Economic evaluation of disease elimination: an extension to the net benefits framework and application to human African trypanosomiasis ======================================================================================================================================= * Marina Antillon * Ching-I Huang * Kat S Rock * Fabrizio Tediosi ## Abstract The net benefits framework has become a mainstay of the cost-effectiveness literature, guiding decision-makers to select among strategies in the presence of budget constraints and imperfect information. However, disease elimination programs are socially desirable but not always cost-effective. Therefore, analytical frameworks are necessary to consider the additional premium for reaching global goals that are beyond the cost-effective use of country resources. We propose a modification to the net benefits framework to consider the implications of switching from an optimal strategy (in terms of cost-per-burden-averted) to a strategy with a higher likelihood of meeting the global target (i.e. elimination of transmission by a specified date). Our expanded framework informs decisions under uncertainty, determines the share of funding necessary to align local and global priorities, enabling local partners to use their resources efficiently while cooperating to meet global health targets. We illustrate the advantages of our framework by considering the economic case of efforts to eliminate transmission by 2030 of *gambiense* human African trypanosomiasis (gHAT), a vector-borne parasitic disease in West and Central Africa. **Significance Statement** Various diseases have now been earmarked for elimination by the global health community. While the health economic implications of elimination have been discussed before, one important topic remains unexplored: uncertainty and its consideration within extant cost-effectiveness frameworks. Here we extend the ubiquitous net benefits framework to consider the comparative efficiency of alternative elimination strategies when these strategies have different probabilities of reaching elimination. We evaluate the premium of elimination, and we apply our method to efforts against human African trypanosomiasis in three settings. This method could be directly applied to simulation-based studies of the cost-effectiveness of other disease elimination efforts, therefore giving the global health community a common metric by which to budget for such initiatives. * gHAT * elimination * economic evaluation * net benefits framework The successful eradication campaigns of smallpox and rinderpest have curried the political support for the elimination and eradication of a host of other diseases. Thus far, regional elimination has been achieved for land-transmitted rabies in Europe and for malaria in large parts of the world. Yet, due to the strong correlation between many of the diseases targeted for elimination and eradication and poor sanitation, health infrastructure, or overall material conditions, such disease campaigns are disproportionately focused in low-resource settings, bringing to the fore important questions about the economic efficiency of such efforts (1, 2). On the one hand, expensive interventions are often justified on the basis of future cessation of activities; one front-loads the expenses on a disease on the premise that public health activities can cease in the not-too-distant future, at which time one may recover investments. Smallpox eradication is claimed to have saved, within just a few years, billions of dollars (3). However, falling just short of elimination or eradication could be the worst of all possible scenarios: one has diverted ever-increasing resources from other purposes, but one has not reached a situation where investment can be recovered. In part because of the risk of failure and in part because of the accelerating per-case costs of elimination and eradication campaigns near the end-game, such efforts might not be considered an efficient use of resources in a typical 10- or 20-year time horizon from the perspective of decision-makers with limited budgets contending with a variety of health challenges. For instance, there were 22-180 wild-type poliomyelitis cases reported annually in recent years but eradication campaigns cost approximately $1B due to the continued need for surveillance and vaccination globally (4, 5). Guinea worm disease (GWD), which has a more restricted geographic reach than polio, is the subject of campaigns that cost approximately $30M when the disease causes only 54 cases yearly (6, 7). The eradication targets of both diseases have experienced delays 23 years and 10 years, respectively, therefore stalling the promise to recover investments (7, 8). To explore, scrutinize, and provide insights into questions surrounding disease elimination and eradication, the epidemiology and health economics fields have a rich toolbox that captures the non-linear transmission dynamics, temporal features, and economic implications of disease control. However, that toolbox presents shortcomings in the face of elimination and eradication: the diseases targeted for elimination and eradication are in hard-to-reach populations usually in low socio-economic strata, but traditional cost-effectiveness is agnostic to concerns of equity, which a few studies have just begun to address (9–13). Separately, a few studies have tried to grapple with questions around the economic implications of disease elimination and eradication by employing game theoretical approaches, conceptualising investment cases for disease eradication with multiple stakeholders (3, 14, 15). For policy-makers and advocates, it may be useful to know the portion of the resources aimed at averting morbidity and mortality versus the additional costs of strategies that promise elimination or eradication. A previously described approach has been to designate different levels of coverage (i.e. vaccine coverage) necessary for “control” and “elimination/eradication”, and the difference in costs between “control” and “elimination/eradication” strategies constitutes the price for elimination (3, 14, 16). While making the problem theoretically tractable, the application is quite narrow: addressing the burden of a disease may require various distinct activities in practice, and parsing the activities that contribute to “control” or “elimination/eradication” is not possible, as each activity contributes to both goals to varying degrees. Moreover, there is a lot of uncertainty in whether strategies would lead to elimination or eradication. In practice one cannot purchase elimination or eradication and then be certain it will happen, as GWD and polio have shown, one merely invests in activities that are conducive to eradication, and therefore, the absence of probabilistic thinking in previous literature fails to capture a key component of the decision-making process. Here we develop a framework that can handle strategies that have different probabilities of elimination, rather than strategies that are conceived as certain to reach elimination or not, as well as strategies where activities are not easily classified as “control” or “elimination” activities. Disease campaigns with eradication objectives can be examined analogously. Specifically, we extend the net-benefits framework, useful for decision-analysis in the presence of uncertainty, in order to simultaneously evaluate cost-effectiveness of public health strategies while explicitly outlining the ‘premium’ of elimination, or the additional resources that are necessary to bring a countries activities in line with global goals. One important feature of our framework is that it is operationalized within a probabilistic simulation framework, which makes it a simple extension to ubiquitously used approaches for decision analysis in the face of uncertainty, thereby being applicable in a wide array of situations. We then apply our new framework to *gambiense* human African trypanosomiasis (gHAT) in three distinct regions of the Democratic Republic of Congo (DRC). The three regions highlight the strengths of our framework and its applications under different circumstances: circumstances of certainty, uncertainty, and where more than two strategies are reasonable candidates to abate disease burden and interrupt disease transmission. ## Economic evaluation framework Our objective is to introduce an extension of the net benefits framework to account for the resource implications of aligning local targets with global goals like elimination. A detailed overview of the relationship between the incremental cost-effectiveness ratio (ICER) and the net benefits framework is found in SI section 1, but we provide a brief summary here as a point of departure for our framework. The keystone metric of value-for-money in cost-effectiveness analysis is ICER, defined as the ratio of the difference in costs, Δ*C*, and the difference in health effects, Δ*E* of two interventions: ![Formula][1] where the change in costs and health effects are computed as the net difference in strategies. Effects are usually denominated in disability-adjusted life-years (DALYs) averted, a metric that is comparable across diseases. For the purpose of our analysis, the effects will be distinguished between DALYs and the probability of elimination of transmission (EOT). Most often, even when one is simply concerned with DALYs and costs, uncertainty in these metrics exists, in particular with regards to populations that are difficult to study or diseases that have been historically neglected. There exists a literature devoted to the difficulties of accounting for parameter uncertainty and ICERs while remaining consistent with the economic principles on which cost-effectiveness is grounded (17, 18). The net benefits framework was therefore developed to circumvent some of the issues surrounding ICERs and uncertainty. The NMB is characterised as a simple arithmetic rearrangement of the ICER, provided that the threshold for cost-effectiveness, *λ*WTP, is explicitly stated: ![Formula][2] The linear additive formulation avoids the mathematical issues with samples of ratios. In a Monte Carlo sample of *N* iterates of the disease and the cost model, the preferred strategy is the one with the highest expected NMB, E(NMB), over all *N* samples. The algorithm therefore presents a measure of certainty that the strategy with the highest expected NMB is optimal over all other strategies: the probability that the strategy is optimal at a *λ*WTP value is given by the proportion of samples where the strategy has a higher NMB (18, 19). The NMB is conditional on the threshold of cost-effectiveness (*λ*WTP), or the willingness-to-pay (WTP) per DALY averted (18, 20). A cost-effective intervention should have a WTP that is equal or smaller to the the least efficient strategy in a decision-maker’s health portfolio (21–24). Recently the World Health Organization (WHO) has advocated that cost-effectiveness results should be shown at a range of WTP values (Table 1). View this table: [Table 1.](http://medrxiv.org/content/early/2021/02/16/2021.02.10.20181974/T1) Table 1. **Contextualizing willingness-to-pay values ![Graphic][3] for low-income settings.** ## Premium of elimination A useful metric easily calculated from the formulation of NMB is the PremiumEOT. We begin our exploration of the premium of elimination with a simple context (almost never seen in practice). If we suppose there are two strategies, one which is certain to reach elimination and a strategy that is certain not to reach elimination, and we suppose that the elimination strategy would not avert any additional DALYs over the non-elimination strategy (suppose that detection and treatment are superb), then the expected PremiumEOT would equal the expected cost difference between the two strategies ![Formula][4] However, in practice any two or more strategies are unlikely to avert the same number of DALYs; in fact, most often the elimination-prone strategy is likely to avert some additional DALYs, albeit at a potentially high cost. The health planner has a certain WTP for those additional DALYs that the elimination strategy averts, though perhaps not a WTP that would completely bridge the gap in costs between the two strategies. To convince that health planner to undertake the costlier, elimination strategy, the overall expected PremiumEOT must be: ![Formula][5] We argue that this is a sensible formulation of PremiumEOT because the more the health planner is willing to pay for DALYs averted, the lower the degree to which the health planner must justify costlier elimination-prone strategies on the grounds of elimination alone. If one strategy has both a higher probability of achieving the elimination and a relatively low incremental cost at reasonable time-horizon, then the premium of elimination is defined to be zero, as no additional resources are needed above those traditionally considered cost-effective to avert disease in order to justify elimination. We now turn our attention, and the plurality of the results and the discussion in the current analysis, to contexts in which the more expensive strategy bolsters the probability of elimination but does not necessarily guarantee it. 