A uniform approach for costing school-based lifestyle interventions is currently lacking. The objective of this study was to develop a template for costing primary school-based lifestyle interventions and apply this to the costing of the “Healthy Primary School of the Future” (HPSF) and the “Physical Activity School” (PAS), which aim to improve physical activity and dietary behaviors. Cost-effectiveness studies were reviewed to identify the cost items. Societal costs were reflected by summing up the education, household and leisure, labor and social security, and health perspectives. Cost inputs for HPSF and PAS were obtained for the first year after implementation. In a scenario analysis, the costs were explored for a hypothetical steady state. From a societal perspective, the per child costs were €2.7/$3.3 (HPSF) and €− 0.3/$− 0.4 (PAS) per day during the first year after implementation, and €1.0/$1.2 and €− 1.3/$− 1.6 in a steady state, respectively (2016 prices). The highest costs were incurred by the education perspective (first year: €8.7/$10.6 (HPSF) and €4.0/$4.9 (PAS); steady state: €6.1/$7.4 (HPSF) and €2.1/$2.6 (PAS)), whereas most of the cost offsets were received by the household and leisure perspective (first year: €− 6.0/$− 7.3 (HPSF) and €− 4.4/$− 5.4 (PAS); steady state: €− 5.0/$− 6.1 (HPSF) and €− 3.4/$− 4.1 (PAS)). The template proved helpful for costing HPSF and PAS from various stakeholder perspectives. The costs for the education sector were fully (PAS) and almost fully (HPSF) compensated by the savings within the household sector. Whether the additional costs of HPSF over PAS represent value for money will depend on their relative effectiveness. . . . Keywords Costs and cost analysis School health services/economics Schools Child Introduction 2016;Tangetal. 2009). A number of studies found that recent primary school-based lifestyle interventions are effective in It is increasingly being recognized that schools have a key role normalizing children’s body mass index (BMI) (Mei et al. in the promotion of children’s health and wellbeing (Hayman 2016; Sobol-Goldberg, Rabinowitz, and Gross 2013). Much less is known about the costs of these interventions (John, Wenig, and Wolfenstetter 2010). This study focuses on Electronic supplementary material The online version of this article performing systematic and standardized cost calculations on (https://doi.org/10.1007/s11121-018-0918-1) contains supplementary school-based lifestyle interventions. material, which is available to authorized users. A wide range of cost estimates have been published, rang- ing from €4to €865 per child per year (adjusted for purchase * Marije Oosterhoff firstname.lastname@example.org power parity) (see for example: (Barrett et al. 2015;Cradock et al. 2017; McAuley et al. 2010; te Velde et al. 2011; Waters et al. 2017), which is likely due to the high variation in inter- Department of Clinical Epidemiology and Medical Technology Assessment (KEMTA), Maastricht University Medical Centre vention components, setup of interventions, and the heteroge- MUMC+/Care and Public Health Research Institute (CAPHRI), neity in costing approaches. The heterogeneity in costing ap- Maastricht University, Maastricht, The Netherlands proaches is, amongst others, reflected by the various methods Department of Social Medicine, Care and Public Health Research for identifying and selecting cost items. While the recurring Institute (CAPHRI), Faculty of Health, Medicine and Life Sciences, resource use of intervention delivery is generally included in Maastricht University, P.O. Box 5800, 6202, AZ cost calculations, a minority of studies included development Maastricht, The Netherlands and evaluation costs, and consensus on this aspect seems to be Department of Family Medicine, Care and Public Health Research lacking (Meng et al. 2013; Moodie et al. 2013; Wang, Li, Institute (CAPHRI), Faculty of Health, Medicine and Life Sciences, Maastricht University, Maastricht, The Netherlands Siahpush, Chen, and Huberty 2017). In addition to the Prev Sci (2018) 19:716–727 717 variability in costing methods, most studies provided little The objective of this study was to develop a template for information on how resource consumption was identified, the costing of primary school-based lifestyle interventions. measured, and valued. For example, most studies listed the The template should enable to perform cost calculations from included cost categories but did not justify why cost categories various stakeholder perspectives and examine the impact of were excluded (e.g., not applicable to the intervention, not in different scenarios, assumptions, or intervention adaptations. alignment with the perspective of the costing study). The lack To illustrate the applicability of this framework, we calculated of standardization and transparency makes it difficult to ascer- the costs of the “Healthy Primary School of the Future” tain whether cost estimates are valid and reliable. (HPSF) and the “Physical Activity School” (PAS). Standardized and transparent cost calculations are of utmost importance to be able to perform cost comparisons between interventions, and to examine which interventions yield the Methods best value in return for the required investments. The method- ological quality of cost estimations is also key for implemen- Firstly, we developed a template for the cost calculation of tation as underestimating the intervention costs will lead to school-based lifestyle interventions. The template was then implementation, but delivery will not be sustained when the used for the cost calculation of HPSF and PAS by selecting costs are found to be higher than initially estimated. In con- the cost items that applied to both interventions and by mea- trast, overestimating the intervention costs may withhold chil- suring and valuing the selected cost items. dren from receiving the benefits of effective interventions. Decision-makers in public health are interested in the social