Cost-Effectiveness of Population-Level Fluoride Policies: A Systematic Review of Health and Economic Outcomes Authors: Diaa Almutairi, Moodhi Al Humaidi, Saeed Mastour Alshahrani, Awatef Albalawi Published: 2026 ABSTRACT National Health Policies have adopted fluoridation-based programmes for the prevention of dental caries in many communities. Assessing their economic value across different settings is important for informing health policy and optimising resource allocation. To synthesise evidence of the economic efficiency of population-level public health interventions that use fluoride to prevent dental caries, compared with no fluoride policy or with alternative dental preventive interventions. Five electronic databases (PubMed/Medline, Cochrane Library, Embase, Global Health and EconLit) were searched for English-published studies on economic evaluations conducted in any country or population, without time restrictions. Eligible studies compared fluoridated water with no fluoridation, fluoridation cessation or alternative fluoride delivery strategies. Data were extracted on study design, perspective, time horizon, cost components, outcome measures and economic results. Given heterogeneity in settings, methods and outcomes, qualitative synthesis was conducted without pooling of the estimates. The review included 15 economic evaluations from high- and middle-income settings. Most studies reported that fluoridation was cost-effective or cost-saving compared with no fluoridation. More favourable economic results were generally observed in larger populations, those with a higher baseline caries risk and in studies focused on children. The evidence suggests cost-effectiveness and favourable budget impact of public fluoridation programmes compared with no fluoridation across diverse settings. However, the magnitude of economic benefit likely varies by population characteristics, study design and analytic assumptions. Therefore, careful interpretation is warranted when applying these findings to policy decisions. These findings provide clinicians and policymakers with robust economic evidence to support the continuation or implementation of fluoridation programmes as a primary preventive measure against dental caries. Introduction Dental caries ranks as the most common noncommunicable dental health problem worldwide.1,2 Its pathophysiology involves the demineralisation of the tooth enamel caused by acid-producing bacteria.3,4 The condition is associated with pain and reduced quality of life.5,6 The impact of dental caries is not restricted to inducing negative health outcomes but extends to adversely affect economic perspectives.7 Prevention of caries could be achieved through the combined implementation of protective measures at individual, professional and community levels.8,9 Higher risk of caries has been associated with lower socioeconomic status and socially disadvantaged communities.10,11 To achieve equity and enhance oral health for all social levels, effective preventive interventions on the community or national scale should be implemented. Fluoride is currently the main effective public health noninvasive strategy for the prevention of dental caries.9 The anticariogenic effects of fluoride are exerted through promoting the remineralisation of early carious lesions, inhibiting the demineralisation of healthy enamel and interfering with bacterial metabolism.12 Fluoride can be administered systemically through water, milk or salt. In addition, topical administration can be achieved through the use of toothpastes, varnishes, gels and mouthwashes,13 as well as fluoride-releasing dental sealants and restorative materials.14,15 Community-based fluoride interventions entail the addition of fluoride to products that are widely consumed, such as milk, salt and drinking water.9 However, adverse effects can arise from fluoride exposure exceeding the recommended thresholds, such as dental and skeletal fluorosis, as well as neurotoxicity.16,17 While professional consensus recommends tooth brushing with fluoridated toothpaste, several challenges hinder the achievement of optimal population-level oral hygiene. These include variability in clinical guidance, reliance on user technique and the affordability of fluoridated toothpastes and powered toothbrushes.18 Passive population-level measures such as community water fluoridation (CWF) can play an important role in overcoming these challenges. Although several concerns have been raised over the safety of fluoride, including claims that CWF leads to uncontrollable fluoride intake, public health recommendations in most countries mandate a wide safety margin far below toxicity thresholds.19 Besides the effectiveness and safety of interventions, other important factors should be considered before deciding on their implementation as a public health policy. These factors include cost-effectiveness and the impacts of the intervention from political, social, ethical and legal perspectives must be considered.20 Economic evaluations (EE) are required to assess the effective use of health care resources.21 The process of EE includes the comparative analysis of the costs and outcomes of 2 alternative strategies.22 EE can be full or partial when only either the costs or the outcomes are evaluated.23 The types of full EE were divided into cost-effectiveness analysis (CEA), cost-utility analysis and cost-benefit analysis (CBA). In a CEA, costs are expressed in monetary values while effects are presented in natural units. A cost-utility analysis is a special type of CEA where the costs are expressed in monetary units and effects in healthy years (eg, quality-adjusted life years [QALYs]). In a CBA, the benefits of healthcare are expressed as the amount of money that an individual is willing to pay/receive for the benefits/disadvantages offered, with both costs and effects expressed in monetary terms.21,24 Full EE is more informative than partial EE, as the latter provides no information on technical and allocative efficiency, while the former compares both factors and provides incremental analyses.23,22 In the context of oral health, the overall use of EE was limited, with a paucity of reviews on the cost-effectiveness of dental interventions. A rigorous synthesis is therefore required to consolidate the available evidence regarding the cost-effectiveness of fluoride programmes to protect against dental caries. The synthesised evidence can inform evidence-based policy decisions on the allocation of resources for population-level caries prevention. Therefore, the present systematic review was conducted in order to synthesise evidence of the economic efficiency of population-level public health interventions that use fluoride to prevent dental caries, compared with no fluoride policy or with alternative dental preventive interventions. Methods Methodology The study protocol was registered at the International Prospective Register of Systematic Reviews (registration number: CRD420251081356). This systematic review followed the Preferred Reporting Items for Systematic Reviews and Meta-analyses guidelines.25 Eligibility criteria Inclusion criteria Studies included were based on population, intervention, control, outcomes and study design elements described below:1.Population: Population of all ages who received or were under a national or local fluoride intervention in any country.2.Intervention: The interventions were any national or local fluoridation intervention, including water fluoridation, salt fluoridation, milk fluoridation and mandated fluoride supplement policy.3.Comparison: The control group for comparison was communities or populations with no fluoride policy (no fluoride-based public health intervention) or those implementing alternative, nonfluoride dental caries prevention strategies.4.Outcomes: Primary outcomes included incremental cost-effectiveness ratios (ICERs), cost per QALY gained, cost per dental caries prevented, changes in the Decayed, Missing and Filled Teeth (DMFT) index and net healthcare cost savings. Secondary outcomes included health equity outcomes, such as differential effects of fluoridation across population subgroups defined by socioeconomic, demographic or geographic characteristics; and long-term economic implications, including lifetime costs and savings and budget impact over extended time horizons.5.Setting: National or community settings in which population-wide fluoride policies were implemented. Studies were required to report economic outcomes (with or without associated health outcomes) using full or partial EE methods at the policy or population levels.6.Study designs: The eligible study designs were model-based or trial-based full or partial EEs. 1.Population: Population of all ages who received or were under a national or local fluoride intervention in any country. Population: Population of all ages who received or were under a national or local fluoride intervention in any country. 