For too many families, delays in performing Cesarean sections have transformed the most joyous occasion in the life of a family, the birth of a child, into a tragedy. Like other human beings, fetuses need a constant supply of oxygen, a supply that can be compromised during labor and delivery. Delays in restoring the oxygen supply by performing a Cesarean can cause brain damage or even kill the child. The daily struggles of the survivors, and that of their families to support them, are heroic. Sadly, the debate has become complicated by misinformation about risks and benefits of Cesarean sections.
To properly address the issue, a few truths must be acknowledged. Clearly, Cesarean sections are not inherently good. Cesareans have risks and some may not be medically justified. Surgery carries with it inherent risks and can never be taken lightly. Many significant complications are associated with Cesareans in the medical literature (wound infection, hemorrhage, skin lesions, urogenital fistula, gastrointestinal problems, omental masses, pelvic lesions, hematoma, urinary tract infection, infertility and death. See http://3.ly/ReThinkingPatientSafetyCesareanRisks. ) Some of these issues were recently addressed in a conference in Ethiopia: “Human Resources for Maternal Survival: Task-Shifting to Non-Physician Clinicians.” (see http://www.ijgo.org/article/PIIS0020729209005311/fulltext ). There has to be a sufficient medical reason in order to justify exposing the mother and fetus to those risks,
Here is a second truth: Cesareans are not inherently bad. Some Cesareans have benefits and are medically necessary. Every human, including a fetus, needs a constant supply of oxygen. The fetus depends entirely on oxygen that passes through the placenta and umbilical cord. (Think of the fetus as a deep-sea diver, and the placenta as the oxygen tank.) Events which impair the function of the placenta (e.g. placental abruption) and umbilical cord (e.g. cord compression) threaten the oxygen supply necessary for fetal metabolism. Without oxygen, the baby’s brain cells can be injured or die within minutes. Restoring the oxygen lifeline to the baby may require a Cesarean.
Deciding when to perform a Cesarean in some deliveries requires analyzing the facts, and balancing risks and benefits. The decision makers must decide which route of delivery is the best for both patients after considering all relevant details. They must consider the desires and wishes of the mother, the details of the mother’s medical history, and the risks and benefits to the mother of performing a Cesarean at that time. They must also consider the available staff and facilities,, the condition of the fetus, the risks and benefits to the fetus, the progress in labor, the prospects for vaginal delivery, and any other risks or benefits for the fetus.
A Cesarean delivery during labor could be considered “unnecessary” under several scenarios. In the most extreme case, if all the relevant factors weighed against the decision to opt for a surgical delivery, and in favor of proceeding with a vaginal delivery, the Cesarean would be contraindicated and should not be performed. Similarly, if some important factors weighed against a surgical delivery, and no relevant factors weighed in favor of a decision to operate, the Cesarean would not be indicated and should not be performed.
Abdominal surgeries carry inherent risks and should not be performed unless there is sufficient reason. We do not hear of abdominal surgeries performed arbitrarily, e.g. the patient wanted a photograph of her small intestines. Like other abdominal surgeries, Cesareans have inherent risks and should not be performed arbitrarily. Unindicated Cesareans, i.e. those lacking surgical justification, are unnecessary and should clearly be avoided. These cases likely represent, at most, a tiny fraction of Cesareans if any.
The interesting cases are the ones in which competing considerations must be balanced. Surgeons must compare Cesarean risks and benefits in particular cases as the labor progresses and the calculus evolves. Thoughtful decisions will consider the availability of staff to administer anesthesia and resuscitate the newborn as well as the potential impact of intervention and non-intervention on two patients. These decisions that balance risks of fetal brain damage and maternal hemorrhage are difficult enough without misinformation.
Sadly, the debate has been clouded by medical literature written to defend birth trauma lawsuits. Many opinions expressed in the public arena about how many and which Cesareans are unnecessary rely on that literature. Some of the misinformation comes from respected sources of medical information including researchers at the National Institutes of Health, of the New England Journal of Medicine, and the American College of Obstetricians and Gynecologist. The misinformation usually appears in articles about whether events in labor and delivery cause, or a Cesarean might prevent, fetal oxygen deprivation and resulting cerebral palsy. Failures to perform and delays in performing a Cesarean are frequently cited as the basis for medical malpractice cases against obstetricians. One of the elements of such cases is that the child’s injuries, the cerebral palsy, could have been prevented with a timely Cesarean. If there was scientific evidence that events in labor and delivery, and particular fetal asphyxia, did not cause cerebral palsy, that evidence could be used to defend malpractice cases.