𝔼 (PremiumEOT) factors in the risk of switching to a more expensive strategy that does not necessarily achieve the EOT goal. We therefore extend our framework to determine whether a costlier strategy is sufficiently efficient at raising the probability of elimination to justify the additional cost by a decision-maker. ### Efficiency of elimination in the presence of uncertainty We propose the following re-formulation, which makes explicit the relationship between the WTP for averting disease burden denominated in DALYs averted, while taking into account the uncertainty in costs, intervention impact on morbidity and mortality, and elimination. We consider a two-strategy scenario and we present how it generalises to the multiple-strategy scenario in the next subsection. Generally, a second strategy will be favoured (cost-effective) compared to the first (comparator) if the additional costs are less than the product of reduction in burden (DALYs averted) and the WTP. Given any parameter set of a disease and cost model, *θ**i*, a strategy is cost-effective if ![Formula][6] where *i ∈* {1, …, *N*}. In this paper, we call the term ![Graphic][7] the justifiable costs (on the grounds of averting disease burden). Our proposal is that if there is a premium that a stake-holder is willing to pay for elimination, above and beyond the justifiable costs, then we include that justifiable premium for elimination as an additive term: ![Formula][8] where the indicator function, 𝕀EOT is 1 if EOT is achieved by only one strategy and 0 if neither or both strategies achieve EOT (for parameter set *θ**i* and strategy *j*): ![Formula][9] The use of ![Graphic][10] incorporates the WTP to raise the probability of elimination. In simulations where strategy A reaches elimination and strategy B does not, the NMB is lower, favouring strategy A. A further discussion of the term representing the justifiable costs of EOT is found in SI section 2. The term ![Graphic][11] is interpreted as the highest price paid to avert an additional DALY, whereas we interpret the new ![Graphic][12] as how much a decision-maker would pay per additional percentage point increase in the probability of elimination. ![Formula][13] The linear additive scale allows us to decompose the resource expenditure into two portions: the portion that is justifiable based on disease burden averted and the portion of the expenditure that is justifiable by concerns about global goals. While elimination and disease control are not separable or independent, which is accounted for in the dynamic transmission model, the linear form in this formulations allows us to separate the elimination benefits related to disease control (decreasing DALYs) and all other benefits of elimination above and beyond averted disease burden. We can then re-arrange these terms to express that a strategy should be adopted if the ![Graphic][14] is greater than the the difference between the costs and the costs justifiable for averting disease burden, analogous to our decision rule in the conventional context of optimal disease control: ![Formula][15] In other words, if the comparator strategy (strategy A) is the preferred strategy under the traditional NMB framework, then in order to select strategy B on the basis of elimination, the WTP for ‘efficient’ elimination must be at least as large as the PremiumEOT, provided that (using parameter set *θ**i*) strategy A does not lead to elimination, but strategy B does (i.e. ![Graphic][16] and ![Graphic][17]). In order to compute the probability that a strategy is cost effective across all parameter sets, rather than a single one we marginalise over all iterations of the model: ![Formula][18] where ![Formula][19] This formulation is directly analogous to the traditional construction of a cost-effectiveness acceptability curve (see SI section 2). Therefore, the probability that strategy *j* is cost-effective at ![Graphic][20] and ![Graphic][21] is the proportion of times in a simulation of N iterates where the strategy *j* has the highest net monetary and elimination benefits among all strategies. After marginalising over all *i* (taking the mean net benefit), then the ![Graphic][22] is the additional resources a health planner would be willing to put toward raising the probability of elimination by one percentage point. Generally, the strategy that ought to be implemented is indicated by: ![Formula][23] ### Scenarios with multiple potential new strategies The reformulation of the net benefits framework easily generalises to the multiple-strategy scenario due to the use of linear additive scale. However, instead of taking a net benefit greater than zero (which is to be interpreted as greater than the comparator intervention) one takes the maximum net benefit among all possible (*J*) alternative strategies, denoted below as *j* ∈ {2 : *J*}. This generalisation follows a similar form as in the conventional case where elimination considerations are absent (see SI section 2): ![Formula][24] where ![Formula][25] and the optimal strategy is indicated by: ![Formula][26] over all *N* samples of *θ* where ![Formula][27] and the differences refer to the difference between the comparator strategy (*j* = 1) and any alternative strategy *j* ∈ 2 : *J* in the analysis. As a preferred strategy changes with higher ![Graphic][28] values, making EOT justifiable by DALYs averted, the premium of elimination shrinks. An alternative derivation of NMEB, paralleling the derivation of the NMB from the ICER, is found in SI section 2. ## Results ### Health Outcomes, Costs, and Traditional ICERs Using a joint transmission and cost model we made projections of the epidemiological impact the resource use over 2020–2040 of four strategies in three locations against gHAT (Table 2 provides an overview on the component interventions, while further details can be found in the Materials and Methods). In Region 1, success or failure of the 2030 EOT goal is certain depending on the selected strategy, but in Regions 2 and 3 success and failure of the EOT goal is uncertain (Table 3). View this table: [Table 2.](http://medrxiv.org/content/early/2021/02/16/2021.02.10.20181974/T2) Table 2. **Strategies for control and elimination of gHAT in a typical endemic health district.