Critique of ACOG Practice Bulletin # 5, July 1999, "Vaginal birth after previous cesarean section"

by Marsden Wagner, MD, MSPH

Editor’s Note: To order copies of this practice bulletin (ISSN 1099-3630), contact:

The American College of Obstetricians and Gynecologists

409 12th St., SW

P.O. Box 96920

             Washington, DC 20090-6920

1. Problems with Final Recommendations.

1.1 Problems with the first recommendation in Level C: "Because uterine rupture may be catastrophic, VBAC should be attempted in institutions equipped to respond to emergencies with physicians immediately available to provide emergency care."

The first problem with this recommendation is that, of the eight final recommendations, it is this one which has the huge impact on maternity care in the United States; and yet, by ACOG’s own admission, there is no evidence to back up this recommendation. This is the reason, of course, this recommendation has been placed in Level C and is thus ACOG’s confession that since there is no data, they will simply have to go on the basis of "expert opinion." This is a sad regression to the days of "trust me, I’m a doctor" in spite of the new direction of medical care to evidence-based practice.

This is not the first time ACOG has been willing to make recommendations without any evidence base. For example, ACOG’s written statement that homebirth is not safe also doesn’t try to reference any data and flies in the face of overwhelming scientific evidence that planned homebirth and planned birth in an out-of-hospital freestanding birth center are perfectly safe options for the great majority of pregnant women.

So this recommendation, "VBAC should be in institutions equipped to respond to emergencies with physicians immediately available," has no data to support it—no studies showing improvements in maternal mortality or perinatal mortality related to the characteristics of institutions or availability of physicians. What ACOG doesn’t say is that one study (2) included in their reference list in their document compared VBAC and repeat cesarean section by three types of hospitals—community, regional and tertiary care—and while community and regional hospitals had more repeat cesarean section and more failed VBAC, no difference in mortality rates for these two procedures by type of institution is reported. This study suggests the ACOG recommendation is scientifically unfounded.

The second problem with the recommendation "VBAC should be attempted in institutions . . ." is that it is based on an unproven assumption: cesarean section can be accomplished faster if the labor takes place at a tertiary care hospital. ACOG presents no data in support of this assumption.

On the other hand, out-of-hospital birth has been shown scientifically to be safe when attended by midwives who, when necessary, can be in close contact with a hospital capable of emergency cesarean section. This is why homebirth and freestanding birth centers have been proven safe in those places in the United States with good communication and respect between home and hospital. The midwife can telephone the hospital and describe the emergency, and while the woman is in transport, the hospital staff is preparing, scrubbing in, etc.

How long does it take to do a cesarean section if the labor is in a tertiary care hospital in the United States? While ACOG has recommended in the past that the "decision to incision" time be no more than 30 minutes, in one study at a university hospital in the United States (3), 52 percent of the emergency cesarean sections for fetal distress had a decision to incision time which exceeded 30 minutes.

One reason for this delay is that in the United States it is not the laboring woman who is in transport, it is the absent doctor who is in transport, the doctor who has been trying to monitor the labor in the hospital by telephone. The ACOG recommendation "to have a physician immediately available" is in reality a criticism of the U.S. system in which the laboring woman’s doctor is usually not available and must be called to come in.

The elegant solution is not to take away valid choices for childbirth from the woman and her family but to change the system. Rather than insisting that the woman having a VBAC be transported at the beginning of labor to a big hospital that is away from her primary caregiver, her family, her friends and familiar community, instead do what is done in the other highly industrialized countries with maternal and perinatal mortality rates lower than ours. Develop a system in the United States in which there is close communication during childbirth between primary care in the community—home, birth center, small hospital—and the big hospital, so that when the woman in labor needs to be transported, the decision to incision time is no greater than if the same woman were laboring in the big hospital and needed to be transported from the delivery suite to the surgical suite for a cesarean section.

A third problem with the recommendation "VBAC should be attempted in institutions . . ." is that it is aimed entirely at the treatment of uterine rupture and not at the prevention of uterine rupture. The increasing rate of uterine rupture in the United States during the past decade has been alarming, which rightly concerns ACOG. But ACOG’s solution is analogous to responding to an increasing rate of drowning at a summer camp by placing some life rings out in the lake rather than teaching children how to swim better.

Attempts have been made to identify risk factors for uterine rupture, but the focus has been on maternal factors such as number and type of uterine scars. There has been insufficient attention in the obstetric literature or by ACOG to any relationship between the management of the VBAC and uterine rupture. The key issue here is pharmacological induction of labor with VBAC, which leads us to look at another final recommendation.

1.2 Problems with the second recommendation in Level B: "Use of oxytocin or prostaglandin gel for VBAC requires close patient monitoring."

