Malaria Vaccines

Published: October 2022

What is the problem?

Malaria is the fourth largest cause of death for children under five in sub-Saharan Africa, accounting for 13% of deaths.1 According to the World Health Organization (WHO), although malaria deaths in Africa declined by around 36% between 2000 and 2014, from 2015 to 2019 progress stalled with estimated annual deaths hovering between around 525,000 and 545,000.2 Malaria is transmitted from person to person by infected mosquitoes3 and involves flu-like symptoms, including fever.4 When a case of malaria is severe, it leads to life-threatening complications, particularly in children.5 It is also believed that malaria can cause permanent disability (hearing impairment, visual impairment, epilepsy, etc.).6

Malaria is preventable with insecticide-treated nets and drugs (among other interventions), and it is treatable with drugs.7

What is the program?

Malaria vaccines attempt to stimulate immune response against the parasites that cause malaria. Vaccines may vary depending on:

• the parasite they target (e.g., Plasmodium falciparum or Plasmodium vivax, to which 95 and 2 percent of malaria cases are attributed, respectively8 )
• the life stage of the parasite they target9
• the age group they target10

Though we briefly describe additional vaccines under development here, the bulk of this report focuses on RTS,S/AS01 (hereafter RTS,S), which targets P. falciparum11 and as of June 2022 is the only malaria vaccine recommended for use by WHO.12

WHO Recommended Vaccine: RTS,S/AS01

WHO recommends that RTS,S be administered in a four-dose schedule to children five months of age or older living in regions with moderate to high transmission13 of malaria as part of a comprehensive malaria control program.14 It is recommended that the first three doses be administered at one-month intervals with the fourth, “booster” dose given 18 months after the third dose.15 WHO also notes that countries may consider seasonal provision of the vaccine using a five-dose schedule in areas with seasonal transmission or peaks.16

Since 2019, RTS,S has been implemented through routine childhood immunization systems in selected areas of Ghana, Kenya, and Malawi through the Malaria Vaccine Implementation Programme (MVIP).17 Gavi, the Vaccine Alliance, began accepting applications in July 2022 from the three pilot countries for support for broader implementation of RTS,S and announced a second application window for additional moderate-to-high transmission countries for 2022/2023.18

Vaccines in the pipeline

As of June 2022, WHO reported that two vaccine candidates—R21/Matrix M and PfSPZ—were approaching late-stage clinical evaluation19 and others were in earlier stages of development.20

R21/Matrix M is similar in design to RTS,S, and phase-2b clinical trials administered it to young children in a four-dose schedule similar to that used for RTS,S.21 PfSPZ differs from RTS,S and R21 in using entire sporozoites as an ingredient and is expected to be tested in older children in a three-dose schedule delivered over four weeks in a phase-2 trial.22 We are also aware of planned phase-2 trials of both R21 and PfSPZ in adult populations.23

Does the program have strong evidence of effectiveness?

Evidence from a phase-3 clinical trial and a pilot evaluation indicates that RTS,S reduces the rate of clinical malaria and severe malaria in young children. We think that there is strong evidence that children who receive a four-dose schedule are less likely to experience clinical or severe malaria. We are less certain about malaria outcomes when a three-dose schedule is administered because children assigned to receive three doses were more likely to experience severe malaria than their unvaccinated peers in the second half of the clinical trial’s follow-up period. This could be indicative of “rebound” in severe malaria, but we are uncertain as to the magnitude of any effect because some of the excess severe malaria measured in the trial may have been due to chance. We think that RTS,S could plausibly avert deaths even given some rebound in risk of severe malaria because it provides higher protection to children who are younger, and could therefore be more vulnerable. We expect that the final results of the pilot evaluation and the results from a further trial phase will provide more information on whether a rebound effect is present.

Results from a clinical trial of R21/Matrix M have indicated a high vaccine efficacy rate. We expect to assess evidence for R21 and other vaccines under development more deeply when results from later-stage clinical trials become available.

Evidence of Effectiveness for RTS,S

Overall, we believe there is strong evidence that RTS,S reduces clinical malaria and severe clinical malaria in young children when administered in a four-dose schedule. The protective effect of vaccines appears to wane over time. We think it is likely that the reduction in clinical malaria averts deaths among the vaccinated population.

There is some evidence that children who receive only three doses (i.e., those who do not receive the “booster” dose) experience a rebound in severe malaria after immunity from vaccination has waned.24 However, we are uncertain about the magnitude of any rebound effect because some of the excess severe malaria measured in the trial may have been due to chance. Because malaria deaths tend to be concentrated in younger children, we believe it is plausible that protecting children against severe malaria when they are very young could avert deaths from malaria, even if those children go on to experience severe malaria at a higher rate than unvaccinated children at older ages. We are aware of ongoing data collection efforts from the MVIP that could address some of these uncertainties.

Phase-3 clinical trial results

Overall, we judge the evidence on the efficacy of RTS,S from phase-3 clinical trials to be of high quality, but an apparent pre-treatment difference in the risk of severe malaria between the trial arms complicates its interpretation.

Internal and external validity
The study was a pre-registered, individually randomized, double blind trial across 11 sites with varying degrees of malaria transmission in seven countries from 2009 through 2014.25 We focus on the modified intent-to-treat analysis of young children, which involved 8,922 participants (aged 5 to 17 months at recruitment) who were assigned to either receive four doses of RTS,S (“four-dose group”), three doses of RTS,S and one dose of an alternative vaccine (“three-dose group”), or four doses of an alternative vaccine and then were followed up for an average of 48 months.26

We have a moderate level of uncertainty about the extent to which findings from trials will generalize to future funding opportunities, though we expect that we may be able to reduce this uncertainty by comparing data on the following metrics from trial vs. implementation settings:

• Malaria transmission dynamics: Although not statistically significant, there is some evidence of higher vaccine efficacy in areas with lower malaria transmission, particularly in averting severe malaria.27 Thus places with higher transmission than the trial sites on average might benefit less (and vice versa).
• Access to vector control interventions: The trial facilitated access to insecticide-treated nets, and as a result around 70% to 80% of children across the treatment and control groups were covered by nets throughout the trial.28 Indoor residual spraying (IRS) was also used in a small number of sites, and those sites appear to have had lower net use,29 so the effective coverage with one or the other is likely higher than the nets figures imply. To the extent that nets or IRS act as substitutes to vaccines, it seems plausible that vaccine efficacy might be higher in populations with lower net or IRS use (and vice versa).
• Access to chemoprevention: A minority of children (less than 10%) in one site were treated with intermittent preventive treatment in infants for malaria, and the study does not report any sites using seasonal malaria chemoprevention.30 We are aware of one study comparing results from combined chemoprevention and RTS,S with results from each intervention alone, which may allow us to estimate effects of co-administration of RTS,S and chemoprevention drugs for particular funding opportunities.31
• Access to care: It seems likely that children in the study had better access to treatment for malaria and other illnesses compared with other children in their community. By design, parents were encouraged to seek care for illness, hospital transport fees were covered, and participants were visited monthly to track serious adverse events throughout the course of the study.32 Among study participants, 99% of those who presented to study clinics with confirmed malaria were treated with artemisinin combination therapy as specified by national guidelines.33 It is possible that vaccines would have larger effects on severe malaria in community settings where careseeking is delayed and/or the care provided might be of lower quality.34

While we believe the study to be high quality overall, subsequent analysis showed that the children assigned to three doses of RTS,S had a higher risk of severe malaria than those in the four-dose group even before the booster dose was administered.35 This difference in pre-booster risk, which WHO reports as due to chance rather than a randomization failure, makes us uncertain of the importance of the fourth dose and thus deaths averted in real-world contexts where coverage of the booster dose is likely to be lower than coverage of earlier doses.36

Average results
Overall, the study found that RTS,S protected against clinical and severe malaria in young children, but that the four-dose group experienced much more substantial protection, particularly against severe malaria.

Among young children who were assigned to four doses, the study reported vaccine efficacy of 36.3% (95% CI: 31.8, 40.5) against episodes of clinical malaria throughout the trial.37 It reported a slightly lower vaccine efficacy of 32.2% (95% CI: 13.7, 46.9) against having experienced at least one episode of severe malaria for the same group throughout the trial.38

Among young children who were assigned to three doses, the study reported vaccine efficacy of 28.3% (95% CI: 23.2, 32.9) against episodes of clinical malaria throughout the trial.39 This group experienced a much lower vaccine efficacy of 1.1% (95% CI: -23.0, 20.5) against having experienced at least one episode of severe malaria throughout the trial.40

The phase-3 trial results showed waning protection over time, particularly for the three-dose group after month 20 of the study (when they were assigned to receive an alternative vaccine rather than a booster dose of RTS,S).

Vaccine efficacy against episodes of clinical malaria declined substantially, from 45.1% protection across both treatment arms in the first 20 months following vaccination to about 3% for the three-dose group and 12% for the four-dose group from month 33 to study end, about 48 months for the median participant (see the table below).41

	Months 0-20 of study	Months 21-32 of study	Months 33-Study end​​
3-dose group	45.1% (41.4, 48.7) Reported as combined group	16.1% (8.5 to 23.0)	2.9% (-6.4 to 11.4)
4-dose group	45.1% (41.4, 48.7) Reported as combined group	37.4% (31.4 to 42.8)	12.3% (3.6 to 20.1)

Vaccine efficacy also declined substantially with regard to the risk of experiencing at least one case of severe malaria, from 33.9% (15.3 to 48.3) protection across both treatment arms in the first 20 months following vaccination to about -44.4% (-119.0 to 4.1) for months 21 to 32 of the study and -57.9% from month 33 to study end for the three-dose group.42 In addition, the estimate for vaccine efficacy from month 21 through the end of the study for the three-dose group suggests a large negative effect, -41.0% (-98.5, -0.8), which is statistically significant at conventional levels (p=0.038).43 It appears to be concentrated in high-transmission sites, where unvaccinated children might rapidly develop natural immunity.44 These results are suggestive of substantial risk of rebound among the three-dose group.

Protection also decreased over time for the four-dose group, though these drops are less substantial than for the three-dose group, and there is no statistically significant evidence of rebound, with vaccine efficacy of -3.2% (-61.8 to 34.1) for months 21 to 32 and -18.8% (-128.0 to 37.6) for months 33 to the study end.45 In addition, the proportion of children affected by severe malaria dropped over time across all three groups.46

Phase-3 trial evidence on mortality
The study found no significant effect on all cause mortality, and point estimates actually suggested vaccine receipt for both the three-dose and four-dose groups was associated with higher mortality risk.47 The study authors speculate that the failure to find a protective effect of vaccination could be due to the higher standard of care given in the trial context.48 As we discussed above, it seems plausible that trial participants had better access to and higher-quality care than others in their communities, and in fact, mortality rates documented in the study appear to have been lower than those in the rest of the community (note, though, that this comparison relies on national averages and imprecise age categorization).49

Additional evidence related to rebound

Additional analysis of phase-3 data
Evidence of pre-booster-dose differences in severe risk between the three-dose and four-dose group and the widely varying vaccine effectiveness estimates for three versus four doses on the incidence of clinical malaria compared with the risk of experiencing severe malaria make it seem plausible that the phase-3 results overestimate the risk of rebound in severe malaria for children who receive three doses. However, we have high uncertainty about the magnitude of this overestimate.

