NIH Cesarean Conference: Interpreting Meeting and Media Reports (Updated, 10/2006)

NOTE: This Alert has been updated to address the Final NIH Panel Statement, which is posted on the NIH website and published in the June 2006 issue of Obstetrics & Gynecology, and to reflect Listening to Mothers II survey results reported in October 2006.


On March 27-29, 2006, the National Institutes of Health (NIH) held a state-of-the-science conference entitled "Cesarean Delivery on Maternal Request." At the end of the meeting, a multi-disciplinary panel issued a draft statement, also called "Cesarean Delivery on Maternal Request," about a commissioned evidence report and other results presented and discussed at the conference. Childbirth Connection leaders followed this process since the planning stage of the report. Childbirth Connection's Executive Director Maureen P. Corry and Director of Programs Carol Sakala, as well as several board members, attended the NIH meeting. The Panel's final statement was published two months after the conference.

This Alert is provided to help women, health professionals and others interpret the evidence report, panel statement and media reporting about them.

The panel's statement and commissioned evidence report failed to report a single study with any data about
  • the extent to which U.S. women are initiating and planning an initial or "primary" cesarean with the understanding there is no medical reason
  • the panel's assertion that "limited evidence suggests that cesarean delivery on maternal request is increasing"
  • the short- and long-term effects on mothers and babies of such a strategy in comparison with planned vaginal birth
  • the many well-established ways for women and providers to increase the likelihood of safe, spontaneous vaginal birth, in contrast to present record-level trends for surgical birth.

Deeply flawed analyses used proxy studies rated as "weak" for nearly all reported outcomes. Flaws included: failure to search for and include many relevant outcomes, including numerous studies that were too small to measure specific outcomes, including numerous studies with no vaginal birth comparison groups, and including numerous studies focusing on the unique question of mode of birth for women with breech presentation (see question 2, below). These poor-quality analyses should not be used to guide policy, practice, education and research.

For help with informed decision making, Childbirth Connection refers women and health professionals to our many resources relating to cesarean section, vaginal birth and choices for women with a previous cesarean, accessible through Related Links at the end of this page.

Emphasis on "maternal request" is generating confusion in the media, within the general public and among health professionals and pregnant women. The only national data collected from women themselves has found that this was a highly infrequent occurrence among women who gave birth in the U.S. in 2005 (details below). Continuing misplaced focus on "maternal request" cesareans draws attention from the legal, clinical, financial and social factors that continue to drive the U.S. cesarean rate to unprecedented heights in an overwhelmingly healthy population.

What is the background to the 2006 State-of-the Science Conference Statement?

An evidence report commissioned by the U.S. Agency for Healthcare Research and Quality was charged with answering four key questions:
  • Key question 1: What are trends in the use of cesarean section, including the extent to which women in the U.S. are themselves initiating and planning primary (initial) cesareans while understanding that there is no medical need?
  • Key question 2: What are short- and long-term benefits and harms to mothers and babies of "cesarean delivery on maternal request" when compared with planned vaginal birth?
  • Key question 3: What factors influence benefits and harms of planned cesarean and planned vaginal birth?
  • Key question 4: What future research is needed to help make appropriate decisions about planned cesarean section by choice with no medical reason versus planned vaginal birth?

Results of the evidence report and invited papers on these and related topics were presented and discussed at a meeting in Bethesda, Maryland, on March 27-29, 2006. The panel developed a draft statement from these deliberations and issued a final statement two months after the conference.

Key Question 1: What is the extent of "cesarean delivery on maternal request" in the United States?

"Cesarean Delivery on Maternal Request" was defined as a planned first or "primary" cesarean initiated by the mother with the understanding that there is no medical need. The evidence report and conference statement did not produce a single study that clarified the extent or trends of maternal request cesarean.

Several recent studies have found that there is no clear medical reason in birth certificate and hospital discharge records for a growing number of initial ("primary") cesareans. Although these data sources provide no information about women's preferences and decision-making processes, some have inappropriately claimed that all such surgeries were initiated by mothers. Lead authors of two such studies stood before the panel to underscore that their work cannot be used to provide any data about "cesarean delivery on maternal request" and to plead for proper use of such studies. Despite this clarification and request, and the evidence report conclusion that "we identified no recent data regarding the rate of [cesarean delivery on maternal request]," the panel statement used figures from such "no indicated risk" cesareans to estimate the extent of maternal request cesareans as from 4 to 18% of all cesareans.

