Maternal death secondary to COVID-19 infection: A case report and review of the literature

Abstract

Maternal death secondary to coronavirus disease 19 (COVID-19) infection in a previously well woman is described. The woman presented with an eight-day history of productive cough and shortness of breath. Rapid deterioration of respiratory function was seen following admission, with associated tachycardia, tachypnoea and hypoxia. Emergency caesarean section was performed followed by transfer to the intensive care unit. COVID-19 PCR throat swab from day 0 was positive. Later, she developed hypoxaemia refractory to mechanical ventilation, proning and paralysis. The woman was transferred for veno-venous Extra Corporeal Membrane Oxygenation therapy but unfortunately died despite rigorous management. We review the conflicting information regarding physiological and immunological changes occurring during pregnancy and how these may affect susceptibility to respiratory viral disease. An overview of the current literature concerning ventilation and intensive care support in pregnant women suffering from COVID-19 is given. Further documentation of such cases is called for to progress understanding and management strategies.

Case report

A 28-year-old woman of Pakistani origin booked at 15 + 5 weeks gestation in her third pregnancy. She had no past medical history, had never smoked and had two previous vaginal deliveries. Her body mass index was 24 kg/m².

Her antenatal course was uncomplicated, and three appointments were not attended but were successfully re-arranged. One routine carbon monoxide test showed an isolated raised recording of 6 ppm at 32 weeks and 5 days of gestation; the woman was a non-smoker but reported that smoking did occur within the household. Glucose tolerance test was normal. Growth scans demonstrated linear growth.

At 36 weeks and 4 days of gestation (April 2020), the woman attended the accident and emergency department via ambulance with an eight-day history of cough productive of brown sputum, chest tightness and shortness of breath on exertion. There was no history of fever. There was a mild tachycardia (104 beats/min) and tachypnoea (27 breaths/min), and saturations were 97% on air. She soon developed a significant tachypnoea of 36 breaths/min. Differential diagnoses included coronavirus disease 19 (COVID-19) infection, pulmonary embolism or bacterial chest infection.

Chest radiograph revealed bilateral pneumonia (Figure 1). Admission blood tests revealed a lymphopaenia (0.7 × 10⁹/L), thrombocytopaenia (138 × 10⁹/L), raised CRP (145 mg/L), raised ferritin (317 mcg/L), raised procalcitonin (2.41 ng/mL) and also raised ALT (40 IU/L) and LDH (377 IU/L). The remaining full blood count, renal markers and electrolytes were unremarkable.

Figure 1.

Admission chest X-ray demonstrating bilateral pneumonia.

Figure 2.

Timeline of events from admission to maternal death.

Electronic fetal monitoring in the emergency department was normal. A single dose of intramuscular dexamethasone was given for fetal lung protection in case of emergency caesarean delivery.

In the 5 h following admission, a significant deterioration in physiology began; after 8 h, the woman was severely unwell. Respiratory rate rose to 46 breaths/min, heart rate to 120 bpm and supplemental oxygen (28% FiO₂) was required to maintain acceptable saturations. There was evidence of peripheral vasoconstriction with drowsiness. The woman was reviewed by both the obstetric and intensive care teams, and a decision was made to proceed to an emergency caesarean section with postoperative intensive care support. Caesarean was indicated to improve maternal condition as opposed to concerns with fetal wellbeing. This was also discussed with the woman who agreed and consented.

Category 2 caesarean section was performed under general anaesthetic 11 h after admission. The baby was delivered in good condition with Apgar scores of 8 at 1 and 9 at 5 min. There was no delayed cord clamping, and no skin-to-skin contact was adopted. Postoperatively, the woman was transferred immediately to the intensive care unit from theatre.

Following transfer to the intensive care unit, the woman remained intubated due to the low PaO₂/FiO₂ ratio. COVID-19 PCR throat swab from day 0 result was positive. On day 1, a unilateral right-sided pneumothorax was noted, and an intercostal drain was placed; despite this, the pneumothorax persisted, and a second ipsilateral drain was inserted. Subsequent interval chest X-rays over the next three days showed increase in size of the pneumothorax along with suspicion of pneumomediastinum.

