Authored by The 1Day Africa research team. This project was a collaboration by Ollie Sayeed, Isla Gibson, Hannah Greene, Iain Macleod Briongos, Karam Elabd, David Tellett, Ryan Duncombe, and Julia Murdza.
Introduction
Malaria continues to carry a large disease burden, with an estimated 249 million cases and 608,000 deaths worldwide as of 2022. Affordability is a crucial factor in the fight against malaria, as the disease is endemic primarily in low-income regions. The global priority should be to decrease the incidence and burden of this disease, with the long-term goal of eliminating malaria country by country and eventually achieving global eradication.
The WHO recommends several malaria prevention measures, but they make clear that none of these disease control interventions are sufficient to serve as a stand-alone intervention. Effective malaria control requires a multipronged approach, incorporating a variety of interventions tailored to the local environment, available infrastructure, and the specific needs of the population. Here, we conduct a review of studies evaluating the cost-effectiveness of several malaria prevention interventions, including vaccines, vector control, insecticide-treated bed nets, and chemoprevention.
This analysis aims to collect, compare, and discuss the existing cost-effectiveness research on malaria interventions. The goal of this systematic review is particularly timely given the recent appearance of the RTS,S and R21 vaccines in the anti-malaria toolkit. As both vaccines are being rolled out across Africa, policymakers will benefit from having a complete picture of the cost-effectiveness of the new vaccines compared to existing interventions.
Methods
Search strategies and selection criteria
We conducted a systematic review of malaria interventions along with existing cost-effectiveness estimates for each one. The scope of our investigation was wide: we accepted all studies relating to the malaria interventions listed below, with no restrictions on date of publication, language of publication, geographical context, or study design, so long as each study included an estimate of cost-effectiveness in terms of dollars per disability adjusted life year (DALY) averted.
To guide the search, we compiled a list of common malaria interventions. First, several kinds of intervention involve controlling mosquito populations rather than administering medicines or vaccines to people at risk of malaria. These include bed nets, both ordinary insecticide-treated nets (ITNs) and “long-lasting” insecticide-treated nets (LLINs) with insecticide incorporated into the net material; indoor residual spraying (IRS), coating indoor areas with insecticide; and larviciding, to kill mosquito larvae. Second, perennial malaria chemoprevention (PMC) and seasonal malaria chemoprevention (SMC) both involve giving antimalarial drugs to children at risk of malaria, either year-round or in seasons with high malaria burden. Finally, and of particular topical interest, the new RTS,S and R21 malaria vaccines form a new category of vaccine-based interventions.
Searches were conducted using the databases JSTOR, PubMed, PubMed Central, and Scopus. A schematic example search expression is given below: the same terms were used in all databases across all available search fields. All duplicates across databases were removed before further processing of the list of studies.
(“Cost-Effectiveness” OR “Cost-Benefit Analysis” OR “Economic Evaluation”) AND (“DALYs” OR “Disability-Adjusted Life Years”) AND (“Malaria” OR “Plasmodium”) AND (“Prevention” OR “Intervention” OR “Vaccine” OR “RTS,S” OR “R21” OR “Bed Nets” OR “Insecticide-Treated Nets” OR “LLINs” OR “Larviciding” OR “Indoor Residual Spraying” OR “IRS” OR “Chemoprevention” OR “IPTi” OR “SMC” OR “PMC”)
For logistical reasons, the search strategy sometimes had to be completed using multiple searches. First, JSTOR’s paper database imposes a limit on the maximum length of its search terms. We broke the full term into separate searches in such a way as to preserve the logical structure of the search, as searching for “A and (B or C)” is equivalent to searching for “A and B” and “A and C.” Second, in response to the WHO’s 2022 renaming of IPTi (intermittent preventive treatment for infants) as PMC (perennial malaria chemoprevention), the PMC part of the search was conducted later. Duplicates across searches within the same database were treated the same way as duplicates across databases.
For results that were themselves review papers citing data from other studies, we discarded the review papers and only included the original studies if they otherwise met the search criteria. All studies surveyed were written in English, and the relevant results were accessible to us without having to assess grey literature sources or contact the study authors for materials. Two records were identified outside of our search strategy through bibliographies. Both were sources that did not themselves appear in our search results but whose cost per DALY findings were cited by papers that did appear in our results.
Data analysis
From each study, we extracted a final cost-effectiveness estimate of the relevant intervention in terms of US dollars per DALY averted. For studies that reported multiple cost-effectiveness estimates from a single point in time – for example, from multiple sites – we calculated the median and interquartile range (IQR) of the estimates reported. For longitudinal studies that reported cost-effectiveness estimates over a span of multiple years, we applied the same process to only the most recent data reported.
We recorded whether each figure represented an estimate of provider cost – the cost directly incurred by the organization delivering the intervention – or societal cost – the total cost to society of delivering the intervention. Studies that reported both types of cost were included as separate data points. We recorded the costing year of each study and adjusted each figure for inflation based on the US Consumer Price Index, giving a final set of figures in January 2025 US dollars.
Each estimate was categorized into one of the types of malaria intervention listed above: ITNs, LLINs, larviciding, IRS, PMC, SMC, and the RTS,S and R21 vaccines. These, in turn, were grouped into larger categories for analytical purposes: ITNs and LLINs into a category of bednets, larviciding, and IRS as vector control, and PMC and SMC as chemoprevention. Within each intervention category and for each type of cost mentioned above, we calculated the median and interquartile range of the cost-effectiveness estimates across the relevant studies. This represents a point estimate and a corresponding range of uncertainty of the cost of saving lives from malaria with each intervention.
