Unclear conclusions on SARS-CoV-2 elimination versus mitigation – Authors' reply

We thank Christian Gold for his interest and thoughtful Correspondence. Country strategies to control SARS-CoV-2 vary across many parameters but can be crudely grouped into countries that aim for elimination (ie, maximum action to control SARS-CoV-2 and stop community transmission as quickly as possible) or mitigation (ie, action increased in a stepwise, targeted way to reduce cases so as not to overwhelm health-care systems).¹ Our classification is based on a qualitative analysis of policy documents, communication by government officials and advisers (eg, through press releases or other public communication), and lockdown dates and their severity with respect to case and mortality reports. Lockdown dates and severity give a good proxy on the overall response that a country showed to the pandemic, including behavioural changes that are possibly not directly linked to restrictions or policy interventions. Two authors jointly reviewed the policy documents and data, classifying each country into one of the two categories. Consensus for any disagreement was reached through discussion and final agreement from the whole author group. The classification of a country's pandemic response into one of two categories is by virtue imperfect, particularly as responses change over time (eg, we considered only the first 12 months of the COVID-19 pandemic) and also vary within countries (eg, many countries used a zoning approach, adopting public health measures dependent on the epidemiological situation of each city, state, or region).² Notably, Japan used local interventions as early as March, 2020, aiming to eliminate virus clusters by spatially targeted interventions.³

Gold suggests that a country-by-country analysis could strengthen our findings. For brevity, in the appendix we present annual data for a selection of countries that took markedly different approaches throughout the first year of the pandemic and have different geographical and political situations: Australia, New Zealand, and Vietnam (ie, classified as following an elimination aim) and Brazil, the UK, and the USA (ie, classified as following a mitigation aim). Focusing on annual data allows us to include non-Organisation for Economic Co-operation and Development countries, unlike in our original analysis,¹ and consider cumulative deaths associated with COVID-19 in 2020, year-on-year growth change in gross domestic product in 2020 compared with in 2019, and the mean stringency index throughout 2020. The stringency index was developed at the University of Oxford, Oxford, UK, and measures the strictness of lockdown-style policies. Overall, countries that opted for an elimination strategy performed better along these three metrics than did countries that opted for a mitigation strategy. Notably, Vietnam aimed for the elimination of SARS-CoV-2 and was able to protect its extensive—and usually open—land borders from reimportations of the virus.

To summarise, we agree with Gold that further analysis could clarify the varying strategies that countries used and how each country's experience with SARS-CoV-2 depends on other relevant variables, such as centrality in the international travel network (ie, how likely people are to travel to or through the country) or how strategic choices interacted (eg, whether countries that aimed for elimination underinvested in vaccines). Nevertheless, drawing analytically causal connections in policy analysis will always be partly “art and craft”.⁴ The overwhelming evidence here points towards elimination strategies having substantially outperformed mitigation approaches throughout the first year of the pandemic, when vaccines were not available.

IK is a member of the Global Preparedness Monitoring Board. JVL is a member of the Lancet COVID-19 Commission Public Health Taskforce. SV is a member of Team Halo (London School of Hygiene & Tropical Medicine, the UN, and Gavi, the Vaccine Alliance). All other authors declare no competing interests.