Search strategy and selection criteria

Publications were collected by searching the MEDLINE (via OVID), Embase Classic+Embase (via OVID), Global Health - OVID, PubMed, and Econ Lit databases on the 29^th of April 2024. The search terms included variants of the following keywords (dengue, Wolbachia, biocontrol, economic evaluation, cost-effectiveness analysis, cost-utility analysis, cost-benefit analysis, cost-consequence analysis, and cost-minimisation analysis) without any date- or language- restrictions. Additional searches of the grey literature were performed as outlined in the S1 File.

The retrieved citations were uploaded to Covidence, a web-based systematic review software [9] to identify and remove duplicates. The titles and abstracts of all the articles were scanned to identify relevant studies by two reviewers (HCT and TMH). The bibliographies of related papers were also searched to identify additional articles not initially retrieved from the databases. The full texts of the identified studies were then reviewed for eligibility by two reviewers (HCT and TMH). Any studies with uncertainty regarding their inclusion were discussed and resolved by the reviewers. The full selection process is illustrated in Fig 1. More detailed information on the search terms and PRISMA checklist are provided in the S2 File [10]. As stated in the relevant conflict of interest section, it is important to note that HCT has received research funding from The World Mosquito Program in the past. The World Mosquito Program had no involvement in this paper (other than being a potential source of studies within the search of the grey literature).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Flow diagram outlining the inclusion and exclusion of the identified studies.

A PRISMA checklist is provided in S2 File [10].

https://doi.org/10.1371/journal.pntd.0014395.g001

Based on the database and grey literature searches (the latter including pre-prints), we included studies that conducted an economic evaluation (i.e., cost-utility analysis, cost-effectiveness analysis, or cost-benefit analysis [11]) related to the use of the Wolbachia-based intervention to control dengue (both suppression and replacement-based strategies). Although no language restrictions were applied at the search stage, non‑English full texts were excluded during screening because translation resources were not available. Reviews/systematic reviews, intervention costing studies, and conference abstracts were also excluded.

In cases where a non–peer‑reviewed study was identified, and a peer‑reviewed version was later published, we also considered the peer‑reviewed version when extracting data, even if it was published after the search timeframe.

Data extraction and output

Methodological and contextual data from each study were systematically extracted by two independent reviewers (HCT and TMH). Key methodological details included the study setting, base case time horizon, the modelling approach used to estimate effectiveness, the primary outcome measure, and the discount rate applied in the base case results were extracted. Additional relevant assumptions were documented, such as the baseline burden of symptomatic cases in the absence of intervention, the effectiveness of the intervention, the size of the release area, the population covered, the cost of the intervention and the corresponding cost year (S3 File).

We grouped the scale of the intervention into three categories: National, large scale (targeting at least 200,000 people), and small scale (targeting under 200,000 people).

The results of the studies were summarised by extracting their reported cost-effectiveness ratios and benefit-cost ratios. The cost-effectiveness ratios were stratified by the different perspectives employed [12]. All of the benefit-cost ratios extracted related to the societal perspective (S1 File: S1 Box).

Cost values (and economic benefits) were adjusted for inflation to 2024 prices using the United States (US) gross domestic product (GDP) deflators [13,14]. Incremental cost-effectiveness ratio (ICER) values were not adjusted and reported with the corresponding cost year.

To provide overarching, high-level conclusions, cost‑effectiveness ratios were compared against a threshold of 0.5 times a country’s per‑capita GDP [15,16] (values taken from the World Bank [17]). This benchmark is gaining traction as an alternative to the earlier (widely criticised) benchmarks that classified interventions as “cost‑effective” if they fell below 1–3 times per‑capita GDP per disability-adjusted life year (DALY) averted [18]. Furthermore, an empirical analysis of historical spending across 174 countries found that 51% implicitly used thresholds of cost per quality-adjusted life year (QALY) gained of 0.5 times their per capita GDP [19]. However, it remains important to recognise that appropriate cost-effectiveness thresholds continue to be debated, and GDP-based benchmarks have well-recognised limitations that should be considered when interpreting these findings [20–23].