2.Intervention: The interventions were any national or local fluoridation intervention, including water fluoridation, salt fluoridation, milk fluoridation and mandated fluoride supplement policy. Intervention: The interventions were any national or local fluoridation intervention, including water fluoridation, salt fluoridation, milk fluoridation and mandated fluoride supplement policy. 3.Comparison: The control group for comparison was communities or populations with no fluoride policy (no fluoride-based public health intervention) or those implementing alternative, nonfluoride dental caries prevention strategies. Comparison: The control group for comparison was communities or populations with no fluoride policy (no fluoride-based public health intervention) or those implementing alternative, nonfluoride dental caries prevention strategies. 4.Outcomes: Primary outcomes included incremental cost-effectiveness ratios (ICERs), cost per QALY gained, cost per dental caries prevented, changes in the Decayed, Missing and Filled Teeth (DMFT) index and net healthcare cost savings. Secondary outcomes included health equity outcomes, such as differential effects of fluoridation across population subgroups defined by socioeconomic, demographic or geographic characteristics; and long-term economic implications, including lifetime costs and savings and budget impact over extended time horizons. Outcomes: Primary outcomes included incremental cost-effectiveness ratios (ICERs), cost per QALY gained, cost per dental caries prevented, changes in the Decayed, Missing and Filled Teeth (DMFT) index and net healthcare cost savings. Secondary outcomes included health equity outcomes, such as differential effects of fluoridation across population subgroups defined by socioeconomic, demographic or geographic characteristics; and long-term economic implications, including lifetime costs and savings and budget impact over extended time horizons. 5.Setting: National or community settings in which population-wide fluoride policies were implemented. Studies were required to report economic outcomes (with or without associated health outcomes) using full or partial EE methods at the policy or population levels. Setting: National or community settings in which population-wide fluoride policies were implemented. Studies were required to report economic outcomes (with or without associated health outcomes) using full or partial EE methods at the policy or population levels. 6.Study designs: The eligible study designs were model-based or trial-based full or partial EEs. Study designs: The eligible study designs were model-based or trial-based full or partial EEs. Exclusion criteria The following articles were excluded: studies that reported on fluoridation programmes without a comparator, case reports, qualitative studies, reviews, editorials, commentaries, letters, replies, conference abstracts, articles not published in English and articles lacking the outcomes of interest. Search strategy The search strategy included terms that were derived from the population, intervention, control, outcomes and study design elements. A comprehensive search was conducted to identify potential studies using the following databases: PubMed/Medline, Cochrane Library, Embase, Global Health and EconLit. There were no date restrictions; the search was carried out up to 5 September 2025. However, language was restricted to papers with English abstracts. The search string for PubMed/Medline was (("fluoridation"[MeSH Terms] OR "fluoridation"[All Fields] OR ("water"[All Fields] AND "fluoridation"[All Fields]) OR "water fluoridation"[All Fields]) OR ("fluoridation"[MeSH Terms] OR "fluoridation"[All Fields] OR ("salt"[All Fields] AND "fluoridation"[All Fields]) OR "salt fluoridation"[All Fields]) OR ("fluoridation"[MeSH Terms] OR "fluoridation"[All Fields] OR ("milk"[All Fields] AND "fluoridation"[All Fields]) OR "milk fluoridation"[All Fields])) AND ("economics"[MeSH Subheading] OR "economics"[All Fields] OR "cost"[All Fields] OR "costs and cost analysis"[MeSH Terms] OR ("costs"[All Fields] AND "cost"[All Fields] AND "analysis"[All Fields]) OR "costs and cost analysis"[All Fields]). Study selection Two reviewers (DA and MH) independently reviewed the titles and abstracts based on the inclusion criteria and a third reviewer (SA) resolved any disagreements. Subsequently, the full texts of studies included after the title and abstract screening were obtained and reviewed independently by the 2 reviewers (DA and MH). The references cited in the reviewed articles were checked to identify any further potential studies. Data extraction Data extraction was performed independently by 2 reviewers, using a predefined data extraction form. Extracted information included the study characteristics (setting, design, type of EE, time horizon for EE, population, eligibility criteria and completion rate), baseline characteristics of the population (age, ethnicity, socioeconomic characteristics and geographic region), details of the intervention, outcomes and information required for risk of bias assessment. The time horizon was defined as the period over which costs and outcomes are observed or modelled in an EE. Any discrepancies were resolved through discussion, with involvement of the third reviewer when necessary. Quality appraisal and risk of bias assessment The 2 examiners independently appraised the methodological quality and risk of bias (ROB) for the included studies using the Joanna Briggs Institute Checklist for EE.26 Methodological quality was graded as proposed by Dos Santos et al.27 as follows: fair (<40% of the items presented), moderate (41%-80% of the items presented) and good (>80% of the items presented). Data presentation and synthesis As marked heterogeneity was observed in estimates of cost-effectiveness, arising from variations in the populations’ characteristics, study design and outcomes, pooling of the data was not performed. Qualitative synthesis was carried out following the synthesis without meta-analysis guidelines.28 Studies were grouped by the type of EE as well as the implemented fluoride intervention, the comparator (no policy/alternative intervention) and the country income based on the World Bank classification.29 Economic thresholds were defined by categorising studies by ICER into dominant and cost-effective interventions. An intervention is judged to be dominant when it is both more effective and results in lower total costs compared to the alternative.30 Cost-effective interventions have a net cost but provide significant health value at thresholds that were defined by national health benchmarks.30 Subgroup and sensitivity analyses were carried out to explore heterogeneity and test the robustness of results. Subgroups were created based on the type of EE (full or partial; CEA or CBA; empirical or modelled; age groups). Results Literature search and study selection The search of online databases yielded 2149 records. A total of 1724 duplicates were removed, leaving 425 records for screening of their titles and abstracts. We then excluded 361 noneligible records. We retrieved the full texts of the remaining 64 records and assessed them against the eligibility criteria for this systematic review. Forty-nine records were excluded. Fifteen records were found eligible and were included in this study (Figure 1).31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45Fig. 1PRISMA flow diagram of the study selection process.Fig 1 dummy alt text PRISMA flow diagram of the study selection process. Study characteristics The included studies were conducted across diverse geographical settings. Five studies were conducted in the United States (USA),31,32,39,40,45 3 in Australia,34,35,42 2 in England43,44 and 1 in each of Brazil,33 Ireland,41 New Zealand,38 Peru36 and South Africa.37 Several methodological approaches were used, including model-based full EE31,34, 35, 36, 37, 38,41,45, observational full EE,42, 43, 44 observational partial EE32,39,40 and ecological analyses.33 Most studies (n = 9) were conducted at the community level,31, 32, 33, 34, 35, 36,39,42,44 whereas the remaining studies adopted a national-level perspective (Table 1).37,38,40,41,43,45Table 1General characteristics of included studies.Table 1 dummy alt textStudyCountrySetting/analytic perspectiveStudy design/economic