Coronavirus disease 2019 (COVID-19), caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), was first reported in New York City (NYC) on 1 March 2020 [1]. NYC was one of the earliest and most profoundly impacted cities in the United States (US) by the COVID-19 pandemic, which continues to be a global crisis. As of 1 December 2020, NYC has recorded more than 290,000 confirmed cases and 19,000 deaths [2].
The Bronx is a borough of NYC and has been the most devastated by COVID-19, with case rates considerably higher than its sister boroughs of Manhattan, Brooklyn, Queens, and Staten Island [3]. The borough is also home to the poorest congressional district in the US [4], and has a higher proportion of racial and ethnic minorities compared to the other NYC boroughs [3]. Socioeconomic disparities have played a major role in increased COVID-related mortality witnessed in the Bronx [5]. Investigating the outcomes of COVID-19 in this vulnerable population at the “epicentre of the epicentre” of disease in the US is critical for our understanding of the complete impact of the COVID-19 pandemic.
Given the historically adverse outcomes associated with the previous severe acute respiratory syndrome coronavirus (SARS-CoV) and Middle East respiratory syndrome coronavirus (MERS-CoV) epidemics during pregnancy [6], COVID-19 infection in obstetric patients is of particular concern. High rates of complications including preterm delivery, intrauterine growth restriction (IUGR), and maternal death have been reported, as well as an increased case fatality rate compared to non-pregnant patients [6]. Despite increasing literature regarding the epidemiology, presentation, and management of COVID-19 disease across most subspecialties, there are only limited reports discussing the anaesthetic implications and subsequent management of obstetric patients infected with COVID-19 [7–12].
The purpose of this study was to explain the anaesthetic management and maternal-fetal outcomes in a cohort of pregnant patients carrying the diagnosis of SARS-CoV-2 who were admitted during the early months of the COVID-19 pandemic to labour and delivery (L&D) units within one of the busiest hospital systems in NYC.
METHODS
Study design and participants
A historical review of medical records was conducted on pregnant patients with concomitant SARS-CoV-2 infection admitted to the L&D units at Montefiore Medical Center’s Jack D. Weiler Hospital and Wakefield Hospital in the Bronx, New York, USA from 1 February 2020 to 1 May 2020. This study was approved by the Albert Einstein College of Medicine Montefiore Medical Center Institutional Review Board (069109). The requirement for written informed consent was waived by the Institutional Review Board. Diagnosis of COVID-19 infection was defined as an on-site laboratory-confirmed positive SARS-CoV-2 result on reverse transcription polymerase chain reaction (RT-PCR) of a nasopharyngeal sample obtained during the hospital stay. All obstetric patients admitted to the L&D unit who tested positive for SARS-CoV-2 were included in this study. Changes to hospital policy were made during the investigation period on 2 April 2020, prompting universal screening of patients admitted for delivery upon entry to the emergency department (ED) or L&D triage. Prior to the implementation of universal screening, patients received an RT-PCR test based on exhibited symptoms and/or risk factors for COVID-19 disease. All newborns received 2 RT-PCR tests for SARS-CoV-2, using nasopharyngeal samples obtained 24 hours and 72 hours after delivery.
Initial guidance regarding the care of obstetric patients testing positive for SARS-CoV-2, including contact precautions, staff training, heightened emphasis of early placement of neuraxial analgesia for labour, and overall considerations for general anaesthesia (GA) were influenced by recommendations set by the Society for Obstetric Anesthesia and Perinatology, first published on 15 March 2020 [13].
