Keywords
COVID-19 - perinatal data - preterm births
Introduction
The pandemic which began in late 2019 following the spread of the infectious disease
known as COVID-19 was caused by the SARS-CoV-2 virus. The effects of the pandemic
on peripartum and
perinatal outcome data have already been studied in numerous publications and meta-analyses.
Griewing et al. (2022) recently reported an example of the impact of the COVID-19
pandemic on the provision of gynecological and obstetric services by a German tertiary
care university
hospital [1].
The spread of COVID-19 pandemic and the social and country-specific impact of the
pandemic were countered by global lockdown measures (of differing lengths and severity).
It may have had a
negative impact on overall birth rates and a positive effect on preterm birth rates
[2]. A meta-analysis already carried out in 2021,
which included 40 studies on this topic, initially found no changes in the rate of
preterm births born before the 37th week of gestation in 15 of the included studies
[3]. However, the analysis did find significant differences between low-income and high-income
countries, with the latter exhibiting a decrease
in preterm births (12 studies) and spontaneous preterm births (two studies) [4].
In an article published in 2021 in Lancet Global Health, Ochoa et al. already called
upon “the perinatal research community to collaboratively take advantage of the unique
natural experiment
provided by the COVID-19 pandemic to accelerate progress in maternal and child health
globally.” Together it would be possible to “learn from experiences from the pandemic
and start
identifying mechanisms that might contribute to a healthier start for future generations.”
[5].
To take up this suggestion, after the recent publication of an analysis of Bavarian
data [6], the Berlin perinatal data are now being
analyzed and interpreted with regard to possible effects of the pandemic on the preterm
birth rate. The main question in this context was: were there fewer preterm births
in the first year of
the pandemic in 2020 in Berlin than in the three years immediately preceding the pandemic?
The aim was also to answer two secondary questions: are the differences in the rate
of preterm births
grouped according to week of gestation (GW) (< 28 + 0 vs. ≥ 28 + 0 GW to < 37 + 0
GW)? Was there a drop in the number of “iatrogenic” (primary/planned caesarean section,
induction of
labor) preterm births in 2020?
Method
Sample cohort
All singletons born in Berlin hospitals between January 1, 2017 and December 31, 2020
were included in the evaluation. Multiple births were excluded from the outset, as
multiple pregnancies
are already at higher risk of preterm birth. The focus of the analysis was on the
birth rates and the percentage of preterm neonates among singleton births born before
37 + 0 weeks of
gestation in the three pre-pandemic years 2017, 2018 and 2019 in Berlin compared with
the corresponding rate for 2020, the first year of the pandemic. The data were sourced
using the
routinely collected data of all hospital births obtained in the context of implementing
measures for quality assurance in accordance with § 136 SGB V [Book V of the German
Social Code]
(previously: perinatal data collection).
Requirements for data analysis
Data analysis was carried out after being advised by the Steering Committee for Quality
Assurance Berlin and the release of the anonymized data obtained as part of hospital-based
quality
assurance.
This retrospective analysis of registry data complies with the standards outlined
in the Berlin Data Protection Law and the guideline of the Charité in Berlin on good
scientific
practice.
Statistical analysis
Data were evaluated using Python libraries, pandas, SciPy and statsmodels. Fisher-Yates
test (Fisher’s exact test) was used for statistical inference tests with four field
scenarios, and
chi-square test to test for one-dimensional homogeneity of distribution. Allocation
of data to specific time periods was done using the child’s date of birth (in the
published annual
evaluation of data, the date of the mother’s discharge is used as the criterion for
allocating data to a specific year). A birth was identified as iatrogenically induced
if the infant was
delivered either by primary caesarean section or following induction of labor without
prior rupture of membranes. A patient was classified as a COVID-19 case if her diagnosis
on admission or
at discharge was recorded as U07.1 or U07.2.
