Objective. Although several studies have documented an increase in maternal temperature associated with use of epidural analgesia during labor, none have investigated the impact of epidural use on the rate of intrapartum fever or the consequences for the fetus and newborn of this elevated maternal temperature. This study evaluates the impact of epidural analgesia use during labor on the rate of intrapartum fever and the performance of neonatal sepsis evaluations and treatment with antibiotics.
Methods. We studied 1657 nulliparous women with term pregnancies and singleton vertex fetuses who were afebrile at admission for delivery. The rates of maternal intrapartum fever >100.4°F, neonatal sepsis evaluation, and neonatal antibiotic treatment according to use of epidural analgesia during labor were determined. Rate ratios and 95% confidence intervals (CI) were calculated. Multiple logistic regression was used to examine associations while controlling for confounding factors.
Results. Intrapartum fever >100.4°F occurred in 14.5% of women receiving an epidural but only 1.0% of women not receiving an epidural (adjusted odds ratio (OR) = 14.5, 95% CI = 6.3, 33.2). Without epidural, the rate of fever remained low regardless of length of labor; with epidural, the rate of fever increased from 7% for labors ≤6 hours to 36% for labors >18 hours. Neonates whose mothers received epidurals were more often evaluated for sepsis (34.0% vs 9.8%; adjusted OR = 4.3, 95% CI = 3.2, 5.9) and treated with antibiotics (15.4% vs 3.8%; adjusted OR = 3.9, 95% CI = 2.1, 6.1). Although 63% of women received epidurals, 96.2% of intrapartum fevers, 85.6% of neonatal sepsis evaluations, and 87.5% of neonatal antibiotic treatment occurred in the epidural group.
Conclusions. Use of epidural analgesia during labor is strongly associated with the occurrence of maternal intrapartum fever, neonatal sepsis evaluations, and neonatal antibiotic treatment.
Maternal temperature during labor is carefully monitored, because fever can be a marker for an infection that could adversely affect both mother and fetus. Several studies have documented that use of epidural analgesia for pain relief during labor is associated with increases in maternal temperature.1-5 Herbst et al, in a study investigating risk factors for fever during labor, found that among the women with fever, 61% had received epidural analgesia, compared with 21% among those with normal temperature.6 None of the studies performed, however, have investigated the impact of epidural analgesia on the rate of maternal fever during labor or the consequences for the fetus of an elevated maternal temperature. This study evaluates the impact of epidural analgesia use during labor on the rate of intrapartum maternal fever and the performance of neonatal sepsis evaluations in a large hospital-based population of nulliparous women.
The base sample for this study was the 1934 nulliparous women enrolled in the active management of labor (ACT) trial conducted at Brigham and Women's Hospital from May 1990 through October 1994. Approval for the study was obtained from the Human Research Committee of Brigham and Women's Hospital, and informed consent was obtained from all participants. Women enrolling were randomly assigned to have labor managed either under a protocol of active management of labor or to usual care. The active management of labor protocol mandated the criteria for the diagnosis of labor, the timing and dose of oxytocin, and one-to-one nursing throughout the course of labor. Epidural analgesia was not part of the trial protocol. It was administered to women in both groups on request. Data were abstracted from the maternal medical record regarding the course of pregnancy and labor and from the newborn record regarding infant outcome. A complete description of the study methodology has been published elsewhere.7
The current analysis was limited to women with singleton, term pregnancies with the infant in a cephalic presentation and spontaneous or induced labor resulting in a liveborn infant (N = 1701). In addition, women were excluded if a maternal fever or infection was present at admission to labor and delivery (n = 7), if they were diabetic (n = 33), or if data on birth weight were missing (n = 4).
Intrapartum fever was defined as a maternal temperature >100.4°F (38°C) during the course of labor. At our institution, a sepsis evaluation consists of performing a blood culture and complete blood count; inclusion of a lumbar puncture is at the discretion of the provider and is not usually performed in asymptomatic infants.