Additional analysis of the phase-3 clinical trial data, referenced in a 2021 WHO report on RTS,S, indicates that the three-dose group had an elevated incidence of severe malaria before the booster or alternative vaccines were administered at 20 months. In particular, while the incidence per person-year of severe malaria in the three-dose group was approximately 0.32 in the eight months preceding the booster dose (months 12 to 20 of the study), it was approximately 0.29 for the control group, and 0.20 for the four-dose group.50 WHO reports that this difference is likely due to chance given that “further analysis by GSK at the request of WHO indicated no problem with randomization” and “the risk of clinical malaria was similar in the 2 arms.”51 This suggests that some of the increased risk observed in the second half of the phase-3 trial among children assigned to three doses may be attributable to pre-existing differences in risk rather than rebound. However, we have not seen or reviewed the data informing WHO’s conclusion.

In addition, the phase-3 data indicate a much larger marginal benefit of four doses versus three in reducing the risk of experiencing severe malaria than in averting clinical malaria. Whereas the vaccine efficacy for clinical malaria for the four-dose group is 36.3%,52 it is about 78% as large (28.3%)53 for the three-dose group. In contrast, the estimated vaccine efficacy for severe malaria is only about 3% as large for the three-dose group as the four-dose group (1.1% vs. 32.2%).54 To the extent that we expect severe malaria to result from the same underlying risks as clinical malaria, the wide difference in relative benefits lends plausibility to the idea that the three-dose group experienced excess severe malaria by chance rather than because of systematically elevated malaria risk once vaccine immunity wore off, or because the four-dose group got unusually lucky in its lack of severe malaria.

An additional study followed a subset of about 19% of phase-3 trial participants (1,739 children) for three additional years and found no evidence of rebound in severe malaria among RTS,S recipients.55 However we do not consider this evidence informative because the extension study involved only three of the original 11 sites, and these three sites appear to have had better than average vaccine efficacy against severe malaria during the initial study period. In particular, the point estimate for efficacy against severe malaria from month 21 to the end of the initial study for the three-dose group within the three extension sites was positive whereas the same estimate across all sites was -41.0%.56

Distribution of malaria deaths by age
Based on existing data on the distribution of malaria deaths by age, we think that even given some risk of rebound for children receiving three doses, RTS,S could avert deaths overall. This is because it offers protection during a higher-risk period of life (approximately six months to two years of age). However, our understanding of the age distribution of malaria mortality is based on older data from settings without vaccines and lower use of other malaria interventions. We are uncertain whether this data generalizes to future implementation settings.

Carneiro et al. 2010 is a meta-analysis that estimates the probability distribution of malaria deaths for children aged one month to ten years old across multiple countries in sub-Saharan Africa.57 It finds that malaria mortality is highly concentrated in children under two years of age, and this skew toward deaths at younger ages is more pronounced for higher malaria transmission contexts than moderate transmission ones.58 Carneiro et al. 2010 implies that in high-transmission contexts there are almost four times the malaria deaths between ages six months and two years (i.e., within the 20-month period when RTS,S was shown to have higher efficacy) as in the following 28 months (i.e., within the period where RTS,S may cause severe malaria rebound).59 In moderate transmission contexts, there are about twice as many deaths in the younger age groups.60 The malaria case rate was also concentrated in younger children, but the median age of both hospital admissions and clinical malaria was higher than the median age of malaria mortality.61 This implies that the malaria case fatality rate is higher in younger children. Taking these ratios at face value implies that RTS,S could effectively avert deaths in moderate and high transmission contexts by averting severe malaria cases for those aged around two and under, even if it did not reduce severe malaria cases experienced in early childhood overall.

However, we have significant uncertainty about this assumption because Carneiro et al. 2010 was published more than a decade ago and relies on even older data.62 We are unsure how much the wider use of malaria interventions in subsequent periods has affected the mortality burden by age.63 In addition, the higher risk of younger children is plausibly driven by two factors: (a) absence of immunity to malaria and (b) increased vulnerability as a result of being very young. To the extent that absence of immunity is the most important factor driving this distribution, we might expect that rebound after age two would be associated with deaths. On the other hand, if getting older is protective in and of itself, then it would be beneficial to delay a child’s experience of severe malaria.

Malaria Vaccine Pilot Evaluation

The Malaria Vaccine Implementation Programme (MVIP) piloted the introduction of RTS,S through the Enhanced Programme on Immunization (EPI) systems of Malawi, Ghana and Kenya beginning in 2019.64 The Malaria Vaccine Pilot Evaluation (MVPE) is designed to leverage random assignment of subnational areas to receive the vaccine first to evaluate the feasibility of delivery of a four-dose schedule, the safety of the vaccine in routine use, and the impact of the malaria vaccine on severe malaria and all-cause mortality at a population level.65 In addition, a planned case control study embedded within the MVPE seeks to estimate vaccine risks and benefits at an individual level, including measuring the effectiveness of three or four doses of RTS,S at preventing severe malaria and the magnitude of a potential rebound effect.66

MVPE data collection is ongoing, but interim results were published in October 2021. These results reported:

• A 30% reduction in hospital admissions with severe malaria among vaccine age-eligible children in implementation versus comparison areas.67
• A 7% reduction in the ratio of deaths among children who were age-eligible for vaccination to those who were not age-eligible for vaccination in implementation versus comparison areas.68 This difference did not reach conventional levels of statistical significance, but the study was not yet powered to detect a mortality effect at this time.69

This evidence bolsters our assumptions that we should expect RTS,S to reduce severe malaria and death in routine use. However, it is challenging to directly compare this data with that from the phase-3 trials because the MVPE is at a population level. We note the following challenges for comparison with phase-3 trials in particular:

• Coverage rates for the first three doses hovered between 60% and 80% in the MVIP during the period from April 2019 to June 2021. The modified ITT results from the phase-3 trial conditioned on receiving at least one dose of vaccine.70
• The average follow-up period represented in the MVPE depends on how numbers of vaccinated children accrued over time, whereas in the phase-3 data follow-up periods are more standardized.
• Because vaccinations began in the second half of 2019, it seems likely that protection from vaccination would not have significantly waned for most of those vaccinated by the time of the 2021 report.

We expect that final MVPE results (including the case control study) will provide additional data that could address these and other uncertainties, and we expect to update this report after reviewing those results.

Translating evidence on RTS,S into inputs for our cost-effectiveness analysis

There is strong evidence that a four-dose schedule of RTS,S reduces clinical and severe malaria. While the degree of protection degrades over time, there is no statistically significant evidence that children assigned to receive four doses go on to experience higher levels of severe malaria than unvaccinated children.

We have higher uncertainty about the effects of three doses on mortality. On the one hand, phase-3 trial results are consistent with substantial risk of rebound for children who do not receive booster doses. On the other hand, there is suggestive evidence that some of the severe malaria in the three-dose group was due to chance.

In addition, we think it plausible that because it provides the most protection under age two, RTS,S averts more deaths than the average protection over four years after vaccination would suggest. This could be the case if the vaccine delays severe malaria onset until children are older and potentially better able to withstand illness.

We discuss how we translate these considerations into quantitative estimates for our cost-effectiveness estimates here. These assumptions are highly uncertain and we expect to update them (perhaps substantially) after reviewing additional results from the MVPE.

Additional benefits

Vaccinating children against malaria may disrupt malaria transmission, reducing incidence of malaria in the wider (untreated) population. One RCT of a different malaria intervention (Cisse et al. 2016) estimates protection to the untreated population. It found that malaria incidence among the untreated population (over the age of ten) was 26% lower (95% CI 18%-33%) in areas where seasonal malaria chemoprevention (SMC) was delivered to children compared to control areas.71 Overall, we take Cisse et al. 2016 as providing some evidence that reducing malaria rates in children reduces incidence of malaria in the untreated population. We have discussed the results from Cisse et al. 2016 in more depth here. However, we are unsure whether transmission dynamics operate similarly with drug-based versus vaccine-based malaria prevention and whether the concentration of protection provided by RTS,S in younger children reduces impact on transmission overall.

Reducing exposure to malaria during childhood may also have an effect on long-term productivity and earnings, although we are uncertain about the magnitude of this effect. We have reviewed the evidence for the developmental effects of malaria prevention here.

Potential offsetting/negative effects

• Possible rebound effects. As discussed above, there is some evidence that children who are vaccinated against malaria experience a rebound in severe malaria after immunity from vaccination has waned. However, we are uncertain about the magnitude of any rebound effect. We are also unsure if children who receive four doses might experience rebound in risk of severe malaria after immunity from the fourth dose wanes.72
• Possible adverse effects. The phase-3 trials results raised three potential safety signals: higher numbers of meningitis cases in some groups (not temporally linked to the timing of vaccination), an increase in the number of cerebral malaria cases in some groups, and an imbalance in the mortality results for girls.73 The interim MVIP results found no strong evidence of an association between malaria vaccination and meningitis,74 cerebral malaria,75 or gender-specific mortality.76 Note, however, that the gender-specific mortality ratios of 0.98 in girls and 0.90 in boys appear consistent with larger protective effects among boys.77 We plan to revisit the safety results when the phase-4 trial results78 and the final MVPE results79 become available.

Evidence on additional vaccines

Results from a phase-2b clinical trial of R21/Matrix M showed six-month vaccine efficacy of 74% (95% CI 63–82) among children receiving R21 plus 25 μg of matrix M and 77% (95% CI 63–82) among children receiving R21 plus 50 μg MM.80 However, direct comparison between RTS,S/AS01 and these results is not possible81 because the latter results come from a setting with highly seasonal malaria transmission82 and vaccine administration was timed to precede peak transmission.83 Six-month vaccine efficacy for RTS,S/AS01 in the same site as the R21 study found similar vaccine efficacy as R21 (VE: 71.9, C: 59.7, 80.4).84 Ongoing phase-3 clinical trials of R21 are expected to be completed in December 2023.85 We expect to assess evidence for R21 and other vaccines under development more deeply when results from phase-3 clinical trials become available.

How cost-effective is the program?

We conducted a preliminary cost-effectiveness analysis. We expect that future funding opportunities related to malaria vaccines could support technical assistance or research opportunities (see more here). The costs of such activities and impact in terms of additional children vaccinated are likely highly variable and opportunity-specific. Therefore, we view this model as illustrative of the types of opportunities that could be within the range of cost-effectiveness of programs we expect to direct funding to.

Note that our cost-effectiveness analyses are simplified models that do not take into account a number of factors. There are limitations to this kind of cost-effectiveness analysis, and we believe that cost-effectiveness estimates such as these should not be taken literally due to the significant uncertainty around them. We provide these estimates (a) for comparative purposes and (b) because working on them helps us ensure that we are thinking through as many of the relevant issues as possible.

This cost-effectiveness analysis is illustrative, and we think it’s likely that our bottom line cost-effectiveness estimate will change substantially with further review and when considering particular funding opportunities.

We expect vaccines to be relatively expensive to procure, and their average protection over four years is only moderate. However, funding opportunities would potentially leverage Gavi funds for vaccine procurement and nationally run enhanced platforms for immunization, which could increase cost-effectiveness.