** View this table: [Table 3.](http://medrxiv.org/content/early/2021/02/16/2021.02.10.20181974/T3) Table 3. **Intermediate outcomes between 2020–2040, cost-effectiveness, and elimination of transmission in three example regions.** If the status (comparator) strategy remains in place (Mean active screening, AS) there will be an average of 477 cases and 215 deaths in Region 1, 23 cases and 13 deaths in Region 2, and 47 cases and 24 deaths in Region 3. In terms of DALYs, those incidence levels give rise to 3,934 DALYs in Region 1,274 DALYs in Region 2, and 501 DALYs in Region 3. Under any strategy in all settings, the burden of disease is expected to decline, but strategies with VC are expected to expedite this decline substantially (see Fig S3). The uncertainty of costs for each strategy overshadows the differences in costs between strategies. Costs per year show that while strategies that include vector control (VC) are more costly in the short-run, the investments begin to yield returns after 2028 in Region 1, after 2025 in Region 2, and after 2031 in Region 3 (Fig S4). Nevertheless, a timehorizon of 20 years is not certain to show cost-savings by these strategies, although total costs are only marginally higher compared to the comparator (Mean AS). Costs are driven by AS activities, and when applicable, by VC activities, so the timing of cessation of these activities plays an important role in the ability of ambitious investments to be recovered (see Fig 5). For all three regions, the current strategy is expected to be the minimum-cost strategy even after accounting for uncertainty, and while it is certain that Mean AS will not deliver EOT in Region 1, it has a 79% probability of reaching EOT in Region 2 and a 46% probability of reaching EOT in Region 3. In Regions 1 and 2, the strategy that yields >99% probability of EOT involves adding VC to the status quo (Mean AS & VC), which is cost-effective if ![Graphic][29] in Region 1 and if ![Graphic][30], 651 in Region 2. In Region 3, Mean AS & VC is weakly dominated (see SI section 7) but a strategy that is certain to reach elimination (Max AS & VC) is justifiable at a very high willingness-to-pay ![Graphic][31]. ### Net Monetary Elimination Benefits: Region 1, where success and failure are certain The probability of EOT in Region 1 is shown in 1A and the results of our decision analysis under the traditional net benefits framework is shown in Figure 1B. After taking into account parameter uncertainty, our analysis shows that Mean AS has an 80% probability of having the minimum cost (optimal at ![Graphic][32], and optimal at ![Graphic][33] However, if ![Graphic][34], the strategy Mean AS & VC is optimal with 47%-55% probability for values of ![Graphic][35] compatible with a setting like DRC. ![Fig. 1.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2021/02/16/2021.02.10.20181974/F1.medium.gif) [Fig. 1.](http://medrxiv.org/content/early/2021/02/16/2021.02.10.20181974/F1) Fig. 1. Cost-effectiveness acceptability curves and cost breakdown for Region 1. The expected PremiumEOT, is shown in 1C. In a policy environment of low ![Graphic][36], any health planner must be able to justify the entire $800,000 cost difference on the basis of EOT alone. When health benefits are not valued monetarily ![Graphic][37], the PremiumEOT is simply equivalent to the difference in costs between the two strategies. If, for instance, ![Graphic][38], a health planner must be able to justify only a PremiumEOT of $524,900, as the other $275,100 that would be justifiable for DALYs averted. In a policy environment of a ![Graphic][39] per DALY averted, the strategy that reaches elimination is entirely justifiable on the health gains achieved (DALYs averted), and the PremiumEOT is therefore $0. Fig 1C shows the optimal choice of strategy for a range of ![Graphic][40] and ![Graphic][41] values. In a policy environment where ![Graphic][42] and ![Graphic][43] per probability point of EOT, the optimal strategy guarantees elimination, as that is the ![Graphic][44] that justifies the $800,000 premium of elimination. ### Net Monetary Elimination Benefits: where success and failure are uncertain Region 2 illustrates a setting in where the comparator strategy (Mean AS) has a 79% probability of EOT, and therefore the binary conception of “control” or “elimination” strategies fails to adequately capture the decision-maker’s dilemma. The conditions for which an economically optimal strategy maximizes the probability of EOT is shown in Fig 2. If ![Graphic][45], a strategy that almost ensures EOT, Mean AS & VC, is optimal if ![Graphic][46] per percentage point, and the PremiumEOT is $234,000 ($0–939,000) (Tables S14). If, however, ![Graphic][47], then the PremiumEOT is $163,000($0–893,000) and the ![Graphic][48]. At ![Graphic][49], then PremiumEOT decreases to $21,000 ($0–822,000) and ![Graphic][50]. Our most complex setting, Region 3, is shown in Figure 3. Under the traditional net benefits framework, either the Mean AS or Max AS strategy are cost-effective at ![Graphic][51] values consistent with historical investment levels in low-income countries (Fig 3C), but these strategies have only a 46% and 55% probability of EOT (inset Fig 3B). Without an investment in elimination justified by benefits that extend beyond the averted DALYs, achievement of EOT is uncertain. ![Fig. 2.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2021/02/16/2021.02.10.20181974/F2.medium.gif) [Fig. 2.](http://medrxiv.org/content/early/2021/02/16/2021.02.10.20181974/F2) Fig. 2. Cost-effectiveness acceptability heatmaps for Region 2. On the left is the cost-effectiveness acceptability heatmap (CEAH). Along the x-axis is the cost-effectiveness threshold for averting disease burden, ![Graphic][52], and along y-axis is the cost-effectiveness threshold for elimination of transmission (EOT), ![Graphic][53], to raise the probability of EOT by 2030 by one percentage point. On the right are the more traditional cost-effectiveness acceptability frontiers (CEAFs) assuming ![Graphic][54] of $0 and $10,000. The inset in the top-right graph is the probability of each strategy’s EOT by 2030. ![Fig. 3.