Abstract
Background
Using data from the German CRONOS registry, we assessed the risk of a complicated course of COVID-19 in women with a SARS-CoV-2-infection during pregnancy, with particular consideration of gestational age, vaccination status, and pandemic dynamics.
Methods
Data acquired in two separate periods (March 2020 to August 2021; January to June 2022) for CRONOS, a prospective, hospital-based observational study (DRKS00021208), were studied with logistic regression models. Odds ratios comparing 32 with 22 weeks of gestation were calculated for relevant COVID-19-specific events occurring within 4 weeks of a positive test result.
Results
Data from 3481 women were evaluated. The risk of all of the defined COVID-19-specific events was low among women who became ill with COVID-19 during the first trimester and rose with increasing gestational age into the early third trimester. For example, the odds ratio for hospitalization because of a COVID-19 infection, comparing 32 versus 22 weeks of gestation, was 1.4 (95% confidence interval [1.2; 1.7]). This risk was lower in the second period of data acquisition than in the first (OR 0.66; 95% CI [0.50; 0.88]), and it was even lower if the pregnant patient had been vaccinated against COVID-19 (OR 0.27; 95% CI [0.18; 0.41]).
Conclusion
These findings can serve as a basis for counseling about prophylactic or therapeutic measures, such as the administration of monoclonal antibodies. They underscore the efficacy of vaccination for pregnant women even during the omicron phase of the pandemic.
COVID-19 is rarely severe in women of reproductive age. In pregnancy, however, infection with SARS-CoV-2 can lead to complications (1) and even to premature birth or stillbirth (2, 3). Starting in April 2020, the hospital-based registry CRONOS (COVID-19 Related Obstetric and Neonatal Outcome Study in Germany) of the German Society of Perinatal Medicine (Deutsche Gesellschaft für Perinatale Medizin, DGPM) records the data of women who test positive for SARS-CoV-2 during gestation (4), with the aim of improving our knowledge of the effects of infection with the virus on the pregnancy. Extracts of the information gained from CRONOS are published weekly (in German) at www.dgpm-online.org and used in the formulation of recommendations and guidelines on the care of pregnant women and their newborn children (5– 8).
This article answers questions as to how COVID-19 influences maternal and perinatal outcomes with regard to gestational age and the role played by the omicron variant of SARS-CoV-2 compared with earlier variants and in relation to vaccination status.
To this end, we analyzed the COVID-19-specific events experienced by pregnant women with COVID-19 within 4 weeks of the diagnosis of SARS-CoV-2 infection during two data collection periods. Period 1 included women whose infections occurred prior to 24 August 2021 (the period up to the first occurrence of the delta variant [9] and before the publication on 17 September 2021 of a general recommendation that pregnant women in Germany should be vaccinated [8]). Period 2 included data from women with infection detected between 17 January 2022 and 16 June 2022 (a period of > 95% omicron dominance [9]).
Method
CRONOS is a multicenter prospective observational study with data from 130 actively recruiting hospitals in Germany and Austria (as of 16 June 2022). The women included were those with acute or previous SARS-CoV-2 infection during pregnancy who, regardless of indication, were cared for in one of the participating obstetric departments. The study was approved by the ethics committees of the study center (UKSH Kiel, AZ: D 451/20) and the participating hospitals. Information about CRONOS has been published at the website www.dgpm-online.org and in the German Clinical Trials Registry (DRKS00021208). The registry’s methodology has been described elsewhere (4, 10– 12). The data set from period 1 was analyzed with regard to the occurrence of unfavorable events depending on gestational age at the time of symptomatic infection (COVID-19). The impact of the omicron variant was estimated by comparing the data from period 1 with the data already available from period 2. To ensure comparability of the two periods, taking into account a considerable number of women not recruted timely to infection and with still ongoing pregnancy in period 2 we selected from both datasets patients who contracted COVID-19 after at least 22 weeks of gestation and came into contact with the hospital (as either outpatients or inpatients) within 4 weeks after infection. The evaluation was restricted to maternal COVID-19-related events.
The following maternal COVID-19-related events were defined:
Hospital admission due to COVID-19 within 4 weeks after infection
Pneumonia, defined by the findings of clinical examination or the need for oxygen therapy
Intensive monitoring, invasive ventilation, or death of the pregnant woman
Iatrogenic delivery due to COVID-19 within 4 weeks after infection
In addition, the following perinatal events were defined for period 1:
The ending of pregnancy (miscarriage or premature birth at gestational age [GA] < 37 + 0 weeks) within 4 weeks after infection
Delivery within 4 weeks after infection followed by transfer of the child to a neonatal intensive care unit (NICU) or by antenatal or postnatal death of the child
The findings were evaluated by means of logistic regression models and calculation of odds ratios to compare diagnosis of infection at GA 32 weeks versus 22 weeks, taking into account potential confounders such as maternal age, body mass index (BMI), comorbidities, and vaccination status.
Detailed description of the analysis strategy, the reason for building odds ratios, data processing, the exclusion criteria, and the rationale for exclusion of women with asymptomatic infections from model analyses can be found in the “Method,” “Prevention of bias,” and “Statistics” sections of the eMethods together with eTables 1 and 2 and eFigures 1– 3.
eTable 1. Selected maternal characteristics and outcome parameters for pregnancy and SARS-CoV-2 infection in 2528 women from data collection period 1, comparing COVID-19 with asymptomatic infection*.
| COVID-19 | Asymptomatic infection | p-value | |||||
| n | % or median | 95% CI or quartiles | n | % or median | 95% CI or quartiles | ||
| Number (n) | 1827 | 701 | |||||
| Primiparity (n (%) and [95% CI]) | 706 | (38.6) | [36.4; 40.9] | 282 | (40.2) | [36.6; 44.0] | 0.76 |
| Multiple pregnancy (n (%) and [95% CI]) | 50 | (2.7) | [2.0; 3.6] | 18 | (2.6) | [1.5; 4.0] | 0.71 |
| Gestational age at time of positive test result (median [IQR]) | 28.0 | [19.0; 34.0] | 37.0 | [32.0; 38.0] | < 0.001 | ||
| – Number of women with positive test result < ga 24 + 0 weeks (n (%) and [95% ci]) | 591 | (32.3) | [30.2; 34.5] | 67 | (9.6) | [7.5; 12.0] | < 0.001 |
| – Number of women with positive test result < ga 37 + 0 weeks (n (%) and [95% ci]) | 1497 | (81.9) | [80.1; 83.7] | 269 | (38.4) | [34.8; 42.1] | < 0.001 |
| Maternal age (median [IQR]) | 31.0 | [28.0; 35.0] | 29.0 | [26.0; 34.0] | < 0.001 | ||
| Maternal BMI in kg/m2 at beginning of pregnancy (median [IQR]) | 24.6 | [21.7; 29.0] | 24.2 | [21.6; 28.6] | 0.11 | ||
| Smoking in pregnancy (n (%) and [95% CI]) | 54 | (3.0) | [2.2; 3.8] | 38 | (5.4) | [3.9; 7.4] | 0.01 |
| Maternal comorbidities at time of infection by severity | 0.02 | ||||||
