Malaria Vaccine Implementation Programme Framework for Policy Decision Mary J Hamel, IVR SAGE 17 April 2018 1
Questions for SAGE on the Framework for Policy Decision 1. Does SAGE agree with the approach? 2. Are the suggested outcomes and matrices useful for policy decision? 3. Does SAGE agree on the suggested next steps? 2
Framework for Policy Decision for RTS,S MPAC and SAGE requested data be collected through the pilot implementations to answer questions on feasibility, safety, impact to inform a policy decision on wider use of RTS,S Framework for Policy Decision aims to describe how data will inform policy at the end of the pilots, in 2022 Also will describe how data could inform 1. Expansion of vaccinations into pilot comparator areas 2. Broader country-wide implementation prior to 2022 1 should emerging findings show: Concerns about safety resolved Implementation data favorable Fourth dose coverage high 3 1. JTEG Background Paper on the RTS,S/AS01 Malaria Vaccine, Sep 2015
Overview of MVIP timelines: Data accumulating over time Phase 1 Phase 2 2017 2018 2019 2020 2021 2022 Ongoing review of MVIP data and regular updates to SAGE/MPAC??? Vaccine implementation Safety data Sentinel hospital surveillance Routine pharmacovigilance Joint regulatory review Authorization decision RTS,S launch GSK EPI MAL 002 pharmacovigilance baseline study 4 th dose for first children Potential policy recommendation Accumulating info GSK EPI MAL 003 Phase 4 study on safety, effectiveness, impact Feasibility data Administrative data monitoring Household surveys Qualitative longitudinal study Baseline Coverage of dose 1-3 Coverage of dose 4 Indicates 3 rd country New vaccine post-introduction evaluation Health economic assessments PIE Vaccine delivery cost analysis Budget impact analysis Impact data Community-based mortality surveillance Sentinel 4 hospital surveillance Impact on severe malaria Impact on mortality
Benefits of Developing a Framework for Policy Decision SAGE and MPAC members will refine ideas on the relative contribution of the collected data (feasibility, safety, impact) to a future policy recommendation Provide clarity on the expected use of the data in anticipation of potential changes in SAGE membership between the time the SAGE/MPAC recommendations were made and the programme end (2022) Funders, potential funders, and manufacturers can refer to the framework for planning purposes Reducing the likelihood of gaps in funding or vaccine availability should the vaccine be recommended for broader use 5
Questions to be Considered for the Framework for Policy Decision What criteria, if met, would likely lead to a recommendation for vaccine use at the end of the pilot programme What to do if conflicting findings from different countries Or if data availability lags considerably from one country What criteria, if met, would likely lead to a recommendation not to implement the vaccine Is it conceivable that there could be an earlier policy recommendation, prior to pilot end If yes, what data would support such a decision 6
Questions to be Considered for the Framework for Policy Decision: Broader Implementation Before Study End What findings would support or delay expansion into the pilot comparator areas What criteria would support favorable implementation data, and broader country-wide implementation of RTS,S High coverage dose 4, safety signals resolved and: No or little adverse effect on other vaccines? Continued use of malaria interventions, or impact data suggesting no negative effect of reduced use? Cost effectiveness? What would be considered high fourth dose coverage Can thresholds of vaccine coverage that predict impact and other criteria be considered a priori and be used to guide decisions on country-wide expansion of vaccine use before pilot end 7
Criteria e.g vaccine coverage MVIP Framework for Decision Making Recommendation for broader use Very Likely e.g. high (?>X%) coverage of doses 1 4, safety concerns resolved Need for nuanced discussion Recommendation for broader use Very Unlikely e.g. poor (<X%) coverage of doses 1 3, <Y% coverage dose 4 or major safety concern 8
Modelers Engaged to Estimate Thresholds of Vaccine Coverage that Predict Impact Through PATH, engaged modellers from Swiss Tropical Institute and Imperial College, London Generating estimates for a range of vaccine coverage that will estimate impact on severe malaria, malaria mortality or cost effectiveness Modelling methods presented to the WHO Immunization and Vaccine-related Implementation Research Advisory Committee (IVIR-AC) March 2018 9
Modelers will Consider Two Scenarios for Vaccine Impact and Cost-effectiveness (CE) Estimates 1. Impact estimates for MVIP pilot areas: a. Estimates of impact and CE will be generated with parasite prevalence that correspond to those in the pilot areas b. Area-specific assumptions on vaccination coverage, costs, and coverage of malaria preventive/curative interventions based on publicly available data 2. Impact estimates for a range of malaria transmission settings where the RTS,S vaccine may be recommended/implemented should there be a policy recommendation: a. Estimates will be generated for parasite prevalence levels representative of those found in sub-saharan Africa (e.g. 10% to 65%) b. A common set of assumptions on vaccination coverage, costs, and coverage of malaria preventive and curative interventions will be applied to all transmission settings based on publically available data 10
