Economic evaluations of human schistosomiasis interventions: a systematic review and identification of associated research

Schistosomiasis is one of the most prevalent neglected Background: tropical diseases (NTDs) with an estimated 229 million people requiring preventive treatment worldwide. Recommendations for preventive chemotherapy strategies have been made by the World Health Organization (WHO) whereby the frequency of treatment is determined by the settings prevalence. Despite recent progress, many countries still need to scale up treatment and important questions remain regarding optimal control strategies. This paper presents a systematic review of the economic evaluations of human schistosomiasis interventions. A systematic review of the literature was conducted on 22nd Methods: August 2019 using the PubMed (MEDLINE) and ISI Web of Science electronic databases. The focus was economic evaluations of schistosomiasis interventions, such as cost-effectiveness and cost-benefit analyses. No date or language stipulations were applied to the searches. We identified 53 relevant health economic analyses of Results: schistosomiasis interventions. Most studies related to Schistosoma followed by Several studies also included other japonicum S. haematobium. NTDs. In Africa, most studies evaluated preventive chemotherapy, whereas in China they mostly evaluated programmes using a combination of interventions (such as chemotherapy, snail control and health education). There was wide variation in the methodology and epidemiological settings investigated. A range of effectiveness metrics were used by the different studies. Due to the variation across the identified studies, it was not Conclusions: possible to make definitive policy recommendations. Although, in general, the current WHO recommended preventive chemotherapy approach to 1-3 4 5 6

the current WHO recommended preventive chemotherapy approach to control schistosomiasis was found to be cost-effective. This finding has important implications for policymakers, advocacy groups and potential funders. However, there are several important inconsistencies and research gaps (such as how the health benefits of interventions are quantified) that need to be addressed to identify the resources required to achieve schistosomiasis control and elimination.

Introduction
Schistosomiasis (also known as bilharzia) is an acute and chronic parasitic disease caused by blood flukes of the genus Schistosoma. People become infected when larval forms of the parasite (released by freshwater snails) penetrate the skin during contact with infested water. It is one of the most prevalent neglected tropical diseases (NTDs) with an estimated 229 million people requiring preventive treatment worldwide 1 . There are two main forms of human schistosomiasis: urogenital caused by S. haematobium, and intestinal caused by S. mansoni, S. guineensis, S. intercalatum, S. japonicum, and S. mekongi. Schistosomiasis can result in anaemia, chronic pain, diarrhea, and malnutrition, causing poor school performance and lower fitness 2 . It is estimated that 89.3% of those requiring treatment for schistosomiasis live in Africa 1 .
In Africa, schistosomiasis control mainly focuses on mass school-based or community-wide preventive chemotherapy using praziquantel 3 : the large-scale distribution of drugs to eligible populations, without diagnosing or testing individuals for current infection. The World Health Organization's (WHO) recommended guidelines for preventive chemotherapy are currently dependent on the prevalence of infection in school-aged children (SAC; 5-14 years old) prior to preventive chemotherapy 4 . Although SAC are generally the focus using targeted school-based preventive chemotherapy, treatment of adults is also recommended depending on the endemicity. Specifically, in low-risk communities (below 10% baseline prevalence in SAC), treatment of SAC once every 3 years is recommended, along with treatment of suspected cases 4 . For moderate-risk communities (10-50% baseline prevalence in SAC), biennial treatment of SAC and at-risk adults is recommended 4 . For high-risk communities (>50% baseline prevalence in SAC), annual treatment of SAC and at-risk adults is recommended 4 . Outside of Africa, it is more common for preventive chemotherapy to be complemented by other interventions, such as health education and snail control 5-8 .
Merck KGaA have committed to donate 250 million tablets of praziquantel annually for the treatment of schistosomiasis, primarily for SAC in Africa 9 . The current WHO goals of morbidity control and elimination as a public health problem 10 , focus on preventive chemotherapy to reduce the prevalence of heavy-intensity infections in SAC and furthermore the interruption of transmission in selected settings (reducing the incidence of infections to zero) 11,12 (Box 1). The post-2020 WHO goals are currently being developed, along with revisions to the current preventive chemotherapy guidelines.
The treatment coverage of SAC has increased notably over the last decade, reaching 61.2% in 2018 1 . However, despite the WHO recommendations, adults are often missed, and their coverage is markedly lower (18.2%) 1 . Although most treatment programmes currently rely solely on preventive chemotherapy, additional operational components, such as the provision of potable water and adequate sanitation, hygiene education, behaviour change and snail control may become essential when moving towards elimination of schistosomiasis 13 .
Despite recent progress, important questions remain regarding optimal schistosomiasis strategies 14 . Economic evaluations have an important role in informing strategies and policy decisions. The aim of this paper is to provide a descriptive overview of the economic evaluations that have been conducted, the areas of research focus and the key findings for human schistosomiasis interventions. Based on these findings, we describe important areas of uncertainty, drivers of variation and remaining research gaps that require further attention within future economic evaluations for schistosomiasis.

