Economic evaluation of the national school food standards across secondary schools in The Midlands, UK (the FUEL study): methodological challenges of undertaking health economics research within non-health settings

Background: Economic evaluations of complex public health interventions are becoming increasingly important. This presents health economists with challenges of adapting methodologies originally designed for healthcare to other contexts, such as education. This study presents an economic evaluation of the UK School Food Standards (SFS), with a particular focus on the methodological challenges involved.

Methods: The economic evaluation was conducted alongside an observational study comparing the SFS-mandated secondary schools to non-mandated schools in the Midlands (UK). Costs of food provision and SFS implementation were collected directly from schools and supplemented by secondary data on schools’ catering expenditure. The outcomes included dietary intake, dental health, health-related quality of life (HRQoL), and educational performance, collected from pupils and secondary data. The analysis comprised a micro-costing, cost-consequence, and an exploratory cost-utility analysis, from school and societal perspectives.

Results: Data were collected from 36 schools and 2,543 pupils. We found mandated schools spent less on food provision compared to non-mandated schools, and pupils attending mandated schools had marginally better HRQoL, dental health, and slightly worse nutritional intake. Mandated schools performed worse according to the educational outcomes. There were large amounts of missing cost data despite repeated data collection attempts, and the results of the cost-utility analysis were uncertain.

Discussion: We found no clear evidence on the cost-effectiveness of the SFS in secondary schools, likely due to substantial variation in implementation and compliance across both mandated and non-mandated schools, as well as multiple challenges, including the COVID-19 pandemic, difficulties in collecting cost data from schools, and the complexity of the study context. This study highlights the challenges of primary cost data collection for evaluating complex interventions and the need to balance data accuracy with the resources required. As economic evaluations of school-based interventions become more common, there is a growing need to refine methods for such evaluations.