Table 1.

Evaluation design frame for interventions chosen

Interventions1,2 and Setting	Target Population3	Cost Results			Key cost issues
		Gross Cost	Net Cost (Net Saving)	Cost Child4
1. Active After School Communities Program [Child Care 5]. Runs 8 weeks in each of the 4 school terms.	Primary school children in Prep to Grade 6 (age 5-11 years). Number ≈ 99 000	$40.3 m [UI: $28.6 m-$56.2 m]6	$36.5 m [UI: $24.9 m-$52.6 m]	$407	i) Extensive & "lumpy" salary costs, particularly for regional physical activity co-ordinators; ii) sub-optimal capacity utilisation; iii) cost data modelled, not empirically-based; and iv) BMI outcomes not commensurate with high cost structure.
2. Multi-faceted program, including education to improve nutrition & increase physical activity, without an active physical education component [School-based].	Children in primary school Grades 1 and 2 (commencing in Grade 1, age 6 years). Number ≈ 114 630	$24.3 m [UI: $12.6 m-$39.2 m]	$9.0 m [UI: net saving of $9.1 m to net cost of $31.7 m]	$211	i) Costs over 2 year-period taken as representative 'annual cost' as reflects concomitant cohorts; ii) included central & school coordination costs for national program, but not teacher time (as integrated into curriculum); iii) assumed uptake by schools does not vary by type of school (public or private); and iv) parent involvement encouraged as part of program but not costed.
3. Multi-faceted program, including education to improve nutrition & increase physical activity, with an active physical education component [School-based].	Children in primary school Grades 1, 2 & 3 (commencing Grade 1, age 6 years). Number ≈ 114 630	$54.2 m [UI: $26.9 m-87.5 m)	($14.0 m) [UI: net saving of $41.9 m to net cost of $1.3 m]	$473	i) Costs over 3 year-period taken as representative 'annual cost' as reflects concomitant cohorts; ii) included central & school coordination costs for national program, but not teacher time (as integrated into curriculum); iii) physical activity component may pose problem for primary schools without specialist physical education teachers; iv) assumed uptake by schools does not vary by type of school; and v) parent involvement encouraged but not costed.
4. Multi-faceted program [School-based] targeted at overweight and obese children.	Overweight or obese children aged 7-10 years (Grades 2-5) at combined primary/secondary school. Number ≈ 17 000 over 4 years (4 200 each year)	$2.2 m [UI: $1.2 m to $4.1 m]	($1.2 m) [UI: net saving of $5.7 m to net cost of $0.38 m]	$129	i) Modelled as implemented over 4 years rather than implementing it to everyone eligible every 4 years; ii) involves a peer-led program using 8th grade students, supported by counsellors, to help obese children in grades 2-5; iii) counsellors costed as publicly funded psychologists employed on part-time basis (different to trial).
5. Education program to reduce consumption of carbonated (fizzy) drinks [School-based].	Children in primary school 2 to 6 (age 7-11 years). Number ≈ 595 000 implemented over 5 years (119,000 each year)	$16.6 m [UI: $7.6 m -$32.2 m]	($26.7 m) [UI: net saving of $112.7 m to net cost of $32.0 m]	$28	i) Capacity calculated on seeing 1/5th of schools each year, not all schools every year; ii) assumes each child receives intervention once during primary school; and iii) assumption of no additional school staff costs as sessions presented by trained project staff.
6. Education program to reduce TV viewing [School-based].	Children in primary school Grades 3 & 4 (age 8-10 years). Number ≈ 268 600	$27.7 m [UI: $12.7 m -$43.3 m]	($43.8 m) [UI: net saving of $81.8 m to net saving of $6.6 m]	$103	i) Modelling included national/state project officers to implement national program and full training costs for teachers, but not teacher time in the classroom (as integrated into curriculum); ii) 50% of schools participate in any one year; iii) assumed uptake by schools does not vary by type of school (public or private); and iv) parent involvement encouraged as part of program but not costed.
7. TravelSmart Schools [Schools/neighbourhoods & community organisations7]	Children in primary school Grades 5 & 6 (age 10-11 years). Number ≈ 267 700	$13.3 m [UI: $6.9 m -$22.8 m]	$12.58 m [UI: $6.1 m- $222.1 m]	$50	i) Large impact on cost-effectiveness from attributing a share of intervention costs to broader congestion, community & environmental objectives; ii) capacity utilisation not an issue for this intervention as 90% of costs are variable; and iii) cost data mostly modelled, not a strong empirical base.
8. Walking School Bus [Schools/neighbourhoods & community organisations].	Primary school children in Prep to Grade 2 (age 5-7 years). Number ≈ 7 840	$22.8 m [UI: $16.6 m-$30.9 m]	$22.53 m [UI: $16.35 m- $30.47 m]	$2908	i) Extensive set-up & overhead costs; ii) poor capacity utilisation; iii) attribution of costs to non obesity objectives; iv) empirical data coming from early developmental period of WSB program in Vic, Australia.