Quality assessment

Quality assessment of the identified relevant studies was undertaken using the CHEQUE tool [24]. This was conducted by two independent reviewers (TMH and HCT) following the procedures outlined for screening and data extraction.

Baseline disease burden

Due to epidemiological and entomological differences, it is expected that the baseline disease burden of dengue varies from area to area. It should also be noted that symptomatic dengue cases are underreported and that the proportion of symptomatic infections reported to surveillance varies between and within countries; therefore, adjustments and modelling approaches are needed when evaluating the population-level impact of dengue interventions. The corresponding actual disease burden estimated will vary depending on the approach employed.

As expected, the assumed baseline incidence of infection varied between the settings investigated (Table 2). However, in some cases, this was not clearly reported. The approach used to approximate the baseline infection burden and adjust for underreporting was also variable (Table 2). However, it is encouraging that all the studies adjusted for underreporting in some form rather than using reported case numbers directly. Several studies adjusted the reported number of cases with expansion factors (the ratio of the estimated true number of symptomatic dengue cases to the reported number of dengue cases). Brady et al. [28] used an unweighted ensemble of multiple previous approaches to obtain consensus estimates of the burden of dengue in Indonesia (taken from O’Reilly et al. [35]). In Turner et al. [29], the baseline burden was based on the incidence of dengue estimated by the Global Burden of Disease (GBD) 2019 study [36] for Vietnam, which was distributed sub-nationally based on the mapping projections by Bhatt et al. [37]. Shepard et al. [26] used burden estimates derived from projections by the GBD study.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Summary of the assumed baseline burden and the base case effectiveness and cost of the Wolbachia intervention.

https://doi.org/10.1371/journal.pntd.0014395.t002

Effectiveness of intervention

A summary of the assumed effectiveness of Wolbachia deployments is presented in Table 2. The only evaluation of a suppression-based programme assumed a reduction in the incidence of dengue cases varying between 40–80%. This was based on the results of field trials of this intervention in Singapore [38] (Table 2).

Several of the evaluations of the replacement-based strategy assumed the effectiveness of Wolbachia deployments in reducing the incidence of symptomatic cases to be 75–77%, based on data from a cluster randomised trial and quasi-experimental studies in Yogyakarta (with the wMel strain) [39,40]. Brady et al. [28] used data from vector competence studies to model Wolbachia-induced changes in force of infection, which could then be projected to areas with different force of infection values and thus produce locally varying estimates of effectiveness. The values varied between 65.7% and 94.4% (within the treated areas).

The duration of the effectiveness of the replacement-based strategy was typically assumed to last 10 years for the base case results. This assumption was often included within the sensitivity analysis. This is a key parameter in the economic analysis of the replacement-based strategy. It should be noted that 10 years was used in the first studies in this area, as that was approximately how long wMel had persisted in northern Australia at the time they were conducted. This could therefore be justifiably increased to closer to 15 years now.

DALY calculations

Seven of the nine studies used DALYs averted as their effectiveness measure (Table 1). DALYs are calculated as the sum of two components: years of healthy life lost due to disability (YLDs) and years of life lost due to premature mortality (YLLs) [41,42]. Within a DALY calculation, YLDs are calculated using a disability weight factor ranging between 0 and 1, which reflects the severity of the disease sequelae, with 0 representing perfect health and 1 representing death. The disability weights used for dengue DALY calculations have changed significantly over time [43]: a summary of the different weight values is provided in S1 File: S2 Box. Interestingly, most of the economic evaluations we identified did not use the disability weights officially designated for dengue in the current GBD studies, which were only used by Zimmermann et al. [31]. Four of the studies used the disability weights estimated by Zeng et al. [44] (Table 1). Soh et al. [34] used a different approach based on what was developed by Meltzer et al. [45] (developed before the GBD 2004 update). Knerer et al. [32] used the disability weights outlined within the GBD 2004 update.