What is the purpose of the misinformation? In February 2003, NIH scientist and author Dr. Karin Nelson told Marie McCullough of the Philadelphia Inquirer that one of the major pieces of literature on the subject, ACOG’s Neonatal Encephalopathy and Cerebral Palsy:
“It’s intended for litigation.”
(“Doctors dispute their role cerebral palsy role” by Marie McCullough, Philadelphia Inquirer, February 10, 2003. Also see Billings Gazette, March 24, 2003, http://www.utsystem.edu/news/clips/dailyclips/2003/0323-0329/UTMB-BG-Doc...) Dr. Nelson should know: she is one of the authors. Dr. Nelson has been a central figure in the body of misinformation on the subject.
One of the first pieces of misinformation on the subject was published over two decades ago. In 1986, Dr. Nelson was the lead author for a major article in the New England Journal of Medicine that has been was cited as proof for the proposition that events in labor and delivery did not cause cerebral palsy. (See “Antecedents of Cerebral Palsy: Multivariate Analysis of Risk,” 315 NEJM 81-86 (1986)). The article was heralded as proof of the proposition that The article was cited repeatedly in Peter Huber’s 1991 book Galileo’s Revenge (See “Chapter 5: Gadgets and Knives,” pages 78 - 88) as proof that cerebral palsy is not caused by events in labor and delivery. The United States Court of Appeals for the Fifth Circuit cited this 1986 article prominently in Tanner v Westbrook, 174 F.3d 542 (5th Cir. 1999), an opinion that threw out a verdict for a birth trauma plaintiff.
Unfortunately, Yet the conclusion for which that article is celebrated, that events in labor do not cause cerebral palsy rests on a logical fallacy. (See Shier and Tilson, “The Temporal Stage Fallacy,”Medicine, Health Care and Philosophy (2006) 9:243–247.) The 1986 article offers no evidence for the conclusion that events in labor and delivery do not damage fetal brains. Instead, through a novel form of argument that compares rates of cerebral palsy associated with various temporal stages of labor and delivery, it creates the impression that events in labor and delivery do not contribute significantly to that condition.
In January 1992, The American College of Obstetricians and Gynecologists contributed a second entry to the body of misinformation when it promulgated ACOG Technical Bulletin 163. That bulletin defines the circumstances under which perinatal asphyxia might be said to cause cerebral palsy. Based on 28 medical articles cited in the footnotes, Technical Bulletin 163 sets forth four criteria (A. pH < 7.0, B. low Apgar, C. multiorgan injury, and D. seizures etc.) that must be satisfied before one can conclude that perinatal asphyxia caused cerebral palsy in a newborn. However, on careful examination, the 28 articles cited in the footnotes fail to support its conclusions. Some articles contradict the very propositions for which they are cited. For instance, the bulletin cites an article by Ruth and Ravio (see ACOG Technical Bulletin 163 footnote # 12: Ruth VJ, Raivio KO. “Perinatal Brain Damage: Predictive Value of Metabolic Acidosis and the Apgar Score,” BMJ1988;297(6640):24-27) for the proposition that a newborn’s pH must be below 7 in order to show that events in labor and delivery may have contributed to a child’s cerebral palsy. However, the article discusses 14 cases, of which two have some indication of asphyxic brain injury. In both cases, the pH is above 7.0 ( See p 26, Table IV, “Clinical Data on Infants who died or who were Clearly Abnormal at 1 year of age;” case #1 is described “asphyxia” with pH = 7.11; and case #11 is described as “Hypoxic-Ischaemic Encephalopathy, Seizures” with pH = 7.15.) (See Lee Tilson, “Exposing Manufactured Scientific Literature,” MTLA Quarterly, Summer 1994, Page 12.) Thus, the evidence in the Ruth and Raivio article contradicts the proposition for which it is cited.
Similarly, in footnote #17, Technical Bulletin 163 cites an article by Perlman and Tack for the proposition that on needed multi organ injury to have asphyxial brain injury. (J Perlman “Renal injury in the asphyxiated newborn infant: relationship to neurologic outcome,” Journal of Pediatrics,1988, vol. 113, no. 5, pp. 875-9). However, in Table 1 “Short- and long-term outcome in term infants with and without oliguria,” the next to the last row indicates that 2 out of 7 term infants with abnormal long-term outcome had normal urine output. Thus, Technical Bulletin 163 cites an article in which about 30% of asphyxiated term newborns have normal urine output and abnormal CNS (Central Nervous System) exams. And the bulletin cites if for the proposition that normal urine output rules out a connection between asphyxia and CNS abnormalities. The evidence in the Perlman and Tack article contradicts the proposition for which it is cited. One would think that with only 28 articles cited in the footnotes, the authors should have been aware of these problems.