By contrast, the first national data about maternal request cesareans from mothers themselves were available and released just prior to the meeting. The data were gathered through the second national Listening to Mothers� survey, conducted by Harris Interactive� in January-February 2006 among U.S. women who gave birth in 2005. The survey is a project of Childbirth Connection, in collaboration with Lamaze International. Of the 1573 women surveyed, just one mother said that she had had a planned initial cesarean by her own choice with the understanding that there was no medical reason.

Thus, working with a leading polling firm to ask mothers themselves results in an estimate of just 0.2% of all cesareans (1/480) by maternal request in 2005, and suggests that the panel's reported estimate of 4 to 18% of all cesareans grossly overestimates this phenomenon.

Two presenters at the NIH conference reported on the timely Listening to Mothers results:  Dr. Fay Menacker of the National Center for Health Statistics in the context of overall trends, and Susan Dentzer, leader of the Health Unit of the Public Broadcasting Service's NewsHour with Jim Lehrer, reporting on consumer perspectives. Although the NIH panel statement referenced much evidence report and conference data characterized as "proxy," "weak" and of "little confidence," they did not reference the only available national data on "maternal request" cesareans according to the NIH definition, which were with reference to the most recent calendar year and collected by a respected survey research firm.

Other Listening to Mothers results also shed light on cesarean trends:
  • survey results suggest that the overall rate continues to steadily rise: whereas the provisional national 2004 cesarean rate was 29.1%, 31.5% of survey participants had cesareans in 2005
  • 9% of the mothers reported feeling pressure from a health professional to have a cesarean, in contrast to 0.06% (1/1573) of all participants who had a "maternal request" cesarean (planned, primary, at mother's request, understanding that there was no medical reason)
  • fully 25% of mothers who had a cesarean reported feeling pressure from a health  professional to  have this procedure, in contrast to just 2% of mothers who had a vaginal birth
  • 42% of participants believed that the current malpractice environment leads maternity care providers to perform cesareans that are not really necessary to avoid being sued
  • 81% felt that is necessary to know every complication associated with cesareans before consenting to have one, yet most were unable to correctly recognize established complications of cesareans
  • 85% felt that a woman with a previous cesarean should be able to have a vaginal birth after cesarean (VBAC) if she so desired, but just 11% of those who had had a previous cesarean did have a VBAC, and most (57%) who did not have a VBAC were denied this option, primarily because caregivers (47%) or hospitals (26%) were unwilling.

Please see the first three links at the top of this page for more details about survey participants' views and experiences with cesareans and about survey methodology.

Key Question 2: What are short-term and long-term benefits and harms to mother and baby associated with cesarean delivery by request versus planned vaginal birth?

Authors of the commissioned evidence report and of other presentations at the meeting did not identify a single study that compared effects of "cesarean delivery on maternal request" with effects of planned vaginal birth.

Authors of the commissioned evidence report used a small assortment of "proxy" studies to answer the second key question. Studies included in the evidence report had very serious limitations and formed the basis for the panel's statement about this question.

Some of the more serious limitations of included studies are:
  • literature searches for the evidence report failed to specify most outcomes of interest (including breastfeeding, mother-infant behaviors and attachment, asthma in babies, placental abruption, ectopic pregnancy, infertility, and numerous fetal/newborn outcomes in future pregnancies); therefore many reported studies that met authors' criteria were undoubtedly overlooked
  • no data were provided about many important outcomes, including ectopic pregnancy, placental abruption, placenta accreta, chronic pelvic pain, and uterine rupture in future pregnancies
  • adhesions are a common, widely recognized consequence of surgery and are associated with a broad range of future reproductive and non-reproductive problems, yet the word is mentioned just once in the evidence report, with reference to needed research
  • a notable proportion of included studies had no vaginal birth comparison group, thus did not meet established criteria and should not have been included
  • a notable proportion of included studies involved only mothers with breech presentation, which in planned vaginal birth groups involves higher rates of injurious interventions, e.g., episiotomy, "assisted" delivery with vacuum extraction or forceps, and unplanned cesarean; for this reason, it was inappropriate to include these studies
  • many included studies had too few participants to draw any conclusion about less common outcomes of interest (with great vulnerability for "type II" errors: reporting "no difference" when outcomes are in fact more likely with one mode of birth than another); they should not have been included in analysis about less common outcomes.

Due to these and other limitations, many studies should not have been included in evidence report analyses to measure specific outcomes, and other more appropriate studies may have been overlooked. The authors appropriately label evidence for most of the reported outcomes as "weak." Reported results are of no value as guidance for policy, practice, education and research.

Key Question 3: What factors influence benefits and harms?