QALY calculations

Two studies used QALYs gained as the main outcome. The QALY utility weights within Suwantika et al. [33] were derived from the literature, adjusted from the DALYs weights estimated by Zeng et al. [44]. The QALY utility weights used by Barbosa et al. [30] were based on the utility weights reported by Suwantika et al. [46]. These were calculated using a retrospective questionnaire administered to 144 patients in three cities (Jakarta, Bandung, and Yogyakarta), which represent regions with a high prevalence of dengue infection in Indonesia. They estimated QALY losses in outpatients, inpatients, and fatal cases to be 0.00004, 0.00018, and 1 (per year), respectively.

Cost of the intervention

A summary of the reported costs and areas covered by the intervention is presented in Table 2 and Fig 2.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. The reported cost per person covered compared to the population covered.

Labels next to each point indicate the population size (millions) for that setting. Only costs related to primary data were considered (hypothetical cost benchmarks/targets were not plotted). Uses the data related to the accelerated scenario for Brady et al. [28].

https://doi.org/10.1371/journal.pntd.0014395.g002

The evaluation of the suppression-based programme in Singapore assumed it would cost US$31.2 million per year (2024 prices) [34]. Note that this is an ongoing cost in the future, as the impacts are not self-sustaining. The cost per person covered was not directly reported; however, based on the population of Singapore (5,076,732 in 2010 [13]), it would cost approximately US$59 per person covered (2024 prices) over the 11-year intervention period examined (Table 2).

The assumed base case total cost per person covered within the evaluations of the replacement-based strategy varied between US$2.62-52.99 (2024 prices) per person covered over the course of the programme (Table 2). The majority of the costs were related to the preparation and release phases of the deployment [28,29], occurring within the first two years. A key driver of the cost was the population size/area covered (the highest cost per person was related to covering a small area/population) (Table 2). In the evaluations of larger deployments, the projected cost was often less than US$10 per person covered. In terms of the source of the cost data, some studies were based on budgets/incurred costs [25–29,31,34], whereas others were based on hypothetical cost benchmarks/targets [32,33].

In addition, Brady et al. [28] projected that conducting releases over a longer sequenced programme could reduce the total cost by 11–38% (albeit with a delay in benefits).

Cost-effectiveness/cost-benefit estimates

The status quo (i.e., continuing existing dengue control measures) was used as the comparator. Wolbachia-based biocontrol interventions were projected to generate notable economic benefits from averted costs of dengue illness (e.g., averted medical costs and prevented productivity losses). The type of economic benefits (cost savings) considered depended on the perspective (S1 File: S1 Box) [12]. Four studies [25,27,29–31] also considered cost savings related to the reduced need for existing dengue-related vector control measures, while the others assumed that these costs would remain unchanged or did not consider them.

In terms of output, rather than ICERs, some studies presented the results in terms of gross and/or net cost-effectiveness ratios:

1. Gross cost-effectiveness ratios were calculated by dividing the investment cost of the intervention by the number of DALYs averted (the cost is not incremental and therefore no cost savings are accounted).
2. Net cost-effectiveness ratios were calculated with the relevant cost savings being deducted from the investment cost of the intervention before it was divided by the number of DALYs averted. Which cost types are included in the cost savings depends on the perspective of the analysis [12,47] (S1 File: S1 Box). This would be equivalent to a traditional ICER-based output and is denoted as ICERs in the main results table (Table 3).

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Reported cost-effectiveness ratios stratified by the assumed perspective (cost year variable).

https://doi.org/10.1371/journal.pntd.0014395.t003

The cost-effectiveness ratios stratified by the different perspectives employed are summarised in Table 3 (with the corresponding economic benefits reported in S1 File: S1 Table). For this intervention, the broader the perspective, the lower the cost-effectiveness ratio. This is because the broader the perspective, the higher the averted cost of illness associated with dengue cases; with the societal perspective including the estimated monetary value of the prevented productivity losses that would have been associated with a dengue case. For the same reason, the projections based on gross cost‑effectiveness ratios (ignoring all cost offsets) appeared less favourable than ICER-based output. Differences in the chosen analytic perspective as well as the types of ratios reported, made further direct comparison across studies challenging. As a result, further quantitative synthesis of the cost‑effectiveness ratios—and any meaningful graphical presentation of these findings—was not feasible.

