A 41-year-old G2P1 with a history of chronic hypertension, obesity (BMI 45), lower extremity edema, and asthma presents with severe-range blood pressures at 37 weeks gestational age. The diagnosis of preeclampsia with severe features is confirmed, an induction of labor is started, and an early epidural is placed due to prior cesarean/TOLAC. Four hours later, the patient is febrile to 101.4°F, tachypneic, oliguric despite fluids, and concern is raised for sepsis. Fetal heart monitoring is progressively non-reassuring, and the decision is made to pursue repeat cesarean delivery under epidural. Immediately following delivery, the patient develops hemodynamic instability and general anesthesia is induced. Point-of-care ultrasound (POCUS) identifies a hyperdynamic and underfilled heart with grossly normal biventricular function. Massive hemorrhage is identified and treated, and the patient experiences a full recovery.
Depending on your experience and practice, this case may induce emotions ranging from anxiety to terror. The rate of maternal morbidity and mortality is rising in the United States due to many factors, including increased maternal age and comorbidities (Obstet Gynecol 2017;130:366-73; NCHS Health E-Stats 2022). As illustrated by this case, the principles of critical care anesthesiology can be applied to better care for increasingly complex parturients in order to decrease the rate of morbidity and mortality.
Identifying patients on the labor floor who are at increased risk of developing complications or severe maternal morbidity (SMM) is an important step in reducing this risk (asamonitor.pub/3FzKZy8). One score that can be used on admission is the obstetric comorbidity index (OB-CMI) (Figure 1) (Am J Obstet Gynecol 2019;221:271.e1-10). OB-CMI scores greater than 7 are associated with a nearly 13% risk of SMM, and a score >9 is associated with a nearly 20% risk (note the case described has a score of 13) (Am J Obstet Gynecol 2019;221:271.e1-10). The OB-CMI score has also been associated with an increased risk of requiring general versus neuraxial anesthesia, particularly for emergent deliveries (Int J Obstet Anesth 2022;51:103546). Parturients with coexisting cardiac disease are at higher risk of new cardiac complications during pregnancy, but this risk varies widely. Several scoring systems, including CARPREG II and ZAHARA, can help to gauge this risk and guide delivery location for these patients (Am Coll Cardiol 2018;71:2419-30; Eur Heart J 2010;31:2124-32).
Despite the utility of these risk indices, many women who develop complications do not have preexisting illness. While traditional scoring systems such as SIRS or qSOFA consistently fail to identify the crashing parturient, other maternal early-warning systems (Figure 2) have been developed to distinguish the normal physiologic changes of pregnancy from the abnormal. A high score triggers providers to urgently evaluate patients to improve time to diagnosis and treatment of conditions such as bleeding, infection, and hypertension (Anesth Analg 2019;129:1613-20; Obstet Gynecol 2014;124:782-6; Anaesthesia 2013;68:354-67). These warning systems may also predict the risk of ICU admission and have been integrated into clinical pathway tools to protocolize treatment of the most common obstetric complications with resultant improvement in severe maternal morbidity (SMM) (J Obstet Gynaecol Can 2017;39:728-33.e3; Am J Obstet Gynecol 2016;214:527.e1-6).
Recognition of sepsis in pregnancy is often delayed due to the mistaken attribution of nonspecific signs and symptoms to the normal physiological changes of pregnancy, blood loss, or labor pain. For example, fever is an unreliable marker in the obstetric population, and while tachycardia or baseline leukocytosis raise suspicion in nonpregnant patients, these changes are common in normal pregnancy. What is clear is that tachypnea, neutropenia, hypothermia, or altered mental status is never normal in pregnancy and should always warrant a workup for sepsis (BJOG 2011;118:1402-3).