Template for Cost Calculation value of school interventions given the fact that most recent school-based lifestyle interventions are aimed at involving The cost items of primary school-based lifestyle interventions multiple stakeholders (e.g., schools, local government, sport were identified from published cost-effectiveness studies on clubs, caterers, child care organizations, health staff), and may these type of programs (Kersley and Knuutila 2011; affect outcomes in others than the child (Middleton, Evans, Oosterhoff et al. 2017). Cost items reflected the incremental Keegan, Bishop, and Evans 2014; Warwick et al. 2005). resource consumption relative to the regular education pro- Additionally, interventions may not only impact on health gram. We focused on the recurrent activities for implementa- outcomes, but may also improve wellbeing by creating posi- tion, delivery, and evaluation. The one-time investments for tive school environments, and improve educational outcomes intervention development and scientific evaluation fell be- as a result of the associations with children’s health status yond the scope of this study. It was aimed to include all cost (Langford et al. 2014). Most economic evaluations have ex- items that reflected the social opportunity costs of an interven- amined the social value of school-based lifestyle interventions tion. Social opportunity cost takes account of the actual pay- by adopting a societal perspective for the calculation of costs ments and the economic costs. The economic costs represent and outcomes. In a societal perspective, all the costs and out- the value of all resources that are not billed, but which are used comes that change as a result of the intervention are consid- for the intervention and are thus no longer available for other ered, regardless of who incurs them or in which sector they purposes (e.g., time investments of volunteers). Cost items fall. A previous study has identified that the costs and out- could reflect a positive cost (cost increase compared to the comes of public health interventions may fall into the regular education program) or a negative cost (cost saving healthcare, education, labor and social security, household compared to the regular education program). Cost items with and leisure, or safety and justice sector (Drost, Paulus, a negative cost should be directly related to intervention de- Ruwaard, and Evers 2014). To effectively inform decision- livery and not to the outcomes of school-based lifestyle inter- makers, it is important to perform cost calculations from a ventions (e.g., provision of lunches at school is a cost offset societal perspective and take account of the various sectors for the household). We refer to this type of cost as a delivery- that are involved or affected by the delivery of school-based related offset. In this study, we made a distinction between a lifestyle interventions. societal perspective and various smaller perspectives, which To determine the support to an intervention, stakeholders contained of the healthcare, education, labor and social secu- require information on the costs they will incur. So far, few rity, and the household and leisure perspective (Drost, Paulus, studies distinguished different stakeholder perspectives, but Ruwaard, and Evers 2014). Cost items were also classified on doing so may be helpful to interpret whether the implementa- the basis of personnel and material costs. tion of an intervention is economically feasible or whether a cost redistribution among stakeholder groups would facilitate The HPSF and PAS Interventions implementation (Moodie, Carter, Swinburn, and Haby 2010; Moodie, Haby, Galvin, Swinburn, and Carter 2009;Moodie, The HPSF and PAS interventions are being evaluated in a Haby, Swinburn, and Carter 2011;Moodieetal. 2013). quasi-experimental study located in the southern region of 718 Prev Sci (2018) 19:716–727 the Netherlands. The interventions are delivered from October for 4 days a week, it was assumed that the primary caregiver of 2015 onwards to children between 4 and 12 years of age the child (usually a parent) would spend less time on provid- (Willeboordse et al. 2016). The HPSF (two schools) and ing after-school care (reduction in caregiver time). We includ- PAS schools (two schools) both target at physical activity ed the value of this time offset, because the primary caregiver and involve a longer lunch break. At PAS, the lunch break is could use this time offset for work or leisure activities. For the extended with approximately half an hour, while at HPSF, the calculations, we assumed that the average family size lunch break is prolonged with about 1 hour. Therefore, chil- amounted to two children per household. Beneficiaries of un- dren attend school to approximately 15:30/15:45 instead of employment benefits were employed to help at HPSF with 15:00. At HPSF, some of the lunch breaks involve an educa- preparing school the lunches. This resulted in a saving for tional component to meet the education hour requirements. the household sector due to the difference between salary With regard to physical activity, additional sports, play, and and unemployment benefits for beneficiaries. In addition, creative activities are provided for at least 4 days a week employing beneficiaries resulted in a cost offset for the social during the lunch break. Activities are being set up by cross- security sector due to the savings from unemployment bene- discipline coordinators. Activities are guided by pedagogical fits and earnings from income taxes. A gross approach to staff from childcare partners with the assistance of volunteers. costing was used by first measuring and valuing costs at the At HPSF, children are also provided with a healthy lunch and level of the school and then disaggregating costs to the level of a morning snack. Lunches are prepared at schools by catering the child. The costs per child were calculated by dividing the personnel and beneficiaries of