PEDIATRICS Vol. 108 No. 5 November 2001, pp. 1099-1102

Maternal Epidural Use and Neonatal Sepsis Evaluation in Afebrile Mothers

Laura Goetzl, MD, MPH^*, Amy Cohen, BS, Fredric Frigoletto Jr, MD^§, Steven A. Ringer, MD, PhD, Janet M. Lang, PhD, ScD^¶, and Ellice Lieberman, MD, DrPH

From the ^* Department of Obstetrics and Gynecology, Baylor College of Medicine, Houston, Texas; and Department of Obstetrics and Gynecology,  Brigham and Women's Hospital and ^§ Massachusetts General Hospital, and  Joint Program in Neonatology (Brigham and Women's Hospital, Children's Hospital, Beth Israel Deaconess Medical Center), Harvard Medical School, and ^¶ Department of Epidemiology and Biostatistics, Boston University School of Public Health, Boston, Massachusetts.

	ABSTRACT

Top Abstract Methods Results Discussion References

Objective.  Epidural use has been associated with a higher rate of neonatal sepsis evaluation. Epidural-related fever explains some of the increase but not the excess of neonatal sepsis evaluations in afebrile women

Methods.  We studied 1109 women who had singleton term pregnancies and who presented in spontaneous labor and were afebrile during labor (<100.4°F). Neonatal sepsis evaluation generally was performed on the basis of the presence of 1 major or 2 minor criteria. Major criteria included rupture of membranes for >24 hours or sustained fetal heart rate of >160 beats per minute. Minor criteria included a maternal temperature of 99.6°F to 100.4°F, rupture of membranes for 12 to 24 hours, maternal admission white blood cell count of >15 000 cells/mL³, or an Apgar score of <7 at 5 minutes.

Results.  Infants of afebrile women with epidural analgesia were more likely to be evaluated for sepsis than infants of women without epidural (20.4% vs 8.9%), although not more likely to have neonatal sepsis. An increased risk of sepsis evaluation persisted in regression analysis (odds ratio: 3.1; 95% confidence interval: 2.0, 4.7) after controlling for confounders and was not explained by longer labors with epidural. Women with epidural were significantly more likely to have major and minor criteria for sepsis evaluation, including fetal tachycardia (4.4% vs 0.4%), rupture of membranes for >24 hours (6.2% vs 3.4%), low-grade fever of 99.6°F to 100.4°F (24.3% vs 5.2%), and rupture of membranes for 12 to 24 hours (21.4% vs 5.2%) than women without epidural.

Conclusions.  Epidural analgesia is associated with increased rates of major and minor criteria for neonatal sepsis evaluations in afebrile women.  Key words:  epidural analgesia, pregnancy, labor, neonatal sepsis evaluation.

Epidural analgesia is the most effective form of pain relief available during labor and is preferred by more than half of women during labor.¹ Like all medical treatments, however, epidural analgesia has potential side effects for women and their infants. The association between epidural use and increases in maternal temperature during labor has been well documented in randomized trials.^2-5 The underlying cause of this increase is not certain, but it has generally been attributed to altered thermoregulation.^6,7 We previously reported an association among epidural use, maternal fever, and increased rates of neonatal sepsis evaluation.⁸

Because intrapartum fever may be a marker for intrauterine infection, our finding that frank maternal fever triggers neonatal evaluation was not surprising. However, we also found that the majority of neonatal sepsis evaluations occur in the infants of afebrile women with epidural. This fact and the lack of documentation for the indications for neonatal sepsis evaluations in the infants of afebrile women were major criticisms of our work.⁹ In direct answer to these criticisms, we went back to the original charts of the women in our study who remained afebrile to evaluate the intrapartum characteristics of low-risk nulliparas and to explore the underlying reasons for the excess neonatal sepsis evaluations in afebrile women with epidural analgesia.