Methods
A systematic review of the literature was conducted on 22 nd August 2019 using the PubMed (MEDLINE) and ISI Web of Science electronic databases. The focus was economic evaluations of schistosomiasis interventions, such as cost-effectiveness and cost-benefit analyses. Although not technically an economic evaluation, studies reporting estimated economic benefits of schistosomiasis interventions were also included. Variants of the following search terms were used to find relevant papers: schistosomiasis and either cost-benefit, cost-effectiveness, economic(s), or economic evaluation. No specific date or language stipulations were applied to the searches. A more detailed summary of the search terms and the PRISMA checklist are supplied as extended data. Studies on schistosomiasis interventions that were identified as a relevant type of economic analysis were included. Studies related to health economic evaluation of schistosomiasis diagnostics/monitoring, cost of illness and willingness to pay for treatment were excluded.
The titles and abstracts of the identified papers were examined initially for relevance by two independent reviewers [HCT and JT]. The full texts for potentially relevant articles were then reviewed to determine eligibility for inclusion. The bibliographies of papers suitable for inclusion were then scanned for studies not originally retrieved from the databases. Discrepancies were solved by consensus among the reviewers. The full selection process is outlined in Figure 1.
The identified health economic studies were grouped into two broad types: • Cost-effectiveness analyses and cost-utility analyses: Cost-effectiveness analysis is a form of economic analysis that compares the relative costs and effectiveness of different courses of action. The effectiveness of the interventions under investigation is measured in terms of natural units (such as life years gained, cases averted, or heavy cases averted). Cost-utility analysis is a specific subtype of cost-effectiveness analysis, where the effectiveness of the intervention is measured using a "utility-based" unit (such as disability-adjusted life years (DALYs) averted and quality-adjusted life years (QALYs) gained) (Box 1).
• Cost-benefit analyses and estimates of economic benefits: Cost-benefit analysis is a form of economic analysis that compares the relative costs and benefits of different courses of action (Box 1). Unlike, costeffectiveness analyses, the benefits of an intervention are expressed in monetary terms, i.e. compares the cost of an intervention to its monetary benefits. We also included studies that only estimated the economic benefits of schistosomiasis interventions.
Due to the number of studies identified and the range of research questions investigated (Figure 1), it was not possible to provide a detailed summary of the results of every study. Instead in line with the aim of the paper, we provide an overview of the studies that have been done and discuss key overarching findings and research gaps that need to be addressed. Due to their significance to policy makers, we outline further details of the studies reporting the cost per DALY averted related to preventive chemotherapy.

Results
We identified 53 relevant health economic analyses of schistosomiasis interventions. An overview of the studies is presented in Table 1. Interestingly, there were notably more economic evaluations of schistosomiasis than those identified in previous reviews for other helminth infections of human health importance 15-17 .
The number of economic evaluations performed on schistosomiasis interventions has gradually increased over time, although a notable number were published pre-2000 (Table 1). Most studies were related to S. japonicum, followed by S. haematobium, and S. mansoni ( Figure 2). Only one study related to S. mekongi (Table 1). Several studies also included other NTDs, particularly soil-transmitted helminths (STH). In Africa, most studies evaluated mass preventive chemotherapy, whereas in China they mostly evaluated programmes using a combination of interventions together (Table 1). Few studies were performed for other settings ( Table 1).
The methods used by the studies to parameterise the costs of preventive chemotherapy were variable: some used primary cost data whereas others performed rough calculations or used assumed crude benchmark values (Table 2). There was also variation regarding the use of financial or economic costs within the studies (Box 1). The economic cost of an intervention is typically higher than the financial cost (Table 2). If praziquantel was assumed to be purchased it would be counted as a financial cost. Depending on the perspective of the analysis, the value of any donated praziquantel would be included as an    economic cost. It should be noted that the shift towards drug distribution being integrated within school systems reduced the delivery costs associated with preventive chemotherapy. Consequently, the results of many of the earlier studies cannot be directly generalised to current control programmes ( Table 2).