| – None (n (%) and [95% CI]) | 1151 | (63.0) | [60.7; 65.2] | 486 | (69.3) | [65.8; 72.7] | |
| – Mild: no or hardly any symptoms (n (%) and [95% CI]) | 432 | (23.6) | [21.7; 25.7] | 148 | (21.1) | [18.1; 24.3] | |
| – Moderate: e.g., regular medication intake (n (%) and [95% CI]) | 169 | (9.3) | [8.0; 10.7] | 43 | (6.1) | [4.5; 8.2] | |
| – Severe: e.g., regular medical examination (n (%) and [95% CI]) | 32 | (1.8) | [1.2; 2.5] | 14 | (2.0) | [1.1; 3.3] | |
| – No data (n (%) and [95% CI]) | 43 | (2.4) | [1.7; 3.2] | 10 | (1.4) | [0.7; 2.6] | |
| Maternal comorbidities at time of infection by type (selected major categories) | |||||||
| – Cardiovascular disease, e.g., arterial hypertension (n (%) and [95% CI]) | 71 | (3.9) | [3.0; 4.9] | 21 | (3.0) | [1.9; 4.5] | 0.34 |
| – Pre-existing diabetes mellitus (n (%) and [95% CI]) | 20 | (1.1) | [0.7; 1.7] | 9 | (1.3) | [0.6; 2.4] | 0.68 |
| – Pulmonary disease, e.g., bronchial asthma (n (%) and [95% CI]) | 60 | (3.3) | [2.5; 4.2] | 12 | (1.7) | [0.9; 3.0] | 0.03 |
| Circumstances of positive test result | < 0.001 | ||||||
| – Acute symptoms (n (%) and [95% CI]) | 1108 | (60.6) | [58.4; 62.9] | 0 | (0.0) | [0.0; 0.5] | |
| – Contact with infected persons (n (%) and [95% CI]) | 296 | (16.2) | [14.5; 18.0] | 99 | (14.1) | [11.6; 16.9] | |
| – Screening in hospital or elsewhere (n (%) and [95% CI]) | 208 | (11.4) | [10.0; 12.9] | 431 | (61.5) | [57.8; 65.1] | |
| – Other/unknown (n (%) and [95% CI]) | 215 | (11.8) | [10.3; 13.3] | 171 | (24.4) | [21.3; 27.7] | |
| Gestational age at time of first hospital contact (median [IQR]) | 36.0 | [31.0; 39.0] | 38.0 | [36.0; 40.0] | < 0.001 | ||
| Hospital contact within 4 weeks of positive test result | < 0.001 | ||||||
| – Number yes (n (%) and [95% CI]) | 1063 | (58.2) | [55.9; 60.5] | 595 | (84.9) | [82.0; 87.5] | |
| – Number no (n (%) and [95% CI]) | 744 | (40.7) | [38.5; 43.0] | 100 | (14.4) | [11.9; 17.2] | |
| – Number with no data (n (%) and [95% CI]) | 20 | (1.09) | [0.7; 1.7] | 5 | (0.71) | [0.2; 1.7] | |
| Time in weeks from infection to first hospital contact (median [IQR]) | 3.0 | [0.0; 11.0] | 0.0 | [0.0; 0.0] | < 0.001 | ||
| – With positive test result in first trimester (GA < 14+0 weeks) | 20.5 | [8.0; 28.0] | 4.0 | [0.0; 29.0] | 0.07 | ||
| – With positive test result in second trimester (GA 14+0 weeks to 27+6 weeks) | 11.0 | [2.0; 17.0] | 8.0 | [0.0; 14.5] | 0.03 | ||
| – With positive test result in third trimester (GA ≥ 28+0 weeks) | 0.0 | [0.0; 3.0] | 0.0 | [0.0; 0.0] | < 0.001 | ||
| Gestational age in weeks at time of hospital admission (median [IQR]) | 36.0 | [31.0; 38.0] | 37.5 | [36.0; 39.0] | < 0.001 | ||
| Hospital admission within 4 weeks of positive test result | < 0.001 | ||||||
| – Number yes (n (%) and [95% CI]) | 738 | (40.4) | [38.1; 42.7] | 525 | (74.9) | [71.5; 78.1] | |
| – Number no (n (%) and [95% CI]) | 1006 | (55.1) | [52.7; 57.4] | 169 | (24.1) | [21.0; 27.5] | |
| – Number with no data (n (%) and [95% CI]) | 83 | (4.5) | [3.6; 5.6] | 7 | (1.0) | [0.4; 2.0] | |
| – Reason for hospitalization: birth or complication of pregnancy (n (%) and [95% CI] | 449 | (24.6) | [22.6; 26.6] | 513 | (73.2) | [69.7; 76.4] | < 0.001 |
| – Reason for hospitalization: other (n (%) and [95% CI]) | 19 | (1.0) | [0.6; 1.6] | 13 | (1.9) | [1.0; 3.2] | 0.15 |
| – Reason for hospitalization: COVID-19 | < 0.001 | ||||||
| – Number yes (n (%) and [95% CI]) | 318 | (17.4) | [15.7; 19.2] | 0 | (0.0) | [0.0; 0.5] | |
| – Number no (n (%) and [95% CI]) | 1503 | (82.3) | [80.4; 84.0] | 701 | (100.0) | [99.5; 100.0] | |
| – Number with no data (n (%) and [95% CI]) | 6 | (0.3) | [0.1; 0.7] | 0 | (0.0) | [0.0; 0.5] | |
| COVID-19-associated treatment or monitoring | < 0.001 | ||||||
| – Number yes (n (%) and [95% CI]) | 348 | (19.0) | [17.3; 20.9] | 0 | (0.0) | [0.0; 0.5] | |
| – Number no (n (%) and [95% CI]) | 1382 | (75.6) | [73.6; 77.6] | 693 | (98.9) | [97.8; 99.5] | |
| – Number with no data (n (%) and [95% CI]) | 97 | (5.3) | [4.3; 6.4] | 8 | (1.1) | [0.5; 2.2] | |
| COVID-19-associated pneumonia (according to clinical judgement) | < 0.001 | ||||||
| – Number yes (n (%) and [95% CI]) | 129 | (7.1) | [5.9; 8.3] | 0 | (0.0) | [0.0; 0.5] | |
| – Number no (n (%) and [95% CI]) | 1548 | (84.7) | [83.0; 86.3] | 693 | (98.9) | [97.8; 99.5] | |
| – Number with no data (n (%) and [95% CI]) | 150 | (8.2) | [7.0; 9.6] | 8 | (1.1) | [0.5; 2.2] | |
| COVID-19-related oxygen therapy | < 0.001 | ||||||
| – Number yes (n (%) and [95% CI]) | 168 | (9.2) | [7.9; 10.6] | 0 | (0.0) | [0.0; 0.5] | |
| – Number no (n (%) and [95% CI]) | 1553 | (85.0) | [83.3; 86.6] | 693 | (98.9) | [97.8; 99.5] | |
| – Number with no data (n (%) and [95% CI]) | 106 | (5.8) | [4.8; 7.0] | 8 | (1.1) | [0.5; 2.2] | |
| COVID-19-associated transfer to intensive care unit (ICU) | < 0.001 | ||||||
| – Number yes (n (%) and [95% CI]) | 102 | (5.6) | [4.6; 6.7] | 0 | (0.0) | [0.0; 0.5] | |
| – Number no (n (%) and [95% CI]) | 1619 | (88.6) | [87.1; 90.0] | 693 | (98.9) | [97.8; 99.5] | |
| – Number with no data (n (%) and [95% CI]) | 106 | (5.8) | [4.8; 7.0] | 8 | (1.1) | [0.5; 2.2] | |
| COVID-19-related invasive ventilation (IV) | < 0.001 | ||||||
| – Number yes (n (%) and [95% CI]) | 45 | (2.5) | [1.8; 3.3] | 0 | (0.0) | [0.0; 0.5] | |
| – Number no (n (%) and [95% CI]) | 1674 | (91.6) | [90.3; 92.9] | 693 | (98.9) | [97.8; 99.5] | |
| – Number with no data (n (%) and [95% CI]) | 108 | (5.9) | [4.9; 7.1] | 8 | (1.1) | [0.5; 2.2] | |
| COVID-19-associated death | < 0.001 | ||||||
| – Number yes (n (%) and [95% CI]) | 4 | (0.2) | [0.1; 0.6] | 0 | (0.0) | [0.0; 0.5] | |
| – Number no (n (%) and [95% CI]) | 1713 | (93.8) | [92.6; 94.8] | 693 | (98.9) | [97.8; 99.5] | |
| – Number with no data (n (%) and [95% CI]) | 110 | (6.0) | [5.0; 7.2] | 8 | (1.1) | [0.5; 2.2] | |
| Venous thromboembolisms (VTE) | 0.001 | ||||||
| – Number yes (n (%) and [95% CI]) | 16 | (0.9) | [0.5; 1.4] | 4 | (0.6) | [0.2; 1.5] | |
| – Number no (n (%) and [95% CI]) | 1507 | (82.5) | [80.7; 84.2] | 534 | (76.2) | [72.8; 79.3] | |
| – Number with no data (n (%) and [95% CI]) | 304 | (16.6) | [15.0; 18.4] | 163 | (23.3) | [20.2; 26.6] | |
*Percentages given with 95% confidence intervals according to Clopper–Pearson. P-values for the difference between asymptomatic cases and women with COVID-19 were calculated using the Mann–Whitney test (for ordinal or constant variables) or Fisher‘s exact test or the chi-squared test (for dichotomous and nominal variables).
IQR, Interquartiles range; 95% CI, 95% confidence interval
eFigure 1.
Flowchart of the processing of the data from period 1, showing the underlying inclusion and exclusion criteria
Influence of gestational age at infection on the maternal and perinatal outcome
Data on 1827 cases of COVID-19 from period 1 were used for modeling. Of these women, 1773 (97%) were unvaccinated, while the vaccination status of the remaining 54 women (3%) was unknown. At the time of data extraction, pregnancy outcome and perinatal outcome data of 1497 women (82.0%) were available. The maternal and perinatal characteristics, together with the 701 asymptomatic women included in period 1, are described in the eMethods and listed in eTables 1 and 2.
eTable 2. Selected perinatal and newborn characteristics and outcome parameters for 2161 women from data collection period 1, comparing COVID-19 with asymptomatic infection*.