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2021/02/16/2021.02.10.20181974/F3.medium.gif) [Fig. 3.](http://medrxiv.org/content/early/2021/02/16/2021.02.10.20181974/F3) Fig. 3. Cost-effectiveness acceptability heatmaps for Region 3. On the left is the cost-effectiveness acceptability heatmap (CEAH). Along the x-axis is the cost-effectiveness threshold for averting disease burden, ![Graphic][55], and along y-axis is the cost-effectiveness threshold for elimination of transmission (EOT), ![Graphic][56], to raise the probability of EOT by 2030 by one percentage point. On the right are the more traditional cost-effectiveness acceptability frontiers (CEAFs). The inset in the top-right graph is the probability of each strategy’s EOT by 2030. At ![Graphic][57], then a decision-maker must be willing to have ![Graphic][58] to bolster the chances of elimination from 46% to 55% or ![Graphic][59] to bolster the probability of elimination from 45% to >99%, representing a PremiumEOT of$691,000 ($0–1,594,000). If, however, the policy environment is one where ![Graphic][60], then the PremiumEOT is $22,000 ($0–259,000) and the ![Graphic][61] to opt for the Max AS strategy, which has a 55% probability of EOT. To maximise the probability of EOT when ![Graphic][62], one would deploy Max AS & VC, for which the PremiumEOT is $586,000 ($0–1,517,000), equivalent to ![Graphic][63]. At ![Graphic][64], the strategy Max AS is cost-effective, so the PremiumEOT is $418,000 ($0–1,407,000) and the ![Graphic][65] to deploy Max AS & VC. ## Discussion Much of the early literature about cost-effectiveness and decision-making in the presence of uncertainty was about analytical methods to do what is now more easily done with simulation (17, 20, 31, 32). Although simulation of a wide array of epidemiological and policy questions is ubiquitous, the interpretation of those simulations must rest within frameworks which decompose uncertainty and frame economic optimization in a manner that is in line with economic theory. The existing tools in the health economics toolbox generally answer the question “are public health efforts economically justified?” but when the global health community has set ambitious goals, such as elimination of infectious diseases, questions of resource use across different administrative levels (local, national, regional, and global) do not fit neatly in existing frameworks and one asks “to what degree are public health objectives economically justified by various objectives”. In the context of infectious disease elimination, there is a tension between reaching goals set by the global health community and maintaining an efficient use of scarce resources within each setting (10). This is particularly salient because the same resources – in-kind, capital, and financial – that vie for elimination efforts could be diverted to address other urgent health goals. We have presented an extension to the net benefits framework to inform decisions that contain an elimination objective that may stand at odds with concerns about efficient resource allocation. Our proposed framework explicitly models the additional premium for elimination activities in the presence of parameter uncertainty. The illustrative analysis shows that in Region 1, EOT is nearly impossible with the comparator strategy, but elimination is cost-effective at a relatively low ![Graphic][66], then yielding ![Graphic][67]. The other regions present more complicated scenarios: EOT is likely in Region 2 (79%) and moderately likely in Region 3 (46%) even with the comparator strategy (Mean AS), but the value-for-money in terms of EOT objectives vary by location. Although the PremiumEOT is highest in Region 1, it is the place where investments justified on the ground of elimination alone are the most efficient, ![Graphic][68]. In Region 2, the PremiumEOT=$210,000 – cheaper than in Region 1 – but on a per-percentage point basis the cost is $10,916 (see Tables S13 and S14). In Region 3, one would pay $6,489 per percentage point of EOT to bolster the chances of elimination from 46% to 55% for a PremiumEOT of $62,000 ($0–269,000), but to maximize the probability of EOT, each percentage point has an incremental cost of $14,113, equivalent to a PremiumEOT of $691,000 ($83,000-1,594,000). While one could justify strategies that maximize the probability of elimination with a sufficiently high ![Graphic][69] in Regions 1 and 2, that would not be possible in Region 3. In Region 3, one could not justify switching to one of the strategies that would ensure EOT (strategies with VC included) purely on the grounds of averting disease burden for the range of reasonable ![Graphic][70] thresholds considered here (see Fig 4 and Table S15). ![Fig. 4.](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2021/02/16/2021.02.10.20181974/F4.medium.gif) [Fig. 4.](http://medrxiv.org/content/early/2021/02/16/2021.02.10.20181974/F4) Fig. 4. Premium of elimination in Region 3, across different values of ![Graphic][71], contextualized in table 1. Existing literature has illustrative analysis of strategies against diseases of person-to-person transmission, so that control vs elimination is a matter of degree of coverage of vaccination (as with smallpox) or treatment (as with HIV in some countries). In the case of gHAT, as well as other vector-borne infections including malaria, multiple modalities of control exist and it is not clear that one activity alone is the key to elimination. Our framework expands the categories of diseases that could be analyzed via a common set of metrics amenable to simulation analysis: ![Graphic][72], and PremiumEOT. In presenting such a flexible and inclusive framework, it is tractable to perform analyses that could inform elimination policy priorities across disease portfolios. With this method, we can attest whether health-budget efficiency alone is enough to justify elimination efforts or whether some of the efforts are being justified on grounds of additional (non-health) benefits or externalities to other populations. Because it would be difficult, if not impossible, to enumerate all non-health benefits or externalities and integrate