unemployment benefits as part total costs by the average number of children per school, be- of their reintegration to the labor market. The lunches consist cause the interventions were provided at the school-level, and of a variety of food products from which children can choose. children were equally exposed to the interventions. All costs Lunch is provided in the classroom or in a central location in are expressed in 2016 euros. Costs were also converted to the school. The lunch sessions are supervised by pedagogical American dollars using the purchase power parity (PPP staff and volunteers. Schools were able to implement addition- 2016, US$ 1 = € 0.82) (https://data.oecd.org/conversion/ al activities based on a list of relevant and evidence-based purchasing-power-parities-ppp.htm). Given the 1-year time additional activities (Willeboordse et al. 2016). In this study, horizon, discounting was not applied. we only focused on the compulsory changes and did not ex- amine the cost of additional activities. Control schools main- Scenario Analyses tained the regular school curriculum. Further details have been published elsewhere (Willeboordse et al. 2016). While the initial delivery of school-based lifestyle interven- tions takes place during a research period, interventions are Cost Analysis of HPSF and PAS only embedded after research has ended. Until interventions become embedded into the school setting, learning-curve ef- We examined the incremental costs, which is defined as the fects and efficiency improvements may occur, because more additional costs of HPSF and PAS in comparison to the regu- experience is obtained with the delivery of interventions (e.g., lar school curriculum. The annual social opportunity costs of some activities may take less time). If interventions are HPSF and PAS were calculated for the first year after their sustained, they become to rely on more structural sources of implementation. Key stakeholders were consulted by means funding. For this reason, it is interesting to examine whether of semi-structured interviews to examine which of the cost organizational improvements are feasible (without items from the template applied to HPSF and PAS, and to compromising the intervention’s relative effectiveness) in or- complement the template. A calculation sheet was created to der to increase funding opportunities. In the first scenario incorporate the cost items, volume, and unit prices. The inputs analysis, the costs of HPSF and PAS were explored for a for volume and unit prices were obtained retrospectively from hypothetical steady state. The steady state reflected a hypo- primary data sources (budget information, accounting data, thetical situation in which the HPSF and PAS interventions are stakeholder interviews, curriculum information). Because the delivered at their full capacity, and learning curves and effi- activities at HPSF and PAS were all provided in addition to the ciency improvements were no longer applicable. Because a regular school curriculum, the resource use at HPSF and PAS steady state was not yet reached, stakeholders were consulted was examined, and the resource use in control schools was set to define the expected learning curves and efficiency to zero. Volumes were based on the average consumption at improvements. HPSF and PAS. For the unit prices, we used both primary data Potential areas for cost reduction were examined with a sources and secondary data sources, with the latter mainly pedagogical staff scenario and an efficiency scenario. In the being used to obtain values for resources that were not directly pedagogical staff scenario, it was assumed that lunch breaks billed (e.g., time investments of volunteers). Because the and activities were guided by teaching assistants instead of pedagogical staff from external childcare partners. The school hours were extended with approximately half an hour Prev Sci (2018) 19:716–727 719 efficiency scenario was constructed based on the views of financial contributions to compensate their time investments stakeholders about the potential areas for further cost reduc- (e.g., time investments of volunteers are costs to the household tions. Stakeholders reported that potential cost reductions sector, and the financial compensation is a cost to the education could be achieved by decreasing the time investments of ex- sector). To avoid double counting when calculating the costs ternal parties providing workshops (not considered as an es- over multiple perspectives, it is important to only count the net sential intervention component) and volunteers (no volunteers costs for each stakeholder, being the value of time investments needed in upper grades), and by reducing the costs of transport minus the received financial contributions. Overheads were not (cost reduction of 50% in the first year after implementation included in the framework as the proportion of overheads may given the limited use of available budgets during the first vary between settings and stakeholder perspectives. However, it year). should be noted that taking account of overheads is highly Furthermore, scenario analyses may be performed when relevant for the costing of interventions. An overview of the parameters are uncertain. In the extended school day scenario, cost items of primary school-based lifestyle interventions is we relaxed the assumption that the additional school hours presented in Table 1. resulted in a cost offset due to the time being freed up for primary caregivers. In the extended school day scenario, we Cost Analysis of HPSF and PAS did not apply a cost offset as it was assumed that the extended school day also caused utility losses as primary caregivers From a societal perspective, the per child costs amounted to might derive utility or wellbeing from being able to provide €2.7/day for HPSF and to €− 0.3/day for PAS during the first after-school care. year after implementation (Table 2, results expressed in American dollars are available online). Because