Anesthesia-related maternal arrest is a feared complication that places the lives of both mother and baby at risk. Literature reviews on the subject have been traditionally hampered by a lack of specifics regarding the care provided and the aging of data by the time it could be collected and analyzed. Valid concerns about the privacy of stricken families, the confidentiality of the healthcare providers involved, and liability risks have likely acted together to prevent the wider dissemination of case specifics in an open forum.
Despite recent advances and changing practice patterns in obstetrical anesthesia, malpractice claim reviews indicate that maternal arrest on labor and delivery continues to result in major morbidity and mortality. The Doctors Company recently reviewed 22 anesthesiology claims that were filed after maternal arrests on labor and delivery wards between 1998 and 2006. Anesthesia care was analyzed at the time of initial medical record review from both Standard of Care and patient safety viewpoints. Characteristics of these claims and expert reviewer comments regarding suggested practice changes that might possibly have avoided the arrests or improved the outcomes are presented here with an aim toward improving maternal safety.
The mothers, aged 17 to 41, suffered the most severe complications post-arrest. Ten out of the 22 died, including 3 who were declared brain dead and removed from ventilators. Eleven suffered degrees of anoxic brain damage, ranging from minimal to severe, but with neurological deficits deemed to be permanent at the time of final claim resolution. Only 1 mother out of the 22 had no apparent residuals post-arrest and was suing primarily for emotional distress.
Surprisingly, the infant outcome in these cases appeared quite different. In all cases where information was available, infants (or children) had been evaluated as developmentally normal at the time of final medical records review. This was true even for the babies born with low Apgar scores, and included not only the 10 babies delivered prior to the maternal arrests, but also the 12 born after their mothers had sustained significant periods of circulatory and/or respiratory arrest.
This disparity suggests that the fetus may well be more resistant to periods of hypoxia and hypotension than is the parturient, and it reaffirms the importance of the anesthesiologist's primary focus being the welfare of the mother. This raises the question as to whether maternal resuscitation should ever be intentionally delayed in order to expedite delivery of the fetus.
The most common scenario in this series (13 patients) was a respiratory arrest following epidural or spinal block. Included in this group were 11 patients who developed unintentionally high neuraxial blockade with resultant apnea and 2 patients who arrested after intravenous sedation was administered post-cesarean section delivery under spinal anesthesia. None of these patients were attached to a maternal monitor with audible alarms at the time of the arrest, making delay in response and resuscitation a frequent reviewer concern.
There were 8 patients in this series who arrested in labor rooms following attempted insertion and dosing of epidural catheters to relieve labor pains. Of these, 7 had subsequent evidence of unintentional subarachnoid blocks, either by positive aspiration of the catheter for cerebrospinal fluid (CSF) (3 patients) or air in the ventricles on CT scanning (4 patients), presumably introduced into the CSF during injection through the needle or catheter.
All 8 of these arrests occurred within the first 30 minutes of initial catheter placement. In half of the cases, anesthesia providers were not in the room at the time, and the arrest was noted first by other healthcare workers. Reasons given by anesthesia providers for leaving the room after catheter placement included wanting to chart at the nursing desk, being called away to place another labor epidural or to attend to another labor patient's needs, or needing to locate drugs or airway equipment.
In the 1 case in which the mother recovered without obvious neurologic impairment, she was placed supine immediately by the anesthesiologist on initial complaint of difficulty breathing and ventilated with oxygen by Ambu-bag as soon as respirations appeared inadequate. The obstetrician accomplished a crash cesarean section within minutes while still in the labor room, with only a benzodiazepine provided before incision. Blood pressure was supported with IV fluid infusion.
The other 7 cases involved the transfer of a mother in respiratory and/or circulatory arrest from the labor room to the operating room for STAT cesarean section due to fetal distress. In 4 of these cases, there were documented delays in the ventilation of the mother for reasons including initial failure to notice maternal arrest, desire to wait for more optimal intubating conditions in the OR, difficulty locating an Ambu-bag or airway device, or an anesthesia provider not being in attendance. The improved outcome in the 1 case involving immediate resuscitation suggests that the rapid establishment of adequate ventilation and blood pressure support might be crucial factors after unintentionally high spinal blockade.
There were 5 cases of maternal respiratory arrest following regional anesthesia administered for elective cesarean-section delivery. These all involved spinal anesthetics, possibly because this is a preferred anesthesia choice for purely elective cases. In 2 instances, the mothers received an intravenous benzodiazepine or opioid after delivery; both had also received spinal opioids. Maternal respiratory arrests occurred after delivery in these cases, although there were possible delays in recognition of the arrests.