| COVID-19 | Asymptomatic infection | p-value | |||||
| n | % or median | 95% CI or quartiles | n | % or median | 95% CI or quartiles | ||
| Number (n) | 1497 | 664 | |||||
| Gestational age at time of positive test result (median [IQR]) | 30.0 | [22.0; 38.0] | 38.0 | [34.0; 39.0] | < 0.001 | ||
| – Number of women with positive test result GA < 24 + 0 weeks (n (%) and [95% ci]) | 417 | (27.9) | [25.6; 30.2] | 57 | (8.6) | [6.6; 11.0] | < 0.001 |
| – Number of women with positive test result GA < 37 + 0 weeks (n (%) and [95% ci]) | 1179 | (78.8) | [76.6; 80.8] | 236 | (35.5) | [31.9;]39.3 | < 0.001 |
| Gestational age at delivery (median [IQR]) | 39.0 | [37.0; 40.0] | 39.0 | [38.0; 40.0] | 0.23 | ||
| Miscarriage (GA < 24 + 0 weeks) (n (%) and [95% ci]) | 17 | (1.1) | [0.7; 1.8] | 11 | (1.7) | [0.8; 2.9] | 0.31 |
| Stillbirth (GA ≥ 24 + 0 weeks) (n (%) and [95% CI]) | 18 | (1.2) | [0.7; 1.9] | 3 | (0.5) | [0.1; 1.3] | 0.19 |
| Premature birth (GA 24 + 0 to 37 + 0 weeks) (n (%) and [95% CI]) | 218 | (14.6) | [12.8; 16.5] | 85 | (12.8) | [10.4; 15.6] | 0.35 |
| 24 + 0 to 27 + 6 weeks (n (%) and [95% CI]) | 14 | (0.9) | [0.5; 1.6] | 7 | (1.1) | [0.4; 2.2] | 0.57 |
| 28 + 0 to 31 + 6 weeks (n (%) and [95% CI]) | 29 | (1.9) | [1.3; 2.8] | 8 | (1.2) | [0.5; 2.4] | 0.31 |
| 32 + 0 to 36 + 6 weeks (n (%) and [95% CI]) | 175 | (11.7) | [10.1; 13.4] | 70 | (10.5) | [8.3; 13.1] | 0.46 |
| Term birth ≥ 37 + 0 weeks (n (%) and [95% CI]) | 1260 | (84.2) | [82.2; 86.0] | 567 | (85.4) | [82.5; 88.0] | 0.48 |
| Abortion or premature birth within 4 weeks of positive test result (whole cohort; women with positive result ≥ 37 + 0 weeks counted as NA) | < 0.001 | ||||||
| – Yes (n (%) and [95% CI]) | 145 | (9.7) | [8.2; 11.3] | 76 | (11.4) | [9.1; 14.1] | |
| – No (n (%) and [95% CI]) | 1034 | (69.1) | [66.7; 71.4] | 160 | (24.1) | [20.9; 27.5] | |
| – NA (n (%) and [95% CI]) | 318 | (21.2) | [19.2; 23.4] | 428 | (64.5) | [60.7; 68.1] | |
| Mode of birth | 0.06 | ||||||
| – Spontaneous delivery (n (%) and [95% CI]) | 849 | (56.7) | [54.2; 59.2] | 404 | (60.8) | [57.0; 64.6] | |
| – Surgical vaginal delivery (n (%) and [95% CI]) | 88 | (5.9) | [4.7; 7.2] | 34 | (5.1) | [3.6; 7.1] | |
| – Primary cesarean section (n (%) and [95% CI]) | 326 | (21.8) | [19.7; 24.0] | 109 | (16.4) | [13.7; 19.5] | |
| – Secondary cesarean section (n (%) and [95% CI]) | 205 | (13.7) | [12.0; 15.5] | 100 | (15.1) | [12.4; 18.0] | |
| – Emergency cesarean section (n (%) and [95% CI]) | 23 | (1.5) | [1.0; 2.3] | 9 | (1.4) | [0.6; 2.6] | |
| – NA (n (%) and [95% CI]) | 6 | (0.4) | [0.1; 0.9] | 8 | (1.2) | [0.5; 2.4] | |
| Iatrogenic delivery (induced birth or cesarean section) (n (%) and [95% CI]) | 0.08 | ||||||
| – Yes (n (%) and [95% CI]) | 586 | (39.1) | [36.7; 41.7] | 228 | (34.3) | [30.7; 38.1] | |
| – No (n (%) and [95% CI]) | 902 | (60.3) | [57.7; 62.7] | 430 | (64.8) | [61.0; 68.4] | |
| – NA (n (%) and [95% CI]) | 9 | (0.6) | [0.3; 1.1] | 6 | (0.9) | [0.3; 2.0] | |
| Iatrogenic delivery due to maternal SARS-CoV-2 infection (n (%) and [95% CI]) | < 0.001 | ||||||
| – Yes (n (%) and [95% CI]) | 63 | (4.2) | [3.2; 5.4] | 5 | (0.8) | [0.2; 1.7] | |
| – No (n (%) and [95% CI]) | 1414 | (94.5) | [93.2; 95.6] | 651 | (98.0) | [96.7; 99.0] | |
| – NA (n (%) and [95% CI]) | 20 | (1.3) | [0.8; 2.1] | 8 | (1.2) | [0.5; 2.4] | |
| Surgical delivery due to maternal SARS-CoV-2 infection (n (%) and [95% CI]) | < 0.001 | ||||||
| – Yes (n (%) and [95% CI]) | 53 | (3.5) | [2.7; 4.6] | 3 | (0.5) | [0.1; 1.3] | |
| – No (n (%) and [95% CI]) | 1427 | (95.3) | [94.1; 96.3] | 653 | (98.3) | [97.1; 99.2] | |
| – NA (n (%) and [95% CI]) | 17 | (1.1) | [0.7; 1.8] | 8 | (1.2) | [0.5; 2.4] | |
| COVID-19-related ending of pregnancy within 4 weeks of positive test result | < 0.001 | ||||||
| – Yes (n (%) and [95% CI]) | 59 | (3.9) | [3.0; 5.1] | 5 | (0.8) | [0.2; 1.7] | |
| – No (n (%) and [95% CI]) | 1426 | (95.3) | [94.1; 96.3] | 654 | (98.5) | [97.2; 99.3] | |
| – NA (n (%) and [95% CI]) | 12 | (0.8) | [0.4; 1.4] | 5 | (0.8) | [0.2; 1.7] | |
| Transfer of newborn to NICU (n (%) and [95% CI]) | 0.26 | ||||||
| – Yes (n (%) and [95% CI]) | 233 | (15.6) | [13.8; 17.5] | 86 | (13.0) | [10.5; 15.7] | |
| – No (n (%) and [95% CI]) | 1210 | (80.8) | [78.7; 82.8] | 556 | (83.7) | [80.7; 86.5] | |
| – NA (n (%) and [95% CI]) | 54 | (3.6) | [2.7; 4.7] | 22 | (3.3) | [2.1; 5.0] | |
| Transfer of newborn to NICU due to maternal SARS-CoV-2 infection | 0.48 | ||||||
| – Yes (n (%) and [95% CI]) | 29 | (1.9) | [1.3; 2.8] | 8 | (1.2) | [0.5; 2.4] | |
| – No (n (%) and [95% CI]) | 1430 | (95.5) | [94.4; 96.5] | 642 | (96.7) | [95.0; 97.9] | |
| – NA (n (%) and [95% CI]) | 38 | (2.5) | [1.8; 3.5] | 14 | (2.1) | [1.2; 3.5] | |
| Transfer of newborn to NICU within 4 weeks of mother’s infection | 0.35 | ||||||
| – Yes (n (%) and [95% CI]) | 133 | (8.9) | [7.5; 10.4] | 72 | (10.8) | [8.6; 13.5] | |
| – No (n (%) and [95% CI]) | 1316 | (87.9) | [86.1; 89.5] | 572 | (86.1) | [83.3; 88.7] | |
| – NA (n (%) and [95% CI]) | 48 | (3.2) | [2.4; 4.2] | 20 | (3.0) | [1.8; 4.6] | |
| Transfer of newborn to NICU or antenatal or postnatal death of newborn within 4 weeks of mother’s infection | 0.003 | ||||||
| – Yes (n (%) and [95% CI]) | 158 | (10.6) | [9.0; 12.2] | 85 | (12.8) | [10.4; 15.6] | |
| – No (n (%) and [95% CI]) | 1322 | (88.3) | [86.6; 89.9] | 559 | (84.2) | [81.2; 86.9] | |
| – NA (n (%) and [95% CI]) | 17 | (1.1) | [0.7; 1.8] | 20 | (3.0) | [1.8; 4.6] | |
| Treatment escalation: respiratory support of newborn | 0.02 | ||||||
| – No treatment (n (%) and [95% CI]) | 1330 | (88.8) | [87.1; 90.4] | 608 | (91.6) | [89.2; 93.6] | |
| – Oxygen therapy (n (%) and [95% CI]) | 22 | (1.5) | [0.9; 2.2] | 3 | (0.5) | [0.1; 1.3] | |
| – High flow (n (%) and [95% CI]) | 5 | (0.3) | [0.1; 0.8] | 3 | (0.5) | [0.1; 1.3] | |
| – CPAP (n (%) and [95% CI]) | 86 | (5.7) | [4.6; 7.0] | 24 | (3.6) | [2.3; 5.3] | |
| – Intubation (n (%) and [95% CI]) | 23 | (1.5) | [1.0; 2.3] | 10 | (1.5) | [0.7; 2.8] | |
| – NA (n (%) and [95% CI]) | 31 | (2.1) | [1.4; 2.9] | 16 | (2.4) | [1.4; 3.9] | |
| Positive SARS-CoV-2 PCR test among all tested newborns | 0.47 | ||||||
| – Yes (n (%) and [95% CI]) | 15 | (4) | [2.2; 6.5] | 6 | (3.0) | [1.1; 6.4] | |
| – No (n (%) and [95% CI]) | 351 | (93.1) | [90.1; 95.4] | 191 | (95.5) | [91.6; 97.9] | |
| – NA (n (%) and [95% CI]) | 11 | (2.9) | [1.5; 5.2] | 3 | (1.5) | [0.3; 4.3] | |
| Neonatal death (n (%) and [95% CI]) | 5 | (0.3) | [0.1; 0.8] | 4 | (0.6) | [0.2; 1.5] | 0.51 |
*Percentages given with 95% confidence intervals according to Clopper–Pearson. P-values for the difference between asymptomatic cases and women with COVID-19 were calculated using the Mann–Whitney test (for ordinal or constant variables) or Fisher‘s exact test or the chi-squared test (for dichotomous and nominal variables).