Data collection
We conducted a manual review of medical records for all COVID-affected patients (n = 92) for information on demographic data, presenting symptoms and vital signs, anaesthetic management, inpatient medications, anticoagulation protocol, clinical outcomes, and imaging studies. Clinical outcomes (e.g. acute kidney injury [AKI], acute respiratory distress syndrome [ARDS]) were reported as defined by the treating physician. We used Clinical Looking Glass proprietary hospital software to collect laboratory data. Patients were stratified into groups based on mode of delivery and COVID-19 disease severity. Modes of delivery included vaginal delivery, scheduled caesarean delivery, and unscheduled caesarean delivery. Disease severity (asymptomatic, mild, moderate, severe, critical) was determined by clinical and laboratory parameters, as defined by criteria set by the World Health Organization (Appendix 1) [14].
Statistical analysis
No sample size was calculated given that all individuals with a diagnosis of COVID-19 during the aforementioned timeframe were included in this study. All patient data were de-identified and compiled in a secured Microsoft Excel spreadsheet. Statistical analysis was performed with Microsoft Excel. Continuous variables were expressed as median (25th–75th percentile; range) and categorical variables were expressed as number/total number (%). ANOVA was used to compare continuous variables and χ2 tests were used to compare categorical variables. All tests were two-sided and P-values < 0.05 were considered to have statistical significance.
RESULTS
Patient characteristics
Between 1 February 2020 and 1 May 2020, 1694 obstetric patients were delivered at Jack D. Weiler Hospital and Wakefield Hospital. Among those, 92 patients (5%) tested positive for SARS-CoV-2 infection. Of the 92 patients, 49 (53%) had vaginal deliveries, 14 (15%) had scheduled caesarean deliveries, and 29 (32%) had unscheduled caesarean deliveries (Table 1). The majority of patients identified as Hispanic or Latino (45 patients [49%]) or Black/non-Hispanic (26 patients [28%]).
TABLE 1
Demographic data on COVID-positive obstetric patients stratified by mode of delivery
The most common comorbidities were hypertension (14 patients [15%]), diabetes (11 patients [12%]), and asthma (10 patients [11%]). Additional comorbidities are shown in Table 1. There was no significant difference in frequency of comorbidities between cohorts stratified by mode of delivery.
Laboratory values were obtained from patients during their hospital stay (Table 2). Median values for albumin, creatinine, fibrinogen, haemoglobin, WBC count, and platelet count were all within normal physiologic limits for pregnant women [15], with no significant variance between groups stratified by disease severity. The highest WBC count (21.6 G L-1) was recorded in a patient with moderate COVID disease and bilateral interstitial pneumonia. The lowest platelet counts were recorded in 1 patient diagnosed with benign gestational thrombocytopenia (platelets: 55 G L-1) and 1 patient diagnosed with HELLP syndrome (platelets: 74 G L-1). Both patients with thrombocytopenia were asymptomatic for COVID-19 disease. Median D-dimer values were elevated across all levels of disease severity, with the highest elevations noted in patients with mild disease.
TABLE 2
Laboratory values obtained from COVID-positive obstetric patients stratified by COVID-19 disease severity
Presenting symptoms
Upon admission to the L&D unit, the most common symptoms were cough (23 patients [25%]), shortness of breath (11 patients [12%]), and fever (10 patients [11%]) (Table 3). Patients who underwent unscheduled caesarean delivery presented with these symptoms at a greater frequency than those who underwent scheduled caesarean delivery or vaginal delivery. 64 patients (70%) were asymp-tomatic upon initial presentation, with more asymptomatic patients in the vaginal and scheduled caesarean groups compared to the unscheduled caesarean group.
TABLE 3
Presenting COVID-19 symptoms upon admission and COVID-19 disease severity stratified by mode of delivery
Anaesthetic management
In our cohort, 83 patients (90%) received neura-xial labour analgesia or surgical anaesthesia prior to delivery. Combined spinal-epidural (CSE) and dural puncture epidural (DPE) were the most common techniques used across all modes of delivery. In the unscheduled caesarean group, 11 patients (38%) received neuraxial labour analgesia for trials of labour prior to caesarean delivery. All epidural catheters placed for neuraxial labour analgesia in this group were successfully converted for surgical anaesthesia without necessitating replacement, and the remainder of patients who did not receive neuraxial labour analgesia had de-novo epidural catheters placed in the operating room (OR) for CSE anaesthesia.
