Epidural Analgesia and Severe Perineal Laceration in a Community-based Obstetric Practice

Methods

Data were obtained from the medical records of all vaginal deliveries between June 1996 and June 2000 at St. Francis Regional Medical Center, a community hospital in Topeka, Kansas, a city of approximately 170,000 people. Anesthesia is available at all times. Obstetric specialists, family physicians, family practice residents, and nurse-midwives staff this community hospital. Consultations with obstetric specialists are not mandatory. A total of 3,074 women were delivered vaginally during the study period. Current analyses were limited to 2,759 patients for whom complete information was available and who had vertex, spontaneous or induced, vaginal deliveries of singleton, live babies of at least 36 weeks’ gestation. Patients with diabetes or severe cardiac disease were excluded because of potential changes in management of second stage of labor with these conditions.

The major outcome variable was an indicator for third- or fourth-degree perineal lacerations as defined by American College of Obstetricians and Gynecologists: a tear extending into the deep transverse perineal muscles and fibers of the anal sphincter (third degree) or rectal mucosa (fourth degree). Other study variables, such as year of birth, number of pregnancies, number of births, vaginal birth after cesarean section, use of episiotomy, type of analgesia (epidural, local, or none), degree of laceration, specialty of the delivering physician (family physician, obstetrician-gynecologist, or nurse-midwife), birth weight, use of inducing agents, maternal age, and use of instrument aides (vacuum, forceps, or none) were taken from literature review and included in the analysis.5 ,6 ,8 ,9 Race of the mother was eliminated as a potential confounder by previous studies.4

Statistical analysis was completed using the SAS software package, version 6.12. Logistic regression (univariate and multivariate) was used to evaluate associations while controlling for possible confounding variables. All tests were performed at the two-sided, α = .05 level. The results of data analyses are reported as odds ratios (OR), with 95% confidence intervals (CI). These reported ratios all had P values of less than .05.

Results

The overall rate of severe perineal (third- and fourth-degree) laceration was 6.38% (n = 176) (Table 1). Epidural analgesia was given to 634 (22.98%) women, 65 (10.25%) of whom had severe perineal lacerations compared with 111 (5.22%) women with severe lacerations who did not have epidural analgesia (Table 2). The characteristics of women who had epidural analgesia differed from those who did not. Women who had epidural analgesia were more likely to be nulliparous, have their labor induced by oxytocin, and have instruments used during their delivery (Table 2). Logistic regression to examine the association of epidural analgesia with severe lacerations, while controlling for potential confounders of episiotomy, parity, induction, maternal age, and birth weight, found epidural analgesia to be a significant predictor of severe perineal injury (OR = 1.528, CI = 1.092–2.137) (Table 3).

We then examined the rate of obstetric intervention according to use of epidural analgesia (Table 2). Use of epidural analgesia was associated with significantly increased operative vaginal delivery. The proportion of operative vaginal deliveries was 16.7% (106 of 634) in women who had epidural analgesia compared with 4.7% (100 of 2,125) in women without epidural analgesia.

A second logistic regression model was constructed to evaluate the effect of higher incidence of instrument use during delivery for women who had epidural analgesia (Table 3). When increased use of instruments was added into the model, epidural analgesia remained a risk factor for severe laceration with an odds ratio of 1.287 but with a statistically insignificant 95% confidence interval of 0.907 to 1.826.

To determine whether increased use of instruments during delivery was the explanation for the observed increase in severe perineal lacerations associated with epidural analgesia, the data were stratified by use of operative vaginal delivery. When the data were separated in this way (Table 4), there were significant differences between the two populations regarding the incidence of severe perineal laceration and use of epidural analgesia. Of the women who did not have an instrument-assisted delivery, severe laceration occurred in 7.77% (41 of 528) of women who had epidural analgesia. Laceration occurred in only 4.49% (91 of 2,025) of women who had neither epidural analgesia nor an instrument-assisted delivery. In contrast, for women who had an instrument-assisted delivery, there was very little difference in the incidence of severe laceration with or without epidural analgesia (22.64% and 20.0%, respectively).

Table 3 shows that use of instruments was a strong predictor of severe laceration (OR = 3.245, 95% CI = 2.162–4.869). Table 5 explores the predictors of instrument use. The logistic regression model shows that epidural analgesia is indeed a strong predictor of instrument use (OR = 3.01, 95% CI = 2.225 to 4.075). Whereas the results in Table 4 appear to show that epidural analgesia is an independent cause of severe lacerations, when instrument use was controlled for in a regression model, epidural analgesia did not show a statistically significant increase in risk for severe tear.