CPAP, Continuous positive airway pressure; GA, gestational age in weeks; IQR, interquartile range; NA, not available; NICU, neonatal intensive care unit; 95% CI, 95% confidence interval
Results
Maternal COVID-19-related outcomes in period 1
Seven hundred thirty-eight women (40.4%) were admitted to the hospital within 4 weeks after a positive test result, 318 (17.4%) of them due to COVID-19. The risk of admission due to COVID-19 rose from 5% (95% confidence interval [3; 10]) in the first trimester to around 22% [17; 26] in the early third trimester. Women diagnosed at GA 30 weeks had the highest risk (Figure 1a, Table 1). The risk of COVID-19 pneumonia went up from < 2% [1; 6] at GA 12 weeks to around 14% [11; 18] in the early third trimester. The highest risk was for women diagnosed at GA 30 weeks (Figure 1b, Table 1). The risk of transfer to an intensive care unit (ICU) or of invasive ventilation (IV) rose from < 0.3% (ICU; [0; 3]) and < 0.1% (IV; [0; 3]) respectively in the first trimester to around 7% [5; 11] for ICU (Figure 1c, Table 1) and around 4% [2; 8] for IV in the early third trimester, with the highest risk at GA 29 weeks (Figure 1d, Table 1). The odds ratios for all comparisons of the results at GA 32 versus 22 weeks are shown in Table 1. The probability of a pregnancy ending within 4 weeks after a positive COVID-19 test result rose between the early second trimester and the early third trimester, reaching 4% at GA 32 weeks (Figure 1e, Table 1). Besides increasing GA, higher BMI before pregnancy and higher maternal age have unfavorable effects on the outcome (table 1). Four patients died, three of whom had become infected in the early third trimester (GA 29–30 weeks).
Figure 1.
The probability (y-axis) of different COVID-19-related events after a positive SARS-CoV-2 test result at defined gestational ages (GA in weeks; x-axis) among the group of 1827 patients with COVID-19 from period 1 (infection between beginning of pandemic and 24 August 2021), showing in each case the regression line (unbroken line) and the 95% confidence intervals (dotted lines; based on logistic regression with restricted cubic spline terms)
a) COVID-19-related hospitalization within 4 weeks of positive test result
b) Pneumonia and/or oxygen therapy
c) Transfer to intensive care unit
d) Invasive ventilation
e) COVID-19-related delivery within 4 weeks of positive test result
Table 1. The risk of unfavorable COVID-19-related maternal and perinatal events*1.
| Predictor variables | OR | 95% CI | p-value | |
| Maternal events | ||||
| COVID-19-related hospital admission within 4 weeks after infection | GA at time of positive test | 1.4 | [1.2; 1.7] | < 0.0001 |
| Maternal pregravid BMI | 1.9 | [1.4; 2.4] | 0.002 | |
| COVID-19 pneumonia | GA at time of positive test | 1.8 | [1.4; 2.4] | < 0.0001 |
| Maternal pregravid BMI | 2.2 | [1.5; 3.1] | 0.0005 | |
| Maternal age | 1.3 | [1.0; 1.7] | 0.02 | |
| Transfer to intensive care unit | GA at time of positive test | 1.8 | [1.1; 2.0] | 0.0002 |
| Maternal pregravid BMI | 2.6 | [1.6; 4.5] | 0.007 | |
| Maternal age | 1.7 | [1.2; 2.3] | 0.001 | |
| Invasive ventilation | GA at time of positive test | 2.1 | [0.8; 5.4] | 0.0001 |
| Maternal pregravid BMI | 3.6 | [1.4; 9.2] | 0.0008 | |
| Maternal age | 1.9 | [1.1; 3.1] | 0.02 | |
| Iatrogenic delivery due to COVID-19 within 4 weeks after infection | GA at time of positive test | 5.7 | [1.8; 18.4] | < 0.0001 |
| Maternal pregravid BMI | 1.4 | [1.1; 1.9] | 0.02 | |
| Maternal age | 1.5 | [1.0; 2.2] | 0.04 | |
| Perinatal events | ||||
| End of pregnancy before GA 37 + 0 weeks, within 4 weeks after infection | GA at time of positive test*2 | 3.0 | [2.2; 4.2] | < 0.0001 |
| Maternal pregravid BMI | 1.3 | [1.1; 1.6] | 0.005 | |
| Maternal age | 1.2 | [0.9; 1.5] | 0.05 | |
| Delivery within 4 weeks after infection with transfer to NICU, or antenatal or postnatal death of child | GA at time of positive test | 1.9 | [1.5; 2.6] | < 0.0001 |
| Maternal pregravid BMI | 1.4 | [1.2; 1.7] | 0.0003 | |
| Maternal diabetes mellitus | 4.9 | [1.7; 14.2] | 0.004 | |
*1 Results of logistic regressions for maternal and perinatal events. The p-values relate to the overall effect of the predictor variables concerned (log-linear and non-log-linear effects). The odds ratios (with 95% CI) relate to the time of infection (GA 32 weeks vs. 22 weeks) or to the quartiles of the sample (age: 35 vs. 28 years; BMI: 29 vs. 22 kg/m²).
*2 Only log-linear effect
BMI, Body mass index; CI, confidence interval; GA, gestational age in weeks; NICU, neonatal intensive care unit; OR, odds ratio
Figure 2.
The probability of a perinatal event (y-axis) of a perinatal event after a positive SARS-CoV-2 test result at defined gestational ages (GA in weeks; x-axis) among the group of 1497 symptomatic patients from period 1 (infection between beginning of pandemic and 24 August 2021), for whom data on the outcome of pregnancy were available at the time of data freezing, showing in each case the regression line (unbroken line) and the 95% confidence intervals (dotted lines; based on logistic regression with restricted cubic spline terms)
a) Premature ending of pregnancy within 4 weeks of positive SARS-CoV-2 test result up to a maximum GA of 36 + 6 weeks
b) Ending of pregnancy within 4 weeks of infection followed by transfer of the neonate to a neonatal intensive care unit (NICU), or antenatal or postnatal death of the child
Perinatal outcomes in data collection period 1
Among the 1497 women with a documented outcome of pregnancy, 17 had a miscarriage before GA 24 weeks and 218 gave birth prematurely (GA 24 + 0 to 36 + 6 weeks). The proportion of the 1179 women with a positive COVID-19 test result before GA 37 + 0 was 18.5%. Eighteen pregnancies (1.2%) ended in a stillbirth; two of these children had lethal chromosomal anomalies. Four live-born children died of various causes during the postnatal period. The risk of experiencing miscarriage or premature birth within 4 weeks after a positive test result rose from < 2% for infection in the first trimester to around 11% in the early third trimester and > 16% at GA 31 weeks (Figure 2a, Table 1). In 602 (40.2%) of the 1497 women, pregnancy ended within 4 weeks of diagnosis of infection. The risk of delivery within 4 weeks followed by transfer of the child to a NICU or death of the child before or after birth increased from around 6% [4; 9] for infection at GA 22 weeks to around 12% [9; 15] at 32 weeks and stayed at that level up to the calculated delivery date (Figure 2b, Table 1). The odds ratios for comparisons of the risk at GA 32 versus 22 weeks are shown in Table 1. Other risk factors were higher BMI, higher maternal age, and pre-existing maternal diabetes (table 1).
Comparison of COVID-19-specific outcomes in data collection periods 1 and 2
During the 18 months of period 1, 953 women became ill with COVID-19 in or after the 22nd week of gestation and contacted a hospital within 4 weeks after infection; the corresponding figure for the 6 months of period 2 was 969 women (efigure 3). While none of the women whose data were captured in period 1 had received vaccinations against COVID-19, 506 (51.2%) of those in period 2 were vaccinated (missing data in 75 cases). The proportion of pregnant women hospitalized for treatment of COVID-19 in period 2 was around a quarter less than in period 1. Moreover, fewer women had pneumonia, needed intensive care or invasive ventilation, or had mandatory delivery due to COVID-19 (table 2). There were no deaths among the women from period 2. Comparison of vaccinated and unvaccinated women from period 2 showed that the former had lower rates of hospitalization and pneumonia and were less likely to undergo delivery due to COVID-19 (table 2). The odds ratio of delivery because of COVID-19 between vaccinated and unvaccinated women was 0.30 ([0.10; 0.84]; p = 0.02) (table 3). There was no difference between the rates—very low in both groups—of intensive care treatment or invasive ventilation (table 2). The odds ratios for comparisons of other outcomes between unvaccinated and vaccinated women from periods 1 and 2 can be found in Table 3.
eFigure 3.
