Technical efficiency and resilience of health systems in Japan during the COVID-19 pandemic

It is widely recognized that in resilient health systems, efficient utilization of limited resources contributes to protecting lives against adverse external shocks. In this study, we estimated technical efficiency of local health systems in Japan during the COVID-19 pandemic, identified efficiency patterns, and compared health outcomes across these patterns to assess the contribution of efficiency to local health system resilience. We conducted estimations using a true fixed-effects model within stochastic frontier analysis, based on panel data for all 47 prefectures over seven COVID-19 waves from 2020 to 2023. In the estimated production function, health services provided to hospitalized and accommodated patients with COVID-19 were positively associated with the case survival proportion after adjusting for contextual factors. The median technical efficiency of prefectures for each wave ranged from 0.75 to 0.82 before declining to 0.49 when the Omicron variant emerged in 2022, but improved to 0.88 and 0.89 in subsequent waves. We categorized the prefectures into three efficiency pattern groups based on the degree of deterioration and recovery of technical efficiency in response to the Omicron shock: (1) small deterioration, (2) large deterioration and large recovery, and (3) large deterioration and small recovery. The case survival proportion did not differ between groups, suggesting that in addition to improving efficiency under fixed health-service inputs, other resilience strategies, such as increasing health-service inputs and leveraging time-invariant regional strengths, may be adopted at the local level. These results indicate that local health system resilience can be achieved through various strategies not limited to efficiency.