Original Article
Effects of Electronic Fetal-Heart-Rate Monitoring, as Compared with Periodic Auscultation, on the Neurologic Development of Premature Infants
N Engl J Med 1990; 322:588-593March 1, 1990
- Abstract
- Abstract
In a multicenter, randomized clinical trial, we assessed the early neurologic development of 93 children born prematurely whose heart rates were monitored electronically during delivery and compared it with that of 96 children born prematurely whose heart rates were periodically monitored by auscultation. All the children were singletons with cephalic presentation, and all weighed ≤ 1750 g at birth. The mental and psychomotor indexes of the Bayley Scales of Infant Development (standardized mean score ±SD, 100±16) and a formal neurologic examination were administered at three follow-up visits (at 4, 8, and 18 months of age, corrected for gestational age).
At 18 months, the mean mental-development scores in the groups receiving electronic fetal monitoring and periodic auscultation were 100.5±2.4 and 104.9±1.8, respectively (P>0.1). The mean psychomotor-development scores in the two groups at 18 months were 94.0±2.4 and 98.3±1.8, respectively (P>0.1). The incidence of cerebral palsy was higher in the electronically monitored group —20 percent as compared with 8 percent in the group that was monitored by auscultation (P<0.03). In the electronicfetal-monitoring group (but not in the periodic-auscultation group), the risk of cerebral palsy increased with the duration of abnormal fetal-heart-rate patterns, as assessed by retrospective review (χ2 trend = 12.71, P<0.001). The median time to delivery after the diagnosis of abnormal fetal-heart-rate patterns was 104 minutes with electronic fetal monitoring, as compared with 60 minutes with periodic auscultation.
We conclude that as compared with a structured program of periodic auscultation, electronic fetal monitoring does not result in improved neurologic development in children born prematurely. (N Engl J Med 1990; 322: 588–93.)
Media in This Article
Figure 1
Distribution of Scores on the Bayley Mental and Psychomotor Development Indexes for Infants Tested at 18 Months (Corrected for Gestational Age) after Intrapartum Electronic Fetal Monitoring or Periodic Auscultation.Figure 2
Duration of Abnormal Fetal Heart Rates during Intrapartum Surveillance with Electronic Fetal Monitoring or Periodic Auscultation among Infants with and without Cerebral Palsy.Article Activity
- Article
ELECTRONIC fetal-heart-rate monitoring is widely used in the United States1 to reduce perinatal mortality and the incidence of disorders of neurologic development.2 The rationale is based primarily on retrospective studies of perinatal mortality and electronic fetal monitoring.3 4 5 However, in eight randomized clinical trials perinatal mortality was not reduced with electronic fetal monitoring.6 7 8 9 10 11 12 13 Little direct experimental evidence has been put forward to support or refute assertions2 , 14 that electronic fetal monitoring reduces the risk of neurologic disorders.
We carried out a randomized clinical trial of electronic fetal monitoring during the delivery of infants born prematurely with birth weights of 1750 g or less. We found that as compared with a structured program of periodic auscultation, electronic fetal monitoring did not reduce perinatal mortality or other adverse perinatal outcomes.13 The present report follows the surviving children from this study to test the hypothesis that electronic fetal monitoring is beneficial for the neurologic development of children with birth weights of 1750 g or less.
Methods
Enrollment Criteria and Randomization
A group of 376 women in premature labor were enrolled at the University of Washington Hospital, Seattle; Grace Hospital, Vancouver, British Columbia; and Madigan Army Hospital, Tacoma, Washington, between November 1981 and February 1985.13 Mothers were excluded before randomization if they were in advanced labor, had a multiple gestation, had a fetus with a noncephalic presentation, had a fetus with an estimated birth weight over 1750 g or a gestational age of more than 32 or less than 26 weeks, were 17 years of age or younger, did not speak English, planned to have a cesarean section before labor, had placenta previa, or had a fetus with a known congenital anomaly. The mothers were randomly assigned with numbered, sealed envelopes to electronic fetal monitoring or periodic auscultation. After birth, the infants were included in the study if they had a birth weight of 1750 g or less. A total of 247 infants with birth weights of 1750 g or less fulfilled the study criteria.
Infants with birth weights of more than 1750 g were subsequently born to some of the women because some of the estimates of birth weights made at randomization were incorrect (42 in the electronic-fetal-monitoring group and 40 in the periodic-auscultation group) or because the premature labor ended and the mother returned to a community hospital for standard intrapartum care (25 in the electronic-fetal-monitoring group and 22 in the periodic-auscultation group). For all 376 mothers (whether or not their infants met the birth-weight criterion), the distributions between the two groups with respect to birth weight and number of days to delivery after randomization were very similar, as is shown by the 20th, 40th, 60th, and 80th percentiles for birth weight: 1040, 1384, 1680, and 2080 g, respectively, in the electronic-fetal-monitoring group and 1076, 1376, 1650, and 1940 g in the periodic-auscultation group. Similarly, for all mothers, the 20th, 40th, 60th, and 80th percentiles for the number of days from randomization to delivery were <1, <1, 1.0, and 2.8 days, respectively, in the electronic-fetal-monitoring group and <1, <1, 1.0, and 3.0 days in the periodic-auscultation group. Since our objective was to study electronic fetal monitoring during the labors of women who gave birth very prematurely, we limited our follow-up and analysis to children with birth weights of 1750 g or less.
