Secondary Logo

Journal Logo

Maternal Salvage With Extracorporeal Life Support: Lessons Learned in a Single Center

Biderman, Philippe MD*; Carmi, Uri MD*; Setton, Eric MD; Fainblut, Michael MD*; Bachar, Oren CCP*; Einav, Sharon MD, MCE‡§

doi: 10.1213/ANE.0000000000002262
Obstetric Anesthesiology

The American Heart Association scientific statement on cardiac arrest in pregnancy did not endorse extracorporeal life support for lack of cohort data. We studied all pregnancy and peripartum cases of extracorporeal life support in 1 medical center (n = 11), including collapse due to infection (n = 6, 55%), thromboembolism (n = 3, 27%), and cardiac disease (n = 2, 18%). Half of the cases (n = 5, 45%) involved extracorporeal cardiopulmonary resuscitation. Most mothers survived (n = 7, 64% [95% confidence interval, 32%–88%]). Deaths were attributable to oxygenator blockage (n = 1) and late sepsis (n = 3). The 2 unique clinical challenges were maintenance of high peripartum cardiac outputs and balancing anticoagulation with hemostasis.

Published ahead of print July 14, 2017.

From the *Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel; Ministry of Health, Jerusalem, Israel; Shaare Zedek Medical Center, Jerusalem, Israel; and §Hebrew University-Hadassah Faculty of Medicine, Ein-Kerem, Jerusalem, Israel.

Accepted for publication April 12, 2017.

Published ahead of print July 14, 2017.

The authors declare no conflicts of interest.

Funding: None.

Work performed at Cardiothoracic Intensive Care, Beilinson Medical Center, Campus Rabin, Petah Tikva, Israel.

Reprints will not be available from the authors.

Address correspondence to Philippe Biderman, MD, Cardiothoracic Intensive Care, Beilinson Medical Center, Campus Rabin, Jabotinsky St 39, Petah Tikva, Israel. Address e-mail to

Maternal critical illness is uncommon, yet approximately 40% of maternal deaths occurring in modern medical settings are both potentially preventable1 and highly reversible, because approximately 50% of parturients are discharged alive from the hospital even after cardiac arrest and resuscitation.2 Clinicians should therefore be aware of strategies that could avert maternal death and/or improve maternal survival in the event of cardiac arrest.

The coincidence of specific maternal susceptibility to the H1N1 influenza virus and the advent of use of extracorporeal life support (ECLS) as a bridge-to-recovery during the period of critical illness in severe cases of H1N1 infection provided a unique opening to establish the use of ECLS during pregnancy. Since then, ECLS has been used in cases of maternal cardiogenic shock, massive pulmonary embolism, and even peripartum hemorrhage.3

ECLS has been suggested by the Society of Obstetric Anesthesia and Perinatology as appropriate therapy for salvage of maternal collapse4 but was not endorsed by the American Heart Association in their scientific statement on cardiac arrest in pregnancy for lack of cohort data on maternal outcome.5 Since the guidelines were published, several authors have presented such cohorts.6,7 To these, we add the experience of an extracorporeal membrane oxygenation (ECMO) referral center in a developed country with a high birth rate and low maternal mortality rate (3.1 births/woman and 5/100,000 live births, respectively).8 We describe a consecutive cohort of 11 women placed on ECLS for salvage due to maternal death/near-demise during the pregnancy and peripartum periods with particular emphasis on clinical management of 2 major peripartum challenges—maintenance of high cardiac output (CO) and balancing anticoagulation with hemostasis.

Back to Top | Article Outline


After waiver of institutional review board approval (authorization 4315), data were extracted from the files of pregnant/peripartum women treated with ECLS in the Rabin Medical Center (RMC; August 2008 to October 2015). All those fulfilling criteria for “peripartum maternal death” or “maternal near miss” with imminent risk of demise were included.9 All treated cases fulfilled the indications for ECLS therapy of the Extracorporeal Life Support Organization—the international consortium that develops, evaluates, and improves ECMO use and sets ECMO guidelines. Venovenous (VV) ECMO was initiated to ameliorate gas exchange, whereas venoarterial (VA) ECMO was initiated in cases requiring both hemodynamic and respiratory support.10

Back to Top | Article Outline

Cannulation and Transfer

Cannulations were performed in the referring hospitals. Standard cannulation procedures were used despite pregnancy (Appendix A). Transfer was undertaken to RMC in a specially outfitted ambulance after initial stabilization.