The cost-effectiveness analyses
We identified 41 cost-effectiveness analyses of schistosomiasis interventions ( Table 1). The majority investigated preventive chemotherapy, with few looking at the cost-effectiveness of water, sanitation and hygiene (WASH) and behavioural change ( Table 1). A range of effectiveness metrics were used by the different studies, including specific types of morbidity averted (such as anaemia), cases prevented, heavy infections averted, decreases in the infection rate (in humans, animal hosts and snails) and DALYs averted.
One of the key areas of analysis was whether selective treatment should be used i.e. where only those that are tested positive for infection (or suspected to be infected) are treated. This strategy uses less praziquantel relative to mass preventive chemotherapy. Some earlier studies found that selective treatment could be more cost-effective than mass treatment 24,48 . However, as the price/value of praziquantel and delivery declined over time, mass school-based or community-wide treatment became more cost-effective than using screening/ selective based approaches, particularly in high prevalence settings 19,31,48 . By adopting cheaper methods of performing selective treatment, such as water contact surveys rather than testing for infection with diagnostics, selective treatment may become more cost-effective in certain settings 30,68 .
The estimated cost per DALY averted for annual mass school-based preventive chemotherapy for schistosomiasis (or both STH and schistosomiasis) ranged widely between US$6-692 in moderate and high prevalence settings ( Most studies evaluating snail control were in China and looked at several types of control method, including environmental modification and molluscicide use. These were typically part of a programme and not a standalone intervention. It is critical not to overgeneralise studies regarding the impact and cost-effectiveness of snail control. For example, the benefit of snail control will likely be higher for zoonotic species such as S. japonicum. Only Lo et al. 28 investigated the cost-effectiveness of snail control for controlling S. haematobium using DALYs as the effectiveness metric. Their results supported the use of snail control within schistosomiasis control strategies, particularly

Box 2. Cost-effectiveness thresholds
To determine whether an intervention is cost-effective using a cost-effectiveness analysis, the cost per DALY averted is compared to a cost-effectiveness threshold. An often misunderstood aspect of cost-effectiveness analysis is that when comparing mutually exclusive interventions (such as school-based vs community-wide preventive chemotherapy), the goal is to find the most effective intervention which has an incremental cost-effectiveness ratio below an established cost-effectiveness threshold. It is not about finding the intervention/ strategy with the lowest cost-effectiveness ratio (i.e. the strategy with the lowest cost per DALY averted). The findings of cost-effectiveness analysis are often miscommunicated to policymakers and many refer to an intervention being the "most cost-effective", when instead they mean it is the optimal intervention for the given cost-effectiveness threshold.
The most appropriate cost-effectiveness thresholds are under debate within the global health field 69-71 . Some studies used the costeffectiveness threshold set by the WHO-CHOICE 72 , namely a cost per DALY averted < 3 times the country's GDP per capita. However, this is now considered to be too high and has been widely criticised 69-71,73,74 . Interestingly, recent analyses have indicated that a significantly lower cost-effectiveness threshold closer to < ½ the country's per capita GDP would be more appropriate for low-income countries 73,75 . The Disease Control Priorities project (Third Edition) also used a more conservative threshold of US$200 per DALY averted to identify priority interventions for consideration in low-income countries 76 . These different thresholds are shown in the A further study by Ndeffo Mbah et al. 27 evaluated the costeffectiveness of a community-based intervention for averting S. haematobium infections and resultant HIV. The intervention integrated the provision of clean water, sanitation, and health education (WSHE) for the entire community with annual praziquantel treatment of SAC. The cost-effectiveness of the community-based intervention was found to vary with the cost of WSHE, the efficacy of WSHE for reducing S. haematobium transmission and the duration of the intervention. The intervention remained cost-effective for a range of WSHE efficacies and became more cost-effective over time. Overall, the results indicated that this integrated community-based approach towards schistosomiasis control could effectively reduce the health and economic burden associated with S. haematobium and HIV infections in sub-Saharan Africa.
The cost-benefit analyses and estimates of economic benefits Schistosomiasis can be debilitating and can negatively impact productivity 2,67,82-91 . A recent systematic review and meta-analysis found that Schistosoma infection/non-treatment was significantly associated with educational, learning and memory deficits in SAC 92 .
A number of studies have estimated the economic benefits or performed cost-benefit analyses of schistosomiasis interventions (Table 1). For example, • Redekop et al. 61 estimated US$17.4 billion in economic benefit would be generated (between 2011-2013) if the WHO 2020 roadmap goals for schistosomiasis were achieved. The majority of this benefit was due to prevented anaemia.
• De Neve et al. 58 estimated the health, financial, and education gains of investing in preventive chemotherapy for schistosomiasis, STH, and lymphatic filariasis in Madagascar. They found that preventive chemotherapy could avert a notable amount of school absenteeism and reduce patients' out-of-pocket expenditure.
• Zhou et al. 35 investigated the cost-benefit of the national schistosomiasis control programme in China (1992-2000).
The net benefit-cost ratio was 6.20.
• Miguel and Kremer 53 found that deworming for STH and schistosomiasis was likely to increase the net present value of wages by over US$30 per treated child based on the estimated rate of return to education in Kenya. Baird et al. 67 also subsequently estimated that mass deworming may generate more in future government revenue than it costs in subsidies.
These studies indicate that schistosomiasis interventions may generate notable economic benefits. These types of analyses were less common compared to cost-effectiveness analyses. This is likely due to the methodological challenges of placing a monetary value on the benefits of schistosomiasis interventions. There was notable variation regarding the methodology of these identified studies and what was used to justify the assumptions made. It should be highlighted that the results of such studies can be highly sensitive to the methodology used. Further standardisation is urgently needed in this area (not just for studies on schistosomiasis).