the fee for Sensitivity Analyses supervision during the lunch break (€1/child/day) was not charged in the first year after implementation, an additional Sensitivity analyses were conducted for the first year after cost reduction or financial contribution of €1.7/day (first year) implementation and for the hypothetical steady state. The is required to make the HPSF intervention cost-saving from a main input parameters were varied to examine whether the societal perspective. The highest costs were incurred by the costs per child were influenced by changes in resource con- education perspective (€8.7 (HPSF) and €4.0 (PAS) per day), sumption and unit prices. Input parameters were varied with ± whereas most of the cost reductions were received by the 50% because volumes and unit prices were not empirically household and leisure perspective (€− 6.0 (HPSF) and €− measured, and the uncertainty around the parameters was 4.4 (PAS)). Given the education perspective, the personnel therefore unknown. costs (HPSF, 69%; PAS, 94%) made up the largest proportion of the costs for HPSF and PAS. Results Scenario Analysis In this study, we developed a template to support cost calcu- lations for lifestyle interventions in the primary school setting With the steady-state assumptions such as the reductions in and illustrated its applicability. time investments of coordinators and the unit costs of lunches (Table 3, results expressed in American dollars are available Template for Cost Calculation online), the per child costs amounted to €1.0/day for HPSF and to €− 1.3/day for PAS. The costs for the education sector Personnel costs contained the time investments of stakeholders (€6.1 (HPSF) and €2.1 (PAS)) were fully compensated for that were compensated (e.g., salary, financial contribution) or PAS by the savings in the household and leisure sector (€− not compensated. School personnel may invest time in the local 5.0 (HPSF) and €− 3.4 (PAS)). In the hypothetical steady planning, coordination, training (and substituting teachers dur- state, the HPSF intervention could become cost-saving from ing the training of personnel), delivery, and evaluation of inter- a societal perspective when the unit costs of lunches do not ventions. The time investments of volunteers and primary care- exceed the cost threshold of €1.0 per child per day, or when givers belonged to the household sector. Negative costs may the costs of pedagogical staff will decrease by 70%. come into play when the school day is extended or when inter- The pedagogical staff scenario led to the highest cost reduc- ventions are offered during the after-school hours, which results tions (to €− 0.6 (HPSF) and €− 1.6 (PAS) in the first year after in time being freed up for the primary caregiver of the child implementation) (Table 4, results expressed in American (reduction in caregiver time). Personnel costs may apply to dollars are available online). Further research should be per- multiple stakeholder perspectives. This occurs when stake- formed to assess whether these scenarios result cost reductions holders outside the education sector receive subsidies or without compromising the intervention’s relative effectiveness. 720 Prev Sci (2018) 19:716–727 Table 1 Cost items of primary school-based lifestyle interventions Cost type Cost item Stakeholder perspective Education Household Labor and social Healthcare and leisure security Personnel School personnel Program coordinator X School project leader X Teacher X Teaching assistant X A A Volunteers, carers, and beneficiaries from the household sector X X X External parties from the leisure sector X X Personnel from the local government (e.g., cross-discipline coordinators) X X Personnel from the labor sector (e.g., pedagogical staff from childcare partners,XX catering personnel) Health staff (e.g., local health department) X X Materials Transport X Accommodations X Food X X Curriculum materials X Monitoring equipment X Advertising and promotion X Accreditation and certification X Training materials for personnel X Communication and administration X There is a potential overlap in costs (see main text). To avoid double counting with the costs in the education sector, it is important to only count the net costs for each stakeholder, being the value of time investments minus the received financial contributions Food costs are a positive cost to the education sector and a negative cost to the household sector and do not indicate an overlap in costs Stakeholders may invest time for the local planning, coordination, training (and substituting teachers during the training of personnel), delivery, and recurrent evaluation. Furthermore, a per child fee may be required to compensate for the time investments of stakeholders. When beneficiaries of unemployment benefits are employed, this may lead to offsets in the household sector and in the social security sector (see main text) The extended school day scenario drives up the societal per distinguished between the education, household and lei- child costs from €2.7 per day to €6.2 (HPSF) and from €− 0.3 sure, labor and social security, and health perspective. The per day to €3.2 (PAS) in the first year after implementation. framework proved useful for calculating the costs of HPSF andPAS:Ithelpedtosystematicallyidentifythe Sensitivity Analysis cost items, distinguish between stakeholder perspectives, and avoid the double counting of cost items. The societal Figures 1 and 2 show how the costs per child vary with the costs were estimated at €2.7 and €− 0.3per dayin thefirst upper and lower bounds on the key inputs (full tables available year after implementation and at €1.0 (HPSF) and €− 1.3 online). The time investments on after-school care in the (PAS) for a hypothetical steady state, respectively. The household, the costs of pedagogical staff, and the unit costs education sector incurred the highest per child costs of food were the main factors of influence. (€8.7 and €6.1/day (HPSF); €4.0 