Back to Top | Article Outline

Statistical Analysis

The study aim was to describe patient outcomes and the unique challenges of pregnancy/peripartum ECLS. The main outcome measure was maternal survival rate. Secondary outcomes included a description of the details of care. We used IBM SPSS Statistics for Windows (v21.0; IBM Corp, Armonk, NY) for descriptive statistics (ie, numbers, proportions, percentages, medians, interquartile ranges, and ranges) and Brixton Health WinPepi freeware for calculating 95% confidence intervals (CIs).

Back to Top | Article Outline


Among the 210 adults that underwent ECLS in RMC during the study period, 11 (6% [95% CI, 1%–6%]) fulfilled study criteria. The causes of maternal collapse were infection (n = 6, 55%), embolism (including amniotic fluid embolie [n = 3, 27%]), and cardiac problems (n = 2, 18%). Six women (55%) had prior comorbidities; in all these cases, there was a clinical association between these comorbidities and the pregnancy/peripartum event that led to their collapse (eg, mitral stenosis and cardiovascular collapse).



Three women were pregnant at the time ECLS treatment was initiated (their gestational age was 13, 33, and 20 weeks, respectively); 2 of these cases culminated in fetal/neonatal death. In all other cases, ECLS treatment was initiated postpartum. Five women were connected to ECMO during cardiopulmonary resuscitation (CPR). Seven women (64%) survived. Additional case details are available in the Table.

Back to Top | Article Outline

Management of ECLS

Our cohort spent a median 6 days on extracorporeal support (interquartile range, 4–16; range, 0–32). One woman with amniotic fluid embolism (verified by clinical event characteristics and an echocardiography finding of severe acute right ventricular failure) died <30 minutes after initiation of ECLS due to acute oxygenator blockage (Table).

Back to Top | Article Outline

Achievement of Target Oxygenation



Ideally, target arterial oxygenation in VV-ECMO should be 80%–85%, but 75%–80% is allowable (Pao2, 45–55).10 There is no clear Sao2 target in the literature or the guidelines for VA-ECMO; some contend that in fact the Svo2 should be the target while others use the same targets as in VV-ECMO.10 In our cohort, several tools were used to achieve postpump oxygen saturation of 85%, despite increased maternal COs generated by the combination of pregnancy and pathological disease processes. Oxygen delivery was improved by increasing median pump flows above the conventional cardiac index of 2.5 L/min/m2 in VA-ECMO and 0.6 cardiac index in VV-ECMO11; median flows were 5.40 L/min on VA-ECMO and 4.10 L/min on VV-ECMO (see Table). This method of improving oxygen delivery is limited because pump flow is physically limited by the size of the cannula and high flows have also been associated with VV blood recirculation,12 hemolysis, and thrombocytopenia.13 We therefore used 3 additional methods designed to lower heart rates and decrease oxygen consumption. These included induction of mild hypothermia (34.0°C–35.0°C) in 4 patients (patients 2, 3, 4, and 9) even in the presence of disseminated intravascular coagulation (DIC) (patients 2, 3, and 4) and pharmacological intervention with analgesics and/or sedation (patients 5, 6, and 7) and short-acting, intravenous β-blockers (ie, esmolol)14 (patients 5 and 6—both being pregnant at this time). The latter was used after verifying normovolemia with echocardiography. The β-blocker dose was titrated to optimize hemodynamics (ie, arterial blood pressure and CO) and surrogate metabolic (eg, Scvo2, lactate, organ function) end points as per Extracorporeal Life Support Organization guidelines.9 Echocardiography enabled continuous matching of pump flow to patient CO while contributing to diagnosis and treatment (eg, myocardial dysfunction, hypovolemia, valvular disease). Because CO measurements using flow techniques are often inaccurate during VV-ECMO (indicators may be recirculated), flow techniques were used to monitor CO only if CO measurements consistently correlated with those observed with echocardiography. The Figure shows the CO plotted against the postpump oxygen saturation during induction of β-blockade in 1 patient; as heart rate decreased, saturation consequently increased.