A sketch of the cost-effectiveness model is below:

• Duration of benefits: We assume vaccination begins at six months of age and protective effects last for four years.
• Deaths due to malaria: We use data from Democratic Republic of the Congo (DRC) for illustration. DRC is a high-burden country with a large number of births in areas categorized as “Category 1 (Greatest need, highest priority)” in the WHO’s framework for allocating malaria vaccines.86 Using data from the Institute for Health Metrics and Evaluation, GBD Results Tool, and adjusting for the targeted age range, we estimate a baseline malaria mortality rate of 2.6 deaths per 1,000 child years.87
• Effect of RTS,S on severe malaria: We use estimates from phase-3 clinical trials of the effect of three-dose and four-dose vaccination schedules on the risk of experiencing severe malaria with subjective adjustments88 to account for higher risk of severe malaria due to chance in the three-dose group.89 This results in estimates that three doses results in 17% fewer children experiencing severe malaria and four doses results in 25% fewer children experiencing severe malaria over the 48-month period after vaccination.90
• Relationship between severe malaria averted and malaria deaths averted: It seems plausible that because RTS,S has larger protective effects at younger ages and malaria deaths tend to be concentrated among the young, it averts more mortality than its effect on experience of severe malaria implies. We include a 110% adjustment to account for this possibility.91
• Cost of the program: We model contributions of three funders: philanthropic funders such as GiveWell, Gavi, and domestic governments.
  • We base the philanthropic contribution off a past GiveWell grant recommendation supporting RTS,S implementation in MVIP countries, yielding an illustrative cost of $11 philanthropic dollars per child vaccinated.
  • We assume that Gavi covers costs of vaccine procurement at $9.44 per dose or about $34 per child vaccinated.
  • We assume that the domestic government contributes $10 per child vaccinated, which includes costs of delivery through routine immunization systems.
  • Given an estimated four years of protection, this yields a cost of $13.63 per child year of protection across all contributors.
• Cost-effectiveness: Using an estimate of $13.63 from all contributors for a child-year of protection, we estimate that the RTS,S vaccine averts a child’s death in DRC for $15,867.

However, we have high uncertainty about the philanthropic cost per additional child vaccinated, the impact of vaccines on malaria mortality, and the counterfactual value of Gavi funds.

• Philanthropic cost per additional child vaccinated: We have high uncertainty about the philanthropic costs that would accompany a particular investment opportunity. Technical assistance and research programs can vary widely in their costs, program activities, and impact on the number of additional children vaccinated. Therefore, we expect that specific funding opportunities could diverge widely from the example modeled here.
• Impact of vaccines on mortality: We have high uncertainty about the level of coverage that would be achieved for fourth doses in implementation settings, the magnitude of “rebound” in severe malaria for those who do not receive a booster dose, and the degree to which malaria mortality would be concentrated in children under two for the vaccinated population. These uncertainties all influence bottom-line mortality averted by malaria vaccination. We hope to address some of these uncertainties by reviewing data from the MVPE when it becomes available.
• Counterfactual value of Gavi funds: Because we expect that vaccine procurement costs would be covered by leveraging Gavi financing and we expect vaccines to be relatively costly, the bottom-line cost-effectiveness estimate depends heavily on the counterfactual value of Gavi funds. This estimate is particularly subjective and uncertain.

Is there room for more funding?

In 2021 the board of Gavi, the Vaccine Alliance, approved US$155.7 million for 2022-2025 to support malaria vaccine introduction, procurement, and delivery in Gavi-eligible countries.92 Given this investment and the potential for increased support from Gavi in subsequent years, we think it unlikely that funding opportunities we would consider would support routine use. Instead, we expect there may be funding opportunities in technical assistance to speed up country implementation and research opportunities. We are uncertain of the likely scale of such opportunities.

Key questions for further investigation

• Where are there gaps in funding for technical assistance or research? What are their costs and expected impact on the number of children vaccinated? Are there research opportunities to improve the use or efficiency of RTS,S?
• What is the impact of malaria vaccines on all-cause mortality impact? We expect to triangulate estimates from the MVPE by also considering the following questions:
  • What is the impact of receiving three or four doses of RTS,S on severe malaria?
  • Do those who receive three doses experience a rebound in severe malaria?
  • What is the impact of waning vaccine immunity on the age distribution of malaria mortality among vaccinated populations?
• What coverage levels are achieved in operational settings?
• Does additional safety monitoring data surface any concerns about adverse events?
• How do other vaccines compare to RTS,S in efficacy, cost, or feasibility of production and implementation?

Our process

• We conducted a light literature review for evidence related to RTS,S, which led us to focus on an in-depth review of the data from phase-3 trials and its extension and on the Full Evidence Report on the RTS,S/AS01 Malaria Vaccine.93 We did not review literature related to the basic science of the vaccines or the biological plausibility of vaccine protection or adverse events. We expect to update our understanding of these areas when data from the MVPE and its associated case control study becomes available.
• We conducted a very light literature review to understand other potential malaria vaccines and read the abstracts of associated papers.
• To construct the CEA, we updated our model of the cost-effectiveness of seasonal malaria chemoprevention. We may update this approach in the future if data becomes available that allows us to estimate the burden of malaria in the age range for which malaria vaccines provide the most protection.
• We had conversations with representatives from the Bill and Melinda Gates Foundation, WHO, PATH,94 and Malaria Consortium about their understanding of the evidence and vaccine landscape, as well as potential funding opportunities.

Sources

Document	Source
Butler 2019	Source
Carneiro et al. 2010	Source (archive)
Chandramohan et al. 2021	Source
Cisse et al. 2016	Source (archive)
Clinical Trials Study Record, "Efficacy of GSK Biologicals' Candidate Malaria Vaccine 257049 Against Malaria Disease in Infants and Children in Africa," 2019	Source (archive)
Clinical Trials Study Record, "Malaria Vaccine Pilot Evaluation (MVPE)," 2021	Source (archive)
Clinical Trials Study Record, "PfSPZ Vaccine Trial in Malian Children," 2022	Source (archive)
Clinical Trials Study Record, "R21/Matrix-M in African Children Against Clinical Malaria," 2022	Source (archive)
Clinical Trials Study Record, "Safety and Immunogenicity of the Malaria Vaccine, R21/MatrixM, in Healthy Thai Adults (R21/Matrix-M)," 2022	Source (archive)
Clinical Trials Study Record, "Safety, Tolerability, Immunogenicity and Protective Efficacy of PfSPZ Vaccine and PfSPZ-CVac in Indonesian Adults Against Naturally-Transmitted Malaria," 2022	Source (archive)
Clinical Trials Study Record, "Strengthening the Evidence for Policy on the RTS,S/AS01 Malaria Vaccine (MVPE-CC)," 2021	Source (archive)
Datoo et al. 2021	Source (archive)
Gavi, "Gavi Board approves funding to support malaria vaccine roll-out in sub-Saharan Africa," 2021	Source (archive)
Gavi, "Gavi opens applications for malaria vaccine rollout support", 2022	Source
GiveWell, Carneiro estimates of child malaria mortality age-distribution for RTS,S BOTEC, 2022	Source
GiveWell, "Mass Distribution of Long-Lasting Insecticide-Treated Nets (LLINs)," 2021	Source
GiveWell, "PATH — RTS,S Malaria Vaccines in Pilot Comparison Areas (January 2022)"	Source
GiveWell, "Seasonal Malaria Chemoprevention," 2018	Source
GiveWell, "Why we can’t take expected value estimates literally (even when they’re unbiased)," 2016	Source
GiveWell's non-verbatim summary of a conversation with PATH and the World Health Organization, January 5, 2022	Source
GiveWell's non-verbatim summary of a conversation with PATH, July 30, 2021	Source
Griffin, Ferguson and Ghani, 2014	Source (archive)
Institute for Health Metrics and Evaluation, GBD Compare Data Visualization, under-5 deaths in sub-Saharan Africa, 2019	Source
Institute for Health Metrics and Evaluation, GBD Results Tool, under-5 deaths in Democratic Republic of the Congo, 2019	Source
Jamison et al., 2006	Source
Laurens 2020	Source
Ledford 2021	Source (archive)
Maitland 2016	Source
Malaria Consortium, SMC at scale - saving lives	Source (archive)
Moorthy and Binka 2021	Source
Mousa et al. 2020	Source (archive)
RTS,S Clinical Trials Partnership 2014, Supplemental materials, Figure S6	Source (archive)
RTS,S Clinical Trials Partnership 2015	Source
Stanford Health Care, "Symptoms of Malaria"	Source (archive)
Tinto et al. 2019	Source
Tinto et al. 2019, supplementary webappendix	Source
World Health Organization, "WHO Guidelines for malaria - 3 June 2022", v4.0	Source (archive)
World Health Organization, "WHO Preferred Product Characteristics (PPC) for Malaria Vaccines", 2014	Source (archive)
World Health Organization, “Global Malaria Programme: Preventive Chemotherapies”	Source (archive)
World Health Organization, “Global Malaria Programme: Treatment”	Source (archive)
World Health Organization, “Global Malaria Programme: Vector Control”	Source (archive)
World Health Organization, Framework for the allocation of limited malaria vaccine supply, 2022	Source (archive)
World Health Organization, Full Evidence Report on the RTS,S/AS01 Malaria Vaccine, 2021	Source (archive)
World Health Organization, Malaria vaccine: WHO position paper, March 2022	Source (archive)
World Health Organization, World Malaria Report 2021	Source

• 1Using search parameters for under-5 deaths in sub-Saharan Africa, 12.88% of total deaths in 2019. Institute for Health Metrics and Evaluation, GBD Compare Data Visualization, under-5 deaths in sub-Saharan Africa, 2019
  
• 2See World Health Organization, World Malaria Report 2021, p. 26, Table 3.2, for estimates of the number of deaths caused by malaria in the WHO African Region from 2000 to 2020. The growth rate between 2000 and 2014 is calculated as the percentage change from an estimated 840,000 deaths in 2000 to an estimated 534,000 deaths in 2014. “Since 2015, the rate of progress in both cases and deaths has stalled in several countries with moderate or high transmission.,” World Health Organization, World Malaria Report 2021, p. 27.
  
• 3"Human infection begins when the malaria vector, a female anopheline mosquito, inoculates plasmodial sporozoites from its salivary gland into humans during a blood meal. The sporozoites mature in the liver and are released into the bloodstream as merozoites. These invade red blood cells, causing malaria fevers. Some forms of the parasites (gametocytes) are ingested by anopheline mosquitoes during feeding and develop into sporozoites, restarting the cycle." Jamison et al., 2006, p. 413.
  
• 4"In the early stages, malaria symptoms are sometimes similar to those of many other infections caused by bacteria, viruses, or parasites. It can start with flu-like symptoms. Symptoms may include:
  • Fever. This is the most common symptom.
  • Chills.
  • Headache.
  • Sweats.
  • Fatigue.
  • Nausea and vomiting.
  • Body aches.
  • Generally feeling sick.”

  Stanford Health Care, "Symptoms of Malaria"
  

• 5“Many children who are admitted will be suffering from life-threatening complications of Plasmodium falciparum malaria, such as coma and convulsions (cerebral malaria), severe anemia (requiring urgent lifesaving transfusion), and rapid breathing (due to severe metabolic acidosis). Approximately 90% of the world’s falciparum infections and deaths occur in sub-Saharan Africa, the latter almost entirely in children younger than 5 years of age.” Maitland 2016.
  
