Even if public health interventions are successful at reducing the spread of COVID-19, there is no guarantee that they will
bring net benefits to the society because of the dynamic nature of the pandemic, e.g., the risk of a second outbreak if those
interventions are stopped too early, and the costs of a continued lockdown. In this analysis, a discrete-time dynamic model
is used to simulate the effect of reducing the effective reproduction number, driven by lockdowns ordered in March 2020 in
four European countries (UK, France, Italy and Spain), on QALYs and hospitalisation costs. These benefits are valued in
monetary terms (€30,000 per QALY assumed) and compared to productivity costs due to reduced economic activity during
the lockdown. An analysis of the optimal duration of lockdown is performed where a net benefit is maximised. The switch
to a soft lockdown is analysed and compared to a continued lockdown or no intervention. Results vary for two assumptions
about hospital capacity of the health system: (a) under unlimited capacity, average benefit ranges from 8.21 to 14.21% of
annual GDP, for UK and Spain, respectively; (b) under limited capacity, average benefits are higher than 30.32% of annual
GDP in all countries. The simulation results imply that the benefits of lockdown are not substantial unless continued until
vaccination of high-risk groups is complete. It is illustrated that lockdown may not bring net benefits under some scenarios
and a soft lockdown will be a more efficient alternative from mid-June 2020 only if the basic reproduction number is maintained
low (not necessarily below 1) and productivity costs are sufficiently reduced.