The rates of intrapartum fever, neonatal sepsis evaluation, and neonatal treatment with antibiotics were calculated for women who received epidural analgesia and those who did not. The crude rate ratios and 95% confidence intervals (CI) for the association of intrapartum fever, neonatal sepsis evaluation, and neonatal treatment with antibiotics associated with epidural use were computed. Multiple logistic regression was used to examine the association of epidural use with these outcomes while adjusting for potential confounding factors. Odds ratios (OR) were calculated from the regression coefficients and CI values from the SE values of those coefficients. The association of epidural analgesia with sepsis work-ups was also evaluated separately for women whose intrapartum temperature never exceeded 100.4°F.
Maternal Fever During Labor
Of the 1657 women included in the analysis, 1047 (63%) received epidural analgesia for pain relief during labor, whereas 610 (37%) did not receive an epidural. Women receiving epidural analgesia were more likely to have had induced rather than spontaneous onset of labor (28% of epidural users were induced vs 12% without epidural), and, on average, their labors were almost 6 hours longer. In addition, women who received an epidural had a slightly more advanced gestational age (mean difference, 2.1 days), and their infants were, on average, 106 g larger. They were also less likely to have had their labor managed with the active management protocol (32% epidural, 49% no epidural; Table 1).
Of the women who received epidural analgesia, 14.5% (n = 152) developed a fever >100.4°F (38°C) during labor compared with only 1.0% (n = 6) of women not receiving epidural analgesia. The crude rate ratio for the association of epidural analgesia with the occurrence of intrapartum fever was 14.8 (95% CI = 6.6, 33.2). Overall, 96% of intrapartum fevers (152/158) occurred within the group of women who received epidural analgesia during labor. In a logistic regression analysis controlling for birth weight, gestational age, induction of labor, premature rupture of the membranes, and treatment with active management of labor, the association was essentially unchanged (OR = 14.5, 95% CI = 6.3, 33.2).
Fig 1 describes the relation between fever and length of labor (defined as time from admission to delivery) in the epidural and no-epidural groups. The rate of fever in the no-epidural group, 1% overall, remained low, regardless of the length of labor. The rate of fever in the group of women who received an epidural was consistently higher than the rate in the group of women who did not receive an epidural, regardless of the length of labor. For women who received an epidural, the rate of fever increased with longer labors, from 7% for women with the shortest labors (≤6 hours) to 36% for women with the longest labors (>18 hours). Greater than 40% of women receiving an epidural had labors lasting at least 12 hours, putting them at high risk for intrapartum fever (Table 2).
NEONATAL SEPSIS EVALUATION AND TREATMENT WITH ANTIBIOTICS
Among the 1657 infants, 25.1% were evaluated for the possible occurrence of sepsis. The rate of evaluation for sepsis was higher in the epidural group: 34.0% (n = 356), as compared with 9.8% (n = 60) in the no-epidural group (Table 3). In a multiple logistic regression analysis performed to evaluate the association of epidural with sepsis work-ups while taking into account potentially confounding factors (birth weight, gestational age, induction of labor, premature rupture of the membranes, and treatment with active management of labor), the OR for the association of epidural use with neonatal sepsis evaluations was 4.3 (95% CI = 3.2, 5.9) (Table 3). Overall, 85.6% of neonatal sepsis work-ups (356/416) occurred within the group of infants with a mother who had received an epidural.
Neonates whose mothers had received an epidural were also four times as likely to be treated with antibiotics because of suspicion of sepsis (15.4% epidural, 3.8% no epidural; Table 3). Controlling for potentially confounding factors in a multiple logistic regression analysis did not appreciably alter the association (OR = 3.9, 95% CI = 2.1, 6.1). Infants in the epidural group were also three times more likely to be treated for a period of ≥3 days. Overall, 87.5% (161/184) of neonatal antibiotic treatment occurred among the group of infants whose mothers had epidural analgesia during labor. Of the total 416 infants evaluated for sepsis, only 4 infants had documented sepsis: 1 infant whose mother did not receive an epidural and 3 whose mothers did.
Because 25% of infants were evaluated for sepsis whereas only 9.5% of mothers developed a fever of >100.4°F during labor, we further examined the data to determine what portion of sepsis work-ups were related to maternal fever (Fig 2). In the presence of an intrapartum maternal fever >100.4°F, most neonates were evaluated for sepsis regardless of epidural use. However, as fever was much more common in the epidural group (14.5% epidural group, 1% no epidural group), the actual number of sepsis evaluations performed was much higher in that group (131 in the epidural group, 5 in the no-epidural group). Even though most infants of women with intrapartum fever had a sepsis work-up, maternal fever accounted for only 32.7% (136/416) of neonatal evaluations for sepsis.