Results
Monthly birth rate and total birth rate
The data of a total of 158368 singleton pregnancies were analyzed. The mean monthly
singleton birth rate for the period from January 1, 2017 to December 31, 2020 in Berlin
was 3299.3 births
per month (95 % CI: 3233.7 – 3364.9) with seasonal variations. A total of 119586 infants
were born in the three pre-pandemic years (mean: 39862 per year), whereas the number
of births in the
first year of the COVID-19 pandemic (2020) was 38782. The total number of births (singletons,
hospital births only) decreased significantly since 2017: 2017 – 40345; 2018 – 39931,
2019 –
39310, 2020 – 38782 births. The birth rate per month in the first year of the pandemic
(2020) differs significantly from the rates for the three pre-pandemic years 2017
to 2019 (chi-square
p = 0.002). The monthly distributions also differ considerably when 2020 is compared
to the period 2017 to 2019: in the summer of 2020, the curve flattens, with around
200 fewer births per
month; this level of flattening is not seen in any of the three previous years ([Fig. 1]). Overall, there were fewer births (around 1000 fewer
than in the investigated previous three years), although for obvious biological reasons,
the decrease in the summer of 2020 cannot be the effect of the pandemic. It can also
clearly not be
explained by a shift in the location where women give birth, with birth settings migrating
to out-of-hospital sites. As [Table 1] shows, the
overall number of births born in an outside-of-hospital setting also decreased slightly
in 2020 compared to the previous years.
Fig. 1 Comparison of monthly birth rates (singleton neonates only) for the years 2017 to
2020.
Table 1
Number of planned births which began out-of-hospital in Berlin for the years 2017
to 2020 (home births and number of births in midwife-led
facilities according to QUAG [The Society for Quality in Out-of-Hospital Births])
[7].
|
Year
|
Home births:
n (%)
|
Births in midwife-led facilities:
n (%)
|
Total number of planned births which began out-of-hospital
|
|
2020
|
463 (34.3)
|
887 (65.7)
|
1350
|
|
2019
|
358 (25.4)
|
1051 (74.6)
|
1409
|
|
2018
|
401 (26.3)
|
1126 (73.7)
|
1527
|
|
2017
|
286 (21)
|
1075 (79)
|
1361
|
Preterm birth rate
The percentage of preterm singleton births with a gestational age of less than 37 + 0
weeks was significantly lower in 2020 compared to the three previous years ([Table 2]; [Fig. 2]). Such a significant difference in the preterm birth rate has not been identified
for
singleton births which occurred in the years 2017–2019 (chi-square test of independence
for 2017–2019 χ2 = 1.85, p = 0.97); in contrast to the decreased birth rate, this is a new
phenomenon.
Table 2
Comparison of preterm birth rates (singleton deliveries) of the years 2017 to 2019
with 2020.
|
Total number of births (n)
|
Preterm births < 37 + 0 weeks of gestation (n/percentage of total births)
|
|
Fisher’s exact test, two-sided odds ratio = 1.08, p = 0.002
|
|
1.1.2020 to 31.12.2020 (1 year)
|
38782
|
2294/5.92
|
|
1.1.2017 to 31.12.2019 (3 years)
|
119586
|
7602/6.36
|
Fig. 2 Difference per thousand of the percentage of births per week of gestation for the
year 2020 compared to the years 2017 to 2019.
When extreme preterm births < 28 + 0 GW and the cohort of preterm neonates born between
28 + 0 and 35 + 0 GW were examined separately, the results showed opposing trends:
there was a
significant trend to more preterm births in the group born < 28 + 0 GW and, in contrast,
significantly fewer preterm births in the group born between 28 + 0 and 35 + 0 GW
([Table 3]).
Table 3
Comparison of preterm singleton birth rates for the subgroups < 28 + 0 and 28 + 0
to < 35 + 0 GW for the years 2017 to 2019 vs.