Fetal-Monitoring Protocols and Withdrawals of Patients
All infants in both study groups were cared for on a one-to-one basis by a study nurse. In the electronic-fetal monitoring group, the fetal-heart-rate patterns were classified as reassuring, nonreassuring, or abnormal. A written protocol outlined the methods of management, including that of sampling the pH of blood from the scalp of fetuses with nonreassuring or abnormal fetal-heart-rate patterns. The presence of a fetal blood pH below 7.2 in conjunction with nonreassuring or abnormal patterns on heart-rate monitoring or the persistence of such patterns for more than 30 minutes (if fetal-blood sampling was unfeasible) were considered indications for rapid delivery by forceps or cesarean section.
In the group monitored by periodic auscultation, a DeLee–Hillis fetoscope or amplified Doppler ultrasonography was used at least every 15 minutes in the first stage of labor and at least every 5 minutes in the second stage. A base-line fetal heart rate was obtained between contractions. Auscultation was then performed for at least 30 seconds immediately after a contraction (as determined by palpation). The auscultated fetal heart rates were classified as reassuring or abnormal. The periodic-auscultation protocol did not include sampling of fetal scalp blood. A written protocol outlined the procedure for the treatment of fetuses with abnormal auscultated heart rates. The persistence of such abnormal rates for more than 30 minutes in the absence of a correctable cause was an indication for rapid delivery by forceps or cesarean section. (Details of the monitoring protocols are available on request.)
Soon after delivery, the records of electronic fetal monitoring and periodic auscultation were reviewed by the study nurses, who coded the fetal-heart-rate patterns with respect to type and duration and categorized them as abnormal or normal. Ten of the electronic-fetal-monitoring recordings that were categorized as abnormal and 20 of those not so classified were randomly selected and reviewed by a perinatologist who was unaware of the perinatal and neonatal events and the findings of the study nurse. Agreement between the study nurses and the perinatologist had a kappa statistic of 0.92 for the presence of abnormal fetal-heart-rate patterns and a correlation coefficient (r) of 0.83 with respect to their duration.
Six infants who received periodic auscultation were withdrawn from the study13 and subsequently underwent routine electronic fetal monitoring. Four of the six had follow-up examinations, and the results are reported according to the monitoring technique assigned at randomization (auscultation). No infant assigned to electronic fetal monitoring was withdrawn from the study.
Long-Term Follow-up and the Bayley Scales of Infant Development
Among the 247 study pregnancies, there were 35 infant deaths: 17 after electronic fetal monitoring and 18 after periodic auscultation.13 We endeavored to study the 212 surviving infants at 4, 8, and 18 months (corrected for gestational age15). Physical therapists (at 4 and 8 months) and clinical psychologists (at 18 months) administered the mental and psychomotor components of the Bayley Scales of Infant Development to assess the infants' "sensory-perceptual acuities, discriminations, and the ability to respond to these; the early acquisition of 'object constancy,' memory, learning, and problem solving ability; vocalizations and the beginnings of verbal communication; and early evidence of the ability to form generalizations and classifications."16 Scores on the Bayley scales have been standardized to a mean (±SD) of 100±16.16
At the four-month and eight-month examinations, an additional neuromotor index for premature infants, the Movement Assessment of Infants,17 was administered. This index evaluates muscle tone, the normal disappearance of primitive reflexes, automatic reactions such as righting oneself and maintaining equilibrium, and volitional movement. A score of 15 or more points on this index indicates a high risk of cerebral palsy.
A developmental pediatrician performed a neurologic examination of each infant at 18 months and diagnosed cases of cerebral palsy, defined as a nonprogressive disorder of movement or posture due to a defect in or damage to the developing brain.18 All the investigators assessing neurologic development, including the pediatricians, were unaware of the monitoring technique used during delivery. In the cases of the infants with cerebral palsy, the hospital records of the pregnancies were reviewed for ascertainment of special related circumstances.
Statistical Analysis
Least-squares multiple regression was used to adjust the scores on the Bayley Scales of Mental and Psychomotor Development to account for the hospital in which delivery took place, sex, birth weight (a continuous variable), premature rupture of membranes, intravenous administration of a uterine tocolytic agent, maternal smoking, race, whether the mother was married or living with a partner (the last five of these variables were dichotomous), and maternal age. The confidence limits for the differences in the adjusted Bayley indexes resulted from an estimate of standard error based on two independent samples.19 The F ratio was used to evaluate whether a mean developmental index for either monitoring technique differed significantly from the common mean after control for its potential confounders.
