Pregnant and postpartum women with SARS-CoV-2 infection in intensive care in Sweden

The Public Health Agency of Sweden has analysed how many pregnant and postpartum women with SARS-CoV-2 infection have been treated in intensive care units (ICU) in Sweden between the March 19 and April 20, 2020, compared with non-pregnant women of similar age. Cases were identified in a special reporting module within the Swedish Intensive Care Registry (SIR). Fifty-three women aged 20-45 years with SARS-CoV-2 were reported in SIR, and thirteen (n=13) of these women were either pregnant or postpartum (<1 week). The results indicate that the risk of being admitted to ICU may be higher in pregnant and postpartum women with laboratory-confirmed SARS-CoV-2 in Sweden, compared to nonpregnant women of similar age.


Introduction
In the beginning of April 2020, the Public Health Agency of Sweden (PHAS) noted that a relatively high number of pregnant and postpartum women with severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) infection were or had been treated in intensive care units (ICU) in Sweden. When analysing the published literature on pregnancy and SARS-CoV-2, gaps in knowledge were identified, especially whether being pregnant represents a risk for increased susceptibility to infection, severity of clinical presentation and adverse outcomes for mothers and neonates. [1][2][3][4][5][6][7] Through dialogue with corresponding authorities in other European countries and the US, it became apparent that none had seen a comparatively increased number of pregnant or postpartum women with SARS-CoV-2 infection requiring intensive care.
As a first step, PHAS analyzed how many pregnant women with SARS-CoV-2 infection had been treated in ICU in Sweden, compared with non-pregnant women of similar age. This analysis was shared and discussed with the National Board of Health and Welfare and with professional medical organisations in Sweden. This is a rapid assessment of the current situation in March and April. Further analyses and research will hopefully shed more light on pregnancy and COVID-19.

Material and Methods
The Swedish Intensive Care Registry (SIR) 8 includes all cases that have received intensive care in Sweden.
Most ICUs also report additional information on patients with laboratory-confirmed SARS-CoV-2 as well as influenza through a special reporting module. Pregnant and postpartum women can be identified through this reporting. During the period between March 19 and April 20, information on all women aged 20-45 years with SARS-CoV-2 reported through this module was collected. Additional information on the pregnant and postpartum women was collected, such as severity of symptoms and risk factors, but because of confidentiality and the small numbers it is not possible to publish detailed, personal information at this point.
For some pregnant/postpartum patients, the main reason for intensive care was not symptoms of SARS-CoV-2 infection, but other conditions. It was not possible to ascertain whether or not SARS-CoV-2 was the primary reason for intensive care for the non-pregnant women. Therefore, a decision was made to include the entire age group, regardless of whether SARS-CoV-2 was the main reason for intensive care admission, as long as the patient had laboratory-confirmed SARS-CoV-2.

Accepted Article
This article is protected by copyright. All rights reserved Population data was obtained from the Swedish population registry. 9 In December 31, 2018, there were 1 671 740 women aged 20-45 years in Sweden. During 2018, the Swedish Birth Registry reported 1 16 079 births (=>27week of gestation). An assumption was made that births were equally distributed throughout the year, resulting in an average of 318 deliveries per day (24 h). The gestational age was assumed to be 40 weeks from the last menstrual period, on average. This resulted in an estimate that 84913 women were likely to be pregnant in Sweden on any given day. Thus, between March 19 and April 20, 2020, 95 089 women were estimated to have been pregnant at some point.
Three sensitivity analyses were also performed. Since the estimate described above only includes pregnancies from gestational age 27 weeks, there is a degree of under-ascertainment regarding the number of pregnancies, as miscarriages and early stillbirths are not included. To compensate for this, a 50 percent higher value for the number of pregnancies was used in the first sensitivity analysis, based on a miscarriage rate of 28% (ranging from 10% at 20 years of age to 40% above 35 years of age), 10 and to be sure to be well above this rate. This may be an unrealistically high number of pregnancies, but it was adopted to avoid an overestimation of risk while interpreting the results.
In the second sensitivity analysis, the number of women requiring ICU was reduced to contain only those who received invasive mechanical ventilation, in order to account for the possibility of a slightly lower threshold for admitting pregnant women to ICU as a precaution. The third sensitivity analysis combined the aspects of the first and second sensitivity analyses.

