Intrapartum fetal monitoring aims to identify fetuses at risk of neonatal and long-term injury resulting from asphyxia. Fetal surveillance with cardiotocography was introduced in the 1960s. Although a positive effect of cardiotocography on neonatal outcomes never has been shown, cardiotocography is widely used.1–8
Fetal monitoring by cardiotocography significantly increases the operative delivery rate, but the addition of fetal blood sampling may prevent this.6 However, performance of fetal blood sampling requires expertise, is invasive, has to be repeated when cardiotocographic abnormalities persist, and still does not guarantee prevention of asphyxia.9 Other tools for fetal surveillance, eg, fetal pulse oximetry, have not been successful.10
In recent years, ST analysis of the fetal electrocardiogram (ECG; STAN, Neoventa Medical, Gothenburg, Sweden) has been introduced.11 This technique detects changes in the ST segment of the fetal ECG, which are related to metabolic acidosis. These are interpreted together with the cardiotocography.12,13
Two randomized trials assessing ST analysis of the fetal ECG showed that this technique decreased metabolic acidosis, instrumental deliveries for fetal distress, and the proportion of neonates born with encephalopathy as compared with cardiotocography alone.14–16 However, in both trials, fetal blood sampling was equally performed in the two study groups. In a recent much smaller trial, the fetal blood sampling percentage was significantly lower in the cardiotocography plus ST analysis group, but improvement in neonatal outcome could not be confirmed.17
Given this controversy, we designed a large pragmatic randomized trial with the aim to quantify the effectiveness of intrapartum fetal monitoring by cardiotocography plus ST analysis of the fetal ECG using a strict protocol for performance of fetal blood sampling.18
MATERIALS AND METHODS
We performed a multicenter randomized pragmatic trial in three academic and six nonacademic teaching hospitals in The Netherlands. The study was situated within the Dutch Obstetric Consortium, which is a collaboration of obstetric clinics in The Netherlands (www.studies-obsgyn.nl).
Before our study, cardiotocography and fetal blood sampling was the standard of care in The Netherlands for fetal surveillance of high-risk deliveries. Before the start of the trial, labor ward personnel of the participating centers was trained in the use of the STAN method by four instructors. After the training, all gynecologists, residents, and midwives in the participating centers became certified STAN-users by passing an examination (STAN training material, Neoventa Medical). Training and certification continued on a regular basis during the trial. All participating centers had at least 2 months of clinical experience with the STAN method before randomization of the first patient.
Laboring women aged 18 years or older with a singleton high-risk pregnancy, a fetus in cephalic presentation, a gestational age greater than 36 weeks, and an indication for internal electronic fetal monitoring were eligible. In The Netherlands, pregnant women at low risk are monitored by midwives or general practitioners at home or in the hospital (primary care), whereas pregnant women at high risk are monitored by gynecologists in the hospital (secondary care). High-risk pregnancies are complicated by preexisting maternal disease, complicated obstetric history, hypertensive disorders, intrauterine growth restriction, ruptured membranes for more than 24 hours, a postdate gestational age, failure to progress, need for pain relief, meconium-stained amniotic fluid, or nonreassuring fetal heart rate at intermittent auscultation by a midwife.
The study was approved by the Institutional Review Board of the University Medical Center Utrecht and had local approval from all other participating hospitals. Eligible women received both oral and written information either at 36 weeks of gestation in the outpatient clinic or in the early stage during labor. After women had given written informed consent, they were randomized on a 1:1 basis through a web-based computer-generated randomization sequence with variable block size to either monitoring by cardiotocography plus ST analysis of the fetal ECG (index group) or cardiotocography alone (control group). Randomization was stratified for center and parity (no compared with one or more previous vaginal deliveries). The (assigned) diagnostic procedure and subsequent labor were managed by the labor ward staff, including gynecologists, residents, and midwives. As a result of the explicit pragmatic nature of the trial, both patients and care givers were not blinded to the allocated interventions.