Discussion
We identified a wide range of economic evaluations of schistosomiasis interventions. Due to the variation in methodology and epidemiological settings, it was not possible to make definitive policy recommendations based on the identified studies. However, the results of the review indicate that annual schistosomiasis treatment interventions are generally estimated to be cost-effective in moderate and high prevalence settings (Box 1) and can generate notable economic benefits. In low prevalence settings aiming for morbidity control or elimination as a public health problem (Box 1), annual mass preventive chemotherapy may not be cost-effective, supporting the current WHO recommendation of less frequent treatment in such settings 10 . There are also a growing number of studies evaluating alternative strategies to those currently recommended, such as snail control and WASH.
Most of the studies related to S. haematobium and S. mansoni evaluated mass preventive chemotherapy. Contrastingly, most of the studies relating to S. japonicum evaluated comprehensive control programmes with multiple components (such as snail control, preventive chemotherapy, health education). These studies were difficult to compare with others as they tended to use metrics based on reductions in infection rates (humans, animal hosts and snails) rather than DALYs or cases/morbidity averted.
There was notable variation in methodology across the different studies (which likely lead to the wide range in the estimated costs per DALY averted ( Table 2)). Key sources of variation included 1) the assumed costs, 2) the epidemiolocal setting, and 3) the methods used to quantify the effectiveness of an intervention.

Variation in the assumed cost of interventions
The average delivery cost of annual preventive chemotherapy using praziquantel was typically assumed (or in some cases estimated) to be between US$0.20-0.50 per treatment 93 . However, the methods used to parameterise the costs of different interventions varied widely making it difficult to understand the relative costs of different interventions and how costs may vary across countries/regions. This variation also made it difficult to directly compare the different studies. An important consideration is whether financial or economic costs are being used (Box 1). Economic costs are considered the gold standard within economic evaluations as they better reflect the sustainability and replicability of interventions. Many studies did not formally state if they were using financial or economic costs. There was also variation regarding whether the cost of donated praziquantel had been included within the analysis as an economic cost (Table 2).
When interpreting and comparing the studies, it is vital to consider that integrating drug distribution through the school system rather than using mobile teams notably reduced the delivery costs associated with preventive chemotherapy. However, many of the earlier health economic studies identified (pre-2000) related to the use of mobile teams to deliver treatments, which is more costly than the currently used school-based platform. Additionally, in the past, the cost of praziquantel was substantially higher 47,94 with many of the studies having praziquantel costing US$0.60 per treatment. However, several forms of generic praziquantel tablets are now available, currently costing around US$0.08 per tablet 95,96 and some countries (such as China 94,97 ) produce their own. Merck KGaA have also committed to donate 250 million tablets of praziquantel annually (primarily for SAC in Africa) 10 . Consequently, the findings of many of the earlier studies cannot be directly generalised to current control programmes.

