Reducing dietary salt intake is an essential population strategy for cardiovascular disease (CVD) prevention, but evidence on healthcare costs and outcomes is limited in Japan. We aimed to conduct a pilot economic evaluation under hypothetical scenarios of applying the salt reduction policies of England to Japan.
We examined salt reduction policies in England: media health promotion campaigns, front-of-pack labeling, and voluntary and mandatory reformulation with best-case and worst-case policy cost scenarios. We assumed that these policies were conducted in Japan for 10 years from 2019. We used published data on epidemiology and healthcare expenditures in Japan and the costs and effects of salt reduction policies in England, and defined the benefits as a decrease in national medical expenditures on CVD. We developed a Markov cohort simulation model of the Japanese population. To estimate the annual net benefits of each policy over 10 years, we subtracted monitoring and policy costs from the benefits. We adopted a health sector perspective and a 2% discount rate.
The cumulative net benefit over 10 years was largest for mandatory reformulation (best case) at 2,015.1 million USD (with costs of USD 48.3 million and benefits of USD 2063.5 million), followed by voluntary reformulation (net benefit: USD 1,895.1 million, cost: USD 48.1 million, benefit: USD 1,943.2 million), mandatory reformulation (worst case, net benefit: USD 1,447.9 million, cost: USD 1,174.5 million, benefit: USD 2,622.3 million), labeling (net benefit: USD 159.5 million, cost: USD 91.6 million, benefit: USD 251.0 million), and a media campaign (net benefit: USD 140.5 million, cost: USD 110.5 million, benefit: USD 251.0 million). There was no change in the superiority or inferiority of policies when the uncertainty of model parameters was considered.
Mandatory reformulation with the best-case cost scenario might be economically preferable to the other alternatives in Japan. In future research, domestic data on costs and effects of salt reduction policies should be incorporated for model refinement.