• 6
  • “CM [cerebral malaria] is associated with hemiparesis, quadriparesis, hearing and visual impairments, speech and language difficulties, behavioral problems, epilepsy, and other problems (table 21.3).” Jamison et al., 2006, p. 417.
  • See Jamison et al., 2006, p. 416, Table 21.3, for estimates of cases of hearing impairment, visual impairment, epilepsy, etc. caused by malaria.

• 7
  • “Preventive chemotherapy is the use of medicines, either alone or in combination, to prevent malaria infection and its consequences. It requires giving a full treatment course of an antimalarial medicine to vulnerable populations at designated time points during the period of greatest malarial risk, regardless of whether the recipient is infected with malaria.” World Health Organization, “Global Malaria Programme: Preventive Chemotherapies”.
  • “Vector control is a highly effective way to reduce malaria transmission and is a vital component of malaria control and elimination strategies. WHO currently recommends deployment of either insecticide-treated nets (ITNs) or indoor residual spraying (IRS) for malaria vector control in most areas at risk of malaria.” World Health Organization, “Global Malaria Programme: Vector Control”.
  • “For effective case management of the disease, both early diagnosis and prompt treatment of malaria are essential. The best available treatment, particularly for P. falciparum malaria, are the artemisinin-based combination therapies (ACTs).” World Health Organization, “Global Malaria Programme: Treatment”.

• 8
  • See World Health Organization, World Malaria Report 2021, pp. 248-58, Annex 5 - I. Reported Malaria Cases by Species, 2000–2020. Total global reported cases for 2020 were calculated by summing rows for P. falciparum (89,146,343; 95%), P. vivax (1,668,965; 2%), mixed cases (2,649,542; 3%), and other species (153,290; 0%)
  • For example, the RTS,S/AS01 vaccine is specifically aimed at P. falciparum: "The RTS,S/AS01 malaria vaccine should be used for the prevention of P. falciparum malaria in children living in regions with moderate to high transmission as defined by WHO." World Health Organization, "WHO Guidelines for malaria - 3 June 2022", v4.0, pp. 131.

• 9“Two vaccine candidates are approaching late-stage clinical evaluation: the R21/MatrixM vaccine candidate targeting PfCSP protein [142] and the attenuated whole sporozoite vaccine PfSPZ [143]. Additional candidates targeting other malaria lifecycle stages include the Rh5 blood-stage vaccine candidate [144] and Pfs25 and Pfs230 vaccine candidates targeting sexual-stage antigens to prevent human-to-mosquito transmission (NCT02942277). New technologies, such as DNAand mRNA-based vaccines [145], the ongoing development of adjuvants [146], and delivery platforms such as virus-like particles (VLPs; the delivery platform used for RTS,S/AS01) and vesicle-based technologies are being explored for use in malaria vaccines. WHO has developed guidelines on the quality, safety, and efficacy of the recombinant malaria vaccines targeting preerythrocytic and blood stages of P. falciparum [147] and a set of preferred product characteristics (PPCs).” World Health Organization, "WHO Guidelines for malaria - 3 June 2022", v4.0, pp. 130-31.
  
• 10See the "target population" cells, World Health Organization, "WHO Preferred Product Characteristics (PPC) for Malaria Vaccines", 2014, pp. 17-18, Table 3a: Preferred Product Characteristics: Disease-Reducing Malaria Vaccines and Table 3b: Preferred Product Characteristics: Transmission Reducing Malaria Vaccines, development timeline: first product targeted between 2025 and 2035.

  Note: As of June 2022, WHO described these guidelines as “currently being updated”: “The PPCs include attributes ranging from safety and efficacy to route of administration, product stability and storage, in order to help support the ongoing development of new malaria vaccines. These PPCs are currently being updated to reflect recent advances in malaria vaccine research and development.” World Health Organization, "WHO Guidelines for malaria - 3 June 2022", v4.0, p. 131.
  

• 11

  “RTS,S/AS01 is a pre-erythrocytic recombinant protein vaccine, based on the RTS,S recombinant antigen. It comprises the hybrid polypeptide RTS, in which regions of the P. falciparum circumsporozoite protein known to induce humoral (R region) and cellular (T region) immune responses are covalently bound to the hepatitis B virus surface antigen (S).” World Health Organization, Malaria vaccine: WHO position paper, March 2022
  

• 12

  “The RTS,S/AS01 vaccine is the first and currently the only malaria vaccine to be recommended for use by WHO.” World Health Organization, "WHO Guidelines for malaria - 3 June 2022", v4.0, p. 130.
  

• 13“33 moderate and high transmission countries: Angola, Benin, Burkina Faso, Burundi, Cameroon, the Central African Republic, Chad, the Congo, Côte d'Ivoire, the Democratic Republic of the Congo, Gabon, the Gambia, Ghana, Guinea, Guinea-Bissau, Equatorial Guinea, Kenya, Liberia, Madagascar, Malawi, Mali, Mauritania, Mozambique, the Niger, Nigeria, Senegal, Sierra Leone, South Sudan… Togo, Uganda, the United Republic of Tanzania, Zambia and Zimbabwe.” World Health Organization, World Malaria Report 2021, p. 43.

• 14

  “Strong recommendation for, High certainty evidence
  Malaria vaccine (2021)
  The RTS,S/AS01 malaria vaccine should be used for the prevention of P. falciparum malaria in children living in regions with moderate to high transmission as defined by WHO.

  • The RTS,S/AS01 malaria vaccine should be provided in a four-dose schedule in children from 5 months of age.[...]
  • RTS,S/AS01 malaria vaccine should be provided as part of a comprehensive malaria control strategy.”

  World Health Organization, "WHO Guidelines for malaria - 3 June 2022", v4.0, p. 16.

• 15

  “The first 3 doses should be administered at one-month intervals, with a fourth dose given 18 months after the third dose.” World Health Organization, Malaria vaccine: WHO position paper, March 2022

• 16
  • “Countries may consider providing the RTS,S/AS01 vaccine seasonally, with a five-dose strategy, in areas with highly seasonal malaria or with perennial malaria transmission with seasonal peaks.
  • Countries that choose to introduce the vaccine in a five-dose seasonal strategy are encouraged to document their experiences, including adverse events following immunization.”

  World Health Organization, "WHO Guidelines for malaria - 3 June 2022", v4.0, p. 16.

• 17
  • “The MVIP is coordinated by WHO in close collaboration with ministries of health (MoH) in the three participating countries – Ghana, Kenya, Malawi – and a range of in-country and international partners. The MoH of the pilot countries have introduced the RTS,S/AS01 vaccine through their childhood immunization services using routine vaccine introduction strategies and methods.” World Health Organization, Full Evidence Report on the RTS,S/AS01 Malaria Vaccine, 2021, p. 7.
  • See vaccine coverage estimates for different time periods in World Health Organization, Full Evidence Report on the RTS,S/AS01 Malaria Vaccine, 2021, p. 47, Table 1: Vaccine coverage estimates for different time periods according to routine administrative data.

• 18“The world’s first-ever mass vaccination against malaria was brought a step closer today as Gavi, the Vaccine Alliance opened a process for countries to apply for funding and support to roll out the new vaccine.
  The opening of the application window follows the WHO’s recommendation for wider routine use of the RTS,S/AS01 malaria vaccine in October 2021 and a subsequent decision by the Gavi Board in December 2021 to approve an initial investment of US$ 155.7 million for the 2022–2025 period. Malaria vaccination was additionally supported by a US$ 56 million investment through a 'de-risk' agreement with manufacturer GSK and innovative financing partner MedAccess. In recognition of the technical requirements of rollout and the need to provide tailored support to countries, a first application window, which closes 13 September, will be limited to the three countries that have taken part in the vaccine’s multi-year pilot programme: Kenya, Ghana and Malawi.
  A second window, which opens at the end of the year and closes in January, is open to other countries with moderate to high transmission of Plasmodium falciparum malaria. These countries can already submit expressions of interest (EoIs) during the first funding window to signal interest and provide them with the needed support to submit quality applications.”
  Gavi, "Gavi opens applications for malaria vaccine rollout support", 2022.

• 19

  “Two vaccine candidates are approaching late-stage clinical evaluation: the R21/MatrixM vaccine candidate targeting PfCSP protein and the attenuated whole sporozoite vaccine PfSPZ.” World Health Organization, "WHO Guidelines for malaria - 3 June 2022", v4.0, p. 130.
  

• 20“Additional candidates targeting other malaria life-cycle stages include the Rh5 blood-stage vaccine candidate [144] and Pfs25 and Pfs230 vaccine candidates targeting sexual-stage antigens to prevent human-to-mosquito transmission (NCT02942277). New technologies, such as DNA- and mRNA-based vaccines [145], the ongoing development of adjuvants [146], and delivery platforms such as virus-like particles (VLPs; the delivery platform used for RTS,S/AS01) and vesicle-based technologies are being explored for use in malaria vaccines.” World Health Organization, "WHO Guidelines for malaria - 3 June 2022", v4.0, pp. 130-31.
  
• 21

  “The R21 vaccine is similar to the RTS,S vaccine, inasmuch as both are virus-like particle-based vaccines based on CSP. R21 particles are formed from a single CSP-hepatitis B surface antigen (HBsAg) fusion protein, which achieves a much higher proportion of CSP displayed on the antigen surface compared with RTS,S. R21 is thus designed to induce a greater anti-CSP antibody response and lower anti-HBsAg antibody response when compared with RTS,S/AS01, and is considered a next-generation RTS,S-like vaccine.57 Like RTS,S/AS01 and PfSPZ Vaccine, R21 targets the parasite’s pre-erythrocytic sporozoite stage. Because it contains a greater amount of CSP per HBsAg particle, R21 may induce higher anti-CSP antibody responses than RTS,S/AS01. As the magnitude of this anti-CSP antibody response to the NANP repeat region seems to correlate with RTS,S/AS01-induced protection,32 the enhanced immune response to CSP induced by R21 may result in superior efficacy and/or duration of protection.” Laurens 2020, p. 486.
  “In this double-blind, randomised, controlled, phase 2b trial, the low-dose circumsporozoite protein-based vaccine R21, with two different doses of adjuvant Matrix-M (MM), was given to children aged 5–17 months in Nanoro, Burkina Faso—a highly seasonal malaria transmission setting. Three vaccinations were administered at 4-week intervals before the malaria season, with a fourth dose 1 year later.”
  Datoo et al. 2021, Summary.