	METHODS

Top Abstract Methods Results Discussion References

The base sample for this study was 1934 women who were enrolled in the active management of labor trial conducted at Brigham & Women's Hospital from 1990 through 1994. Women were assigned randomly to either usual care or the active management of labor protocol. Epidural analgesia was available on request in both the usual care and active management groups. A complete description of the study methodology has been published elsewhere.¹⁰ Most temperatures were measured orally. The few axillary temperatures were corrected upward a single degree Fahrenheit to be comparable to oral.

The current analysis was limited to women who had singleton, term pregnancies with cephalic presentation and the spontaneous onset of labor (n = 1329). From this group, women were excluded when they were diabetic (n = 33) or when birth weight was missing (n = 4). Given that low-grade maternal temperature was an important study outcome, women also were excluded when they had an admission temperature of >99.5°F (n = 27) or no maternal temperatures were obtained (n = 30). The remaining 1235 women composed our study population. Women who were afebrile during labor (maximum temperature of <100.4°F) were the focus of our analysis (n = 1109).

At our institution, initial evaluation for neonatal sepsis consists of a blood culture and a complete blood count of the neonate. A lumbar puncture is performed on the basis of the level of clinical suspicion. The decision to perform a neonatal sepsis evaluation is based on standardized criteria developed by the neonatology staff and is performed when 1 major or 2 minor criteria are present. At the time of the study, major criteria (apart from a maternal temperature of >100.4°F) included rupture of membranes for >24 hours or a sustained fetal heart rate of >160 beats per minute. Sustained was defined for purposes of this study as a change of baseline for longer than 15 minutes. Minor criteria included a low-grade maternal temperature of 99.5°F to 100.4°F, rupture of membranes for 12 to 24 hours, maternal admission white blood cell count of >15 000 cells/mL³, or an Apgar score of <7 at 5 minutes. Maternal group B streptococcal status was not routinely evaluated at the time of this study.

Differences in maternal characteristics according to epidural use were evaluated using the Student t test for continuous variables and ² or Fisher exact test for categorical variables. The proportion of infants who were evaluated for sepsis among afebrile women with and without epidural analgesia was compared. In addition to observed workups, the presence of major and minor criteria for neonatal sepsis workup was determined and an expected rate of sepsis evaluation was calculated. An evaluation was expected in the presence of 1 major or 2 minor criteria because this is the standard applied in the clinical setting. Logistic regression was used to examine the rate of neonatal sepsis evaluation while controlling for baseline differences between the epidural and no-epidural groups and to calculate adjusted odds ratios (OR) and 95% confidence intervals (CI).

	RESULTS

Top Abstract Methods Results Discussion References

The final study group consisted of 1109 afebrile women. Of these, 613 women (55.3%) received epidural analgesia and 496 women (44.7%) did not. A comparison of the baseline characteristics of these 2 groups is presented in Table 1. There were some notable differences between women who received epidural analgesia and those who did not. On average, women who requested epidural were 1.4 days farther along in gestational age, their infants were 58 g heavier, and they tended to be admitted with a less advanced cervical dilation (3.0 vs 4.2 cm). In addition, women who requested epidural were somewhat more likely to have smoked during pregnancy (6.2% vs 3.6%) and were less likely to have had their labor managed using the active management protocol (46.2% vs 56.1%). The rate of premature rupture of membranes (PROM) was similar for women who later received epidural and those who did not (17.6% vs 20.0%).

Infants of afebrile women with epidural analgesia were more likely to be evaluated for sepsis than infants of women without epidural (20.4% vs 8.9%; P = .001). A >3-fold increased odds of sepsis evaluation persisted in logistic regression analysis (OR: 3.1; 95% CI: 2.0, 4.7) controlling for gestational age, birth weight, maternal smoking, active management of labor, admission cervical examination, and PROM. There were no cases of documented sepsis.