and €2.1/day (PAS)), whichwerefullycompensatedbythe savingsinthe household sector for PAS (€− 4.4/day and €− 3.4/day), Discussion and almost fully compensated for HPSF (€− 6.0/day and €− 5.0/day). The objective of this study was to develop a template for Cost estimates are dependent on the methodological the costing of primary school-based lifestyle interventions approaches and assumptions that are being used for mak- in order to improve the quality and comparability of cost ing cost calculations. For example, we included delivery- calculations. The framework listed the cost items of pri- related offsets as they may affect the opportunity costs to mary school-based lifestyle interventions and stakeholders. The impact of the extended school hours Prev Sci (2018) 19:716–727 721 Table 2 Per child costs of HPSF and PAS for the first year after implementation Cost items for the first year after Time investment Volume (per school) Unit price Stakeholder perspective implementation Education Household and Labor and Healthcare Societal leisure social perspective security HPSF PAS HPSF PAS HPSF PAS HPSF PAS HPSF PAS Personnel Program coordinator Coordination 4 schools: 1 FTE €112,000/FTE €84 €84 School project leaders Coordination HPSF: 0.5 FTE; PAS: 0.4 FTE €65,000/FTE €97 €78 A A Volunteers Assisting during lunch Financial compensation: €38 €24 €53 €17 break and activities HPSF: €12,645 PAS: €8145 Time investment Time investment HPSF: 12 volunteers, 1 h/day, €14/hour 4 days/week (upper grades 5 days/week), 40 weeks PAS: 7 volunteers, 1 h/day, 3 days/week (upper grades 4 days/week), 40 weeks Primary caregivers Per child fee for 4 times/week, 40 weeks €1/day not charged to €160 €160 €− 160 €− 160 supervision during carers lunch break Parental evaluation 5 times/year, 1 h, 10 persons €14/h €2 €2 committee Value of the extended 0.5 h freed up, 4 times/week €14/hour €− 562 €− 562 school hours 2 children / household Beneficiaries of unemployment benefits Preparing lunches as 1 person, 15 h/week NA part of reintegration to the labor market 1 B B External parties from the Giving workshops €6675 €20 €20 €0 €0 leisure sector 1 C C Cross-discipline coordinators Organizing activities 1 FTE/4 schools €50,000/FTE €37 €37 €0 €0 from the local government 1 D D Pedagogical staff from Guidinglunchbreak HPSF: 12 persons, 2 h/day, €65,000/FTE €524 €204 €0 €0 childcare partners and activities 4 days/week (upper grades 5 days/week), 40 weeks PAS: 8 persons, 1.5 h/day, 3 days/week (upper grades 4 days/week), 40 weeks 722 Prev Sci (2018) 19:716–727 Table 2 (continued) Materials Transport €7097 €21 €21 Accommodations 1 h, 4 days/week €17.5/h €8 €8 3 F F Food (including personnel HPSF: 4 times/week, 40 weeks Cost: €2.46/child/day €394 €− 298 from caterer) Offset: €− 1.86/day Curriculum materials 1 set per year €2500/school €7 €7 Monitoring equipment 1 survey €1200/survey €4 €4 Total costs Net costs (per child/year) €1394 €647 €− 965 €− 704 €0 €0 €0 €0 €429 €− 57 Personnel €959 €606 €− 667 €− 704 €0 €0 €0 €0 €292 €− 96 Materials €434 €41 €− 298 €0 €0 €0 €0 €0 €137 €41 Net costs (per child/day) €8.7 €4.0 €− 6.0 €− 4.4 €0 €0 €0 €0 €2.7 €− 0.3 HPSF Healthy Primary School of the Future, PAS Physical Activity School; FTE full-time equivalent Discrepancies between the sum of cost items may be due to rounding Budget “the Healthy Primary School of the Future” Productivity costs of paid labor (Zorginstituut Nederland 2015) Accounting data “the Healthy Primary School of the Future” Productivity costs of unpaid labor (Zorginstituut Nederland 2015) Household expenses on children’s lunches (NIBUD 2017) Minimum wage (“Minimumloon 2016”) Unemployment benefits (Rijksoverheid 2016) Income tax (Belastingdienst 2016) Value of time investment minus financial compensation B–D Financial contributions fully compensated the time investments Assumed that offsets due to the reintegration of beneficiaries only applied to the steady state Food costs are a positive cost to the education sector and a negative cost to the household sector Prev Sci (2018) 19:716–727 723 Table 3 Per child costs of HPSF and PAS for a hypothetical steady state Cost items for the steady Activities Volume (per school) Unit price Stakeholder perspective state Education Household and Labor and Healthcare Societal leisure social perspective security HPSF PAS HPSF PAS HPSF PAS HPSF PAS HPSF PAS Personnel A 1 Program coordinator Coordination 4 schools: 0.25 FTE €112,000/FTE €21 €21 A 1 School project leaders Coordination 0.25 FTE €65,000/FTE €49 €49 B B Volunteers Assisting during lunch Financial compensation: €38 €24 €53 €17 break and activities HPSF: €12,645 PAS: €8145 Time investment Time investment €14/hour HPSF: 12 volunteers, 1 h/day, 4 days/week (upper grades 5 days/week), 40 weeks PAS: 7 volunteers, 1 h/day, 3 days/week (upper grades 4 days/week), 40 weeks Primary caregivers Parental evaluation 5 times/year, 1 h, 10 persons €14/h €2 €2 committee Value of the extended 0.5 h freed up, 4 times/week 2 €14/hour €− 562 €− 562 school hours children/household 6 C C Beneficiaries of Preparing lunches as 1 person, 15 h/week Income level: €1525/month €− 3 €0 €− 22 €0 unemployment benefits part of reintegration Unemployment benefits: 7 6 to the labor market €1320/month ; €8.80/hour Income taxes: 36.55% 1 D D External parties from the Giving workshops €6675 €20 €20 €0 €0 leisure sector 1 E E Pedagogical staff from Guiding lunch break HPSF: 12 persons, 2 h/day, €65,000/FTE €524 €204 €0 €0 childcare partners and activities 4 days/week (upper grades 5 days/week), 40 weeks PAS: 8 persons, 1.5 h/day, 3 days/week (upper grades 4 days/week), 40 weeks Materials A F F Food (including personnel from caterer) HPSF: 4 times/week, 40 weeks Cost: €2/child/day Offset: €320 €0 €− 298 €0 €1.86/day Curriculum materials 1 set per year €2500/school €7 €7 Monitoring equipment 1 survey €1200/survey €4 €4 Total costs Net costs (per child/year) €982 €328 €− 807 €− 544 €− 22 €0 €0 €0 €153 €215 724 Prev Sci (2018) 19:716–727 was