Side Effects of Epidurals: Research Data
Author’s note: All of the following data, unless otherwise cited, is collected and summarized from three recent reviews of this topic. This article is intended as a summary of available data. For more complete information, you are encouraged to seek out the original review articles, and to further trace them back to the original studies they review.
Please note the section at the end on “limitations of research data.” Also, note that the risks illustrations only include risks, and do not cover benefits or treatment implications.
See also the summary chart on the parent education part of the website, which covers ways to prevent and treat side effects.
Some terms used in the summaries below
· Relative risk (RR). The chance of this side effect occurring in women who have had in epidural in labor versus women who have not. For example, a RR of 2.0 means that women who have an epidural are two times more likely to have this side effect.
· Randomized, controlled trials (RCT). Participants in a study (laboring women) are randomly assigned to either the epidural group, or the no epidural group. The no epidural group may receive no medication or may receive another treatment such as parenteral opioids (e.g. IV narcotics). After treatment, the two groups are compared, differences in outcomes are assumed to be due to epidural use.
o Intent to treat: analysis is based on what group the participants were originally assigned to, no matter what medication they actually ended up receiving during labor.
o Crossover. With epidural studies included in Lieberman’s review, of those assigned to “epidural groups”, 2 - 35% of the members never actually received an epidural. Of women assigned to “no epidural groups”, as many as 56% of the participants chose to get an epidural. So, intent to treat analyses may not really reflect the impact of epidural.
· Observational studies: After the participants have given birth, studies are done which compare the side effects experienced by women who had epidurals with women who did not, and conclusions are drawn about what effect the epidural had on these women.
· Meta-analysis. Leighton combined results from 14 studies to give a broad summary of the likelihood of various side effects.
Effects on the Laboring Mother
· Pain Relief:
o Trials evaluated pain relief using a visual analog scale: patients are asked to mark on a scale how strong their pain is, from 1 (no distress) to 100 (unbearable distress). Meta-analysis indicates that with epidural, pain ratings were an average of 40 points lower during first-stage, and 29 points lower during second stage. (Leighton)
o Philipsen and Jensen found that 59% of the women in the epidural group reported a “painless” labor and delivery
· Patient satisfaction:
o Philipsen and Jensen found that 73% of women who used epidural would choose it for subsequent labors.
o Lower satisfaction for those who had planned natural childbirth: In Kannan’s study of 47 women who had planned not to use pain medication during childbirth, 23 women did not use medication, and 24 chose epidural. The women who requested epidural reported significantly lower pain scores. However, 88% of them reported being less satisfied with their birth experience than those who did not use epidural, despite lower pain intensity.
· Incomplete pain relief: “In spite of large doses epidural block may fail to provide adequate analgesia in up to 25% of patients due to difficulty in blocking sacral roots.” (Eldor, J. Combined Spinal-Epidural Anesthesia? On CSEN, the Global Regional Anesthesia website. http://www.csen.com/anesthesia/book/#ch11)
· Hypotension (low blood pressure). 16 studies: 0 – 50%. (Mayberry) Meta-analysis: RR: 74.2 (Leighton*)
o Risks: Can cause decreased oxygen flow to fetus, decreased fetal heart rate.
o Benefits: This side effect is generally considered a risk of epidural, but it may be a benefit in certain cases of maternal hypertension.
· Fever > 38°C or 100.4 °F.
o 4 RCT’s, 6 observational: Without epidural, 0 – 5%. With: 4 – 24%. RR: 1.5 to 70.8 (Table VIII in Lieberman) Meta-analysis: RR 5.6 (Leighton*)
o The rate of fever increased with longer labors, from 5% with labor < 3 h to 28% with labor > 6 h. (Gonen) Typical increase of .07 degrees C per hour of epidural. (Vinson)
o Lieberman et al and Gonen et al found that more than 95% of fever in their term populations occurred in women who had received epidural. Epidural-related fever is generally believed to result from thermoregulatory alterations rather than infection. (Lieberman)
o Secondary effects / risks of fever (from Lieberman):
§ Increased risk of instrumental / c-s: One study showed women with temps > 99.5°F were 3 times as likely to have a c-section (25% vs. 7%) and 3 times more likely to have instrumental delivery (25% vs. 9%)
§ Mom assumed to have infection, treated with antibiotics: 3 times more likely with epidural (20% vs. 6%)
§ Neonatal outcomes: Infants of women with fever were 3 times more likely to have 1 minute Apgars <7, and 10 times more likely to be hypotonic after delivery; 4 times more likely to require bag and mask resuscitation, and 6 times more likely to be given oxygen in nursery.