Flowchart of the data selection, showing the underlying inclusion and exclusion criteria, for the comparison of periods 1 and 2 taking account of vaccination status
Table 2. The frequency of relevant COVID-19-specific events in period 1 (from the beginning of the pandemic to 24 August 2021) and period 2 (17 January 2022 to 16 June 2022) among women who became ill with COVID-19 after GA22 weeks and were entered in the registry via contact with a recruiting hospital within 4 weeks after infection.
| Period 1 unvaccinated | Period 2 unvaccinated | Period 2 vaccinated | P-value* | Adj. p-value | |
| Number (n) | 923 | 388 | 506 | ||
| Gestational age at time of infection (median [IQR]) | 35.0 [31.0; 38.0] | 36.0 [33.0; 38.0] | 36.0 [32.0; 38.0] | < 0.001 | |
| Hospital admission due to COVID-19 (n (%) | 271 (29.4) | 84 (21.6) | 35 (6.9) | < 0.0001 | < 0.0001 |
| COVID-19-associated treatment or monitoring (n [%]) | 282 (31.5) | 71 (18.6) | 27 (5.3) | < 0.0001 | < 0.0001 |
| COVID-19-associated pneumonia or oxygen therapy (n [%]) | 163 (18.3) | 13 (4.0) | 4 (0.8) | < 0.0001 | < 0.0001 |
| COVID-19-related intensive care treatment (ICU) (n [%]) | 92 (10.3) | 3 (0.8) | 4 (0.8) | < 0.0001 | < 0.0001 |
| COVID-related invasive ventilation (IV) (n [%]) | 39 (4.4) | 0 (0.0) | 1 (0.2) | 0.008 | 0.01 |
| COVID-19-associated death (n [%]) | 4 (0.5) | 0 (0.0) | 0 (0.0) | 0.19 | 0.08 |
| COVID-19 as reason for delivery (n [%]) | 59 (7.4) | 12 (3.8) | 5 (1.2) | 0.0001 | 0.0001 |
*P-values for unadjusted comparison of the three groups and adjusted for GA at the time of infection
Adj., adjusted; GA, gestational age; ICU intensive care unit; IQR, interquartile range
Table 3. The risk of relevant defined outcomes, calculated from the data of women who in data collection period 1 (from the beginning of the pandemic to 24 August 2021) and data collection period 2 (17 January 2022 to 16 June 2022) became ill with COVID-19 after GA 22 weeks and were entered in the registry via contact with a recruiting hospital within 4 weeks after infection*.
| Period 2 unvaccinated vs. period 1 unvaccinated | Period 2 vaccinated vs. period 2 unvaccinated | ||||||
| OR | [95% CI] | p | OR | [95% CI] | p | ||
| Hospital admission owing to COVID-19 | Unadjusted | 0.77 | [0.58; 1.04] | 0.09 | 0.23 | [0.15; 0.36] | < 0.0001 |
| Adjusted | 0.66 | [0.50; 0.88] | 0.004 | 0.27 | [0.18; 0.41] | < 0.0001 | |
| COVID-19-associated treatment or monitoring | Unadjusted | 0.56 | [0.42; 0.76] | 0.0002 | 0.22 | [0.14; 0.35] | < 0.0001 |
| Adjusted | 0.5 | [0.37; 0.67] | < 0.0001 | 0.25 | [0.15; 0.39] | < 0.0001 | |
| COVID-19-associated pneumonia or oxygen therapy | Unadjusted | 0.21 | [0.12; 0.39] | < 0.0001 | 0.18 | [0.06; 0.56] | 0.003 |
| Adjusted | 0.19 | [0.10; 0.33] | < 0.0001 | 0.2 | [0.06; 0.62] | 0.005 | |
| COVID-19-related intensive care treatment (ICU) | Unadjusted | 0.08 | [0.03; 0.26] | < 0.0001 | 0.9 | [0.20; 4.07] | 0.89 |
| Adjusted | 0.07 | [0.02; 0.22] | < 0.0001 | 1.00 | [0.22; 4.51] | 0.99 | |
| COVID-related invasive ventilation (IV) | Unadjusted | NA | 0.73 | NA | 0.82 | ||
| Adjusted | NA | 0.75 | NA | 0.83 | |||
| COVID-19 as reason for delivery | Unadjusted | 0.51 | [0.27; 0.96] | 0.04 | 0.3 | [0.10; 0.85] | 0.04 |
| Adjusted | 0.5 | [0.26; 0.94] | 0.03 | 0.3 | [0.10; 0.84] | 0.02 | |
*The odds ratios (with 95% confidence interval) relate to the comparisons between unvaccinated women in period 2 and unvaccinated women in period 1 and between vaccinated and unvaccinated women in period 2. The p-values for pairwise comparisons are either unadjusted or adjusted for GA at the time of infection.
GA, Gestational age; ND, no data; OR, odds ratio; vs., versus; 95% CI, 95% confidence interval
Discussion
Pregnant women constitute a group of patients that require particular consideration in the COVID-19 pandemic. Medications for COVID-19 have mostly not been tested in pregnancy, and great care must be taken to ensure they are indicated (6, 13), taking account of the actual COVID-19-specific risk. The study reported here, conducted at hospitals in Germany and at Linz, Austria, enables assessment of the risk of severe COVID-19 depending on the GA at the time of infection, the virus variant involved, and the woman’s vaccination status. For unvaccinated pregnant women with COVID-19, the risk of hospitalization and transfer to an ICU rises with increasing GA at the time of infection with SARS-CoV-2. In period 1, one of every five women infected at GA 30 weeks was hospitalized and one in 15 received intensive care treatment. The 4.0% rate of transfer to ICU among all women observed in the pre-omicron phase of the pandemic corresponds with findings from other countries. In a study of 2352 women with a positive SARS-CoV-2 test result from 17 hospitals in the USA, Metz et al. found a 3.7% rate of maternal transfer to ICU (14). In a Scottish prospective cohort study, 2.1% of 4950 women with confirmed SARS-CoV-2 infection had to be transferred to ICU. None of the women infected in the first trimester of pregnancy were involved, but 4.3% of those in the third trimester (15). Furthermore, modeling of the CRONOS data shows that the rate of ICU treatment goes down again for women infected later than GA 30 weeks. The reason for this may be that as fetal maturity increases, obstetricians are more generous in offering birth in severe COVID-19. One can theorize, with great caution, that delivery has a positive effect on the severity of COVID-19, for example through optimization of treatment options (e.g., the possibility of prone positioning of women on ventilation) (12). A noticeable consequence of these iatrogenic deliveries is an increasing proportion of premature births. In comparison with women not infected with SARS-CoV-2, Metz et al. describe a 3.7% higher rate of premature birth (14). In our sample, one in every seven women delivered prematurely—a rate somewhat lower than that stated by Metz et al. (17.7%) and slightly higher than found by Stock et al. (10.2%). The Scottish study showed an elevated rate of perinatal mortality (stillbirth or neonatal death), namely 8 per 1000 births versus the background rate of 5.6 per 1000 for all births during the same period. Stock et al. also evaluated in isolation the extended perinatal mortality among 5766 vaccinated women, finding a lower rate of 4.3 per 1000 births (15). However, their study also documented a lower vaccination rate in pregnant women: in October 2021, 32.3% of pregnant women were fully vaccinated compared with 77.4% for non-pregnant women of the same age. The finding that acceptance of vaccination is not high among pregnant women coincides with our own experience (16) and the results of the present study. In the 6 months of period 2, covering the omicron phase of the pandemic, almost half of the women in the CRONOS Registry were unvaccinated. Although the observed rate of severe illness in the omicron phase is very low, a protective effect of vaccination can be inferred from the data. With reference to the existing evidence on the safety of mRNA vaccines in pregnancy (8, 17, 18), the results presented here can be put to good use in informing pregnant women about the risks posed by infection of the unvaccinated.