Ethical approval
This study was completed as part of PHAS responsibility for public health issues at a national level and its subsequent work on surveillance of COVID-19 during the pandemic and was exempt from formal ethical approval.

Accepted Article
This article is protected by copyright. All rights reserved In total, 53 women aged 20-45 years with SARS-CoV-2 admitted in ICU were reported during the period between March 19 and April 20. Thirteen of these women were pregnant (11 of 13) or had recently given birth (2 of 13) on admission (within one week postpartum). Their age varied between 20-35 years, and gestational age between week 13-40. Risk factors reported for some of the women were gestational diabetes and obesity. For seven women, outcome of the pregnancy is known and of these, five had delivered the baby by caesarean section (CS). The indication for CS is not known in detail for all, but for two, the indication reported was obstetric and for two, the indication reported was SARS-CoV-2 symptoms.
All of the pregnant or postpartum women required intensive care. In addition, 7 of the 13 women respectively.
An analysis similar to the first sensitivity analysis was done for the 2015-2016 influenza season included 180 903 pregnant women, and resulted in an incidence of intensive care with laboratory-confirmed influenza of 3.9 per 100 000, compared with 1.8 per 100 000 for non-pregnant women. The relative risks are presented in Table 1. Relative risk indicates the increased probability of receiving intensive care in conjunction with laboratory-confirmed SARS-CoV-2 for pregnant or postpartum women, compared with non-pregnant women in the same age group.

Discussion
We identified that the risk of requiring intensive care may be higher in pregnant/postpartum women with laboratory-confirmed SARS-CoV-2, compared to non-pregnant women in the same age group, even after

Accepted Article
This article is protected by copyright. All rights reserved accounting for miscarriages and early stillbirths (<27 weeks) in the denominator. This risk was higher than that calculated for 2015-2016 seasonal influenza epidemic. The increased risk remained when the analysis was restricted to only those women in need of mechanical ventilation.
Our analysis has obvious limitations. It is based on a small number of pregnant and postpartum women with SARS-CoV-2. For some of these, SARS-CoV-2 symptoms were not the main reason for ICU admission, even though all had laboratory-confirmed SARS-CoV-2 infection. As the same detailed information on the primary reason for intensive care was not available for non-pregnant women, the analysis was done without excluding those women in ICU who were not primarily admitted because of SARS-CoV-2. The baseline information to calculate risk are assumed and we do not have exact information. Furthermore, other confounding factors, such as pre-existing comorbidities and socio-economic factors, could not be analyzed in detail and analysis stratified by gestational age or different trimesters of pregnancy was not possible. Some of the pregnant women, but not all, exhibited risk factors like hypertension, overweight or obesity, and gestational diabetes. In addition, we had no details on reason for admission, and it is possible that pregnant and postpartum women are sometimes admitted for precautionary purposes. Still, the increased risk remained when only patients requiring mechanical ventilation were included. The actual number of women in our analysis is very small and could reflect heightened baseline incidence rate. Our findings need to be confirmed by other studies.
Although the limitations described above need to be taken into consideration, the results generated immediate recommendations from PHAS, including suggestions on possible preventive measures.
Information on pregnant women receiving intensive care in Sweden with COVID-19 will be continuously monitored and more refined analyses will be performed. Moreover, a joint research project has been initiated to elucidate the impact of COVID-19 during pregnancy on maternal and neonatal outcomes, using data from the Swedish Pregnancy Register, 12 the Swedish Neonatal Quality Register (SNQ) and SmiNet. 13

Conclusion
The risk of requiring intensive care may be higher in pregnant women with laboratory-confirmed SARS-CoV-2 in Sweden, compared to non-pregnant women of similar age. Pregnant women should be cautious considering the potential severe consequences of SARS-CoV-2 infection and those with additional risk factors such as overweight or obesity, hypertension and gestational diabetes should take extra precautions. This study needs to be replicated in other countries and more detailed information on symptoms, treatment and outcomes for pregnant and postpartum women managed in ICU is needed.

Accepted Article
This article is protected by copyright. All rights reserved