· Impaired motor ability. 8 studies used the Bromage scale to assess leg strength or the rectus abdominus muscle test to assess ability to sit up. The overall incidence of no motor block or minimal motor block was high, 76 – 100%. This indicates that women would be able to move around or walk with an epidural; however, in Olofsson, considerably less than one third of women chose to be out of bed, citing fatigue or a feeling of insecurity. (Mayberry)
· Inability to Urinate. 3 studies: 0 – 68% (Mayberry) Reason for concern: A full bladder may impede uterine contractility, may cause bladder trauma, or a lack of postpartum bladder tone. Treatment: Catheterization.
· Nausea and vomiting. 7 studies examined nausea: 0 – 30%, average is 7.3%. 5 studies examined vomiting: 0-13%, average is 4.6% (Mayberry) Meta-analysis: 1.46 – slightly more likely with epidural than without. (Leighton*)
· Shivering. 2 studies examined: only one case was documented in each study. (Mayberry)
· Pruritus. (Itching.) 17 studies examined this: When 1 or more opioids was included in the medication given, incidence ranged from 8 – 100%, with an average of 62%. When no opioid was given, incidence was 0-4%. Most cases appeared to be mild, as very few mothers requested treatment for itching. (Mayberry)
· Sedation / Drowsiness. 5 studies: 1 – 56%, average 21%. (Mayberry)
· Perineal lacerations (3rd and 4th degree). 6 studies: Without epidural – 3-34%. With epidural – 5 to 37%. Relative risks: 1.0 – 2.7. (Lieberman) Severe lacerations are more likely with instrumental delivery, and since data support an increased risk of instrumental delivery with epidural, these increases are likely closely related.
Effects on the Fetus during Labor
· Fetal Malposition (occiput posterior, transverse). 3 RCT’s: 1) 19% with epidural vs. 4% without, 2) 22% vs. 18% 3) 16% vs. 14%. However, trials 2 and 3 had high crossover rates, making those numbers harder to interpret. 2 observationals. Overall, relative risk ranges from .8 to 4.2. (Lieberman, table IX.) It’s unclear whether epidurals cause malposition, or whether women with a malpositioned baby are more likely to choose epidural because of increased pain.
· Fetal heart rate abnormalities. Meta-analysis of 4 studies: no significant difference with epidural. (Leighton*)
· Tachycardia / Fetal heart rate greater than 160 bpm. Rojansky et al: At the end of first stage: 7% with epidural vs. 2% without. In second stage: 16% with epidural, 13% without. Mayer et al: 6% vs 0%. (Lieberman) Likely due to increased risk of maternal fever, as FHR is highly correlated with maternal temperature.
· Bradycardia / Fetal heart rate <100: after 11% of initial or repeat injections of anesthetic into epidural space. (Stavrou 1990, cited in Thorp)
· Fetal heart rate: Late or variable decelerations Rojansky: end of first stage: 47% vs. 13%, relative risk 3.3. Second stage: 66 vs. 41%, relative risk 1.6. (Lieberman)
· Meconium stained amniotic fluid. None of 5 studies found any difference in meconium staining between epidural and non-epidural groups (Lieberman)
Effects on the Progress of Labor
· Increased use of oxytocin to stimulate contractions and speed labor. 5 randomized controlled trials representing 3679 women: For women receiving narcotic analgesia, oxytocin rates were 11 – 64%. For women receiving epidurals, oxytocin rates were 27 – 78%. Mayberry’s analysis indicates augmentation is 1.28 times more likely with epidural. (Mayberry) Meta-analysis: RR 2.8 (Leighton*)
· Decreased rate of dilation Thorp, et al. Before epidural, both randomized groups dilating at an average of .52 cm/hr. After analgesia, epidural group was 1.9 cm / hr versus 2.7 cm / hr without epidural. (cited in Lieberman)
· Diminished urge to push / decreased ability to actively push during 2nd stage. Mayberry describes this issue, but does not state the incidence (how commonly it occurs). For care recommendations, see below.
· Longer labor overall. 3 RCTs: .9, 1.5, and 2.4 hours longer with epidural. 7 Observational: 1.7 to 5.1 hours longer. (table VI and VII in Lieberman)Note: differences in length of labor between studies can depend on issues like: whether start time is based on admission to the hospital, frequency of contractions, or a particular cervical dilation.