To determine the reason for this higher rate of sepsis evaluation, we examined the occurrence of clinical criteria for sepsis evaluation in both groups (Table 2). Epidural use was associated with an increased rate of expected sepsis evaluation on the basis of the presence of major and minor criteria. The criteria for neonatal sepsis evaluation (1 major or 2 minor criteria) were met in 22.4% of infants whose mothers received epidural analgesia but only 8.3% of those whose mothers did not receive epidural analgesia (Table 3). When evaluating the specific criteria that contribute to this difference in the expected rate of evaluation, we found that afebrile women who received epidural analgesia had a 1.8-fold increased risk of rupture of membranes for >24 hours (6.2% vs 3.4%) and a 10.9-fold increased risk of fetal tachycardia (4.4% vs 0.4%). In addition, afebrile women who received epidural analgesia had a 4.6-fold increase in low-grade temperature elevations (24.3% vs 5.2%) and a 1.5-fold increase in rupture of membranes for 12 to 24 hours (21.4% vs 13.9%). The rates of maternal leukocytosis and low 5-minute Apgar score were not higher in the epidural group. One possible confounder that has been advanced to explain the association of epidural with low-grade fever (99.5°F-100.4°F) is that women with increasing temperature are more uncomfortable and so may be more likely to request epidural.¹¹ However, only 13 women (2.1%) who received epidural analgesia had a low-grade temperature elevation (>99.5°F) at the time of epidural placement. In addition, the mean time from epidural to low-grade temperature was 4.3 hours (±3.8 hours) in women who received epidural analgesia.

Labor was significantly longer in women with epidural analgesia (10.7 hours vs 6.7 hours; P = .0001). As increased length of labor is associated with epidural use in randomized trials,^12,13 possibly because of the direct effects of epidural, we did not control for this factor in our logistic regression. However, because longer labor could be an important factor that mediates the association of epidural with sepsis evaluation, we wanted to examine whether, independent of length of labor, there was an association between epidural and sepsis evaluation. To determine this, we performed a stratified analysis. Although afebrile women with epidural analgesia labored longer, neonatal sepsis evaluations were more common with epidural, regardless of length of labor (Fig 1). Significant increases in rates of neonatal sepsis evaluations were seen in the epidural group even when labor was <6 hours (12.6% vs 6.1%) and were also seen in the epidural group at 6 to 12 hours (19.4% vs 9.2%) and 12 to 18 hours (46.8% vs 19.1%). Although an increased rate of neonatal sepsis evaluation was seen in the epidural group at >18 hours of labor (73.5% vs 55.6%), this large difference was no longer significant (P = .3), likely because of the small number of women who labored this long and who did not have an epidural (n = 9).

View larger version (32K): [in this window] [in a new window]   Fig. 1.   Sepsis evaluations in neonates of afebrile women according to length of labor.

Finally, we examined the expected rate of sepsis evaluation in the overall population of 1235 women (1109 afebrile and 126 febrile). For the overall population, the rate of expected neonatal sepsis evaluation based on major and minor criteria increases to 35.3% in women with epidural analgesia (compared with 22.4% expected among afebrile women). In contrast, there is only a slight increase in women without epidural analgesia to 8.6% (from 8.3% expected among afebrile women). The percentage of infants who actually underwent evaluation is similar to the rate predicted by major and minor criteria in both the afebrile group and the overall population (Table 3).

	DISCUSSION

Top Abstract Methods Results Discussion References

Epidural analgesia is associated with a higher rate of neonatal sepsis evaluation in term, low-risk women.⁸ Although part of this association is attributable to the previously documented relationship between epidural and fever, in our population, infants of afebrile women with epidural analgesia also have a higher rate of evaluation.⁸ This follow-up study presents data gathered to elucidate the underlying factors that contribute to the higher rate of neonatal sepsis evaluations observed in the infants of afebrile women with epidural.

The observed rate of neonatal sepsis evaluation in infants of afebrile women without epidural analgesia was substantially lower (8.9%) than the rate observed in afebrile women with epidural (20.4%). Our analyses indicate that this increased rate can be explained by the association that we showed between epidural analgesia and several of the criteria that prompt neonatal evaluation, including prolonged rupture of membranes, fetal tachycardia, and low-grade fever. Importantly, there were no cases of documented neonatal sepsis in either group.

