Country-specific COVID-19 incidences - A biased view of the epidemiological situation

Incidence levels – the number of new COVID-19 cases within a given time period – are currently widely used as indicators to define high-risk areas. However, the differences observed between neighbouring countries and regions in Europe are greatly influenced by differences in testing strategies. An article published on 23 April in The Lancet Regional Health – Europe highlights the need for coherent data and a more nuanced approach to allow meaningful comparisons between countries. Written collaboratively by Prof. Paul Wilmes, spokesperson for the Research Luxembourg COVID-19 Task Force, Dr Joël Mossong and Dr Thomas Dentzer from the Luxembourg Health Directorate, this article also summarises some of the currently available figures for Luxembourg and the surrounding European countries.

Since the beginning of the pandemic, Luxembourg has pursued a broad testing strategy including mass screening of its population and systematic contact tracing: on average every resident has been tested at least 3.6 times. "At 2.6%, the country’s overall positivity rate – the percentage of all tests performed that are actually positive – has remained comparatively low," details Dr Joël Mossong, epidemiologist at the Luxembourg Health Directorate. "Meanwhile rates in other European countries vary, between 5.6% in neighbouring Germany and 17% in The Netherlands where test coverage is relatively limited."

In addition to large-scale testing, Luxembourg has also performed a representative serological sampling on a weekly basis among its residents since last November. The data up until 15 January 2021 indicates a seroprevalence of 7.7%, which by extrapolation to the entire population corresponds to 48,264 expected cases. "Compared to the 48,630 cases recorded until this date, this suggests that case detection in Luxembourg is very high, with only around 0.8% of the cases having gone undetected," explains Dr Thomas Dentzer, Chief Strategy Officer at the Luxembourg Health Directorate. While this type of information is missing for most countries, similar data indicate that in contrast 62.3% of cases might have gone undetected in Belgium.

Overall, these figures highlight that country-specific COVID-19 incidence levels might be more a reflection of different case ascertainment rates than reflect actual epidemiological risks for infection with SARS-CoV-2. Given the inherent biases linked to differing test strategies and to account for the disparities in testing coverage, a more nuanced approach not solely based on incidence levels would help to more accurately define high risk countries and regions. "Positivity rates, hospital and intensive care unit (ICU) occupancies, as well as data on seroprevalence should be factored into more inclusive risk assessments," stresses Prof. Paul Wilmes from the University of Luxembourg. The impacts of differing vaccination strategies and frequencies of new virus variants will also need to be taken into account. "The availability of coherent data about different countries, collected and shared in a well-coordinated, pan-European manner, will be fundamental to allow meaningful comparisons in the coming weeks and months," concludes Prof. Wilmes.

Reference: Paul Wilmes, Joël Mossong, Thomas G. Dentzer, Generalisation of COVID-19 incidences provides a biased view of the actual epidemiological situation, The Lancet Regional Health - Europe, Volume 5, 2021.

Press release by the Ministry of Health and the  University of Luxembourg