Simulation of the costs and consequences of potential vaccines for Plasmodium falciparum malaria

Malaria is one of the major public health problems for low income countries, a major
global health priority, and it has also a dramatic economic impact. Funding for
malaria control is on the rise and both international donors and governments of
malaria endemic countries need tools and evidence to assess which are the best and
most efficient strategies to control malaria.
Standard tools traditionally used to assess the public health and economic impact of
malaria control interventions, such as efficacy trials and static cost-effectiveness
analyses, capture only short term effects. They fail to take into account long term and
dynamic effects due to the complex dynamic of malaria, and to the interactions
between intervention effectiveness and health systems.
This thesis is part of a wider research project, conducted by the Swiss Tropical
Institute, aimed at developing integrated mathematical models for predicting the
epidemiologic and economic effects of malaria control interventions. The thesis
specifically combines innovative mathematical models of malaria epidemiology with
innovative modeling of the health system and of the costs and effects of malaria
control interventions. These approaches are applied to simulate the epidemiological
impact and the cost-effectiveness of hypothetical malaria vaccines. Chapter 1 describes why malaria is a public health priority, the increasing relevance
of conducting economic analyses in the health sector, the economic evaluation
framework, and the economic consequences of malaria.
Chapter 2 presents an approach to dynamically modeling the case management of
malaria in Sub-Saharan Africa.
Chapter 3 describes an approach to costing the delivery of a hypothetical malaria
vaccine through the Expanded Programme on Immunization (EPI), on the basis of the
information available on the likely characteristics of the vaccine most advanced in
development. The results show that, although the vaccine price determines most of
the total delivery costs, other costs are relevant and should be taken into account
before planning its inclusion into the EPI.
Chapter 4 and 5 combine modeling of malaria transmission and control with
predictions of parasitologic and clinical outcomes, to assess the epidemiological
effects and the potential short and long term cost-effectiveness of a pre-erythrocytic
vaccine delivered via the EPI. The results suggest a significant impact on morbidity
and mortality for a range of assumptions about the vaccine characteristics, but only small effects on transmission intensities. They also suggest that at moderate to low
vaccine prices, a pre-erythrocytic vaccine providing partial protection, and delivered
via the EPI, may be a cost-effective intervention in countries where malaria is
endemic.
Chapter 6 simulates the cost-effectiveness of three different vaccine types: Preerythrocytic
vaccines (PEV), Blood stage vaccines (BSV), mosquito-stage
transmission-blocking vaccines (MSTBV), and combinations of these, each delivered
via a range of delivery modalities (EPI, EPI with booster, and mass vaccination
combined with EPI). The simulations presented in this Chapter show that PEV are
more effective and cost-effective in low transmission settings. In contrast to PEV,
BSV are predicted to be more effective and cost-effective at higher transmission
settings than low transmission. Combinations of BSV and PEV are predicted to be
more efficient than PEV, in particular in moderate to high transmission settings, but
compared to BSV, combinations are more cost-effective in mostly moderate to low
transmission settings. Combinations of MSTBV and PEV or PEV and BSV do not
increase the effectiveness or the cost-effectiveness compared to PEV and BSV alone
when delivered through the EPI. However, when applied with EPI and mass
vaccinations, combinations with MSTBV provide substantial incremental health
benefits at low incremental costs in all transmission settings. This highlights the
importance of developing other vaccine candidates as they have potential to facilitate
a PEV/BSV combination vaccine to be more beneficial. Chapter 6 simulations
indicate that the transmission setting and the vaccine delivery modality adopted are
important determinants of the cost-effectiveness of malaria vaccines. Alternative
vaccine delivery modalities to the EPI may sometimes, but not always, be more costeffective
than the EPI. In general, at moderate vaccine prices, most vaccines and
delivery modalities simulated are likely to present cost-effectiveness ratios, which
compare favorably with those of other malaria interventions. Chapter 7 discusses the implications of approaches and results presented in the thesis,
their limitations and potentials. The approach used in this research represents the first
attempt to develop dynamic models of malaria transmission and disease to evaluate
the cost-effectiveness of malaria control interventions. Combining advanced
stochastic simulation modeling of malaria epidemiology with health system dynamic
modeling is a crucial innovation proposed by the approaches presented in this thesis.
In fact, while it is well known that the interactions between malaria and health
systems take place under temporal and spatial heterogeneity, integration of health
system metrics in epidemiological modeling is rarely done. The cost-effectiveness analyses are based on an approach to model the health system characteristics of the
settings where a new intervention, such as a malaria vaccine, will be implemented,
The rationale of this approach rests on: a) the need to capture the long term health and
economic impact due to the interactions between malaria control interventions and
the health system - e.g. the impact on the health system of variations in transmission
intensity due to an intervention; b) the recognition that policy makers are more
interested in cost-effectiveness predictions that are specifically tailored to their health
system context rather than on a hypothetical one.
The approaches developed provide a platform that could be used to model the effects
of integrated strategies for malaria control. The increase in computer power available
makes possible simulating complex scenarios with several dimensions/variables in a
relatively short time. This, coupled with the increasing availability of information on
malaria endemic countries health systems, should be exploited to further modeling
health system dynamics, which is fundamental to assess integrated malaria control
strategies.
The models and the approaches presented could be applied to inform decisions at
several levels. Further applications might include simulating the epidemiology, the
costs and consequences of packages of interventions, allowing estimating both
effectiveness and (technical and allocative) efficiency. This would, thus, help policy
makers to determine which intervention or, most likely, which package of
interventions, might be most effective and efficient in a particular area. Additionally,
it would be possible to simulate the implications of coverage extension of malaria
control interventions, and of different strategies and service delivery modalities that can reach the poorest.
The approaches developed could also allow identification of areas where intensified
malaria control is the only feasible option, areas where malaria elimination is more
likely to be achieved, the incremental cost-effectiveness of proceeding to elimination
once a high level of control has been achieved, the optimal transmission levels at
which to change strategy, and, in principle, economies of scope and or synergies in
effectiveness and cost-effectiveness of new strategies. These are all research areas
that have been identified as fundamental in the research agenda to be set up following
the recent call for malaria elimination.