included, assuming that the time freed up for the primary caregiver led to productivity gains. Previous cost-effectiveness studies on after-school interventions al- so accounted for the cost offsets on after-school care (Cradock et al. 2017; Wang, Sekine, Chen, Yamagami, and Kagamimori 2008). However, this approach can be debated as it only incorporates the costs of after-school care. Providing after-school care may also lead to gains in utility or wellbeing in primary caregivers (van den Berg, Brouwer, and Koopmanschap 2004). Combining these utility losses with the productivity gains would result in a lower cost offset. In addition, time offsets may not fully be translated in productivity gains. The scenario analysis showed that the offsets of the extended school day were of major influence on the intervention costs. Future re- search on the value of after-school care is warranted to be able to generate cost estimates that reflect true social opportunity costs. Assumptions can also apply to the context of the cost analysis. Cost calculations are often made for interven- tions in a steady state, meaning that the intervention is implemented at its full effectiveness potential. These cost estimates can be used in economic evaluations to examine the full cost-effectiveness potential of interventions. On the other hand, estimating the intervention costs with the associated learning curves may be more relevant to inform implementation processes. Given the time and resource consuming processes for assessing learning curves, sever- al studies assumed that interventions operated at steady- state conditions by excluding the one-off costs for start-up (Cobiac, Vos, and Veerman 2010; Moodie et al. 2010; Moodie et al. 2011). The resulting estimates do not take account of the potential learning effects and cost efficien- cies (e.g., changes in time investments, improved capacity utilization, reduction in unit costs) that may occur over time (Anderson et al. 2007). In this study, stakeholders were consulted to define the expected learning curves and efficiency improvements. With both approaches (e.g., only excluding one-off costs versus defining a hy- pothetical steady state), the interventions costs may be underestimated or overestimated. The intervention costs should be monitored to examine whether the approaches can be used to produce accurate cost estimates. The costs of HPSF and PAS were estimated for various perspectives in order to examine the cost distribution among stakeholders groups. The results showed large un- balances between the costs for education sector and the household sector (e.g., education perspective, €8.7/day; household perspective, €− 6.0/day (HPSF, first year)). In the Netherlands, the costs of the usual education program amounts to €32, per child per day (updated to 2016 prices) (Ministerie voor Onderwijs Cultuur en Wetenschap (OCW) 2014). The delivery of HPSF Table 3 (continued) Personnel €651 €317 €− 510 €− 544 €− 22 €0 €0 €0 €120 €− 226 Materials €331 €11 €− 298 €0 €0 €0 €0 €0 €33 €11 Net costs (per child/day) €6.1 €2.1 €− 5.0 €− 3.4 €− 0.1 €0 €0 €0 €1.0 €− 1.3 HPSF Healthy Primary School of the Future; PAS Physical Activity School; FTE full-time equivalent. Discrepancies between the sum of cost items may be due to rounding. Budget “the Healthy Primary School of the Future” Productivity costs of paid labor (Zorginstituut Nederland 2015) Accounting data “the Healthy Primary School of the Future” Productivity costs of unpaid labor (Zorginstituut Nederland 2015) Household expenses on children’s lunches (NIBUD 2017) Minimum wage (“Minimumloon 2016”) Unemployment benefits (Rijksoverheid 2016) Income tax (Belastingdienst 2016) Steady state assumption Value of time investment minus financial compensation Household: income level minus unemployment benefits; social security: savings from unemployment benefits and earnings from income taxes D–E Financial contributions fully compensated the time investments Food costs are a positive cost to the education sector and a negative cost to the household sector Prev Sci (2018) 19:716–727 725 Table 4 Scenario analysis (per child societal costs) Scenario Description First year after Hypothetical implementation steady state Base-case HPSF: €429 (€2.7/day) PAS: €− 55 (€− 0.3/day) 1.1. Hypothetical A situation that occurs on the long run, HPSF: €153 (€1.0/day) steady state when interventions are delivered at PAS: €− 214 (€− ;1.3/day) their full capacity and learning curves, and efficiency improvements do not longer occur. Assumptions were defined by stakeholders. 2.2. Pedagogical A scenario analysis to examine areas for further HPSF: €− ;95 (€− HPSF: €− ;371(€− staff scenario cost reductions. Lunch breaks and activities were ;0.6/day) ;2.3/day) guided by teaching assistants instead of pedagogical PAS: €− ;259 (€− AS: €− ;417 (€− ;2.6/day) staff from external childcare partners. ;1.6/day) 2.3. Efficiency scenario Views of stakeholders about the potential areas HPSF: €353 (€2.2/day) HPSF: €88 (€0.6/day) for further cost reductions PAS: €− ;107 (€− PAS: €− ;254 (€− ;1.6/day) ;0.7/day) 2.4. Extended school day A scenario analysis to examine the uncertainty about the HPSF: €991 (€6.2/day) HPSF: €715 (€4.5/day) scenario cost item. It was assumed that the time offset for the PAS: €507 (€3.2/day) PAS: €348 (€2.2/day) primary caregiver due to the extended school day (saving) was fully compensated by the utility loss from being able to provide after-school care. HPSF Healthy Primary School of the Future; PAS Physical Activity School (incremental cost, €8.7/child/day) and PAS (incremental than the total of the general per child fees for supervision cost, €4.0/child/day) is associated with a 27 and 13% in- during the lunch break (€1/child/day) and the household crease in the daily educational expenses, respectively. expenses that are usually required for lunches and morn- These results indicate that cost redistributions among ing snacks for children in the regular school curriculum stakeholder