Variation in the epidemiolocal setting
Pre-control endemicity is a notable driver in the estimated cost-effectiveness of schistosomiasis interventions. As pre-control endemicity increases, the intervention generally becomes more cost-effective. This needs to be considered when comparing studies and using them to inform policy.
In contrast, the importance of the age profile of infection is more subtle and often ignored. Although the assumed age profile of infection would have a relativity small impact on the estimated cost-effectiveness of the recommended schoolbased treatment, it can have notable implications regarding the relative benefit of alternative interventions, particularly regarding the benefit of targeting additional age groups with preventive chemotherapy. Consequently, it needs to be considered in studies investigating the cost-effectiveness of alternative schistosomiasis interventions to school-based treatment (Table 3) and has substantial consequences regarding the generalisability of economic evaluations and their conclusions. Figure 3 illustrates some of the available age profile data. It should be noted that the shape of the typical age profile (and variation around them) differs for each species. The importance of these age profiles of infection and their variation is often not accounted for in modelling studies and policy recommendations 98 . There is a real danger that overgeneralising in this area could lead to highly inefficient recommendations. Caution is needed when parameterising this aspect of models when performing economic evaluations. For example, if the data relating to adults are not representative, such as being from high-risk adults only, the assumed average burden of adult infection and costeffectiveness of switching to community-wide treatment will be overestimated 99 .
A further factor is whether other species are included within the economic evaluation (Table 1 and Table 2). Based on the available studies, we cannot make inferences on the relative cost-effectiveness for different species. Generally, accounting for other species would increase the cost-effectiveness or cost-benefit of preventive chemotherapy. In addition, Lo et al. highlighted that the optimal strategy for schistosomiasis can depend on if it is co-endemic with STH 57 .

Variation in the methods used to quantify the effectiveness of interventions
A key area of uncertainty and variation is how the health gains/effectiveness of interventions were quantified. This is important as the chosen method impacts the outcome of an economic evaluation, potentially leading to different policy recommendations. For example, when basing effectiveness on the number of worm years or heavy cases years averted (both metrics related to infection intensity), models find that the total effectiveness tends to increase with the transmission setting 98 . In contrast, when the effectiveness is based on reductions in the prevalence of infection, the estimated effectiveness can be greater in lower transmission settings 98 . This would imply that when modelling reductions in morbidity are based on reductions in prevalence, the results could find that it is more cost-effective to treat in lower transmission settings -which is likely to be misleading in terms of morbidity control. The effectiveness metric also has a particularly notable influence on the relative benefit of annual community-wide to school-based treatment: metrics based on reduction in infection intensity tend to estimate a lower benefit relative to those based on infection prevalence 98 . Pre-SAC: pre-school-aged children (2-4 years old), SAC: school-aged children (5-14 years old), adults: 15+ years old, EPG: eggs per gram.
Note that these show the mean infection intensity for the entire age group (i.e. including non-infected individuals). * The values from Turner et al. 98 were converted from worm burden to EPG. Typically, the ideal metric for evaluating different control strategies in low-and middle-income countries is the number of DALYs averted (Box 1). As this metric is used for a wide range of diseases, the cost-effectiveness estimates can be directly compared to other healthcare interventions. This makes it possible to have standardised thresholds to class whether an intervention is cost-effective (Box 2) which is rarely possible for disease-specific metrics. However, the DALY burden of schistosomiasis is often calculated by simply applying a disability weight (representing the disability of an `average' prevalent case of schistosomiasis) to the prevalence of infection. Although this approach may be a suitable method for approximating the disease burden of schistosomiasis at a given point in time, we would argue that it is misleading to apply this framework to evaluate the effectiveness of schistosomiasis interventions 98 . This is because the relationship between schistosome infection and morbidity is complex and often not merely due to the presence or absence of infection 98, 113 .
Only quantifying health gains based on reductions in prevalence would implicitly assume that curing a very light infection has a health benefit whereas reducing a heavy infection to a light infection has none.
The DALY disability weights have also been controversial for schistosomiasis (defined in terms of 0 being perfect health and 1 being death (Box 1)) 2,114,115 . The Global Burden of Disease (GBD) 1990 study gave schistosomiasis a disability weight of 0.005 116 , implying a very minimal level of disability for "average" schistosomiasis. In contrast, a higher weight of 0.02 was proposed by King et al. 2,117 . The GBD now has an average weight of 0.006 for mild infection, with higher disability weights for specific forms of morbidity (such as anaemia, bladder pathology, and haematemesis) 118 119 .
The method used to quantify the effectiveness of interventions and the associated uncertainty needs greater consideration when using schistosomiasis related economic evaluations for informing policy. The choice of metrics will need to reflect the goal and will therefore likely need to be changed as settings move from aiming for morbidity control to interruption of transmission. In most African settings, this will likely mean that the metrics used are currently related to the level of morbidity, whereas in China the metrics will need to reflect the level transmission.

