Abstract
Objective: Pregnancy is a risk factor for acute respiratory failure (ARF) following severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection. We hypothesised that SARS-CoV-2 viral load in the respiratory tract might be higher in pregnant intensive care unit (ICU) patients with ARF than in non-pregnant ICU patients with ARF as a consequence of immunological adaptation during pregnancy.
Design: Single-centre, retrospective observational case–control study.
Setting: Adult level 3 ICU in a French university hospital.
Participants: Eligible participants were adults with ARF associated with coronavirus disease 2019 (COVID-19) pneumonia.
Main outcome measure: The primary endpoint of the study was viral load in pregnant and non-pregnant patients.
Results: 251 patients were included in the study, including 17 pregnant patients. Median gestational age at ICU admission amounted to 28 + 3/7 weeks (interquartile range [IQR], 26 + 1/7 to 31 + 5/7 weeks). Twelve patients (71%) had an emergency caesarean delivery due to maternal respiratory failure. Pregnancy was independently associated with higher viral load (-4.6 ± 1.9 cycle threshold; P < 0.05). No clustering or over-represented mutations were noted regarding SARS-CoV-2 sequences of pregnant women. Emergency caesarean delivery was independently associated with a modest but significant improvement in arterial oxygenation, amounting to 32 ± 12 mmHg in patients needing invasive mechanical ventilation. ICU mortality was significantly lower in pregnant patients (0 v 35%; P < 0.05). Age, Simplified Acute Physiology Score (SAPS) II score, and acute respiratory distress syndrome were independent risk factors for ICU mortality, while pregnancy status and virological variables were not.
Conclusions: Viral load was substantially higher in pregnant ICU patients with COVID-19 and ARF compared with non-pregnant ICU patients with COVID-19 and ARF. Pregnancy was not independently associated with ICU mortality after adjustment for age and disease severity.
Pregnancy has emerged as an important risk factor for acute respiratory failure (ARF) and intensive care unit (ICU) admission following severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection.1 About 8% of pregnant women with coronavirus disease 2019 (COVID-19) require ICU admission,2, 3 with ICU mortality ranging between 0% and 15%,4, 5, 6, 7, 8, 9, 10, 11 but with substantial neonatal morbidity.5, 6, 8, 12
Management of pregnant ICU patients with severe COVID-19 and a viable fetus relies on expert opinions,13, 14, 15, 16 and is based on individualised delivery planning depending on gestational age. Whether pregnant patients with COVID-19 and ARF benefit from emergency delivery remains largely conjectural at this time, although a recent study on 17 pregnant patients with COVID-19 and acute respiratory distress syndrome (ARDS) who required invasive mechanical ventilation suggested that emergency delivery slowed the deterioration of oxygenation over time.11 Furthermore, since pregnant patients are usually excluded from clinical trials of therapeutic agents for SARS-CoV-2 infection, the benefit of antiviral administration in this population remains unknown. This issue is critical since the susceptibility of pregnant patients to severe forms of COVID-19 may be related to immunological impairment during pregnancy.17
We hypothesised that COVID-19-associated ARF during pregnancy is a particular phenotype of ARF, and that viral load in the respiratory tract might be higher in pregnant patients than in non-pregnant patients as a consequence of pregnancy-associated immunological alterations.
Methods
Study aims
The primary objective of the study was to identify whether viral load is higher in pregnant ICU patients with COVID-19 and ARF, independent from confounding factors, than in control non-pregnant ICU patients with COVID-19 and ARF.
The secondary objectives of the study were:
-
•
to identify whether severe forms of COVID-19 during pregnancy are related to specific mutations in the viral genome;
-
•
to evaluate the impact of emergency caesarean delivery on arterial oxygenation in patients undergoing invasive mechanical ventilation; and
-
•
to identify whether pregnancy is an independent risk factor for ICU mortality in patients with COVID-19 and ARF.
Study design
We undertook a single-centre, retrospective observational case–control study of patients admitted between 1 January and 15 September 2021 to an adult ICU in a French university hospital with a level 4 regional neonatal intensive care unit (NICU). The study was conducted in accordance with the amended Declaration of Helsinki, complied with the STROBE criteria,18 and was approved by our institutional Ethics Committee (Hospices Civils de Lyon, CSE HCL – #21_640). The patients were informed of their right to oppose to the retrospective use of their data for non- interventional research, according to French regulation.
