Perimortem cesarean delivery (PMCD) is rarely performed, with fewer than 300 cases reported in the English-language literature. Nonetheless, it is an important topic for two reasons. First, there is a clear imperative to promptly decide and act when PMCD is indicated. As in all emergency care, knowledge must precede the crisis rather than await it. Second, indications for PMCD have broadened considerably since the 1980s, and the procedure may attain a more prominent role in the future.

Cesarean delivery is one of the oldest surgical procedures in history, with literature dating to at least 800 BC. Prior to the 20th century, the phrase "postmortem cesarean" would have been redundant, as the procedure was essentially undertaken only in dead or moribund mothers. Initially, the Roman decree (Lex Cesare, [Law of Caesar]) that unborn infants should be separated from their mothers' bodies was for purposes of religious ritual rather than for any real hope of survival of either the mother or the child. Some infants appear to have survived because the law specified "with the hope of preserving citizens to the State," and failure to obey the mandate was grounds for "legal suspicion that a living child had been killed."

Indeed, several mythological and historical figures reportedly were born in this fashion, including Apollo's delivery of his son, the Greek physician Asklepios, "from the womb of dead Koronis." Bacchus was supposedly born this way, as was Scipio Africanus (the Roman general who defeated Hannibal). Pliny the Elder dates this event at 237 BC. Pope Gregory XIV had this distinction, as did 15th-century Genoese admiral, Andrea Doria. Some attribute the birth of Edward VI as occurring after the death of the unfortunate Jane Seymour, although others claim she lived several days after the delivery. Shakespeare referred to cesarean delivery in Macbeth.

Judaic writings from the first through third centuries AD identified this practice and found the mother not liable for a sacrifice of purification, as were women who delivered vaginally. Christian leaders made it a religious issue in the 1280 Council of Cologne as a means of assuring the infant's baptism.

The first documented record of maternal survival after a cesarean delivery is that of the Swiss sow gelder Jacob Nufer, who sectioned his own wife for the delivery of their firstborn in 1500. None of the available references detail what must have been the desperate circumstances leading to this intervention, but this hearty lady survived to have 5 spontaneous vaginal deliveries, including a set of twins. She became the first recorded patient with a successful postcesarean trial of labor and the first survivor of the procedure.

By the time of the Renaissance, PMCD had become such a standard practice that, in 1749, King Charles of Sicily ruled that failure to perform the procedure was punishable by death, and one instance of the law being applied to a physician is recorded.

During the late 19th and early 20th centuries, case reports began to appear of PMCDs successfully salvaging the fetus, and the procedure began to be seriously entertained as a legitimate medical intervention. The salvage rate was low well into the 20th century; therefore, wisdom dictated that all possible attempts be made to resuscitate the mother with the infant still in utero unless demise was clearcut and inevitable.

During the 1980s, several authors reported unexpected maternal recoveries after "postmortem" deliveries. This led to a consideration of the possibility that PMCD might actually improve, rather than worsen, a mother's chance of survival during a collapse. Uteroplacental blood flow may require up to 30% of a woman's cardiac output during pregnancy, some of which may be retrieved for other visceral organ perfusion after delivery. Indeed, several animal and laboratory models and a growing body of clinical evidence suggest that cardiac compressions are significantly more effective once they are relieved of the caval compression associated with a term pregnancy and of the tremendous circulatory demands of a placenta and fetus.

A 30% decrease occurs in stroke volume and cardiac output in a pregnant woman lying supine. This reduction is largely a result of complete occlusion of the inferior vena cava, which occurs in 90% of women in late pregnancy. In addition, a 20% reduction in functional residual capacity at term and a higher metabolic rate occur, leading to decreased oxygen reserves and faster onset of anoxia following apnea.

Delivery of the near-term fetus provides a 30-80% improvement in cardiac output and, in conjunction with other resuscitative measures, may provide sufficient circulatory improvement to adequately support CNS function during an arrest. This has led to the current belief of PMCD as an appropriate resuscitative intervention for both mother and infant. In light of this, intervening promptly and appropriately is critical to maximize the survival possibilities for both patients simultaneously.