In women assigned to the index group, a scalp electrode was applied to the fetal head and connected to a STAN S21 or S31 fetal heart monitor. This scalp electrode allowed both for registration of the cardiotocography and ST analysis of the fetal ECG. The cardiotocography was classified according to the STAN clinical guidelines (based on the guidelines of the International Federation of Gynecology and Obstetrics).19 Clinical management was supported by the computerized automatic ST interval assessment, the ST log, and the STAN clinical guidelines (Appendices 1 and 2, available online at http://links.lww.com/AOG/A178).18
Performance of fetal blood sampling was restricted to three situations: 1) start of STAN registration with an intermediary or abnormal cardiotocography trace; 2) abnormal cardiotocography trace for more than 60 minutes during the first stage without ST events; and 3) poor ECG signal quality in the presence of an intermediary or abnormal cardiotocography trace. Poor signal quality was defined as absence of ST information for more than 4 minutes or less than one average ECG complex per minute. If the pH of a fetal blood sampling was below 7.20, immediate delivery was recommended unless the cause of fetal distress could be alleviated. If the pH was between 7.20 and 7.25, the advice was to repeat fetal blood sampling after 30 minutes. If the pH was above 7.25, the fetal condition was considered well enough to follow the STAN clinical guidelines.
In women assigned to the control group, a scalp electrode was applied to the fetal head and connected to a conventional fetal heart rate monitor. The cardiotocography was classified and interpreted according to the STAN clinical guidelines as normal, intermediary, abnormal, or (pre)terminal.19 Fetal blood sampling was performed on indication by the obstetric caregiver in case of an intermediary or abnormal cardiotocography trace. Clinical decisions were based on cardiotocography, fetal blood sampling, or both results using the same thresholds for interventions as in the index group. In case of a fetal blood sampling result with pH above 7.25, fetal blood sampling could be repeated according to cardiotocography patterns on discretion of the caregiver.
In both groups, immediately after birth, the umbilical cord was doubly clamped to sample both arterial and venous cord blood. In case arterial and venous pH in cord blood differed less than 0.03, the results were considered to be from a venous sample, because the latter is more easily obtained.
For both groups, protocol monitoring was achieved by quarterly case meetings in which controversial cases were discussed and feedback was given on potential protocol violations. In each center, a research nurse or midwife and a gynecologist were responsible for training, study monitoring, and data entry into a Web-based database.
Primary outcome measure was the incidence of metabolic acidosis, defined as an umbilical cord artery blood pH below 7.05 and a base deficit calculated in the extracellular fluid compartment (BDecf) above 12 mmol/L according to the Siggaard-Andersen acid-base chart algorithm.20,21
We also defined metabolic acidosis as an umbilical cord artery blood pH below 7.05 combined with a base deficit calculated in blood (BDblood) above 12 mmol/L.
Base deficit in blood is reported by most umbilical cord blood analyzers, and therefore these values are often used in clinical practice. However, BDecf seems to better reflect the true metabolic component of acidosis, which is associated with neonatal morbidity.21–23
Other secondary outcomes were the number of Apgar scores below 4 and 7 at 1 and 5 minutes, respectively, total neonatal admissions, admissions to a neonatal intensive care unit (NICU, Level III), operative deliveries (cesarean delivery, instrumental vaginal delivery, or both), and number of cases with fetal blood sampling.
Two neonatologists (F.G. and M.J.B.) who were blinded to randomization allocation independently assessed all neonatal admission letters and charts to evaluate whether signs of moderate or severe neonatal hypoxic–ischemic encephalopathy had developed according to Sarnat and Sarnat.24 Because retrospective application of the Sarnat grading system causes large variability regarding the evaluation of mild encephalopathy (Sarnat grade 1), without a strong relation of this grade to adverse neurologic outcome,25 we restricted grading to moderate and severe hypoxic–ischemic encephalopathy (Sarnat grades 2 and 3, respectively).
Safety monitoring was performed by a Data Safety Monitoring Committee.18 Three safety conditions were monitored: 1) umbilical cord artery blood pH below 7.00 and a base deficit above 12 mmol/L, 2) Apgar score below 7 after 5 minutes, and 3) admission to a NICU. Occurrence of either condition 1 and 3 or condition 2 and 3 was defined as a serious adverse event.
We hypothesized an incidence of our primary outcome of 3.5%.26 An absolute reduction in the metabolic acidosis rate of 1.4% in favor of the index group was considered clinically relevant.14,15,17 Using a two-sided alpha of 0.05 and power of 0.80 implied randomization of 4,638 women (2,319 per group). Allowing for 10% loss to follow-up, our target sample size was 5,200 women.
The statistical analyses were performed according to intention to treat. For dichotomous outcomes, the relative risk (RR) with 95% confidence interval (CI) was estimated, adjusted for the stratified randomization by center and parity, by including the latter variables in the multivariable regression analysis. For the primary outcome, the number needed to treat was also calculated.