them into one CEA, one can, at minimum, present a lower bound for the utility that would have to be conferred on other sectors or populations to consider a strategy efficient. This could motivate advocating for transfers across government sectors or across health budgets in different jurisdictions, a theme that has only been recently considered in the literature (15, 33). ## Limitations The total costs of elimination will inevitably be affected by the size of the population that must be treated, and the size of the relevant population to attain a durable elimination gain involves the concepts of critical community size, meta-population, and importation probability that we do not examine here. Some diseases are worth eliminating in small patches because even one imported case will not re-establish transmission, while other diseases are only worth eliminating if you can eliminate entirely transmission in very large interconnected networks of settlements (34–36). Unlike many neglected tropical diseases, gHAT interventions have been very heterogeneous, even across the same administrative district, and so two regions with the same transmission in 2017 may have quite different underlying epidemiology. We captured this uncertainty by utilising posterior parameters from various regions, however the present results are not designed to be representative of a single area. Tailored models, fitted to longitudinal case and intervention data will yield more specific analyses for specific gHAT strategies. It is worth noting that the premium of elimination is not a subsidy, although this number could be used to inform a subsidy. Incurring outlays in the short-term may require financing products even if a setting is expected to save money in the long run. One form of short-term financing might be a subsidy, and this subsidy might go above and beyond the premium if the country cannot secure financial inflows at the right time, for instance in the form of loans. While we do not address issues surrounding the elimination of diseases that mainly afflict the extreme poor, there is a budding literature regarding such concerns termed ‘equityenhanced CEA’, ‘extended CEA’, and ‘distributional CEA’ which look at how elimination and control can improve not only health but correct past inequities (9, 10, 12, 37). ## Materials and Methods ### An application: human African trypanosomiasis *Gambiense* human African trypanosomiasis (gHAT) is a parasitic infection caused by *Trypanosoma brucei gambiense* and transmitted by tsetse (biting flies). gHAT infections are almost always fatal if untreated, and at the peak of the epidemic at the turn of this century it is suspected that up to tens of thousands of cases went undetected and untreated (38, 39). In 2012, the World Health Organization marked gHAT elimination of transmission by 2030 (40). While gHAT has historically burdened 24 countries, the Democratic Republic of Congo (DRC) remains the most affected, with over 74% of the worldwide caseload (41). Here we employed a previously published model of gHAT transmission fitted to historic data from three health zones in DRC: Kwamouth, in Mai Ndombe province; Mosango, in Kwilu province; Yumbi, in Mai Ndombe province. Details about these health zones are in SI section 3 and in a previous publication (42). Previously published models are based on epidemiological data provided by the WHO atlas of HAT (43). We selected these locations as they provide interesting illustrative examples of the NMEB framework. ### Health effects, costs, and cost-effectiveness All modelling choices are described in previous publications (42, 44, 45) and summarised in the supplement (SI section 3). The model provided projections under alternative strategies for 2020–2040 of future case-reporting as well as unobservable features such as transmission events, disease burden, and unreported deaths (44). Four strategies made up of combinations of interventions are shown in Table 2 and illustrated in SI Fig S1. We then applied a model of the resource use for these strategies (45) to estimate the costs and health burden accrued and averted in terms of cases, deaths, and DALYs. Costs were denominated in 2018 US$. Both costs and health effects are discounted at a rate of 3% in accordance with standard practice (26) and we performed our main analysis from the perspective of the healthcare providers collectively over a 20-year time-horizon (2020-2040). Uncertainty was accounted for in two ways: 1) uncertainty in all model parameters was propagated via Monte Carlo simulation, drawing 10,000 random samples from probability distributions chosen to characterise the extant uncertainty in each parameter in accordance with established practice (19), and 2) the model-simulated for stochasticity in case detections. Our initial outcome metric to evaluate cost-effectiveness is the ICER. Because we are concerned with cost-effectiveness and uncertainty, we construct cost-effectiveness acceptability frontiers(CEAFs), which denote the optimal strategy (in terms of cost-effectiveness) at a range of willingness to pay values (20). Lastly, we develop the cost-effectiveness acceptability heatmaps (CEAHs), a form of two-way CEAF with both ![Graphic][73] and ![Graphic][74]. We use no pre-defined thresholds for WTP values, as we aim to provide guidance rather than prescription. ## Supporting information Supplemental Text [[supplements/181974_file02.pdf]](pending:yes) ## Data Availability All data and code to reproduce the analysis contained in this paper are available through OpenScienceFramework (OSF). [https://osf.io/9375x/](https://osf.io/9375x/) ## Acknowledgements We wish to thank Ron Crump for original epidemiological model fit to data (42), and Dr. Erick Miaka of the Programme National de Lutte contre la Trypanosomiase Humaine Africaine (PNLTHA) in DRC and Dr. Jose Ramon Franco at the World Health Organization (WHO) for providing the data (43) which was used for original model fitting (42) and ongoing discussions on gHAT with our group. This work was funded by the Bill and Melinda Gates Foundation (BMGF): Human African Trypanosomiasis Modelling and Economic Predictions for Policy (HAT MEPP) project (OPP1177824) and the NTD Modelling Consortium (OPP1184344, OPP1156227, and OPP1186851). ## Footnotes * M.A. performed the formal analysis and created visualisations. M.A., K.S.R. and F.T. conceptualized the study and wrote the original draft. K.S.R. and F.T. acquired funding and supervised the study. C.H. performed simulations and reviewed and edited the final draft. * The authors declare no conflict of interest. * Received February 10, 2021. * Revision received February 10, 2021. * Accepted February 16, 2021. * © 2021, Posted by Cold Spring Harbor Laboratory This pre-print is available under a Creative Commons License (Attribution 4.0 International), CC BY 4.0, as described at [http://creativecommons.org/licenses/by/4.0/](http://creativecommons.org/licenses/by/4.0/) ## References 1. 