In summary, the perspective on PMCD has evolved through 23 centuries from a means of providing appropriate burial and/or ritual for both mother and baby to a way of saving a child's life when maternal death is inevitable to a method of optimizing resuscitation for both mother and baby.

One of the reasons PMCD has become a more realistic procedure relates to the prevailing etiologies of maternal demise. In a 1986 review, Katz et al highlight the shift over the past century from primarily chronic, mostly infectious causes of mortality to primarily acute, mostly cardiorespiratory causes of mortality. The chronically ill mother may be hypoperfusing or inadequately nourishing her unborn child for months, thus making a good outcome of any delivery less likely. Conversely, an emergent event, such as pulmonary embolus, leaves the infant with good reserves and allows a less-than-optimal delivery setting to produce a good outcome.

In addition to changing diagnoses, the evolution of medicine allows more hope of success. The ability to monitor high-risk patients and to intervene in advance of a crisis has largely developed in the last half of the 20th century. Cardiac and respiratory support is available, at least for the short term, in virtually every setting in which medicine is practiced. Emergency transport services allow prompt medical attention to life-threatening conditions and, in a clinical setting in which time is the single most critical factor determining success, may be the most important development in medicine.

Several factors discussed in the literature must be considered when deciding whether to undertake PMCD. The first is estimated gestational age (EGA) of the fetus. This information is occasionally difficult to obtain in an emergency situation, and a time-consuming ultrasonographic estimate is not practical. Thus an "eyeball estimate" may be necessary. As a general rule, the uterus reaches the umbilicus at 20 weeks of gestational age and grows at the rate of approximately 1 cm in length for every week thereafter. Thus, in a relatively thin woman, a fundal height of 8 cm above the umbilicus likely represents a pregnancy of 28 weeks' gestation.

Depending on the resources of the institution, fetal salvageability under ideal circumstances (availability of all skilled personnel and a controlled setting) may be anywhere from 23-28 weeks' EGA and should be considered in the decision regarding PMCD. If the fetus is known to be 23 weeks' EGA and survival of a 23-week-old fetus in the institution's nursery has never occurred, PMCD is probably not indicated for the sake of the fetus. PMCD may be of less help to the mother as well, compared to third-trimester intervention. Cardiovascular effects of pregnancy are less pronounced prior to 28 weeks; thus, delivery will not achieve as dramatic a cardiovascular improvement as at a later EGA. In the first and early second trimesters, aggressive maternal support is the only indicated intervention. At least one case of a complete maternal and fetal recovery after a prolonged arrest at 15 weeks' EGA has been reported.

A second concern relates to length of time between arrest and delivery. The latest reported survival was of an infant delivered 22 minutes after documented maternal cardiac arrest. Most of the authors support a 5-minute rule. Best outcomes in terms of infant neurologic status appear to occur if the infant is delivered within 5 minutes of maternal cardiac arrest. This means the decision to operate must be made and surgery started by 4 minutes into the code. This concern is poorly understood by most clinicians faced with the decision. The tendency is to wait while maternal resuscitative efforts are optimized and to attempt PMCD only in the face of a terminated code. Although the evidence is not based on a huge number of cases, a compelling argument for prompt intervention can be made on both maternal and fetal grounds.

The third critical factor to success is adequacy of other resuscitative efforts in the interim. Although chest compressions may provide only 30% of baseline cardiac output, some oxygenation is clearly better than none. Displacement of the uterus leftward until surgery is actually started allows for better blood return from the inferior vena cava. Because the fetus lives on the steep portion of the oxygen dissociation curve, relatively minor changes in maternal oxygenation result in far more dramatic changes for the fetus.

Fourth, trained personnel to accomplish and maintain neonatal resuscitation must be available. Successfully extracting a 28-week-old infant only to have no qualified person available to assume resuscitation efforts is tragic. All infants in this setting may be presumed to require postdelivery support, and the delivery team will most likely be fully occupied with the mother as well as suboptimally experienced with infant resuscitation techniques.