After delaying for years while Cytotec induction spread like wildfire in the United States, resulting in thousands of VBAC Cytotec inductions leading to hundreds of uterine ruptures and dozens of dead newborns, ACOG finally closed the barn door on Cytotec induction of VBAC, but only after research showed a rate of uterine rupture with Cytotec induction of VBAC 28 times higher than the rate of VBAC uterine rupture without Cytotec induction. (4)

But the ACOG recommendation on VBAC under review here does not shut the barn door on the use of other pharmacological agents to induce VBAC, as this recommendation (under Level B, meaning based on limited or inconsistent scientific evidence) states: "Use of oxytocin or prostaglandin gel for VBAC requires close patient monitoring"—note: this means it is okay to use it as long as you closely monitor.

Read carefully the paragraph titled "Induction" in this ACOG document as it is an example of trying to torture the data until they confess to what you want them to say.

"Induction or augmentation with oxytocin has been suspected as a factor responsible for uterine rupture. A meta-analysis found no relationship between the use of oxytocin and rupture of the uterine scar. However, other studies indicate that high infusion rates of oxytocin place women at greater risk. Although there are studies that suggest that prostaglandin gel applied to the cervix or vagina appears to be safe, there are occasional reports of uterine rupture with prostaglandin preparations."

Note: "A meta-analysis found no relationship between the use of oxytocin and rupture of the uterine scar." This statement is false. The meta-analysis they are referring to—by Rosen et al. (5)—found a uterine rupture rate of 2.3 percent with oxytocin induction of VBAC compared with 1.5 percent with no oxytocin induction of VBAC. So there is a relationship, but it does not reach a level of statistical significance. It would take a larger sample size to reach statistical significance because uterine rupture is an unusual event. But it is what scientists call a "trend," and this trend is consistent with a relationship found in other studies. As ACOG states in this same paragraph, "However, other studies indicate that high infusion rates of oxytocin place women at greater risk" [of uterine rupture].

So the evidence, while not totally conclusive, strongly suggests that using oxytocin with VBAC increases the chance for uterine rupture. And yet ACOG does not recommend against oxytocin VBAC induction. Why? If the evidence is inconclusive, the cautious, conservative approach would be to follow the basic rule of medical practice, "First do no harm." But induction is extremely obstetrician friendly, as it allows the practitioner some control of a busy practice through scheduling the induction at a convenient time rather than waiting for spontaneous labor, which is 24 / 7. Proof? Data from the Centers for Disease Control (CDC) (6) show induction of labor in the United States doubled from 10 percent to 20 percent of all births the past 10 years. The same CDC data also show an increasing trend throughout the last decade for more births Monday through Friday. And this is the same decade in which uterine rupture also increased. The CDC data strongly support the possibility that the increase in uterine rupture the past 10 years is due to increased induction of labor including induction of VBAC.

2. ACOG’s Process for Making these Recommendations.

2.1 Who made these recommendations?

In the summary, Level C recommendations are based "primarily on consensus and expert opinion." Who was involved in the consensus and who are the experts? There is an urgent need for transparency here. I contacted ACOG, but they were unwilling to say who was in the group making this practice bulletin. Since the first recommendation in Level C—"Because uterine rupture may be catastrophic, VBAC should be attempted in institutions equipped to respond to emergencies with physicians immediately available to provide emergency care"—affects the care of a large group of pregnant women at the time of birth, consensus needs to include not only obstetricians but also midwives, nurses, family physicians, perinatal epidemiologists and consumers. Were any of these groups represented, or were only obstetricians? This is particularly important when a recommendation, such as this one, is of the "you need more of us" variety. Since this recommendation is friendly (an advantage) to obstetricians and unfriendly (a disadvantage) to family physicians, midwives and many women, the recommendation easily could be seen as self-aggrandizing for obstetricians.

Here, an editorial from The Lancet, January 8, 2000, is instructive (1). "Advocacy guidelines developed by a single-specialty group in isolation may be counterproductive, because those disciplines and professions that were not involved in the development of the guidelines but may be required to implement the recommendations mount their attacks and lodge their disclaimers. Some of the guidelines may be of the Good Old Boys Sat at Table (GOBSAT) variety, based on received wisdom rather than current scientific evidence, and may be biased by undeclared conflicts of interests. . . . Studies have shown that the balance of disciplines within a guideline-development group has considerable influence on the guideline recommendations. Widespread multidisciplinary participation is essential not only to ensure that the guideline is valid, but also that it is valued by all the members of the multidisciplinary team, in order to be incorporated successfully into practice."

2.2 Inappropriate influence of non-medical factors such as fear of litigation in making the recommendations.

Near the beginning of the document, mention is made of ACOG’s fear of litigation: "Physicians in the United States, facing increased medical-legal pressures . . . ." Furthermore, in this document, ACOG’s fear of litigation focuses on fear of VBAC litigation: "Increasingly, these adverse events during trial of labor have led to malpractice suits." ACOG’s fear of VBAC itself is revealed in the Figure 1 algorithm, which includes the need to "counsel patient regarding benefits and risks of VBAC" but does not include a similar need to counsel patient regarding benefits and risks of the woman’s other choice, cesarean section. To what extent do ACOG’s fear of VBAC and fear of litigation influence the recommendations in this document? To understand the importance of this question, it is necessary to understand ACOG.