The evidence report concluded that evidence about "effect modifiers" that alter the impact of planned vaginal and planned cesarean birth routes is "sparse," and it failed to identify any of the many well-established clinical practices that influence the outcome of planned vaginal birth.

Established factors that increase the likelihood of attaining a vaginal birth and avoiding harms of assisted delivery (with vacuum extraction and forceps) or unplanned cesarean include:
  • choosing a caregiver and a place of birth with a conservative practice style with respect to avoidable interventions
  • receiving continuous labor support from a caregiver who comes into the hospital for this purpose (in the current Cochrane review, this is associated with 26% less likelihood of cesarean and 41% less likelihood of assisted delivery)
  • working with caregivers to delay hospital admission until active labor
  • avoiding epidural analgesia
  • avoiding electronic fetal monitoring
  • avoiding early intentional breaking of membranes
  • avoiding arbitrary time limits after the cervix is fully dilated
  • pushing in an upright or side-lying position
  • providing access to vaginal birth after cesarean (VBAC) for women with a previous cesarean
  • offering hands-on-belly maneuvers (external version) to turn breech presentation babies to a head-first position
  • offering intensive counseling for women with an extreme fear of vaginal birth.

Established factors that increase the safety of vaginal birth include:
  • avoiding episiotomy
  • avoiding back-lying pushing positions
  • avoiding assisted delivery when possible, and applying skillful judicious assisted delivery when needed
  • avoiding forceful staff-directed pushing and relying instead on the mother's own pushing reflexes
  • avoiding pressure by a member of the staff on the mother's abdomen (fundal pressure)
  • encouraging a regimen of pelvic floor muscle exercises preventively in pregnancy and to promote recovery after birth.

Implementing these measures broadly would improve the quality of maternity care and the outcomes of planned vaginal birth. All mothers should have access to safest vaginal birth practices. We should not ask mothers to choose between vaginal birth with avoidable harms and cesarean section.

Key Question 4: What research is needed to inform decisions about "maternal request" cesarean versus planned vaginal birth?

Because "cesarean delivery on maternal request" applies to an infinitesimal segment of the childbearing population (an estimated 2606 women out of about 4 million who gave birth in the U.S. in 2005), it is a low priority to devote further finite resources to this population. However, due to the increasingly casual use of cesarean section, including with no apparent reason at all, it is important to ensure that all stakeholders have access to valid information about relative harms of planned vaginal birth and planned cesarean in healthy mothers with no or a questionable medical need.

The panel recommends surveys of women, providers, insurers and facilities to understand the extent of maternal request cesarean. This information should be derived from women themselves, and the recent Listening to Mothers survey offers a current national measure: 1 mother in 1573 surveyed (0.06%) chose a "maternal request" cesarean by NIH definition. There will be value in monitoring trends when this survey is repeated in the future.

The panel recommends a billable CPT code (Current Procedural Terminology) for cesarean delivery on maternal request. Such coding would not be a trustworthy measure of mothers' intent and decision-making processes, and it would create a convenient mechanism for billing and reimbursement for elective surgery in the absence of evidence to support such a practice.

The panel recommends research devoted to strategies to increase the likelihood of "successful" vaginal birth. There is a critical need to implement the established factors listed above under Key Question 3 that influence the likelihood of achieving vaginal birth and of avoiding harms of numerous obstetric practices. Translational research to improve practice is a priority.

Similarly, the panel recommends further identification of factors that can improve the management of labor. This is an essential research priority to ensure that all childbearing women have access to safest vaginal birth practice.

The panel recommends large prospective cohort studies with long-term follow-up of mothers and children. Such studies offer important opportunities to measure outcomes of planned vaginal birth and planned cesarean birth. Because outcomes of interest may play out over many years with delay in needed results, great expense for follow-up, and inevitable failure to follow-up many initial participants, retrospective cohort and case-control study designs must complement prospective cohort studies. This is encompassed in the panel's recommendation to consider large databases and case-control studies for some outcomes.

Although the panel recommends exploring randomized controlled trials, most outcomes of interest could not be measured well or at all through such methodology. This is a poor research design for mode of birth due to the:
  • unwillingness of many women to enroll and unrepresentativeness of those who would agree to be randomized to vaginal or cesarean birth
  • ethical concerns about assigning healthy mothers and babies to major elective surgery
  • very large numbers needed to measure most outcomes and compounding of this problem by the many women who would be assigned to vaginal birth yet have unplanned cesareans in the present environment
  • expense and challenge of following up this large group during the months, years and decades that effects may come into play
  • fact that effects of ill-considered obstetric practices such as episiotomy and non-physiologic pushing practices would be attributed to "vaginal birth" (see Key Question 3, above)
  • need to improve current standards for measuring pelvic floor outcomes before confident results can be obtained for those outcomes.