The ratio of the odds of cerebral palsy in children who were monitored electronically before birth to those in children who were auscultated was calculated with unconditional logistic regression20 to adjust simultaneously for the risk factors listed above. Ninety-five percent confidence intervals were calculated from the asymptotic-covariance matrix, and the corresponding P values from a likelihood-ratio statistic. The chi-square test for trend21 was used to evaluate trends in proportions. The chi-square test for heterogeneity was used to test for the uniformity of the odds ratio across strata.22 The Mann—Whitney U test was used to compare the median test scores on the Movement Assessment of Infants. Two-tailed P values are reported throughout, although the reader should bear in mind that our original hypotheses considered simply the benefits of electronic fetal monitoring.
Human Subjects Review
This study was approved by the institutional review board at the University of Washington. In addition, an external advisory board composed of medical and public health experts met on three occasions, reviewed the data, and advised continuation of the study. The mothers participating in the study gave written consent. Because of the longitudinal nature of the study, many of the long-term follow-up data were not collected until after enrollment was complete.
Results
The characteristics of the 247 mothers in the study before delivery or randomization (e.g., delivery hospital, marital status, age, premature rupture of membranes, race, and sex) and those during delivery and after randomization (e.g., administration of betamethasone, infant's birth weight, and administration of a tocolytic agent) were the same in the electronic-fetal-monitoring group and the periodic-auscultation group.13 Of the 212 surviving infants, 189 (89 percent) had at least one follow-up visit, and 173 (82 percent) had an 18-month follow-up visit. For the 173 infants, both antepartum and intrapartum characteristics were examined. Premature rupture of membranes was more common in the electronic-fetal-monitoring group, whereas the administration of a tocolytic agent was more common in the periodic-auscultation group (Table 1Table 1
Characteristics of the 173 Infants Who Had 18-Month Follow-up Visits after Electronic Fetal Monitoring or Periodic Auscultation.). Neonatal characteristics such as the Apgar score and the arterial pH of umbilical-cord blood did not vary substantially between the infants who had an 18-month follow-up visit and those who did not (Table 2Table 2
Birth Weight and Neonatal Characteristics of the 212 Surviving Infants in the Two Study Groups, According to Whether They Had a Follow-up Visit at 18 Months.); however, the tendency was for the infants who had an 18-month follow-up visit to have a lower birth weight and a greater likelihood of intracranial hemorrhage of grade 3 or 4.The mean scores on the Bayley indexes of mental and psychomotor development differed little between the two study groups at 4, 8, and 18 months (Table 3Table 3
Scores on the Bayley Scales of Infant Development at 4, 8, and 18 Months, According to Study Group.*, Fig. 1Figure 1Distribution of Scores on the Bayley Mental and Psychomotor Development Indexes for Infants Tested at 18 Months (Corrected for Gestational Age) after Intrapartum Electronic Fetal Monitoring or Periodic Auscultation.). At 18 months, the adjusted difference in the mean mental-development scores between the two groups was −3.3 (95 percent confidence interval, −8.9 to 2.3), and the adjusted difference in the mean psychomotor-development score was −4.4 (95 percent confidence interval, −10.4 to 1.6). Excluding the 18-month scores on the Bayley indexes for the four children who were withdrawn from the periodic-auscultation protocol did not affect the adjusted differences between the electronic-fetal-monitoring and the periodic-auscultation study groups.At 18 months, scores below 83 or below 68 on the mental-development index were more common after electronic fetal monitoring (20 percent [16 of 82] and 9 percent [7 of 82], respectively) than after periodic auscultation (9 percent [8 of 91] and 2 percent [2 of 91], respectively) (for the comparison between groups of the proportion of scores below 83 on the mental-development index: χ2 with 1 df = 4.13, P<0.05). On the psychomotor-development index, scores below 83 or below 68 were also more common after electronic fetal monitoring (23 percent [19 of 82] and 15 percent [12 of 82], respectively) than after periodic auscultation (16 percent [15 of 91] and 7 percent [6 of 91], respectively), but the differences were smaller.
At the four-month visit, the median number of risk points on the Movement Assessment of Infants was higher in the electronic-fetal-monitoring group than in the periodic-auscultation group (8 vs. 6, P = 0.048). Similarly, at the eight-month visit, the median number of risk points was higher in the electronic-fetal-monitoring group than in the periodic-auscultation group (11 vs. 8, P = 0.0088).