Five patients (5%) ultimately required caesarean delivery under GA, all of whom were in the unscheduled caesarean group (Table 4). Two of those patients were asymptomatic for COVID-19 disease, 2 had severe disease, and 1 had critical disease. Of the 2 asymptomatic patients, 1 patient initially received CSE for surgical anaesthesia and was subsequently converted to GA after needing emergent hysterectomy for placenta percreta in the setting of significant postpartum haemorrhage. The other asymptomatic patient underwent delivery under GA secondary to rapidly falling low platelet count associated with HELLP syndrome, which ultimately contraindicated neuraxial placement.
TABLE 4
Anaesthetic management of COVID-positive obstetric patients stratified by COVID-19 disease severity
Of the 2 patients with severe COVID-19 disease, both were admitted to the hospital due to respiratory distress and pre-emptively intubated for caesarean delivery under GA before further respiratory decompensation could occur. The 1 patient with critical COVID-19 disease was initially admitted for expectant management with intravenous insulin for euglycaemic diabetic ketoacidosis (DKA) and observation in the setting of mild COVID-19 symptoms. However, the patient’s respiratory status rapidly deteriorated, necessitating emergent intubation and subsequent caesarean delivery under GA. All 3 patients with severe or critical COVID-19 disease were intubated by the anaesthesiology team in the OR and underwent caesarean delivery immediately after induction. Following delivery, all 3 patients were transferred to the intensive care unit (ICU), where they remained on mechanical ventilation. The 2 patients with severe disease were extubated within the same day and the 1 patient with critical disease was extubated 2 days postpartum. All 3 patients were ultimately discharged in medically stable condition, with no subsequent hospital readmissions (Table 5).
TABLE 5
Maternal outcomes stratified by mode of delivery
Antepartum corticosteroids, intrapartum antibiotics, and anticoagulants were used in 10 (11%), 74 (80%), and 27 (29%) patients, respectively (Table 4). There was no significant difference in frequency of usage of these medications between groups when stratified by disease severity.
Maternal and fetal outcomes
The most common obstetric complication was preeclampsia (26 patients [28%]) (Table 5). Other common obstetric complications included gestational diabetes (17 patients [19%]) and postpartum haemorrhage (14 patients [15%]). Ten patients (11%) were diagnosed with pneumonia, most of whom underwent unscheduled caesarean delivery. No patient in our cohort suffered from stroke, deep vein thrombosis (DVT), pulmonary embolism (PE), or disseminated intravascular coagulation (DIC) during delivery hospitalization. After discharge, 2 patients (2%) were readmitted to the hospital, both for complications of preeclampsia. No maternal deaths occurred during delivery hospitalization and all patients were ultimately discharged home.
Of the 92 deliveries, 25 newborn infants (27%) required admission to the neonatal intensive care unit (NICU) (Table 6). Eight infants (9%) were diagnosed with intrauterine growth restriction (IUGR) and 4 (4%) had a 5-minute Apgar score less than 7. Two infants (2%) experienced respiratory distress after delivery. Infants born to patients in the unscheduled caesarean group were more likely to have been preterm deliveries, with 31% of unscheduled caesarean deliveries (9/29 deliveries) occurring prior to 34 weeks gestation compared to 10% of vaginal deliveries (5/49 deliveries) (Table 5). There were 0 (0%) documented instances of SARS-CoV-2 vertical transmission amongst our cohort, and 2 (2%) fetal deaths. The first fetal death was an intrauterine fetal demise at 29 weeks gestation secondary to chronic hypertension and superimposed preeclampsia complicated by HELLP syndrome. The second fetal death was a miscarriage at 19 weeks gestation. Both patients who experienced fetal deaths underwent inductions of labour with DPE for labour analgesia.