frameworkIntervention (I) and comparator (C)Evidence sourceDiscount rate/price y/currencyCost categoriesTarget populationTime horizonCampain et al.34AustraliaCommunity/SocietalModel-based CBAI: CWFC: no fluoridationModelling7%, 2005, AUDDirect treatment, direct societal and indirect societalWhole population (ages 6-66+)LifetimeChoi and Simon45United StatesNational/HealthcareModel-based CEAI: CWF cessationC: status quo (CWF)Modelling3%, 2024, USDDirect treatment and direct societalChildren (0-19 y)5-10 yCiketic et al.35AustraliaCommunity/HealthcareModel-based CEAI: CWFC: no fluoridationModelling3%, 2003, AUDDirect treatment and direct societalWhole population (modelled 2-100 y)LifetimeCronin et al.41IrelandNational/HealthcareModel-based CEAI: CWFC: no fluoridationModelling4%, 2017, EURDirect treatment, direct societal and indirect societalSchool children (5, 8 and 12 y)LifetimeFyfe et al.38New ZealandNational/SocietalModel-based CEAI: CWFC: no fluoridationModelling3.5%, 2011/12, NZDDirect treatment, direct societal and indirect societalWhole population and children (<18 y)15 yGabardo et al.33BrazilCommunity/NREcological observational CBAI: CWFC: no fluoridationEmpiricalNR, 2002, USDDirect societalSchool children (12 y)NRGriffin et al.40United StatesNational/HealthcareObservational partial EE (cost analysis)I: dietary F supplementsC: CWFEmpiricalNR, 2011, USDDirect societalChildren (0.5-16 y)1 yKroon and Van Wyk37South AfricaNational/HealthcareModel-based CEA and CBAI: CWFC: no fluoridationModellingNR, 2006/2011, USDDirect treatment and direct societalWhole population (children <15 focus)4 yMariño et al.36PeruCommunity/SocietalModel-based CEAI: Salt fluoridationC: no interventionModelling3%, 2009, PENDirect treatment, direct societal and indirect societalSchool children (6-12 y)6 yMaupomé et al.32United StatesCommunity/HealthcareObservational partial EE (cost analysis)I: CWFC: no fluoridationEmpiricalNR, 1995, USDDirect treatment costs onlyWhole population (HMO members ages 5-98)6 yMeyer et al.39United StatesCommunity/HealthcareObservational partial EE (cost analysis)I: CWF cessationC: precessation CWFEmpiricalNR, 2012, USDDirect treatment costs onlyChildren/adolescents (0-18 y)9 yMoore et al.43EnglandNational/HealthcareObservational CEAI: CWFC: no fluoridationEmpirical1.5% (base); 3.5% (sensitivity), 2020, GBPDirect treatment and direct societalWhole population (adults/adolescents 12+)10 yO’Connell et al.31United StatesCommunity/SocietalModel-based CBAI: CWFC: no fluoridationModelling3%, 2003, USDDirect treatment, direct societal and indirect societalWhole population (aged 5+)LifetimeWhittaker et al.44EnglandCommunity/HealthcareObservational CEAI: CWFC: no fluoridationEmpirical3.5%, 2014, GBPDirect treatment and direct societalSchool children (birth to age 11)5-6 yZhao et al.42AustraliaCommunity/HealthcareObservational CBAI: CWFC: no fluoridationEmpiricalNR (uses 2.1%-3.4% inflation adjustments), 2021, AUDDirect treatment, direct societal and indirect societalWhole population (aboriginal communities)20 yCost categories: direct treatment = dental/healthcare costs; direct societal = intervention and programme costs; indirect societal = productivity, patient time or travel costs.AUD, Australian dollar; C, comparator; CBA, cost-benefit analysis; CEA, cost-effectiveness analysis; CUA, cost-utility analysis; CWF, community water fluoridation; EE, economic evaluation; EUR, euro; GBP, British pound sterling; HMO, health maintenance organisation; I, intervention; NR, not reported; NZD, New Zealand dollar; PEN, Peruvian sol; QALY, quality-adjusted life year; USD, United States dollar. Table 1General characteristics of included studies.Table 1 dummy alt textStudyCountrySetting/analytic perspectiveStudy design/economic frameworkIntervention (I) and comparator (C)Evidence sourceDiscount rate/price y/currencyCost categoriesTarget populationTime horizonCampain et al.34AustraliaCommunity/SocietalModel-based CBAI: CWFC: no fluoridationModelling7%, 2005, AUDDirect treatment, direct societal and indirect societalWhole population (ages 6-66+)LifetimeChoi and Simon45United StatesNational/HealthcareModel-based CEAI: CWF cessationC: status quo (CWF)Modelling3%, 2024, USDDirect treatment and direct societalChildren (0-19 y)5-10 yCiketic et al.35AustraliaCommunity/HealthcareModel-based CEAI: CWFC: no fluoridationModelling3%, 2003, AUDDirect treatment and direct societalWhole population (modelled 2-100 y)LifetimeCronin et al.41IrelandNational/HealthcareModel-based CEAI: CWFC: no fluoridationModelling4%, 2017, EURDirect treatment, direct societal and indirect societalSchool children (5, 8 and 12 y)LifetimeFyfe et al.38New ZealandNational/SocietalModel-based CEAI: CWFC: no fluoridationModelling3.5%, 2011/12, NZDDirect treatment, direct societal and indirect societalWhole population and children (<18 y)15 yGabardo et al.33BrazilCommunity/NREcological observational CBAI: CWFC: no fluoridationEmpiricalNR, 2002, USDDirect societalSchool children (12 y)NRGriffin et al.40United StatesNational/HealthcareObservational partial EE (cost analysis)I: dietary F supplementsC: CWFEmpiricalNR, 2011, USDDirect societalChildren (0.5-16 y)1 yKroon and Van Wyk37South AfricaNational/HealthcareModel-based CEA and CBAI: CWFC: no fluoridationModellingNR, 2006/2011, USDDirect treatment and direct societalWhole population (children <15 focus)4 yMariño et al.36PeruCommunity/SocietalModel-based CEAI: Salt fluoridationC: no interventionModelling3%, 2009, PENDirect treatment, direct societal and indirect societalSchool children (6-12 y)6 yMaupomé et al.32United StatesCommunity/HealthcareObservational partial EE (cost analysis)I: CWFC: no fluoridationEmpiricalNR, 1995, USDDirect treatment costs onlyWhole population (HMO members ages 5-98)6 yMeyer et al.39United StatesCommunity/HealthcareObservational partial EE (cost analysis)I: CWF cessationC: precessation CWFEmpiricalNR, 2012, USDDirect treatment costs onlyChildren/adolescents (0-18 y)9 yMoore et al.43EnglandNational/HealthcareObservational CEAI: CWFC: no fluoridationEmpirical1.5% (base); 3.5% (sensitivity), 2020, GBPDirect treatment and direct societalWhole population (adults/adolescents 12+)10 yO’Connell et al.31United StatesCommunity/SocietalModel-based CBAI: CWFC: no fluoridationModelling3%, 2003, USDDirect treatment, direct societal and indirect societalWhole population (aged 5+)LifetimeWhittaker et al.44EnglandCommunity/HealthcareObservational CEAI: CWFC: no fluoridationEmpirical3.5%, 2014, GBPDirect treatment and direct societalSchool children (birth to age 11)5-6 yZhao et al.42AustraliaCommunity/HealthcareObservational CBAI: CWFC: no fluoridationEmpiricalNR (uses 2.1%-3.4% inflation adjustments), 2021, AUDDirect treatment, direct societal and indirect societalWhole population (aboriginal communities)20 yCost categories: direct treatment = dental/healthcare costs; direct societal = intervention and programme costs; indirect societal = productivity, patient time or travel costs.AUD, Australian dollar; C, comparator; CBA, cost-benefit analysis; CEA, cost-effectiveness analysis; CUA, cost-utility analysis; CWF, community water fluoridation; EE, economic evaluation; EUR, euro; GBP, British pound sterling; HMO, health maintenance organisation; I, intervention; NR, not reported; NZD, New Zealand dollar; PEN, Peruvian sol; QALY, quality-adjusted life year; USD, United States dollar. General characteristics of included studies. Cost categories: direct treatment = dental/healthcare costs; direct societal = intervention and programme costs; indirect societal = productivity, patient time or travel costs. AUD, Australian dollar; C, comparator; CBA, cost-benefit analysis; CEA, cost-effectiveness analysis; CUA, cost-utility analysis; CWF, community water fluoridation; EE, economic evaluation; EUR, euro; GBP, British pound sterling; HMO, health maintenance organisation; I, intervention; NR, not reported; NZD, New Zealand dollar; PEN, Peruvian sol; QALY, quality-adjusted life year; USD, United States dollar. Most studies compared CWF with no fluoridation, while a smaller number of studies evaluated fluoridation cessation39,45 or alternative fluoride delivery strategies.36,40 The analytic perspective was predominantly that of the healthcare system or health payer, while a societal perspective was adopted in 4 studies.31,34,36,38 Direct dental treatment costs were consistently included, while broader societal and indirect costs, such as productivity losses or patient travel, were incorporated less consistently and more commonly in model-based analyses (Table 1).31,34,36,38,41 Target populations ranged from whole-population analyses31,32,34,35,37,38,42,43 