Reduction in preterm birth rates during and after the COVID‐19 lockdown in Queensland Australia

Background Preventative strategies for preterm birth are lacking. Recent evidence proposed COVID‐19 lockdowns may have contributed to changes in preterm birth. Aims To determine the prevalence of preterm birth and birth outcomes during and after the COVID‐19 lockdown at the Sunshine Coast University Hospital and the overall state of Queensland, Australia. Methods Retrospective cohort analysis of all births in Queensland including the Sunshine Coast University Hospital, during two epochs, April 1–May 31, 2020 (lockdown) and June 1–July 31, 2020 (post‐lockdown), compared to antecedent calendar‐matched periods in 2018–2019. Prevalence of preterm birth, stillbirth, and late terminations were examined. Results There were 64 989 births in Queensland from April to July 2018–2020. At the Sunshine Coast University Hospital, there was a significantly higher chance of birth at term during both lockdown (odds ratio (OR) 1.81, 95% CI 1.17, 2.79; P = 0.007) and post‐lockdown (OR 2.01, 95% CI 1.27, 3.18; P = 0.003). At the same centre, prevalence of preterm birth was 5.5% (30/547) during lockdown, compared to 9.1% (100/1095) in previous years, a 40.0% relative reduction (P = 0.016). At this centre during lockdown, emergency caesareans concurrently decreased (P < 0.01) and instrumental vaginal births increased (P < 0.01). In Queensland overall, there was a nonsignificant decrease in the prevalence of preterm birth during lockdown. Conclusions There is a link between lockdown and a reduction in the prevalence of preterm birth on the Sunshine Coast. The cause is speculative at present, although increased influenza vaccination rates, decreased transmission of infections, and improved air quality may have been favourable in reducing preterm birth. Further research is needed to determine a causal link.


INTRODUCTION
Preterm birth (PTB) is a major problem worldwide and is the leading cause of death in children under five. 1 Despite advances in obstetric care, there has been no sustained reduction in the prevalence of PTB and preventative measures remain only partially successful.
The World Health Organization declared COVID-19 a pandemic on 11 March, 2020. 2 This was followed by lockdowns across the globe, causing unprecedented effects on society and human behaviour in ways not seen before. In Queensland, Australia, the restrictions to mitigate the spread of COVID-19 culminated in a nine-week state-wide lockdown from March 30 to June 1, 2020, with a phased easing of restrictions. 3 During lockdown, all shops, restaurants, schools, universities, and non-essential workplaces were closed. 3 Residents were prohibited from leaving their homes except for essential reasons (Fig. 1). 3 To minimise the risk of virus transmission, models of antenatal care at the Sunshine Coast University Hospital (SCUH) were adapted. 4 Women were risk-triaged to either virtual or in-person appointments utilising COVID-19 safe practices. 4 While lockdown in Queensland has since lifted, restrictions continue to change rapidly throughout Australia in response to COVID-19 cases.
In the wake of the pandemic, research emerged demonstrating the varied impact of lockdown on the prevalence of PTB worldwide. [5][6][7][8][9][10][11][12][13][14] Our primary aim was to determine the prevalence of PTB during and after the COVID-19 lockdown at SCUH and the overall state of Queensland, Australia. Our secondary aims were to investigate birth outcomes including stillbirth and late termination of pregnancy, and compare perinatal characteristics, mode of delivery, and proposed aetiologies for preterm birth at SCUH during and after lockdown.

