Skip to main content

Table 1 Sensitivity and scenario description: strategies and benefit

From: Cost effectiveness of a multi-component school-based physical activity intervention targeting adolescents: the ‘Physical Activity 4 Everyone’ cluster randomized trial

Test to be modelled Detailed assumptions Justification
Sensitivity analyses   
 (i) Variation in the intervention cost Higher estimate of the assumed opportunity cost of school staff participation in PA strategy (PAS) 4 & 5 and implementation support strategy (ISS) 1 Plausible variation in the cost
 (ii) Variation in the intervention cost Lower estimate of the assumed opportunity cost of school staff participation in PAS 4 & 5 and ISS 1 Plausible variation in the cost
 (iii) Varying the magnitude of the effect size Assumes benefit of the overall intervention varies between the calculated confidence interval of the effect size in daily minutes of MVPA Plausible variation in the effect size
 (iv) Extending the benefit of physical activity recess and lunchtime activities to students beyond the target year. Assumes benefit of PAIS 4 is extended to 10 % of students beyond the target year, with a reduced effect on daily minutes of MVPA compared to students in the target year. Reduced effect estimate was based on the accelerometer data within the recess and lunchtime segment from the efficacy trial (unpublished). It was likely these specific components of the intervention would impact students more broadly and not be isolated to those students within the evaluation cohort.
The number of additional students that may benefit from whole of school recess and lunchtime activities was conservatively estimated based on 10 % of a multiple of 3X the mean number of students in the target year (n = 132).
 (v) Extending the benefit of multiple strategies to all students Assumes benefit of PAS 1, PAS 5 and ISS 1, 2 and 3 are extended to all students (100 %) outside the target year (in Grades 7–10), with a reduced effect on daily minutes of MVPA compared to students in the target year. As above, due to the nature of the intervention strategies (teacher training, school environment and broader school community links) the intervention impact would likely not be isolated to the evaluation cohort. For example, once PE teachers are trained on how to maximise MVPA in PE, these strategies would likely be applied to all classes at no additional cost. The same assumption applies for other strategies such as a school Physical activity policy, executive support, change agent, and use of resources. As such the cost of these strategies would not increase, however we have assumed there is potential for more students to benefit from a school implementing such strategies.
The assumed effect size for the extension cohort was based on the results of the sensitivity analysis conducted within the efficacy trial (undertaken using imputation of missing data).
Scenario analysis  
 State wide roll out (current model) Total cost of the intervention is based on the current implementation support model.
Assumes benefit to 100 % of students, with an effect size based on the results of the sensitivity analysis conducted within the efficacy trial (undertaken using imputation of missing data).
The number of students (n = 254,923) is based on a calculation from 487 NSW schools with Grades 7–10.
 State wide roll out- Alternate (real world) model The total cost of the intervention is modified to reflect (a) an alternate model of school support - existing in-school teacher to support role out (1/2 day per week (0.5 FTE/ ½ day per week) and (b) a reduction in the equipment cost per school. Whilst the offer of an equipment pack was an attractive selling point for schools to consent to the intervention, evaluation of this specific strategy highlighted that schools within the intervention group were well stocked with equipment. As such, the provision equipment was not deemed an essential component of the trial. Based on this observation, the assumption that reducing the intervention costs by removing the provision of equipment, would not substantively alter the impact of the intervention.
Assumes benefit to 100 % of students, with an effect size based on the results of the sensitivity analysis conducted within the efficacy trial (undertaken using imputation of missing data).
The number of students (n = 254,923) is based on a calculation from 487 NSW schools with Grades 7–10.