Patients
Eligible participants were adults consecutively admitted to the ICU with ARF associated with COVID-19 confirmed by a SARS-CoV-2 real-time reverse transcriptase polymerase chain reaction (RT-PCR) test in respiratory swab or a rapid antigen test, and pneumonia on chest imaging. Exclusion criteria were previous inclusion during a prior ICU stay, time between first symptoms and ICU admission > 21 days, and lack of consent to participate. Patients were classified into the pregnant group if they were pregnant at ICU admission or if they were admitted to the ICU immediately after delivery.
Study endpoints
The primary endpoint was viral load assessed by the cycle threshold value on the first semi-quantitative RT-PCR performed during the hospital stay. The secondary endpoints of the study were:
-
•
mutations in SARS-CoV-2 genomes;
-
•
worst arterial partial pressure of oxygen (PaO2)/fraction of inspired oxygen (FiO2) ratio during the 24 hours before intubation, and on the calendar day after intubation in pregnant patients undergoing delivery and non-pregnant patients; and
-
•
ICU mortality.
Statistical analysis
Statistical analyses were performed using R version 4.0.1. A two-sided P < 0.05 was chosen for statistical significance. We anticipated that the study would have at least an 80% power to detect an effect size of at least 0.8 on the primary judgment criteria if at least 200 controls and 15 pregnant patients were included.
Categorical variables were expressed as count (percentage) and continuous data were presented as medians (interquartile range [IQR]). Data were compared between groups with the Fisher exact test for categorical variables and the t test or Mann–Whitney test for continuous variables, as indicated. Multivariable analyses were performed by selection of variables, with P < 0.2 in univariate analysis and other variables deemed relevant, and by a backward selection algorithm. The confounding effect of sex and age was adjusted in all models. Additional information regarding the methods is provided in the Online Appendix.
Results
Population characteristics
Missing data per variables are reported in the Online Appendix, table 1. Data collection stopped with the end of the fourth wave of COVID-19 in our area. Two- hundred and fifty-one patients were included in the study, of which 17 were pregnant (Online Appendix, figure 1). Pregnant patients were significantly younger, had a lower Simplified Acute Physiology Score (SAPS) II score and higher PaO2/FiO2 ratio during the first 24 hours in the ICU than the non-pregnant patients (Table 1). Requirement for invasive mechanical ventilation was not statistically different between groups (59% v 56% in pregnant and non-pregnant patients respectively).
Table 1.
Patients’ characteristics
| Variables | Whole population | No pregnancy | Pregnancy |
|---|---|---|---|
| Total number of patients | 251 | 234 | 17 |
| Age, years, median (IQR) | 61 (52-71) | 62 (54-71) | 33 (32-40)⁎ |
| Sex, male | 164 (65%) | 164 (70%) | 0⁎ |
| Antiviral treatment | |||
| Casivirimab/indevimab | 2 (1%) | 2 (1%) | 0 |
| Convalescent plasma therapy | 12 (5%) | 11 (5%) | 1 (6%) |
| SARS-CoV-2 vaccination | |||
| One dose or non-vaccinated | 243 (97%) | 227 (97%) | 16 (94%) |
| Fully vaccinated | 8 (3%) | 7 (3%) | 1 (6%) |
| Immunosuppression | 17 (7%) | 17 (7%) | 0 |
| BMI, kg/m2, median (IQR) | 29 (26-34) | 30 (26-34) | 29 (27-34) |
| Time between first symptoms and ICU admission, days, median (IQR) | 8 (6-11) | 8 (6-11) | 7 (6-10) |
| SAPS II score at ICU admission, median (IQR) | 38 (28-51) | 39 (29-52) | 24 (21-27)⁎ |
| SOFA score at ICU admission, median (IQR) | 3 (2-4) | 3 (2-5) | 3 (2-3) |