By definition, a randomized controlled trial cannot measure "maternal request" cesarean, as women who enroll do not have a strong preference for mode of birth and agree to join a study group by chance.

The panel highlighted the importance of better understanding about effects of mode of birth on maternal death, placental abnormalities in future pregnancies, pelvic floor outcomes, psychological outcomes, neonatal outcomes and cost implications, all of which are important matters for informed decision making.

How sound are the conclusions drawn by the panel?

Panel conclusion 1: "The incidence of cesarean delivery without medical or obstetric indications is increasing in the United States, and a component of this is cesarean delivery on maternal request. Given the tools available, the magnitude of this component is difficult to quantify."

Comment: The rate of cesarean delivery with no apparent medical indication is on the rise in the U.S., but the only evidence from mothers themselves -- a national survey of childbirth in the U.S. in 2005 conducted by a leading survey research firm -- found that the component attributable to "maternal request" is infinitesimal. We must look to other factors to explain the inexorably rising national cesarean rate.

Panel conclusion 2: "There is insufficient evidence to evaluate fully the benefits and risks of cesarean delivery on maternal request as compared to planned vaginal delivery, and more research is needed."

Comment: Although no research expressly measuring this comparison was identified, existing evidence suggests that maternal request cesarean is a minor phenomenon and not a continuing priority for research and resource allocation. Focus should be directed to provider and health system factors contributing to changing standards for use of cesarean section.

Panel conclusion 3: "Until quality evidence becomes available, any decision to perform a cesarean delivery on maternal request should be carefully individualized and consistent with ethical principles."

Comment: Without clear justification based on confident results about this comparison regarding the full range of outcomes of interest, extreme caution is warranted when considering elective major abdominal surgery in healthy mothers and babies.

Panel conclusion 4: "Given that the risks of placenta previa and accreta rise with each cesarean delivery, cesarean delivery on maternal request is not recommended for women desiring several children."

Comment: As risks increase with every cesarean birth, most women desire and have more than one child, and many women lack access to vaginal birth after cesarean, primary cesareans should be avoided whenever possible. The 2002 U.S. National Survey of Family Growth clarifies that about 80% of women giving birth for the first time will go on to have 1, 2, 3 or more additional births, and a woman's future reproductive path cannot be known at the first birth. The number of children ever born to women in the 40-44 age range suggests that cesarean section should be avoided whenever possible:

 4 or more15%

Panel conclusion 5: "Cesarean delivery on maternal request should not be performed prior to 39 weeks or without verification of lung maturity, because of the significant danger of neonatal respiratory complications."

Comment: There is no fail-safe way to date gestational age in the absence of firm knowledge about conception from mothers. A study in the evidence report by John J. Morrison and colleagues found that babies born through week 39 have increased risk for overall respiratory morbidity in comparison with babies born beyond that time. And a presentation at the conference by Dr. Lucky Jain identified increased problems with newborn transition to life outside the womb with earlier birth and when babies do not experience the transition at the end of pregnancy just before labor begins and labor itself. Elective delivery in healthy mothers and babies before the onset of physiologic labor is a questionable practice.

Panel conclusion 6: "Maternal request for cesarean delivery should not be motivated by unavailability of effective pain management. Efforts must be made to assure availability of pain management services for all women."

Comment: Epidural and/or spinal/epidural analgesia is widely available and used in U.S. hospitals in women giving birth vaginally. Highly rated non-pharmacologic pain relief measures such as tubs, showers and birth balls, as well as continuous supportive care from labor companions, are often unavailable and/or underutilized in U.S. maternity care settings. To make informed decisions, women should have accurate knowledge from better-quality studies about the intensity and duration of postpartum pain, which are worse overall following cesarean delivery than vaginal birth, and in a notable proportion of cesarean mothers persist over a long period.

Panel conclusion 7: "NIH or another appropriate Federal agency should establish and maintain a Web site to provide up-to-date information on the benefits and risks of all modes of delivery."

Comment: Extensive evidence-based resources are available to provide information and guidance to women on this Childbirth Connection website, including in-depth help with cesarean versus vaginal birth, vaginal birth after cesarean (VBAC) versus repeat cesarean, and  preventing pelvic floor problems when giving birth. The cesarean Topic is based upon results of the only systematic review to compare the full range of harms of cesarean and vaginal birth, and the VBAC Topic is based upon that review as well as results of a previous federal evidence report on VBAC. Please see related links, below, for links to the many relevant resources on this website.