Cerebral palsy was diagnosed at 18 months in 20 percent of the children who had electronic fetal monitoring (16 of 82) and 8 percent of the children who had periodic auscultation (7 of 91) (χ2 with 1 df = 5.20, P<0.025). At the three study sites, the proportions of children with cerebral palsy after electronic fetal monitoring and periodic auscultation, respectively, were 21 percent (8 of 38) and 10 percent (4 of 41) at University Hospital, 18 percent (7 of 38) and 7 percent (3 of 46) at Grace Hospital, and 17 percent (1 of 6) and 0 (0 of 4) at Madigan Army Hospital (χ2 for heterogeneity = 0.18, P>0.5). Overall, the odds of cerebral palsy were 2.9 times higher after electronic fetal monitoring than after periodic auscultation (95 percent confidence interval, 1.2 to 7.3; P<0.025). After adjustment for potentially confounding antepartum and intrapartum risk factors, the odds ratio was 3.8 (95 percent confidence interval, 1.3 to 11.4; P<0.012).
Thirty-nine children with birth weights of 1750 g or less (23 in the electronically monitored group and 16 in the auscultated group) had no 18-month follow-up visit. Of these children, 16 had follow-up visits at four or eight months, and none had signs consistent with a diagnosis of cerebral palsy or more than eight risk points on the Movement Assessment of Infants. Among the 23 infants who had no follow-up visit, one neonate in the electronic-fetal-monitoring group had an intracranial hemorrhage of grade 3 or 4 and hydrocephalus. None of the other 22 infants who had no follow-up visits had an intracranial hemorrhage of grade 3 or 4, hydrocephalus, or seizures during the neonatal period.
The infants with cerebral palsy in the electronic-fetal-monitoring group were no different from those in the periodic-auscultation group with respect to the type and severity of cerebral palsy. The percentages of infants with each type of cerebral palsy who had electronic fetal monitoring were 56 percent (5 of 9) for spastic diplegia, 78 percent (7 of 9) for spastic hemiplegia, 75 percent (3 of 4) for spastic quadriplegia, and 100 percent (1 of 1) for spastic quadriplegia with athetosis (χ2 with 3 df = 1.60, P>0.5). The percentages of infants who had electronic fetal monitoring, according to severity of disease, were 71 percent (5 of 7) with severe disease, 67 percent (4 of 6) with moderate disease, and 70 percent (7 of 10) with mild disease (χ2 with 2 df = 0.03, P>0.9).
The medical records of the 16 infants with cerebral palsy who had electronic fetal monitoring were reviewed to ascertain whether special circumstances were related to the development of cerebral palsy. In only one pregnancy were abnormal fetal-heart-rate patterns followed up by pH testing of the fetal scalp that might have prompted a decision to delay and observe rather than deliver immediately. In this case, the scalp pH was 7.39, and the infant was delivered by forceps one half hour later. In another case, the fetal heart rate was recorded as difficult to detect and interpret during labor. A decision not to intervene was made during labor on the basis of the parents' feelings about the extreme prematurity of the fetus and the low likelihood of survival. This child had cerebral palsy. Another child in the electronic-fetal-monitoring group had congenital microcephaly.
The medical records of the seven infants with cerebral palsy who had periodic auscultation were also reviewed. One child had severe hydronephrosis and urinary anomalies that necessitated urethrotomy. The course was complicated by the collapse of the respiratory and circulatory systems, hyperbilirubinemia (354 μmol per liter), and necrotizing enterocolitis. Another child was beaten by a baby sitter at four months of age, resulting in a ruptured spleen and multiple fractures, including a skull fracture.
The follow-up of the infants with birth weights over 1750 g was not formally included in this study; nevertheless, because University Hospital and Grace Hospital are referral centers, we were aware of four children with birth weights over 1750 g in whom cerebral palsy developed after their participation in the study. One child weighing 1780 g with moderate diplegia had had electronic fetal monitoring. Two children (1760 and 1940 g) with severe quadriplegia had had periodic auscultation, as had a third child (1810 g) who had mild diplegia. This last child had normal scores on the Bayley indexes (mental development, 126; motor development, 107) but was classified as having cerebral palsy on the basis of increased tone and reflexes in both lower extremities and a diminished capacity for fine movement of the hand.