Discussion

The purpose of this study was to determine whether epidural analgesia was an independent risk factor for severe perineal laceration during vaginal delivery at a community-based obstetrical practice. The most recent study to explore this issue was by Robinson et al4 in a tertiary maternity hospital setting in 1999. They found initially that epidural analgesia was associated with increased rates of severe laceration in a population of nulliparous women, but after controlling for use of instruments during delivery, the association was no longer significant. This study also found that epidural analgesia was associated with an increased use of instruments during delivery, which led to the conclusion that epidural analgesia led to an increased use of operative delivery, and that the instruments, in turn, caused an increase in severe perineal laceration.

Our current analyses agree with these results (Table 3). The Robinson et al study,4 however, did not include a logistic regression model to determine risk factors for use of instruments in vaginal delivery. In addition, the Robinson et al study, like many, was performed at a tertiary maternity hospital. In contrast to the tertiary hospital setting of that study, the community hospital population of this study had a significantly lower incidence of instrument use (7.47% vs 16.6%), episiotomy (22.07% vs 47.6%), severe laceration (6.38% vs 14.0%), and epidural analgesia (22.98% vs 70.9%).

A study by Janssen et al in 200110 found that tertiary level maternity hospitals are associated with higher rates of epidural analgesia, cesarean delivery, and instrument-aided delivery. Thus, the findings by Robinson et al4 might not be applicable to community hospitals.

In 1988 Legino et al8 found that 22% of women who sustained a severe perineal laceration received epidural analgesia. Only 7% of women who did not have severe perineal trauma had epidural analgesia, however. The study did not put forth any explanation for the observed higher rates of severe perineal laceration in the epidural analgesia group.

Conversely, Walker et al6 in 1991 found no significant interaction between the incidence of epidural analgesia and the use of forceps, episiotomy, or parity. Further, the study concluded that use of epidural analgesia does not appear to be related to the incidence of perineal laceration in a population of nulliparous and multiparous patients. Combs et al5 agreed with this conclusion in 1990, when their study found no effect of epidural analgesia on severe perineal laceration in a population of 2,832 women having operative deliveries.

Finally, the oldest study, by Bickers7 in 1970, concluded that epidural analgesia was protective for perineal laceration. Because that study did not delineate between severe laceration (third and fourth degree) and less serious laceration (first and second degree), the lower laceration rate in the study can be possibly attributed to the lack of first- and second-degree lacerations in the epidural analgesia group. This group had increased rates of episiotomy. To our knowledge, these studies are the only ones to explore the association between epidural analgesia and severe perineal laceration.

Ours is the only reported study to use a community-based obstetric teaching practice to analyze the effect of epidural analgesia on incidence of severe laceration. In our study, women who gave birth vaginally at term with epidural analgesia had significantly higher incidences of severe laceration and instrument use. Epidural analgesia was found to be a strong predictor of instrument use (OR 3.01, 95% CI - 2.225–4.075). Operative vaginal delivery is well documented to be a significant cause of third- and fourth-degree laceration in both this study and many others12–16 (OR 3.245, 95% CI 2.162–4.869). Given these facts, it is strongly suggested that epidural analgesia can be linked to increased numbers of patients who have severe perineal trauma as a result of an increase in operative vaginal delivery.

It should be noted, however, that the patients who did not have instruments used during delivery had a 58% higher incidence of severe perineal laceration when epidural analgesia was used (Table 4). This finding suggests that epidural analgesia might be either an independent risk factor or linked to third- and fourth-degree perineal laceration by some other mechanisms. Despite the trend shown by the data in Table 4, epidural analgesia was not found to be a statistically significant independent risk factor when instrument use was included in the regression model.

Conclusion

Epidural analgesia has become an effective and popular method of pain control during vaginal delivery. Much attention is given to its safety and efficacy, but many patients and health care providers are unaware of its potential negative consequences. The effectiveness of epidural analgesia is not without the cost of an increased likelihood of instrument use during a delivery and the possibility of a severe laceration, such as a tear extending into the deep transverse perineal muscles and fibers of the anal sphincter (third degree) or rectal mucosa (fourth degree) (Tables 3 and 5). Patients and providers should be informed of the potential drawbacks of epidural analgesia, such as severe perineal laceration and its considerable long-term consequences, such as incontinence, dyspareunia, and rectovaginal fistula.

Epidural analgesia is associated with an increase in severe perineal trauma as a result of its tripling the risk of instrument use. Instrument use in vaginal delivery more than triples the risk of severe perineal laceration.