Health gains and financial risk protection afforded by public financing of selected malaria interventions in Ethiopia: an extended cost-effectiveness analysis

Malaria prevention and control has been prioritized over the past decade in many national health sector plans. As a result, remarkable progress was made worldwide in reducing incidence and mortality from malaria [1, 2]. Due to the expansion of effective strategies, between 2001 and 2013, malaria incidence has dropped by 30% [1, 2]. Despite such progress, malaria remains a major public health burden with a huge impact on the socio-economic development of many countries [1, 2]. Nearly one-half of the world population lives in malaria-endemic countries [3]. In 2016 alone, there were an estimated 216 million cases and 445,000 deaths attributable to malaria worldwide [4]. Sub-Saharan Africa accounts for 90% of both cases and deaths due to malaria [4]. Malaria control is unequally distributed across socioeconomic groups and the rates of insecticide- and drug-resistance are increasing. Further scale-up of cost-effective malaria interventions with sustainable financing mechanisms is therefore urgently needed [5].

Ethiopia has made notable progress towards malaria control [6, 7]. Nationally, the prevalence of malaria has declined from 5 to 3% over 2010–2015 [5, 8, 9]. During the same period, malaria-related deaths were reduced by 40% [5]. Scale-up of effective anti-malaria interventions at the primary health care level and improved community engagement were major contributing factors to this progress [10]. There is little evidence from Ethiopia about other factors that might have contributed to malaria decline (e.g. climate change, housing structures and urbanization). However, despite significant progress, much remains to be done in the fight against malaria in Ethiopia, where about 2.6 million cases and 5000 deaths were estimated for the year 2016 [4]. Additionally, the 2015 malaria indicator survey shows that only 40% of the population at risk correctly use insecticide-treated bed nets [9].

Malaria prevention and control are major priorities for Ethiopia’s health sector transformation plan (HSTP) [11]. The primary strategies include rolling out long-lasting insecticide-treated bed nets (LLIN) and insecticide residual spray (IRS) for at-risk population [10, 12]. Similarly, artemisinin-based combination therapy (ACT) is recommended as first-line treatment of uncomplicated malaria [10, 12]. Ethiopia has committed to end malaria by 2030 and adopted global malaria control and elimination strategies [12]. As the country moves towards elimination by 2030, tests that are more sensitive will be required to detect subclinical malaria infection to prevent disease transmission [13]. A malaria vaccine (i.e. RTS,S/AS01) could help curb the malaria burden. However, the efficacy of the vaccine is partial and presents rapid waning immunity [14, 15].

Malaria is endemic in many regions of Ethiopia with marked seasonal and geographic variation. Nearly 60% of the total population reside in high-risk areas [10, 12]. In addition to its public health impact, malaria imposes a large financial burden on households, consuming on average 7% of household income [16, 17]. Marginalized and economically vulnerable populations are also at a higher risk of acquiring malaria and of experiencing fatal consequences because of limited health care access and the inability to pay for it [1, 18, 19]. Malaria spending is estimated to cost Ethiopia about $200 million annually or 10% of its total health expenditure [20]. Hence, reducing malaria disease burden has the potential to improve socioeconomic development [21].

The recent attention to universal health coverage (UHC) has provided context to explore mechanisms that would expand access to malaria prevention and treatment services in Ethiopia [22]. This would also help address the high rate (33%) of out-of-pocket (OOP) payments [20]. Given that a quarter of the Ethiopian population lives below the national poverty line [23], OOP malaria treatment costs can be an important barrier to access effective treatment and in pushing households into impoverishment in Ethiopia. Accounting for non-health benefits is essential to reduce health inequalities and contribute to the objectives of UHC [22]. Financial risk protection (FRP) is an important policy objective and can improve access to all needed quality health services without financial hardship [24, 25].

In this paper, the aim is to estimate the potential health, FRP, and equity benefits of universal public finance of scaling up selected malaria prevention and treatment interventions in Ethiopia [26]. This will support policymakers in jointly considering health gains, FRP and equity benefits in resource allocation related decisions.

Using extended cost-effectiveness analysis (ECEA), we consider the costs and health impact of malaria interventions across population subgroups and estimate the FRP impact on households in Ethiopia [26]. Building on a recent ECEA of malaria vaccine [28], and using a static disease model, are quantified, across socioeconomic groups (i.e. income quintiles), for each of four malaria interventions (ACT, LLIN, IRS, and malaria vaccine): the number of malaria-related deaths and OOP expenditures averted; the corresponding household FRP provided; and the implementation costs. Furthermore, ECEA is also applied across malaria transmission intensities to account for geographic variation of malaria (see Additional file 1: Appendix Table S2).