This study evaluated data from 130 departments of obstetrics at hospitals providing different levels of care. Around one third of all births in Germany take place in these institutions. Almost all pregnant women in Germany sooner or later attend a hospital. Despite this distinct strength, the study also features limitations. Although complete documentation including the status of women after infection in early pregnancy is possible, it cannot be guaranteed for various reasons (e.g., bias in invitation for follow-up visits or in study inclusion). Furthermore, the number of undetected infections with SARS-CoV-2 among women in the early weeks of pregnancy is particularly high. Overestimation of the proportion of severe cases from the first trimester to the beginning of the third trimester can therefore be assumed and may explain the observed relative reduction in severe illness at the end of pregnancy. Asymptomatic cases are very often detected when the women concerned attend a hospital due to (pre-existing) complications of pregnancy or to give birth. In the early weeks of pregnancy these “incidental COVIDs” contribute to an elevation of the premature birth rate with COVID-19, not because of COVID-19. Owing to the over-representation in late pregnancy, however, asymptomatic women still distort the results regarding GA-related events even at this time. To counteract this bias, our study focused on COVID-19 patients who displayed symptoms. The core messages of the study, especially the elevated relative risk for women with COVID-19 as pregnancy progresses, can thus be considered valid. The quality of the data was guaranteed by monitoring, weekly assessment, and close communication with the individual hospitals regarding the documentation of severe events (serious maternal illness, death of child, infection of child).
Conclusion
Advanced pregnancy (GA > 22 weeks) at the time of infection, together with high BMI and high maternal age, is a major risk factor for a severe course of COVID-19. The virus variant also plays a part: in the omicron phase of the pandemic there were fewer cases of severe illness in pregnant women than was the case with earlier variants. In addition to the less virulent variant, accumulated experience in managing the disease and improved treatment strategies may have played a role. Furthermore, vaccinated women less often need COVID-19-related inpatient treatment or are delivered because of COVID-19 than unvaccinated women. The data underscore the recommendation of the German Standing Commission on Vaccination (STIKO) that pregnant women should (from the second trimester) be vaccinated against COVID-19 (8).
The data presented here may support physicians in their attempts to increase acceptance of vaccination for pregnant women or when considering drug treatments such as administration of monoclonal antibodies (13). Given the uncertainty regarding new variants of the virus and the possible occurrence of new waves of COVID-19 in fall 2022, the coming weeks and months should be used not only to offer vaccinations to women of reproductive age but also to make sure they understand the risks involved in contracting a SARS-CoV-2 infection in pregnancy, with considerable consequences for the unborn child.
The CRONOS Registry shows the feasibility of representative documentation of data from selected groups in Germany under the direction of professional associations. Continuation of this established registry infrastructure, also between pandemic phases, will—provided adequate funding of data acquisition and analysis is forthcoming—be of considerable medical benefit in the care of women who are pregnant, have recently given birth, or are breastfeeding (7). The membership of the CRONOS Network is shown in the eBox.
eBOX. CRONOS Network.
Ulrich Pecks, Nadine Mand, Thomas Kolben, Mario Rüdiger, Peter Oppelt, Janine Zöllkau, Astrid Dempfle, CRONOS Network: Michael Abou-Dakn, Clara Backes, Panagiota Batrani, Susanne Beckmann, Martin A. Berghaeuser, Cahit Birdir, Thordis Bloecker, Nina Bock, Michael K. Bohlmann, Dörthe Brüggmann, Bettina Buchfink, Lisa Butscher-Sprung, Maria Delius, Annika Dittrich, Iris Dressler-Steinbach, Elisabeth Edeler, Franz Edler von Koch, Hans-Walter Eibach, Sabine Enengl, Julia Engel, Charlotte Engelbrecht, Sarah Engler-Hauschild, Franziska Fettke, Andreas Funk, Julia Günther, Alexander Hein, Cathleen Heinemann, Lars Hellmeyer, Ulrich Hennig, Tino Hentrich, Vanessa Hepp, Antonella Iannaccone, Tizian Jahreis, Peter Jakubowski, Anja Jebens, Magdalena C. Jegen, Sophie Kage, Constantin S. von Kaisenberg, Corinna Keil, Carolin Kladt, Celine Kohll, Hans-Christian Kolberg, Katrina Kraft, Mirjam Kunze, Bettina Kuschel, Ioannis Kyvernitakis, Katharina Lang, Ann C. Longardt, Laura Lüber, Katharina Lüdemann, Rolf F. Maier, Sara Fill Malfertheiner, Jula Manz, Maike Manz, Filiz Markfeld Erol, Sabrina Marquardt, Valerie Meister, Nina Mingers, Christine A. Morfeld, Mirjam Morgen, Cora F. Mueller, Thorsten Orlikowsky, Monika Palz-Fleige, Olaf Parchmann, Babett Ramsauer, Athena Rejaey, Lydia Remtisch, Stefan P. Renner, Manuela F. Richter, Charlotte Rohlwink, Jens Rohne, Claudia Roll, Anne Runkel, Ute M. Schaefer-Graf, Ralf L Schild, Saskia Schmidt, Markus Schmidt, Michael O. Schneider, Susanne Schrey-Petersen, Carolin Stegmann, Diana G. Schwarz, Sven Seeger, Gregor Seliger, Diana A. Solomon, Marina Sourouni, Sonja Spieß, Johanna Stelbrink, Johannes Stubert, Jeannette Teeuwen-Mutter, Anna Treptow, Nadja Uhlig, Silke Wegener, Carolina Weßling, Katharina Weizsäcker, Julia Weyermanns, Jennifer L. Winkler, Vincent Winkler, Feline Wowretzko, Katja Zmarsly, Joachim Zucker-Reimann and many more.
Supplementary Material
eMethods
Method
Starting on 3 April 2020, staff of the participating hospitals can enter data on pregnant women who test positive for SARS-CoV-2 in the CRONOS COVID-19 registry. Since 98% of women expecting a child in Germany make contact with a hospital at some time during their pregnancy, at the latest when childbirth is impending, CRONOS aims to achieve complete recruitment of all women who are known to have been infected with SARS-CoV-2 during the current pregnancy or have COVID-19 at the time of presentation to the hospital. At the time of writing (16 June 2022), 175 hospitals in Germany and Kepler University Hospital in Linz, Austria, have pledged to take part. The 130 hospitals that have so far contributed data to CRONOS hosted 264 364 births in 2021, i.e., 33.2% of the 795 517 births in Germany in that year.
An electronic case report form (eCRF) was developed specifically for CRONOS using the cloud-based electronic data capture platform provided by castoredc.com (Amsterdam, Netherlands). This form comprises 13 categories with a total of 252 questions on SARS-CoV-2 testing, COVID-19 symptoms, general and obstetric history, comorbidities, results of investigations in the hospital, treatment and course of the illness, childbirth, perinatal outcome, and the puerperium. The principal investigator established a central monitoring unit at University Hospital Schleswig-Holstein, Kiel. The incoming data are assessed on a weekly basis. In the event of implausible data entries or documentation of serious events such as severe maternal complications necessitating intensive care, stillbirth, or infection of the newborn child, the hospital concerned is contacted directly to verify the data. On 15 December 2020 the eCRF was updated to include extra questions, e.g., vaccinations against COVID-19.
Prevention of bias
To avoid selection bias, the registry aims to achieve non-selective complete documentation, including follow-up data. Recruitment adherence is reinforced by weekly assessments and feedback to all participating hospitals. Furthermore, the quality of data entry is ensured by means of plausibility analyses and by contact with the originating hospital whenever “red flag” data (e.g., severe maternal complications or stillbirth) are entered. To minimize potential sources of bias, a large number of potentially relevant confounders are collected and considered in the statistical analyses. Part of the distortion may be attributable to collider bias (e1). As the due date approaches, the virtually complete screening in the hospitals yields a representative sample. However, women who become infected relatively early in pregnancy (GA < 35 weeks) are greatly under-represented in the registry. These women go to the hospital (and are thus included in the registry) either because they need treatment for COVID-19 or because they have complications of pregnancy. Asymptomatic women, in particular, are almost always documented owing to complications of pregnancy. Therefore, among women with asymptomatic COVID-19 early in pregnancy those with complications are hugely over-represented. This can lead to apparently paradoxical associations especially in the group of asymptomatic women, e.g., large numbers of premature births. For this reason the data analysis focused on women with COVID-19, i.e., those who showed symptoms in (temporal) connection with the infection or in whom a SARS-CoV-2 test was requested because of symptoms. In this group, it can be assumed that the proportion of cases of severe illness is overestimated among women with infection earlier in the course of pregnancy: women with severe illness in the first trimester are highly likely to be included in the registry. Many more women with mild symptoms will remain undetected in the first trimester, tilting the balance in the favor of severe cases.