Back to Top | Article Outline

Invasive Procedures and Anticoagulation

Anticoagulation either immediately after surgery (cesarean delivery and/or hysterectomy) or in the presence of DIC (eg, after amniotic fluid embolism/hemorrhage) is an additional challenge encountered during peripartum ECMO management with no consensus of anticoagulation strategy in the literature.6 To strike the delicate balance between the risk of postoperative bleeding in patients with DIC with the risk of clotting the ECMO circuit/thrombosis, several strategies were used. Surgical hemostasis was performed as meticulously as possible despite the urgency of the situation since bleeding may begin to manifest only after restoration of blood pressure. Large-bore cannulas, required for ECLS, were inserted under ultrasound guidance by an experienced operator. In the presence of DIC (patients 1–4), a 2500 IU bolus of heparin was administered intravenously during cannulation. Then heparin was withheld until bleeding ceased and standard coagulation tests (drawn twice daily) normalized. High pump flows and heparin-coated tubing were used to modify the risk of ECMO obstruction. In cases without DIC, a 5000 IU bolus of heparin was administered intravenously during cannulation, immediately followed by a heparin drip titrated according to patient condition and circumstances (eg, during weaning of VA-ECMO with low blood flow, the activated clotting time goal was 220–240 seconds versus the standard 180–220 seconds). No maternal death was attributable to hemorrhage. Epidural catheters were left in situ until hemostasis was either restored or correctible without risk of oxygenator/line obstruction.

Back to Top | Article Outline

Maternal and Neonatal Outcomes

Two maternal deaths occurred during ECLS (one shortly after initiation) and 2 more occurred after withdrawal of ECLS (overall 36% mortality [95% CI, 12%–68%]). Two of the women who died underwent ECLS for respiratory failure and 2 for circulatory failure. The 3 late deaths occurred due to infections.

Three women were treated with ECLS during pregnancy but only 2 carried their pregnancy throughout ECLS therapy. In all other cases, the fetus was either delivered shortly before or during initiation of extracorporeal support. There were 3 fetal deaths (all in nonviable pregnancies) but no neonatal deaths. In 1 case, the fetus was aborted semielectively at week 12 of pregnancy due to severe maternal mitral stenosis, but several days later the woman developed fatal heart failure requiring ECLS. In the second case, ECLS was initiated on week 13 of pregnancy during ongoing CPR secondary to H1N1 infection. This fetus was spontaneously aborted within a day of ECLS. In the third case, ECLS was initiated at week 20 of pregnancy for severe Haemophilus influenzae pneumonia with secondary acute respiratory distress syndrome. This fetus was carried throughout the 4 days of ECLS but died 3 days later.

Back to Top | Article Outline


In this limited case series of ECLS for salvage of maternal cardiac arrest and/or critical illness, maternal mortality was 36% (29% with circulatory failure and 50% with respiratory failure). These outcomes are not as good as those described in isolated case reports of maternal ECLS,3 but are reminiscent of the outcomes of the general population undergoing ECLS for respiratory failure and better than those reported for circulatory failure.10 The 2 unique clinical challenges encountered in this population were the need to maintain relatively high peripartum COs and the need to balance between anticoagulation (to prevent pump obstruction) and hemostasis (to prevent surgical/obstetric/epidural hemorrhage in the presence of DIC). The methods used to decrease oxygen consumption were induction of mild hypothermia even in the presence of DIC and pharmacological intervention with analgesics and/or sedation and short-acting, intravenous β-blockers. Treatment with heparin was adapted to patient condition and circumstances with no undue complications. Although this report reflects the experience of only 1 medical center and may therefore not be generalizable, this medical center is an ECMO referral center in a country that has a low maternal mortality rate despite its high birth rate. Our results suggest that ECLS can be successfully tailored to the unique needs of the pregnant/parturient population and thus should play a role in salvage of maternal collapse.

Back to Top | Article Outline


Cannulation Procedures and Patient Transport to the ECMO Center

The ECMO center sends their trained staff to the referring hospital. This staff includes physician specialists (ie, intensivist/anesthesiologist and cardiothoracic/vascular surgeon) and an ECMO technician. This team cannulates the patient on location.

Unless cannulation is performed during CPR, patients are sedated with low-dose remifentanil and propofol. VA-ECMO is used in circulatory shock refractory to maximal conventional treatment and in refractory cardiac arrest. Blood is aspirated from a vein (eg, femoral) using a centrifugal pump (Rotaflow; Maquet, Cardiopulmonary GmBH, Rastatt, Germany), directed through a membrane oxygenator (Quadrox; Maquet, Cardiopulmonary GmBH) and reinjected through an artery (also usually femoral) into the aorta allowing the ECMO to function as cardiopulmonary bypass (ie, both hemodynamic and respiratory support). An additional perfusion cannula is inserted distally into the femoral artery to ensure ongoing blood supply to the cannulated inferior limb. Femoral-to-internal jugular or femoral-to-femoral VV-ECMO is used in cases of severe respiratory failure when optimized mechanical ventilation has failed to provide sufficient oxygenation (the ratio of arterial oxygen partial pressure to fractional inspired oxygen [Pao2/Fio2 ratio] <80) and/or for CO2 removal in cases with severe retention despite plateau ventilation pressures >30 cm H2O.