Among the surviving study infants seen at 18 months, 26 percent (21 of 82) had abnormal fetal-heart-rate patterns as diagnosed by a retrospective examination of electronic-fetal-monitoring tracings, and 16 percent (15 of 91) had abnormal fetal heart rates as diagnosed by a retrospective examination of the periodic-auscultation data sheets (Fig. 2Figure 2
Duration of Abnormal Fetal Heart Rates during Intrapartum Surveillance with Electronic Fetal Monitoring or Periodic Auscultation among Infants with and without Cerebral Palsy.). The 21 abnormal fetal-heart-rate patterns in the electronic-fetal-monitoring group (median, 40.5 minutes; range, 10 to 320) lasted approximately as long as the 15 abnormal patterns in the periodic-auscultation group (median, 45.5 minutes; range, 3 to 247). The time from the onset of the abnormal pattern to delivery was longer in the electronic-fetal-monitoring group (median, 104.5 minutes; range, 15 to 1312) than in the periodic-auscultation group (median, 60.5 minutes; range, 6 to 351).The duration of abnormal fetal-heart-rate patterns was related to the risk of cerebral palsy among the infants who had electronic fetal monitoring (Fig. 2). In this group, the risks of cerebral palsy on the basis of abnormal fetal heart rates lasting less than 1 minute, 1 to 10 minutes, 11 to 30 minutes, 31 to 90 minutes, or 91 minutes or more were 13 percent (8 of 61), 0 (0 of 1), 14 percent (1 of 7), 43 percent (3 of 7), and 67 percent (4 of 6), respectively (χ2 for trend = 11.86, P<0.001). Similarly, after the onset of abnormal fetal heart rates, the risks of cerebral palsy on the basis of times to delivery of 1 to 10 minutes, 11 to 30 minutes, 31 to 90 minutes, or 91 minutes or more were, respectively, 0 (0 of 0), 0 (0 of 3), 17 percent (1 of 6), and 58 percent (7 of 12) (χ2 for trend = 4.94, P<0.05). Two children in the periodic-auscultation group had cerebral palsy after the onset of abnormal fetal heart rates — one after 46 minutes and the other after 13 minutes.
Abnormal fetal-heart-rate patterns were a more frequent indication for delivery by cesarean section or forceps in the electronic-fetal-monitoring group (5 percent [4 of 82] and 4 percent [3 of 82], respectively) than in the periodic-auscultation group (3 percent [3 of 91] and 1 percent [1 of 91], respectively). When the indication for delivery was not considered, delivery by cesarean section or forceps occurred no more often in the electronic-fetal-monitoring group (12 percent [10 of 82] and 10 percent [8 of 82], respectively) than in the periodic-auscultation group (14 percent [13 of 91] and 8 percent [7 of 91], respectively). All cesarean sections except one (for severe preeclampsia in the periodic-auscultation group) occurred during advanced labor, when delivery within hours was inevitable.
Discussion
This randomized trial of fetal monitoring during premature labor suggests that the infants' neurologic development was not improved with electronic fetal monitoring and that such monitoring may have been less effective than a structured program of periodic auscultation. Another randomized clinical trial of electronic fetal monitoring assessed the neurologic outcomes of infants born at term and also found no improvement with electronic fetal monitoring.6 , 23 Although our initial hypothesis examined only the potential superiority of electronic fetal monitoring over periodic auscultation, in most comparisons the results of periodic auscultation were actually superior. The mean scores on the Bayley scales were higher in the periodic-auscultation group in every comparison we made: at 4, 8, and 18 months, and on the mental and psychomotor indexes (Table 3). Low scores on the Bayley scales (below 83) were more common after electronic fetal monitoring. In addition, there was an unanticipated 2.9-fold increase in the odds of having cerebral palsy among infants weighing 1750 g or less who had electronic fetal monitoring as compared with infants who had periodic auscultation. This increase was observed at each of the three study hospitals.
Overall, as compared with other American perinatal centers,24 our hospitals had a higher observed prevalence of cerebral palsy. This may be partly related to the early age at diagnosis (18 months, with correction for gestational age). However, later neurologic assessment is unlikely to reverse entirely the excess incidence of cerebral palsy that we observed with electronic fetal monitoring, since it was not restricted to mild cases.
The duration of abnormal fetal heart rates and the time from their onset to delivery were strongly related to the risk of cerebral palsy in the electronically monitored group. Whether earlier delivery may have averted untoward outcomes in some of these pregnancies is uncertain. In a case–control study of suboptimal responses to fetal distress (defined according to electronic-fetal-monitoring criteria), the risk of cerebral palsy was not increased.25 In the present study, the clinicians may have been reassured despite the presence of abnormal fetal heart rates, by certain inherent aspects of electronic fetal monitoring such as the paper record itself or by the base-line variability, which cannot be measured with periodic auscultation. The recognized low specificity of electronic fetal monitoring and its high likelihood of yielding a false positive diagnosis of fetal distress26 may also have led clinicians to deviate from the procedures for managing abnormal fetal-heart-rate patterns suggested by the protocol. Other investigators have expressed similar opinions.27
Our results could have been biased if the intrapartum monitoring technique had affected the attainment of a birth weight over 1750 g; however, there was no evidence that this occurred. The birth-weight distributions for infants of all 376 mothers randomly assigned to the two experimental groups differed little (two-tailed P = 0.87 by the Kolmogorov–Smirnov two-sample test), and the proportions of infants who had birth weights over 1750 g were similar in the electronic-fetal-monitoring group (36 percent [68 of 190]) and the periodic-auscultation group (33 percent [61 of 186]). In the electronic-fetal-monitoring group, all cesarean sections were performed during advanced labor, when delivery within hours was inevitable. Thus, it is unlikely that electronic monitoring could have increased (by the early cesarean delivery of infants at risk) the number of cases of cerebral palsy observed in infants weighing 1750 g or less (as opposed to higher birth weights). Four children with cerebral palsy and birth weights over 1750 g have come to our attention. The lack of follow-up data on most of the children with birth weights over 1750 g does not permit meaningful conclusions about the effects of intrapartum monitoring in this birth-weight range.