to child-focused cohorts.33,36,39, 40, 41,44,45 Time horizons widely varied across the studies, ranging from short-term analyses to lifetime modeling (Table 1). Risk of bias Overall, the methodological quality of the included economic evaluations was moderate to high, with clear differences according to study design (Figure 2). Full EEs showed low ROB across most domains. Bias was minimised through the conduction of incremental analyses, the use of validated effectiveness inputs, discounting of future costs and outcomes and extensive deterministic or probabilistic sensitivity analyses.31,34, 35, 36, 37, 38,41, 42, 43, 44, 45Fig. 2Assessment of methodological quality and risk of bias using the JBI checklist for economic evaluations. Yes = low risk of bias for domain; no = high risk of bias for domain; unclear = insufficient information.Fig 2 dummy alt text Assessment of methodological quality and risk of bias using the JBI checklist for economic evaluations. Yes = low risk of bias for domain; no = high risk of bias for domain; unclear = insufficient information. Meanwhile, observational and ecological studies showed higher or unclear ROB in several key domains, primarily due to design-related limitations. The most frequently unmet domains were incremental analysis (Q8) and investigation of uncertainty (Q9).32,33,39,40 Also, adjusting for differential timing (Q7) was unmet in 2 studies,33,40 while identification of all relevant costs and outcomes for alternatives (Q3) showed high risk in 3 studies.33,39,40 Overall, methodological quality was good (score 81%-100%) for all studies except for 3 moderate-quality studies (scores 45.5%-72.7%).33,39,40 Qualitative synthesis The studies included in this review used a spectrum of outcome metrics that can be categorised into 3 relevant outcomes: clinical indices, health system utilisation and health utility. Clinical indices measure the dental caries burden, counting the sum of decayed, missing and filled teeth. These indicators included dmft (for primary teeth), DMFT (for permanent teeth) or d3vcmft/D3vcMFT (decayed, missing or filled teeth recorded at the dentine level, with or without cavitation). Health system utilisation is concerned with averted procedures, which are inherently correlated with clinical indices. Health utility measures population-level health impacts, which can be assessed using several metrics. Disability-Adjusted Life Years (DALYs) quantify the years of life lived with caries as a disability, while QALYs capture the reduced well-being associated with caries (eg, pain, abscesses or tooth loss). Despite the heterogeneity of these endpoints, they demonstrate a high degree of directional consistency: a reduction in clinical caries indices leads to a reduction in invasive procedures, which ultimately manifests as a gain in health utility (QALYs) or a reduction in disease burden (DALYs). As EEs showed substantial heterogeneity in economic frameworks, outcome metrics, analytic perspectives and time horizons across studies, quantitative synthesis was not done. Studies were synthesised narratively based on the direction and consistency of economic conclusions (cost-saving, cost-effective or not cost-effective), rather than on numerical estimates. The synthesis comprised 4 analytic domains: economic efficiency, clinical effectiveness, budget/fiscal impact and equity. Economic efficiency The economic efficiency outcome was synthesised using 8 studies reporting ICERs (Table 2). The 8 studies had good overall methodological quality. Analyses of cost-effectiveness (CEA) reported ICERs per dental outcome (ie, DMFT/dmft prevented and invasive dental treatments avoided) or per DALY.Table 2Incremental economic efficiency outcomes (ICER-based evidence only).Table 2 dummy alt textStudyEconomic frameworkCountry income levelOutcome metricOverall methodological qualityResultsEconomic conclusionChoi and Simon45CEA, model-basedHighIncremental costs and QALYs (cessation vs continuation)GoodCessation: −2.9 million QALYs; cost is +$9.8 billion over 5 yContinuation of fluoridation dominates cessationCiketic et al.35CEA, model-basedHighICER per DALYGoodICER: $3608 per DALY without treatment savings. ICER: −$64,127 per DALY when dental treatment cost offsets were includedDominant when treatment savings are included; cost-effective otherwiseCronin et al.41CEA, model-basedHighICER per d₃vcmft/D₃vcMFT preventedGoodICER: €14.09 per d3vcmft/D3vcMFT without treatment savings. ICER: −€115.67 per d3vcmft/D3vcMFT when dental treatment cost was includedDominant when treatment savings are included; cost-effective otherwiseFyfe et al.38CEA, model-basedHighNet cost per dmft/DMFT avertedGoodICER: −$0.55 to −$6.09 per dmft/DMFT averted, depending on population size and age groupDominantKroon and Van Wyk37CEA, model-basedUpper middleCost per DMFT savedGoodICER: US$3.32-US$11.08Cost-effectiveMariño et al.36CEA, model-basedUpper middleICER per DMFT avoidedGoodICER: S/. 10.73 to S/. 14.35 saving per DMFT avoided, depending on the discount rate and effectiveness scenariosDominantMoore et al.43CEA, empiricalHighICER per invasive dental treatment avoidedGoodICER: £94.55 per invasive dental treatment avoidedCost-effectiveWhittaker et al.44CEA, empiricalHighICER per QALY gainedGoodICER: £6245-£2973 per QALY gainedCost-effectiveDominant indicates lower costs and greater health benefits compared with the comparator.CEA, cost-effectiveness analysis; CUA, cost-utility analysis; DALY, disability-adjusted life year; DMFT/dmft, decayed, missing and filled teeth; ICER, incremental cost-effectiveness ratio; QALY, quality-adjusted life year. Table 2Incremental economic efficiency outcomes (ICER-based evidence only).Table 2 dummy alt textStudyEconomic frameworkCountry income levelOutcome metricOverall methodological qualityResultsEconomic conclusionChoi and Simon45CEA, model-basedHighIncremental costs and QALYs (cessation vs continuation)GoodCessation: −2.9 million QALYs; cost is +$9.8 billion over 5 yContinuation of fluoridation dominates cessationCiketic et al.35CEA, model-basedHighICER per DALYGoodICER: $3608 per DALY without treatment savings. ICER: −$64,127 per DALY when dental treatment cost offsets were includedDominant when treatment savings are included; cost-effective otherwiseCronin et al.41CEA, model-basedHighICER per d₃vcmft/D₃vcMFT preventedGoodICER: €14.09 per d3vcmft/D3vcMFT without treatment savings. ICER: −€115.67 per d3vcmft/D3vcMFT when dental treatment cost was includedDominant when treatment savings are included; cost-effective otherwiseFyfe et al.38CEA, model-basedHighNet cost per dmft/DMFT avertedGoodICER: −$0.55 to −$6.09 per dmft/DMFT averted, depending on population size and age groupDominantKroon and Van Wyk37CEA, model-basedUpper middleCost per DMFT savedGoodICER: US$3.32-US$11.08Cost-effectiveMariño et al.36CEA, model-basedUpper middleICER per DMFT avoidedGoodICER: S/. 10.73 to S/. 14.35 saving per DMFT avoided, depending on the discount rate and effectiveness scenariosDominantMoore et al.43CEA, empiricalHighICER per invasive dental treatment avoidedGoodICER: £94.55 per invasive dental treatment avoidedCost-effectiveWhittaker et al.44CEA, empiricalHighICER per QALY gainedGoodICER: £6245-£2973 per QALY gainedCost-effectiveDominant indicates lower costs and greater health benefits compared with the comparator.CEA, cost-effectiveness analysis; CUA, cost-utility analysis; DALY, disability-adjusted life year; DMFT/dmft, decayed, missing and filled teeth; ICER, incremental cost-effectiveness ratio; QALY, quality-adjusted life year. Incremental economic efficiency outcomes (ICER-based evidence only). Dominant indicates lower costs and greater health benefits compared with the comparator. CEA, cost-effectiveness analysis; CUA, cost-utility analysis; DALY, disability-adjusted life year; DMFT/dmft, decayed, missing and filled teeth; ICER, incremental cost-effectiveness ratio; QALY, quality-adjusted life year. Synthesis by ICER outcome revealed that 5 studies (62.5%) reported a dominant ICER, indicating that fluoridation both improved health outcomes and reduced total costs relative to no fluoridation.35,36,38,41,45 Two of these studies35,41 noted that dominance was elicited when treatment savings were included, while the intervention was cost-effective otherwise. Three studies (37.5%) reported a cost-effective but nondominant ICER.37,43,44 Different willingness to pay thresholds were used in studies reporting cost-effectiveness. Kroon and Van Wyk37 considered the intervention viable if the cost per person per year to save 1 DMFT was less than the cost of a 2-surface restoration ($32.52). Whittaker et al.44 utilised a willingness