After cannulation, the patient is stabilized and then transferred (while connected to the ECMO) to an ambulance, which has been configured by the Israeli National Emergency Medical Services (MDA) for ECMO transport.

Back to Top | Article Outline


Name: Philippe Biderman, MD.

Contribution: This author helped acquire data, conceive and design the study, draft the manuscript, and revise it critically for important intellectual content.

Name: Uri Carmi, MD.

Contribution: This author helped assist Philippe Biderman to acquire data.

Name: Eric Setton, MD.

Contribution: This author helped assist Philippe Biderman to acquire data.

Name: Michael Fainblut, MD.

Contribution: This author helped assist Philippe Biderman to acquire data.

Name: Oren Bachar, CCP.

Contribution: This author helped assist Philippe Biderman to acquire data.

Name: Sharon Einav, MD, MCE.

Contribution: This author helped conceive and design the study, draft the manuscript, and revise it critically for important intellectual content.

This manuscript was handled by: Jill M. Mhyre, MD.

Back to Top | Article Outline


1. Berg CJ, Harper MA, Atkinson SM, et al. Preventability of pregnancy-related deaths: results of a state-wide review. Obstet Gynecol. 2005;106:1228–1234.
2. Mhyre JM, Tsen LC, Einav S, Kuklina EV, Leffert LR, Bateman BT. Cardiac arrest during hospitalization for delivery in the United States, 1998–2011. Anesthesiology. 2014;120:810–818.
3. Moore SA, Dietl CA, Coleman DM. Extracorporeal life support during pregnancy. J Thorac Cardiovasc Surg. 2016;151:1154–1160.
4. Lipman S, Cohen S, Einav S, et al.; Society for Obstetric Anesthesia and Perinatology. The Society for Obstetric Anesthesia and Perinatology consensus statement on the management of cardiac arrest in pregnancy. Anesth Analg. 2014;118:1003–1016.
5. Jeejeebhoy FM, Zelop CM, Lipman S, et al.; American Heart Association Emergency Cardiovascular Care Committee, Council on Cardiopulmonary, Critical Care, Perioperative and Resuscitation, Council on Cardiovascular Diseases in the Young, and Council on Clinical Cardiology. Cardiac arrest in pregnancy: a scientific statement from the American Heart Association. Circulation. 2015;132:1747–1773.
6. Sharma NS, Wille KM, Bellot SC, Diaz-Guzman E. Modern use of extracorporeal life support in pregnancy and postpartum. ASAIO J. 2015;61:110–114.
7. Agerstrand C, Abrams D, Biscotti M, et al. Extracorporeal membrane oxygenation for cardiopulmonary failure during pregnancy and postpartum. Ann Thorac Surg. 2016;102:774–779.
8. The World Bank. Available at: Accessed 2015.
9. Pattinson R, Say L, Souza JP, Broek Nv, Rooney C; WHO Working Group on Maternal Mortality and Morbidity ClassificationsWHO maternal death and near-miss classifications. Bull World Health Organ. 2009;87:734.
10. Extracorporeal Life Support Organization. Available at: Accessed 2015.
11. Lücker A, Weber B, Jenny P. A dynamic model of oxygen transport from capillaries to tissue with moving red blood cells. Am J Physiol Heart Circ Physiol. 2015;308:H206–H216.
12. Locker GJ, Losert H, Schellongowski P, et al. Bedside exclusion of clinically significant recirculation volume during venovenous ECMO using conventional blood gas analyses. J Clin Anesth. 2003;15:441–445.
13. Lou S, MacLaren G, Best D, Delzoppo C, Butt W. Hemolysis in pediatric patients receiving centrifugal-pump extracorporeal membrane oxygenation: prevalence, risk factors, and outcomes. Crit Care Med. 2014;42:1213–1220.
14. Guarracino F, Zangrillo A, Ruggeri L, et al. β-Blockers to optimize peripheral oxygenation during extracorporeal membrane oxygenation: a case series. J Cardiothorac Vasc Anesth. 2012;26:58–63.
Copyright © 2017 International Anesthesia Research Society