A Controlled Trial of a Program for the Active Management of Labor

All nulliparous patients admitted to the labor and delivery suite of Northwestern Memorial Hospital in spontaneous labor after 37 or more completed weeks of gestation were eligible for participation in this study. Brochures describing the program had been given to potential participants during the previous month of prenatal care. Exclusion criteria included multiple gestation, noncephalic presentation, and previous uterine surgery. Medical complications such as diabetes and hypertension did not exclude patients from participating. Spontaneous labor was defined as the presence of regular, painful uterine contractions at least once every five minutes in association with either complete cervical effacement or spontaneous rupture of membranes. The degree of cervical dilation was not considered in the diagnosis of labor, but the rate of cervical dilation was assessed once labor began. Discharge to the home or sedation was encouraged for patients who did not meet the criteria for the diagnosis of labor, although additional time was allotted for observation in selected cases. If amniotomy was performed or augmentation of labor with oxytocin was begun before the criteria for labor were satisfied, the patient was excluded from participation in the study.

After the diagnosis of labor was made, informed consent was obtained. Patients were then assigned to either the active-management protocol or traditional management of labor (control group). Randomized assignment, without stratification, was based on a permuted-blocks design, with numbers placed sequentially in opaque, sealed envelopes. The fetal heart rate was monitored electronically in all cases. The frequency and intensity of contractions were assessed by external tokodynamometry, and the use of intrauterine pressure catheters was encouraged whenever there was doubt about the adequacy of contractions. Umbilical-cord arterial and venous blood was obtained routinely at delivery for the measurement of blood gases.

Patients assigned to the active-management protocol had amniotomy performed within one hour of the diagnosis of labor if the membranes were still intact. Cervical examinations were performed hourly for the first three hours, then every two hours. If the rate of cervical dilation during the first stage of labor was less than 1 cm per hour, progress in labor was considered inadequate and augmentation with oxytocin was initiated at a rate of 6 mU per minute. (Oxytocin was also used when the descent of the fetal head was arrested for one hour in the second stage.) The rate of infusion of oxytocin was increased by 6 mU per minute every 15 minutes until there were seven contractions every 15 minutes or until a maximal rate of 36 mU per minute was reached. The occurrence of more than seven contractions per 15 minutes was defined as tachysystole and managed by decreasing the oxytocin infusion rate by 6 mU per minute. When abnormal fetal heart-rate patterns were of concern and occurred in the absence of tachysystole, the capillary pH was measured in samples of blood from the fetal scalp; the oxytocin infusion rate was not changed. In such cases, the fetal-scalp pH measurements were interpreted, and the decision whether to proceed with operative intervention was made by the attending obstetrician. The oxytocin infusion was stopped altogether only for prolonged decelerations of the fetal heart rate or when the results of scalp sampling suggested the need for immediate delivery.

For the patients assigned to traditional management, the timing of amniotomy, the frequency of cervical examinations, and the criteria for identifying inadequate progress were left up to the attending obstetrician. In general, the absence of further cervical dilation for two hours was considered an arrest of progress, and a rate of dilation of less than 1 cm per hour was considered protraction of labor. When arrest or protraction was identified, oxytocin infusion was initiated at 1 mU per minute, and the dose was increased by 1 or 2 mU per minute at 15-minute intervals until uterine contractions were occurring at a frequency of eight per 20 minutes. More frequent uterine contractions were managed by decreasing the rate of infusion of oxytocin or stopping the infusion. When abnormal fetal heart-rate patterns were a matter of concern, the capillary pH was measured in samples of blood from the fetal scalp or the oxytocin infusion rate was decreased.

Secondary arrest disorders were managed similarly in both study groups. Arrest of cervical dilation for two hours in the first stage of labor despite an appropriate frequency and intensity of contractions or despite the infusion of oxytocin at the maximal rate allowed in the protocol was considered justification for cesarean delivery. Arrest of the descent of the fetus's head for one hour in the second stage of labor despite appropriate augmentation of labor with oxytocin was considered an indication for either operative vaginal delivery or abdominal delivery.