A full body CT scan showed a large volume pneumomediastinum, bilateral pneumothoraces, extensive-dependent consolidation and bilateral ground glass opacities. There was also a large volume of gas within the abdominal wall, along with a small abdominal wound haematoma. After discussion with the local thoracic referral centre, a decision was made to prone the woman, despite the risk of haemodynamic instability due to the pneumomediastinum. Proning was uneventful, and the PaO₂/FiO₂ ratio improved from 10 to 20 kPa. A dialogue was started with the local extracorporeal membrane oxygenation (ECMO) centre in order to ensure timely intervention if proning was latterly ineffective. Following the initial improvement, worsening respiratory acidosis occurred over the next 12 h. A formal referral for ECMO was made, and the woman was accepted and transferred to the local ECMO centre within the next 12 h.

The woman was initially transferred on conventional ventilation to the local ECMO centre however required ECMO support after three days. The situation remained static over the following four days with the woman completely dependent on ECMO. Subsequent positive microbiology results included Candida from sputum and Klebsiella pneumonia from blood cultures for which appropriate antimicrobials and line changes were undertaken. The woman was also noted to be thrombocytopaenic. The previously imaged abdominal wall haematoma was evacuated, and the woman required additional blood products due to postoperative bleeding. On the 12th day following transfer, a significant left-sided pneumothorax occurred, and a left-sided intercostal drain was placed. Unfortunately, a haemothorax developed following drain placement which necessitated a resuscitative thoracostomy. Immediately following the evacuation of the haemothorax, a loss of cardiac output occurred, the woman was found to be in ventricular fibrillation, and this was altered to a ventricular escape rhythm following pericardial direct current cardioversion. No specific single bleeding point was identifiable; the cause of bleeding was attributed to profound coagulopathy. Coagulopathy was thought due to bone marrow failure in critical illness and therapeutic anticoagulation for ECMO.

Haemostasis was achieved with packing of the chest cavity. Within the next 24 h, the woman suffered a cardiac arrest which was refractory to medical intervention and she sadly died after cessation of resuscitative efforts. The case was discussed with the medical examiner, and the agreed cause of death was recorded as: 1a COVID-19 pneumonitis, 2 Pregnancy.

Discussion

Introduction

COVID-19 has prompted every healthcare speciality to review the emerging evidence to allow optimal outcomes for infected women. The case described is one of only four pregnancy-related cases in the UK to have progressed through rigorous critical care interventions on to ECMO to date. ¹ Below, we consider potential reasons for such significant deterioration, adding to the literature information about ECMO support and proning in the immediately post-natal period.

Maternal predisposition to severe viral respiratory disease in pregnancy

Owing to disproportionately increased severe morbidity and mortality demonstrated in pregnant women affected with severe acute respiratory syndrome (SARS), Middle East respiratory syndrome and influenza, concerns regarding the risk of COVID-19 in pregnancy were significant.

Reported data in general however have not demonstrated similarly increased risk of mortality to pregnant women. Pregnancy induces physiological changes to allow placental perfusion and accommodation of a growing fetus. As the uterus expands throughout pregnancy, the diaphragm is shifted upwards, and the shape of the thorax altered. ² The expiratory reserve volume and residual volumes fall (functional residual capacity), however, an increase in inspiratory capacity through the expansion of the thorax circumferentially allows total lung capacity to remain almost unchanged.^2–4

The closing capacity remains relatively unchanged also and so may encroach on the functional residual capacity causing small airway closure during tidal volume breaths, especially when supine.^3,4

More frequent occurrence of small airway collapse risks quicker progression from focal to widespread pneumonia and more rapid decompensation to type 1 respiratory failure. ⁴

Furthermore, progesterone is known to promote oedema of mucosal surfaces, increasing alveolar and small airway congestion secondary to infection. ⁴ Difficulty in clearing such respiratory secretions could lead to worsening disease with the risk of superadded bacterial infection.

The immune system adapts, as opposed to supresses to allow maternal tolerance of pregnancy. Literature is conflicting as to whether this may lend to more severe disease in pregnant women.