| Worst PaO2/FiO2 ratio during the first 24 hours in ICU, mmHg, median (IQR) | 82 (65-131) | 80 (64-130) | 118 (74-158)† |
| Most invasive ventilatory support during ICU stay | |||
| Standard oxygen therapy | 41 (16%) | 37 (16%) | 4 (23%) |
| High flow oxygen | 69 (28%) | 66 (28%) | 3 (18%) |
| Invasive mechanical ventilation | 141 (56%) | 131 (56%) | 10 (59%) |
| ARDS during ICU stay | 136 (54%) | 126 (54%) | 10 (59%) |
| ARDS severity†’ | |||
| Mild | 1 (1%) | 0 | 1 (10%) |
| Moderate | 26 (19%) | 22 (17%) | 4 (40%) |
| Severe | 109 (80%) | 104 (83%) | 5 (50%) |
| High flow oxygen failure | 98 (62%) | 89 (61%) | 9 (75%) |
| Duration of high flow oxygen therapy, days, median (IQR) | 2 (1-4) | 2 (1-4) | 1 (1-2) |
| Worst PaO2/FiO2 ratio during the 24 hours before intubation, mmHg, median (IQR) | 53 (49-57) | 52 (49-55) | 57 (55-65)⁎ |
| Worst PaO2/FiO2 ratio during the calendar day following intubation, mmHg, median (IQR) | 95 (74-123) | 94 (74-120) | 108 (77-147) |
| Prone position during ICU stay‡ | 135 (96%) | 126 (96%) | 9 (90%) |
| Prone position sessions during ICU stay,‡ median (IQR) | 10 (5-17) | 11 (6-18) | 3 (2-4)⁎ |
| Inhaled nitric oxide‡ | 64 (46%) | 60 (46%) | 4 (40%) |
| ECMO during ICU stay | 41 (16%) | 39 (17%) | 2 (12%) |
| Vasopressor during ICU stay | 136 (54%) | 127 (54%) | 9 (53%) |
| RRT during ICU stay | 62 (25%) | 62 (26%) | 01 |
| Invasive mechanical ventilation duration, days, median (IQR) | 19 (10-36) | 19 (10-38) | 11 (6-15) |
| Ventilator-free days at day 60 from ICU admission, median (IQR) | 42 (0-60) | 39 (0-60) | 54 (47-60) |
| ICU length of stay, days, median (IQR) | 10 (4-32) | 10 (4-32) | 7 (4-15) |
| ICU mortality | 82 (33%) | 82 (35%) | 0† |
| SARS CoV-2 variant† | |||
| Alpha | 79 (54%) | 75 (56%) | 4 (33%) |
| Delta | 37 (25%) | 30 (22%) | 7 (59%) |
| Other variants | 30 (21%) | 29 (22%) | 1 (8%) |
| Time between symptoms and RT-PCR, days, median (IQR) | 11 (7-14) | 11 (7-14) | 9 (6-10) |
| Sampling site | |||
| Upper respiratory tract | 127 (67%) | 115 (66%) | 10 (83%) |
| Lower respiratory tract | 62 (33%) | 60 (34%) | 2 (17%) |
| Viral load (cycle threshold), median (IQR) | 27 (22-32) | 27 (23-33) | 22 (21-25)† |
ARDS = acute respiratory distress syndrome; BMI = body mass index; ECMO = extracorporeal membrane oxygenation; Fio2 = fraction of inspired oxygen; ICU = intensive care unit; IQR = interquartile range; Pao2 = arterial partial pressure of oxygen; RRT = renal replacement therapy; RT-PCR = real-time reverse transcriptase polymerase chain reaction; SAPS = Simplified Acute Physiology Score; SARS-CoV-2 = severe acute respiratory syndrome coronavirus 2; SOFA = Sequential Organ Failure Assessment.
P < 0.001.
P < 0.05.
In patients needing invasive mechanical ventilation.
Biological data
Semi-quantitative RT-PCR and quantitative RT-PCR tests were highly correlated in a subset of 37 patients (Spearman correlation coefficient, -0.86; P < 0.0001; Online Appendix, figure 2). Distribution of SARS-CoV-2 variants were significantly different between pregnant and non-pregnant patients (Table 1). The Delta variant was more frequently identified in pregnant patients (59% v 22%), while the Alpha variant was more frequently identified in non-pregnant patients (33% v 56%). Viral load was significantly higher in pregnant patients, with a median difference amounting to 5 cycle thresholds (95% CI, 2-8). Pregnancy was independently associated with higher viral load (-4.6 ± 1.9 cycle threshold; P < 0.05; Table 2) after adjustment for confounders. Time between symptoms onset and RT-PCR, SARS-CoV-2 Delta variant, and upper respiratory tract site of sampling were also independently associated with viral load (Table 2), while age and sex were not.
Table 2.