9. Reduction of TV advertising of high fat and/or high sugar foods & drinks to children [Media & marketing].	All Australian children aged 5-14 years. Number ≈ 2.4 million	$0.13 m [UI: $0.12 m-$0.14 m]	($299 m) [UI: net saving of between $133 m- $484 m]	$0.54	Key issue is exclusion of costs other than cost of monitoring/enforcing compliance with revised regulation. Excluded costs include: changing the regulations; any additional food costs to families in switch from non-core to core; impact on revenue stream of advertising companies & producers of non-core foods.
10. Family-based GP program targeted at overweight and moderately obese children [Primary care services g]	Overweight or moderately obese children aged 5-9 years Number ≈ 9 685	$6.3 m [UI: $5.3 m-$7.4 m]	$2.95 m [UI: net saving of $1.06 m to net cost of $7.0 m]	$650	i) Intervention design includes costs of family participation, but no additional benefits are included from weight loss to family members other than the child; ii) potential for piggy-backing this intervention into other GP-based interventions; iii) low % of fixed costs means cost drivers for affordability (patient numbers), do not impact much on cost-effectiveness; and iv) majority of cost incurred by government, but cost impact on family still significant, with time costs a major factor.
11. Family-based targeted program for obese children [Primary care services9 + hospital setting delivery by multidisciplinary team]	Obese children aged 10-11 years. Number ≈ 5 800	$11.0 m [UI: $6.8 m-$18.3 m]	($4.0 m) [UI: net saving of $19.0 m to net cost of $2.4 m]	$1,896	i) Recruitment component adjusted from screening in schools in RCT, to opportunistic recruitment via GPs; ii) assumed 50% of 6,000 GPs already have calibrated scales and stadiometers necessary to measure weight; iii) 'intention to treat' approach adopted for costing (i.e. non completion still involved intervention costs), but full completion of visits required before benefits attributed.
12. Orlistat therapy for obese adolescents [Primary care services9].	Obese adolescents aged 12-16 years. Number ≈ 3 256	$6.3 m [UI: $1.4 m-$20.0 m]	$4.9 m [UI: $1.1 m-$15.9 m]	$1,935	i) Modelling incorporated opportunistic recruitment in Australian primary care setting (by GPs with dieticians providing dietary advice) and conservative adherence rates (65%); ii) only patients responsive to Orlistat assumed to continue past 2-week run-in period; iii) costs to parents in accompanying adolescent included; iv) high proportion of costs falling on patients/families (as Orlistat not on Pharmaceutical Benefits Schedule) impacts on access.
13. Laparoscopic adjustable gastric banding (LAGB) for morbidly obese adolescents [Hospital10].	Severely obese adolescents, aged 14-19 years, with private health insurance Number ≈ 4 120	$130 m [UI: $52 m-$265 m]	$53.4 m [UI: $20.1 m-$116.8 m]	$31,553	i) Definition of 'current practice'; ii) inclusion in costing of ongoing follow-up, including regular consultations and 2 LAGB replacements over lifetime; iii) cost data coming from early developmental period of LAGB (case series of 28 patients) extrapolated to eligible adolescent population; and iv) management of co-morbidities assumed to be same for intervention and comparator.
Protocol issues common to all interventions: i) Inclusion of time costs for adults/carers, but not children/adolescents; ii) exclusion of production gains/loses; iii) exclusion of unrelated costs in rest of life; iv) 'steady-state' costing, with program modeled in accordance with efficacy potential, assuming trained staff and infrastructure available; v) costs offsets based on mean reduction in BMI continuing over life of the child; vi) early set-up & development costs excluded (i.e. costs incurred before intervention commences, such as development of training packages); vii) annuitisation of capital, including human capital costs like training; and viii) full pathway costing, including recruitment and coordination.

Notes:

¹ Current practice comparator defined as "no intervention" as programs either focussed on children previously inactive and/or minimal activity previously existed.

² The intervention period is defined as one representative year of "steady-state" operation, with "rest-of-life" modelling for all associated costs and benefits.

³ Number of children participating in the intervention based on Australian population figures in 2001 and likely take-up rates. For some interventions, not all of the children/adolescents participating receive a health benefit from the intervention.

⁴Cost per child estimates do not include cost offsets.

⁵ Includes child care centres, family day care and outside school hours care.

⁶ 95% uncertainty interval

⁷ Includes State/Territory government, local government, community groups, recreation and sporting bodies, and private organisations.

⁸ Current regulations limit adverts to 5 minutes every 30 minutes during 5 hrs of designated child slots & prohibits advertisements during 2.5 hrs per week of designated pre-school timeslots.

⁹ Includes general medical practice (GPs), community health centres and other community-based and private-sector services.

¹⁰ The hospital setting was not included in 'Healthy Weight 2008' [21], but this clinical intervention was included in the project for purposes of comparison and benchmarking.