Attention should be given to other potential sources of bias that may explain our results. There was little evidence that the process of randomization failed, since the antepartum and intrapartum risk factors were equivalent for the mothers and infants studied in the electronic-fetal-monitoring group and the periodic-auscultation group.13 This was also generally true of the 173 surviving infants who had 18-month follow-up visits. Among these infants, two risk factors (premature rupture of membranes and administration of a tocolytic agent) differed between the two study groups. Simultaneous control for the potentially confounding effects of these and other antepartum and intrapartum risk factors had little effect on the results on the Bayley scales and produced somewhat elevated estimates of the risk of cerebral palsy in the electronic-fetal-monitoring group. Incomplete follow-up is unlikely to have affected our results, since no important differences in risk factors were apparent between the 173 infants who had follow-up visits and the 39 surviving infants who had no follow-up; only a single surviving study infant (who had electronic monitoring) with serious adverse neonatal events (intracranial hemorrhage of grade 3 or 4 with hydrocephalus) did not have a follow-up visit. In addition, there is little reason to expect that a particular technique of intrapartum monitoring would have a singular effect on infant follow-up. The neurologic assessments were protected from bias in ascertainment because the clinicians who performed the testing were unaware of the intrapartum monitoring technique used with each patient. After random assignment to periodic auscultation, six women in labor were withdrawn from the study and switched to continuous electronic fetal monitoring at the request of the physician or patient. There were no deaths or adverse neonatal outcomes among the infants of these six women; four of the infants had follow-up visits and were considered for purposes of data analysis to belong to the periodic-auscultation group to which they were assigned at randomization. The exclusion of these women and their infants from the analysis did not affect our results.
This study suggests that clinicians should be attentive to the potentially adverse outcomes associated with electronic fetal monitoring. We found that, as compared with a structured protocol of periodic fetal auscultation, electronic fetal monitoring did not improve the neurologic development of children born prematurely.
Supported in part by grants from the National Center for Health Services Research and Health Care Technology (R01 HT 00003) and the National Center for Health Services Research and Health Care Technology Assessment (R01 HS 04848).
Source Information
From the Departments of Obstetrics and Gynecology (K.K.S., D.A.L., J.A.W., M.A.S.), Pediatrics (F.C.B.), Internal Medicine (E.B.L.), and Biostatistics (G.v.B., J.P.H.), University of Washington, Seattle, and the Department of Pediatrics. Grace Hospital, Vancouver, B.C., Canada (M.W.). Address reprint requests to Dr. Shy at the Department of Obstetrics and Gynecology, RH-20, University of Washington, Seattle, WA 98195.
- References
References
1
Placek
Web of Science | Medline2
Hess
CrossRef | Web of Science | Medline3
Paul
Web of Science | Medline4
Gabert
Web of Science | Medline5
Neutra
Full Text | Web of Science | Medline6
Haverkamp
Web of Science | Medline7
Renou
Web of Science | Medline8
Kelso
Web of Science | Medline9
Haverkamp
Web of Science | Medline10
Wood
Web of Science | Medline11
MacDonald
Web of Science | Medline12
Leveno
Full Text | Web of Science | Medline13
Luthy
Web of Science | Medline14
Parer
CrossRef | Web of Science | Medline15
Parmelee
Web of Science | Medline16
Bayley N. Manual for Bayley scales of infant development. New York: Psychological Corporation, 1969.
17
Chandler LS, Andrews MS, Swanson MW. Movement assessment of infants. Seattle: University of Washington, 1980.
18
Bax
CrossRef | Medline19
Armitage P. Statistical methods in medical research. New York: John Wiley, 1971:118–23.
20
Breslow NE, Day NE. Statistical methods in cancer research. Vol. I. The analysis of case–control studies. Lyon, France: World Health Organization: International Agency for Research on Cancer, 1980. (IARC scientific publications no. 32.)
21
Mantel
CrossRef | Web of Science22
Zelen
Web of Science23
Murphy
Web of Science | Medline24
Nelson
Web of Science | Medline25
Niswander
CrossRef | Web of Science | Medline26
van den Berg
CrossRef | Medline27
Prentice
CrossRef | Web of Science | Medline
- Citing Articles (48)
Citing Articles
1
G Justus Hofmeyr, Jon F Barrett, Caroline A Crowther, G Justus Hofmeyr. 2011. Planned caesarean section for women with a twin pregnancy. .
