NEJM Study on Pain Medication and C-Section

Summary, Analysis, Concerns

Childbirth Connection visitors are hearing about a new study about the timing of labor pain medication. Many media reports are providing misleading coverage. Childbirth Connection has prepared the following information to help visitors interpret this study, which Cynthia Wong and colleagues published in the New England Journal of Medicine on February 17, 2005.

Please see Childbirth Connection's clear, simple advice for women about labor pain relief.

What did the February 2005 study do?

The focus of the new study was on how to use the intervention, rather than whether to use it. The researchers looked at two approaches to regional analgesia, which refers to giving pain medications within the spinal column. Their objective was to understand whether waiting until later in labor could offer benefits, including reduced likelihood of cesarean section.

Only healthy women participated in the study. Each woman who agreed to participate was assigned by chance to one of two study groups:
  • "early" regional group got one type of regional analgesia before their cervix had dilated to 4 centimeters, then later got another type of regional analgesia
  • "late" regional group got pain medication by injection into the muscle and by intravenous drip first, and regional analgesia later, generally after their cervix had dilated to 4 or more centimeters

The early regional group had two types of regional analgesia. Early in labor, women in this group had spinal analgesia, which is injected into the fluid that surrounds the spinal cord and given just once. When they asked for more pain relief, they had epidural analgesia. An epidural is when pain medication is delivered through a small tube (catheter) into the epidural space just inside the tough outer membrane covering the spinal cord. This multi-stage technique, which is not widely available in U.S. maternity settings, is sometimes called combined spinal-epidural analgesia.

The late regional group received a narcotic, hydromorphone, through an injection into the muscle and through an intravenous drip early in labor, followed later in labor with epidural analgesia, regional medication through a tube into the epidural space.

Did the study do a good job measuring differences between earlier and later regional analgesia?

Unfortunately, it is difficult to answer the main question, whether delaying regional analgesia reduced the likelihood of cesarean section, for two reasons.

First, other things could have influenced the results:
  • the late group received hydromorphone, a powerful drug that influenced just this group
  • over 90% of mothers in both groups received synthetic oxytocin, a drug that intensifies contractions of the uterus, beginning in most cases early in labor; this could reduce the use of cesarean section in both groups and lead to misleading results about effects of regional analgesia.

Second, the combined spinal-epidural technique is not widely used in the United States and differs in fundamental respects from epidural analgesia. We do not know if results of the combined technique apply to the more common use of epidural alone.

Have other studies found that delaying regional analgesia lowers risk for cesarean section?

Lieberman and O'Donoghue (2002) carried out a rigorous systematic review of the best available research to understand whether delaying use of epidural until later in labor reduced the likelihood of cesarean birth. They concluded that such a delay may lead to fewer cesareans, but that existing research could not give a clear answer.

Does the study show that epidurals are safe?

No. Women in both groups had epidural analgesia later in labor. The study did not compare groups that did and did not have this type of pain relief. Therefore, this study sheds no light on the safety or effectiveness of epidurals. It is wrong to conclude from this study that epidurals are "safe".

What is the best evidence about the safety of epidural analgesia?

There is ample rigorous research showing that epidurals have many adverse consequences, including systematic reviews from Childbirth Connection's "The Nature and Management of Labor Pain" project (published in American Journal of Obstetrics and Gynecology, May 2002). It is important that all pregnant women understand pros and cons of epidurals and other methods of labor pain relief.

In a systematic review of adverse effects of epidural analgesia, Lieberman and O'Donoghue (2002) found that epidurals increase:
  • risk of tachycardia (abnormally rapid heart beat) in the fetus
  • length of pushing phase of labor
  • use of vaccum extraction or forceps
  • risk of serious perineal tears into the anal muscle in mothers
  • risk of fever in mothers
  • evaluation of newborns for infection, and early separation of mothers and babies
  • provision of antibiotics to newborns.

This review also found that mothers with epidural were less likely to have a spontaneous birth (with neither cesarean nor use of vacuum extraction or forceps), and that their babies scored worse on Brazelton Neonatal Behavioral Assessment Scale.

A systematic review by Mayberry and colleagues (2002) looked at the impact of epidurals on the labor experience, and found that epidurals increase likelihood of:
  • trouble urinating
  • sedation
  • low blood pressure
  • immobility, even when encouraged to move about
  • itching.

Mayberry and colleagues found that use of epidural analgesia was associated with use of many other interventions to monitor, prevent or treat epidural side effects. This regional analgesia technique involved routine use of electronic fetal monitoring, intravenous drip, and frequent blood pressure monitoring, and increased the likelihood that mothers would have other labor interventions, such as synthetic oxytocin, bladder catheters, and drugs for low blood pressure (2002).

Does the February 2005 study show that epidurals do not increase risk for cesarean section?

No. Because women in both groups had epidurals, this study sheds no light on whether an epidural increases risk for cesarean section. Lieberman and O'Donogue's systematic review of randomized controlled trials and observational research looked at the best available studies for answering this question and concluded that available studies did not permit a clear answer (2002). For now, the best we can say is that having an epidural may increase risk for cesarean section.

What are some concerns about the quality of care provided in this study?

Childbirth Connection has concerns about the quality of care that participants in this study received. Due to these concerns, to difficulty understanding the impact of hydromorphone and synthetic oxytocin on study results, and to limited use of combined spinal-epidural technique, this study should not serve as a model for the care of healthy birthing women.