· Longer first stage. 6 RCT’s: -0.4 to 2.6 hrs longer. 11 observational: .6 to 4.8 hrs longer. (table VI / VII in Lieberman) Meta-analysis: 26 minutes longer on average. (Leighton*)
· Longer second stage. 7 trials: all found longer second stage, with the difference ranging from to 61 minutes longer. (Lieberman) 18 observational: 0 to 55 minutes longer (table VII in Lieberman) Meta-analysis: 15 minutes longer on average. (Leighton*)
Effects on Mode of Delivery
· Reduced chance of spontaneous vaginal delivery (no cesarean or instrumental delivery needed) 6 of 9 studies indicate that less than 50% of women who received an epidural had a spontaneous vaginal delivery (Lieberman)
· Instrumental vaginal delivery (forceps and vacuum extractor).
o Rate without epidural ranges a great deal: 4% to 60% depending on the study. (Lieberman)
o 10 RCTs: Rates 7 – 80%with epidural. RR 1.1 – 5.3. (Crossover rates affect the accuracy of these results.) Observational studies: 26 out of 27 showed a statistically significant association between epidural and instrumental delivery. Rates: 6 – 76%. Relative risk ranged from 1.3 to 4.8. (Lieberman)
o Meta-analysis: RR of 2.08. (Leighton*)
o Observational studies limited to women with vaginal births indicate that 3-23% of women without epidural had instrumental deliveries, vs. 5 – 52% of women with epidural, leading to RR 1.7 – 14.0. (table V in Lieberman.)
· Cesarean section (for any reason) Meta-analysis indicated No significant difference between patients randomized to receive parenteral opioids and epidurals. (Leighton. See note above about concerns about the accuracy of this analysis.*)
· Cesarean for non-reassuring fetal status: 5 RCT’s and 7 observational studies: There is not a significantly higher rate for women with epidural versus for women without epidural (Lieberman)
· Cesarean for dystocia / failure to progress.
§ 10 RCT’s show relative risk ranging from .7 to 11.2 (c-section .7 times less likely to 11.2 times more likely with epidural anesthesia than without.) Note that crossover rates influenced several of these results.
§ 33 observational trials show RR’s from 0 to 6.5. (Of the 10 observations of low risk nulliparas in spontaneous labor, RR’s ranged from 1.6 – 6.5.) The general trend of studies indicates a potential increased risk of c-section; however, because epidurals are something that laboring women choose, it is not clear whether epidural itself increased their chance of c-section, or whether there was something inherent about these births that would have made c-section more likely even without epidural (e.g. the specific populations studied, the type of labors which lead mothers to choose epidural, the specific setting for care, or the management of the labors) (Lieberman)
o More details on studies:
§ Highest relative risk found in RCT: Thorp’s RCT of 93 women (where only one woman in the no epidural group received an epidural) found an 11-fold increase in cesarean rate amongst epidural group, and stopped the trial early because of this finding.
Lowest relative risk
found in RCT: .7
§ A study which controls for some additional factors: Lieberman et al (1996) evaluated 1733 term, low-risk women with spontaneous onset of labor. The overall cesarean delivery rate was 4% for women without epidural, and 17% with epidural. They divided them into five groups based on characteristics which predict the choice to have an epidural: such as dilation at admission, station of fetal head at admission, rate of dilation, gestational age, infant birth weight, mother’s race, weight, and height. In all 5 groups, the cesarean delivery rate was higher among women who received an epidural, suggesting that the association was present regardless of the characteristics of a woman’s labor. In a logistic regression analysis controlling for these factors, epidural was associated with a 3.7 fold increase in the rate of cesarean. (cited in Lieberman 2002)
Effects on Maternal Postpartum Recovery
· Postpartum hemorrhage and retained placenta. At one hospital, no difference. At a second hospital, risk of postpartum hemorrhage was 15% with epidural and 3% without. Another study found the rate of postpartum hemorrhage was twice as high among women who received epidurals (10% vs. 5%) Amongst women with retained placenta, 51% had used epidural. (Lieberman)
· Postpartum urinary retention. 2 studies of symptomatic urinary retention requiring treatment: 4% with epidural, 1% without; 2.7% with, .1% without. 4 studies of asymptomatic (i.e. a high residual volume in the bladder after voiding): 2 studies found no association, 1 a RR of 1.8, one RR of 4.7 (Lieberman)
· Urinary incontinence. Viktrup and Lose: Immediate postpartum - 27% with epidural vs. 13% without. At 3 months, 16% vs. 4%; at one year, 7% vs. 3%. (cited in Lieberman)
· Backache (long-term) 5 studies: current data do not support an association with epidural (Lieberman) Macarthur et al found more back pain with epidural on day 1 postpartum (53% vs. 43%), but no difference on day 7, week 6, or at 1 year. (cited in Leighton)
· Postpartum Headache. 2 studies indicate that inadvertent dural punctures occur in 1.6 – 1.8% of women; 23% of these women had new onset of chronic symptoms including headache, migraine, or neck ache, starting within 3 months after childbirth, and lasting from nine weeks to over 8 years. (Thorp) Mayberry states that these headaches can be severe enough to temporarily interfere with women’s normal activities, including infant care for up to 48 hours.
