info@efcni.org

Measures against the COVID pandemic did not affect rates of preterm births in three Scandinavian countries

© unsplash.com

 A study from Denmark, Norway, and Sweden combined data from each country’s birth registry and evaluated whether the rate of preterm births was affected by lockdown measures against the COVID19 pandemic. Some prior studies indicated lower rates, but overall the results were contradictive. The new analysis by Laura Oakley and colleagues incorporated a detailed strategy to account for potential confounders and eventually could not confirm a relationship between COVID-related restriction and preterm birth rates.

Different indications exist about the effects of the measures to counteract the COVID-19 pandemic in high-income countries, besides their actual purpose of mitigating infection rates of SARS-CoV2. Effects include improved air quality, improved prevention of infections, altered health conscious behaviour on the positive side; on the negative side are reduced health care access, additional anxiety about a potential COVID-19 infection, insecurities around employment, and fewer social contacts. All of these are under discussion to affect preterm birth rates.

Previous reports on the effect of COVID-19 measures on preterm births in 2020 have been inconsistent overall, but included some evidence for a decreased preterm birth rate. In contrast to previous studies that used data from selected healthcare facilities or only regional data, the new study by Laura Oakley and team analysed data from national registries from Denmark, Norway, and Sweden. The data comprised over 1.5 million births between January 2014 and December 2020, of which nearly 900,000 were included in the analysis.

The included countries had different levels of lockdown with Norway and Denmark setting up stricter lockdown rules than Sweden. For the analysis, the start point of these measures was set to March 12, 2020, around which lockdown measures were implemented. Generally, rates of preterm birth are not constant but show some degree of both seasonal and yearly fluctuation. This fact was taken into account with a complex strategy for data analysis.

The rate of preterm births, defined as delivery before 37 weeks of pregnancy was compared between the weeks before March 12 to the weeks after March 12 of the year 2020. The same comparison was calculated for each year separately from 2014 to 2019. Finally, each of these differences from the years 2014 to 2019 were compared to the difference of 2020 to account for variability in the preterm birth rate in a so-called difference-in-difference analysis. This whole analysis was done comparing two weeks prior to post-March 12 in each year, and then repeated with four, eight, 12, and 16 weeks. All of these analyses were then repeated with different subgroups (medically indicated preterm birth, spontaneous preterm birth, and very preterm birth before 32nd week of pregnancy) to understand whether these were affected differently. All of the aforementioned analyses were first done per country (due to data privacy rules) and then the results were combined.

At large, the rates of preterm births were as low as 5.6% and 5.7% in Denmark/Norway and in Sweden, respectively, with an overall downward trend over the years. No indication in neither of all the different analyses that the preterm birth rate was affected by the restrictions active against the pandemic. In more detail, trends for Norway were negative for the longer time intervals; however, the authors did not interpret this as an effect of the COVID-19 restrictions.

Main advantages of the study are the inclusion of population-based data, the inclusion of seasonal and yearly changes in the preterm birth rate, and an in-depth analysis of the data including subgroup analysis, that was possible because of the large dataset available. All of these make the results reliable. However, the rate of extremely preterm deliveries could not be analysed separately due to low number of such events. Of note, direct effects of COVID-19 infection on pregnancy outcomes was not the scope of this analysis.

 

Paper available at: American Journal of Obstetrics and Gynecology

Full list of authors: Laura L Oakley, Anne K Örtqvist, Jonas Kinge, Anne Vinkel Hansen, Tanja Gram Petersen, Jonas Söderling, Kjetil E Telle, Maria C Magnus, Laust Hvas Mortensen, Anne-Marie Nybo Andersen, Olof Stephansson, Siri E Håberg

DOI:  https://doi.org/10.1016/j.ajog.2021.11.034