Statistics
For the purpose of descriptive analysis, continuous variables were expressed in terms of median and quartiles, discrete variables as number and percentage. The 95% confidence intervals for proportions were calculated using the Clopper–Pearson method (e2). Differences between investigated groups were tested using the Mann–Whitney U-test for continuous and ordinal discrete variables and the chi-squared test or Fisher’s exact test for dichotomous and nominal variables. The numbers of asymptomatic women and women with COVID-19 were depicted in bar charts according to the GA at the time of infection. In the sample of pregnant women ill with COVID-19, the risk for various COVID-19-related and perinatal events depending on GA at infection, maternal age, BMI before or at the beginning of pregnancy, specific previous illnesses (cardiovascular disease, e.g., arterial hypertension; pre-existing diabetes mellitus; pulmonary disease, e.g., bronchial asthma), and impairment by comorbidity (on a four-point ordinal scale) was modeled by logistic regression. Non-log-linear (non-exponential) effects of GA at the time of infection (and, if applicable, those of age and BMI) were modelled by restricted cubic spline terms (e3). Non-significant variables (at alpha = 0.05) were excluded by backward selection. P-values for the overall effect of a parameter (linear and non-linear effects) were calculated using the likelihood-ratio test for nested models. Because for almost all of the endpoints considered the risk does not increase log-linearly with the GA at the time of a positive test result, it is not possible to give one odds ratio (OR) that expresses the change in risk over the whole pregnancy as a single number. Therefore, on obstetric and neonatological grounds we selected time points during gestation defined by extrauterine viability of the fetus (GA 22 + 0 and 24 + 0 weeks) and by the time of pregnancy-prolonging management in the case of severe pregnancy-relevant pathology (32 + 0 and 34 + 0 weeks). Independently, our experience of COVID-19 from the previous year led us to expect a peak of severe illness with COVID-19 at GA 32 weeks, and for this reason odds ratios were calculated from the respective final model for infection at GA 32 weeks versus infection at GA 22 weeks. Odds ratios were also calculated for age 35 years versus 28 years and for BMI 29 kg/m² versus 22 kg/m² (corresponding to the upper and lower quartiles of the sample). The inverse logit function was used for estimation of the absolute risks depending on GA at the time of infection, using the respective median for each of the other variables contained in the model. The graphs of these risks show the 95% confidence range. The risk for various COVID-19-related events depending on data collection period or vaccination status (in period 2) was also modeled by logistic regression and adjusted for GA at time of infection (as restricted cubic spline terms).
Results
Overall, data on 3481 women were evaluated. In data collection period 1, from 3 April 2020 to 24 August 2021, data on 2819 patients were registered. The entries were checked for plausibility by means of comparison and search algorithms. Eight duplicates were identified, verified, and removed. One case did not fulfill the inclusion criteria. After exclusion of patients with no information on GA at the time of infection (n = 98) and of patients in whom antibody screening revealed a past infection at an unknown time (n = 62), 2650 datasets remained for further analysis. Of these women, 2341 (88.3%) were diagnosed by PCR testing, 62 (2.3%) by rapid flow testing, and 45 (1.7%) by antibody testing in combination with previous clinical symptoms. For 202 women (7.6%) there were no data on test procedures. The datasets were divided into women with COVID-19 and those who were asymptomatic. In 122 cases (4.6%) the presence or absence of symptoms was not specified. Among the 1827 women who became ill with COVID-19 the diagnosis was made by PCR testing in 1618 cases (88.6%), by rapid flow testing in 49 cases (2.7%), and by antibody testing in combination with previous clinical symptoms in 45 cases (2.5%). The data on 115 women (6.3%) included no mention of testing procedures. At the time of data acquisition, 367 women had not yet given birth. The data processing is summarized in eFigure 1. The numbers of women with COVID-19 and of asymptomatic women from period 1 can be found in eFigure 2. Important supplementary maternal and neonatal variables with descriptive comparisons of defined events in asymptomatically infected women and women who became ill with COVID-19 can be found in eTables 1 and 2.
Of the 1063 women who became ill with COVID-19 and made contact with a hospital within 4 weeks after infection, 953 were infected after GA 22 weeks, while the vaccination status of 30 women was unknown. The data on 923 unvaccinated women were obtained for further comparisons with data from period 2. Data from period 2 (17 January 2022 to 16 June 2022) with a positive test result were used for comparison with the omicron variant of the virus. Following the same criteria, a final cohort of 969 women was drawn from the data of 2218 women. Of these women, 388 were unvaccinated, 506 were vaccinated, and in 75 cases the registry data contained no mention of vaccination status. The algorithm for data procession is shown schematically in eFigure 3.
eFigure 2.
The number of SARS-CoV-2-positive pregnant women in the cohort from data collection period 1 (y-axis) by gestational age in weeks (GA) at the time of the positive test result (x-axis). Black bars, women with COVID-19; gray bars, asymptomatic women. Note the continuous rise in the number of women with COVID-19 with increasing GA. However, asymptomatic women are over-represented after GA 37 weeks. These “incidental COVIDs” were mostly detected on screening after hospitalization for delivery.
Acknowledgments
Translated from the original German by David Roseveare.
Footnotes
Conflict of interest statement
Prof. Pecks has received funding for parts of the CRONOS project from the Krumme Foundation (F379155) and the state of Schleswig-Holstein (K128002). Prof. Pecks has received fees for lectures on the topic of this article from Roche Diagnostics and Jenapharm. Prof. Pecks is research officer and board member of the German Society of Perinatal Medicine (Deutsche Gesellschaft für Perinatale Medizin, DGPM) and in this capacity manages the CRONOS Registry.
Prof. Rüdiger is an unpaid member of the board of the DGPM.
The remaining authors declare that no conflict of interest exists.
References
- 1.Zambrano LD, Ellington S, Strid P, et al. Update: Characteristics of symptomatic women of reproductive age with laboratory-confirmed SARS-CoV-2 infection by pregnancy status—United States, January 22-October 3, 2020. MMWR Morb Mortal Wkly Rep. 2020;69:1641–1647. doi: 10.15585/mmwr.mm6944e3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Allotey J, Stallings E, Bonet M, et al. Clinical manifestations, risk factors, and maternal and perinatal outcomes of coronavirus disease 2019 in pregnancy: living systematic review and meta-analysis. BMJ. 2020;370 doi: 10.1136/bmj.m3320. m3320. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Wei SQ, Bilodeau-Bertrand M, Liu S, Auger N. The impact of COVID-19 on pregnancy outcomes: a systematic review and meta-analysis. CMAJ. 2021;193:E540–E548. doi: 10.1503/cmaj.202604. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Pecks U, Kuschel B, Mense L, Oppelt P, Rüdiger M, CRONOS-Netzwerk: Pregnancy and SARS CoV-2 infection in Germany—the CRONOS registry Dtsch Arztebl Int. 2020;117:841–842. doi: 10.3238/arztebl.2020.0841. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Zöllkau J, Hagenbeck C, Hecher K, et al. Update on recommendations for SARS-CoV-2/COVID-19 during pregnancy, birth and childbed. Z Geburtshilfe Neonatol. 2020;224:217–222. doi: 10.1055/a-1199-6404. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Zöllkau J, Hagenbeck C, Hecher K, et al. [Recommendations for SARS-CoV-2/COVID-19 during pregnancy, birth and childbed—Update November 2021 (Long Version)] Z Geburtshilfe Neonatol. 2022;226:e1–e35. doi: 10.1055/a-1688-9398. [DOI] [PubMed] [Google Scholar]
- 7.AWMF 015/092. S2k Leitlinie SARS-CoV-2 in der Schwangerschaft, Geburt und Wochenbett. www.awmf.org/leitlinien/detail/ll/015-092.html . 2022 [Google Scholar]
- 8.Takla A, Matysiak-Klose D, Bogdan C, et al. Empfehlung und Begründung der STIKO zur Impfung gegen COVID-19 von Schwangeren und Stillenden. Epidemiol Bull. 2021;38:10–36. [Google Scholar]
- 9.Control ECfDPa. Data on SARS-CoV-2 variants in the EU/EEA. www.ecdc.europa.eu/en/publications-data/data-virus-variants-covid-19-eueea (last accessed on 30 March 2022) [Google Scholar]
- 10.Kleinwechter HJ, Weber KS, Mingers N, et al. Gestational diabetes mellitus and COVID-19: results from the COVID-19-Related Obstetric and Neonatal Outcome Study (CRONOS) Am J Obstet Gynecol. 2022 doi: 10.1016/j.ajog.2022.05.027. S0002 9378(22)00372-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Ziert Y, Abou-Dakn M, Backes C, et al. Maternal and neonatal outcomes of pregnancies with COVID-19 after medically assisted reproduction: results from the prospective COVID-19-Related Obstetrical and Neonatal Outcome Study. Am J Obstet Gynecol. 2022 S0002 9378(22)00300-3 doi: 10.1016/j.ajog.2022.04.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Sitter M, Pecks U, Rüdiger M, et al. Pregnant and postpartum women requiring intensive care treatment for COVID-19-first data from the CRONOS-Registry. J Clin Med. 2022;11 doi: 10.3390/jcm11030701. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Zöllkau J, Reuken PA, Schleußner E, Groten T. Monoclonal SARS-CoV-2 antibodies in pregnancy—a case series. Dtsch Arztebl Int. 2022;119:113–114. doi: 10.3238/arztebl.m2022.0110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Metz TD, Clifton RG, Hughes BL, et al. Association of SARS-CoV-2 infection with serious maternal morbidity and mortality from obstetric complications. JAMA. 2022;327:748–759. doi: 10.1001/jama.2022.1190. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Stock SJ, Carruthers J, Calvert C, et al. SARS-CoV-2 infection and COVID-19 vaccination rates in pregnant women in Scotland. Nat Med. 2022;28:504–512. doi: 10.1038/s41591-021-01666-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Schaal NK, Zöllkau J, Hepp P, Fehm T, Hagenbeck C. Pregnant and breastfeeding women’s attitudes and fears regarding the COVID-19 vaccination. Arch Gynecol Obstet. 2021:1–8. doi: 10.1007/s00404-021-06297-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Dagan N, Barda N, Biron-Shental T, et al. Effectiveness of the BNT162b2 mRNA COVID-19 vaccine in pregnancy. Nat Med. 2021;27:1693–1695. doi: 10.1038/s41591-021-01490-8. [DOI] [PubMed] [Google Scholar]
- 18.Goldshtein I, Steinberg DM, Kuint J, et al. Association of BNT162b2 COVID-19 vaccination during pregnancy with neonatal and early infant outcomes. JAMA Pediatr. 2022;176:470–477. doi: 10.1001/jamapediatrics.2022.0001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- E1.Tönnies T, Kahl S, Kuss O. Collider bias in observational studies: consequences for medical research Part 30 of a series on evaluation of scientific publications. Dtsch Arztebl Int. 2022;119:107–112. doi: 10.3238/arztebl.m2022.0076. [DOI] [PMC free article] [PubMed] [Google Scholar]
- E2.Clopper CJ, Pearson ES. The use of confidence or fiducial limits illustrated in the case of the binomial. Biometrika. 1934;26 [Google Scholar]
- E3.Harrell Jr FE. Regression modeling strategies. Cham: Springer International Publishing. (Springer Series in Statistics) 2015 [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eMethods
Method
Starting on 3 April 2020, staff of the participating hospitals can enter data on pregnant women who test positive for SARS-CoV-2 in the CRONOS COVID-19 registry. Since 98% of women expecting a child in Germany make contact with a hospital at some time during their pregnancy, at the latest when childbirth is impending, CRONOS aims to achieve complete recruitment of all women who are known to have been infected with SARS-CoV-2 during the current pregnancy or have COVID-19 at the time of presentation to the hospital. At the time of writing (16 June 2022), 175 hospitals in Germany and Kepler University Hospital in Linz, Austria, have pledged to take part. The 130 hospitals that have so far contributed data to CRONOS hosted 264 364 births in 2021, i.e., 33.2% of the 795 517 births in Germany in that year.