Limitations of this analysis
A potential source of bias of the search strategy is that it did not capture economic evaluations published outside of the searched electronic databases, such as grey literature, policy documents/reports, and non-English language publications. In addition, texts without available abstracts or those not clearly identifiable as an economic evaluation may be missed. Efforts were made to minimise this bias by searching the bibliographies of selected studies. There could also be a degree of publication bias, with economic evaluations with negative or less favourable results being less likely to be published. A number of the studies did not have the full text published in English. These were included within the results based on information from their abstracts (Table 1) but further analysis was limited.

Research needs for future health economic analyses
There are important inconsistencies and research gaps that need to be addressed as we move towards the post-2020 WHO goals.
In the following subsections we outline several key research needs and considerations for future economic evaluations.
Preventive chemotherapy cost data The costs of preventive chemotherapy vary across different settings 15-17 , both within and between countries. A driver is the size of the targeted population [120][121][122] . This is because the delivery costs of preventive chemotherapy can have economies of scale, such that as the number of people treated increases, the cost per treatment tends to decrease 123,124 . Another key driver for the costs is the targeted age group. There is very little primary data on the relative cost of school vs community-based preventive chemotherapy 16 . Many studies in this area make assumptions on the relative cost based on little data. This is an important research gap that needs to be filled to allow further analysis to inform whether and when to switch to community-wide preventive chemotherapy.
It should be noted that preventive chemotherapy delivery costs are not constant and will likely increase over time as countries develop and as expectations regarding the quality of distribution increase. Crucially, the cost per treatment of programmes will also likely increase considerably as they approach their "last mile", particularly with an end goal of interruption of transmission. This is due to the increase in costs associated with expanding programmes to include harder-to-reach areas and groups 123,125 . Hence, a greater understanding of the variation in the costs of preventive chemotherapy across settings and quantifying its impact within subsequent economic evaluations is an important research gap for future studies 123 .
A notable research gap is the lack of understanding of the costs of integrated NTD control 126,127 and how integration may influence the costs and cost-effectiveness of implementing different control strategies 123 .