CrossRef2
Lina F. Chalak, Dwight J. Rouse. (2011) Neuroprotective Approaches: Before and After Delivery. Clinics in Perinatology 38:3, 455-470
CrossRef3
D Ayres-de-Campos, D Arteiro, C Costa-Santos, J Bernardes. (2011) Knowledge of adverse neonatal outcome alters clinicians’ interpretation of the intrapartum cardiotocograph. BJOG: An International Journal of Obstetrics & Gynaecology 118:8, 978-984
CrossRef4
Molly J. Stout, Alison G. Cahill. (2011) Electronic Fetal Monitoring: Past, Present, and Future. Clinics in Perinatology 38:1, 127-142
CrossRef5
Alfredo G Esposto. (2009) Tort Reform, Defensive Medicine, and the Diffusion of Diagnostic Technologies. Eastern Economic Journal 34:2, 141-157
CrossRef6
Hamisu M. Salihu. (2008) Epidemiology of Stillbirth and Fetal Central Nervous System Injury. Seminars in Perinatology 32:4, 232-238
CrossRef7
Ira Adams-Chapman. (2008) Long-Term Neurologic Outcome of Infants Born by Cesarean Section. Clinics in Perinatology 35:2, 437-454
CrossRef8
C. Tempfer, L. Hefler, P. Husslein. (2007) Modern intrapartum fetal monitoring: room for improvement?. Archives of Gynecology and Obstetrics 276:2, 99-100
CrossRef9
Greene, Michael F., . (2006) Obstetricians Still Await a Deus ex Machina. New England Journal of Medicine 355:21, 2247-2248
Full Text10
Ernest M. Graham, Scott M. Petersen, Dana K. Christo, Harold E. Fox. (2006) Intrapartum Electronic Fetal Heart Rate Monitoring and the Prevention of Perinatal Brain Injury. Obstetrics & Gynecology 108:3, Part 1, 656-666
CrossRef11
Zarko Alfirevic, Declan Devane, Gillian ML Gyte, Zarko Alfirevic. 2006. Continuous cardiotocography (CTG) as a form of electronic fetal monitoring (EFM) for fetal assessment during labour. .
CrossRef12
Stephen B Thacker, Donna Stroup, Man-huei Chang, Stephen B Thacker. 2006. Continuous electronic heart rate monitoring for fetal assessment during labor. .
CrossRef13
Zarko Alfirevic, Andrew Weeks, Zarko Alfirevic. 2006. Oral misoprostol for induction of labour. .
CrossRef14
Nadav Schwartz, Bruce K. Young. (2006) Intrapartum fetal monitoring today. Journal of Perinatal Medicine 34:2, 99-107
CrossRef15
Nelson, Karin B., . (2003) Can We Prevent Cerebral Palsy?. New England Journal of Medicine 349:18, 1765-1769
Full Text16
H. David Banta, Stephen B. Thacker. (2001) Historical Controversy in Health Technology Assessment:. Obstetrical and Gynecological Survey 56:11, 707-719
CrossRef17
SB Thacker, D Stroup, M Chang, Sonja Henderson. 2001. Continuous electronic heart rate monitoring for fetal assessment during labor. .
CrossRef18
Ken L Bassett, Nitya Iyer, Arminee Kazanjian. (2000) Defensive medicine during hospital obstetrical care: a by-product of the technological age. Social Science & Medicine 51:4, 523-537
CrossRef19
Lisa M. Hollier. (2000) Can neurological injury be timed?. Seminars in Perinatology 24:3, 204-214
CrossRef20
J. Hornbuckle, A. Vail, K. R. Abrams, J. G. Thornton. (2000) Bayesian interpretation of trials: the example of intrapartum electronic fetal heart rate monitoring. BJOG: An International Journal of Obstetrics and Gynaecology 107:1, 3-10
CrossRef21
Eric H. Dellinger, Frank H. Boehm, Martin M. Crane. (2000) Electronic fetal heart rate monitoring: Early neonatal outcomes associated with normal rate, fetal stress, and fetal distress. American Journal of Obstetrics and Gynecology 182:1, 214-220
CrossRef22
Theodore G. Cheek, Stacy Lewin. (1999) New developments in fetal monitoring for the anesthesiologist. Current Opinion in Anaesthesiology 12:3, 289-293
CrossRef23
David Mechanic. (1999) Issues in promoting health. Social Science & Medicine 48:6, 711-718
CrossRef24
C. Loghis, E. Salamalekis, N. Vitorato. (1999) Umbilical cord blood gas analysis in augmented labour. Journal of Obstetrics & Gynaecology 19:1, 38-40
CrossRef25