In the other 3 cases, the mothers received no intravenous anesthetics. One mother arrested immediately after the spinal was placed, with suspected preeclampsia and volume depletion as contributing factors. The other 2 cases involved apparent high spinals, with delay in recognition and/or resuscitation also potential problems.
Morbid obesity, which is known to complicate regional anesthesia, was documented in 3 out of the 8 labor epidural cases and 1 out of the 5 cesarean sections. These proportions would appear to be higher than those present in most labor and delivery populations and suggest that morbid obesity may be a significant relative risk factor for maternal arrest following regional blocks.
Three mothers in this series carried the diagnosis of preeclampsia. Two arrested at the time of induction of anesthesia for cesarean section (1 spinal, 1 general anesthetic). Reviewers raised the possibility of relative hypovolemia in these cases and questioned whether invasive monitoring might have provided useful additional information.
There were 7 cases involving arrests in mothers after massive postpartum hemorrhage—3 after normal spontaneous vaginal deliveries and 4 after cesarean section births. Predisposing diagnoses, when available, included placenta accreta, placental abruption, and traumatic arterial laceration. Knowing there had been a maternal arrest due to hemorrhage, reviewers attempted to identify ways in which the treatment might have been optimized, although it was acknowledged that the size and facilities of the obstetric units involved were varied.
A frequent reviewer impression was that the hemorrhage was so excessive by the time it was diagnosed, it was extremely difficult for the anesthesia provider to "catch up" with the continuing blood loss. Postpartum hemorrhage was not always initially apparent through vaginal bleeding, as it was often primarily internal. The initial presentation was frequently hypotension and/or tachycardia in the mother, which was usually treated first with intravenous crystalloid and pressors. It was not always clear when continued bleeding should have been suspected as the cause of the maternal vital sign instability.
In some instances, delays in transfusing mothers were related to problems obtaining or transporting blood products from the blood bank, or to an inability to run the blood products through available intravenous lines more rapidly. Some cases involved delays in waiting for crossmatched blood when possibly O-negative or type-specific blood might have been available. Reviewers commented that several patients might have benefited from earlier consideration of additional blood components—including fresh frozen plasma, platelets, or cryoprecipitate. Laboratory tests of serial hemoglobin and hematocrits, coagulation panels, or disseminated intravascular coagulation (DIC) screens were not always ordered.
Better communication might have facilitated transfusion in some cases. On retrospective reviews, potentially improvable delays were identified in informing blood banks of the need for products, in calling for additional medical assistance, or in notifying obstetricians that postpartum patients were hemorrhaging and that surgical intervention (such as exploration, uterine ligation, or hysterectomy) might be required. Potentially useful equipment, such as central line kits or rapid infusion devices, were sometimes available in the facility, but labor and delivery ward personnel might not have known how or where to obtain them.
Included in these maternal arrest cases were 5 general anesthetics and 17 regional blocks. This likely reflects an overall shift toward the use of regional anesthesia in obstetrics. In 3 of the general anesthetics cases, the arrest followed severe postpartum hemorrhage. In those cases, the choice of anesthesia did not likely affect the ultimate outcome significantly. General anesthesia was chosen in 2 cases for postpartum hysterectomy for patients who were already severely bleeding. Only 1 general anesthetic involved a difficult intubation and loss of the airway—traditionally one of the more feared complications of emergency cesarean sections. Interestingly, aspiration of gastric contents, traditionally listed as a leading cause of death in obstetric anesthesia,1 was not seen in this series of arrests.
The physiological changes of pregnancy undoubtedly contribute to the high incidence of anoxic brain damage and death cases after maternal arrests. The size of the full-term uterus decreases functional residual capacity (FRC) in the mother, leading to a much more rapid development of hypoxia during periods of apnea than would be expected in the woman's non-pregnant state. The increased oxygen demand of pregnancy further shortens the interval of apnea tolerated before arterial desaturation results. Although a pre-oxygenated, non-pregnant woman may sustain a several-minute period of apnea without desaturating, that same patient at 9 months' gestation breathing room air might not.
Maternal circulation is compromised in the supine position due to compression of the vena cava and aorta by the uterus, decreasing venous return and cardiac output. The necessity of placing an already unstable mother supine, to combat rising spinal levels, to transport her to the operating room, or to manage the airway, may further complicate successful resuscitation.