Results
The initial search using the terms described above yielded a total of 685 results across the four databases JSTOR, PubMed, PubMed Central, and Scopus, 325 of which were duplicates. The 360 remaining results were filtered for relevance to malaria cost-effectiveness and led to a final list of 28 studies containing at least one estimate of cost per DALY averted. Five of these included data on multiple interventions, giving 38 individual cost-effectiveness estimates in total. The majority of studies took place in sub-Saharan Africa, along with one study each in Pakistan, Bangladesh and Myanmar.
The data set, including inflation data and a full list of cost-effectiveness estimates from each study, can be found here. Each estimate is reported as a cost in 2025 US dollars per DALY averted. A negative cost per DALY averted, as seen in one study of bednets, represents an intervention that saved money for the healthcare provider. For each category of intervention, we calculated the median and interquartile range (IQR) of the cost-effectiveness estimates in the data set, separating out the R21 and RTS,S vaccines given their topical importance.
Combining provider costs and societal costs, the most cost-effective intervention category was chemoprevention, with a median cost-effectiveness of $17.66 per DALY averted and an IQR of $19.70. In second place, although only represented by one study, the R21 vaccine had a cost-effectiveness of $42.58 per DALY. Bednets, traditionally seen as the most cost-effective intervention, came in third with a median cost-effectiveness of $52.67 and IQR $83.30. The RTS,S vaccine and vector control had notably higher cost-effectiveness values, with a median of $251.01 (IQR $155.75) for RTS,S and $396.82 (IQR $959.20) for vector control.
Figure 1 shows a breakdown of cost-effectiveness estimates between provider costs and societal costs for each category of intervention in the data set. Individual points represent individual studies, and bars represent the median cost-effectiveness across studies within each category. Points are colour-coded according to whether the study reported provider costs or societal costs.
Figure 2 shows the same breakdown of cost-effectiveness between provider and societal costs, excluding the three outliers above $425 per DALY averted from the graph while still including them in the median calculations. We see that the two measures of cost are broadly correlated, with both measures showing that chemoprevention and nets are highly cost-effective and that the RTS,S vaccine and vector control are less cost-effective.
The very low societal cost-effectiveness of chemoprevention is represented by a single study of intermittent preventive treatment of malaria in pregnancy (IPTp), with an inflation-adjusted cost of $1.88 per DALY averted; but even aside from this outlier study, the twelve chemoprevention studies in the data set from a provider perspective lead us to a median cost-effectiveness estimate lower than for any other intervention category.
Discussion
The different measures of cost in the results above lead to similar conclusions. Chemoprevention is the most cost-effective category of interventions when measured by either provider costs or societal costs, surpassing the more famous intervention of distributing insecticide-treated bednets. Vector control and the RTS,S vaccine are less cost-effective, with high median costs per DALY averted under both measures of cost.
The R21 vaccine, which has been rolled out on a large scale over the past year and has been a focus of 1Day Africa’s malaria advocacy, is seen to be considerably more cost-effective than the earlier RTS,S vaccine. We suggest this effect is driven mostly by the difference in manufacturing costs between the two vaccines, with R21 manufactured by the Serum Institute of India, a global leader in high-volume, low-cost vaccine manufacturing. But our conclusions about R21 are based on a single study that only records provider costs: further work is needed on the cost-effectiveness of R21, including measurements of the societal costs of the vaccine.
There are several other uncertainties in our interpretation of these results. The studies in our systematic review covered different countries with different per capita incomes, logistical costs, and quality of public health infrastructure. Provider costs, making up the majority of cost estimates in the data set, fail to account for differences in other logistical costs borne by patients themselves. Societal costs, which account for costs borne by people other than the healthcare provider, will, in turn, vary in importance by the income of the country, as an income-unadjusted $1 cost is significantly less affordable to someone earning $1 a day than to someone earning $5 a day. The ongoing costs of a program also likely decrease over time compared to the initial costs of starting a program, so studies of established programs may not be representative of the cost-effectiveness of new programs and vice versa.
The studies in our data set took place in different years and calculated their numerical costs using different costing years, ranging from 1995 to 2023. All costs were reported in US dollars indexed to the relevant costing year, except for one study that reported costs in Ugandan shillings along with the then-current US dollar exchange rate. We adjusted each of these costs for inflation up to January 2025 using the US Consumer Price Index. Still, different countries have experienced different inflation rates since their costing years, and individual interventions may have undergone faster or slower changes in price than average inflation.
The populations targeted by each intervention likely vary in their average age, both because different countries vary in their age distributions and because different interventions vary in which age groups they target. Vaccines, chemoprevention, and bed nets, for example, tend to target children, but less individually targeted interventions like vector control benefit all age groups. Because saving the life of a child averts more DALYs than saving the life of an adult with fewer expected future life years, this complicates direct DALY comparisons between interventions aimed at different age groups.
Finally, the epidemiology of malaria itself will vary between countries in the data set, with higher or lower prevalence and varying seasonality. Interventions tend to be less cost-effective in contexts where background malaria prevalence is lower, given the higher number of treatments needed to save a life in expectation. The seasonality of malaria transmission can increase the cost-effectiveness of interventions like seasonal malaria chemoprevention, timed to correspond to peaks in the malaria calendar, which is less cost-effective in regions with perennial malaria transmission.
Conclusion
With the roll-out of the RTS,S and R21 vaccines, the world has more tools for fighting malaria than ever before. This systematic review incorporates the most recent data about R21, RTS,S and other malaria interventions to calculate a current best estimate of their cost-effectiveness. We find that chemoprevention stands out as the most cost-effective category of malaria intervention, saving one disability-adjusted year of life for an average of $17.66 spent.
We welcome questions and comments. Please reach out to malaria.cea@1daysooner.org.