Variables associated with cycle threshold values
| Variables | Univariate model coefficient ± SE | Univariate P value | Multivariable model coefficient ± SE | Multivariable P value |
|---|---|---|---|---|
| Intercept | - | - | 23.1 ± 2.4 | < 0.001 |
| Time between symptoms onset and RT-PCR (per day) | 0.6 ± 0.1 | < 0.001 | 0.6 ± 0.1 | < 0.001 |
| Site of sampling (ref = lower respiratory tract) | 0.3 ± 1.0 | 0.75 | 2.5 ± 1.0 | < 0.05 |
| Pregnancy (ref = No) | -5.5 ± 1.8 | < 0.01 | -4.6 ± 1.9 | < 0.05 |
| SARS-CoV-2 variant (ref = Alpha) | ||||
| Delta | -3.2 ± 1.1 | < 0.01 | -2.9 ± 1.0 | < 0.01 |
| Other variants | -0.5 ± 1.1 | 0.66 | 0.7 ± 1.0 | ns |
| Immunosuppression (ref = No) | -6.0 ± 1.8 | < 0.01 | -3.2 ± 1.6 | 0.05 |
| Age (per 10 years) | 0.0 ± 0.4 | 0.92 | -0.5 ± 0.4 | ns |
| Sex (ref = male) | -2.5 ± 1.0 | < 0.05 | 0.1 ± 0.9 | ns |
ns = not significant; ref = reference; RT-PCR = real-time reverse transcriptase polymerase chain reaction; SARS-CoV-2 = severe acute respiratory syndrome coronavirus 2; SE = standard error
The analysis of SARS-CoV-2 sequences did not identify clustering regarding the phylogeny of sequences (Online Appendix, figure 3). In addition, no significant overrepresented mutation was identified in SARS-CoV-2 genomes of pregnant women. Neutralising autoantibodies against interferon (IFN)-α and IFN-ω were detected in the serum of one patient among the 13 analysed (8%).
Management and clinical course of pregnant patients and newborns
Ten pregnant patients (59%) were mechanically ventilated, among whom nine (52%) were intubated because of high flow oxygen therapy failure, and one to allow for urgent delivery in relation with ARF. Emergency caesarean delivery was performed in 12 patients (71%), exclusively for maternal respiratory failure. Patient characteristics as a function of emergency caesarean delivery status are reported in Table 3. Patients in the emergency caesarean delivery group were more severely hypoxaemic, and all patients with worst PaO2/ FiO2 ratio below 150 mmHg during the first 24 hours in the ICU underwent emergency caesarean delivery (Figure 1, A). Overall, preterm birth was observed in 81% of the patients and was more frequent in the emergency caesarean delivery group (92% v 50%). Most preterm birth occurred before 32 weeks’ gestational age (69%), and newborn admission to the NICU was common (65%), but no newborn fatalities were reported (Table 3).
Table 3.
Maternal and fetal/neonatal characteristics as a function of emergency caesarean delivery status
| Variables | Pregnant patients | Non-emergency caesarean delivery | Emergency caesarean delivery |
|---|---|---|---|
| Total number of patients | 17 | 5 | 12 |
| Maternal age at ICU admission, median (IQR) | 33 (32-40) | 33 (32-35) | 33 (32-40) |
| Maternal comorbid conditions | |||
| None | 11 (65%) | 2 (40%) | 9 (75%) |
| Obesity | 3 (17%) | 1 (20%) | 2 (17%) |
| Other | 3 (17%) | 2 (40%) | 1 (8%) |
| Worst PaO2/FiO2 ratio during first 24 hours in the ICU, mmHg, median (IQR) | 118 (74-158) | 177 (158-248) | 106 (73-129) |
| Invasive mechanical ventilation | 10 (59%) | 1 (20%) | 9 (75%) |
| Antenatal steroids for fetal maturation | 15 (88%) | 4 (80%) | 11 (92%) |
| Time between antenatal steroids and delivery, days, median (IQR) | 3 (1-5) | 73 (71-80) | 2 (1-3) |
| Time between ICU admission and delivery, days, median (IQR) | 2 (1-12) | 71 (59-76) | 1 (1-2) |
| Anaesthesia modality during emergency caesarean delivery | |||
| General anaesthesia | 8 (67%) | na | 8 (67%) |
| Spinal anaesthesia | 4 (33%) | na | 4 (33%) |
| Lowest maternal SpO2 during emergency caesarean delivery, %, median (IQR) | 88 (77-90) | na | 88 (77-90) |
| Gestational age at ICU admission, weak + day, median (IQR) | 28 + 3/7 (26 + 1/7 to 31 + 5/7) |
28 + 1/7 (25 + 0/7 to 28 + 5/7) |
28 + 3/7 (26 + 2/7 to 31 + 4/7) |
| Gestational age at delivery, weak + day, median (IQR) | 30 + 3/7 (27 + 2/7 to 34 + 4/7) |
37 + 3/7 (33 + 2/7 to 39 + 4/7) |
29 + 3/7 (27 + 2/7 to 31 + 4/7) |
| Preterm birth before 37 weeks' gestational age | 13 (81%) | 2 (50%) | 11 (92%) |
| Preterm birth before 32 weeks' gestational age | 11 (69%) | 1 (25%) | 10 (83%) |
| Fetal death in utero | 1 (6%) | 1 (20%) | 0 |
| Apgar score 1 min after delivery, median (IQR) | 6 (2-9) | 9 (9-10) | 5 (1-7) |
| Apgar score 5 min after delivery, median (IQR) | 9 (6-10) | 10 (10-10) | 9 (6-9) |
| Apgar score 10 min after delivery, median (IQR) | 10 (8-10) | 10 (10-10) | 9 (7-10) |
| Newborn body weight, g, median (IQR) | 1560 (935-2735) | 3490 (3480-3655) | 1375 (865-1715) |
| Newborn admission to NICU | 11 (65%) | 0 | 11 (92%) |
| Oxygen therapy in newborns | 11 (65%) | 0 | 11 (92%) |
| Time between delivery and oxygen weaning in newborn, days, median (IQR) | 48 (28-65) | na | 48 (28-65) |
| Newborn hospital death | 0 | 0 | 0 |
ARDS = acute respiratory distress syndrome; BMI = body mass index; FiO2 = fraction of inspired oxygen; ICU = intensive care unit; IQR = interquartile range; na = not applicable; NICU = neonatal intensive care unit; PaO2 = arterial partial pressure of oxygen; SpO2 = oxygen saturation measured by pulse oximetry. Statistical analyses were not performed between groups owing to the small sample size.