Finally, a vital factor beyond the control of the delivery team is the nature of the maternal condition. Acute conditions result in better outcomes than chronic, and conditions primarily involving cardiopulmonary collapse respond better than more systemic organ failures.

Documentation of fetal heart tones prior to PMCD is not required, partly because doing so is time-consuming and may negatively impact the infant's outcome and partly because maternal indications for the procedure are pressing regardless of fetal status.

To summarize, conditions that affect the success of PMCD are gestational age of the infant (>24-28 wk is optimal), length of time from arrest to delivery (≤5 min is optimal), adequacy of maternal resuscitative efforts, availability of neonatal resuscitation experts, and nature of the underlying maternal condition.

A special case relates to the "scheduled PMCD." This involves a woman who would be considered legally "brain-dead" but is being maintained on artificial support solely for the purpose of allowing further fetal development. While successful cases of this sort have been reported from as early as 6 weeks' EGA, it can be argued that extraordinary support measures for the sole purpose of providing a fetal incubator constitute experimental interventions and require full informed consent of the next-of-kin.

The most likely timeframe for both successful support and ethical imperative to support is at 24-27 weeks' EGA, when a few days make a large difference to fetal outcome. Support beyond likely fetal ex utero survival is controversial, as is support from an early EGA. Dillon et al make a strong distinction between true brain death and persistent vegetative state, and they argue that termination of support measures is ethically defensible only in the former case.

The niceties of preoperative preparation are unlikely to be practical in this setting. If a urinary drainage catheter has not already been placed, do not spare the time to place one. Similarly, a detailed assessment of fetal well-being is impractical. Even seeking and assessing for fetal heart tones is unnecessary because this can be difficult and fraught with error (in the large patient or in a noisy room) and because the delivery is performed as much for maternal indications as fetal. A "splash prep" of antiseptic solution across the abdomen is ritually satisfying but of uncertain clinical value. In essence, preparation consists of acquiring some basic equipment listed below and baring the patient's abdomen.

Continue full cardiopulmonary resuscitation measures during the delivery. This optimizes oxygen delivery to both patients.

Most young obstetricians perform Pfannenstiel incisions almost exclusively for cesarean deliveries. Although obstetricians are taught that midline incisions allow faster entry into the abdominal cavity, many have found that with optimal lighting, equipment, and assistance, a Pfannenstiel incision takes little or no longer time to perform and produces a stronger, more aesthetically pleasing scar.

This absolutely does not apply to the incision made under suboptimal conditions. The equipment available is likely to be minimal and not laid out neatly with a scrub technician standing by. While many spectators may be present, no real assistants are likely to be available. Lighting may be poor and not deployable where needed within the field. Given these restrictions, a midline abdominal incision remains the appropriate choice for performance of PMCD.

Do not abandon regard for surgical technique despite the limitations of the setting. Given the possibility of maternal survival, take care to protect bowel and bladder from injury if possible. Protect the fetus from the large laceration that is a probable consequence of reckless uterine entry. Deliver the infant with attention to planes of anatomic function so that permanent nerve damage from overextension does not occur. A person trained in neonatal resuscitation should be available to assume medical responsibility for the infant as soon as delivery is accomplished.

If desired, a loop of cord may be clamped off at each end and saved for later evaluation of cord gases. The closed loop of cord may sit for as long as 60 minutes without significant degradation of gases. As with all deliveries, collect cord blood so that routine neonatal hematologic studies can be performed without necessitating blood draws from the infant.

Always remove the placenta prior to closure. If the mother survives the initial collapse, bleeding or infection with residual placenta as a nidus worsens her chances of eventual survival. If she does not survive the initial episode, the placenta may deliver spontaneously in the morgue if not extracted at time of surgery. This is inconvenient at best.

Undertake closure based on maternal circumstances. If the cardiorespiratory resuscitation team believes that a chance of survival exists, perform a careful, layered closure as with any cesarean delivery. Attention to meticulous closure technique is vital because suboptimal perfusion may not reveal areas of bleeding that could later become problematic when circulatory function is restored. In addition, disseminated intravascular coagulation is a common sequela of massive hemodynamic challenge and may create postoperative problems if closure is not meticulous. If the maternal condition is believed to be hopeless, than a rapid closure for purposes of esthetics is indicated as a kindness to the family.

