Abstract
Background: During the COVID-19 pandemic, governments faced uncertainty about how to balance the intensity of non-pharmaceutical interventions with economic costs. Existing national preparedness indices describe pandemic preparedness capacity but do not indicate how optimal intervention intensity should vary with national conditions.Methods: We analysed COVID-19 case data from 110 countries during the early pandemic response with an integrated health–economic approach combining mechanistic transmission modelling with multi-objective optimisation considering three vaccine implementation timings. We evaluated national indicators to assess which indicators best predict cost-optimal intervention intensity.Findings: No single national indicator significantly predicts the cost-optimal intervention level, but Socio-demographic index (SDI) and GDP show relatively strong associations; higher SDI and GDP correspond to lower cost-optimal intervention intensity. The excess cost ratio (the relative increase over the optimal cost) averaged 65·5% across 110 countries (median 29·9%). There was an unstable under-intervention zone in excess cost near the cost-optimal solution. The safe intervention zone to reduce excess cost and its uncertainty is 6-20 percentage points more reduction in transmission compared to the theoretical cost-optimal solution.Interpretation: A risk-averse public health strategy is to target interventions stronger than the theoretical optimum to avoid economic burden. Implementing a strong intervention policy has less uncertainty about excess costs than implementing a weak policy. Especially, low- and middle-income countries require a higher level of intervention to minimise the economic burden than high income countries.