Evaluation of alternative interventions and strategies
As more data on the costs and effectiveness of interventions become available, economic evaluations of alternative control strategies to preventive chemotherapy need to be conducted, particularly expanding to African settings. This includes WASH 128 , vaccines 129 , behaviour change, paediatric treatment, snail control and alternative preventive chemotherapy delivery platforms. Further studies are also needed to quantify the costs and effectiveness of alternative delivery strategies, beyond schoolbased and community-wide preventive chemotherapy. In particular, the coverage and adherence of different age groups when using various treatment delivery platforms needs to be assessed 130 . For example, the number of high-risk adults treated through a school-based delivery system and the number of non-enrolled children missed needs to be quantified. It will also be important to investigate the cost-effectiveness of targeting high-risk adults within the community in comparison to targeting the whole community. If a sufficient coverage of high-risk adults could be achieved in addition to school-based treatment, it could be a more cost-effective alternative to mass community-wide preventive chemotherapy. This option is often missed in economics evaluations (due to the lack of data), with most studies only evaluating mass school-based treatment and community-wide preventive chemotherapy.
It is important to consider that as schistosomiasis programmes become more integrated with other NTD programmes or health/education interventions, the costs will likely go down (due to economies of scope 123 ), but the quality of the programme and treatment coverage may also decline.
Importantly, on-going mass treatment may no longer be optimal in low transmission settings and other strategies (such as selective treatment) may need to be considered. Furthermore, the level of treatment coverage that can be maintained with passive treatment at public health facilities needs to be investigated before stopping mass treatment.
Quantifying the health benefits of interventions Currently, we would argue that it is difficult to accurately capture the impact of treatment on the morbidity related to schistosomiasis 98 . There is an urgent need for the development of frameworks that can accurately estimate the number of DALYs averted by schistosomiasis interventions. An ideal framework will need to account for: i. The differences in how pathogenic different levels of infection are in different age groups (similar to the approach used for STH 131 ). A crucial area of uncertainty for this is the relative burden of light infections 98 and how this varies for different species.
ii. Which forms of morbidity are permanent vs reversible with treatment 98 . This would need to include what morbidity is present in individuals without a current infection 132 .
Without a better framework, results regarding the benefit of expanding treatment beyond SAC will be highly dependent on assumptions from limited empirical evidence. In particular, overestimating the relative burden of light infections could overestimate the benefit and cost-effectiveness of community-wide treatment (and vice versa) 98 . Further work is needed to accurately quantify the excess mortality related to schistosomiasis and to reassess the death estimates due to schistosomiasis.
Epidemiological data, one health and hotspots Better epidemiological data are needed to parametrise models used for schistosomiasis economic evaluations. This includes prevalence and intensity of infection data across all ages to inform pre-control age profiles of infection, treatment coverage and adherence data, and WASH data. By collecting this data, economic evaluations can be more accurate and informative.
Animal populations have also been shown to be infected with schistosomiasis. The impact of zoonotic transmission to human infection differs by species and this influences the impact of different interventions. Some studies investigating S. japonicum included targeting the zoonotic reservoir (Table 1) but further work is needed to investigate the costs and benefits of applying a one health approach across a range of settings 133 .
Despite treatment, there are hotspots where infection remains at persistently high levels 14, [134][135][136] . Hotspots can be caused by various programmatic factors such as poor treatment coverage/ adherence, movement of infected individuals and intense water contact. Additionally, it is possible that these may be caused by declining drug efficacy as some individuals remain infected following multiple treatment rounds 137,138 . These factors pose threats to the effectiveness of treatment programmes and need to be considered in future economic evaluations. Ignoring these could lead to the long-term cost-effectiveness of interventions being overestimated. Economic evaluations also have a role in informing the optimal strategies for identifying and controlling these hotspots.
The impact of the goal of the intervention The goal of a programme is an important consideration when interpreting economic evaluations 15,17 . Some areas, particularly in Asia, are aiming to move beyond morbidity control and elimination as a public health problem to interruption of transmission 139 (Box 1). Modelling studies have indicated that breaking transmission may theoretically be possible with mass preventive chemotherapy alone at high coverage and compliance 98,140 . Field studies and trials are currently underway to confirm if this is feasible in practice 141,142 . Moving towards elimination will likely require more intensive strategies, including treatment in low prevalence settings 14 . It should be noted that an increase in programmatic costs would be required-at least in the short term. More intensive and expensive interventions may not be cost-effective in the context of morbidity control but could be when breaking transmission. For example, in some settings, annual school-based preventive chemotherapy may be the optimal strategy in terms of controlling schistosomiasisrelated morbidity, but other more intensive strategies may be more cost-effective when the goal is the interruption of transmission.

Conclusions
A wide range of economic evaluations of schistosomiasis interventions have been conducted. Based on the identified studies, annual preventive chemotherapy has generally been found to be cost-effective in moderate to high prevalence settings, thereby supporting the WHO recommendation of less frequent treatment in low prevalence settings. However, the cost-effectiveness of mass preventive chemotherapy varies depending on the setting and it is difficult to generalise across species and regions/countries. There are also a growing number of studies evaluating alternative strategies to schoolbased preventive chemotherapy. Due to the variation in methodology and epidemiological settings, it was not possible to make definitive policy recommendations based on the identified studies. There are several important research gaps that need to be addressed as we move towards the post-2020 WHO goals. In particular, evaluations of interventions other than mass preventive chemotherapy are needed (especially in low transmission settings). Further work is also needed to develop frameworks that can more accurately quantify the health benefits of different strategies. It is also important that future health economics evaluations accurately account for the underlying epidemiology (such as the variation in the age profile of infection) and for the factors which may be driving persistent hotspots.

Data availability
Underlying data All data underlying the results are available as part of the article and no additional source data are required. I don't know exactly how many systematic reviews of economic evaluations or costing evaluations related to schistosomiasis exist, but the paper made almost no references to other reviews. Perhaps almost no such reviews exist. However, I know of one review of the unit costs (rather than the cost-effectiveness) of schistosomiasis: Salari (2019) . Could the authors delineate how their observations of the cost et al. components compare to those found by Salari ? et al.

Extended data
Once again, congratulations on a job well done. Such synthesizing efforts are far from straightforward, but are invaluable to understand how policy should move forward.