In this paper, the health and financial benefits of UPF for malaria interventions were estimated across Ethiopian households at all income levels. Overall, all four interventions showed substantial benefits, with ACT and LLIN accounting for the larger shares of malaria-related deaths and CHE cases averted.

All the interventions showed a greater number of deaths averted among the poorest 40% of the population, averted similar OOP expenditures across all income groups, and relatively higher FRP benefits for the poorest 40%. Even if the poor had lower access for care and higher baseline malaria risk, for each of the intervention greater benefits would go toward the poor. This suggests that the malaria interventions analysed in this paper benefit the worse-off and poor populations in remote areas of Ethiopia, who suffer the disease risk at most. Given the relatively lower malaria burden, the four malaria interventions would avert fewer deaths annually, as compared to, other interventions addressing childhood diarrhoea and pneumonia for example [51, 52]. Rapid decline of malaria deaths in Ethiopia over the last two decades and a relatively lower prevalence were the main reasons [6]. Among the four interventions, LLIN and ACT were the two strategies with the highest impact on malaria mortality. In contrast, the malaria vaccine would prevent the smallest number of deaths averted (i.e. 38 per year) as compared to the other interventions. This is largely because the vaccine would be relatively less efficacious [14, 41]: only 2% of malaria-related child deaths would be prevented from the vaccine in this study.

Even though the rich had more access to health services and less malaria burden, the private OOP savings would be similar across all income quintiles. This might be due to the fact that the poor and rich are spending similar OOP expenditures for malaria care. In absolute terms, the gains in private OOP expenditures could be lower as compared to findings from other Ethiopian ECEAs [51‐53]. This might be due to less OOP payments for malaria care as compared to the other diseases. As for the FRP benefits, LLIN and ACT prevented a higher number of CHE cases, and for all interventions, the greatest number of CHE cases averted would occur in the poorest income quintile. In addition, the annual cost of implementing IRS at a 10% incremental coverage for the at-risk population was about $33 million, 2 times higher than that of the LLIN intervention. This corresponds to more than 16% of malaria-related health care spending in Ethiopia [20]. Lastly, though ACT, LLIN, IRS, and malaria vaccine are critical for malaria control and elimination, these interventions would need to be combined with other interventions, such as behavioural change, correct use and implementation, to yield full impact.

Nevertheless, the analysis presented here has several limitations. First, the disease model was static and did not address the dynamics of malaria transmission. Second, because of the unavailability of key input parameters by socioeconomic group, proxy input parameters were used. For example, the percentage who sought treatment for fever in the past 2 weeks was used as a proxy indicator for seeking malaria care. This might have overestimated malaria cases as there are other causes of fever among individuals (besides malaria). The Ethiopian 2016 DHS, the Malaria indicator survey and the ACT malaria consortium guidance on health equity analysis use health care utilisation due to fever in the past 2 weeks as a proxy for seeking care for malaria [9, 34, 35]. Third, due to the lack of disaggregated data, constant rates for case fatality ratio, intervention effectiveness, and inpatient cost inputs were assumed across quintiles. Fourth, unit costs for the vaccine were not specific to Ethiopia. However, despite the limitations, the analysis is crucial as the findings could assist policymakers decide on which health interventions to rollout to reduce malaria disease burden affecting 60% of the Ethiopian population [9].

The ECEA can also answer some of the equity concerns by providing valuable information on how malaria prevention or treatment strategies would decrease both malaria burden and financial risk incurred by households across various socioeconomic groups in Ethiopia. This study shows that malaria interventions could improve FRP across all income groups, especially among the bottom income groups in Ethiopia. Furthermore, this analysis can help reorienting malaria interventions to target elimination across selected segments of the population, especially among the poor.

All four malaria interventions would save more lives among the poor than among the rich. Preventing and treating malaria provides substantial health benefits and FRP, especially among poor Ethiopians. ACT and LLINs would generate the largest impact on malaria-related deaths averted and FRP benefits. Improving health equity and reducing poverty are major objectives of the Sustainable Development Goals, and the findings of the study presented here would provide insight for policymakers on how to prioritize malaria interventions for targeted population groups including the poorest.