Eleven years later, with great fanfare, ACOG published Neonatal Encephalopathy and Cerebral Palsy in January 2003. With more organizational endorsements, ACOG again weighed in on the subject with a visually impressive eight chapter booklet with over 400 footnotes. Sadly, the logic of the publication is just as flawed as its predecessors. Even the executive summary contains contradictions (about whether perinatal asphyxia causes “dyskinetic” cerebral palsy).
Less than two weeks after publication, Dr. Karin Nelson, one of the authors, was quoted in the Philadelphia Inquirer as admitting that the booklet was “intended for litigation.” (“Doctors dispute their role cerebral palsy role” by Marie McCullough, Philadelphia Inquirer, February 10, 2003, and Billings Gazette, March 24, 2003, see http://www.utsystem.edu/news/clips/dailyclips/2003/0323-0329/UTMB-BG-Doc...)
Much of the literature on the causes of cerebral palsy including the 1986 New England Journal of Medicine article cited above and many of the footnotes in ACOG’s 2003 Neonatal Encephalopathy and Cerebral Palsy are based on conclusions drawn from analysis of the National Collaborative Perinatal Project. That data was collected from over 50,000 deliveries between 1959 and 1965 at 12 major university hospital systems in the United States. That project is the basis for much of the medical literature purporting to show that fetal heart monitors do not reliably identify the labors and deliveries in which Cesareans should be performed. Many authors cite articles based on this project for the conclusion that Cesarean deliveries have no benefit. The argument claims that the promise of fetal monitoring has not been realized.
The flaw with this argument? Dr. Hon did not invent fetal heart monitors until the 1970’s. Not a single one of the 50,000 + deliveries in the National Collaborative Perinatal Project used a fetal heart monitor. The results of that study are irrelevant to whether the information from fetal heart monitors should be considered in deciding on whether to perform a Cesarean during a delivery. Indeed, the 1986 New England Journal of Medicine article discussed above was based on the National Collaborative Perinatal Project. Fetal heart monitors were not available. Yet this body of literature is cited repeatedly for the propositions that performing Cesareans based on fetal heart monitor tracings has no benefit.
When are Cesareans indicated? Providing an overview of the thousands of medical articles on when Cesareans are indicated would be an enormous undertaking. In the alternative, let me offer a thought experiment for discussion.
Most birth asphyxia cases with which I have been involved share a common pattern. After fetal heart monitors display an abnormal fetal heart pattern, the nurses and staff undertake what is called “intrauterine resuscitation.” Procedures such as giving the mother a flow rate of oxygen higher than normal, turning the mother onto her left side to hopefully increase circulation of oxygen to the fetus, turning off hormones that strengthen contractions, or other procedures are sometimes taken to restore the fetal heart rate to normal. Everyone is relieved when the fetal heart pattern returns to normal. The indications of possible asphyxia are gone, the problem is solved. Even if the earlier abnormal fetal heart pattern correctly identified fetal asphyxia, it is now gone. The fetus is not longer asphyxic. Therefore, it is reasoned, there is no longer any reason to expose the mother to the risks of surgery by performing a Cesarean.
Typically, in the birth asphyxia cases I have litigated, there are several episodes of abnormal fetal heart rate patterns that are relieved by intrauterine resuscitation. At some hours in labor, an abnormal fetal heart pattern indicating possible fetal asphyxia appears. The medical staff and nurses use intrauterine resuscitation to relieve the distress and the pattern will return to normal. Labor will progress for some time before a second episode of abnormal tracings. When the second episode of abnormal tracings appear, the staff again uses intrauterine resuscitation to restore the fetal heart tracing to normal. Abnormal patterns indicating possible asphyxia continue to appear a third, fourth and fifth time. Each time, the patterns are relieved each time with intrauterine resuscitation. The medical staff repeats this intrauterine resuscitation so long as it works.
The thought process seems to be that since Intrauterine resuscitation carries no risks and surgery carries potentially significant risks, intrauterine resuscitation is the preferred method of relieving signs of fetal asphyxia. When fetal asphyxia appears, undertake intrauterine resuscitation first. Perform surgery only if an intrauterine resuscitation fails. This prevents unnecessary Cesareans. Cesareans are not performed unless intrauterine resuscitation no longer relieves fetal distress.
