疫学：ヨーロッパにおけるSARS-CoV-2の伝播がロックダウンでどのように変化したか

Large-scale lockdowns and other non-pharmaceutical interventions in Europe have been successful in reducing the transmission levels of SARS-CoV-2 enough to control the epidemic growth, suggests a modelling study published online in Nature today. The estimated reductions in transmission are based on combined data from 11 European countries up to early May 2020, offering a general view of the current situation that may not fully account for differences in approaches.

From 2–29 March 2020, European countries began implementing major non-pharmaceutical methods (such as school closures and national lockdowns) to control the COVID-19 epidemic. Measuring the effectiveness of these interventions is important, given their economic and social impacts, and may indicate which course of action is needed to maintain control. Estimating the reproduction number — the average number of cases an infected person is likely to cause while they are infectious — is a useful measure. However, the reproduction number can be difficult to calculate using case data, as a larger proportion of infections are likely to go unreported. An alternative way to track an epidemic is to calculate infection levels retrospectively by analysing reported deaths. Although death data may also be subject to under- or misreporting, it is considered more reliable than case data, and can also be useful to estimate the share of unreported cases.

Seth Flaxman and colleagues use death data to infer changes in the course of the COVID-19 epidemic as a result of non-pharmaceutical interventions. They analyse data from 11 countries in Europe, including the United Kingdom, Spain, Italy, Germany and Belgium, up until 4 May 2020. They estimate that by that date between 12 and 15 million individuals in these countries have been infected with SARS-CoV-2 (3.2% to 4% of the population, with large country-to-country fluctuations). By comparing the number of observed deaths against those predicted by their model in the absence of interventions, the authors suggest that approximately 3.1 million deaths have been averted due to non-pharmaceutical measures. They calculate that the reproduction number has dropped to below one as a result of the interventions, decreasing by an average of 82%, although the values vary from country to country.

The use of pooled data and the short time separation between subsequent non-pharmaceutical interventions makes it difficult to determine the effect of individual sanctions on the suppression of the COVID-19 epidemic. One limitation of the model is that it assumes that each measure had the same effect on all countries, whereas in reality, there was variation in how effective lockdown was in different countries. However, the authors conclude that lockdown has a substantial effect in reducing the reproductive number below one and has helped to contain the spread of COVID-19 as of the beginning of May.