In ‘A social cost–benefit framework for COVID-19 policy decisions', Heatley proposes a framework for cost–benefit analyses for policy decisions regarding COVID-19 responses. He illustrates this in a worked example, considering the New Zealand Cabinet’s decision in April 2020 to extend alert level 4 restrictions by five days. In his analysis he compares the extension to a counterfactual scenario where alert level 3 is resumed immediately, finding that this scenario outperforms the extension made by Cabinet. However, using a simple extension of Heatley’s method that better captures the stochastic dynamics towards the end of an outbreak, I find the opposite: Cabinet’s decision significantly outperforms Heatley’s counterfactual. I conclude that an improved treatment of disease dynamics better describes the costs and benefits of the policy choices facing Cabinet in April 2020. This improved framework or similar may be of use to future decision-makers.