For the replacement strategy, the estimated societal benefit-cost ratios ranged between 0.19-4.68 (Table 3 and Fig 3). The only country setting below with a benefit-cost ratio below 1 related to Port Vila, Vanuatu, which only targeted approximately 54,000 people, and had a high corresponding cost per person covered (Fig 3). The reported economic benefits are highlighted in S1 File: S1 Table. In some cases, studies that did not disaggregate cost-effectiveness ratios by different perspectives reported disaggregated economic benefits (S1 File: S1 Table). When comparing the reported cost-effectiveness ratios to established cost-effective thresholds (such as <0.5 of the country’s per capita GDP [15,16]), it shows that overall, Wolbachia replacement-based programmes were found to be cost-effective or even often cost-saving when targeting high-burden cities (Table 3). The cost-effectiveness was not as promising for smaller-scale settings (releases covering approximately 50,000–200,000 people), and for such settings, the cost-effectiveness ratio is more likely to be above the economic threshold. This is consistent with the societal benefit-cost ratios results. This is likely because such settings have a notably higher cost per person covered (Table 2). How the costs for such settings could change if they were included within an expansion of a nearby larger programme requires further investigation.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. The projected societal benefit-cost ratios across different study settings.

Bars show the estimated societal benefit–cost ratio for each implementation setting, reflecting the extent to which projected societal benefits outweigh programme costs. The dashed horizontal line at 1.0 indicates the break‑even point where benefits equal costs.

https://doi.org/10.1371/journal.pntd.0014395.g003

When using the societal perspective, a national suppression-based strategy in Singapore was also projected to generate cost-savings, with societal benefit-cost ratios >1 [34] (Table 3 and S1 File: S1 Table). However, it had a notably higher total cost per person covered (approximately US$59 per person covered - Table 2) compared to the replacement-based interventions at a similar scale (Fig 2).

Quality assessment

Quality assessment was conducted with the CHEQUE tool [24]. Overall, the included studies demonstrated generally sound methodological and reporting standards, with average scores reflecting acceptable adherence to CHEQUE criteria (S1 File: S2 Table). Key analytical features—such as target population, time horizon, perspective, and discount rate—were typically reported clearly across studies. However, several recurring areas requiring improvement were identified (S1 File: S1 Fig). Most notably, studies frequently lacked clear documentation of model validation procedures and often provided insufficient justification for key modelling choices and assumptions; in addition, the software used to conduct the analysis was rarely stated. Furthermore, incremental effects, incremental costs, and the corresponding ICERs were often not fully reported, limiting interpretability and hindering cross-study comparison. Additional gaps included incomplete reporting of resource use and limited summaries of the broader consequences of the interventions. Finally, equity considerations and ethical implications of the economic evaluations were generally not fully considered in the studies, highlighting a persistent gap in how dengue economic evaluations incorporate potential distributional impacts/equity considerations.

Comparison to other dengue interventions

Reviews summarising the economic evaluations of other types of dengue interventions have been conducted [57–60]. These found that other types of dengue interventions can also be cost-effective. Interestingly, an overarching review of the economic evidence of Aedes-borne arboviruses found that “the current economic evidence of Aedes-borne arbovirus lacks consistency on many methodological areas” [60]. This is consistent with our results.

When comparing interventions, it is important to consider not only their cost-effectiveness ratio but also their overall population-level impact on dengue. For example, it was estimated that when using the WHO’s pre-vaccination screening recommendation, the Dengvaxia vaccine would also be cost-saving when using a societal perspective [61]. However, its overall impact in terms of the reduction in hospitalisation was much more limited than that of Wolbachia-based interventions [61].

Implementation and equity

This review focused on the estimated cost-effectiveness of Wolbachia‑based biocontrol interventions. However, real-world implementation of Wolbachia-based biocontrol interventions is shaped by operational feasibility, community engagement, regulatory pathways, and health‑system capacity. These factors are as critical as economic evidence when considering the adoption of this intervention. Country ownership, the availability of the technology from different sources, and financing arrangements require careful consideration to ensure programs can be scaled and maintained in an equitable manner.