An electronic case report form (eCRF) was developed specifically for CRONOS using the cloud-based electronic data capture platform provided by castoredc.com (Amsterdam, Netherlands). This form comprises 13 categories with a total of 252 questions on SARS-CoV-2 testing, COVID-19 symptoms, general and obstetric history, comorbidities, results of investigations in the hospital, treatment and course of the illness, childbirth, perinatal outcome, and the puerperium. The principal investigator established a central monitoring unit at University Hospital Schleswig-Holstein, Kiel. The incoming data are assessed on a weekly basis. In the event of implausible data entries or documentation of serious events such as severe maternal complications necessitating intensive care, stillbirth, or infection of the newborn child, the hospital concerned is contacted directly to verify the data. On 15 December 2020 the eCRF was updated to include extra questions, e.g., vaccinations against COVID-19.
Prevention of bias
To avoid selection bias, the registry aims to achieve non-selective complete documentation, including follow-up data. Recruitment adherence is reinforced by weekly assessments and feedback to all participating hospitals. Furthermore, the quality of data entry is ensured by means of plausibility analyses and by contact with the originating hospital whenever “red flag” data (e.g., severe maternal complications or stillbirth) are entered. To minimize potential sources of bias, a large number of potentially relevant confounders are collected and considered in the statistical analyses. Part of the distortion may be attributable to collider bias (e1). As the due date approaches, the virtually complete screening in the hospitals yields a representative sample. However, women who become infected relatively early in pregnancy (GA < 35 weeks) are greatly under-represented in the registry. These women go to the hospital (and are thus included in the registry) either because they need treatment for COVID-19 or because they have complications of pregnancy. Asymptomatic women, in particular, are almost always documented owing to complications of pregnancy. Therefore, among women with asymptomatic COVID-19 early in pregnancy those with complications are hugely over-represented. This can lead to apparently paradoxical associations especially in the group of asymptomatic women, e.g., large numbers of premature births. For this reason the data analysis focused on women with COVID-19, i.e., those who showed symptoms in (temporal) connection with the infection or in whom a SARS-CoV-2 test was requested because of symptoms. In this group, it can be assumed that the proportion of cases of severe illness is overestimated among women with infection earlier in the course of pregnancy: women with severe illness in the first trimester are highly likely to be included in the registry. Many more women with mild symptoms will remain undetected in the first trimester, tilting the balance in the favor of severe cases.
Statistics
For the purpose of descriptive analysis, continuous variables were expressed in terms of median and quartiles, discrete variables as number and percentage. The 95% confidence intervals for proportions were calculated using the Clopper–Pearson method (e2). Differences between investigated groups were tested using the Mann–Whitney U-test for continuous and ordinal discrete variables and the chi-squared test or Fisher’s exact test for dichotomous and nominal variables. The numbers of asymptomatic women and women with COVID-19 were depicted in bar charts according to the GA at the time of infection. In the sample of pregnant women ill with COVID-19, the risk for various COVID-19-related and perinatal events depending on GA at infection, maternal age, BMI before or at the beginning of pregnancy, specific previous illnesses (cardiovascular disease, e.g., arterial hypertension; pre-existing diabetes mellitus; pulmonary disease, e.g., bronchial asthma), and impairment by comorbidity (on a four-point ordinal scale) was modeled by logistic regression. Non-log-linear (non-exponential) effects of GA at the time of infection (and, if applicable, those of age and BMI) were modelled by restricted cubic spline terms (e3). Non-significant variables (at alpha = 0.05) were excluded by backward selection. P-values for the overall effect of a parameter (linear and non-linear effects) were calculated using the likelihood-ratio test for nested models. Because for almost all of the endpoints considered the risk does not increase log-linearly with the GA at the time of a positive test result, it is not possible to give one odds ratio (OR) that expresses the change in risk over the whole pregnancy as a single number. Therefore, on obstetric and neonatological grounds we selected time points during gestation defined by extrauterine viability of the fetus (GA 22 + 0 and 24 + 0 weeks) and by the time of pregnancy-prolonging management in the case of severe pregnancy-relevant pathology (32 + 0 and 34 + 0 weeks). Independently, our experience of COVID-19 from the previous year led us to expect a peak of severe illness with COVID-19 at GA 32 weeks, and for this reason odds ratios were calculated from the respective final model for infection at GA 32 weeks versus infection at GA 22 weeks. Odds ratios were also calculated for age 35 years versus 28 years and for BMI 29 kg/m² versus 22 kg/m² (corresponding to the upper and lower quartiles of the sample). The inverse logit function was used for estimation of the absolute risks depending on GA at the time of infection, using the respective median for each of the other variables contained in the model. The graphs of these risks show the 95% confidence range. The risk for various COVID-19-related events depending on data collection period or vaccination status (in period 2) was also modeled by logistic regression and adjusted for GA at time of infection (as restricted cubic spline terms).
Results
Overall, data on 3481 women were evaluated. In data collection period 1, from 3 April 2020 to 24 August 2021, data on 2819 patients were registered. The entries were checked for plausibility by means of comparison and search algorithms. Eight duplicates were identified, verified, and removed. One case did not fulfill the inclusion criteria. After exclusion of patients with no information on GA at the time of infection (n = 98) and of patients in whom antibody screening revealed a past infection at an unknown time (n = 62), 2650 datasets remained for further analysis. Of these women, 2341 (88.3%) were diagnosed by PCR testing, 62 (2.3%) by rapid flow testing, and 45 (1.7%) by antibody testing in combination with previous clinical symptoms. For 202 women (7.6%) there were no data on test procedures. The datasets were divided into women with COVID-19 and those who were asymptomatic. In 122 cases (4.6%) the presence or absence of symptoms was not specified. Among the 1827 women who became ill with COVID-19 the diagnosis was made by PCR testing in 1618 cases (88.6%), by rapid flow testing in 49 cases (2.7%), and by antibody testing in combination with previous clinical symptoms in 45 cases (2.5%). The data on 115 women (6.3%) included no mention of testing procedures. At the time of data acquisition, 367 women had not yet given birth. The data processing is summarized in eFigure 1. The numbers of women with COVID-19 and of asymptomatic women from period 1 can be found in eFigure 2. Important supplementary maternal and neonatal variables with descriptive comparisons of defined events in asymptomatically infected women and women who became ill with COVID-19 can be found in eTables 1 and 2.
Of the 1063 women who became ill with COVID-19 and made contact with a hospital within 4 weeks after infection, 953 were infected after GA 22 weeks, while the vaccination status of 30 women was unknown. The data on 923 unvaccinated women were obtained for further comparisons with data from period 2. Data from period 2 (17 January 2022 to 16 June 2022) with a positive test result were used for comparison with the omicron variant of the virus. Following the same criteria, a final cohort of 969 women was drawn from the data of 2218 women. Of these women, 388 were unvaccinated, 506 were vaccinated, and in 75 cases the registry data contained no mention of vaccination status. The algorithm for data procession is shown schematically in eFigure 3.