TABLE 6
Fetal outcomes
| Complication | No./total no. (%) |
|---|---|
| NICU admission | 25/92 (27) |
| IUGR | 8/92 (9) |
| 5-minute Apgar score < 7 | 4/92 (4) |
| Respiratory distress | 2/92 (2) |
| SARS-CoV-2 vertical transmission | 0/92 (0) |
| Fetal death | 2/92 (2) |
Eight patients in the unscheduled caesarean group had a length of stay (LOS) greater than 5 days, including the 3 patients who were intubated for caesarean delivery under GA for rapidly worsening COVID-19 disease. Details of their presentation, anaesthetic management, laboratory test results, and outcomes are shown in Appendix 2.
DISCUSSION
In this study, we report the anaesthetic management and short-term maternal-fetal outcomes in a cohort of 92 obstetric patients in the Bronx, New York, USA who tested positive for SARS-CoV-2.
Outcomes and complications
Caesarean section was the primary mode of delivery in the symptomatic COVID-19 patients within our cohort, with a greater caesarean delivery rate than the New York State (NYS) average [16]. A significant number of these caesarean deliveries were unscheduled secondary to complications from preeclampsia, COVID-related pneumonia, or both. Nearly every patient with pneumonia underwent unscheduled caesarean delivery, 3 of whom experienced hypoxic respiratory decompensation ultimately necessitating intubation and delivery under GA.
Preeclampsia was the most common complication, with rates in our cohort greater than rates in the general population on a nationwide, statewide, and citywide level [17-19]. Existing literature has documented a potential association between preeclampsia and coronavirus infection in pregnancy [6]. Although higher rates of preeclampsia are seen among patients in the Bronx [20], SARS-CoV-2 infection must be considered as a potential contributor to the rates of preeclampsia above baseline levels in our cohort.
Increased rates of preterm delivery were also observed in our cohort compared to general populations without COVID-19 in NYS and NYC [20, 21]. A majority of these preterm births occurred in the unscheduled caesarean group, which had a higher proportion of patients with moderate to critical disease. These findings support existing studies documenting an increase in preterm deliveries in patients with symptomatic COVID-19 disease [7, 8]. Likely attributable to the increased rate of preterm births, NICU admissions were also elevated compared to NICU admission rates in the general population [22].
Existing research has demonstrated possible evidence for vertical transmission of SARS-CoV-2 infection [23]. Within our cohort, SARS-CoV-2 was not detected in any newborn following delivery.
Asthma, hypertension, and diabetes were the most common comorbidities among patients in our cohort. Rates of these comorbidities were consistent with rates of asthma, hypertension, and diabetes in the general population of adults without COVID-19 aged 18-49 in NYC [24–26]. Although it is well established that these comorbidities are associated with increased mortality and decreased survival in COVID-19 disease [27], they were not correlated with an increased frequency of negative outcomes in our obstetric patient cohort.
Anaesthetic management
Nearly every patient in our obstetric cohort received neuraxial anaesthesia. Given the risks associated with SARS-CoV-2 aerosol transmission during invasive endotracheal procedures [28], GA was avoided in all but the most critically ill of patients. Since reliance on spinal anaesthesia confers the additional risk of conversion to GA if the block wears off intraoperatively, CSE and DPE (in which epidural catheters are maintained for the duration of anaesthetic delivery) were the preferred techniques in patients who underwent caesarean delivery. When compared to plain epidural analgesia/anaesthesia (which was rarely used in our cohort), CSE and DPE have lower rates of maternal and fetal side effects, decreased time to analgesia, and improved quality of block [29]. Catheter failure is also less common in CSE and DPE compared to epidural anaesthesia wherein CSF confirmation is not obtained [30]. Although CSE and DPE have similarly low failure rates, placement success of CSE catheters remains uncertain in the first 1–2 hours after the patient receives an immediate intrathecal provision of analgesia. Thus, we found added security in performing DPE – especially at a teaching institution – wherein failed catheters can be promptly identified, troubleshot, and replaced if necessary.
