# Is a voluntary healthy food policy effective? evaluating effects on foods and drinks for sale in hospitals and resulting policy changes

## TL;DR

A voluntary healthy food policy in New Zealand hospitals failed to significantly improve the availability of healthier food options, with unhealthy items still dominating.

## Contribution

This study evaluates the real-world effectiveness of a voluntary healthy food policy in healthcare settings using systematic audits.

## Key findings

- No organization met the policy's criteria for healthy food availability, with 38.9% of items rated as unhealthy.
- Organizations adopting the voluntary policy had fewer unhealthy items compared to those with their own policies, but unhealthy items still dominated.
- Unhealthy and less healthy items were more likely to be promoted and priced lower than healthier options.

## Abstract

Healthy food and drink guidelines for public sector settings can improve the healthiness of food environments. This study aimed to assess the implementation and impact of the voluntary National Healthy Food and Drink Policy (the Policy) introduced in New Zealand in 2016 to encourage provision of healthier food and drink options for staff and visitors at healthcare facilities.

A customised digital audit tool was used to collate data on foods and drinks available for sale in healthcare organisations and to systematically classify items as green (‘healthy’), amber (‘less healthy’), or red (‘unhealthy’) according to Policy criteria. On-site audits were undertaken between March 2021 and June 2022 at 19 District Health Boards (organisations responsible for providing public health services) and one central government agency. Forty-three sites were audited, encompassing 229 retail settings (serviced food outlets and vending machines). In total, 8485 foods/drinks were classified according to Policy criteria. The primary outcome was alignment with Policy guidance on the availability of green, amber, and red category food/drink items (≥ 55% green and 0% red items). Secondary outcomes were proportions of green, amber, and red category items, promotional practices, and price. Chi-square tests were used to compare results between categorical variables.

No organisation met the criteria for alignment with the Policy. Across all sites, 38.9% of food/drink items were rated red (not permitted), 39.0% were amber, and 22.1% were green. Organisations that adopted the voluntary Policy offered more healthy foods/drinks than those with their own organisational policy, but the proportion of red items remained high: 32.3% versus 47.5% (p < 0.0001). About one-fifth (21.3%) of all items were promoted, with red (24.6%) and amber (22.2%) items significantly more likely to be promoted than green items (14.0%) (p < 0.001). Green items were also significantly more costly on average (NZ$6.00) than either red (NZ$4.00) or amber (NZ$4.70) items (p < 0.0001).

Comprehensive and systematic evaluation showed that a voluntary Policy was not effective in ensuring provision of healthier food/drink options in New Zealand hospitals. The adoption of a single, mandatory Policy, accompanied by dedicated support and regular evaluations, could better support Policy implementation.

The online version contains supplementary material available at 10.1186/s12916-025-04122-x.