Figure 1.
Changes of arterial partial pressure of oxygen (PaO2 )/fraction of inspired oxygen (FiO2 ) ratio in pregnant patients (A), and in all intubated patients before and after intubation/delivery (B), Patients in the emergency delivery group had a caesaren delivery immediately after intubation
HFO = high flow oxygen; ICU = intensive care unit; O2 = oxygen; T1 = 24 hours before intubation; T2 = calendar day after intubation. * P < 0.05 v ICU day 1. † P < 0.05 v 24 hours before delivery.
Impact of delivery on arterial oxygenation
In pregnant patients, the PaO2/FiO2 ratio decreased significantly from the first 24 hours in the ICU to the 24 hours preceding emergency caesarean delivery and increased significantly after delivery except in one patient (Figure 1, A).
To take into account the confounding effect of mechanical ventilation in oxygenation improvement after delivery, pregnant patients needing invasive mechanical ventilation before delivery were compared with non-pregnant patients needing invasive mechanical ventilation for the worst PaO2/ FiO2 ratio during the 24 hours preceding intubation and on the calendar day following intubation (and delivery in pregnant patients). A significant interaction was found between the timing of measurement and pregnancy status (Figure 1, B), with a linear interaction coefficient amounting to 32 ± 12 mmHg; that is, the increase between before and after intubation in pregnant patients was higher by 32 ± 12 mmHg compared with non-pregnant patients (Online Appendix, table 2).
Mortality risk factors
Overall ICU mortality amounted to 33% and was significantly lower in pregnant than in non-pregnant patients (0% v 35%; P < 0.05). Univariate risk factors for ICU mortality are provided in the Online Appendix, table 3. Age, SAPS II score, and ARDS were independent risk factors for ICU mortality, while pregnancy status or virological variables were not (Table 4).
Table 4.
Multivariable risk factors for intensive care unit (ICU) mortality
| Variables | Odds ratio (95% CI) | P |
|---|---|---|
| Age (per one-year increment) | 1.06 (1.03-1.09) | < 0.001 |
| SAPS II subscore at ICU admission (per one-unit increment)⁎ | 1.04 (1.01-1.06) | < 0.001 |
| ARDS during ICU stay (ref = No) | 9.91 (4.46-22.01) | < 0.001 |
| Sex (ref = male) | 0.98 (0.49-1.95) | 0.95 |
ARDS = acute respiratory distress syndrome; ref = reference; SAPS = Simplified Acute Physiology Score. The following variables were entered in the multivariable model: pregnancy status, age, sex, body mass index, SAPS II subscore without age component, Sequential Organ Failure Assessment (SOFA) score at ICU admission, worst arterial partial pressure of oxygen (Pao2)/fraction of inspired oxygen (FiO2) ratio during the first 24 hours in the ICU, ARDS occurrence during ICU stay, viral load, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variant, time between first symptoms and ICU admission and immunosuppression status. Model calibration assessed with the Hosmer-Lemeshow test, P = 0.21. Model discrimination assessed with the C-statistic, 0.84.
Without age component.
Discussion
After adjustment for confounding factors, our study shows that the viral load in pregnant patients with ARF associated with COVID-19 was significantly and substantially higher than in non-pregnant patients with ARF associated with COVID-19. In addition, viral molecular characteristics were not significantly different between pregnant and non-pregnant patients, suggesting that host-related factors associated with pregnancy are involved in disease severity. Furthermore, emergency caesarean delivery was independently associated with a modest but significant improvement in arterial oxygenation, amounting to 32 ±12 mmHg in patients needing invasive mechanical ventilation. Moreover, pregnancy was not an independent risk factor for ICU mortality in patients with ARF associated with COVID-19.