to pay threshold of £20,000 per QALY, which is the standard threshold used to determine whether interventions are acceptable to the NHS.46 Meanwhile, Moore et al.43 did not define a willingness to pay threshold; instead, cost-effectiveness was assessed as the mean cost per episode of invasive dental treatment avoided (Table 2). Stratification by country income level and study design showed that studies from high-income countries (HICs) contributed the majority of evidence (n = 6 studies).35,38,41,43, 44, 45 Within HICs, 4 model-based studies projected dominance,35,38,41,45 while 2 empirical studies from England43,44 reported positive but highly cost-effective ICERs. Upper middle-income countries (UMICs; n = 2 studies) provided model-based evaluations. Both UMIC studies reported favourable economic outcomes: dominance in Peru36 and a cost-effective ICER in South Africa (Table 2).37 Model-based studies (n = 6) projected fluoridation to be either dominant or, in the absence of treatment cost offsets, highly cost-effective.35, 36, 37, 38,41,45 These findings were consistent across both HIC and UMIC settings. The 2 empirical studies, both conducted in England, also concluded that fluoridation was cost-effective but reported positive ICERs rather than dominance (Table 2). The greater magnitude of economic benefit in model-based analyses compared to empirical studies might reflect the capturing of long-term cost offsets by the former. Dental outcomes Studies showed lower caries experience or reduced treatment needs in fluoridated populations.32,33,36,38, 39, 40, 41,43 Moreover, large increases in caries prevalence were predicted under fluoridation cessation scenarios.44,45 Subgroup analysis by age shows that fluoridation was associated with greater reductions in caries in children than in adults in multiple studies.36,38,41 Moreover, studies reported that young children (<7 years) experienced the largest increases in caries-related procedures following fluoridation cessation.39 Early-life exposure was associated with larger lifetime oral health benefits44 (Table 3).Table 3Clinical outcomes and age-specific effects.Table 3 dummy alt textStudyEconomic frameworkCountry income levelClinical outcome metricResultsDirection of effectAge-specific/subgroup effectsOverall methodological qualityCronin et al.41CEA, model-basedHighd₃vcmft/D₃vcMFTThe mean annual caries averted per child was 0.19 (5-y), 0.11 (8-y) and 0.11 (12-y).↓ cariesStronger effect in childrenGoodFyfe et al.38CEA, model-basedHighdmft/DMFTThe mean difference in caries experience between fluoridated and nonfluoridated areas was 1.0 dmft/DMFT for children (ages 2-17) and 0.8 DMFT for the total population↓ cariesChildren > adultsGoodMariño et al.36CEA, model-basedUpper middleDMFT (modelled)Children with access to salt fluoridation from ages 6 to 12 would show a decline to a mean DMFT of 3.07↓ cariesPaediatric populationsGoodMoore et al.43CEA, empiricalHighDMFT; invasive dental treatmentsThe predicted mean DMFT was 2% lower in the optimally fluoridated group compared to the nonoptimally fluoridated group (13.15 vs 13.36), a difference of 0.212 teeth↓ small but significantAdolescents and adultsGoodMaupomé et al.32Partia EE, empiricalHighDental treatment experienceChildren in fluoridated areas experienced fewer restorative dental treatments and lower treatment intensity compared with those in nonfluoridated areas.↓ in fluoridated areasNot reportedGoodMeyer et al.39Partia EE, empiricalHighCaries-related proceduresFollowing fluoridation cessation, caries-related dental procedures increased substantially, with the largest increases observed among children younger than 7 y↑ after cessationStrongest in children <7 yModerateGabardo et al.33CBA, empiricalUpper middleDMFT (ecological)Cities with a high Human Development Index (HDI-M ≥0.8) had a mean DMF-T of 3.62, while those with a lower HDI-M (<0.8) had a mean DMF-T of 5.18↓ with earlier exposureSocioeconomic gradient observedModerateGriffin et al.40Partia EE, empiricalHighCaries outcomesCommunities served by water fluoridation achieved comparable or better caries prevention compared with supplement-based delivery↓ with fluoridationNot reportedModerateChoi and Simon45CEA, model-basedHighCaries prevalence; number of decayed teeth (modelled)Eliminating fluoridation would increase the total number of decayed teeth by 25.4 million over 5 y and 53.8 million over 10 y↑ with cessationLarger increases in children and the publicly insuredGoodWhittaker et al.44CEA, empiricalHighOral health states underlying QALYsThe birth cohort had 0.13 fewer decayed, missing, or filled teeth and the older school cohort had 0.11 fewer dmft↓ oral health with no fluoridationChildhood exposure drives lifetime effectsGoodDMFT/dmft indicates decayed, missing and filled teeth in permanent/primary dentition; d₃vcmft/D₃vcMFT indicates cavitated dentine caries in primary/permanent teeth. Arrows indicate the direction of effect associated with fluoridation relative to the comparator (↓ reduction; ↑ increase). Results labelled as ‘modelled’ are derived from predictive or decision-analytic models; all other results are based on observed clinical or administrative data. Table 3Clinical outcomes and age-specific effects.Table 3 dummy alt textStudyEconomic frameworkCountry income levelClinical outcome metricResultsDirection of effectAge-specific/subgroup effectsOverall methodological qualityCronin et al.41CEA, model-basedHighd₃vcmft/D₃vcMFTThe mean annual caries averted per child was 0.19 (5-y), 0.11 (8-y) and 0.11 (12-y).↓ cariesStronger effect in childrenGoodFyfe et al.38CEA, model-basedHighdmft/DMFTThe mean difference in caries experience between fluoridated and nonfluoridated areas was 1.0 dmft/DMFT for children (ages 2-17) and 0.8 DMFT for the total population↓ cariesChildren > adultsGoodMariño et al.36CEA, model-basedUpper middleDMFT (modelled)Children with access to salt fluoridation from ages 6 to 12 would show a decline to a mean DMFT of 3.07↓ cariesPaediatric populationsGoodMoore et al.43CEA, empiricalHighDMFT; invasive dental treatmentsThe predicted mean DMFT was 2% lower in the optimally fluoridated group compared to the nonoptimally fluoridated group (13.15 vs 13.36), a difference of 0.212 teeth↓ small but significantAdolescents and adultsGoodMaupomé et al.32Partia EE, empiricalHighDental treatment experienceChildren in fluoridated areas experienced fewer restorative dental treatments and lower treatment intensity compared with those in nonfluoridated areas.↓ in fluoridated areasNot reportedGoodMeyer et al.39Partia EE, empiricalHighCaries-related proceduresFollowing fluoridation cessation, caries-related dental procedures increased substantially, with the largest increases observed among children younger than 7 y↑ after cessationStrongest in children <7 yModerateGabardo et al.33CBA, empiricalUpper middleDMFT (ecological)Cities with a high Human Development Index (HDI-M ≥0.8) had a mean DMF-T of 3.62, while those with a lower HDI-M (<0.8) had a mean DMF-T of 5.18↓ with earlier exposureSocioeconomic gradient observedModerateGriffin et al.40Partia EE, empiricalHighCaries outcomesCommunities served by water fluoridation achieved comparable or better caries prevention compared with supplement-based delivery↓ with fluoridationNot reportedModerateChoi and Simon45CEA, model-basedHighCaries prevalence; number of decayed teeth (modelled)Eliminating fluoridation would increase the total number of decayed teeth by 25.4 million over 5 y and 53.8 million over 10 y↑ with cessationLarger increases in children and the publicly insuredGoodWhittaker et al.44CEA, empiricalHighOral health states underlying QALYsThe birth cohort had 0.13 fewer decayed, missing, or filled teeth and the older school cohort had 0.11 fewer dmft↓ oral health with no fluoridationChildhood exposure drives lifetime effectsGoodDMFT/dmft indicates decayed, missing and filled teeth in permanent/primary dentition; d₃vcmft/D₃vcMFT indicates cavitated dentine caries in primary/permanent teeth. Arrows indicate the direction of effect associated with fluoridation relative to the comparator (↓ reduction; ↑ increase). Results labelled as ‘modelled’ are derived from predictive or decision-analytic models; all other results are based on observed clinical or administrative data. Clinical outcomes and age-specific effects. DMFT/dmft indicates decayed, missing and filled teeth in permanent/primary dentition; d₃vcmft/D₃vcMFT indicates cavitated dentine caries in primary/permanent teeth. Arrows indicate the direction of effect associated with fluoridation relative to the comparator (↓ reduction; ↑ increase). Results labelled as ‘modelled’ are derived from predictive or decision-analytic models; all other results are based on observed clinical or administrative data. The direction of effect was uniform across both model-based and empirical studies. All empirical estimates were statistically significant and aligned with the protective effect predicted by models. On stratification by country income level, only 2 studies were conducted in UMICs.33,36 The direction of evidence was consistent across both HIC and UMIC settings (Table 3). Net costs, budget impact and long-term fiscal effects All studies evaluating net costs or budget impact reported that fluoridation was associated with net fiscal benefits or cost savings and reductions in dental service expenditure. Meanwhile, fluoridation cessation was consistently associated with substantial increases in caries-related treatment costs.39,45 The favourable budgetary outcomes were reported across short-, medium- and long-term time horizons.31,34, 35, 36,38,41 Cost-benefit analyses in remote and small communities similarly demonstrated positive net present benefits over extended infrastructure lifespans.42 This fiscal benefit was observed regardless of the study design (model-based vs empirical) and income setting (HICs vs UMICs), though model-based analyses tended to project larger long-term savings than short-term empirical evaluations. When stratified by overall methodological quality, all studies with good quality reported favourable fiscal outcomes associated with fluoridation. The single study rated as moderate quality reported increased healthcare costs following fluoridation cessation and therefore showed the same direction of effect observed in higher-quality studies.39 When stratified by time horizon, favourable budgetary effects were observed across short-term (annual), medium-term (5-10 years) and long-term (≥20 years or lifetime) horizons. While some analyses indicated reduced cost savings in very small or remote systems,38,42 fluoridation generally remained cost-beneficial even under conservative assumptions (Table 4).34,37,38,42Table 4Net costs, budget impact and long-term fiscal effects.Table 4 dummy alt textStudyEconomic frameworkCountry income levelBudget/Cost outcomeResultsTime horizonOverall methodological qualityFiscal interpretationCampain et al.34CBA, model-basedHighNet annual per-person healthcare cost savingsEstimated savings ranged from AUD 69.86 to AUD 249.45 per person.AnnualGoodCost-savingCiketic et al.35CEA, model-basedHighTotal net savings when dental treatment cost offsets are includedTotal net saving of AU$ 665,686,529 (95% CI: −$973.6 to −$381.3 million)LifetimeGoodNet savingsCronin et al.41CEA, model-basedHighLifetime dental treatment savingsAverage net saving of −€17.68 per childLifetimeGoodNet savingsFyfe et al.38CEA, model-basedHighNet programme costs by community sizeThe annual equivalised per capita costs averted were $4.82 for the total population and $5.21 for children. CWF was cost-saving in all community sizes at the base caseAnnualGoodCost-saving (except in very small systems under sensitivity scenarios)Kroon and Van Wyk37CBA and CEA, model-basedUpper middleCBR/low annual per-person costCBR varied from 0.1 (50% reduction) to 0.34 (15% reduction). The average per capita annual cost was US$0.28 in 2006 and US$0.35 in 2011, with a 23.2% increase over the 5-y study periodAnnualGoodCost-beneficial (low per-capita cost)Mariño et al.36CEA, model-basedUpper middleOverall net cost savings across all scenariosOverall net saving of S/. 711,193 (US$ 236,277) for a hypothetical population of 25,000 children over 6 yAnnualGoodNet savingsMaupomé et al.32Partial EE, empiricalHighTotal and restorative dental treatment costsOver the 6-y study period, age-adjusted total cost differences (NF minus CWF) per person ranged from −$152.31 to $196.02. Lower restorative costs in the youngest (<18 y) and oldest (>58 y) cohorts. The highest cost differences were observed around age 75, where NF costs were approximately $75 higher than CWF costsMulti-yearGoodReduced healthcare expenditureMoore et al.43CEA, empiricalHighNHS dental service expenditureNHS dental treatment costs for optimally fluoridated patients were 5.5% lower (a saving of £22.26 per person over 10 y). The study estimated a 36% return on investment for the public sector, totalling £16.88 million in savings between 2010 and 2020.10 yGoodPositive return on investmentO’Connell et al.31CBA, model-basedHighNet healthcare cost savings and BCR (savings relative to programme costs)Annual savings of $148.9 million (credible range: $115.1-$187.2 million) in 2003 dollars, representing an average annual saving of $60.78 per person. BCR ranged from 21.82 for small water systems to 135.00 for large systemsAnnualGoodCost-saving; high return on investmentWhittaker et al.44CEA, empiricalHighDownstream reductions in NHS dental costsReductions in NHS dental service costs exceeded the cost of providing fluoridation. The mean cost difference was −£21.55 for the birth cohort and −£64.01 for the older school cohort5-6 yGoodNet budgetary benefitZhao et al.42CBA, empiricalHighNPB over infrastructure lifespanThe NPB was estimated at $2.02 million (typical community of 300-499 people) and $26.63 million (communities ≥2000) over 20 y20 yGoodCost-beneficialMeyer et al.39Partial EE, empiricalHighCaries-related treatment costs after cessationSignificant cost increases are associated with CWF cessation. Mean caries-related treatment costs per patient were 28% to 111% higher in the suboptimal CWF group after adjusting for inflationPostcessationModerateCosts increase after cessationChoi and Simon45CEA, model-basedHighNational dental care costs if fluoridation ceasedMaintaining the status quo (fluoridation) prevents a national health care cost increase of approximately $9.8 billion over 5 y and $19.4 billion over 10 y. These costs are primarily driven by treatment for increased tooth decay and associated complications5-10 yGoodCessation fiscally costlyMonetary values are reported in the currency and price year stated in the original studies. Negative values indicate net cost savings. Time horizons refer to the period over which costs or savings were estimated.AUD, Australian dollars; AU$, Australian dollars; CBR, cost-benefit ratio; CI, confidence interval; CWF, community water fluoridation; NF, nonfluoridated; NPB, net present benefit; ROI, return on investment. Table 4Net costs, budget impact and long-term fiscal effects.Table 4 dummy alt textStudyEconomic frameworkCountry income levelBudget/Cost outcomeResultsTime horizonOverall methodological qualityFiscal interpretationCampain et al.34CBA, model-basedHighNet annual per-person healthcare cost savingsEstimated savings ranged from AUD 69.86 to AUD 249.45 per person.AnnualGoodCost-savingCiketic et al.35CEA, model-basedHighTotal net savings when dental treatment cost offsets are includedTotal net saving of AU$ 665,686,529 (95% CI: −$973.6 to −$381.3 million)LifetimeGoodNet savingsCronin et al.41CEA, model-basedHighLifetime dental treatment savingsAverage net saving of −€17.68 per childLifetimeGoodNet savingsFyfe et al.38CEA, model-basedHighNet programme costs by community sizeThe annual equivalised per capita costs averted were $4.82 for the total population and $5.21 for children. CWF was cost-saving in all community sizes at the base caseAnnualGoodCost-saving (except in very small systems under sensitivity scenarios)Kroon and Van Wyk37CBA and CEA, model-basedUpper middleCBR/low annual per-person costCBR varied from 0.1 (50% reduction) to 0.34 (15% reduction). The average per capita annual cost was US$0.28 in 2006 and US$0.35 in 2011, with a 23.2% increase over the 5-y study periodAnnualGoodCost-beneficial (low per-capita cost)Mariño et al.36CEA, model-basedUpper middleOverall net cost savings across all scenariosOverall net saving of S/. 