2020.
|
Number of births (n)
|
Preterm births < 28 + 0 GW (n/percentage of total births)
|
Preterm births ≥ 28 + 0 to < 35 + 0 GW (n/percentage of total births)
|
|
* Two-sided Fisher’s exact test: odds ratio = 0.85; p = 0,03; ** Two-sided Fisher’s
exact test: odds ratio = 0.92; p = 0.04
|
|
1.1.2020 – 31.12.2020 (1 year)
|
38782
|
245/0.63*
|
777/2.00**
|
|
1.1.2017 – 31.12.2019 (3 years)
|
119586
|
642/0.54*
|
2603/2.18**
|
[Fig. 2] elaborates on this presentation of results. In this graph, the number of births
per thousand births differentiated according to week of
gestation (GW) was calculated for the period 2017–2019 (p17) and for 2020 (p20). The
graph shows the difference between the two groups (diffT = p20–p17) and the difference
per week of
gestation (GW). It clearly shows the GW areas in which there were fewer preterm births
in 2020 compared to the same period in the years 2017 to 2019 (> 30 + 0 GW, increasing
by week 37 of
gestation), and the periods in which comparatively more preterm births occurred (24 + 0
to 28 + 0 GW). It was noticeable that the drop in preterm births was particularly
high closer to the
expected date of delivery; correspondingly, there was also a contrary (positive) spike
in the difference per thousand around the time of the expected delivery, during which
a greater
percentage of births (+0.5% – +0.8% in the 39th and 40th GW) occurred in 2020.
Delivery by caesarean section and induction of labor
The percentage of secondary deliveries by caesarean section did not differ significantly
between the two periods under comparison (2017–2019 vs. 2020), not even when only
the rate of
caesarean sections for singleton births before 37 + 0 GW were compared (2020 = 13.86%;
2017–2019 = 14.52%; p = 0.21). The total rate of primary caesarean sections, irrespective
of the week
of gestation, also did not differ between the period from 2017 – 2019 compared to
2020 (p = 0.23); however, significantly fewer infants below the age of 37 + 0 weeks
of gestation were born
by primary caesarean section (2020 = 5.84%; 2017–2019 = 7.91%; p = 0.00).
When births by primary caesarean section and births following induction of labor without
prior rupture of membranes were viewed separately, there was a significant decrease
in
“iatrogenically-induced” preterm births before 35 + 0 weeks of gestation but a significant
increase in the number of births between > 35 + 0 and 39 + 0 weeks of gestation. The
number of
inductions of labor remained approximately the same for both periods, with 482 inductions
in 2020 compared to a total of 1483 (494/year) in the three years before the pandemic
(p = 0.113)
([Fig. 3]; [Table 4]).
Fig. 3 Percentage of iatrogenic births according to gestational age and time period (green
area: lower rate of iatrogenic births in 2020, orange area: higher rate in 2020).
The rate is
displayed for 50 births and above in the respective week of gestation.
Table 4
Comparison of the rates of iatrogenic births for the two subgroups > 28 + 0 – 35 + 0
and > 35 + 0 – ≤ 39 + 0 weeks of gestation for the
years 2017–2019 and in 2020.
|
Number of neonates born > 28 + 0 – 35 + 0 and > 35 + 0 – 39 + 0 (n/n)
|
Iatrogenic births > 28 + 0 – 35 + 0 GW (n/percentage of total births/p (Fisher’s exact
test)
|
Iatrogenic births > 35 + 0 – 39 + 0 GW (n/percentage of total births/p (Fisher’s exact
test)
|
|
1.1.2020 to 31.12.2020 (1 year)
|
834/11670
|
137/16.4%/p = 0.016*
|
4040/34.61%/p = 0.0053**
|
|
1.1.2017 to 31.12.2019 (3 years)
|
2790/36097
|
562/20.1%
|
11989/33.21%
|
COVID infections
According to the data recorded during perinatal data collection, a SARS-CoV-2 infection
was detected in 125 of the 38782 births born in Berlin in 2020. The percentage of
primary and
secondary caesarean section deliveries in this (small) group was slightly higher compared
to the total cohort of other deliveries born in 2020 but the increase was not significant
(29.6%
deliveries by caesarean section in COVID-19 patients, 26.7% in patients with no COVID-19
infection, p = 0.48). A primary or secondary caesarean section before 37 + 0 weeks
of gestation was
carried out significantly more often in COVID-19 patients (10.3% for patients without
vs. 24.3% with COVID-19 infection, p = 0.012) ([Fig. 4]).