Virological factors associated with a severe form of COVID-19 during pregnancy
In a prospective cohort study including 127 pregnancies, viral load was shown to be higher in symptomatic pregnant patients compared with asymptomatic ones.19 In the present study, viral load was independently and substantially higher in pregnant than in non-pregnant patients. This result is consistent with an impaired immune response to the SARS- CoV-2 during pregnancy and relates to several suspected mechanisms, including shift in CD4+ T cell population toward the Th2 phenotype, decrease in circulating natural killer and in circulating plasmacytoid dendritic cells, and alterations in the innate immune system, among others.17 The effect size of pregnancy regarding viral load was of similar magnitude as the effect size of immunosuppression (-4.5 ± 1.9 v-3.1 ± 1.6; Table 2). Based on viral molecular analyses, pregnant women were not infected by viruses carrying specific mutations differing from non-pregnant patients. This suggests that the increased viral load in pregnant women with ARF associated with COVID-19 was related to the immune host-response in this population. Interestingly, the relatively low occurrence of neutralising auto-antibodies against IFN-α and IFN-ω in our study suggests that other impaired immune mechanisms are responsible for high viral load and COVID-19 severity during pregnancy.20, 21 The over-representation of the SARS-CoV-2 Delta variant in pregnant patients in our study is also in line with a recent study identifying higher rate of serious morbidity and adverse perinatal outcomes in the period following its widespread diffusion compared with the preceding period.22 The higher viral load with the Delta variant in our study is also in line with previous reports,23 and this effect was taken into account in our multivariable analysis.
Oxygenation improvement associated with delivery
We report a high rate of emergency caesarean delivery in line with previous studies on ICU patients,6, 10 suggesting that the management of these patients regarding emergency caesarean delivery was similar to previous studies. Improvement of respiratory parameters after delivery in pregnant patients with ARF remains controversial. One study performed in ten patients who needed invasive mechanical ventilation identified a significant improvement in the oxygenation index at 12-15 hours after delivery, but most patients were mildly to moderately hypoxaemic and some of the respiratory conditions leading to intubation were rapidly reversible.24 Another case series of ten patients mechanically ventilated for ARF observed modest effects of delivery on maternal respiratory function with a 28% reduction in FiO2 at 24 hours after delivery.25 Finally, a recent study on 17 patients with ARF associated with COVID-19 showed that emergency caesarean delivery did not improve oxygenation but only slowed the deterioration of oxygenation over time.11 All these studies lacked a control group, and the use of control patients with the same risk factor for ARF in our study allowed us to identify a modest but significant positive impact of emergency caesarean delivery on oxygenation, the clinical relevance of which may be questionable.
ICU mortality of pregnant patients
In our study, ICU mortality of pregnant patients was in the lower range of previous studies.4, 5, 6, 7, 8, 9, 10, 11 However, policy regarding ICU admission may vary among national health care systems and over time as a function of the epidemic dynamics, hence baseline risk of death at ICU admission may be highly heterogeneous among studies. Nevertheless, the rate of requirement of invasive mechanical ventilation for pregnant patients in our study was within the range of previously published studies, suggesting similar respiratory severity. Surprisingly, no virological factors were associated with ICU mortality in our study, as opposed to previous studies performed during the first COVID-19 pandemic wave (with different SARS-CoV-2 variants; population, pregnant v non-pregnant; and design, repetitive v single measurements26, 27). Finally, pregnancy status was not independently associated with ICU mortality, suggesting that the low mortality rate of this subgroup is mostly explained by younger age according to our multivariable analysis, although our strategy regarding fetal extraction on occurrence of maternal respiratory worsening may have influenced outcome.
Strength and weaknesses
Our study had some limitations. A moderate rate of missing virological data should be acknowledged, although this has been taken into account by adequate statistical methods. The sample size of the pregnancy group was relatively small, as a consequence of both the rarity of the disease and the single-centre design, although the coexistence of the ECMO referral centre and of a level 4 NICU on site should have ensured that virtually all pregnant patients with ARF and a viable fetus in the regional area would have been referred to our centre. The control group differed substantially from the pregnancy group on important characteristics (age and sex notably), and this may have biased the results, although sex and age were adjusted for in our multivariable analysis. The impact of delivery on oxygenation could only be assessed in intubated patients, and our data do not provide any insight regarding non-intubated pregnant patients. The viral load was assessed by a semi-quantitative technique as the normalised quantitative reference technique is not performed routinely. However, both techniques were highly correlated in a subset of our samples. Finally, most of the patients were unvaccinated, and the impact of vaccination on viral load in pregnant women with ARF associated with COVID-19 remains unknown.