1. T Fürst, et al., Global health policy and neglected tropical diseases: Then, now, and in the years to come. PLOS Neglected Trop. Dis. 11, e0005759 (2017). 2. 2. M Bangert, DH Molyneux, SW Lindsay, C Fitzpatrick, D Engels, The cross-cutting contribution of the end of neglected tropical diseases to the sustainable development goals. Infect. Dis. Poverty 6, 1–20 (2017). 3. 3. S Barrett, Economic considerations for the eradication endgame. Philos. Transactions Royal Soc. B: Biol. Sci. 368, 20120149–20120149 (2013). 4. 4. JS Lickness, et al., Surveillance to Track Progress Toward Polio Eradication — Worldwide, 2018–2019. MMWR. Morb. Mortal. Wkly. Rep. 69, 623–629 (2020). 5. 5. M Zimmermann, B Hagedorn, H Lyons, Projection of Costs of Polio Eradication Compared to Permanent Control. J. Infect. Dis. 221, 561–565 (2020). [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/infdis/jiz488&link_type=DOI) 6. 6. C Fitzpatrick, et al., The cost-effectiveness of an eradication programme in the end game: Evidence from guinea worm disease. PLoS Neglected Trop. Dis. 11 (2017). 7. 7. L Roberts, Battle to wipe out Guinea worm stumbles. World Health Organization delays target date for eradicating the parasite to 2030. Nature 574, 157–158 (2019). 8. 8. O Razum, et al., Polio: From eradication to systematic, sustained control. BMJ Glob. Heal. 4, 1–4 (2019). [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1136/bmjgh-2018-001157&link_type=DOI) 9. 9. MW Merritt, CS Sutherland, F Tediosi, Ethical Considerations for Global Health Decision-Making: Justice-Enhanced Cost-Effectiveness Analysis of New Technologies for Try-panosoma brucei gambiense. Public Heal. Ethics 11, 275–292 (2018). 10. 10. TC Bailey, MW Merritt, F Tediosi, Investing in justice: Ethics, evidence, and the eradication investment cases for lymphatic filariasis and onchocerciasis. Am. J. Public Heal. 105, 629– 636 (2015). 11. 11. CS Sutherland, F Tediosi, Is the elimination of sleeping sickness’ affordable? Who will pay the price? Assessing the financial burden for the elimination of human African trypanosomiasis Trypanosoma brucei gambiense in sub-Saharan Africa. BMJ Glob. Heal. 4, 1–11 (2019). [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1136/bmjgh-2018-001157&link_type=DOI) 12. 12. M Asaria, S Griffin, R Cookson, Distributional cost-effectiveness analysis: A tutorial. Med. Decis. Mak. 36, 8–19 (2016). 13. 13. J Round, M Paulden, Incorporating equity in economic evaluations: a multi-attribute equity state approach. Eur. J. Heal. Econ. 19, 489–498 (2018). 14. 14. P Klepac, I Megiddo, BT Grenfell, R Laxminarayan, Self-enforcing regional vaccination agreements. J. The Royal Soc. Interface 13, 20150907 (2016). 15. 15. S Walker, S Griffin, M Asaria, A Tsuchiya, M Sculpher, Striving for a Societal Perspective: A Framework for Economic Evaluations When Costs and Effects Fall on Multiple Sectors and Decision Makers. Appl. Heal. Econ. Heal. Policy 17, 577–590 (2019). 16. 16. P Klepac, R Laxminarayan, BT Grenfell, Synthesizing epidemiological and economic optima for control of immunizing infections. Proc. Natl. Acad. Sci. 108, 14366–14370 (2011). [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NDoicG5hcyI7czo1OiJyZXNpZCI7czoxMjoiMTA4LzM0LzE0MzY2IjtzOjQ6ImF0b20iO3M6NTA6Ii9tZWRyeGl2L2Vhcmx5LzIwMjEvMDIvMTYvMjAyMS4wMi4xMC4yMDE4MTk3NC5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 17. 17. AA Stinnett, AD Paltiel, Estimating CE Ratios under Second-order Uncertainty. Med. Decis. Mak. 17, 483–489 (1997). [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1177/0272989X9701700414&link_type=DOI) [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=A1997XZ78300014&link_type=ISI) 18. 18. AA Stinnett, J Mullahy, Net Health Benefits: A New Framework for the Analysis of Uncertainty in Cost-Effectiveness Analysis. Med. Decis. Mak. 18, S68–S80 (1998). [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1177/0272989X9801800209&link_type=DOI) [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000073088900009&link_type=ISI) 19. 19. A Briggs, K Claxton, M Sculpher, Decision Modelling for Health Economic Evaluation. (Oxford University Press, Oxford, UK), First edition, (2006). 20. 20. E Fenwick, K Claxton, M Sculpher, Representing uncertainty: The role of cost-effectiveness acceptability curves. Heal. Econ. 10, 779–787 (2001). 21. 21. E Marseille, B Larson, DS Kazi, JG Kahn, S Rosen, Thresholds for the cost–effectiveness of interventions: Alternative approaches. Bull. World Heal. Organ. 93, 118–124 (2015). 22. 22. MY Bertram, et al., Disease control programme support costs: an update of WHO-CHOICE methodology, price databases and quantity assumptions. Cost Eff. Resour. Allocation 15, 21 (2017). 23. 23. AJ Culyer, Cost-effectiveness thresholds in health care: a bookshelf guide to their meaning and use. Heal. Econ. Policy Law 11, 415–432 (2016). 24. 24. B Woods, P Revill, M Sculpher, K Claxton, Country-Level Cost-Effectiveness Thresholds: Initial Estimates and the Need for Further Research. Value Heal. 19, 929–935 (2016). 25. 25. J Ochalek, J Lomas, K Claxton, Estimating health opportunity costs in low-income and middle-income countries: A novel approach and evidence from cross-country data. BMJ Glob. Heal. 3 (2018). 26. 26.1. TTT Edejer, et al. World Health Organization, Making Choices in Health: WHO guide to cost-effectiveness analysis eds. TTT Edejer, et al. (Geneva, Switzerland), (2003). 27. 27.WHO Department of Control of Neglected Tropical Diseases, WHO interim guidelines for the treatment of gambiense human African trypanosomiasis. (2019). 28. 28. I Tirados, et al., Tsetse Control and Gambian Sleeping Sickness; Implications for Control Strategy. PLOS Neglected Trop. Dis. 9, e0003822 (2015). 29. 