Fig. 4 Percentage of caesarean section deliveries in 2020 born to COVID-19 patients compared
to other patients whose child was born in 2020 according to the week of gestation
(GW) at
delivery.
Discussion
The important findings of our data analysis are:
-
The total number of live-born singletons born in Berlin maternity hospitals and wards
decreased in the first year of the pandemic.
-
The percentage of preterm births in the cohort of singleton pregnancies born below
the age of 37 + 0 weeks of gestation was significantly lower in 2020 compared to the
three previous
years, with significantly more preterm neonates delivered before 28 + 0 weeks of gestation
and significantly fewer preterm births occurring between 28 + 0 and 35 + 0 weeks of
gestation.
-
Significantly fewer infants aged less than 37 + 0 weeks of gestation were born by
primary caesarean section in 2020.
-
The incidence of induction of labor remained approximately the same.
Stumpfe et al. (2022) recently reported on the limited impact of SARS-CoV-2-related
lockdowns and the consequent decrease in population mobility on the rate of preterm
births in Bavaria. They
also carried out a secondary analysis of obstetric clinical quality assurance parameters
and compared the results for the years 2010 to 2020. They found significantly lower
values during both
lockdown phases for unadjusted preterm birth rates but did not group their data according
to weeks of gestation [6].
What results have been reported by working groups in other European countries and
North America, and what are their explanations for the impact of the pandemic on birth
rates and preterm
birth rates?
A selective search of the literature which we carried out at the end of December 2021
and which only included retrospective cohort and registry studies on indirect consequences
of the
COVID-19 pandemic on perinatal outcome data showed contradictory findings with regard
to pandemic and lockdown-related differences in the prevalence of undesirable pregnancy-related
events
such as preterm births and stillbirths [8]. The Danish group working group pf Aabakke et al. (2021) found no significant differences
in
obstetric or neonatal outcomes, nor did Son et al. (2021) in the USA [9]
[10]. Simon
and colleagues reported in 2021 for the first time about a decrease in preterm births
in a large French cohort of non-infected or non-COVID-19-tested pregnant women and
concluded that
wide-ranging social changes such as lockdown situations could be associated with a
positive impact on perinatal morbidity rates [11]. In
Australia, Rolnik et al. (2021) reported fewer preterm births in women who had been
subjected to strict limitations to contain the spread of the pandemic [12]. Liu et al. (2021) evaluated Canadian data and found a decrease in the number of
singleton births occurring in early gestation for the period of the pandemic they
were
able to review and ascribed it, at least in part, to fewer inductions of labor and
fewer caesarean deliveries, which would indicate that this was partially due to iatrogenic
causes [13].
According to a review of the literature by Townsend et al. (2021), the planned and
unplanned use of antenatal and maternity services decreased during the COVID-19 pandemic
in highly developed
industrial nations; instead, there was an increase in telemedicine appointments and
virtual care [14]. In their review article published
in Lancet Global Health, Chmilewska et al. (2021) emphasized that when considering
the impact of the COVID-19 pandemic on women’s health, the strength and quality of
the respective healthcare
system in different countries (available resources in low-income vs. industrialized
nations) was a significant influencing factor. They included 15 studies in their evaluation
and found no
changes in preterm birth rates prior to the 37th week of gestation [2]. They did, however, find differences between high-income and
low-income countries and noted a decrease in preterm births (12 studies) and of spontaneous
preterm births (two studies) in high-income countries [3]. Simon et al. (2022) were of the opinion that interpreting the results on preterm
births was difficult, particularly because of the effects of country-specific healthcare
measures and restrictions such as lockdowns on non-infected pregnant women [4].