Since all respiratory arrests after labor epidurals in these cases occurred within the first 30 minutes after catheter insertion, increased monitoring during this time period would seem a worthwhile consideration. This could be visual—with the anesthesia provider, nurse, or a designee in the room with the patient—or through electronic monitoring of pulse oximetry, capnography, or ventilation, with an alarm audible to responsible personnel. Some birthing facilities have labor and delivery rooms equipped with pulse oximeters that read continuously at the nursing stations, yet that is not currently standard. Many of the maternal arrests following labor epidurals occurred on wards in which only fetal monitoring was transmitted continuously to nurses. The outcomes in those cases suggest that by the time hypoxia due to apnea becomes apparent on a fetal tracing, it might already be too late to prevent anoxic brain damage in the mother.
In 4 cases of planned labor epidurals, the anesthesiologist or CRNA observed symptoms or signs consistent with unintentional spinal blockade prior to the arrest (such as positive aspiration for CSF or maternal complaints of sudden headache or difficulty breathing). Since most labor epidural patients in this series arrested after unplanned subarachnoid blocks, patients for whom there are suspicions of "wet tap" may be at increased risk and might benefit from closer observation and/or monitoring.
Case reviews suggest that keeping pulse oximeter or end-tidal carbon dioxide monitor alarms in an audible mode continuously during cesarean sections is advisable, even after delivery of the newborn. As of October 2005, the American Society of Anesthesiologists (ASA) standards for basic anesthesia monitoring include the statement that whenever pulse oximeters or capnometers are utilized, the low threshold alarms should be audible.2
Rapid recognition of maternal respiratory arrest and restoration of oxygenation and ventilation should be key goals. Airway devices such as self-inflating bag/mask systems, oral and nasal airways, laryngeal mask airways (LMAs), and intubation equipment should be immediately available if needed in labor rooms, with all nurses and anesthesia providers acquainted with their location. The recently revised ASA Practice Guidelines for Obstetrical Anesthesia contain more complete lists of potentially useful equipment.3
Because an anesthesia provider may not always be the first to arrive on the scene, labor and delivery nurses should also be able to assess the adequacy of ventilation, establish an airway, and begin ventilation if necessary. Supplemental oxygen should be available in the labor room and immediately available in portable tank form, should transportation of an apneic patient become necessary. Since hypoxia will likely develop rapidly in a full-term apneic patient, adequate ventilation and oxygenation of the mother should ideally be established before transporting her to another location.
As with any resuscitation, maternal blood pressure and circulation should be evaluated and supported, if necessary, with fluids and pressors. CPR should be started as soon as maternal circulation appears inadequate. Since aortocaval compression and an elevated hemidiaphragm can complicate standard CPR, the American Heart Association suggests displacing the uterus to the left by tilting the patient, and performing chest compressions higher on the sternum (slightly above the center).4
Massive hemorrhage on labor and delivery is a rare occurrence, and as a result, many anesthesia providers have little or no experience managing it. Yet, the incidence of major hemorrhage in the obstetric population appears to be increasing over time. The increased rate of repeat cesarean sections, with the associated rise in incidence of placenta previa and placenta accreta, may largely account for this.5 One New York hospital, after experiencing two maternal hemorrhage-related deaths, created a multidisciplinary patient safety team specifically designed to handle labor and delivery patients experiencing major bleeding episodes.5
Their obstetric rapid response team includes members of the trauma team, as the individuals identified in that hospital with the most experience in establishing large-bore intravenous lines and massive volume and blood replacement. Efforts were made to identify high-risk patients, who were advised about auto-donation of blood and type and screened in advance of delivery. A cell saver blood scavenging device was used, when necessary, after the fetus was delivered and after peritoneal lavage.5
With these and other interventions, that hospital was able to significantly decrease the number of maternal deaths, despite an increase in the total number of cases of major obstetrical hemorrhage. This suggests that having a pre-planned and coordinated multi-departmental approach to maternal hemorrhage may well advance patient safety.
While maternal arrest is, fortunately, a very rare complication, the above cases are a testament to the fact that it still can and does occur—even when currently acceptable anesthesia practices are followed. Anesthesia providers and labor and delivery staff should consider planning their own response to a "worst-case scenario" before it happens to them. A few questions that those providing obstetrical anesthesia may wish to consider:
Every case included here was devastating on many levels to the patients, families, and healthcproviders involved. While it is tempting to search for "mistakes" in each individual scenario, the major issues identified were rarely unique. It is hoped that through taking a "systems" approach and focusing instead on the common factors that these cases share, similar occurrences might be prevented and maternal safety improved.