While sepsis in any population is deadly, sepsis in pregnancy portends the potential for rapid decline: 50% of parturients who die of sepsis will do so in 24 hours (Acta Obstet Gynecol Scand 2009;88:647-53). Thus, early identification and intervention is key. However early warning systems have failed to predict early sepsis in parturients (Int J Obstet Anesth 2018;36:96-107). To date, the best predictive scoring system in the obstetric population is the Sequential Organ Failure Assessment (SOFA) score (Intensive Care Med 2021;47:1181-1247). This score, along with a high index of suspicion, is needed to identify if early signs of end-organ dysfunction (i.e., respiratory dysfunction, coagulopathy, kidney injury, cardiovascular dysfunction, hepatic dysfunction, or altered mental status) are secondary to sepsis, baseline comorbidities, or other disorders of pregnancy (Intensive Care Med 2021;47:1181-1247; Anesth Analg 2013;117:944-50). Once these patients with organ dysfunction and suspected infection are identified, the key principles of sepsis management must be applied, including early collection of blood cultures and source control, initiation of I.V. antibiotics with broad-spectrum coverage, resuscitation with a balanced crystalloid solution, monitoring for fluid responsiveness, and targeting vasopressors to a mean arterial pressure (MAP) >65 mmHg. In parturients, these principles must be customized: source control should include evacuation of the uterus when intrauterine infection is suspected. Antibiotics must cover group A Streptococcus, Escherichia coli, staphylococcus, and Gram-negative bacteria, organisms commonly linked to sepsis in pregnancy. Care must also be taken to avoid volume overload and pulmonary edema, particularly in those patients with preexisting cardiac disease, current treatment with oxytocin, and the presence of preeclampsia to avoid worsening outcomes (Crit Care 2015;19:251). Cardiac and lung ultrasound can evaluate volume status and show early signs of volume overload (B-lines). Finally, target MAP goals should be personalized to the patient, as 65 mmHg may be too high in a pregnant patient without prior comorbidities or too low to perfuse a distressed fetus. Adequate perfusion can be monitored by trending the fetal heart tracing and maternal lactate clearance.
During cesarean delivery with hemodynamic collapse, one of the most important questions is distinguishing hemorrhage from amniotic fluid embolism (AFE) or other obstructive shock. In patients with coexisting cardiac disease, the differential can be additionally complicated by decompensation of preexisting cardiac conditions. POCUS can provide critical diagnostic information regarding cardiac filling, dynamic movement, and any structural problems (BJA Educ 2018;18:271-6).
Early recognition of postpartum hemorrhage (PPH) has been shown to improve outcomes, as these patients are at risk of SMM (Anesth Analg 2010;110:1368-73). However, early recognition is complicated by the fact that typical signs and symptoms of hypovolemic shock (tachycardia and hypotension) may not appear until 25% of a parturient’s blood volume has been lost. Prompt use of POCUS can show hypovolemia before vital sign changes are noted and coagulation derangements have begun (PLoS One 2013;8:e57594). Informed by critical care and trauma hemorrhage protocols, systematic responses to PPH, based on estimated blood loss, have improved outcomes (asamonitor.pub/3kIm5qp). These protocols delineate increasing interventions based on severity of blood loss or development of disseminated intravascular coagulation (“DIC”). Point-of-care testing, such as thromboelastography (“TEG”) or rotational thromboelastometry (“ROTEM”), can be key in treating coagulation derangements; however, treatment of presumed coagulopathy should not be delayed for lab results (Anesth Analg 2019;128:414-23). Specific OB massive transfusion protocols can be helpful to automate the provision of blood products in cases of massive hemorrhage.
AFE rarely presents with the classic triad of hypoxia, hypotension, and coagulopathy. POCUS may help distinguish the cause of hemodynamic collapse when both AFE and PPH are possible by demonstrating RV strain or frank RV failure, as pulmonary pressures rise drastically during this anaphylactoid event (Am J Obstet Gynecol 2020;222:48-52). Bedside cardiac ultrasound can be useful in assessing the degree of RV failure and recovery, as the duration of hemodynamic instability is often relatively brief. RV support in the form of inotropes (such as epinephrine, milrinone, or dobutamine), inhaled pulmonary vasodilators (such as epoprostenol or nitric oxide), and, in the case of profound hemodynamic collapse, ECMO, are all appropriate therapies to consider. As is often the case in critical illness, early intervention is key to improving survival.
The complexity of illness is increasing in pregnancy, and the use of validated screening tools may help to identify those at risk for SMM and critical illness in pregnancy. Application of critical care principles in the assessment of risk, identification of early deterioration, management of sepsis, the use of POCUS, and evidence-based protocols are all vital tools in differentiating causes for SMM and avoiding maternal mortality. The anesthesiologist is uniquely suited to quickly recognize changes in patient status that may otherwise be assumed to be normal in pregnancy. This early recognition of deterioration allows for prompt treatment and care escalation so the crashing parturient can be saved.