Background: To compare the likely costs and benefits of a range of potential policy interventions in Fiji
and Tonga targeted at diet-related noncommunicable diseases (NCDs), in order to support more
evidence-based decision-making.
Method: A relatively simple and quick macro-simulation methodology was developed. Logic models were
developed by local stakeholders and used to identify costs and dietary impacts of policy changes. Costs
were confined to government costs, and excluded cost offsets. The best available evidence was combined
with local data to model impacts on deaths from noncommunicable diseases over the lifetime of the
target population. Given that the modelling necessarily entailed assumptions to compensate for gaps
in data and evidence, use was made of probabilistic uncertainty analysis.
Results: Costs of implementing policy changes were generally low, with the exception of some requiring
additional long-term staffing or construction activities. The most effective policy options in Fiji and Tonga
targeted access to local produce and high-fat meats respectively, and were estimated to avert approximately
3% of diet-related NCD deaths in each population. Many policies had substantially lower benefits.
Cost-effectiveness was higher for the low-cost policies. Similar policies produced markedly different
results in the two countries.
Conclusion: Despite the crudeness of the method, the consistent modelling approach used across all the
options, allowed reasonable comparisons to be made between the potential policy costs and impacts. This
type of modelling can be used to support more evidence-based and informed decision-making about
policy interventions and facilitate greater use of policy to achieve a reduction in NCDs.