Our overall rate of neonatal sepsis evaluation of 15.2% in term afebrile women reflects the care practices of our institution at the time of the study. Although criteria may differ elsewhere, these criteria are based on accepted risk factors for neonatal infection¹⁴ and were designed to result in a cautious standard for evaluation. Although many infants are evaluated, fewer than 50% of the infants evaluated received antibiotic therapy. Maternal epidural use was not a factor considered in the criteria for neonatal sepsis evaluation, but in our data set maternal epidural use is associated with labor characteristics that are considered risk factors for infection.

Philip et al⁴ analyzed the relationship among epidural, maternal fever, and neonatal sepsis evaluation in women who were randomized to epidural analgesia. The association between maternal fever and epidural analgesia was again confirmed with a 4-fold increase in fever in women with epidural analgesia, controlling for length of labor and nulliparity. Although Philip et al did not analyze the relationship between epidural analgesia and neonatal sepsis evaluation directly using regression analysis, the rate of neonatal sepsis evaluation in the setting of maternal fever was 96% versus 13% in afebrile women. However, in contrast to our findings, there was no increased rate of neonatal sepsis evaluation in their subset of afebrile women with epidural analgesia.

Philip et al attributed the difference in results to bias in our nonrandomized study. More likely, these different results stem from different practice protocols used in the decision to perform a neonatal sepsis evaluation. Neonatologists at our institution used combinations of low-grade fever and prolonged rupture of membranes to flag infants of afebrile women for an increased sepsis risk, whereas the specific indications for infants of afebrile mothers who underwent sepsis evaluations in the study by Philip et al were not given. General indications that were reported include temperature instability, dusky spells, lethargy, and hypoglycemia.

As our study is observational, women chose whether to receive epidural analgesia. This self-selection results in some differences between the epidural and the no-epidural groups. We used a combination of regression techniques and stratified analysis to control for these differences while examining the association of epidural with sepsis evaluation. Our results are supported by the fact that the factors that account for the majority of the association of epidural with sepsis evaluation have been shown to be effects of epidural in randomized trials. The association of epidural use with an increased length of labor has been well demonstrated.^12,13 Longer labor contributes to the increased likelihood of meeting criteria for sepsis evaluation related to prolonged rupture of membranes. However, although increased length of labor may be partially explanatory, we demonstrated that significant excess rates of neonatal evaluations are still observed in afebrile women even at the shortest length of labor. These excess neonatal evaluations likely are attributable to the higher rate of maternal low-grade temperature elevation and fetal tachycardia that we observed in the epidural group. The association between epidural and increased maternal temperature also has been well documented in randomized trials.^2-5 In addition, fetal tachycardia has been demonstrated to accompany the increases in maternal temperature that accompany epidural use.⁷

Although increases in length of labor could result in added risk of infection as a result of increased length of rupture of membranes and increased opportunity for vaginal examinations, we did not find any increase in the rate of documented neonatal infections. When vaginal examinations are controlled in multivariate analysis, epidural remains a strong predictor of neonatal sepsis evaluation (OR: 2.4; 95% CI: 1.5, 3.7). We were unable to assess adequately histologic chorioamnionitis, as the placenta is not routinely submitted for pathologic examination at our institution.

Our results point out a potentially significant limitation in the criteria used to prompt neonatal sepsis evaluations in the infants of women who receive epidural analgesia. These criteria should be reevaluated to determine whether it is possible to decrease the rate of evaluation safely, but the situation is complex. Difficulty in distinguishing physiologic effects related to epidural from true infection persists for obstetrician and neonatologist alike. Maternal temperature elevation can be associated with either epidural use or infection. Prolonged duration of rupture of membranes and increased length of labor in women with epidural analgesia may raise concerns about a potential increase in the opportunity for infection. Development of methods for distinguishing infection from epidural-related fever would reduce the need for evaluation of these infants.

Epidural analgesia remains the most effective means of pain relief for the laboring woman and should be available to women. However, more complete information related to potential side effects of epidural analgesia also should be available to women. Discussion of pain control during labor should occur during pregnancy so that women have an opportunity to weigh risks and benefits before the onset of labor.

	ACKNOWLEDGMENTS

This study was supported by National Institute of Health and Human Development Grant No. RO1-HD26813.

	FOOTNOTES

Received for publication Aug 21, 2000; accepted Apr 23, 2001.

Reprint requests to (L.G.) Department of Obstetrics and Gynecology, Division of Maternal-Fetal Medicine, Baylor College of Medicine, 6550 Fannin St, 9th Floor, Houston, TX 77098. E-mail: lgoetzl{at}bcm.tmc.edu

	ABBREVIATIONS

OR, odds ratio; CI, confidence interval; PROM, premature rupture of membranes.

	REFERENCES

Top Abstract Methods Results Discussion References

1. Hawkins JL, Beaty BR, Gibbs CP Update on U.S. OB anesthesia practices. Anesthesiology 1977; 87:135-143
2. Ramin SM, Gambling DR, Lucas MJ, Sharma SK, Sidawi JE, Leveno KJ Randomized trial of epidural versus intravenous analgesia during labor. Obstet Gynecol 1995; 86:783-789 [CrossRef][Medline]
3. Sharma SK, Sidawi JE, Ramin SM, Lucas MJ, Leveno KJ, Cunningham FG Cesarean delivery: a randomized trial of epidural versus patient-controlled meperidine analgesia during labor. Anesthesiology 1997; 87:487-494 [CrossRef][Medline]
4. Philip J, Alexander JM, Sharma SK, Leveno KJ, McIntire DD, Wiley J Epidural analgesia during labor and maternal fever. Anesthesiology 1999; 90:1271-1275 [CrossRef][Medline]
5. Halpern SH, Leighton BL, Ohlsson A, Barrett JFR, Rice A Effect of epidural vs. parenteral opioid analgesia on the progress of labor: a meta-analysis. JAMA 1998; 280:2105-2110 [Abstract/Free Full Text]
6. Macaulay JH, Bond K, Steer PJ Epidural analgesia in labor and fetal hyperthermia. Obstet Gynecol 1992; 80:665-669 [Medline]
7. Fusi L, Maresh MJA, Steer PJ, Beard RW Maternal pyrexia associated with the use of epidural analgesia in labour. Lancet 1989; 8657:1250-1252
8. Lieberman E, Lang JM, Frigoletto F, Richardson DK, Ringer SA, Cohen A Epidural analgesia, intrapartum fever, and neonatal sepsis evaluation. Pediatrics 1997; 99:415-419 [Abstract/Free Full Text]
9. Camaan W Intrapartum epidural analgesia and neonatal sepsis evaluations: a casual or causal association? Anesthesiology 1999; 90:1250-1252 [CrossRef][Medline]
10. Frigoletto FD, Lieberman E, Lang JM, A clinical trial of active management of labor. N Engl J Med 1995; 333:745-750 [Abstract/Free Full Text]
11. Dolak JA, Brown RE Epidural analgesia and neonatal fever [letter]. Pediatrics 1998; 101:492
12. Thorp JA, Hu DH, Albin RM, The effect of intrapartum epidural analgesia on nulliparous labor: a randomized, controlled, prospective trial. Am J Obstet Gynecol 1993; 169:851-858 [Medline]
13. Alexander JM, Lucas MJ, Ramin SM, McIntire DD, Leveno KJ The course of labor with and without epidural anesthesia. Am J Obstet Gynecol 1998; 178:516-520 [CrossRef][Medline]
14. American Academy of Pediatrics, Committee on Infectious Diseases and Committee on Fetus and Newborn. Revised guidelines for the prevention of early-onset group B streptococcal (GBS) infection. 1997;99:489-496