Study population
This study focused on pregnant women in two geographical locations in Australia. Firstly, the state of Queensland, which services the health needs of over 5 million people, and secondly, SCUH, which provides care for the 400 000 people living in this area of Queensland. 15 SCUH is a level five neonatal centre and cares for infants ≥29 weeks gestation or ≥1000 g birth weight. Of pregnant women in Queensland, 3.0% were aged less than 20 years and 4.0% were aged over 40 years, 16 7.5% identified as Aboriginal or Torres Strait Islander, 16 21.0% were obese, 17

Study design
Retrospective cohort analysis of births in Queensland, Australia was carried out during two epochs, April 1-May 31, 2020 (lockdown) and June 1-July 31, 2020 (post-lockdown), compared to calendar-matched periods from the preceding two years.
There was a gap in the existing literature lacking analysis of birth outcomes after restrictions easing, thus a post-lockdown period was included. The prevalence of birth outcomes including PTB, stillbirth, and late terminations were examined.
All PTB less than 37 weeks gestation were analysed, including sub-group analyses of extreme preterm (20 + 0-27 + 6 weeks), very preterm (28 + 0-31 + 6 weeks), and moderate-to-late preterm (32 + 0-36 + 6 weeks). To facilitate comparison of birth data at SCUH to Queensland overall, and to wholly represent all births after 20 weeks, no births were excluded because of multiple births, stillbirths after 20 weeks, or terminations after 20 weeks. At SCUH, perinatal characteristics were collected, including maternal age, body mass index (BMI), parity, and mode of delivery, proposed aetiology of birth outcomes including spontaneous PTB, iatrogenic PTB, stillbirth, and late termination. We subclassified spontaneous PTB into preterm premature rupture of membranes, preterm labour, and multiple pregnancy. Subclassifications for iatrogenic PTB included intrauterine growth restriction, hypertension, multiple pregnancy, fetal distress, placental abruption and antepartum haemorrhage, and other maternal comorbidities.

Statistical analysis
All data were deidentified prior to analysis. Univariate logistic regression was performed to identify potential predictors of preterm or term birth. Variables with P < 0.20 were included in multivariate logistic regression analysis. The prevalence of PTB, stillbirth, and late terminations were calculated. Percent change was calculated to assess for difference between time periods.
Descriptive statistics were reported as mean and standard deviation for normally distributed continuous data, median and interquartile range (IQR) for non-normally distributed data, and frequencies and percentages for categorical data. Continuous data were tested for normality using the Shapiro-Wilk test and analysed using the t-test. Categorical data were analysed using either a Mann-Whitney U-test, χ 2 , or Fisher's exact test.
P-values < 0.05 were considered significant. All analyses were conducted using R statistical software.

Patient and public involvement
Patient and public involvement began in the early stages of planning with the Preterm Infants Parents Association (PIPA), a registered Australian charity for families with premature babies.
The PIPA team reviewed the study design and research questions.
A plain language summary will be distributed to the PIPA network. All births at SCUH and in the state of Queensland overall are summarised in Table 1. During lockdown, the prevalence of PTB at SCUH was 5.5% (30/547), compared to 9.1% (100/1095) during the same period in previous years, representing a 40.0% relative reduction (P = 0.016). This reduction in PTB at SCUH appeared to be driven by moderate-to-late PTB (32 + 0-36 + 6 weeks). The prevalence of moderate-to-late PTB was 4.6% (25/547) during lockdown compared to 7.5% (82/1095) in previous years, a 39.0% relative reduction (P = 0.034). In contrast, the prevalence of PTB in Queensland overall was 8.6% (870/10 154) during lockdown compared to 8.9% (1871/20 939) in previous years, a nonsignificant 4.1% relative reduction in PTB (P = 0.33).

RESULTS
After lockdown restrictions eased, the prevalence of PTB at SCUH was 5.0% (26/524) during the post-lockdown period compared to 7.1% (80/1130) in previous years, a nonsignificant 29.9% relative reduction in PTB (P = 0.12). This reduction appeared to be driven by moderate-to-late PTB, as the prevalence in this group was 3.8% (20/524) during post-lockdown, compared to 6.4% (72/1130) in previous years, a 40.0% relative reduction (P = 0.045). In the state of Queensland overall, there was no significant difference in PTB observed in the post-lockdown period compared to previous years. Table 2 outlines the perinatal characteristics of births at SCUH between April 1-July 31, 2018-2020. Most pregnant women were aged 20-40 years and only 4.3% of women were in the higher risk age groups for PTB of ≤20 years or ≥40 years. 18 There was a general increase in BMI in both the lockdown and post-lockdown groups in 2020 when compared to 2018-2019, consistent with the rising prevalence of obesity. 17 At SCUH during lockdown, we observed a 29.5% relative decrease in emergency caesarean sections (P < 0.01) and concurrent 41.5% relative increase in instrumental vaginal births (P < 0.01) when compared to previous years.
Although the overall prevalence of PTB decreased during lockdown at SCUH, there was no change identified in the ratio between spontaneous or iatrogenic (medically indicated) PTB during lockdown or post-lockdown ( Table 3). A nonsignificant increase in stillbirth was observed during lockdown (P = 0.09) and post-lockdown (P = 0.08) when compared to previous years at SCUH, although the numbers were low. There was no change in the prevalence of late terminations, compared to previous years at SCUH.

DISCUSSION
Australia enforced strict lockdowns to mitigate the spread of COVID-19, reaching a stringency level of 71.3 in a 1-to-100 scale on the Oxford Government Response Tracker. 19 Residents were not permitted to leave their home except for essential reasons. 3 At SCUH in Queensland Australia, there was a significantly higher chance of birth at term during lockdown and post-lockdown, when compared to previous years. During lockdown, a 40.0% relative reduction in the prevalence of PTB was observed at SCUH. This finding is consistent with similar studies worldwide which reported a significantly lower prevalence of PTB during lockdown. 6,7,9,11,12 In contrast to the reduction in PTB observed at SCUH, there was a nonsignificant decrease in PTB observed in the state of Queensland overall during lockdown. Our study and another study from Brisbane, 6 both metropolitan areas in Queensland, observed reductions in PTB. However, in our study, the reduction in PTB was not observed for the state of Queensland overall. It is possible the effect of lockdown on PTB was more apparent in metropolitan areas, although this requires further investigation. At SCUH, the reduction in the prevalence of PTB during lockdown was driven by moderate-to-late PTB, in line with other studies. 6 From June 2020, severe restrictions were lifted, although mild restrictions remained. In the post-lockdown period, the reduction in moderate-to-late PTB was sustained, which may have been due to ongoing mild restrictions, or that altered maternal behaviours enforced during lockdown persisted even after restrictions eased.
The obstetric antecedents of PTB include delivery for materno-fetal indications (iatrogenic), spontaneous preterm labour, and preterm premature rupture of membranes. 18 In high-income countries, 30.0% of PTB are iatrogenic, while 70.0% are spontaneous. 18 The aetiology of most spontaneous PTB is unknown, the most common cause identified is intrauterine infection, inflammation, and placental anomalies. 18 We did not observe a significant difference in the ratio between   free-of-charge. 17 We also recognise that vaccination uptake is lower in people living in regional and remote areas; a large proportion of Queenslanders live outside cities. 17  Our study is the first to examine state-wide PTB data during the COVID-19 pandemic in Australia. This allowed for comparison of one health region to a large state overall, including metropolitan, regional, and rural areas. This paper is one of only three studies from Australia to determine the prevalence of PTB during the pandemic. Further, our study is the first paper to highlight the possible association between reduced PTB during lockdown and increased influenza vaccination rates in Queensland, Australia.
For births at SCUH, we analysed perinatal characteristics including maternal demographics and comorbidities, mode of delivery, and birth outcomes including spontaneous and iatrogenic PTB, stillbirth, and late terminations.
Our study has limitations, which should be considered when interpreting results. As this study was observational, it was not possible to account for unmeasured confounding variables. Although detailed data were collected on all births at SCUH, state-wide data for Queensland was limited to birth outcomes and gestational age,

DATA AVAILABILITY STATEMENT
The authors confirm that the data supporting the findings of this study are available within the article and its Supplementary Materials.

SUPPORTING INFORMATION
Additional supporting information may be found online in the Supporting Information section at the end of the article.