William W.K. To, W.C. Leung. (1998) The Incidence of Abnormal Findings from Intrapartum Cardiotocogram Monitoring in Term and Preterm Labours. The Australian and New Zealand Journal of Obstetrics and Gynaecology 38:3, 258-261
CrossRef26
Jason Gardosi. (1998) Systematic reviews: insufficient evidence on which to base medicine. BJOG: An International Journal of Obstetrics and Gynaecology 105:1, 1-5
CrossRef27
Karin B. Nelson, Judith K. Grether. (1997) Cerebral palsy in low-birthweight infants: Etiology and strategies for prevention. Mental Retardation and Developmental Disabilities Research Reviews 3:2, 112-117
CrossRef28
Samuel Lurie, Ariel Weissman, Gari Blumberg, Zion Hagay. (1996) Fetal Oximetry Monitoring. Obstetrical & Gynecological Survey 51:8, 498-502
CrossRef29
(1996) Electronic Fetal Monitoring in Predicting Cerebral Palsy. New England Journal of Medicine 335:4, 287-289
Full Text30
Jason Gardosi. (1996) 10 Monitoring technology and the clinical perspective. Baillière's Clinical Obstetrics and Gynaecology 10:2, 325-339
CrossRef31
Luis A. Cibils. (1996) On intrapartum fetal monitoring. American Journal of Obstetrics and Gynecology 174:4, 1382-1389
CrossRef32
MacDonald, Dermot, . (1996) Cerebral Palsy and Intrapartum Fetal Monitoring. New England Journal of Medicine 334:10, 659-660
Full Text33
Marie-Josee Trepanier, Patricia Niday, Barbara Davies, Ann Sprague, Carl Nimrod, Corinne Dulberg, Nancy Watters. (1996) Evaluation of a Fetal Monitoring Education Program. Journal of Obstetric, Gynecologic, <html_ent glyph="@amp;" ascii="&"/> Neonatal Nursing 25:2, 137-144
CrossRef34
Ken Bassett. (1996) Anthropology, clinical pathology and the electronic fetal monitor: Lessons from the heart. Social Science & Medicine 42:2, 281-292
CrossRef35
Jørgen Falck Larsen. (1996) Why has conventional intrapartum cardiotocography not given the expected results?. Journal of Perinatal Medicine 24:1, 15-23
CrossRef36
Mary E. D'Alton, Lynn L. Simpson. (1995) Syndromes in twins. Seminars in Perinatology 19:5, 375-386
CrossRef37
P GROVES, N ORIOL. (1995) How useful is intrapartum electronic fetal heart rate monitoring?. International Journal of Obstetric Anesthesia 4:3, 161-167
CrossRef38
Robert C. Goodlin. (1995) Do concepts of causes and prevention of cerebral palsy require revision?. American Journal of Obstetrics and Gynecology 172:6, 1830-1836
CrossRef39
Barry S. Schifrin. (1994) To the Editor:. Birth 21:4, 236-237
CrossRef40
Kuban, K.C.K.Leviton, Alan. (1994) Cerebral Palsy. New England Journal of Medicine 330:3, 188-195
Full Text41
Irene H. Butter. (1993) Premature Adoption and Routinization of Medical Technology: Illustrations from Childbirth Technology. Journal of Social Issues 49:2, 11-34
CrossRef42
Malcolm Smith, Richard Simon, Dennis Cain, Richard S. Ungerleider. (1993) Children and cancer. A perspective from the cancer therapy evaluation program, national cancer institute. Cancer 71:S10, 3422-3428
CrossRef43
David H. Chestnut. (1993) Fetal monitoring and anaesthesia for fetal distress. Canadian Journal of Anaesthesia 40:S1, R74-R80
CrossRef44
Denise M. Oleske, Gerald L. Glandon, Daniel J. Tancredi, Mehdi Nassirpour, John R. Noak. (1992) Information Dissemination and the Cesarean Birth Rate: The Illinois Experience. International Journal of Technology Assessment in Health Care 8:04, 708
CrossRef45
Joanne Guay. (1991) Monitorage du fœtus et réanimation du nouveau-né: ce que l’anesthésiste doit savoir. Canadian Journal of Anaesthesia 38:S1, R74-R88
CrossRef46
M. S. Rogers, A. M. Z. Chang. (1991) Perinatal asphyxia: A Bayesian analysis of prediction and prevention. Journal of Obstetrics & Gynaecology 11:1, 34-40
CrossRef47
(1990) Electronic Fetal-Heart-Rate Monitoring as Compared with Periodic Auscultation and the Neurologic Development of Premature Infants. New England Journal of Medicine 323:5, 345-348
Full Text48
Freeman, Roger, . (1990) Intrapartum Fetal Monitoring — A Disappointing Story. New England Journal of Medicine 322:9, 624-626
Full Text
- Letters