Consider antibiotic prophylaxis if maternal survival seems likely and in light of the probable absence of sterile technique. Although no series define an optimal choice for this setting, the rules of "dirty" surgery should apply, and any broad-spectrum penicillin or cephalosporin should suffice.

The person best suited to perform the PMCD is the most experienced obstetrical surgeon available. That the first reported cesarean success was in the hands of a sow gelder, presumably comfortable with the feel of live tissue and familiar with concepts of vascular control, is probably no accident. Sow gelders are relatively unavailable these days, but most communities have obstetricians. Their expertise in estimating risk-success factors (eg, gestational age) and their operative experience serve both patients well. If the maternal collapse occurs in any setting other than a labor and delivery suite, calling for obstetric and pediatric support as soon as the possibility of PMCD becomes known is prudent.

The following is a review of PMCD technique:

• Perform only preoperative preparations that will not delay the procedure.

·         Perform a midline abdominal incision, umbilicus to pubis (longer if necessary), and extend it to and through the uterine wall.

·         Protect the bowel, bladder, and other maternal organs to the extent possible under the circumstances.

·         Provide due attention to the position of the fetus and deliver by careful, anatomically attentive extraction.

·         Extract the placenta prior to closure.

·         Close all layers as meticulously as maternal condition dictates.

·         Inspect carefully for unexpected damage and repair if indicated.

·         Pay particular regard to hemostasis.

·         If maternal survival seems likely, consider 24 hours of antibiotic prophylaxis.

The optimal setting for PMCD is the operating room (OR) with an open section pack and trained OR crew. If this is not available, several equipment packages have been recommended. These can be quite elaborate or as simple as a "Urgent delivery tray" .

The essential items, for the most part, are readily available and include a scalpel, laparotomy sponges, bandage scissors or straight Mayos, 3 or 4 Kelly clamps, suture material (preferably #0 or #1 chromic or Vicryl), needle driver, forceps (preferably Russians, Bonneys, or heavy toothed), bulb suction, warmer for the infant, and full neonatal resuscitation support (eg, drugs, endotracheal tubes).

Unfortunately, the practice of medicine has been irreversibly tainted by the fear of being held liable for actions undertaken with the best possible knowledge and compassion but with a bad outcome. This fear is primarily accountable for the flight of many obstetricians into gynecology-only practices, in which litigation risk is much lower. Fear of legal repercussions may also lead to withholding PMCD when it might be effective.

Physicians may take comfort in knowing that, according to the medical literature on the subject, a suit has never been filed on the basis of wrongful performance of PMCD. In fact, only one legal penalty has been levied: the aforementioned death penalty in the 18th century for failure to perform the procedure.

Generally, PMCD is deemed to fall under the same guidelines as any emergency procedure in which consent is not possible. While some medical literature opines that consent should be obtained if possible, determining who is legally qualified to grant consent may be difficult. Roseate discussions of "talking to the husband" fail to consider the complexities of current relationships. A patient has the right to refuse a cesarean delivery, even if her baby is in extreme jeopardy. This is based on the ethical principle of maternal autonomy. Nonetheless, in a setting in which maternal consent is not an issue, no other opinion should be deemed as legally binding. When the issue is the ventilator-dependent brain-dead patient being kept alive solely as a nursery, next-of-kin decisions clearly become relevant, and legal and spiritual counsel should be sought.

PMCD should be undertaken as part of maternal and fetal resuscitation in any gestation advanced beyond 24-28 weeks. It should be undertaken within 4 minutes of full cardiac arrest if possible, and the technique should be subject to time and environmental constraints. The most experienced obstetrician available is the optimal person to perform the procedure, and a person trained in neonatal resuscitation should be available if at all possible. Medicolegal considerations have thus far been only theoretic, and the law is likely to support this procedure under the emergency care rubric should it ever be tried. Mothers who are brain-dead may undergo extraordinary support measures for prolonged periods, but this should involve extensive discussion with next-of-kin.