Effects for the Newborn Baby
Apgar scores <7 at
1 minute, 5 minutes.
· Low Umbilical cord pH. 6 RCT’s and 4 observationals: No study found a significant difference with or without epidural. (Lieberman) Meta-analysis of 5 studies shows no significant difference (Leighton*)
· Neonatal evaluation for bacterial infection and/or neonatal antibiotic treatment. Lieberman: Neonates whose mothers’ had epidurals were more likely to be evaluated for sepsis (34% vs. 9.8%) and to be treated with antibiotics because of suspicion of sepsis (15.4% vs. 3.8%) However, the rate of sepsis was low in both groups (.3% epidural, .2% non-epidural) Philip: sepsis evaluation 25% with epidural, 16% without; antibiotics 19% with, 11% without.
· Hyperbilirubinemia (jaundice). 7 studies: All found 1.5 to 2.0 – fold increase in the rate for infants after epidurals. (Lieberman)
· Neonatal behavioral and neurologic outcomes. 11 studies examined:
o In the 6 studies that examined epidural vs. no/minimal medication, 3 found no significant differences. Lieberman et al found that infants in the epidural group were less responsive to the human voice, Murray et al found that epidural was associated with lower scores overall at day 1 with differences in motoric processes, response to stress and state control. Differences remained at day 5, but not at one month. Sepkowski et al found that epidural was associated with lower scores on orientation and motor clusters.
o In 6 studies that examined epidural vs. parenteral opioids: 3 found no significant differences. Wiener found that epidural group habituated to sound more quickly and had poorer muscle tone. In another study, Wiener found epidural group had decreased reflexes and poorer muscle tone. Kangas-Saarela found epidural infants habituated to sound and oriented to inanimate sound better. (Lieberman) Sample sizes were small.
· Effects on breastfeeding. 2 studies: Kiehl found women who had received epidural were less likely to be breastfeeding at 6 months (30% vs. 50%). Halpern found that the drugs used in labor did not predict difficulty in initiating breastfeeding or level of breastfeeding at 6 to 8 weeks postpartum. However, there are issues with both study designs which make these results difficult to interpret. (Lieberman)
Research Evaluations of Care Recommendations for Minimizing Side Effects of Epidurals
Coping with the Diminished Urge to Push: Directed pushing vs. Laboring Down. Common management is directed, strong, sustained pushing efforts with prolonged breath holding. “There is no research to support the necessity of this practice.” (Mayberry) This approach carries some risk for fetal compromise because of decreased oxygen, risk of increased maternal fatigue, and risk of short- and long-term pelvic damage.
Another option is ‘delayed pushing’ or ‘laboring down’. The woman simply allows the baby to descend on its own, not beginning active pushing until the baby’s head is crowning, and/or she feels the urge to push. Recent studies show no adverse fetal effects of delayed pushing for up to three hours, plus a reduced risk of instrumental or cesarean delivery. There may be an increase of fever with delayed pushing. (cited in Mayberry)
Timing of epidural / Delaying epidural administration to reduce side effects:
· Impact on risk of cesarean: All 9 studies examining the association of epidural timing with cesarean delivery among nulliparous women reported a higher rate of cesarean with earlier epidural. Most studies reported relative risks between 1.6 and 2.2, although not all the differences were statistically significant. Thorp examined the effect of labor characteristics on this: For slow dilators (< 1 cm/hr in early labor), epidural before 5 cm dilation was associated with a higher risk of cesarean delivery than late epidural (24% vs. 7%). However, for fast dilators (≥ 1 cm/hr in early labor), the timing of the epidural made no statistically significant difference in the rate of cesarean delivery (14% early epidural vs. 11% late epidural)
· Instrumental delivery: Data suggest that there’s a modest increase in the chance of instrumental vaginal delivery for early epidural.
· Progress of labor: Thorp found a longer length labor for slow dilators with early epidural rather than late, but no difference for women who dilated quickly in early labor. Robinson et al found no significant difference in the length of second stage for early and late epidurals, Sheiner reported shorter early labors among women who received early epidurals, but the difference was not statistically significant. The inconsistent findings may be caused by several confounding factors.
· Fetal malposition: Robinson found malpositioning amongst nulliparas to be 32% with early epidural and 12% with later epidurals, for multiparas 16% vs. 3%. Sheiner found no statistical difference between early and late epidural. (Lieberman)
· Fever. Because the risk of fever, and the severity of fever increase with duration of epidural, the less time spent with epidural, the lower chance for fever.
“Light” versus “standard” epidural: 4 RCT’s. No significant differences in the proportion of women with instrumental vaginal or cesarean delivery, or length of labor, or fetal outcome. First stage labor was consistently longer with “light” epidural. (Lieberman)
Intermittent versus continuous infusion. 1 observational study and 6 RCT’s. Little evidence for large differences in maternal or infant outcome based on the use intermittent or continuous infusion. (Lieberman)
Turning down the epidural in second stage: 6 RCT’s, showing various results (see Lieberman for full comments on issues of concern with each study). Chestnut et al found that when epidural anesthesia was stopped at 8 cm, there was no difference in rate of cesarean – 13% in each group; but there was a significantly higher rate of instrumental delivery when the epidural was continued (46% vs. 24%) and second stage was an average of 30 minutes longer (124 minutes vs. 94). There was no significant difference in fetal malposition, fetal pH or meconium staining.
Philipsen and Jensen’s trial included 111 women. In one group, epidural was discontinued after 8 cm dilation. There was a 60% higher rate of c-s among women receiving epidural (18% vs. 11%); this difference is not considered statistically significant because of the small sample size. (cited in Lieberman)
Cool cloths to reduce risk of fever: An anecdotal method for reducing risk of epidural fever is to use cold compresses on mom’s forehead or the back of her neck. I did not find research information on this. What I did find was: a) the recommendation that tepid water should be used rather than cold water, as cold can cause the blood vessels in the skin to constrict, making it more difficult for heat to escape from the body. b) information about studies of the efficacy of sponge bathing with tepid water to reduce childhood fevers. 2 research trials found a greater and more rapid fall in temperature after tepid sponging than after administration of temperature-lowering drugs. One study found that acetaminophen was more effective in rapidly lowering temperature. One found that sponging and medication together were no more effective than medication alone. Two studies showed that tepid sponging causes distress in children. However, laboring women tend to greatly appreciate it. A general conclusion could be: cool (not cold compresses) feel good for moms, and may carry an additional benefit of reducing risk of fever. (Mahar, et al Tepid sponging to reduce temperature in febrile children in a tropical climate. Clin Pediatr (Phila), 1994; 33(4):227-31. Aksoylar, et al Evaluation of sponging and antipyretic medication to reduce body temperature in febrile children. Acta Paediatra Jpn, 1997; 39(2): 215-7. Agbolosu, et al. Efficacy of tepid sponging versus paracetamol in reducing temperature in febrile children. Ann Trop Paediatr, 1997; 17(3):283-8. Newman, J. Evaluation of sponging to reduce body temperature in febrile children. Can Med Assoc J, 1985; 132(6):641-2.)
Major reviews examined:
Leighton BL, Halpern SH.. The effects of epidural anesthesia on labor, maternal, and neonatal outcomes: A systematic review. Am J Obstet Gynecol 2002; 186:S69-77. Reviewed RCT’s and prospective cohort studies in which epidural anesthesia was compared with parenteral opioids in labor. All studies were published in English between 1980 and 2001, enrolled only healthy women with uneventful pregnancies, and met additional criteria for quality. A total of 14 RCT that enrolled 4324 women met their inclusion criteria. A meta-analysis was done, combining the results of these trials to reach the conclusions presented.