Nearly all of the healthy study participants (early group: 92%, late group: 95%) received synthetic oxytocin. For most, use of this drug began early in labor. Due to adverse effects associated with synthetic oxytocin (click "Full Prescribing Information"), its routine use in healthy women is not appropriate.

Despite this near-universal "stimulation" of labor and inclusion of only healthy women, over one-third in both groups experienced either surgical delivery cesarean section or use of vacuum extraction or forceps when giving birth.

The "late" group received hydromorphone (trade name Dilaudid�), a narcotic that is stronger, quicker-acting, and more sedating than morphine. The Food and Drug Administration "label" or guidelines for hydromorphone in the Physician's Desk Reference (PDR), states that it is contraindicated (not to be used) for labor and birth.

Drug-free help with pain relief (for example, continuous supportive care in labor and use of tubs, showers and "birth balls") was not offered to the "late" group, yet these options have good safety profiles and many who use them give high ratings for pain relief. Providing continuous labor support would reflect a serious commitment to limiting use of cesarean section in healthy women, as a systematic review on effects of continuous labor support has found that, across many higher-quality studies, the presence of a companion with an exclusive focus on labor support reduced risk of cesarean by 26%.

What are additional concerns about this study?

The report opens by describing current policy of the American College of Obstetricians and Gynecologists (ACOG) on timing of epidurals. ACOG recommends not providing epidurals until a woman's cervix has dilated to 4 or 5 centimeters, as earlier use of epidurals may increase risk of cesarean section. The authors discuss their concern that this policy may prevent women from access to effective pain relief in early labor, and hypothesize that initiating combined spinal-epidural in early labor would not increase risk of cesarean. Because the policy addresses epidural analgesia, the new study does not evaluate the wisdom of this policy.

The study was not seriously set up to measure differences between the groups. The bar for detecting differences was set so high as to ensure that the conclusion would be: no difference in cesarean rates between the 2 groups. The researchers enrolled just enough women in the study to be able to detect an improbable difference between groups of 50 percent or more in the rate of cesarean delivery. Such dramatic results are not achieved with a single intervention, especially one in which both groups received such similar treatment (e.g., epidurals, synthetic oxytocin). The study was not set up to detect differences in cesarean rates that might be less than 50%, which the authors trivialize as "small".

A casual glance at Tables 1 and 2 could lead people to assume that this study compared having regional analgesia to not having it. In these tables the study groups are labeled as regional versus injection/intravenous groups rather than early versus later regional analgesia. The article does not provide specific numbers that clarify that virtually all women in both groups had epidurals. Similarly, the authors note that although epidural has been associated with increased use of synthetic oxytocin, this study found no differences in groups; in fact, the study cannot shed light on whether epidural involved increased used of labor stimulation as both groups received epidurals.

Media coverage of the study has been confusing and could lead many to conclude erroneously that the study could in fact shed light on the safety of epidural and whether it is associated with cesareans. Some headlines erroneously state that the study shows that epidural is not associated with increased risk of cesarean and is a safe option for laboring women, without qualification.

The study was not endorsed and funded by an external group, such as the National Institutes of Health, but was supported by the department of anesthesiology with which most authors were affiliated.

The sweeping conclusion at the end of the discussion section goes far beyond what can be determined from the present study:

"In summary, the results of this randomized trial suggest that nulliparous women in spontaneous labor or with spontaneous rupture of membranes who request pain relief early in labor can receive neuraxial [regional] analgesia at this time without adverse consequences." (p. 665). This is misleading in three major respects:
  • The study was not designed to measure whether there are adverse effects with receiving regional anesthesia, which in fact have been well-documented.
  • Other treatment that study participants received makes it difficult to understand the impact of delaying regional analgesia.
  • It is wrong to apply results of this experience with early spinal injections to regional analgesia generally, including early use of epidurals.

Additional labor pain resources on Childbirth Connection's website:


Lieberman E, O'Donoghue C. Unintended effects of epidural analgesia during labor: a systematic review. Am J Obstet Gynecol 2002;186:S31-68.

Mayberry LJ, Clemmens D, De A. Epidural analgesia side effects, co-interventions, and care of women during childbirth: a systematic review. Am J Obstet Gynecol 2002;186:S81-93.

Wong CA, Scavone BM, Peaceman AM, McCarthy RJ, Sullivan JT, Diaz NT et al. The risk of cesarean delivery with neuraxial analgesia given early versus late in labor. N Engl J Med 2005;352:655-65.
Most recent page update: 6/30/2008

© 2016 National Partnership for Women & Families. All rights reserved.

Founded in 1918, Childbirth Connection has joined forces with and become a core program of the National Partnership for Women & Families. Together, these two women's health powerhouses are transforming maternity care in the United States.
News and Features
Featured Resource

Check out our resource, "Hormonal Physiology of Childbearing: Evidence and Implications for Women, Babies, and Maternity Care"
Read more

Special Announcement

Childbirth Connection has joined forces with and become a core program of the National Partnership for Women & Families.
Read more

Our Vision

We want all women and babies receive the best possible maternity care.
Play video

Get Involved

Read our 2020 Vision, Blueprint for Action, blog and more
Sign up for email updates
Find us on Facebook
Follow us on Twitter
Support us