Outcomes and Outcome Metrics to be Generated Outcomes: Severe malaria cases averted Severe hospitalized malaria averted Malaria deaths averted DALYs averted Outcome metrics: Events averted per 100,000 vaccinated Events averted per dose Events averted per 100,000 population 0-5 year olds; target age group Percent reduction in events Cost per event averted 11
Illustrative Example of Outputs: Events Averted by Malaria Transmission (not based on actual estimates) Figure 1: Events averted per 100,000 population for a single vaccine coverage scenario, across a range of transmission settings. This figure can be produced for specific population groups and vaccine coverage scenarios, and 95% credible intervals can be included. 12
Illustrative Example of Output: Events Averted by Dose 4 Coverage (not based on actual estimates) Figure 2: Events averted per 100,000 population for a single transmission setting, across a range of scenarios for coverage of the fourth vaccine dose. In this example, the coverage of the third dose is fixed, and the fourth dose coverage varies along the X-axis. This figure can be produced for specific population groups and transmission settings (for example in a series of plots for PfPR 2-10 = 10 40%) and different levels of coverage of the first three vaccine doses. 13
Illustrative Examples of Outputs: Cost Per Event Averted (not based on actual estimates) Figure 4: Cost per event averted for a range of transmission settings, for three vaccine coverage scenarios, where coverage of doses 1 3 and dose 4 are both varied. A range of different vaccine coverage assumptions can be included. 14
Timeline Timelines and Activities for Framework Activity 1Q-2Q 2018 Seeking input on the Framework for Policy Decision (Presented to the IVIR-AC in March 2018, PAG March 2018, SAGE/MPAC April 2018) 2Q-3Q 2018 Modelers will generate estimates for inclusion in the Framework for Policy Decision (Presentation to IVIR-AC September 2018), modelled estimates of criteria thresholds to be incorporated into the Framework Convene working group, including members from PAG and MPAC/SAGE, to deliberate on Framework Present the working group s report and recommendations on the Framework to PAG, SAGE and MPAC for discussion in fall 2018 or spring 2019 15
Questions for SAGE on the Framework for Policy Decision 1. Does SAGE agree with the approach? 2. Are the suggested outcomes and matrices useful for policy decision? 3. Does SAGE agree on the following suggested next steps? I. Additional SAGE and MPAC members join the PAG working group to consider and deliberate on the questions posed within the Framework (~2 from each?) II. III. The working group report back with those considerations and presents to PAG, MPAC and SAGE at future meeting, aiming for fall 2018 or spring 2019 Next step for Chairs of SAGE and MPAC to provide to the MVIP secretariat the names of those available to participate on such a working group 16
17 Thank you
18 Extra slides
Summary of IVIR-AC informal feedback Produce multiple outcome metrics for each outcome. Some examples include events averted per 100,000 vaccinated children, events averted per dose, events averted per 100,000 population (all ages or 0-5 year olds). Present impact results for 3 doses vs. no vaccination Present impact results for 4 doses vs. no vaccination, where 3 rd and 4 th dose coverage is the same. Present impact results for multiple levels of 4 th dose coverage for a given 3 rd dose vaccine coverage. This will assess the incremental impact of differing vaccine coverage for the 4 th dose. Once formal feedback from the IVIR-AC is received, it will be incorporated into the modeling plans as appropriate. 19
Vaccine coverage assumptions (Example) Scenario Dose 1 coverage 1 (10% drop-off between 1-3, 20% drop-off between 3 and 4) 2 (10% drop-off between 1-3, 10% drop-off between 3 and 4) 3 (5% drop-off between 1-3, 20% drop-off between 3 and 4) 4 (5% drop-off between 1-3, 10% drop-off between 3 and 4) 5 (5% drop-off between 1-3, 5% drop-off 20 between 3 and 4) Dose 2 dropout rate Dose 3 dropout rate Cumulative coverage doses 1-3 Dose 4 dropout rate Cumulative coverage doses 1-4 50% 10% 10% 41% 20% 32% 60% 10% 10% 49% 20% 39% 70% 10% 10% 57% 20% 45% 80% 10% 10% 65% 20% 52% 90% 10% 10% 73% 20% 58% 50% 10% 10% 41% 10% 36% 60% 10% 10% 49% 10% 44% 70% 10% 10% 57% 10% 51% 80% 10% 10% 65% 10% 58% 90% 10% 10% 73% 10% 66% 50% 5% 5% 45% 20% 36% 60% 5% 5% 54% 20% 43% 70% 5% 5% 63% 20% 51% 80% 5% 5% 72% 20% 58% 90% 5% 5% 81% 20% 65% 50% 5% 5% 45% 10% 41% 60% 5% 5% 54% 10% 49% 70% 5% 5% 63% 10% 57% 80% 5% 5% 72% 10% 65% 90% 5% 5% 81% 10% 73% 50% 5% 5% 45% 5% 43% 60% 5% 5% 54% 5% 51% 70% 5% 5% 63% 5% 60% 80% 5% 5% 72% 5% 69%
21
Why Develop a Framework for Policy Decision? Provides the opportunity for SAGE and MPAC members to refine ideas on the relative contribution of the collected data (feasibility, safety, impact) to a future policy recommendation Provide clarity on the expected use of the data in anticipation of potential changes in SAGE and MPAC membership between the time the SAGE/MPAC recommendations were made (2015) and the programme end (2022) Funders, potential funders, and manufacturers can refer to the framework for planning purposes Reducing the likelihood of gaps in funding or vaccine availability should the vaccine be recommended for broader use 22
Illustrative Examples of Outputs: Varied Coverage of Dose 3, 4 (not based on actual estimates) A. B. Total population Under 5s 23 Figure 3: Percentage reduction of event averted by coverage at 3 rd and 4 th doses for a given PfPr 2-10 by total population (A) and among 0-5 year olds (B), following 5 years of RTS,S implementation. This figure can be produced for multiple populations including number of children vaccinated.