711,193 (US$ 236,277) for a hypothetical population of 25,000 children over 6 yAnnualGoodNet savingsMaupomé et al.32Partial EE, empiricalHighTotal and restorative dental treatment costsOver the 6-y study period, age-adjusted total cost differences (NF minus CWF) per person ranged from −$152.31 to $196.02. Lower restorative costs in the youngest (<18 y) and oldest (>58 y) cohorts. The highest cost differences were observed around age 75, where NF costs were approximately $75 higher than CWF costsMulti-yearGoodReduced healthcare expenditureMoore et al.43CEA, empiricalHighNHS dental service expenditureNHS dental treatment costs for optimally fluoridated patients were 5.5% lower (a saving of £22.26 per person over 10 y). The study estimated a 36% return on investment for the public sector, totalling £16.88 million in savings between 2010 and 2020.10 yGoodPositive return on investmentO’Connell et al.31CBA, model-basedHighNet healthcare cost savings and BCR (savings relative to programme costs)Annual savings of $148.9 million (credible range: $115.1-$187.2 million) in 2003 dollars, representing an average annual saving of $60.78 per person. BCR ranged from 21.82 for small water systems to 135.00 for large systemsAnnualGoodCost-saving; high return on investmentWhittaker et al.44CEA, empiricalHighDownstream reductions in NHS dental costsReductions in NHS dental service costs exceeded the cost of providing fluoridation. The mean cost difference was −£21.55 for the birth cohort and −£64.01 for the older school cohort5-6 yGoodNet budgetary benefitZhao et al.42CBA, empiricalHighNPB over infrastructure lifespanThe NPB was estimated at $2.02 million (typical community of 300-499 people) and $26.63 million (communities ≥2000) over 20 y20 yGoodCost-beneficialMeyer et al.39Partial EE, empiricalHighCaries-related treatment costs after cessationSignificant cost increases are associated with CWF cessation. Mean caries-related treatment costs per patient were 28% to 111% higher in the suboptimal CWF group after adjusting for inflationPostcessationModerateCosts increase after cessationChoi and Simon45CEA, model-basedHighNational dental care costs if fluoridation ceasedMaintaining the status quo (fluoridation) prevents a national health care cost increase of approximately $9.8 billion over 5 y and $19.4 billion over 10 y. These costs are primarily driven by treatment for increased tooth decay and associated complications5-10 yGoodCessation fiscally costlyMonetary values are reported in the currency and price year stated in the original studies. Negative values indicate net cost savings. Time horizons refer to the period over which costs or savings were estimated.AUD, Australian dollars; AU$, Australian dollars; CBR, cost-benefit ratio; CI, confidence interval; CWF, community water fluoridation; NF, nonfluoridated; NPB, net present benefit; ROI, return on investment. Net costs, budget impact and long-term fiscal effects. Monetary values are reported in the currency and price year stated in the original studies. Negative values indicate net cost savings. Time horizons refer to the period over which costs or savings were estimated. AUD, Australian dollars; AU$, Australian dollars; CBR, cost-benefit ratio; CI, confidence interval; CWF, community water fluoridation; NF, nonfluoridated; NPB, net present benefit; ROI, return on investment. Health equity and distributional outcomes Four studies reported equity-related results. Evidence from HICs (n = 3 studies) showed mixed results. A natural experiment using data from Medicaid dental claims showed that cessation was associated with substantial increases in caries-related treatment utilisation and costs among Medicaid-enrolled children, particularly those younger than 7 years.39 A national decision-analytic model similarly projected that fluoridation cessation would disproportionately increase caries prevalence and treatment costs among low-income and publicly insured children, further widening existing inequalities.45 On the other hand, a UK study found no compelling evidence that water fluoridation reduced social inequalities in dental health during the study period.43 The single empirical study from a UMIC setting found that water fluoridation coverage was influenced by social development rather than need, with the most socially developed areas receiving the oral health benefits first. In contrast, less developed areas remained at a disadvantage33 (Table 5).Table 5Health equity and distributional outcomes.Table 5 dummy alt textStudyEconomic frameworkCountry income levelEquity dimension assessedResultsDirection of effectPopulation subgroup(s)Overall methodological qualityGabardo et al.33CBA, empiricalUpper middleArea-level socioeconomic development (HDI-M)Municipalities with higher HDI-M had lower mean DMFT; fluoridation was associated with lower DMFT overall, but socioeconomic gradients persistedThe SES gradient persistsMunicipalities stratified by HDI-MModerateMeyer et al.39Partial EE, empiricalHighSocioeconomic vulnerability (Medicaid population)Following fluoridation cessation, caries-related treatment costs and utilisation increased among Medicaid-enrolled children; between-group SES comparisons were not assessed↑ adverse effects on the disadvantaged groupsMedicaid-insured children, especially <7 yModerateMoore et al.43CEA, empiricalHighArea-level deprivation (IMD deciles)Outcomes were analysed by patient deprivation decile; no compelling evidence was found that water fluoridation reduced social inequalities in dental health during the study periodNo clear change in inequalitiesPatients stratified by IMD decilesGoodChoi and Simon45CEA, model-basedHighSocioeconomic status; insuranceModelled analyses projected that fluoridation cessation would disproportionately increase caries prevalence and treatment costs among low-income and publicly insured children↑ inequalities after cessationLow-income and publicly insured childrenGoodDMFT, decayed, missing and filled teeth (permanent dentition); HDI-M, Municipal Human Development Index; IMD, index of multiple deprivation; SES, socioeconomic status. Table 5Health equity and distributional outcomes.Table 5 dummy alt textStudyEconomic frameworkCountry income levelEquity dimension assessedResultsDirection of effectPopulation subgroup(s)Overall methodological qualityGabardo et al.33CBA, empiricalUpper middleArea-level socioeconomic development (HDI-M)Municipalities with higher HDI-M had lower mean DMFT; fluoridation was associated with lower DMFT overall, but socioeconomic gradients persistedThe SES gradient persistsMunicipalities stratified by HDI-MModerateMeyer et al.39Partial EE, empiricalHighSocioeconomic vulnerability (Medicaid population)Following fluoridation cessation, caries-related treatment costs and utilisation increased among Medicaid-enrolled children; between-group SES comparisons were not assessed↑ adverse effects on the disadvantaged groupsMedicaid-insured children, especially <7 yModerateMoore et al.43CEA, empiricalHighArea-level deprivation (IMD deciles)Outcomes were analysed by patient deprivation decile; no compelling evidence was found that water fluoridation reduced social inequalities in dental health during the study periodNo clear change in inequalitiesPatients stratified by IMD decilesGoodChoi and Simon45CEA, model-basedHighSocioeconomic status; insuranceModelled analyses projected that fluoridation cessation would disproportionately increase caries prevalence and treatment costs among low-income and publicly insured children↑ inequalities after cessationLow-income and publicly insured childrenGoodDMFT, decayed, missing and filled teeth (permanent dentition); HDI-M, Municipal Human Development Index; IMD, index of multiple deprivation; SES, socioeconomic status. Health equity and distributional outcomes. DMFT, decayed, missing and filled teeth (permanent dentition); HDI-M, Municipal Human Development Index; IMD, index of multiple deprivation; SES, socioeconomic status. Discussion Summary of the main findings This systematic review synthesised evidence regarding the economic evaluation of using fluoridation interventions against no fluoridation or a different fluoridation alternative. While the included studies showed substantial variations in their analytical frameworks, perspectives and populations, the results indicated that fluoridation introduction/continuation was cost-effective or cost-saving compared to no fluoridation or fluoridation cessation. The benefits of fluoridation interventions were derived from reduced rates of dental caries and downstream treatment costs. As regards the budget impact of fluoridation, several studies reported net cost savings over short- to long-term horizons.31,34, 35, 36,38,41 The savings from averted dental treatment often exceeded the costs of the implementation of the fluoridation programme; thus, fluoridation showed dominance (lower costs with improved outcomes) compared to no-fluoridation or fluoridation-cessation scenarios. These findings indicate that fluoridation is not only economically efficient but also affordable and fiscally sustainable across diverse settings. Equity considerations were less extensively evaluated in the included studies and showed mixed results. Two studies suggested that cessation of fluoridation can increase the adverse effects and inequalities in the disadvantaged populations (low-income, publicly insured groups).39,45 On the other hand, a large cohort study found no strong evidence that