The equity implications surrounding Wolbachia‑based biocontrol interventions are also complex. On the one hand, significantly reducing the incidence of dengue can benefit poorer populations that face high out‑of‑pocket costs associated with the disease. In addition, as an area‑based intervention that does not require any action from residents to benefit, Wolbachia-based interventions are expected to provide broadly equitable protection for all individuals living within the release zone, including across socio‑economic groups. However, while targeting the intervention to high‑burden urban areas can maximise health gains and economic benefits, concentrating deployment in cities risks widening geographic and socioeconomic inequities if rural and peri-urban populations are left without access to disease control. This highlights that comprehensive dengue control in endemic countries will require multiple complementary interventions and should not rely on a single solution.

Areas that need further research

Capturing the effectiveness and uncertainty regarding the long-term impact

At this time, there are limited data related to the effectiveness of suppression-based strategies, with the only evaluation assuming a wide range of effectiveness between 40–80%. There is a need for further effectiveness data, particularly outside of Singapore. It is important to note that historically, Singapore has implemented one of the world’s most intensive vector control programmes, with sustained efforts that drove Aedes populations to levels [72]. While initially highly effective in suppressing transmission and dengue incidence, this also resulted in a progressively larger pool of susceptible individuals, which has been identified as a key driver of the scale of subsequent outbreaks when new serotypes or substantial viral introductions occur and requiring increasing amounts of control measures to keep transmission at the same level [72,73]. These epidemiological implications of these dynamics [73] warrant greater consideration when evaluating the projected long-term impact of Wolbachia‑based interventions.

Several of the evaluations of the replacement-based strategy based the assumed effectiveness of Wolbachia deployments on the results of a cluster randomised trial and quasi-experimental studies in Yogyakarta (with the wMel strain) [39,40]. There are at least three factors that can result in an underestimated efficacy from this type of trial: human movement, mosquito movement, and coupled transmission dynamics between trial arms [74]. Regarding the latter, Cavany et al. [74] highlighted the importance of transmission dynamic modelling in designing and interpreting future trials. A reanalysis of the Yogyakarta trial using spatiotemporally resolved data on the distribution of Wolbachia mosquitoes and the mobility of participants estimated an increased intervention efficacy to >80% [75]. In Colombia, following pilot releases in 2015–2016, staged city-wide Wolbachia deployments were undertaken in the cities of Bello, Medellín, and Itagüí between October 2016 and April 2022 [76,77]. A quasi-experimental study using interrupted time series analysis showed that notified dengue case incidence in the three cities declined by 95–97% compared to the prior decade [77].

That said, it is important to note that the successful dispersal of Wolbachia-infected mosquitoes can be heterogeneous and influenced by local environmental factors [28,78]. The effectiveness in other settings may therefore be lower than that found in the Yogyakarta study trial, and this potential variation requires further investigation. For example, an analysis of pilot releases of wMel-infected mosquitoes in Niterói, Brazil, estimated a 69% reduction in dengue incidence (95% confidence interval (CI): 54–79%) [79]. In contrast, a spatiotemporal modelling study of a release programme in Rio de Janeiro, Brazil, reported only a 38% (95% CI: 32–44%) reduction in dengue incidence [80]. In addition, a study utilising the wAlbB strain in Malaysia estimated a 40.3% (95% CI: 5.1–64.6%) reduction in dengue cases [4]. It should be noted that there is likely a gap between early measures of efficacy found within trials/pilot interventions and the real-world effectiveness of large-scale programmes. Therefore, it is important for future studies to explore a range of effectiveness values lower than the assumed baseline. It will be important that future economic evaluations in this area incorporate updated field data and become less reliant on long term projections.

A further source of uncertainty is the duration of the impact of the replacement-based strategy (which was a key driver of the projected cost-effectiveness). While field evidence from northern Australia demonstrates that wMel-infected mosquitoes remain effective more than a decade after their initial release, indicating long-term stability of Wolbachia in Ae. aegypti populations [81], more data are needed from a range of epidemiological settings. In addition, several factors could potentially reduce the long-term impact of Wolbachia-based interventions [28]. These include reinvasion by Wolbachia uninfected mosquitoes, evolution of viral resistance, temperature effects on viral blocking efficacy and inheritability, and selection of more virulent dengue virus strains. These aspects require further investigation.

Finally, it is important to consider that climate change is likely to expand the geographical distribution and transmission intensity of several vector-borne human infectious diseases, including dengue. This could potentially increase the baseline burden that could be averted by Wolbachia interventions, increasing its public health impact. On the other hand, it is also possible that climate change will have an impact on the long-term effectiveness of Wolbachia-based interventions [82–84]. In an analysis of this area, Vásquez et al. [84] concluded that this technology is generally robust to near-term (2030s) climate change. However, accelerated warming may challenge this in the 2050s and beyond.

The cost of the intervention

Currently, there are limited published costing data related to the use of Wolbachia interventions. Many current primary cost estimates are based on mid-sized project sites. There is potential for the costs associated with Wolbachia deployments to be reduced over time through advances in mass mosquito production, economies of scale, and alternative implementation models [29]. The actual costs incurred for large-scale deployments require further investigation. An important area for further research is understanding the marginal cost of expanding existing programmes.

Brady et al. [28] highlighted how the costs of Wolbachia deployments show a relatively complex relationship with cost per person varying by population density in the release area and the scale of deployment. They found that because areas with higher human density require more mosquito release numbers per unit area (as they typically have higher natural mosquito population sizes), the cost per km² covered increases as the human density increases. However, despite this, the cost per person covered will decrease because more people are covered in these high-density urban areas – which reduces the cost per person covered due to economies of scale. These relationships must be considered when projecting the cost of this intervention. There is a need to develop costing models to better account for how costs change as the intervention is scaled up within countries. These complexities in cost scaling (by both area and population covered) are a distinctive challenge associated with environmental interventions, leading to cost-effectiveness projections that are more highly context dependent.

There is also an ongoing need for the development/implementation of methods that reduce the cost of Wolbachia-based interventions. This is highlighted by Tiley et al. [8], who estimated that for a replacement-based strategy to be deployable in enough areas to make major contributions to reducing the global dengue burden by 25% (in line with 2030 WHO targets), the cost must ultimately be reduced to between US$0.24-7.63 (2020 prices) per person protected, in order to be cost-neutral from a health sector perspective. Hollingsworth et al. [85] recently developed a stochastic dynamic programming framework for determining optimal release schedules for Wolbachia-transinfected mosquitoes that balances the cost of dengue infection with the costs of rearing and releasing transinfected mosquitoes. Such an approach could help optimize programmes – reducing its cost. Novel ways to conduct the mosquito releases also need to be considered.

Broader currently unquantified benefits

As well as the investigated averted cost of illness associated with controlling dengue, there are other potential benefits that are not being quantified. For example, dengue can lead to a loss of tourism revenue [86]. In addition, dengue outbreaks can cause congestion in intensive care wards, potentially having negative consequences on the care of patients with other conditions due to the deterioration of overall service quality. The benefits of Wolbachia-based interventions in reducing this have not been accounted for.

The current economic evaluations have only investigated the public health impact of Wolbachia-based interventions against dengue. However, Wolbachia infected mosquitoes are also refractory to the Zika virus, Chikungunya virus and, Yellow Fever virus [87,88]. This means that the overall public health impact of Wolbachia-based interventions will be larger, increasing its cost-effectiveness/value for money. The extent of this will vary across different settings and the endemicity of these diseases.

Further evaluation of Wolbachia-based strategies and the combination with other interventions

We found nine studies evaluating Wolbachia-based interventions, eight of which related to a replacement-based strategy (mostly based on data related to the wMel Wolbachia strain). That said, there remains a need to evaluate Wolbachia replacement-based programmes in other settings as well as the use of other strains/distribution methods as they become available. There is also a need to further evaluate the cost-effectiveness of suppression-based strategies. It is also important to consider combinations of Wolbachia interventions. A key difference from both an operational and cost perspective is that suppression‑based Wolbachia strategies require ongoing releases, whereas replacement‑based strategies do not. This fundamental distinction has major implications for feasibility in LMIC settings. Implementing suppression at the same geographic scale as replacement would likely be substantially more costly. However, suppression approaches may still be suitable for targeted, localised hotspots where sustained releases can be focused on smaller areas with high transmission intensity. Further analysis is needed to understand how these methods perform operationally at different spatial scales and what cost structures emerge under real‑world implementation conditions.