29. F Courtin, et al., Reducing human-tsetse contact significantly enhances the efficacy of sleeping sickness active screening campaigns: A promising result in the context of elimination. PLoS Neglected Trop. Dis. 9, 1–12 (2015). 30. 30. MH Mahamat, et al., Adding tsetse control to medical activities contributes to decreasing transmission of sleeping sickness in the Mandoul focus (Chad). PLoS Neglected Trop. Dis. 11, 1–19 (2017). 31. 31. BA van Hout, MJ Al, GS Gordon, FFH Rutten, Costs, Effects and C / E-Ratios Alongside. Heal. Econ. 3, 309–319 (1994). 32. 32. ZN Lothgren M, Definition, Interpretation and Calculation of Cost-effectiveness. Heal. Econ. 630, 623–630 (2000). 33. 33. M Remme, M Martinez-Alvarez, A Vassall, Cost-Effectiveness Thresholds in Global Health: Taking a Multisectoral Perspective. Value Heal. 20, 699–704 (2017). 34. 34. MJ Keeling, BT Grenfell, Disease extinction and community size: Modeling the persistence of measles. Science 275, 65–67 (1997). [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6Mzoic2NpIjtzOjU6InJlc2lkIjtzOjExOiIyNzUvNTI5Ni82NSI7czo0OiJhdG9tIjtzOjUwOiIvbWVkcnhpdi9lYXJseS8yMDIxLzAyLzE2LzIwMjEuMDIuMTAuMjAxODE5NzQuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9) 35. 35. BT Grenfell, O. Bjørnstad, BF Finkenstädt, Dynamics of measles epidemics: Scaling noise, determinism, and predictability with the TSIR model. Ecol. Monogr. 72, 185–202 (2002). [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1890/0012-9615(2002)072[0185:DOMESN]2.0.CO;2&link_type=DOI) [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000175147800003&link_type=ISI) 36. 36. CJE Metcalf, K Hampson, AJ Tatem, BT Grenfell, O. Bjørnstad, Persistence in Epidemic Metapopulations: Quantifying the Rescue Effects for Measles, Mumps, Rubella and Whooping Cough. PLoS ONE 8, 1–7 (2013). [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1371/journal.pone.0054583&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=23620825&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2021%2F02%2F16%2F2021.02.10.20181974.atom) 37. 37. S Verguet, R Laxminarayan, DT Jamison, Universal Public Finance of Tuberculosis Treatment in India: An Extended Cost-Effectiveness Analysis. Heal. Econ. 24, 318–332 (2015). 38. 38. M Ekwanzala, et al., In the heart of darkness: Sleeping sickness in Zaire. Lancet 348, 1427– 1430 (1996). [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/S0140-6736(96)06088-6&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=8937285&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2021%2F02%2F16%2F2021.02.10.20181974.atom) [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=A1996VU98600017&link_type=ISI) 39. 39. D Mumba, et al., Prevalence of human African trypanosomiasis in the democratic republic of the Congo. PLoS Neglected Trop. Dis. 5, 1–5 (2011). 40. 40.WHO Expert Committee on human African trypanosomiasis, Control and surveillance of human African trypanosomiasis: report of a WHO expert committee, Technical report (2013). 41. 41. JR Franco, et al., Monitoring the elimination of human African trypanosomiasis at continental and country level: Update to 2018. PLOS Neglected Trop. Dis. 14, e0008261 (2020). 42. 42. RE Crump, et al., Quantifying epidemiological drivers of gambiense human African Trypanosomiasis across the Democratic Republic of Congo. medRxiv (2020). 43. 43. PP Simarro, et al., The Atlas of human African trypanosomiasis: a contribution to global mapping of neglected tropical diseases. Int. journal health geographics 9, 57 (2010). 44. 44. CI Huang, et al., Shrinking the gHAT map : identifying target regions for enhanced control of gambiense human African trypanosomiasis in the Democratic Republic of Congo. medRxiv (2020). 45. 45. M Antillon, et al., Economic evaluation of gambiense human African trypanosomiasis elimination campaigns in five distinct transmission settings in the Democratic Republic of Congo. medRxiv, 2020.08.25.20181982 (2020). [1]: /embed/graphic-1.gif [2]: /embed/graphic-2.gif [3]: T1/embed/inline-graphic-1.gif [4]: /embed/graphic-4.gif [5]: /embed/graphic-5.gif [6]: /embed/graphic-6.gif [7]: /embed/inline-graphic-2.gif [8]: /embed/graphic-7.gif [9]: /embed/graphic-8.gif [10]: /embed/inline-graphic-3.gif [11]: /embed/inline-graphic-4.gif [12]: /embed/inline-graphic-5.gif [13]: /embed/graphic-9.gif [14]: /embed/inline-graphic-6.gif [15]: /embed/graphic-10.gif [16]: /embed/inline-graphic-7.gif [17]: /embed/inline-graphic-8.gif [18]: /embed/graphic-11.gif [19]: /embed/graphic-12.gif [20]: /embed/inline-graphic-9.gif [21]: /embed/inline-graphic-10.gif [22]: /embed/inline-graphic-11.gif [23]: /embed/graphic-13.gif [24]: /embed/graphic-14.gif [25]: /embed/graphic-15.gif [26]: /embed/graphic-16.gif [27]: /embed/graphic-17.gif [28]: /embed/inline-graphic-12.gif [29]: /embed/inline-graphic-13.gif [30]: /embed/inline-graphic-14.gif [31]: /embed/inline-graphic-15.gif [32]: /embed/inline-graphic-16.gif [33]: /embed/inline-graphic-17.gif [34]: /embed/inline-graphic-18.gif [35]: /embed/inline-graphic-19.gif [36]: /embed/inline-graphic-20.gif [37]: /embed/inline-graphic-21.gif [38]: /embed/inline-graphic-22.gif [39]: /embed/inline-graphic-23.gif [40]: /embed/inline-graphic-24.gif [41]: /embed/inline-graphic-25.gif [42]: /embed/inline-graphic-26.gif [43]: /embed/inline-graphic-27.gif [44]: /embed/inline-graphic-28.gif [45]: /embed/inline-graphic-29.gif [46]: /embed/inline-graphic-30.gif [47]: /embed/inline-graphic-31.gif [48]: /embed/inline-graphic-32.gif [49]: /embed/inline-graphic-33.gif [50]: /embed/inline-graphic-34.gif [51]: /embed/inline-graphic-35.gif [52]: F2/embed/inline-graphic-36.gif [53]: F2/embed/inline-graphic-37.gif [54]: F2/embed/inline-graphic-38.gif [55]: F3/embed/inline-graphic-39.gif [56]: F3/embed/inline-graphic-40.gif [57]: /embed/inline-graphic-41.gif [58]: /embed/inline-graphic-42.gif [59]: /embed/inline-graphic-43.gif [60]: /embed/inline-graphic-44.gif [61]: /embed/inline-graphic-45.gif [62]: /embed/inline-graphic-46.gif [63]: /embed/inline-graphic-47.gif [64]: /embed/inline-graphic-48.gif [65]: /embed/inline-graphic-49.gif [66]: /embed/inline-graphic-50.gif [67]: /embed/inline-graphic-51.gif [68]: /embed/inline-graphic-52.gif [69]: /embed/inline-graphic-53.gif [70]: /embed/inline-graphic-54.gif [71]: F4/embed/inline-graphic-55.gif [72]: /embed/inline-graphic-56.gif [73]: /embed/inline-graphic-57.gif [74]: /embed/inline-graphic-58.gif