groups might be required to facilitate and sup- (€1.8/child/day) (NIBUD 2017). port the implementation of the interventions. For example, In addition to the offsets of the extended school day, the it can be argued that a per child fee may be charged to sensitivity analysis showed that personnel costs had a major primary caregivers who can afford it, as the societal costs influence on the costs of the HPSF and PAS interventions. In of €2.7 (HPSF first year after implementation) are lower several countries, primary education, childcare, and after- Fig. 1 Tornado diagram (€/child/year). HPSF Healthy Primary School of the Future, PAS Physical Activity School 726 Prev Sci (2018) 19:716–727 Fig. 2 Tornado diagram (€/child/day). HPSF Healthy Primary School of the Future, PAS Physical Activity School LLMV00; and by Maastricht University. None of the funding bodies school care are becoming more integrated with each other had a role in the design of the study or the writing of this manuscript. (Oostdam, Tavecchio, Nøhr, and Ex 2014). A close collabo- Nor will the funding bodies have a role in the future data collection, ration between multiple professionals and a redistribution of analysis, interpretation of data, and the writing of publications. their tasks (e.g., supervision during lunch break, assisting in the classroom, providing after-school care) may possibly lead Compliance with Ethical Standards the way to obtain further cost reductions and facilitate the implementation of lifestyle interventions in primary schools. Conflict of Interest The authors declare that they have no conflict of interest. The results showed that the delivery of HPSF is a more costly alternative compared to PAS. In addition to the cost- Ethical Approval For this type of study, formal consent is not required. ing study, the relative effectiveness of both interventions on outcomes including health, wellbeing, and educational out- Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http:// comes will be evaluated to examine whether the additional creativecommons.org/licenses/by/4.0/), which permits unrestricted use, costs of HPSF and PAS represent value for money when distribution, and reproduction in any medium, provided you give compared to the regular school curriculum. appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Conclusions References The template proved helpful in calculating the costs of school- Anderson, I., Lopez, A., Wilson, A., Vos, T., Doran, C., Barendregt, J., based lifestyle interventions and distinguishing between vari- Bertram, M. (2007). Assessing cost-effectiveness in the prevention ous stakeholder perspectives. The results showed that the de- of non-communicable disease (ACE–Prevention) Project 2005–09. livery of HPSF is a more costly alternative compared to PAS. Retrieved from https://public-health.uq.edu.au/files/622/ACE-P_ Econ_Protocol_no_append.pdf. The education sector incurred the highest per child costs, Barrett, J. L., Gortmaker, S. L., Long, M. W., Ward, Z. J., Resch, S. C., which were fully compensated by the savings in the household Moodie, M. L., & Cradock, A. L. (2015). Cost effectiveness of an sector for PAS, and almost fully compensated for HPSF. elementary school active physical education policy. American Whether the additional costs of HPSF over PAS represent Journal of Preventive Medicine, 49, 148–159. https://doi.org/10. 1016/j.amepre.2015.02.005. value for money depends on the relative effectiveness of the Belastingdienst. (2016). Overzicht tarieven en schijven. Retrieved August interventions. 28th, 2017, from https://www.belastingdienst.nl/wps/wcm/connect/ bldcontentnl/belastingdienst/prive/inkomstenbelasting/ Acknowledgements The authors would like to thank Andrew Simons, heffingskortingen_boxen_tarieven/boxen_en_tarieven/overzicht_ Peter van Bloemendaal, Gert-Jan van Druten, Maurice van Graven, tarieven_en_schijven/. Marco Reumkens, Ruben Smits, and Diana Vaessen for their input and Cobiac, L. J., Vos, T., & Veerman, J. L. (2010). Cost-effectiveness of contribution to this study. interventions to promote fruit and vegetable consumption. PLoS One, 5, e14148. https://doi.org/10.1371/journal.pone.0014148. Funding This study was funded by the Limburg provincial authorities, Cradock, A. L., Barrett, J. L., Kenney, E. L., Giles, C. M., Ward, Z. J., Long, M. W., & Gortmaker, S. L. (2017). Using cost-effectiveness Project Number 200130003; by Friesland Campina, Project Number Prev Sci (2018) 19:716–727 727 analysis to prioritize policy and programmatic approaches to phys- for primary school children—TravelSMART Schools Curriculum program. Journal of Physical Activity and Health, 8,503–515. ical activity promotion and obesity prevention in childhood. Preventive Medicine, 95(Suppl), S17–s27. https://doi.org/10.1016/ Moodie, M. L., Herbert, J. K., de Silva-Sanigorski, A. M., Mavoa, H. M., j.ypmed.2016.10.017. Keating, C. L., Carter, R. C., Swinburn, B. A. (2013). The cost- Drost, R. M. W. A., Paulus, A., Ruwaard, D., & Evers, S. (2014). effectiveness of a successful community-based obesity prevention Handleiding Intersectorale Kosten en Baten van (Preventieve) program: the be active eat well program. Obesity (Silver Spring), 21, Interventies: Maastricht University, Faculty of Health, medicine 2072–2080. doi: https://doi.org/10.1002/oby.20472. and life sciences CAPHRI, school for public health and primary care Nationaal Instituut voor Budgetvoorlichting [NIBUD]. (2017). Wat geeft care, Department of Health Services. u uit aan voeding? Retrieved May 12th, 2017, from https://www. Hayman, J. (2016). Personal, social, health and economic education: the nibud.nl/consumenten/wat-geeft-u-uit-aan-voeding/. bridge between public health and education. International Journal Oostdam, R., Tavecchio, L., Nøhr, K., & Ex, C. (2014). Een integraal of Health Promotion and Education, 54,157–161. https://doi.org/ kindcentrum als open leer- en ontwikkelingsgemeenschap; Tijd voor 10.1080/14635240.2015.1111770. een samenhangende