• 22

  “In laboratory studies, the vaccine, called PfSPZ, has proven the most effective malaria vaccine developed so far, giving healthy volunteers complete protection.
  PfSPZ works by eliciting an immune response against the malaria parasite Plasmodium falciparum. It is made of sporozoites (SPZ), the stage in the malaria parasite’s life cycle that infected mosquitoes inject into people during a bite. Sanaria isolates and purifies billions of sporozoites from farmed mosquitoes.
  The vaccine is unique in using whole parasites as its ingredient; most candidate malaria vaccines include only a small number of genetically engineered parasite proteins. The abundance of proteins in the whole parasite vaccine explains why it provokes such a strong immune response.” Butler 2019.
  “In this randomized, double-blind, placebo-controlled trial, 268 healthy Malian children aged 6-10 years, residing in Bancoumana and surrounding villages, will be administered three doses of 9.0x10^5 Pf sporozoites (PfSPZ) of PfSPZ Vaccine (or placebo) at 1, 8, and 29-days using direct venous inoculation (DVI).” Clinical Trials Study Record, "PfSPZ Vaccine Trial in Malian Children," 2022

• 23
  • “The study is a double-blind, randomized, placebo-controlled, Phase 2 clinical trial that will assess the safety, tolerability, immunogenicity and protective efficacy of PfSPZ Vaccine and PfSPZ-CVac against naturally occurring malaria in 372 healthy Indonesian soldiers aged 18-55 years who will be deployed in malarious eastern Indonesia.” Clinical Trials Study Record, "Safety, Tolerability, Immunogenicity and Protective Efficacy of PfSPZ Vaccine and PfSPZ-CVac in Indonesian Adults Against Naturally-Transmitted Malaria," 2022.
  • “This is a randomized, open label, single centre, Phase 2 study. 120 healthy non-pregnant Thai adults, aged 18-55 years, inclusive, will be recruited.
    Each participant will be randomized into one of the following study arms in a ratio of 5:5:2, as follows:
    1. R21/MatrixM + Dihydroartimisinin (DHA)-Piperaquine (PIP)+ primaquine (PQ) (Group 1, n=50) will receive R21/MatrixM + 3 doses DHA-PIP+PQ at Month 0, Month 1 and Month 2
    2. R21/MatrixM only (Group 2, n=50) will receive R21/MatrixM standard dose at Month 0, Month 1 and Month 2
    3. DHA-PIP+PQ only (Group 3, n=20) will receive 3 doses DHA-PIP+PQ at Month 0, Month 1 and Month 2.”

    Clinical Trials Study Record, "Safety and Immunogenicity of the Malaria Vaccine, R21/MatrixM, in Healthy Thai Adults (R21/Matrix-M)," 2022.

• 24“In addition, among 5–17-month-old-children who only received three doses of RTS,S, the initial reduction in severe malaria was counterbalanced by an increase in severe malaria around 18 months after the initial vaccine course, presumably due to waning immunity. This age shift effect has been noted among recipients of other malaria-control interventions when the intervention is withdrawn. Presumably when the intervention group is then compared to a contemporaneously followed control group in the same population who did not receive the intervention and who develop immunity through repeated episodes of natural infection, the intervention group is at comparatively higher risk of malaria and severe disease for a limited period.

  This age shift in severe malaria was most marked in higher transmission settings, possibly because participants in the control group developed immunity through natural infection more rapidly. Importantly, an age shift in severe malaria was not observed up to the end of the follow-up period among children vaccinated at 5-17 months of age who received a fourth dose.”
  World Health Organization, Full Evidence Report on the RTS,S/AS01 Malaria Vaccine, 2021, p. 64.

• 25
  • "We undertook this phase 3, double-blind, observer-blind, individually randomised controlled trial between March 27, 2009, and Jan 31, 2014, at 11 centres in seven countries in sub-Saharan Africa that are situated in areas with different intensities of malaria transmission." RTS,S Clinical Trials Partnership 2015, p. 32.
  • Clinical Trials Study Record, "Efficacy of GSK Biologicals' Candidate Malaria Vaccine 257049 Against Malaria Disease in Infants and Children in Africa," 2019.

• 26
  • “From March 27, 2009, until Jan 31, 2011, children (age 5–17 months) and young infants (age 6–12 weeks) were enrolled at 11 centres in seven countries in sub-Saharan Africa. Participants were randomly assigned (1:1:1) at first vaccination by block randomisation with minimisation by centre to receive three doses of RTS,S/AS01 at months 0, 1, and 2 and a booster dose at month 20 (R3R group); three doses of RTS,S/AS01 and a dose of comparator vaccine at month 20 (R3C group); or a comparator vaccine at months 0, 1, 2, and 20 (C3C [control group]). . . . 8922 children and 6537 young infants were included in the modified intention-to-treat analyses. Children were followed up for a median of 48 months (IQR 39–50) and young infants for 38 months (34–41) after dose 1. . . . ” RTS,S Clinical Trials Partnership 2015, p. 31.
  • Infants were included in the study, but results were weaker than those found for young children: “In young infants, compared with 6170 episodes of clinical malaria that met the primary case definition in the C3C group, 4993 episodes occurred in the R3R group (VE 25·9%, 95% CI 19·9–31·5) and 5444 occurred in the R3C group (18·3%, 11·7–24·4); and compared with 116 infants who experienced at least one episode of severe malaria in the C3C group, 96 infants experienced at least one episode of severe malaria in the R3R group (17·3%, 95% CI −9·4 to 37·5) and 104 in the R3C group (10·3%, −17·9 to 31·8)” RTS,S Clinical Trials Partnership 2015, p. 31.
  • We focus specifically on young children in this report because young children are the recommended target population for the vaccine: “The RTS,S/AS01 malaria vaccine should be provided in a four-dose schedule in children from 5 months of age.” World Health Organization, "WHO Guidelines for malaria - 3 June 2022", v4.0, p. 16.
  • Definition of modified intent-to-treat (ITT): “The modified ITT population included all participants who received at least one dose of vaccine.” RTS,S Clinical Trials Partnership 2015
  • Alternative (comparator) vaccines: The group that received four doses of RTS,S is identified in the study as the R3R group, the group that received three doses is identified in the study as the R3C group, and the control group is identified as the C3C group.
    “One group received RTS,S/AS01 at months 0, 1, and 2, followed by a booster dose at month 20 (R3R group); a second group received the RTS,S/AS01 primary vaccination series with meningococcal serogroup C conjugate vaccine (Menjugate, Novartis, Basel, Switzerland) instead of an RTS,S/AS01 booster (R3C group); and the third group received only comparator vaccines: rabies vaccine (Verorab, Sanofi Pasteur, Paris, France) for children and Menjugate for young infants (C3C [control group]; appendix p 19).” RTS,S Clinical Trials Partnership 2015

• 27“[A]lthough not significant, there is a pattern for the higher point estimates of VE to occur in the lower transmission settings as is the case for VE against clinical malaria and negative point estimates (with large CIs) to occur in the R3C high transmission areas.” RTS,S Clinical Trials Partnership 2015, p. 39.

• 28

  “The research team ensured that insecticide treated bednet use was optimized in each study population. At study start in two study sites (Kilifi, Kenya and Bagamoyo, Tanzania) this was achieved through close collaboration with the respective National Malaria Control Programmes. In the other sites, impregnated bednets were distributed by the study teams to all children who underwent screening, regardless of whether they were eligible for the study. During the course of the study three sites (Agogo, Siaya and Lilongwe) replaced any damaged nets upon the parents’ request. Other centres relied upon the National Malaria Control Programme for the ongoing replacement of bednets.” RTS,S Clinical Trials Partnership 2015, "Supplementary Appendix," pp. 6-7.
  For information about net coverage, see the charts for ITN coverage in RTS,S Clinical Trials Partnership 2015, "Supplementary Appendix," pp. 25-26, Figure S4. Malaria control measures in place at each study site.

• 29

  Two sites (Manhica and Lambarene) had IRS coverage of around 30% of children at the beginning of the study, and those sites also had bednets coverage of only around 60-70% of children at the end of the study. Note also that less than 10% of the trial population overall was covered by IRS. We expect that similar dynamics might apply as in the case of nets. See RTS,S Clinical Trials Partnership 2015, "Supplementary Appendix," pp. 25-26, Figure S4. Malaria control measures in place at each study site.

• 30
  • For IPTi coverage in the intent-to-treat population, see RTS,S Clinical Trials Partnership 2015, "Supplementary Appendix," p. 25, Figure S4. Malaria control measures in place at each study site.
  • SMC: Only one site, Nanoro, Burkina Faso is described as having seasonal risk and according to Malaria Consortium, SMC was not widely scaled throughout most of the trial period.
  • “The location of each participating study site is shown on this previously published map showing the spatial distribution of P. falciparum malaria endemicity. The data are the model-based geostatistical point estimates of the annual mean P. falciparum parasite rate age-standardized for 2-10 years for 2007 within the stable spatial limits of P. falciparum malaria transmission, displayed as a continuum of yellow to red from 0%–100% (see map legend). The rest of the land area was defined as unstable risk (medium grey areas) or no risk (light grey). Nanoro, Burkina Faso has highly seasonal malaria transmission.” RTS,S Clinical Trials Partnership 2015, "Supplementary Appendix," p. 18.
  • “SMC was recommended by WHO in 2012. As of 2014, fewer than 5% (2.99 million) of all eligible children were benefiting from SMC.” Malaria Consortium, SMC at scale - saving lives.

• 31"We randomly assigned 6861 children 5 to 17 months of age to receive sulfadoxine–pyrimethamine and amodiaquine (2287 children [chemoprevention-alone group]), RTS,S/AS01E (2288 children [vaccine-alone group]), or chemoprevention and RTS,S/AS01E (2286 children [combination group])." Chandramohan et al. 2021, Abstract.
  
• 32
  • “During the informed consent process, parents were asked to bring their child to a study health facility as soon as possible if their child fell sick during the study. Malaria was captured by passive case detection. Passive case detection (PCD) is the detection of malaria disease by self-presentation to health facility in the study area. All participating children who presented to a health facility in the study area were evaluated as potential cases of malaria using a standardised algorithm. All parents were asked whether the child had had a fever within the previous 24 hours and all children had their temperature measured. A blood sample was taken for testing for malaria parasites in all children who had had a history of fever during the previous 24 hours or who had a measured axillary temperature ≥ 37.5°C at the time of presentation. Children who needed inpatient treatment were provided transport to a hospital participating in the study.” RTS,S Clinical Trials Partnership 2015, "Supplementary Appendix," p. 8.
  • “SAEs were collected for all participating children throughout the study period, from the time of parental consent. At every visit/contact, information was sought on the occurrence of AEs/SAEs. SAEs were identified by surveillance at health facilities in the study area and through monthly home visits.” RTS,S Clinical Trials Partnership 2015, "Supplementary Appendix," p. 8.

• 33“Treatment of malaria was conducted in accordance with national guidelines. Overall, 99% of children and young infants who presented with confirmed malaria to study clinics received treatment with artemisinin combination therapy (ACT) (Figure S3). In eight of the 11 study sites, the first line treatment for uncomplicated malaria was artemether-lumefantrine whilst in the three other sites (Agogo and Kintampo, Ghana; Nanoro, Burkina Faso), it was artesunate-amodiaquine. The study protocol specified that children admitted to hospital with severe malaria would receive intravenous quinine. During the course of the study, information on the superiority of artesunate over quinine for the treatment of severe malaria became available and as this change in treatment was introduced at country level, artemisinin preparations were used in preference to quinine.” RTS,S Clinical Trials Partnership 2015, "Supplementary Appendix," p. 9.