Further evaluation of the use of Wolbachia-based interventions in combination with other preventive interventions is needed. This is particularly important as new vaccines become available [89]. In this context, it is important to consider that although Wolbachia-based interventions can be highly cost-effective, it does not mean that it is possible to use them everywhere across endemic areas, particularly in more rural areas. The different cost structures of Wolbachia-based interventions and vaccination become highly relevant here. Wolbachia programme costs depend primarily on the geographic area that must be covered, whereas vaccination costs scale with the number of individuals vaccinated. These structural differences have notable implications for where each intervention is likely to be most cost-effective. There is a need to investigate the combination of Wolbachia-based interventions and vaccination more comprehensively, considering the use of different interventions in different areas within countries (rural vs. urban areas).

Limitations

A potential source of bias of the search strategy is that it would not capture studies published outside of the searched electronic databases (i.e., grey literature such as policy documents/reports, and non-English language publications etc.). Efforts were made to minimise this bias by searching the bibliographies of selected studies, and searching the grey literature as outlined in the S1 File. There could also be a degree of publication bias, with economic evaluations of this intervention with unfavourable results being less likely to be published

To provide overarching, high‑level conclusions, cost‑effectiveness ratios were compared against a threshold of 0.5 times a country’s per‑capita GDP. It is important to acknowledge that this type of standardised benchmark has notable limitations, particularly when applied across diverse settings. Ideally, policymakers should draw on locally appropriate thresholds that better reflect country‑specific opportunity costs, and/or budget constraints when making decisions.

The costs were adjusted for inflation using US inflation rates. It was not possible to apply a more precise inflation adjustment that accounted for the proportion of tradable versus non‑tradable (local) resources within each cost estimate. It should also be noted that the technology and approach used to implement Wolbachia-based interventions are advancing over time. Due to this, the older studies captured by this review are less likely to be representative of more recent/future programmes, particularly with regard to the intervention costs and the final product method used for Wolbachia deployment.

Conclusion

Based on the currently available evidence, there is consensus that Wolbachia-based replacement interventions can be cost-effective for dengue control, when they are appropriately targeted, particularly in high-density urban environments (such as cities). For such high burden settings, the estimated net cost-effectiveness ratios in terms of the cost per DALY averted from the health sector perspective were typically below 0.5 the per capita GDP cost-effectiveness threshold. When taking the societal perspective and including the monetary value of the productivity gains associated with the averted cases, the economic benefits of a Wolbachia-based replacement intervention often outweigh its cost. A suppression-based strategy was also found to be cost-saving in Singapore from a societal perspective, however it had a notably higher cost per person covered compared to the replacement-based interventions.

That said, in settings where the baseline burden is not sufficiently high, this type of intervention is unlikely to be cost-effective. The exact conditions which determine which areas should be targeted will be setting specific. It is important to note that the successful dispersal of Wolbachia-infected mosquitoes can be heterogeneous and influenced by local environmental factors. Further studies are needed to investigate and quantify how this heterogeneity affects long-term projections of the cost-effectiveness of Wolbachia-based interventions across a wider range of operational and epidemiological settings.

This review focused on the cost‑effectiveness and value for money of Wolbachia‑based biocontrol interventions, synthesising the estimates currently available. While these economic considerations are important, they represent only one part of the decision-making landscape. Policymakers/stakeholders must also weigh a broader set of factors—including governance arrangements, operational logistics, community acceptance, and regulatory requirements—when assessing whether and how to implement these interventions. In addition, it is important to consider that there will be no single solution to controlling dengue, and it remains vital to consider/evaluate other interventions (such as new vaccines, other novel vector control methods, therapeutics, etc, as they become available).