Roll Back Malaria Progress & Impact Series
Mathematical Modelling to Support Malaria Control and Elimination
Mathematical Modelling to Support Malaria Control and Elimination, the fifth report of the Roll Back Malaria Progress & Impact Series, provides an overview of mathematical modelling, explains its history in relation to epidemiology and malaria, and details its implications and uses for global and national malaria control and elimination planning. The report aims to expand the dialogue within the global malaria community - and among public health decisionmakers in particular - on when and how mathematical modelling can help inform malaria control programmes and policies.
A summary of key messages
- Mathematical modelling has long been applied to malaria control and is particularly relevant today in light of rapid country progress in reaching high intervention coverage targets and given the intensifying global efforts to achieve the malaria-related Millennium Development Goals.
- Modelling is especially well-suited to helping inform decision-making around malaria control because of the disease's complex biological systems, the considerable infrastructural and cost requirements of prevention and elimination, and the rapid pace of change in global planning and national programming to halt malaria.
- Findings from modelling can provide valuable data to help inform a range of decision-making on issues such as identifying potential effective combinations of interventions; setting feasible coverage targets; devising target product profiles for new interventions; anticipating the implications of introducing new interventions; understanding the potential for malaria resurgence; and establishing how malaria surveillance can best enhance global or country progress.
- While models have limitations, their findings can be useful to the full range of constituents of the global malaria control community, including global and regional malaria control policy-makers, program planners at the country and local levels, academic researchers, those involved in product research and development, and donors.
Using modelling to inform decision-making
Mathematical modelling uses computer-based models to describe, explain, or predict behaviour or phenomena in the real world. It is particularly useful in investigating questions or testing ideas within complex systems. For this reason, modelling can be especially helpful in informing decision-making in global malaria control and eradication efforts because they involve extensive changes to a complex web of interconnected biological systems. Establishing optimal policies and programmes to support these efforts is complicated by the potential for parasites and vectors to evolve, the waxing and waning of human immunity, behavioural changes in human and vector populations, and interactions among large numbers of heterogeneous sub-populations of the organisms involved.
As countries scale up malaria control efforts and reach high intervention coverage targets, they are faced with the question of
what to do next. The strategy for maintaining and enhancing the achieved reductions in transmission is not obvious. It is often not
clear whether maintaining current coverage levels would continue to reduce transmission, stabilize transmission at a new level, or
slowly give way to an increase in transmission. Mathematical modelling can build on available data, test multiple scenarios and combinations of intervention strategies, and make verifiable predictions on what can be expected from these strategies.
Box 1 offers a practical example of how mathematical modelling can be applied in a country context.
An example of the role modelling can play in malaria control: Understanding the potential implications of combining ITNs and IRS
Several African countries have achieved high coverage of insecticide-treated nets (ITNs) and are now considering the potential benefit (in terms of reducing disease burden or interrupting transmission) of adding indoor residual spraying (IRS) as an additional means of vector control. Different mathematical models are required in order to produce outputs that would reasonably inform understanding about this issue. The overall models need to:
- accurately describe the malaria transmission cycle including malaria infections in humans and mosquitoes;
- account for the effects of malaria infection in humans on clinical disease, morbidity, and mortality;
- include the effects of the health systems on malaria transmission and disease;
- account for the effects of ITNs and IRS on the malaria transmission cycle;
- use available data in model inputs (such as pre-intervention transmission level, predominant vector species, population age structure, first line treatment drug for malaria) and outputs (such as incidence of infection, age-prevalence of parasitaemia, age-incidence of mortality) to estimate parameter values for the model;
- use additional data for model outputs to ensure that it can reproduce data that it has not been fit to;
- include a set coverage level of ITNs to see the corresponding disease burden;
- add various coverage levels of IRS to see the effects on reduction in disease burden and transmission;
- compare different insecticides to see what is most appropriate to the situation, especially when insecticide resistance is taken into account;
- use cost data to determine the cost-effectiveness of adding IRS.
Adding a model for the evolution of resistance would allow for testing resistance management strategies with the combination of ITNs and IRS.
To produce relevant, robust findings, mathematical modelling should at the outset involve partnership and good communication between technical experts in mathematical modelling, experts in malaria field and laboratory science, and health policy decision-makers. Models produce the most useful data when they are formulated with important biological, economic, and practical realities in mind and when their results are interpreted with care, making them another useful tool in the fight against malaria.