During the 12-month period before this study, the rate of cesarean section at Northwestern Memorial Hospital was 20.9 percent overall and 22.4 percent for nulliparous women. For nulliparous patients of private physicians, the rate was 23.4 percent, and for nulliparous clinic patients it was 17.0 percent. For the purpose of the study design, a lower overall rate was assumed for a population of women delivering at term that excluded patients with breech presentation, multiple gestation, and induction of labor. A sample-size calculation determined that, assuming a cesarean-section rate of 18 percent in the traditional-management group and a significance level of 0.05, 700 patients would be required for the study to achieve a power of 0.8 for detecting a decrease in the rate of cesarean section to 11 percent in the active-management group. Analyses were performed with Systat (Systat, Evanston, Ill.) and SPSS/PC (SPSS, Chicago) software and included Student's t-test, the chi-square test, and stepwise multiple logistic-regression analysis. Continuous data are expressed as means ±SD. Statistical significance was assumed at a level of P<0.05. The protocol was approved by the institutional review board of Northwestern University and the research committee of Northwestern Memorial Hospital.

This study was conducted between February 5, 1990, and March 1, 1991. During the study period, 966 nulliparous patients were determined to be eligible and were offered participation. Of these, 720 (75 percent) agreed to participate and were randomly assigned to treatment groups. Fifteen of these patients were subsequently excluded from the study and from the final analysis because of breech presentation (4 patients) or cessation of labor before amniotomy (i.e., reevaluation of the diagnosis of labor; 11 patients). There remained 351 patients in the active-management group and 354 in the traditional-management (control) group. Five additional patients assigned to the active-management group withdrew before amniotomy. Although their labor was managed in a manner similar to that of the control patients, these five patients were included in the analysis according to their original assignment to active management. Two of these five patients ultimately underwent cesarean section. There was only infrequently an uneven distribution of the patients of individual physicians between the two groups, and this factor did not affect the analysis.

Table 1. Table 1. Demographic Characteristics of the Study Patients.*

Demographic data are presented in Table 1. In addition to the variables listed, there were no significant differences between the two groups in the incidence of prenatal obstetrical or medical complications, height or weight before pregnancy, or total weight gain during pregnancy.

Table 2. Table 2. Characteristics of Labor in the Study Groups.*

Table 2 summarizes the characteristics of the study participants' labor. Those in the active-management group did not require augmentation of labor with oxytocin or conduction anesthesia any more frequently than the control patients. For women in the active-management group who required augmentation, however, the maximal rate of infusion of oxytocin was significantly higher than in the control group.

Table 3. Table 3. Type of Delivery According to Study Group.*

There was a 26 percent reduction in the overall rate of cesarean section in the active-management group (10.5 percent, as compared with 14.1 percent in the control group; P = 0.18) (Table 3). The reduction was due primarily to a decrease in the frequency of arrest disorders. This difference was even greater for the patients of private physicians (rate of cesarean section, 11.1 percent vs. 16.3 percent). After we controlled for potential confounding variables (maternal age, height, weight before pregnancy, weight gain during pregnancy, type of payer, gestational age, infant's birth weight, use of epidural anesthesia, and degree of dilation, effacement, and membrane status on admission), the reduction in the rate of cesarean delivery was statistically significant (P<0.05; odds ratio, 0.57; 95 percent confidence interval, 0.36 to 0.95). The rates of operative vaginal delivery did not differ significantly between the active-management and control groups.

Table 4. Table 4. Labor Intervals for Patients with Vaginal Deliveries, According to Study Group.*

Table 4 summarizes the progress of labor in patients who delivered vaginally. The interval from admission to amniotomy was significantly shorter for patients in the active-management group, and therefore the degree of cervical dilation at the time of membrane rupture was lower (3.9 cm vs. 4.6 cm, P<0.001). Both the first stage of labor and the total duration of labor were significantly shorter for patients whose labor was actively managed; earlier amniotomy, earlier use of oxytocin, and a more rapid response to augmentation all appeared to contribute to this reduction. Only 16 (4.6 percent) of the patients in the active-management group had not delivered by 12 hours after admission, as compared with 67 (18.9 percent) of the control patients (X² = 32.9, P<.001).

Table 5. Table 5. Complications of Labor and Delivery, According to Study Group.*

As shown in Table 5, there were no increases in complications of labor associated with active management of labor. In fact, active management was associated with a 53 percent reduction in maternal infectious morbidity (chorioamnionitis and endometritis). The rate of infusion of oxytocin was lowered or the infusion was stopped altogether less frequently for patients in the active-management group than for control patients. Despite the relatively large number of patients in each group who required a decrease in the oxytocin infusion rate, few operative interventions were necessary because of abnormalities in the fetal heart rate.