visie op opvoeding, educatie en opvang. John, J., Wenig, C. M., & Wolfenstetter, S. B. (2010). Recent economic Pedagogiek, 34,60–73. https://doi.org/10.5117/PED2014.1.OOST. findings on childhood obesity: cost-of-illness and cost-effectiveness Oosterhoff, M., Bosma, H., Van Schayck, C. P., Evers, S. M. A. A., of interventions. Current Opinion in Clinical Nutrition and Dirksen, C. D., & Joore, M. A. (2017). Recommendations for eco- Metabolic Care, 13,305–313. https://doi.org/10.1097/MCO. nomic evaluations on lifestyle interventions in primary schools: is- 0b013e328337fe18. sues and ways forward. Manuscript submitted for publication. Kersley, H., & Knuutila, A. (2011). The benefits of procuring school Rijksoverheid. (2016). Uitkeringsbedragen per 1 januari 2016.Retrieved meals through the food for life partnership. London: New August 28th, 2017, 2017, from https://www.rijksoverheid.nl/ Economics Foundation Retrieved from http://neweconomics.org/ documenten/publicaties/2015/12/17/bedragen-uitkeringen-2016. 2011/05/benefits-procuring-school-meals-food-life-partnership/. Sobol-Goldberg, S., Rabinowitz, J., & Gross, R. (2013). School-based obe- Langford, R., Bonell, C. P., Jones, H. E., Pouliou, T., Murphy, S. M., sity prevention programs: a meta-analysis of randomized controlled Waters, E., Campbell, R. (2014). The WHO Health Promoting trials. Obesity, 21, 2422–2428. https://doi.org/10.1002/oby.20515. School framework for improving the health and well-being of stu- Tang, K. C., Nutbeam, D., Aldinger, C., St Leger, L., Bundy, D., dents and their academic achievement. Cochrane Database of Hoffmann, A. M., & Heckert, K. (2009). Schools for health, educa- Systematic Reviews (4). doi: ARTN CD008958. tion and development: a call for action. Health Promotion McAuley, K. A., Taylor, R. W., Farmer, V. L., Hansen, P., Williams, S. International, 24,68–77. https://doi.org/10.1093/heapro/dan037. M., Booker, C. S., & Mann, J. I. (2010). Economic evaluation of a te Velde, S. J., Lennert Veerman, J., Tak, N. I., Bosmans, J. E., Klepp, K. community-based obesity prevention program in children: the I., & Brug, J. (2011). Modeling the long term health outcomes and APPLE project. Obesity (Silver Spring), 18,131–136. https://doi. cost-effectiveness of two interventions promoting fruit and vegeta- org/10.1038/oby.2009.148. ble intake among schoolchildren. Economics & Human Biology, 9, Mei, H., Xiong, Y., Xie, S., Guo, S., Li, Y., Guo, B., & Zhang, J. (2016). 14–22. https://doi.org/10.1016/j.ehb.2010.09.001. The impact of long-term school-based physical activity interven- van den Berg, B., Brouwer, W. B. F., & Koopmanschap, M. A. (2004). tions on body mass index of primary school children—a meta- Economic valuation of informal care. The European Journal of analysis of randomized controlled trials. BMC Public Health, 16, Health Economics, formerly: HEPAC, 5,36–45. https://doi.org/10. 1–12. https://doi.org/10.1186/s12889-016-2829-z. 1007/s10198-003-0189-y. Meng, L., Xu, H., Liu, A., van Raaij, J., Bemelmans, W., Hu, X., & Ma, G. (2013). The costs and cost-effectiveness of a school-based com- Wang, H., Li, T., Siahpush, M., Chen, L. W., & Huberty, J. (2017). Cost- prehensive intervention study on childhood obesity in China. PLoS effectiveness of ready for recess to promote physical activity in One, 8, e77971. https://doi.org/10.1371/journal.pone.0077971. children. Journal of School Health, 87,278–285. https://doi.org/ Middleton, G., Evans, A., Keegan, R., Bishop, D., & Evans, D. (2014). 10.1111/josh.12495. The importance of parents and teachers as stakeholders in school- Wang, H., Sekine, M., Chen, X., Yamagami, T., & Kagamimori, S. based healthy eating programs. In Y. B. Larock & D. C. Gustave (2008). Lifestyle at 3 years of age and quality of life (QOL) in (Eds.), Health education: parental and educators' perspectives, cur- first-year junior high school students in Japan: results of the rent practices and needs assessment. Health Care Issues, Costs and Toyama Birth Cohort Study. Quality of Life Research, 17, 257– Acces. New York, NY: NOVA Science Publishers Retrieved from 265. https://doi.org/10.1007/s11136-007-9301-6. http://eprints.lincoln.ac.uk/11965/. Warwick, I., Aggleton, P., Chase, E., Schagen, S., Blenkinsop, S., Schagen, Ministerie voor Onderwijs Cultuur en Wetenschap [OCW]. (2014). I., & Eggers, M. (2005). Evaluating healthy schools: perceptions of Kerncijfers 2009–2013.Retrieved from https://www.rijksoverheid. impact among school-based respondents. Health Education Research, nl/documenten/jaarverslagen/2012/05/16/kerncijfers-2007-2011. 20,697–708. https://doi.org/10.1093/her/cyh024. Moodie, M. L., Carter, R. C., Swinburn, B. A., & Haby, M. M. (2010). Waters, E., Gibbs, L., Tadic, M., Ukoumunne, O. C., Magarey, A., Okely, The cost-effectiveness of Australia’s active after-school communi- A. D., . . . Gold, L. (2017). Cluster randomised trial of a school- ties program. Obesity (Silver Spring), 18,1585–1592. https://doi. community child health promotion and obesity prevention interven- org/10.1038/oby.2009.401. tion: findings from the evaluation of fun ‘n healthy in Moreland! Moodie, M. L., Haby, M., Galvin, L., Swinburn, B., & Carter, R. (2009). BMC Public Health, 18, 92. doi: https://doi.org/10.1186/s12889- Cost-effectiveness of active transport for primary school children— 017-4625-9. Walking School Bus program. International Journal of Behavioural Willeboordse, M., Jansen, M. W., van den Heijkant, S. N., Simons, A., Nutrition and Physical Activity, 14,63. https://doi.org/10.1186/ Winkens,B.,de Groot,R.H.M.,et al.(2016).The healthyprimary 1479-5868-6-63. school of the future: study protocol of a quasi-experimental study. BMC Moodie, M. L., Haby, M. M., Swinburn, B., & Carter, R. (2011). Public Health, 16,1–13. https://doi.org/10.1186/s12889-016-3301-9. Assessing cost-effectiveness in obesity: active transport program
Prevention Science – Springer Journals
Published: May 31, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera