2024/11/14 by Asma Khalil, Ian Painter, Vivienne Souter · 2 voices
Medicine · #COVID-19 Impact on Reproduction #COVID-19 and healthcare impacts #Kawasaki Disease and Coronary Complications
paper · pdf · doi:10.1002/uog.29126
Infection with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in pregnancy is associated with increased maternal morbidity and mortality1, 2. Associations have also been reported with placental complications, including stillbirth and pre-eclampsia2, and cases of vertical transmission of the virus from the mother to the fetus have been documented, although these appear to be rare3. Until recently, there has been very little evidence to suggest an association between the coronavirus disease 2019 (COVID-19) pandemic and congenital anomalies4-6. In a study of 35 123 singleton births in California, USA, a positive maternal COVID-19 test was associated with increased odds of congenital anomalies (odds ratio (OR), 1.69 (95% CI, 1.15–2.50))7. Other research found no association between COVID-19 and congenital anomalies8. However, in November 2023, Wang et al. reported an increase in the incidence of situs inversus in a large birth cohort following the lifting of pandemic-associated restrictions in China9. In light of this report, we sought to evaluate whether there was a change in the incidence of congenital heart defects among live births in the USA during the COVID-19 pandemic. We obtained USA birth certificate data for singleton live births from the Centers for Disease Control and Prevention's (CDC) publicly available national birth files between December 2016 and November 2022 to evaluate the effect of the COVID-19 pandemic on the incidence of congenital heart defects. The USA government requires a birth certificate to be submitted for all live births irrespective of gestational age or viability. Data are entered into a standardized form by staff at the hospital or birth center where the baby was delivered. The completed forms are then submitted to the local state vital statistics office where checks are performed and any queries are submitted to the relevant hospital or birth center. Patient identifiers are removed before the birth certificate data are submitted to the National Center for Health Statistics. Further details on the procedure for collecting birth certificate data and the type of information collected are available on the CDC website10. USA birth certificates collect information for one congenital cardiac field, namely cyanotic heart defects, defined as transposition of the great arteries, tetralogy of Fallot, pulmonary or pulmonic valvular atresia, tricuspid atresia, truncus arteriosus, total/partial anomalous pulmonary venous return with or without obstruction, coarctation of the aorta, hypoplastic left heart syndrome and interrupted aortic arch11. An interrupted time series analysis with monthly intervals was used to evaluate the change in incidence of cyanotic congenital heart defects in USA birth certificate data over time, using the COVID-19 pandemic as the interruption. The pre-interruption period was defined as 1 December 2016 to 30 November 2019 and the post-interruption period was defined as 1 December 2020 to 30 November 2022, giving 36 timepoints before and 24 timepoints after the interruption. We excluded the period between 1 December 2019 and 30 November 2020 for several reasons. First, the earliest identified case of SARS-CoV-2 infection in the USA was on 19 January 202012. Second, if SARS-CoV-2 infection was associated causally with congenital heart defects, then exposure would have to take place early in pregnancy, during the period of embryogenesis, and we would not expect to observe an increase in congenital heart defects in live births fewer than 6 months after the exposure (since most births occur at 37 weeks' gestation or later). Additionally, the COVID-19 pandemic evolved during 2020 and population testing was limited, so the exact timing of maximum SARS-CoV-2 exposure in pregnant people in the first trimester of pregnancy is unknown. Lastly, the COVID-19 vaccine was first made available in the USA in December 2020 and was recommended for all pregnant people by the American College of Obstetricians and Gynecologists in July 2021, and it could have affected both exposure and symptoms in pregnant people13. A secondary segmented regression analysis was conducted with a shorter post-interruption period from 1 December 2020 to 30 November 2021 followed by an additional segment from 1 December 2021 to 30 November 2022, with a slope equal to that during the pre-interruption period, to examine whether a shift in trend occurred once a large majority of the population had been vaccinated against or infected with SARS-CoV-2. The incidence of Down syndrome in live births is also collected in USA birth certificate data and was evaluated during the pre- and post-interruption periods as a control. The interrupted time series analysis used a segmented Poisson regression to assess trends in the incidence of congenital heart defects, expressed as OR. We used robust standard errors14. We checked for seasonality during the pre-interruption period given disruptions to seasonal patterns of infectious disease during the post-interruption period. We checked for autocorrelation using the Durbin–Watson test (DHARMa package). We calculated the change in odds between the pre- and post-interruption periods as the level change in OR between the pre-interruption trend projected out to the start of the post-interruption period and the post-interruption estimate of risk at the start of the post-interruption period. Estimates of the slope in each time period were presented as OR over a 12-month interval with 95% CI. Statistical analysis was performed using the R package version 4.3.115. Since Down syndrome reporting includes both confirmed and pending cases, we conducted separate analyses for confirmed cases only and confirmed or pending cases. As we observed changes over time in maternal body mass index (BMI), prepregnancy diabetes mellitus and prenatal care, all of which might impact the incidence or detection of congenital heart defects, we also conducted logistic regression analyses adjusted for maternal BMI (six categories plus unknown as recorded in the birth certificate), prepregnancy diabetes mellitus, prepregnancy hypertension, maternal age, parity and the month in gestation at which prenatal care was initiated (four categories plus unknown as recorded in the birth certificate) (Table 1). Unknown values were included as a category in the adjusted analysis. There were 11 054 415 singleton births recorded during the pre-interruption period and 7 087 583 singleton births recorded during the post-interruption period. Of these, 43 651 (0.39%) and 26 957 (0.38%) births, respectively, were recorded as non-reporting or unknown for cyanotic congenital heart disease and were excluded from the analysis, leaving 11 010 764 and 7 060 626 births, respectively. Cases that were non-reporting for cyanotic congenital heart disease were generally non-reporting for all congenital anomalies, including Down syndrome. However, 27 births that were non-reporting for cyanotic congenital heart disease but were recorded as reporting for Down syndrome were excluded from the analysis. Compared with the pre-interruption period, birthing individuals in the post-interruption period were slightly more likely to be aged 35 years or older (20.0% vs 18.0%; P < 0.001), had higher BMI (median, 26.5 kg/m2 vs 25.7 kg/m2; P < 0.001), were more likely to not have prenatal care (2.1% vs 1.7%; P < 0.001) and were more likely to have prepregnancy diabetes mellitus (1.1% vs 0.9%; P < 0.001) or prepregnancy hypertension (2.8% vs 2.0%; P < 0.001) (Table 1). Birthing individuals in the post-interruption period were also more likely to be nulliparous (39.2% vs 38.4%; P < 0.001). Compared with the pre-interruption period, the incidence of cyanotic congenital heart disease during the post-interruption period was significantly higher (65.4 vs 56.5 per 100 000 live births), while the incidence of confirmed Down syndrome (23.0 vs 24.4 per 100 000 live births) or confirmed plus pending Down syndrome (28.7 vs 30.3 per 100 000 live births) was lower, but this drop was not statistically significant. During the post-interruption period, the incidence of unknown status was higher for both cyanotic congenital heart disease and Down syndrome (200.1 vs 143.7 per 100 000 live births for both). Table 2 shows the incidence per 100 000 live births of cyanotic congenital heart disease and Down syndrome in USA national birth certificate data by year. No seasonality was observed in the pre-interruption period for either cyanotic congenital heart disease or Down syndrome (P = 0.404 and P = 0.365, respectively), and the Durbin–Watson test for autocorrelation in residuals was not statistically significant (P = 0.612 and P = 0.967, respectively). An increasing trend in the incidence of cyanotic congenital heart disease was observed during the pre-interruption period, with the odds increasing by 3.4% per year (OR, 1.03 (95% CI, 1.01–1.06)) (Table 3, Figure 1a). The estimated incidence at the start of the post-interruption period was 11.1% higher than that projected from the pre-interruption period (OR, 1.11 (95% CI, 1.02–1.20)). We observed a similar level change in the adjusted model (adjusted OR (aOR), 1.12 (95% CI, 1.01–1.23)). A statistically significant change in slope from the pre-interruption period to the post-interruption period was observed (P = 0.027 on adjusted analysis), with the slope in the post-interruption period estimated to be decreasing (OR per year, 0.96 (95% CI, 0.91–1.01)). However, this slope was not significantly different from 1, indicating that the slope in the post-interruption period might be flat. We did not observe a level change in the incidence of confirmed Down syndrome between the pre- or post-interruption periods on unadjusted or adjusted analysis (OR, 0.96 (95% CI, 0.81–1.15) and aOR, 0.97 (95% CI, 0.84–1.12), respectively) (Figure 1b). No difference was observed in the incidence of confirmed Down syndrome between the pre-trend projection and the start of the post-interruption period. Estimates for confirmed or pending Down syndrome were similar. Thus, our data demonstrate that the incidence of cyanotic congenital heart defects was significantly higher in the post-interruption period compared with the pre-interruption period (OR, 1.11 (95% CI, 1.02–1.20); P = 0.012). There was no corresponding increase in the incidence of Down syndrome over the same period (OR, 0.96 (95% CI, 0.81–1.15); P = 0.648). The reasons underlying the finding of an increase in cyanotic congenital heart defects in the USA during the COVID-19 pandemic are uncertain, but this finding raises the possibility that infection with SARS-CoV-2 could be associated with fetal cyanotic cardiac abnormalities. It has been postulated that SARS-CoV-2 transmission from mother to fetus in early pregnancy is mediated through the angiotensin-converting enzyme 2 (ACE2) found in developing human embryos in the early stages of development (gametes, zygotes and four-cell embryos)4, 5. It could be hypothesized that, through the same mechanism, the virus could penetrate fetal cells at an early stage in development and affect cell transformation and growth in the human heart, which is one of the earliest organs to develop and begin functioning during the embryonic period. Caution is needed when interpreting this observation, as there are other possible explanations. These include changes in access to care during the pandemic, such as limited prenatal ultrasound screening, or a change in birth certificate completion or reporting. It could be postulated that lifestyle and dietary changes could affect congenital anomalies, but, even after adjusting for maternal BMI, prepregnancy diabetes mellitus, prepregnancy hypertension and the month in gestation at which prenatal care was initiated, the increase in incidence of cyanotic congenital heart disease persisted. However, the possibility of an unrecognized confounding exposure cannot be excluded. If the increase in odds of cyanotic heart defects was due to infection with SARS-CoV-2 during the course of pregnancy, the observed increase in odds is likely to be a substantial underestimate of the direct effect of infection. During the post-interruption time period, seroprevalence-based estimates of any prior infection in the 18–49-year age group increased from 12.6% in December 2020 to 70.0% in April 2022 (the time period in which term pregnancies with SARS-CoV-2 infection during the first 3 months of gestation would be captured in the post-interruption time period)16, 17. If the observed increase in odds is entirely due to infection during the first 3 months of pregnancy, we calculate that, to get a pre–post OR of 1.11, under the assumption that infections during the first 3 months of pregnancy occur at the same rate as that in the 18–49-year age group, the OR from infection would need to be 2.1. If antibodies from prior infection or vaccination were protective against the effects of infection on cyanotic cardiac abnormalities, then the direct effect of infection during the first 3 months would be even higher. If we assume that the rate during the second half of the post-interruption period (from December 2021 to November 2022) followed the same trend as the pre-interruption period, then the estimated level change for the period December 2020 to November 2021 was 14% (OR, 1.14 (95% CI, 1.03–1.27)). This study benefits from the use of a large national dataset in which we were able to evaluate the incidence of cyanotic congenital heart anomalies. We were unable to assess the incidence of cyanotic congenital cardiac defects by subtype or obtain the corresponding data on maternal SARS-CoV-2 infection. We were also unable to access corresponding data on the incidence of maternal COVID-19 infection in this cohort. Another limitation is that the analysis included only live births and did not capture stillbirths, miscarriages or terminations of pregnancy. However, two studies in the USA did not observe a decrease in termination of pregnancy or stillbirth during the pandemic that could have contributed to an increase in live births with cyanotic heart defects18, 19. Additionally, although the data on congenital birth defects in national birth files likely have limitations, there is nothing to suggest that these limitations were different over the period of the COVID-19 pandemic. We have recently addressed the need for robust evidence linking pregestational and early-pregnancy SARS-CoV-2 infections with congenital anomalies, highlighting the inadequate tracking of infection history and methodological flaws in published studies, as well as the restricted access to testing and burden of undiagnosed infections, particularly in low- and middle-income countries20. Thus, given the important implications of a possible link between the COVID-19 pandemic and fetal cardiac abnormalities, we suggest that these results warrant further research into the association between maternal COVID-19 disease and fetal cyanotic congenital heart defects. The data that support the findings of this study are available from the corresponding author upon reasonable request.