• 34We read the abstract of one study suggesting delayed careseeking is strongly associated with clinical malaria onset but have not reviewed the study in detail:
  “A search using Ovid MEDLINE and Embase was initially conducted to identify studies on severe Plasmodium falciparum malaria that included information on treatment delay, such as fever duration (inception to 22nd September 2017). Studies identified included 5 case–control and 8 other observational clinical studies of SM and UM cases. Risk of bias was assessed using the Newcastle–Ottawa scale, and all studies were ranked as ‘Good’, scoring ≥7/10. Individual-patient data (IPD) were pooled from 13 studies of 3,989 (94.1% aged 15 years) SM patients and 5,780 (79.6% aged 15 years) UM cases in Benin, Malaysia, Mozambique, Tanzania, The Gambia, Uganda, Yemen, and Zambia. Definitions of SM were standardised across studies to compare treatment delay in patients with UM and different SM phenotypes using age-adjusted mixed-effects regression. The odds of any SM phenotype were significantly higher in children with longer delays between initial symptoms and arrival at the health facility (odds ratio [OR] = 1.33, 95% CI: 1.07–1.64 for a delay of >24 hours versus ≤24 hours; p = 0.009). Reported illness duration was a strong predictor of presenting with severe malarial anaemia (SMA) in children, with an OR of 2.79 (95% CI:1.92–4.06; p 0.001) for a delay of 2–3 days and 5.46 (95% CI: 3.49–8.53; p 0.001) for a delay of >7 days, compared with receiving treatment within 24 hours from symptom onset. We estimate that 42.8% of childhood SMA cases and 48.5% of adult SMA cases in the study areas would have been averted if all individuals were able to access treatment within the first day of symptom onset, if the association is fully causal. In studies specifically recording onset of nonsevere symptoms, long treatment delay was moderately associated with other SM phenotypes (OR [95% CI] >3 to ≤4 days versus ≤24 hours: cerebral malaria [CM] = 2.42 [1.24–4.72], p = 0.01; respiratory distress syndrome [RDS] = 4.09 [1.70–9.82], p = 0.002). In addition to unmeasured confounding, which is commonly present in observational studies, a key limitation is that many severe cases and deaths occur outside healthcare facilities in endemic countries, where the effect of delayed or no treatment is difficult to quantify.” Mousa et al. 2020.
  
• 35

  “First, at the time of the initial analysis of severe malaria risk in 5–17-month-old children between the 3 and 4 dose groups, it was assumed that up until the time of the 4th dose, the 3 and 4 dose groups were equivalent, and thus were treated as a single group in analysis. However additional analysis revealed that, in the pre 4th dose period, there was a higher risk of severe malaria in those randomized to the 3-dose arm than those randomized to the 4-dose arm (Figure 14).” World Health Organization, Full Evidence Report on the RTS,S/AS01 Malaria Vaccine, 2021, pp. 64-65.

• 36
  • “Further analysis by GSK at the request of WHO indicated no problem with randomization, the difference therefore arose by chance. The risk of clinical malaria was similar in the 2 arms. However, this unexpected difference may have complicated the interpretation of the data over the whole study period and contributed to a potential overestimation of the importance of the 4th dose.” World Health Organization, Full Evidence Report on the RTS,S/AS01 Malaria Vaccine, 2021, p. 65.
  • Preliminary results from MVIP countries found that coverage of the booster dose averaged about 52% of coverage of the first dose across all three countries, although coverage of the fourth dose appeared to improve over time in Malawi and Ghana.
    • Malawi: RTS,S 1: 77%, RTS,S 4: 39% = coverage of fourth dose is about 51% of first dose coverage
    • Ghana: RTS,S 1: 70%, RTS,S 4: 38% = coverage of fourth dose is about 54% of first dose coverage;
    • Kenya RTS,S 1: 80%, RTS,S 4: 41% = coverage of fourth dose is about 51% of first dose coverage
    • Change over time: Malawi fourth dose coverage rose from 28% in 2020 to 46% in 2021. Ghana fourth dose coverage rose from 30% in 2020 to 42% in 2021.

    See World Health Organization, Full Evidence Report on the RTS,S/AS01 Malaria Vaccine, 2021, p. 47, Table 1.

• 37

  “In the modified ITT population, from month 0 until study end, compared with 9585 episodes of clinical malaria that met the primary case definition in children in the C3C group, 6616 episodes occurred in children in the R3R [4 dose] group (VE 36·3%, 95% CI 31·8–40·5) and 7396 in the R3C [3 dose] group (28·3%, 23·3–32·9; table 1; appendix pp 28, 69).” RTS,S Clinical Trials Partnership 2015, p. 38.

• 38

  “In the modified ITT population, from month 0 until study end, compared with 171 children in the C3C group, 116 children in the R3R [4 dose] group (VE 32·2%, 95% CI 13·7 to 46·9) and 169 in the R3C [3 dose] group (1·1%, −23·0 to 20·5) experienced at least one episode of severe malaria that met the primary case definition (table 2; appendix p 69).” RTS,S Clinical Trials Partnership 2015, p. 38.
  

• 39

  “In the modified ITT population, from month 0 until study end, compared with 9585 episodes of clinical malaria that met the primary case definition in children in the C3C group, 6616 episodes occurred in children in the R3R [4 dose] group (VE 36·3%, 95% CI 31·8–40·5) and 7396 in the R3C [3 dose] group (28·3%, 23·3–32·9; table 1; appendix pp 28, 69).” RTS,S Clinical Trials Partnership 2015, p. 38.

• 40“In the modified ITT population, from month 0 until study end, compared with 171 children in the C3C group, 116 children in the R3R [4 dose] group (VE 32·2%, 95% CI 13·7 to 46·9) and 169 in the R3C [3 dose] group (1·1%, −23·0 to 20·5) experienced at least one episode of severe malaria that met the primary case definition (table 2; appendix p 69).” RTS,S Clinical Trials Partnership 2015, p. 38.
  
• 41“8922 children and 6537 young infants were included in the modified intention-to-treat analyses. Children were followed up for a median of 48 months (IQR 39–50) ” RTS,S Clinical Trials Partnership 2015, p. 31.

  The results provided in the table are taken from RTS,S Clinical Trials Partnership 2015, p. 36, Table 1.
  

• 42

  See RTS,S Clinical Trials Partnership 2015, p. 38, Table 2.

• 43

  See RTS,S Clinical Trials Partnership 2015, p. 38, Table 2.

• 44
  • “[A]lthough not significant, there is a pattern for the higher point estimates of VE to occur in the lower transmission settings as is the case for VE against clinical malaria and negative point estimates (with large CIs) to occur in the R3C high transmission areas.” RTS,S Clinical Trials Partnership 2015, p. 39.
  • “This increased risk occurred predominantly in sites with a higher level of malaria transmission. Why children who received the RTS,S/AS01 primary vaccination series but who did not receive a booster dose were at increased risk of severe malaria during the latter part of the study is uncertain. This finding might have occurred by chance; the number of cases was low and a similar pattern was not noted for cases of uncomplicated malaria. However, vaccination, by providing protection against malaria infection, might have reduced the natural acquisition of immunity obtained through repeated infections, making these children more susceptible when the vaccine effect waned.” RTS,S Clinical Trials Partnership 2015, p. 42.

• 45

  See RTS,S Clinical Trials Partnership 2015, p. 38, Table 2.

• 46For months 0-20, 21-32 and 33-study end the proportions affected were:
  • For the control group children: 0.04, 0.02, 0.01
  • For the three-dose group children: 0.03 (combined with four-dose group results), 0.02, 0.01
  • For the four-dose group children: 0.03 (combined with three-dose group results), 0.02, and 0.01

  See RTS,S Clinical Trials Partnership 2015, p. 38, Table 2.
  

• 47
  • “No significant effect on overall mortality, malaria mortality, pneumonia, or sepsis was noted in either age category. The latter finding is surprising because malaria seems to be an important risk factor for invasive bacterial infections. Failure to detect an effect of RTS,S/AS01 on these secondary outcomes, including mortality, might have been caused in part by the high level of clinical care provided during the trial, including high coverage with insecticide-treated bednets and enhanced access to effective treatment of malaria and other disorders.” RTS,S Clinical Trials Partnership 2015, pp. 42-43.
  • For point estimates, see RTS,S Clinical Trials Partnership 2015, "Supplementary Appendix," p. 76, Table S13.

• 48
  • “No significant effect on overall mortality, malaria mortality, pneumonia, or sepsis was noted in either age category. The latter finding is surprising because malaria seems to be an important risk factor for invasive bacterial infections. Failure to detect an effect of RTS,S/AS01 on these secondary outcomes, including mortality, might have been caused in part by the high level of clinical care provided during the trial, including high coverage with insecticide-treated bednets and enhanced access to effective treatment of malaria and other disorders.” RTS,S Clinical Trials Partnership 2015, pp. 42-43.
  • For point estimates, see RTS,S Clinical Trials Partnership 2015, "Supplementary Appendix," p. 76, Table S13.

• 49The study found all-cause mortality in the control groups was 1.5% among children aged 5-17. ee RTS,S Clinical Trials Partnership 2015, "Supplementary Appendix," p. 76, table S13, all-cause mortality, case definition 1, C3C results columns. Data over the same period (2009 to 2013) for the countries involved in the study (Burkina Faso, Ghana, Gabon, Kenya, Malawi, Mozambique, and Tanzania) showed an average post-neonatal under-5 mortality rate of 5.0%, and a post-infancy under-5 mortality rate of 2.7%. See national data here.
  
• 50

  Rates are approximations based on eyeballing figure 14 from World Health Organization, Full Evidence Report on the RTS,S/AS01 Malaria Vaccine, 2021, p. 65.The figure does not report point estimates or confidence intervals, and we are unsure if the rates of difference between the groups are statistically significant. In addition, the figure does not define the population included (e.g., ITT or per protocol), so it may not correspond exactly to the population reported in phase-3 trial results.

• 51

  World Health Organization, Full Evidence Report on the RTS,S/AS01 Malaria Vaccine, 2021, p. 65.

• 52

  "In the modified ITT population, from month 0 until study end, compared with 9585 episodes of clinical malaria that met the primary case definition in children in the C3C group, 6616 episodes occurred in children in the R3R [“4 dose”] group (VE 36·3%, 95% CI 31·8–40·5) and 7396 in the R3C [“3 dose”] group (28·3%, 23·3–32·9; table 1; appendix pp 28, 69)." RTS,S Clinical Trials Partnership 2015, p. 38.

• 53

  "In the modified ITT population, from month 0 until study end, compared with 9585 episodes of clinical malaria that met the primary case definition in children in the C3C group, 6616 episodes occurred in children in the R3R [“4 dose”] group (VE 36·3%, 95% CI 31·8–40·5) and 7396 in the R3C [“3 dose”] group (28·3%, 23·3–32·9; table 1; appendix pp 28, 69)." RTS,S Clinical Trials Partnership 2015, p. 38.

• 54

  "In the modified ITT population, from month 0 until study end, compared with 171 children in the C3C group, 116 children in the R3R [“4 dose”] group (VE 32·2%, 95% CI 13·7 to 46·9) and 169 in the R3C [“3 dose”] group (1·1%, −23·0 to 20·5) experienced at least one episode of severe malaria that met the primary case definition (table 2; appendix p 69)." RTS,S Clinical Trials Partnership 2015, p. 38.