Many litigated birth asphyxia cases involve repeated attempts at intrauterine resuscitation. The Cesarean is not performed until the fourth, fifth, or sixth attempt at resuscitation relieves the signs of fetal distress on the fetal heart monitor. Unfortunately, when the fetus is compromised, sometimes a Cesarean cannot be performed quickly enough. After the fetus is stressed to its limits of toleration and beyond, after all the fetal reserves have been depleted, the fetus may not have the ability to wait for the surgical team to be assembled. At some institutions, assembling a team may take fifteen to twenty five minutes before a Cesarean can be attempted. How many humans can hold their breaths this long? Hypoxic stress is apparent. Should the fetus be exposed to hypoxic stress beyond the capabilities of Olympic athletes?
The proposition that intrauterine resuscitation must fail before Cesareans are indicated and before surgical teams can be assembled creates an unconscionable risk of fetal asphyxia, injury and death.
Consider an alternate approach. In this sequence of events, when do we know that we have a problem? When do we have a good reason for assembling a team to perform a Cesarean? Let’s consider the same set of circumstances in a different light.
The first episode of intrauterine resuscitation was successful. The nursing and medical staff were concerned enough about an abnormal fetal heart tracing to perform an intrauterine resuscitation. There was a reason for believing that an abnormality on the fetal heart monitor strips indicated that the fetus needed oxygen. More oxygen was delivered to the fetus. The abnormal tracing disappeared. What can be deduced from these events?
The medical and nursing staff are concerned that the fetus needs oxygen and administers intrauterine resuscitation. More oxygen is delivered to the fetus and the problem disappears. This is good evidence that the cause of the abnormal tracings was a lack of oxygen. If oxygen solved the problem, eliminated the abnormal tracing, it is reasonable to infer that a lack of oxygen caused the problem.
The second intrauterine resuscitation is also successful. Twice, oxygen has relieved an abnormal fetal heart tracing. The evidence that the fetus is not receiving enough oxygen has become obvious. Twice, oxygen has cured problems on the fetal monitor strip. Is it possible that the abnormal fetal heart tracing might occur a third time? If the abnormal tracing occurs a third or fourth or fifth time, are we sure that each occurrence can be cured by intrauterine resuscitation? If the fetus is going to be exposed to the risk of additional episodes of asphyxia, should the health care team be in a position to relieve it before the fetus is injured or killed?
Given these circumstances, shouldn’t the obstetrician be notified immediately? What reasons are there for withholding the information from the obstetrician? Shouldn’t a surgical team be assembled and available in case subsequent problems on the fetal monitor strip are not relieved by intrauterine resuscitation?
Everyone should be concerned about mothers exposed to the risk of an unnecessary Cesarean. Similarly, everyone should be concerned about the risks of brain damage and death when a fetus is exposed to unnecessary asphyxia while a surgical team is assembled.
When a surgical team is not assembled, a gamble is placed. The medical staff gambles that the next time there is a problem on the monitor indicating asphyxia, yet another round of intrauterine resuscitation will solve the problem and relieve the asphyxia. Every time this gamble is made, the fetus faces significant stress and an uncertainty as to whether it can be relieved. At some point, after enough episodes of stress, the stakes are raised. The stakes become the baby’s brain and life. The gamble is that a surgical team can be assembled fast enough to prevent brain damage or death.
I submit that we need to rethink the evidence. A successful intrauterine resuscitation strongly indicates that the fetus needs oxygen. A successful intrauterine resuscitation is not an “all clear” sign, but a warning sign. It is as clear of a warning sign as nature can give us that the fetus is not getting enough oxygen. If oxygen solves the problem, a lack of oxygen might have caused it.
A successful intrauterine resuscitation does not mean that subsequent Cesareans are unnecessary. The opposite is true. A successful intrauterine resuscitation confirms that the fetus did not receive enough oxygen and is at risk of suffering asphyxia for the rest of that labor and delivery.
Unnecessary Cesareans are bad and to be avoided. They expose the mother to unnecessary and significant risks.
The failure to perform a timely Cesarean should also be avoided. That failure exposes the fetus to the risk of unnecessary brain damage and death. Unfortunately, the misinformation injected into the debate has made it more difficult to recognize when Cesareans are necessary.
Lee Tilson litigated medical malpractice cases for decades and was drafted into the patient safety movement by medical errors that adversely affected two family members. Listed in the Best Lawyers in America, Lee has been a frequent public speaker, has written numerous articles and book chapters, has presented at grand rounds, has taught Health Law, was invited to a World Health Organization meeting on patient safety, founded the Birth Trauma Litigation Group and has served on the board of the Hypertrophic Cardiomyopathy Association. Lee founded and blogs at www.rethinkingpatientsafety.com.