Nevertheless, the strengths of our study include consecutive patient enrolment over a well defined period, explicit inclusion and exclusion criteria, use of a sizable control group of non-pregnant ICU patients with COVID-19, availability of valuable semi-quantitative RT-PCR and genomic sequencing data for most of the study patients, and a consistent management algorithm of pregnant patients throughout the study period.
Clinical implications
The substantially higher viral load in pregnant patients with ARF associated with COVID-19 (of similar magnitude with that of immunosuppressed patients) would suggest a potential benefit of early antiviral treatment, as reported recently,28 but to date, this remains speculative. Regarding timing of delivery, our data confirm previous reports about the modest impact of delivery on maternal oxygenation,11, 25 although a slight increase in oxygenation related to delivery may be of clinical importance in mechanically ventilated patients with severe hypoxaemia as a life-saving procedure.
Conclusions
Viral load was significantly and substantially higher in pregnant patients with COVID-19 and ARF compared with non-pregnant patients. The SARS-CoV-2 Delta variant was over-represented in pregnant patients with ARF and may have intrinsically triggered a severe form of COVID-19 during pregnancy. Emergency caesarean delivery in patients needing invasive mechanical ventilation was independently associated with a modest but significant improvement in arterial oxygenation and may be viewed as a life-saving procedure in severely hypoxaemic patients.
Competing interests
All authors declare that they do not have any potential conflict of interest in relation to this manuscript.
Acknowledgements
We thank Loredana Baboi for her help in data acquisition. This work was funded by Hospices Civils de Lyon.
Supplementary Information
References
- 1.Zambrano L.D., 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.Di Toro F., Gjoka M., Di Lorenzo G., et al. Impact of COVID-19 on maternal and neonatal outcomes: a systematic review and meta-analysis. Clin Microbiol Infect. 2021;27:36–46. doi: 10.1016/j.cmi.2020.10.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Villar J., Ariff S., Gunier R.B., et al. Maternal and neonatal morbidity and mortality among pregnant women with and without COVID-19 infection: the INTERCOVID multinational cohort study. JAMA Pediatr. 2021;175:817–826. doi: 10.1001/jamapediatrics.2021.1050. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Knight M., Bunch K., Vousden N., et al. Characteristics and outcomes of pregnant women admitted to hospital with confirmed SARS-CoV-2 infection in UK: national population based cohort study. BMJ. 2020;369 doi: 10.1136/bmj.m2107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Vousden N., Bunch K., Morris E., et al. The incidence, characteristics and outcomes of pregnant women hospitalized with symptomatic and asymptomatic SARS-CoV-2 infection in the UK from March to September 2020: A national cohort study using the UK Obstetric Surveillance System (UKOSS) PLoS One. 2021;16(e0251):123. doi: 10.1371/journal.pone.0251123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Pierce-Williams R.A.M., Burd J., Felder L., et al. Clinical course of severe and critical coronavirus disease 2019 in hospitalized pregnancies: a United States cohort study. Am J Obstet Gynecol MFM. 2020;2 doi: 10.1016/j.ajogmf.2020.100134. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Delahoy M.J., Whitaker M., O’Halloran A., et al. Characteristics and maternal and birth outcomes of hospitalized pregnant women with laboratory-confirmed COVID-19 — COVID-NET, 13 States, March 1 - August 22, 2020. MMWR Morb Mortal Wkly Rep. 2020;69:1347–1354. doi: 10.15585/mmwr.mm6938e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Blitz M.J., Grünebaum A., Tekbali A., et al. Intensive care unit admissions for pregnant and nonpregnant women with coronavirus disease 2019. Am J Obstet Gynecol. 2020;223:290–291. doi: 10.1016/j.ajog.2020.05.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Collin J., Byström E., Carnahan A., Ahrne M. Public Health Agency of Sweden’s brief report: pregnant and postpartum women with severe acute respiratory syndrome coronavirus 2 infection in intensive care in Sweden. Acta Obstet Gynecol Scand. 2020;99:819–822. doi: 10.1111/aogs.13901. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Easter S.R., Gupta S., Brenner S.K., Leaf D.E. Outcomes of critically ill pregnant women with COVID-19 in the United States. Am J Respir Crit Care Med. 2021;203(1):22–25. doi: 10.1164/rccm.202006-2182LE. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Pineles B.L., Stephens A., Narendran L.M., et al. The relationship between delivery and the PaO2/FiO2 ratio in COVID-19: a cohort study. BJOG. 2022;129:493–499. doi: 10.1111/1471-0528.16858. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Metz T.D., Clifton R.G., Hughes B.L., et al. Disease severity and perinatal outcomes of pregnant patients with coronavirus disease 2019 (COVID-19) Obstet Gynecol. 2021;137:571–580. doi: 10.1097/AOG.0000000000004339. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Chen D., Yang H., Cao Y., et al. Expert consensus for managing pregnant women and neonates born to mothers with suspected or confirmed novel coronavirus (COVID-19) infection. Int J Gynaecol Obstet. 2020;149:130–136. doi: 10.1002/ijgo.13146. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Rose C.H., Wyatt M.A., Narang K., et al. Timing of delivery with coronavirus disease 2019 pneumonia requiring intensive care unit admission. Am J Obstet Gynecol MFM. 2021;3:100373. doi: 10.1016/j.ajogmf.2021.100373. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.National Institutes of Health NIH; 2021. COVID-19 treatment guidelines. Special considerations in pregnancy. https://www.covid19treatmentguidelines.nih.gov/special-populations/pregnancy (viewed Nov 2021) [PubMed]
- 16.ACOG practice bulletin No. 211: critical care in pregnancyObstet Gynecol. 2019;133:e303–e319. doi: 10.1097/AOG.0000000000003241. [DOI] [PubMed] [Google Scholar]
- 17.Wastnedge E.A.N., Reynolds R.M., van Boeckel S.R., et al. Pregnancy and COVID-19. Physiol Rev. 2021;101:303–318. doi: 10.1152/physrev.00024.2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.von Elm E., Altman D.G., Egger M., et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet. 2007;370:1453–1457. doi: 10.1016/S0140-6736(07)61602-X. [DOI] [PubMed] [Google Scholar]
- 19.Edlow A.G., Li J.Z., Collier A.R.Y., et al. Assessment of maternal and neonatal SARS-CoV-2 viral load, transplacental antibody transfer, and placental pathology in pregnancies during the COVID-19 pandemic. JAMA Netw Open. 2020;3:e2030455. doi: 10.1001/jamanetworkopen.2020.30455. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Bastard P., Gervais A., Le Voyer T., et al. Autoantibodies neutralizing type I IFNs are present in ~4% of uninfected individuals over 70 years old and account for ~20% of COVID-19 deaths. Sci Immunol. 2021;6:eabl4340. doi: 10.1126/sciimmunol.abl4340. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Zhang Q., Bastard P., COVID Human Genetic Effort, et al. Human genetic and immunological determinants of critical COVID-19 pneumonia. Nature. 2022;603:587–598. doi: 10.1038/s41586-022-04447-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Seasely A.R., Blanchard C.T., Arora N., et al. Maternal and perinatal outcomes associated with the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) Delta (B.1.617.2) variant. Obstet Gynecol. 2021;138:842–844. doi: 10.1097/AOG.0000000000004607. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Ong S.W.X., Chiew C.J., Ang L.W., et al. Clinical and virological features of SARS-CoV-2 variants of concern: a retrospective cohort study comparing B.1.1.7 (Alpha), B.1.315 (Beta), and B.1.617.2 (Delta) Clin Infect Dis. 2021 doi: 10.1093/cid/ciab721. [Epub ahead of print] [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Lapinsky S.E., Rojas-Suarez J.A., Crozier T.M., et al. Mechanical ventilation in critically-ill pregnant women: a case series. Int J Obstet Anesth. 2015;24:323–328. doi: 10.1016/j.ijoa.2015.06.009. [DOI] [PubMed] [Google Scholar]
- 25.Tomlinson M.W., Caruthers T.J., Whitty J.E., et al. Does delivery improve maternal condition in the respiratory-compromised gravida? Obstet Gynecol. 1998;91:108–111. doi: 10.1016/s0029-7844(97)00585-1. [DOI] [PubMed] [Google Scholar]
- 26.Bitker L., Dhelft F., Chauvelot L., et al. Protracted viral shedding and viral load are associated with ICU mortality in COVID-19 patients with acute respiratory failure. Ann Intensive Care. 2020;10:167. doi: 10.1186/s13613-020-00783-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Magleby R., Westblade L.F., Trzebucki A., et al. Impact of severe acute respiratory syndrome coronavirus 2 viral load on risk of intubation and mortality among hospitalized patients with coronavirus disease 2019. Clin Infect Dis. 2021;73:e4197–e4205. doi: 10.1093/cid/ciaa851. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.van Genderen M.E., van Uitert E., Duvekot J.J., Gommers D., COVPREG Study Group Management and outcome of critically ill pregnant women with COVID-19. Intensive Care Med. 2022;48:613–615. doi: 10.1007/s00134-022-06653-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials