Assessing the lockdown effect from excess mortalities

Background and AimsThe reported case numbers of COVID-19 are often used to estimate the growth rate of infections. We use the excess mortality instead to show the effect of most restrictive non-pharmaceutical interventions (mrNPIs) as compared to less restrictive NPIs (lrNPIs) concerning growth rate and death counts. MethodsWe estimate the COVID-19 growth rate for Austria, France, Germany, Italy, Netherlands, South Korea, Spain, Sweden, UK and USA from the excess mortality. We use the average growth rate obtained for Sweden and South Korea, the only countries with lrNPIs only, to estimate additional death numbers in the other countries, had mrNPIs not been applied. ResultsThe growth rate estimated from excess mortality decreased faster for countries with mrNPIs than for Sweden and South Korea, suggesting that the mrNPIs do have a non-negligible effect. Implementing lrNPIs instead of mrNPIs results in up to 3 times higher death numbers. This is not visible when the growth rate is calculated using the reported case numbers of COVID-19 instead of the excess mortality. ConclusionThe conclusion for the spreading of COVID-19 obtained from reported COVID-19 cases in previous studies are most likely biased. Using our method, a more realistic estimate of the growth rate is obtained. Conclusions made for the reproduction number derived from the reported case numbers, such as the apparent insignificance of mrNPIs (lockdowns), might therefore be wrong and will have to be reevaluated using the growth rates obtained with our method.