What are likely impacts of the exaggerated focus on maternal request cesareans that is furthered by the recent report, meeting and panel statement?

As reviewers failed to provide any evidence about incidence, trends or effects of maternal request cesareans (defined as planned primary cesareans initiated by mothers with the understanding that there was not medical need), it is misleading and confusing for NIH to retain the reference to "maternal request" in the report, meeting and panel statement.

A growing number of pregnant women, their caregivers and others are likely to believe that this is a notable and safe trend, and they may initiate maternal request cesarean as a result of this focus, thus further driving up cesarean rates. This would expose mothers and babies to unnecessary risks and drive up health care costs.

Continued misplaced focus on "maternal request" cesareans draws attention away from the legal, clinical, financial and social factors that are playing a major role in the escalating cesarean rate.

Although the verbal summary of the panel recommendations at the NIH conference underscored that it is not appropriate for health professionals to recommend cesarean section in the absence of a clear medical indication, this guidance was not included in the panel's written statement.

What factors contribute to the steadily rising cesarean rate in the U.S.?

Listening to Mothers survey results suggest that maternal request plays a negligible role in driving up cesarean rates. Factors that undoubtedly play a substantial role include:
  • casual attitudes about moving to a cesarean before there is truly a need and when other things might be done to avoid it
  • pressures on caregivers to practice "defensive medicine"
  • failure to offer vaginal birth after cesarean (VBAC) to women with a previous cesarean 
  • loss of skills or unwillingness to offer vaginal birth to women in some situations, such as with baby in a breech rather than head-first position and with twins
  • underuse of care that can enhance the natural progress of labor, such as continuous labor support and encouraging upright positions and mobility
  • side effects of medical interventions such as electronic fetal monitoring and labor induction in many circumstances, which increase risk for cesarean
  • the growing perception that a cesarean birth, and especially a planned cesarean, is "safe," without recognizing the many surgical risks.

Please download our fact sheet, which provides more detail about these factors.

With appropriate attention to these factors, and to the clinical factors that alter risk and are listed above under Key Question 3, the cesarean rate in the U.S. could safely plummet to a much lower rate than roughly one mother in three at the present time. Very low rates are possible for the majority of pregnant women who are healthy. In the largest prospective studies of North American women giving birth in birth centers (Rooks et al. New England Journal of Medicine 1989) and at home (Johnson and Daviss BMJ 2005), 4% cesarean rates were achieved with no indication of extra risk from the low rates or birth settings, when compared with low-risk women in hospitals. With the will and the skill, very low rates are possible and of benefit to most women.

Why were good science and best interests of mothers and babies thwarted in the evidence report and conference process?

Major resources were allocated to address the priority question of the safety of increasing casual and fully elective use of cesarean section in comparison with planned vaginal birth. Women, providers, policy makers and others would all benefit from high-quality guidance from the federal government on these matters. Unfortunately, the process and products are instead contributing to confusion and controversy.

The inside story of what happened remains to be told. Here are some important questions:
  • Why was the focus exclusively on mothers' demand for cesarean while failing to address provider and other health system pressures, such as liability fears and convenient scheduling and staffing?
  • Why was the "maternal request" frame retained in the face of no evidence report data whatsoever about incidence or trends in the U.S. and about effects in any context?
  • Why did the final panel statement fail to reference actual data on "cesarean delivery on maternal request" as defined by NIH, available from Listening to Mothers, a national U.S. survey of women who gave birth in 2005, which was conducted by the leading survey research firm Harris Interactive and was reported in two presentations at the NIH conference, while the panel statement provided much detail about other evidence described as "proxy," "weak," and of "little confidence"?
  • Why were biases in the entire process consistently in the direction of favoring cesarean delivery and discrediting vaginal birth, beginning with focused searching for just a few of the possible outcomes?
  • Why did responses to the third key question fail to identify any of the many well-established effect modifiers that promote safe, spontaneous vaginal birth and favorably alter the risk profile for mothers and babies in the present pregnancy and over the long term?
  • In planning, carrying out and reporting on the evidence report and conference, why were such valuable and important perspectives as family medicine, midwifery, nursing and consumer advocacy entirely excluded or given extremely marginal roles?

© 2016 Childbirth Connection. All rights reserved.

Childbirth Connection is a national not-for-profit organization founded in 1918 as Maternity Center Association. Our mission is to improve the quality of maternity care through research, education, advocacy and policy. Childbirth Connection promotes safe, effective and satisfying evidence-based maternity care and is a voice for the needs and interests of childbearing families.
Most recent page update: 6/2/2008