water fluoridation reduced social inequalities in dental health.43 These discrepancies could be attributed to variations in implementation contexts, population risk profiles and measurement sensitivities rather than a lack of effectiveness. An ecological study in Brazil found that the introduction of CWF followed the inverse care law, where the most socially developed areas received the oral health benefits first. In contrast, less developed regions remained at a disadvantage.33 This reflects existing inequalities in social development rather than the intervention’s failure to reduce them. In addition, the diffusion of fluoride through foods and beverages processed in fluoridated areas can provide benefits to nonfluoridated communities.31,45 This diffusion can dilute the differences between the socioeconomic groups, potentially masking the intervention’s impact on social equity. Future EEs should prioritise study designs that can better address equity questions. For instance, natural experiments compare outcomes before and after a policy change; thus, researchers can directly observe whether vulnerable groups bear a disproportionate burden when the intervention is removed/introduced. In addition, quasi-experimental difference-in-difference (DiD) designs compare intervention and control groups over time to eliminate statistical noise from other sources of fluoride. Several factors may affect the cost-effectiveness of fluoridation programmes, including the population size, the decline in the risk of caries and the age groups. O’Connell and colleagues31 estimated that about 80% of the additional savings that could be achieved by expanding CWF in Colorado, the USA, would be realised in the large water systems serving more than 20,000 individuals. Fyfe and coworkers38 found that sensitivity analysis for communities under 5000 showed a positive net cost (indicating less cost-effectiveness) under scenarios with high discount rates or short equipment lifespans. Cronin and collaborators41 reported that in the smallest CWF schemes serving <1000 people, the cost of supply outweighed treatment savings, whereas CWF was cost-saving for all communities with >1000 people, with the lowest per capita supply costs in the largest population centers (>100,000). Zhao and coworkers42 found that benefit-cost ratios increased significantly with population size, ranging from 1.1 for communities with <300 people to 16.0 for those with ≥2000. Moreover, they observed that while CWF was cost-effective for all groups, the payback period was 15 years for the smallest communities compared to only 2.2 years for the largest. Our findings agree with those of the scoping review on CWF-related economic evaluations conducted by Mariño and Zaror,13 who found that CWF was a cost-saving strategy – compared with no fluoridation – regardless of the adopted perspective, time horizon or discount rate. Meanwhile, the baseline caries risk of the residents may contribute substantially to the economic impact of the intervention, regardless of the population size. Several studies estimated greater cost-effectiveness in communities with a higher risk of dental caries,35 particularly the low-income or disadvantaged remote ones38,42,43 or those publicly insured or uninsured.45 As regards the differential effect by age groups, studies analysing cessation39 or prescription shifts40 highlight that children under the age of 6 or 7 are the most sensitive to changes in water fluoridation status. For school-aged children (ages 5-12), CWF consistently results in net cost savings across different national contexts, including Ireland, New Zealand and the UK.38,41,44 Debate exists regarding the cost-effectiveness of CWF for older adults. However, several studies indicated that CWF remains a highly cost-effective measure for older adults by supporting healthy aging and reducing the fiscal burden of complex dental care.31,32,34 Overall completeness, applicability and quality of evidence The included studies showed some limitations that may affect the quality of evidence and the estimates derived from cost-effectiveness analyses. A potential limitation is the limited inclusion of costs related to adverse effects, specifically dental fluorosis. Only 1 study45 explicitly included fluorosis treatment costs in their microsimulation model, valuing moderate-to-severe fluorosis at a mean of $1468 per case (range: 1050-1850). Cronin et al.41 estimated that annual fluorosis treatment costs would need to reach €414 per child to entirely negate the savings from caries prevention. Meanwhile, most studies excluded the costs of fluorosis. The reasons cited for this omission comprised negligible prevalence in their country,42 the lack of significant difference in fluorosis rates between fluoridated and nonfluoridated populations,38 and the inability to assess fluorosis in preschool children.44 While the omission of these costs may lead to a slight overestimation of net benefits, the available data suggest that the economic benefits of caries prevention remain robust unless fluorosis treatment costs are exceptionally high. Future research should incorporate the costs of fluorosis treatment into economic models to provide more robust and comprehensive recommendations to national policy decision-makers. Additionally, other methodological variations in economic modeling were observed, particularly regarding the application of discount rates to future costs and health outcomes. In the case of public health interventions, implementation costs are immediate while the benefits accrue over the years. Therefore, the nonapplication of discounting rates can result in an overestimation of the net benefits and interventions appear more cost-effective than they would be. While these studies still align with the overall directional trend of the review, their absolute magnitude of ‘net savings’ should be interpreted with caution. Moreover, several studies relied on different clinical endpoints for calculating ICER, such as DMFT or averted treatment procedure, rather than quality-of-life measures, limiting comparability across the studies. The current systematic review also showed some limitations. First, the evidence was derived only from published analyses of the included studies. Unpublished data and grey literature were not included. In addition, we included only studies published in English, which may introduce bias by excluding non-English studies that arise mainly from low- and middle-income countries (LMICs) with widespread fluoridation programmes. Several systematic reviews have reported that excluding non-English studies has a minimal effect on overall conclusions.47, 48, 49 However, we think that excluding LMIC studies is more likely to inflate ICERs for public health interventions, thus underestimating the intervention’s cost-effectiveness. In the settings of LMICs, intervention costs are usually lower (eg, cheaper infrastructure and labour), whereas caries burdens are higher, resulting in more favourable ICERs compared to high-income settings. Future cost-effectiveness analyses should endeavour to incorporate evidence published in other languages to fully capture global economic evidence. Conclusions, implications for practice, policy and future research The evidence from this systematic review suggests that continued or expanded investment in CWF may represent a cost-effective public health intervention. The degree of cost-effectiveness is likely context-dependent, influenced by local implementation costs and baseline caries prevalence. Cessation of fluoridation may lead to the reversal of these benefits, particularly in disadvantaged populations. However, these findings should be interpreted cautiously in light of the heterogeneity in study designs and moderate quality of evidence. To address the existing uncertainties, future research should prioritise the inclusion of distributional cost-effectiveness analyses to assess equity issues related to fluoridation programmes. Longitudinal evaluations in adults are recommended to assess cumulative, lifetime benefits. Future studies should also utilise standardised outcome measures – such as QALYs – to enable comparability across different settings and studies. CRediT authorship contribution statement Diaa Almutairi: Conceptualization, Methodology, Writing – review & editing. Moodhi Al Humaidi: Conceptualization, Methodology, Writing – review & editing. Saeed Mastour Alshahrani: Conceptualization, Methodology, Writing – review & editing. Awatef Albalawi: Conceptualization, Methodology, Writing – review & editing. Conflict of interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.