Modulation in part includes disruption of TNF alpha activity. TNF alpha has been thought to play a role in COVID-19-related cytokine storm syndrome which is thought to occur in select women with critical COVID-19-related disease.^5,6 Reduction in these levels would thus be thought to be protective to pregnant women. Conversely, Liu et al. ⁷ postulate that an increase in a number of additional pro-inflammatory cytokines are present in the first and third trimester of pregnancy, hence the cytokine storm brought about by COVID-19 infection may lead to more severe disease in such women.

Maternal death secondary to COVID-19

As case numbers increase, reports into the effects of COVID-19 on pregnant women are increasing.

Despite this, reports of maternal deaths remain minimal, with few maternal deaths documented in the literature to date. Evidence from China at the start of the outbreak concluded no increased risk of critical disease secondary to COVID-19. ⁸ Authors of reports documenting mortalities since this however warn against complacency. Although numbers may be smaller than expected, mortality remains a possibility and should be considered carefully in each case.

Strategies for ventilatory support

As the pandemic continues to unfold, the understanding of ‘optimal’ ventilation is evolving through the critical care community. The initial ventilation strategies adopted by many were based upon existing Acute Respiratory Distress practice; however, there is evidence that the physiological respiratory changes are far more heterogeneous in COVID-19 patients. ⁹ Early and repeated prone positioning is broadly considered to be beneficial by correcting a ventilation perfusion mismatch and may serve to additionally reduce diaphragmatic pressure in this group. ¹⁰ Peri-partum proning is possible with meticulous care taken to support the chest and pelvis and offload excessive abdominal pressure. This may be conducted in both awake and sedated COVID-19 patients and has shown some promise however the duration of proning may well be limited in a pregnant or post-partum woman owing to discomfort. ¹¹

Other common strategies such as higher positive end expiratory pressure (PEEP), strict or liberal tidal volumes and airway pressure release ventilation may be used as personalised ventilation methods. Factors that may help to decide between likely optimal strategies include lung compliance, recruitability, haemodynamics and CT changes. It is important to consider the underlying physiology of pregnancy when setting parameters for ventilation. For example, uterine blood flow may be limited by permissive hypercapnoea, and higher PEEP may cause a significant reduction in preload leading to haemodynamic instability.

Extracorporeal membrane oxygenation in the peripartum period

ECMO during the peripartum period outside of the COVID-19 pandemic has been described in small numbers. There appears to be a slightly lower rate of major complication and higher rate of survival in this patient group when compared to all women receiving ECMO. ¹² This is likely to be due to the relatively healthy and younger female population. In the UK, there are currently five ECMO referral centres managed through a national referral service that has been established during the pandemic.

The current inclusion criteria for ECMO in the NHS are focussed on selecting individuals who are most likely to benefit from support (Figure 3). ¹³ The majority of the most unwell COVID-19 patients have multiple co-morbidities and frailty, which may exclude them; however, given that pregnant women are usually younger with few co-morbidities, ECMO should be considered.

Figure 3.

Updated NHS England ECMO referral criteria. ¹³

Bacterial co-infection in COVID-19

Bacterial co-infection on presentation with COVID-19 appears to be unlikely, and so empirical antibiotics are not routinely recommended for mild cases.^14,15 Co-infection after prolonged mechanical ventilation, however, is common and has historically been associated with poor outcomes in other viral infections. The true extent of bacterial co-infection in mechanically ventilated COVID-19 patients is difficult to quantify due the widespread use of empirical antibiotics in intensive care reducing culture sensitivity. The causative organism in bacterial co-infection varies with geographical location and appears to be in line with previously reported local pathogens. ¹⁶ Owing to the small numbers of positive samples in these studies it is difficult to be confident in identifying the most common pathogens associated with COVID-19.

Conclusion

Few cases of maternal death secondary to COVID-19 have been reported in the literature to date. Clinicians should recognise however that maternal death is a known complication of infection in these women and should not be overlooked in their management planning. We suggest early involvement of the multidisciplinary team and forward planning in the case of deterioration to optimise care. Secondary to small numbers, management strategies are adopted from intensive care of peri-partum women or non-pregnancy-related COVID-19-infected illness. Further reporting of cases of critical disease or maternal death is encouraged to widen knowledge in this area. Further understanding of pathophysiology may accompany such reporting.

ORCID iDs

Annabel Stout https://orcid.org/0000-0003-4164-8308

Robert Crichton https://orcid.org/0000-0002-8973-4750