Various subjects had missing values. Because these are often selectively missing, which was also the case in our study (Appendix 3, available online at http://links.lww.com/AOG/A178), it is well documented that a complete case analysis likely yields biased results.27–29 Hence, we multiply imputed missing values (10 times) before doing the analysis, using the AregImpute method in S-plus (Insightfull Corp, Seattle, WA). Results of these analyses on the 10 imputed data sets were then pooled according to standard methods using Rubin's rule.30 All analyses including the multiple imputation were performed in S-plus 6.1.
The study was performed between January 2006 and July 2008. During the trial, we monitored whether umbilical cord blood samples were adequately performed and outcome results were available. Because these data appeared to be incomplete for 20% (instead of the assumed 10%), the trial was extended to randomization of 5,681 women, a decision made before any comparison of groups.
We randomly allocated 2,832 women to the index group and 2,849 to the control group. After randomization, 14 women were excluded (five in the index group, nine in the control group), because they did not meet the inclusion criteria. Data for 5,667 women (2,827 in the index group, 2,840 in the control group) were analyzed according to intention to treat (Fig. 1). Baseline characteristics of these women are summarized in Table 1.
Fetal blood sampling was performed less in the index group (10.6%) than in the control group (20.4%) (RR 0.52, 95% CI 0.46–0.59). The overall rates of cesarean or instrumental vaginal deliveries were comparable (RR 0.96, 95% CI 0.87–1.06) with slightly more operative deliveries for suspected fetal distress in the index group, whereas there were slightly less operative deliveries for other indications in this group (Table 2).
Neonatal outcomes are shown in Table 3. The incidence of the primary outcome metabolic acidosis based on pH and BDecf was lower in the index group (0.7% compared with 1.1%) compared with the control group (RR 0.70, 95% CI 0.38–1.28, number needed to treat 252).
The rate of metabolic acidosis based on pH and BDblood was significantly lower in the index group than in the control group (1.6% compared with 2.6%; RR 0.63, 95% CI 0.42–0.94, number needed to treat 100). For an umbilical cord artery pH below 7.05 as well as 7.00, there were also reduced incidences in the index group (RR 0.67, 95% CI 0.46–0.97 and RR 0.56, 95% CI 0.31–1.01, respectively). The rates of low Apgar scores after 1 and 5 minutes, total neonatal admissions, and admissions to a NICU did not differ between groups (Table 3).
In total, there were four newborns with signs of moderate or severe hypoxic–ischemic encephalopathy (0.1%). In the index group, two newborns were graded with a Sarnat 2 and one with a Sarnat 3 (also a case of perinatal death). In the control group, one newborn had Sarnat grade 2 (Table 3).
In total, there were five perinatal deaths (Table 3). Three (one in the index group and two in the control group) were caused by congenital malformations (one transposition of the great arteries, one intracranial teratoma, and one hypoplastic left heart syndrome). The other two deaths were both in the index group. One woman, with a history of cesarean delivery, had multiple fetal blood sampling for persisting abnormal cardiotocography and two significant ST events. After 90 minutes of active pushing, a ventouse delivery for fetal distress failed and a cesarean delivery was performed. A newborn with Apgar scores of 0, 0, and 3 after 1, 5, and 10 minutes, respectively, was delivered. Rupture of the uterus appeared to be the cause of distress. The newborn was admitted to the NICU and died of severe perinatal asphyxia and neonatal encephalopathy (Sarnat grade 3). In the second case, immediately after applying a scalp electrode for internal fetal monitoring, the cardiotocography showed a (pre)terminal trace, whereas it had been normal before. At emergency cesarean delivery, a severely asphyxiated newborn with Apgar scores of 0 after 1, 5, and 10 minutes was born.
Our trial shows that fetal monitoring by cardiotocography combined with ST analysis of the fetal ECG decreases the incidence of acidosis by 30% to 44% depending on its definition. However, the reduction in our primary outcome is not significant as a result of a low incidence of metabolic acidosis calculated in the extracellular fluid compartment in both groups. Moreover, we do not find an effect on Apgar scores, neonatal admissions, moderate to severe hypoxic–ischemic encephalopathy incidence, or operative deliveries.
Our results were achieved with a 48% lower incidence of fetal blood sampling in the index group, which is in accordance with previous trials.17,31,32 Fetal blood sampling is a relatively invasive procedure, which has to be repeated as cardiotocography abnormalities persist. Because the results of our trial imply that less fetal blood sampling performance is needed in addition to ST analysis of the fetal ECG, this may be considered a positive effect.