• 55
  • The original study enrolled 8922 children, and the extension study followed 1739 of them (19.5%).
    "8922 children and 6537 young infants were included in the modified intention-to-treat analyses. Children were followed up for a median of 48 months (IQR 39–50) and young infants for 38 months (34–41) after dose 1. . . . ”RTS,S Clinical Trials Partnership 2015, p. 31.
  • "We enrolled 1739 older children (aged 5–7 years) and 1345 younger children (aged 3–5 years). During the 3-year extension, 66 severe malaria cases were reported, resulting in severe malaria incidence of 0·004 cases per person-years at risk (PPY; 95% CI 0–0·033) in the four-dose group, 0·007 PPY (0·001–0·052) in the three-dose group, and 0·009 PPY (0·001–0·066) in the control group in the older children category and a vaccine efficacy against severe malaria that did not contribute significantly to the overall efficacy (four-dose group 53·7% [95% CI −13·7 to 81·1], p=0·093; three-dose group 23·3% [–67·1 to 64·8], p=0·50).
    Overall, severe malaria incidence was low in all groups, with no evidence of rebound in RTS,S/AS01 recipients, despite an increased incidence of clinical malaria in older children who received RTS,S/AS01 compared with the comparator group in Nanoro. No safety signal was identified."
    Tinto et al. 2019.

• 56

  Tinto et al. 2019, Supplementary webappendix, Appendix 6, shows that vaccine efficacy in the “post dose 4 (initial study)" period for the three-dose group was positive. In contrast, vaccine efficacy for the threedose group during the same time period was −41.0% across all sites. See RTS,S Clinical Trials Partnership 2015, p. 38, Table 2.
  

• 57

  “We have, therefore, undertaken a pooled analysis of existing data from multiple sites to enable a comprehensive overview of the age-patterns of malaria outcomes under different epidemiological conditions in sub-Saharan Africa. . . . The percentage of each outcome by age (0–10 years - excluding neonates for malaria-diagnosed mortality) was calculated for each study.” Carneiro et al. 2010.

• 58

  “Malaria-diagnosed mortality is more focussed in younger children than admissions with malaria in all settings for which there are comparative data. Once more there is a distinctive shift of the peak age towards infants as transmission becomes more intense. The median age for malaria-diagnosed mortality ranges from 12 months (IQR: 6, 22) at high, not markedly seasonal transmission to 28 months (IQR: 15, 51) at medium and markedly seasonal transmission.” Carneiro et al. 2010.

• 59

  See our analysis here: GiveWell, Carneiro estimates of child malaria mortality age-distribution for RTS,S BOTEC, 2022.

• 60

  See our analysis here: GiveWell, Carneiro estimates of child malaria mortality age-distribution for RTS,S BOTEC, 2022.

• 61

  "Clinical malaria is relatively evenly distributed across all ages with a shift towards younger age groups as transmission intensity increases, both in areas of non-marked and marked seasonality. The median age for clinical malaria ranges from 32 months (Inter-quartile range (IQR): 15, 61) in settings of highly intense and not markedly seasonal transmission, to 72 months (IQR: 45, 97) in settings of low intensity and markedly seasonal transmission. . . . Hospital admissions with malaria parasites are more concentrated in younger children than is clinical malaria in all settings, and these severe cases become more concentrated in younger ages with increasing transmission intensity and less seasonality. The median age ranges from 17 months (IQR: 10, 29) at high, not markedly seasonal transmission to 36 months (IQR: 20, 60) at low and markedly seasonal transmission.
  Malaria-diagnosed mortality is more focussed in younger children than admissions with malaria in all settings for which there are comparative data. Once more there is a distinctive shift of the peak age towards infants as transmission becomes more intense. The median age for malaria-diagnosed mortality ranges from 12 months (IQR: 6, 22) at high, not markedly seasonal transmission to 28 months (IQR: 15, 51) at medium and markedly seasonal transmission.” Carneiro et al. 2010.

• 62

  Included studies covered periods ranging from 1982 to 2007. See Carneiro et al. 2010, Table S1.
  

• 63

  We have seen one paper suggesting that mortality burden shifts to older children when transmission drops, but have not investigated this topic:
  "With declining transmission, there have been shifts in cases to older ages. For example, in south-western Senegal, a 30-fold drop in malaria incidence between 1996 and 2010 was accompanied by a shift in the age distribution of cases, with 34% of cases in the under-fives in 1996 falling to ~5% in 2010 (ref. 3). Similarly, in western Gambia, a rapid fall in the proportion of malaria admissions between 2003 and 2007 was accompanied by an increase in the mean age of paediatric malaria admissions from 3.9 years to 5.6 years. This changing distribution of cases is most likely due to the slower development of naturally acquired immunity."
  Griffin, Ferguson and Ghani, 2014.

• 64
  • “The MVIP is coordinated by WHO in close collaboration with ministries of health (MoH) in the three participating countries —Ghana, Kenya, Malawi—and a range of in-country and international partners. The MoH of the pilot countries have introduced the RTS,S/AS01 vaccine through their childhood immunization services using routine vaccine introduction strategies and methods. . . . Within the pilot region in each country, districts or similar areas were randomized to introduce the vaccine in 2019, or to delay introduction until a decision is reached about safety and effectiveness.” World Health Organization, Full Evidence Report on the RTS,S/AS01 Malaria Vaccine, 2021, p. 7.
  • “The RTS,S/AS01 malaria vaccine is being introduced sub-nationally in phased pilot introductions through the EPI programmes in Malawi Ghana and Kenya." Clinical Trials Study Record, "Malaria Vaccine Pilot Evaluation (MVPE)," 2021.
  • See vaccine coverage estimates for different time periods in World Health Organization, Full Evidence Report on the RTS,S/AS01 Malaria Vaccine, 2021, p. 47, Table 1.

• 65
  • “The RTS,S/AS01 malaria vaccine is being introduced sub-nationally in phased pilot introductions through the EPI programmes in Malawi Ghana and Kenya. Vaccine introduction is by the respective MoH in selected areas randomly assigned to receive the vaccine at the beginning of the pilots. In the context of this programmatic activity, the Malaria Vaccine Pilot Evaluation (MVPE) registered here as observational evaluations during early vaccine introduction, include a series of 3 household surveys, and sentinel hospital and community mortality surveillance, building on routine systems. These observational evaluations will measure:
    1. The programmatic feasibility of delivering a 4 dose schedule;
    2. Safety in routine use, with focus on cerebral malaria and meningitis;
    3. The impact of the malaria vaccine in routine use on severe malaria and all-cause mortality.” Clinical Trials Study Record, "Malaria Vaccine Pilot Evaluation (MVPE)," 2021.
  • “The Malaria Vaccine Pilot Evaluation-Case Control (MVPE-CC) registered here as observational study is embedded within MVPE comprising case-control studies of clinical and mortality outcomes. Each case will require four controls, and caregiver informed consent will be required prior to study activities.

    These observational case control studies will measure as complementary information to what is being collected through MVPE:

    1. Safety among children who received the malaria vaccine, with focus on cerebral malaria, meningitis and severe malaria
    2. The impact of the malaria vaccine on all-cause mortality for boys and girls, AND
    3. Promote use of case-control approaches by Expanded Programmes on Immunization (EPI) and malaria control programmes.”

    Clinical Trials Study Record, "Strengthening the Evidence for Policy on the RTS,S/AS01 Malaria Vaccine (MVPE-CC)," 2021

• 66“Primary Outcome Measures :
  1. Number of children admitted with a diagnosis of probable and confirmed meningitis [ Time Frame: Through study completion, an average of 1 year ]...
  2. Number of children admitted with a diagnosis of severe malaria [ Time Frame: Through study completion, an average of 1 year ]...
  3. The number of deaths of any cause [ Time Frame: Through study completion, an average of 1 year ]...
  4. Number of deaths in children by gender [ Time Frame: Through study completion, an average of 1 year ]...

  Secondary Outcome Measures :

  1. Number of children with a diagnosis of severe malaria in relation to the 4th dose of RTS,S [ Time Frame: Through study completion, an average of 1 year ]...
  2. Number of children with a diagnosis of cerebral malaria in relation to the 4th dose of RTS,S [ Time Frame: Through study completion, an average of 1 year ]...
  3. Number of deaths of any cause in relation to the 4th dose of RTS,S [ Time Frame: Through study completion, an average of 1 year ]...
  4. Number of deaths among girls in relation to the 4th dose of RTS,S [ Time Frame: Through study completion, an average of 1 year ]...” Clinical Trials Study Record, "Strengthening the Evidence for Policy on the RTS,S/AS01 Malaria Vaccine (MVPE-CC)," 2021.

• 67

  Among children eligible to have received all three primary doses of RTS,S/AS01, there were a total of 1107 admissions with severe malaria (P. falciparum infection with severe anaemia, or respiratory distress, or with impaired consciousness or convulsions but not meeting criteria for meningitis), 418 from implementation areas and 689 from comparison areas. Among children who were not eligible to have received any doses of RTS,S/AS01 there were 1313 patients admitted from implementation areas and 1390 from comparison areas. The incidence rate ratio comparing incidence of admission with severe malaria between implementation and comparison areas was 0.70 (95%CI 0.54, 0.92), a reduction of 30% (95%CI 8%, 46%) in the context of overall vaccine coverage during the first two years of vaccine introduction of approximately 60-70%. World Health Organization, Full Evidence Report on the RTS,S/AS01 Malaria Vaccine, 2021, p. 45.
  

• 68
  • WHO’s report specifies a 7% reduction, however GiveWell’s calculations imply a reduction of approximately 10%. We expect to follow up about this discrepancy with deeper investigation.
  • WHO: "Excluding deaths due to injury, among children eligible to have received three doses of RTS,S/AS01, there were a total of 2864 deaths reported, 1421 from implementing regions and 1443 from comparison regions. In children who were not eligible to have received the vaccine there were 4218 deaths in implementing regions and 3874 in comparison regions. The mortality ratio in the vaccine-eligible age group between implementing and comparison regions, was 0.93 (95%CI 0.84,1.03), a 7% reduction (95%CI -3%,16%)." World Health Organization, Full Evidence Report on the RTS,S/AS01 Malaria Vaccine, 2021, p. 42.
  • GiveWell: We calculate a 10% reduction in the ratio as (1421/4218) divided by (1443/3874).

• 69

  "As expected, there was insufficient power at this point to detect an effect on mortality (~13 500 child deaths were recorded through the mortality surveillance system, while to achieve 90% power to demonstrate a 10% reduction in mortality, 24 000 deaths will need to have accumulated). Nonetheless, the 7% impact on mortality (not statistically significant) measured through the MVPE is consistent with what would be expected if malaria contributes to about 30% of deaths in young children (based on a 25% reduction in severe malaria as a proxy for malaria related mortality)."
  World Health Organization, Full Evidence Report on the RTS,S/AS01 Malaria Vaccine, 2021, p. 80.
  

• 70
  • See World Health Organization, Full Evidence Report on the RTS,S/AS01 Malaria Vaccine, 2021, p. 47, Table 1.
  • "The modified ITT population included all participants who received at least one dose of vaccine." RTS,S Clinical Trials Partnership 2015, p. 35.

• 71

  “Among age groups too old to receive SMC, incidence was reduced by 26% (95% CI 18%–33%, p 0.001) (RDT-confirmed), or 29% (95% CI 21%–35%, p 0.001) (confirmed and unconfirmed cases) in areas where SMC was delivered to children compared to control areas.” Cisse et al. 2016.