Even in the absence of maternal fever >100.4°F, sepsis evaluations were performed at a higher rate in the epidural group. The rate in the epidural group was 25.1% compared with 9.1% in the no-epidural group. In a multiple logistic regression (adjusting for birth weight, gestational age, induction of labor, premature rupture of the membranes, and treatment with active management of labor) performed within the group with no intrapartum fever, infants of women who had received an epidural remained more than three times as likely to be evaluated for the presence of sepsis (OR = 3.2, 95% CI = 2.3, 4.4).
Among women without intrapartum fever, those who had received an epidural were also more likely to have labors >12 hours long (38% with epidural, 8% without epidural) and ruptured membranes >18 hours (19.2% with epidural, 8.7% without epidural), factors that could contribute to the decision to evaluate for sepsis. However, longer length of labor did not entirely account for the higher rate of sepsis evaluations, because these infants were more likely to have a sepsis work-up regardless of length of labor (Fig 3). A multiple logistic regression analysis indicated that even when taking into account length of labor (in addition to the other factors controlled above), infants of women who had received epidural remained twice as likely to be evaluated for sepsis (OR = 2.1, 95% CI = 1.4, 2.9).
Summary of Outcomes
Table 4 summarizes the extent to which maternal intrapartum fever and neonatal adverse outcomes were associated with epidural use. The percent of cases associated with epidural use was calculated by comparing the number of observed cases of an adverse outcome in the entire sample with the number that would be expected if the rate of that outcome in the entire group were the same as that in the no-epidural group. For example, the rate of intrapartum fever in the no-epidural group was 1%, meaning that among the 1657 women in the study population, 17 cases would be expected. However, 158 cases were observed. Therefore, the percent of cases associated with epidural use was the number of excess cases (158−17) divided by the number of observed (158), or 89%. The data suggest that the vast majority of maternal fever, neonatal sepsis evaluation, and newborn antibiotic treatment is associated with use of epidural analgesia during labor.
In our population, almost all intrapartum fevers (96%) occur in women who have received epidural analgesia for pain relief. Only 1% of women without epidural analgesia develop an intrapartum fever >100.4°F. We also noted a fourfold increase in both the performance of sepsis evaluations and the treatment with antibiotics among the infants of women who received epidural analgesia during labor. In the presence of fever, the rate of sepsis evaluation was similar in the two groups. However, as the frequency of fever was substantially higher in the epidural group, a much greater number of sepsis evaluations were performed in that group. In the absence of fever, infants in the epidural group were nearly three times as likely to be evaluated for sepsis.
At our institution, criteria mandating the performance of a sepsis evaluation are explicitly delineated.8 Conventionally (at our institution and others), infants of mothers with intrapartum fevers >100.4°F are routinely evaluated for the presence of sepsis.9 In the absence of fever >100.4°F, other combinations of factors indicating the need for neonatal evaluation are specifically defined. Risk factors considered in determining whether a sepsis evaluation will be performed include the presence of membranes ruptured for >12 hours, maternal white blood cell count >15 000/mm3, lower levels of intrapartum temperature elevation, and neonatal symptoms in the delivery room (poor color or tone). Maternal use of epidural analgesia is not a factor in determining whether a sepsis evaluation should be performed. Application of this standard in our institution results in 25% of term, singleton infants being evaluated for sepsis. It is clear that this is a cautious standard, because fewer than half of the babies evaluated actually received treatment with antibiotics.
The increase in sepsis work-ups in the epidural group in the absence of an intrapartum temperature >100.4°F is not fully accounted for by the longer length of labor known to be associated with epidural analgesia10 and likely also relates to other factors such as temperature elevations lower than the 100.4°F threshold. Because our data indicated only the presence or absence of fever >100.4°F, we were unable to document whether women receiving epidural more frequently had temperature elevations not reaching 100.4°F.
An alternative explanation for our findings is that women with fever differentially receive epidural analgesia. Although we do not have information in this dataset on the onset of fever relative to epidural administration, in a similar dataset of all nulliparous women delivering at our institution during a 6-month period (N = 1954), 96% of fever in the epidural group occurred after the administration of the epidural. On average, the onset of fever was >6 hours after the initiation of epidural analgesia.
Several previous studies have documented an increase in maternal temperature associated with epidural use.1-6 Estimates of the rate of rise of maternal temperature in the presence of epidural analgesia have been between 0.08°C per hour 3 and 0.14°C per hour.1 It has been suggested that fetal harm may result from maternal temperature elevation during labor. In a study measuring fetal skin temperature during labor, Macaulay et al5 found that in 9% of cases (3/33) in which the mother was given epidural analgesia, the fetal skin temperature reached 39°C, as compared with the no-epidural group, in which no fetuses had a skin temperature that high. They suggest that as the fetal core temperature is likely to be 0.75°C higher than fetal skin temperature, core temperature may sometimes reach 40°C, a temperature that in adults, they note, is associated with a greatly increased risk of heatstroke.
Although the mechanism by which epidural analgesia may result in these temperature increases is not entirely clear, suggested mechanisms generally relate to thermoregulatory changes induced by the epidural administration1,4,5 rather than to infection. During labor, however, it is difficult to determine whether a fever is of infectious origin, because traditional markers are not useful; white blood cell counts tend to be elevated,11,12 and it is not practical to use abdominal tenderness as a marker. Because it is not possible to distinguish confidently whether infection is present, in the presence of significant fever, antibiotics are administered to treat both the mother and the fetus in the event that the fever is of infectious origin.
In the presence of maternal intrapartum fever, neonates are also often treated. Neonatal sepsis work-ups and administration of antibiotics are not entirely benign, however. Treated neonates receive a combined regimen of ampicillin and gentamicin. Although gentamicin is relatively safe, a number of potential dose-related side effects, including nephrotoxicity and ototoxicity, do exist.13 In addition, parents may be affected psychologically when their newborn is admitted to the intensive care unit for tests and spends a minimum of 48 hours receiving intravenous medication. Finally, evaluation and treatment of these infants uses substantial additional health care resources, particularly if antibiotic treatment extends neonatal hospital stay.
These data indicate that in our population, the majority of cases of maternal fever during labor are related to epidural use. Because intrapartum fever in the absence of epidural is rare, even with lower rates of epidural use, most fevers would still occur among women receiving an epidural. Most of these fevers are likely to be noninfectious in origin.1,4,5 Because it currently is difficult to clearly differentiate infectious from noninfectious fever during labor, however, intrapartum fever frequently results in evaluation and treatment of neonates for possible sepsis. Given the cost, risk, and pain to the newborn, the high proportion of sepsis work-ups that may be attributable to epidural use is cause for concern. Our results suggest that existing criteria for neonatal sepsis evaluation and antibiotic treatment should be re-examined, perhaps using a higher fever threshold for women with epidural. In addition, there should be additional study of ways to limit epidural-related temperature elevations by adjusting the ambient temperature of the labor room or by cooling the mother by sponging or fan.1,5Finally, the possible consequences of fever resulting from epidural use should be discussed by women and their health care providers when making the decision about the method of pain relief to be used during labor.
This work was supported by National Institute of Child Health and Human Development Grant #RO1-HD26813.
We thank Dr. Philip J. Steer for conversations regarding the thermoregulatory effects of epidural and Dr. Ruth Tuomala for helpful comments on the manuscript.
- Received April 16, 1996.
- Accepted July 1, 1996.
Reprint requests to (E.L.) Department of Obstetrics and Gynecology, Brigham and Women's Hospital, 75 Francis Street, Boston, MA 02115.
- RR =
- rate ratio •
- CI =
- confidence interval •
- OR =
- odds ratio
- ↵Fusi L, Maresh MJA, Steer PJ, Beard RW. Maternal pyrexia associated with the use of epidural analgesia in labour. Lancet. 1989;1250–1252
- Camann WR,
- Hortvet LA,
- Hughes N,
- Bader AM,
- Datta S
- ↵Guerina N. Bacterial and fungal infections. In: Cloherty JP, Stark AR, eds. Manual of Neonatal Care. 3rd ed, pp 146–169. Boston, MA: Little Brown; 1993
- Copyright © 1997 American Academy of Pediatrics