Lieberman E, O’Donoghue C. Unintended effects of epidural anesthesia during labor: A systematic review. Am J Obstet Gynecol 2002; 186:S31-68. A total of 1900 articles were examined, and evaluated for inclusion in the review based on the authors’ criteria. They limited their review to original reports in English, in peer review journals since 1980; they included both randomized trials and observational studies; they excluded studies with no control group, studies that evaluate specific drug regimens, studies that examine epidurals for anesthesia during cesareans, studies conducted exclusively on high-risk populations, studies where population selection renders results uninformative, studies with analytic choices that make results impossible to interpret, and studies that examine outcomes only for the overall population of delivering women.
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. More than 700 publications were identified; they narrowed that down to 150 studies that addressed one or more of the common side effects and co-interventions, plus 75 articles addressing relevant clinical or nursing care information related to unintended effects of epidurals. They only included prospective, randomized, controlled trials published between 1990 and 2000. These studies were then further limited by pre-established criteria: evidence of little or no crossover effect, minimal loss of subjects after random allocation to comparison groups, and satisfactory description of the randomization procedures. In the final review, they included 19 studies, with a total sample size of 2708 women.
Gonen R, Korobochka R, Degani S, Gaitini L. Association between epidural anesthesia and intrapartum fever. Am J Perinatol 2000; 17: 127-30.
Kannan S, Jamison RN, Datta S. Maternal Satisfaction and pain control in women electing natural childbirth. Reg Anesth Pain Med 2001, 26: 468-72.
Lieberman E, Lang JM, Cohen A, D’Agostino R, Datta S, Frigoletto FD. Association of epidural anesthesia with cesarean delivery in nulliparas. Obstet Gynecol 1996; 88: 993-1000
Thorp JA, Breedlove, G. Epidural Analgesia in Labor: An Evaluation of the Risks and Benefits. Birth 23:2, 1996.
Thorp JA, Hu DH, Albin RM, McNitt J, Meyer BA, Cohen GR, et al. The effect of intrapartum epidural analgesia on nulliparous labor: a randomized, controlled, prospective trial. Am J Obstet Gynecol 1993; 851-8.
Some interesting trivia: “Women who choose epidural are more likely to be slightly shorter, to have larger infants, and to be further along in gestation… are admitted to the hospital earlier in labor, and dilate more slowly just after admission compared to women who do not go on to receive epidural. The rate of epidural use has been noted to decrease directly with greater cervical dilation at admission.” (Sources cited in Lieberman)
* Limitations of research data. In reviewing research, it’s important to be aware of the limitations. For example, if you wanted to find the increased risk of cesarean after epidural, and you examined only one article based on one trial, that article might show anything from a .8 relative risk (i.e. c-section was slightly less likely with epidural than without) to a 11.2 RR (c-section was 11.2 times more likely). Clearly, this discrepancy in results requires further examination of the researcher’s methods, sample populations, and so on.
Review articles can give a clearer picture by examining several studies, and commenting on strengths and weaknesses of each. However, even reviews can suffer from an incomplete examination of the issues. Leighton and Halpern, for example, are aware of crossover rates and include data on them; however, when they did their meta-analysis of all the results, they base it on “intent to treat” and don’t allow for the substantial effect crossover has on interpreting the meaning of these results.*
Lieberman and O’Donoghue do an excellent job of examining crossover in randomized, controlled studies (RCT’s). A good example of the impact of this is Clark et al. In their examination of the results by ‘intent to treat’ (examining the results based on what group the members were assigned to), you would read that there was no difference in the cesarean rate for women assigned to receive epidural versus opioid analgesia. However, a stunning 52% of the opioids group actually received epidurals. And, of the 17 reported cesareans for dystocia in the ‘opioid group’, 14 of those women had actually received an epidural. Thus, as Lieberman states, “When such a high proportion of subjects do not get the treatment to which they were assigned, the intention-to-treat analysis, though technically correct, is impossible to interpret.”
Another reason for the wide variation in research results (in RCT’s and in observational studies) is that birth is a complex process, and is influenced by many different factors which are difficult to control for or evaluate. For example, anyone who has worked with birthing women knows that their labor progress can be affected by psychological, social, and emotional factors. A choice of when to use pain medication, and quite possibly what side effects are experienced, can be influenced by such factors as different birthing environments, labor support from spouses and family members, interaction with medical staff, the mother’s personal history, etc. Therefore, all results merely indicate trends in what side effects are more likely with epidural than without, and indicate what treatments are more likely to be effective.