• 72
  • Taken at face value, the phase-3 clinical trial results suggest that rebound is an issue 20 months or more after initial vaccination (see discussion above). Because these results followed children for approximately 48 months and booster doses were administered at month 20, the phase-3 results do not contain much data from four-dose children after we would expect immunity to wane. Results from the extension study do not suggest risk of rebound for the four-dose group. However, as discussed above, we do not consider extension results to be likely to be representative of the trial sites as a whole.
  • “During the 3-year extension, 66 severe malaria cases were reported, resulting in severe malaria incidence of 0·004 cases per person-years at risk (PPY; 95% CI 0–0·033) in the four-dose group, 0·007 PPY (0·001–0·052) in the three-dose group, and 0·009 PPY (0·001–0·066) in the control group in the older children category and a vaccine efficacy against severe malaria that did not contribute significantly to the overall efficacy (four-dose group 53·7% [95% CI −13·7 to 81·1], p=0·093; three-dose group 23·3% [–67·1 to 64·8], p=0·50).” Tinto et al. 2019, Summary.

• 73

  “Three potential safety signals were noted in the Phase 3 trial. First, in children in the older age category, a higher number of meningitis cases occurred in the malaria vaccine group compared to the control group. However, excess meningitis cases were not temporally related to the timing of vaccine doses, were clustered at 2 of 11 trial sites, and there were a range of etiologies in the cases identified. In addition, an excess of meningitis was not seen in children vaccinated in the younger age group. Whether the increase in meningitis was due to chance or represented a true adverse effect of the vaccine was unknown. Second, in children in the older age group, in the context of a statistically significant decrease in all forms of severe malaria combined, there was an increased number of cerebral malaria cases (a subset of severe malaria) in the malaria vaccine groups compared with the control group. This finding was from an unplanned post-hoc analysis and its significance in relation to vaccination was unclear. An excess of cerebral malaria was not seen in children vaccinated in the younger age group. Third, and also in an unplanned post hoc analysis, there was an imbalance in mortality among girls, with about 2-fold higher deaths among girls who received RTS,S/AS01 than among girls who received comparator vaccines (p=0.001); the ratio of deaths among boys was slightly lower in the RTS,S/AS01 arms versus the control arm. A relationship between the RTS,S/AS01 vaccine and these findings has not been established. The EMA and WHO advisory bodies concluded that all these described safety signals may have arisen by chance.” World Health Organization, Full Evidence Report on the RTS,S/AS01 Malaria Vaccine, 2021, p. 6.

• 74

  “The incidence rate ratio comparing rates of admission with meningitis in implementation and comparison areas, among vaccine-eligible children, was 0.81 (95%CI 0.43, 1.55). There was therefore no evidence that introduction of the malaria vaccine led to an increase in the incidence of hospital admission with meningitis, and there were sufficient cases, and high coverage of the vaccine, to detect an excess of the magnitude observed in the Phase 3 trial.” World Health Organization, Full Evidence Report on the RTS,S/AS01 Malaria Vaccine, 2021, p. 40.

• 75

  “Therefore, there was no evidence that introduction of the malaria vaccine led to an increase in the incidence of hospital admission with cerebral malaria, and there were sufficient cases to detect an excess of the magnitude observed in the Phase 3 trial, if it was present.” World Health Organization, Full Evidence Report on the RTS,S/AS01 Malaria Vaccine, 2021, p. 41.

• 76

  “There was no evidence that the mortality ratio differed between girls and boys (p 0.343). The mortality ratio in girls was 0.98 and in boys 0.90, yielding a relative mortality ratio (girls:boys) of 1.08 (95%CI 0.92,1.28). When analysis was extended to children eligible to have received at least one dose of vaccine, similar results were obtained (ratio of mortality ratios: 1.08 (95%CI 0.93, 1.25), p value for the interaction 0.321). Similar results were also obtained when the analysis was repeated for different age groups of eligible children (mortality ratio girls:boys, in eligible children under 18 months of age, was 1.10, 95%CI 0.94, 1.29, and in eligible children aged 18 months and above, 0.95, 95%CI 0.70, 1.31). Therefore, there was no evidence that the effect of RTS,S/AS01 introduction on all-cause mortality differed between girls and boys in this age group, and there were sufficient deaths to detect an excess of the magnitude observed in the phase 3 trial, if it was present.” World Health Organization, Full Evidence Report on the RTS,S/AS01 Malaria Vaccine, 2021, p. 42.
  

• 77

  “There was no evidence that the mortality ratio differed between girls and boys (p 0.343). The mortality ratio in girls was 0.98 and in boys 0.90, yielding a relative mortality ratio (girls:boys) of 1.08 (95%CI 0.92,1.28). When analysis was extended to children eligible to have received at least one dose of vaccine, similar results were obtained (ratio of mortality ratios: 1.08 (95%CI 0.93, 1.25), p value for the interaction 0.321). Similar results were also obtained when the analysis was repeated for different age groups of eligible children (mortality ratio girls:boys, in eligible children under 18 months of age, was 1.10, 95%CI 0.94, 1.29, and in eligible children aged 18 months and above, 0.95, 95%CI 0.70, 1.31). Therefore, there was no evidence that the effect of RTS,S/AS01 introduction on all-cause mortality differed between girls and boys in this age group, and there were sufficient deaths to detect an excess of the magnitude observed in the phase 3 trial, if it was present.” World Health Organization, Full Evidence Report on the RTS,S/AS01 Malaria Vaccine, 2021, p. 42.

• 78

  “The Phase 4 studies are designed to: a) assess a potential association between vaccination with RTS,S/AS01 and the safety signals observed in the Phase 3 trial; and b) assess any potential association between vaccination and other adverse events of special interest (Phase 4 AESIs); which include rare potential immune-mediated disorders, and other AEFI leading to hospitalization or death (these outcomes were selected as part of a general safety evaluation, and are not related to specific prior safety signals); and c) assess vaccine effectiveness.” World Health Organization, Full Evidence Report on the RTS,S/AS01 Malaria Vaccine, 2021, p. 32.

• 79

  “Hospital-based surveillance systematically documented admissions to the paediatric ward in order to capture information on impact (malaria-specific mortality, severe malaria) and safety (changes in the hospital-based incidence rates of meningitis, cerebral malaria, febrile convulsions, other illnesses, all-cause and malaria-specific mortality. Relevant demographic, vaccination and clinical data were captured in a CRF on all children under 5 years of age admitted to the paediatric wards of sentinel hospitals. Consolidated, quality assured, inpatient surveillance systems were supported by evaluation partners in each country with minimum standards assured to enable systematic, standardized clinical and laboratory assessment and management of all admissions.” World Health Organization, Full Evidence Report on the RTS,S/AS01 Malaria Vaccine, 2021, p. 31.

• 80

  “In this double-blind, randomised, controlled, phase 2b trial, the low-dose circumsporozoite protein-based vaccine R21, with two different doses of adjuvant Matrix-M (MM), was given to children aged 5–17 months in Nanoro, Burkina Faso—a highly seasonal malaria transmission setting. . . . Group 1 received 5 μg R21 plus 25 μg MM, group 2 received 5 μg R21 plus 50 μg MM, and group 3, the control group, received rabies vaccinations. . . . From May 7 to June 13, 2019, 498 children aged 5–17 months were screened, and 48 were excluded. 450 children were enrolled and received at least one vaccination. 150 children were allocated to group 1, 150 children were allocated to group 2, and 150 children were allocated to group 3. . . . At the 6-month primary efficacy analysis, 43 (29%) of 146 participants in group 1, 38 (26%) of 146 participants in group 2, and 105 (71%) of 147 participants in group 3 developed clinical malaria. Vaccine efficacy was 74% (95% CI 63–82) in group 1 and 77% (67–84) in group 2 at 6 months.” Datoo et al. 2021.

• 81

  “While the reported efficacy in this trial is higher than many of the published RTS,S trials, there is no direct comparison between vaccinations timed to occur with the beginning of each malaria season, and case detection methods differ. The question of superiority therefore remains unanswered.” Moorthy and Binka 2021.

• 82
  • “In this double-blind, randomised, controlled, phase 2b trial, the low-dose circumsporozoite protein-based vaccine R21, with two different doses of adjuvant Matrix-M (MM), was given to children aged 5–17 months in Nanoro, Burkina Faso—a highly seasonal malaria transmission setting.” Datoo et al. 2021.
  • “The R21 trial lasted for one year, but Burkina Faso is plagued by malaria for only about six months of each year, notes Stephen Hoffman, chief executive of Sanaria, a company in Rockville, Maryland, that is also developing malaria vaccines. During the second half of the study, there was only one case of malaria in the control group that did not receive the vaccine, Hoffman notes, making it impossible to judge whether the benefits of the vaccine lasted for the full year.” Ledford 2021.

• 83

  “Doses were administered before the seasonal peak of malaria transmission starting in July.” Datoo et al. 2021.

• 84

  RTS,S Clinical Trials Partnership 2014, Supplemental materials, Figure S6, Supplementary Figure 6.a, Panel A (Children 5-17 months: 1-6 months, Nanoro).

• 85

  "Estimated Study Completion Date: December 2023." Clinical Trials Study Record, "R21/Matrix-M in African Children Against Clinical Malaria," 2022.

• 86

  See World Health Organization, Framework for the allocation of limited malaria vaccine supply, 2022, p. 21, Table 2.

• 87These figures are available through the Institute for Health Metrics and Evaluation, GBD Results Tool, under-5 deaths in Democratic Republic of the Congo, 2019. See our adjusted figure here. Note that we approximate mortality rates for ages 6 to 59 months and make the simplifying assumption that these describe mortality during the period for which we model vaccine efficacy (roughly 6-54 months).
  
• 88

  See our informal write up here describing the rationale for these adjustments.

• 89

  The phase-3 clinical trials reported outcomes for the incidence of clinical malaria and the risk of experiencing any case of severe malaria. We use the latter estimate in our cost-effectiveness estimate because (a) it measures severe malaria, which we think is more closely correlated with mortality outcomes, and (b) using an incidence measure implicitly assumes that the risk of mortality is equal for the first and later exposures to malaria. To the extent that vaccination provides protection for the overall number of severe malaria cases experienced (and a higher number of cases contributes to mortality outcomes), we may be underestimating the benefits of vaccination.
  

• 90See our estimates here.
  
• 91

  See our informal write up here.
  

• 92

  "The Board of Gavi, the Vaccine Alliance today approved an investment to support the malaria vaccine introduction, procurement and delivery for Gavi-eligible countries in sub-Saharan Africa in 2022-2025. An initial investment of US$ 155.7 million for 2022-2025 will initiate the implementation of this additional tool in the fight against malaria. The introduction of the RTS,S malaria vaccine to currently recommended malaria control interventions could help drive down child mortality in Africa – a continent that bears the heaviest malaria burden." Gavi, "Gavi Board approves funding to support malaria vaccine roll-out in sub-Saharan Africa," 2021.

• 93

  See RTS,S Clinical Trials Partnership 2015, Tinto et al. 2019, and World Health Organization, Full Evidence Report on the RTS,S/AS01 Malaria Vaccine, 2021.

• 94

  See GiveWell's non-verbatim summary of a conversation with PATH and the World Health Organization, January 5, 2022 and GiveWell's non-verbatim summary of a conversation with PATH, July 30, 2021.