Anticoagulation protocol
Postpartum D-dimer levels were elevated in our cohort, raising concern for an increased risk of coagulopathy and thromboembolic events – already a heightened concern given the physiologic hypercoagulable state of pregnancy. Despite having no formalized anticoagulation protocol for COVID-positive patients during the timeframe studied (Montefiore protocols were ultimately implemented on 4 May 2020), none of our cohort patients developed thrombotic complications such as deep vein thrombosis (DVT), pulmonary embolus (PE), or stroke. Anticoagulation protocols for pregnant patients carrying a COVID-positive diagnosis are at the institutions’ discretion. However, many have used low-molecular weight heparin or other anticoagulant medications carrying the potential to preclude the safe provision of neuraxial analgesia in urgent situations. Multidisciplinary collaboration is pivotal whenever anticoagulation is introduced in the pregnant population, and even more so in the context of COVID-19, where concerns for appropriate anaesthetic management have considerable implications for both patient and healthcare worker safety. COVID-positive patients who have received prophylactic and/or therapeutic anticoagulation may ultimately require emergency provision of GA (with subsequent potential for increased SARS-CoV-2 exposure to healthcare workers via aerosoli-zation), or could be given neuraxial placement in close proximity to having received anticoagulation agents (portending increased risk of spinal and/or epidural hematoma). Given our findings of zero COVID-positive patients in our cohort developing thrombotic complications during their peripartum stay, we recommend further study examining the overall utility of anticoagulation protocols for the entirety of COVID-positive pregnant patients (i.e. the necessity of anticoagulating those with asymptomatic presentations), and suggest that future decisions to initiate anticoagulation in this context be made on a case-by-case basis.
Limitations
Our findings must be interpreted in the context of several limitations. The evolving nature of the COVID-19 pandemic and the disproportionate burden it placed on hospitals in NYC meant that testing protocols and availability of testing materials and personal protective equipment changed on a near-daily basis. Unsurprisingly, the implantation of universal testing protocols midway through the investigation period increased the detection of asymptomatic COVID patients within our obstetric population. Had these protocols been implemented from the beginning of the study, it is likely that more patients would have been included, and a greater proportion of these may likely have been asymptomatic carriers.
We did not compare our cohort with a control group, and the number of patients with severe or critical COVID disease was limited. Further analysis with a larger sample size and a control group will be necessary to account for confounding variables. Additionally, D-dimer and fibrinogen were not obtained among all patients. Given their elevated risk for thromboembolic events, obtaining D-dimer and fibrinogen for every obstetric patient who tested positive for SARS-CoV-2 would have been beneficial.
Finally, due to the demands of the pandemic, most patients and newborn infants did not receive in-person follow-up after their hospital stay, and greater follow-up will be necessary to ascertain the long-term sequelae of COVID-19 disease in pregnancy. Medical records confirmed the zero rate of hospital readmissions to Montefiore Medical Center, though we cannot confirm that any of these patients may have presented to another institution.
Conclusions
Given the number of patients discovered to be asymptomatic carriers of COVID-19, universal testing is critical for the containment of the disease and protection of patients and staff in settings with high disease prevalence. Pregnant patients and their families must be advised to take extreme caution in avoiding SARS-CoV-2 infection due to the increased risk for adverse outcomes including preeclampsia, preterm labour, unscheduled caesarean delivery, and invasive mechanical ventilation. Regular checkups of pregnant women are warranted during the prenatal period and in L&D to identify those most vulnerable to contracting COVID-19 and suffering disease complications. Considering the number of obstetric patients who will likely be infected with SARS-CoV-2 in the future, reporting the details of their anaesthetic management and outcomes carries significant importance for the provision of effective and evidence-based care.