Table 6. Table 6. Neonatal Characteristics, According to Study Group.*

Variables indicating neonatal outcome are listed in Table 6. In addition to arterial pH, other umbilical blood gas indexes we compared were arterial partial pressure of oxygen, partial pressure of carbon dioxide, and base deficit, and venous pH, partial pressure of oxygen, partial pressure of carbon dioxide, and base deficit. No significant differences were detected between the active-management and control groups. No statistical or clinical differences remained in any of these measures after we controlled for the use of oxytocin, type of delivery, and the indication for operative delivery. Examination of the newborns, both at delivery and at discharge, did not detect any increase in morbidity (e.g., fractures, cephalhematomas, or hyperbilirubinemia) in the infants of mothers whose labor was actively managed.

In this prospective, randomized trial, the active management of labor for nulliparous patients (with early amniotomy followed by augmentation with oxytocin, when needed) was associated with a statistically significant decrease in the rate of cesarean section with no detectable increase in maternal or fetal morbidity. This reduction was primarily due to a decrease in the frequency of dystocia. This method of labor management was especially effective for private patients, a group recognized as being at increased risk for dystocia.^{9 , 10} It is important to emphasize that this study was designed to test only three components of the system of active management of labor — early amniotomy, early diagnosis of inadequate progress in labor, and the use of oxytocin at a higher dose than is usual. The criteria for the diagnosis of labor used in the active management of labor may also contribute to the efficacy of the program, but these criteria were also used for the control group to ensure comparability between patient populations. In addition, the unblinded nature of the study may have resulted in a drift in the management of labor in the control group toward a more active approach (Hawthorne effect), thereby contributing to a lower cesarean-section rate for this group as well. The original protocol for the active management of labor emphasized the organizational components of the program in addition to its medical components. We did not incorporate prenatal patient education about the active management of labor or rigorous peer review, nor were nurse-midwives used to manage labor. The use of conduction anesthesia was common, whereas in previous studies it was infrequent. These departures from the original protocol were made in order to evaluate whether a program of active management of labor could be used in American hospitals that serve primarily private patients. Consequently, the comparison of the complete protocol for the active management of labor with traditional management might have yielded an even greater difference in the rates of cesarean delivery.

Among the patients who delivered vaginally, amniotomy was performed earlier in the active-management group, but the interval from amniotomy to delivery was similar in both groups. Among the patients who received oxytocin and delivered vaginally, those in the active-management group had a shorter interval from the diagnosis of labor to the beginning of the oxytocin infusion and a significantly but less dramatically shorter interval from the initiation of augmentation to delivery. Both early amniotomy and the early diagnosis and treatment of dysfunctional labor therefore appear to be important in shortening labor. This suggests that the efficacy of active management is not due solely to the higher doses of oxytocin used, but rather to the program of labor management as a whole.

An earlier study comparing the rate of neonatal seizures at Parkland Hospital in Dallas with that at the National Maternity Hospital in Dublin, Ireland, suggested that the active management of labor may be associated with an increased risk of neurologic injury to the fetus.⁷ Although our study did not have sufficient power to evaluate this association, we found no increase in either seizures in newborns or other neurologic complications associated with active management of labor. The results of umbilical blood gas analysis, the test that evaluates a newborn's status at the time of delivery most objectively, were similar in the two groups and comparable to previously reported values for infants delivered at term.¹¹ Furthermore, interventions in response to abnormalities of the fetal heart rate were actually less common in the active-management group. Consequently, we found no evidence to suggest that the active management of labor presents an increased risk to the fetus as compared with the traditional method.

Several factors may explain how the active management of labor prevents dystocia. The definition of labor in our study, which included regular painful contractions and complete cervical effacement, may be more accurate for diagnosing labor than other criteria. More accurate diagnosis might permit physicians to avoid interventions that predispose patients to arrested or protracted labor. Early intervention with amniotomy and oxytocin, if needed, may also correct uterine dysfunction before muscular fatigue occurs and thus lower the incidence of dysfunctional labor. The avoidance of intrapartum infection, which is associated with prolonged labor, may also decrease the incidence of dysfunctional labor. Prenatal education, one-on-one companionship during labor, and reassurance about the length of labor may all contribute to decreased anxiety on the part of the patient and thus increase the likelihood of vaginal delivery.¹² Finally, peer review of clinical practice and the identification of a decreased rate of cesarean delivery as an institutional goal may also contribute to the achievement of more vaginal deliveries.