Stansfield et al. (2022) investigated the effects of the COVID-19 pandemic and the
repeated lockdowns in Melbourne, Australia, in 2020 and 2021 on the total number of
live births and
particularly on preterm births in a metropolitan hospital group. They noted a decrease
in the total rate of live births after a relaxation of lockdown measures and a strong
increase in the
number of births in one phase between lockdowns. The percentage and number of preterm
births (< 37 + 0 weeks of gestation) decreased at the start of the first lockdown,
although the biggest
decrease occurred after the end of the second lockdown. Births < 34 + 0 weeks of gestation
also decreased, but no significant changes were found for births < 28 weeks of gestation
[2]. Stansfield and colleagues (2022) were of the opinion that these changes in the
preterm birth rate could be traced back to several factors –
including economic instability, decreased access to healthcare services, and changes
in health-related behaviors as well as more and better hygiene measures. In addition,
the restrictive
measures in force may have affected maternal behavior with regard to sleep, smoking
and alcohol consumption as well as physical activity and also general mental health
and may have led to
changes in the rate of preterm births [2].
For the city of New York, which could serve as the best comparison for Berlin, Weinberger
et al. reported in 2022 that the total number of births decreased in 2020 compared
to 2019, just as
it did in Berlin [15]. The rate of extreme preterm births (< 28 + 0 GW) also decreased significantly,
dropping from 5.6 (in 2019) to
4.7 per 1000 births in 2020 (p < 0.0001), although the rate of “moderate” preterm
births (between 28 + 0 and 35 + 0 weeks of gestation) did not change. Weinberger et
al. (2022) interpreted
these data as suggesting that various biological effects of the SARS-CoV-2 infection
and environmental changes occurring during the pandemic had competing impacts on preterm
birth rates. A
SARS-CoV-2 infection might increase the preterm birth rate by increasing the prevalence
of preeclampsia and the number of medically induced preterm births, but the impact
on the rate of
spontaneous preterm births is not known. The restrictive measures used to limit the
pandemic (lockdowns, quarantine, less travel, etc.) may (also) have reduced maternal
exposure to infectious
agents which could trigger an infection-related preterm birth (Weinberger et al.,
2022). According to Weinberger et al. (2022), the effects of the COVID-19 pandemic
on the preterm birth rate
probably reflect a balance between an increase in preeclampsia-induced preterm births
and a decrease in the number of infection-related spontaneous preterm births [15].
In 2022, Dench et al. published a data analysis showing changes in preterm singleton
birth rates, using the national birth data of the US National Center for Health Statistics
for the years
2010 to 2020 which includes 100% of all births registered in the United States including
home births (a total of 41394390 singleton births) [14]. The US data indicate that preterm births decreased after the lockdown in March
2020. The decrease was primarily due to changes in caesarean deliveries or deliveries
following induction of labor (iatrogenically induced preterm births). Dench et al.
(2022) interpret the slight decrease in preterm births born without caesarean delivery
or induction of labor
as an indication of the effect of lockdown [16]. Klumper et al. (2021) investigated preterm births during the first phase of the
pandemic
in The Netherlands and suspected that a pandemic-related reduction in the rate of
“iatrogenic” births, i.e., infants delivered by primary caesarean section or after
induction of labor without
prior rupture of membranes, was the main reason for the decrease in the number of
preterm births [17].
Limitations: The evaluations presented here are based on a secondary analysis of standard
data routinely collected in German labor wards, which could negatively affect the
quality of the
documentation. This also applies to pregnant women identified as having SARS-CoV-2
infection.
Conclusion
A careful interpretation of these initial data which also takes the above-mentioned
limitations into consideration and includes internationally published results suggests
that in 2020, social
and environmental effects of lockdown measures as well as iatrogenic and biological
factors affected the rate of preterm births. More data analyses of existing registry
data will be needed to
determine possible causes for the apparent decrease in the rates of preterm births
in the first year of the pandemic and find out whether this will lead to more insights
which could be used to
shape future strategies to reduce the rates of preterm births. The next step should
consist of a detailed and critical evaluation of the routinely collected, extensive
perinatal data of all
hospital births collected across all of Germany as part of quality assurance, which
would include the years 2021 and 2022. An analysis of possible pandemic-related changes
in out-of-hospital
births would also be very interesting and important.
This pandemic has placed a severe strain on all of society. But if the pandemic provides
an opportunity to carry out an in-depth analysis of data that will lead to the detection
of important
factors which could result in a long-term reduction of preterm births, then we should
absolutely take advantage of this unique time and view it as an opportunity “to make
something good out of
bad.”
