrate the health consequences from failing to rapidly introduce a vaccine. Unlike COVID-19, tuberculosis is a disease of the poor, which does not have the associated novelty, nor the same impact on highincome countries. Therefore, tuberculosis vaccines need concerted, sustained policy attention to overcome these barriers.

Our study has limitations. We successfully calibrated 105 of the 135 LMICs, representing 93% of global tuberculosis incidence, due in part to missing data from 20 countries that prevented . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 17, 2022. ; https://doi.org/10.1101/2022.04.16.22273762 doi: medRxiv preprint calibration from being attempted. Excluding 30 countries will slightly underestimate the number of cases, deaths, and treatments averted, and may bias the generalizability of the relative impact results.

Aligning with the WHO PPC, we assumed the adolescent/adult vaccine would be efficacious ("take") in both infected and uninfected individuals. However, the majority of previous and current trials have only enrolled either IGRA+ or IGRA-individuals. 6, 7 Ideally, novel vaccines will be safe and effective in both infected and uninfected individuals, as it would be costly and logistically difficult to include testing for tuberculosis infection before vaccine administration. If the realised vaccine will only be efficacious in either IGRA+ or IGRA-, our results will be overestimates of impact, as show in previous work. 8 We also assumed that vaccine efficacy was equivalent in PLHIV and HIV-naïve. However, vaccines are not always as efficacious in immunocompromised individuals, 23, 24 which would reduce the vaccine impact in countries incorporating the HIV structure.

We evaluated three vaccine scenarios for the adolescent/adult vaccine and two vaccine scenarios for the infant vaccine to provide estimates of vaccine impact under more and less ambitious introduction years and scale-up trends. Our more ambitious scenario, (Accelerated Scale-up), is less realistic, as it assumes a vaccine candidate would be ready for licensure, the supply exists, and that countries are positioned to make an introduction decision resulting in immediate uptake, all within the next 3 years. No specific risk groups were vaccinated in our model, however, initial delivery of novel tuberculosis vaccines within countries is likely to be through a targeted approach-a strategy shown to have a large population impact per vaccinated individual. [25] [26] [27] [28] [29] Countries may decide to initially target groups at the highest risk of developing tuberculosis disease or that contribute the most to transmission, while others may focus on vaccinating vulnerable age groups. Understanding how a new tuberculosis vaccine may be introduced in different settings is an important area for future research.

There are remaining gaps that modelling can help to address to provide evidence for investing in tuberculosis vaccine development and delivery to inform the Full Vaccine of Vaccine Assessment. 30 Estimates of the cost-effectiveness, budget impact, and wider benefits of specific . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 17, 2022. ; https://doi.org/10.1101/2022.04.16.22273762 doi: medRxiv preprint tuberculosis vaccine candidates would support research investment decision making. Future modelling research can help to better understand potential vaccine effectiveness considering a variety of factors, such as age, gender, duration of vaccine protection, and specific risk groups. We included an access-to-care structure in the model to account for differences in tuberculosis burden and healthcare access, which could be used to investigate vaccine targeting by income.

Additionally, to maximise the potential evidence available to individual countries, it would be beneficial to create more detailed individual country models to inform vaccine introduction decision making.

Novel tuberculosis vaccines could have a substantial impact, which will vary depending on vaccine and delivery characteristics. Vaccination campaigns will be crucial for rapid impact and . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 17, 2022. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 17, 2022. ; https://doi.org/10.1101/2022.04.16.22273762 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 17, 2022. ; https://doi.org/10.1101/2022.04.16.22273762 doi: medRxiv preprint 

World Health Organization. Global tuberculosis report 2021. Geneva: World Health Organization

The impact of COVID-19 on TB: a review of the data

World Health Organization. WHO | WHO End TB Strategy

New data shows COVID-19 combined with funding shortfalls are devastating efforts to end TB by 2030

World Health Organization. WHO Preferred Product Characteristics for New Tuberculosis Vaccines

Final Analysis of a Trial of M72/AS01E Vaccine to Prevent Tuberculosis

Prevention of M. tuberculosis Infection with H4:1C31 Vaccine or BCG Revaccination

Potential impact of tuberculosis vaccines in China, South Africa, and India

Impact and cost-effectiveness of new tuberculosis vaccines in low-and middle-income countries

The natural history of TB disease-a synthesis of data to quantify progression and regression across the spectrum

Subclinical TB disease: lessons and questions from field to lab

The effect of household poverty on tuberculosis

HIV and Tuberculosis: a Deadly Human Syndemic

Antiretroviral Therapy for Prevention of Tuberculosis in Adults with HIV: A Systematic Review and Meta-Analysis

hmer: History Matching and Emulation Package

Model parameter estimation and uncertainty analysis: a report of the ISPOR-SMDM Modeling Good Research Practices Task Force Working Group-6. Med Decis Mak Int

Vaccination and Immunization Statistics

World Bank Country and Lending Groups

Systematic review of mathematical models exploring the epidemiological impact of future TB vaccines

New tuberculosis vaccines: advances in clinical development and modelling

Data. Coronavirus (COVID-19) Vaccinations. Our World Data 2020

Insights on common vaccinations in HIV-infection: efficacy and safety

Long-term safety and immunogenicity of the M72/AS01E candidate tuberculosis vaccine in HIV-positive and -negative Indian adults

Impact of Targeted Tuberculosis Vaccination Among a Mining Population in South Africa: A Model-Based Study

Potential impact of spatially targeted adult tuberculosis vaccine in Gujarat, India

Epidemiological impact of targeted interventions for people with diabetes mellitus on tuberculosis transmission in India: Modelling based predictions

Age-targeted tuberculosis vaccination in China and implications for vaccine development: a modelling study

The epidemiologic impact and cost-effectiveness of new tuberculosis vaccines on multidrug-resistant tuberculosis in India and China

The Full Value of Vaccine Assessments (FVVA): A Framework to Assess and Communicate the Value of Vaccines for Investment and Introduction Decision Making

We are grateful for the support of the World Health Organization for funding this research (2020/985800-0). We thank all the attendees at the WHO meetings on the Full Value Assessment of TB Vaccines for insightful advice and direction. We thank Philippe Glaziou (WHO) for reviewing and providing helpful suggestions on the paper.

Accelerated