ACOG is not a college in the sense of an institution of higher learning, nor is it a scientific body, it is a "professional organization" that in reality is one kind of trade union. Like every trade union, ACOG has two goals—promote the interests of its members, and promote a better product (in this case, well-being of women). But if there is conflict between these two goals, the interests of obstetricians come first.

Proof that ACOG puts members’ interests first? In September 1998 ACOG published Committee Opinion # 207, "Liability Implications of Recording Procedures or Treatments," which includes the statement: "Recording solely for the purpose of patient memorabilia or marketing is not without liability. . . . The Committee strongly discourages any recording of medical and surgical procedures for patient memorabilia." In other words, ACOG recommends that doctors and hospitals refuse permission for women and families to make a videotape of their baby’s hospital birth. Fear of litigation against their members has higher priority for ACOG than women’s rights and family values—the need of the family to record one of the most important events in their lives. This is why ACOG recommendations cannot always be considered the gospel and the recommendations in this document are suspect. Should the United Auto Workers have the final say on standards of auto safety?

2.3 Impact of recommendations goes far beyond obstetric practice.

The first Level C recommendation, "VBAC should be attempted in institutions equipped to respond to emergencies with physicians immediately available to provide emergency care," has a huge impact on the system of maternity care in the United States. If this policy is followed, it drastically reduces or eliminates several options available to women with previous cesarean section, including having their birth at home or in a freestanding birth center or in a small community hospital. Because of all the unnecessary cesarean section in the past, American women with a scarred uterus are a significant minority of pregnant women—in the neighborhood of 15 percent.

If the nearest large hospital is at some distance, it makes a family-centered birth difficult or impossible and is likely to eliminate continuity of care throughout pregnancy and birth. Scientific data show such continuity of care significantly improves birth outcomes.

In addition to this impact on women and families and birth outcomes, this recommendation also has a major impact on community-based midwives, family physicians, birth centers and small hospitals.

3. Conclusions

Two of the eight recommendations are without scientific justification. One recommendation—VBAC only with surgeons right there—has a tragic, unnecessary, negative impact on maternity care and maternity care providers in the United States. The second—oxytocin and prostaglandin induction is OK during VBAC—is quite likely dangerous for birthing women

We see a double standard in ACOG recommendations, depending on whether or not they are doctor friendly (are of benefit to ACOG members). If a recommendation has no evidence to support it but is of benefit to obstetricians—VBAC only in hospitals with surgeons standing by—ACOG will make the recommendation. Other solutions to the same problem that are not obstetrician friendly—facilitation of communication, collaboration and transport between primary and tertiary birth care—will not be recommended even though they would benefit many women. As a second example of this double standard, if there is inconsistent evidence on an issue—oxytocin or prostaglandin induction for VBAC—ACOG will make the doctor-friendly recommendation, approving its use even though the data show a trend that is likely to put some women at risk. ACOG recommendations consistently put the needs of ACOG members before the needs of women unless there is overwhelming evidence that may finally force an ACOG recommendation which is not doctor friendly—no Cytotec induction for VBAC.

We also see little evidence that ACOG is trying to find out why there is increasing uterine rupture so that it can be prevented in the future. ACOG should be making every effort to promote research on the management of those labors the past 10 years that ended with uterine ruptures, including: percent of ruptures associated with VBAC; percent of ruptures associated with the use of Cytotec, oxytocin or prostaglandin gels. There is also an urgent need for far more research on the relationship between characteristics of places of birth (home, birth centers, community hospitals, regional hospitals, tertiary care hospitals) and uterine rupture. Armed with this kind of data, ACOG could make evidence-based recommendations, be they doctor friendly or not.

ACOG’s primary allegiance to the needs of its members over the needs of women and families requires their recommendations to be suspect unless confirmed by overwhelming scientific evidence. As ACOG recommendations come from a single-specialty organization, they always must be carefully evaluated as to bias and should never be the sole basis, nor even the most important justification, for maternity care policy in the United States.

Marsden Wagner, MD, is a neonatologist and perinatal epidemiologist. He was responsible for maternal and child health in the European Regional Office of the World Health Organization for 14 years. Now living in Washington, D.C., he travels the world talking about appropriate uses of technology in birth and utilizing midwives for the best outcome.


  1. Editorial. (2000). Development of practice guidelines. The Lancet 355: 82-83.
  2. McMahon, M. (1996). Comparison of a trial of labor with an elective second cesarean section. New Eng J Med 335 (10): 689-695.
  3. Chauhan, S., et al. (1997). J Reprod Med 42 (6): 347-352.
  4. Plaut, M., et al. (1999). Uterine rupture associated with the use of misoprostol (Cytotec) in the gravid patient with a previous cesarean section. Amer J Obstet Gynecol 180 (6): 1535-1542.
  5. Rosen, M., et al. (1991). Vaginal birth after cesarean: a meta-analysis of morbidity and mortality. Obstet Gynecol 77 (3): 465-470.
Copyright © 1997-2006 Mother Care. All photos © Mother Care & Terri McKinney Photography. All rights reserved. Revised: Wednesday, January 04, 2006.