We did not find a significant reduction in total operative deliveries, unlike some of the previous trials, which showed an 8–12% reduction in all operative deliveries.14,15,32 This may be attributable to the fact that in our study with a relatively high incidence of fetal blood sampling, especially in the control group, the number of unnecessary interventions resulting from false-positive cardiotocography results was low in both groups. It may be speculated that a randomized trial in a routine setting without fetal blood sampling such as in the United States indeed results in a considerable reduction of interventions.
Apart from a trend toward fewer total neonatal admissions in the index group, we found no effect on clinical neonatal outcomes such as Apgar scores or hypoxic–ischemic encephalopathy. It is known that most neonates born at term with low pH do well and associations between pH at birth and neonatal morbidity and mortality are found consistently only when the cord artery pH is below 7.00.23,33,34 In such cases, approximately 10% of children will develop neonatal seizures and approximately one third of those will have long-term impairment.35–38 In our study, only 18 newborns in the index group (0.6%) and 34 in the control group (1.2%) had umbilical cord artery pH below 7.00, thus limiting the power to detect asphyxia-related cerebral injury. The long-term neurodevelopmental outcome of patients in our ongoing follow-up study therefore has to be awaited.
To appreciate the present results, three issues need to be addressed. First, our trial was powered on the assumption that 3.5% of neonates would have a metabolic acidosis at birth.26 However, the overall incidence, specifically in the extracellular fluid compartment, was considerably lower. This may probably be explained by the fact that women diagnosed with acute signs of fetal distress at admission were not approached for participation.
Second, as described, management of obstetric care in The Netherlands is different than in other countries. Because fetal surveillance is only applied in “high-risk” women, our results may be less applicable to practice in other countries. However, the indications for fetal monitoring in our trial varied from relatively “low risk” (need for pain relief) to “higher risk” (meconium-stained amniotic fluid). Moreover, the incidence of acidosis in our trial, although lower than expected, resembles that in other countries and studies, and a trend to a lower acidosis rate using ST analysis was observed in all subgroups.
The third issue concerns the definition of metabolic acidosis. In clinical practice, the definition of metabolic acidosis is usually based on a base deficit threshold of 12 mmol/L.39 Because BDecf is lower than BDblood,21 this will lead to a substantially lower rate of metabolic acidosis when the algorithm for BDecf is used. Although BDecf seems to better reflect the true metabolic component of acidosis,21–23 there is no consensus regarding which type of algorithm should be used. This does not only create clinical difficulties with respect to the definition of acidosis, but also introduces the problem of (in)comparability of study results when cord blood gases are used as outcomes.21
Data from four earlier trials have not been conclusive showing ambiguous results with a lower incidence of metabolic acidosis in the ST group in the two large trials14,15 and a higher incidence of metabolic acidosis in the two smaller trials.17,31 Metaanalysis of these trials, including 9,671 women, showed reduction in metabolic acidosis with a RR of 0.73 (95% CI 0.49–1.09) in favor of ST analysis,32 which is similar to our findings. Both the results of this metaanalysis and our study indicate that the use of the STAN method reduces the incidence of acidosis, at least in settings in which fetal blood sampling is being used.
In conclusion, we found that intrapartum fetal monitoring by cardiotocography plus ST analysis of the fetal ECG substantially decreases the incidence of (metabolic) acidosis. There was no effect on Apgar scores, neonatal admissions, moderate to severe hypoxic–ischemic encephalopathy, or operative deliveries. Also, long-term outcome has to be awaited before a final judgment on the value of ST analysis of the fetal ECG can be made.
1. Hornbuckle J, Vail A, Abrams KR, Thornton JG. Bayesian interpretation of trials: the example of intrapartum electronic fetal heart rate monitoring. BJOG 2000;107:3–10.
2. Parer JT, King T. Fetal heart rate monitoring: is it salvageable? Am J Obstet Gynecol 2000;182:982–7.
3. Thacker SB, Stroup DF, Peterson HB. Efficacy and safety of intrapartum electronic fetal monitoring: an update. Obstet Gynecol 1995;86:613–20.
4. Low JA, Pickersgill H, Killen H, Derrick EJ. The prediction and prevention of intrapartum fetal asphyxia in term pregnancies. Am J Obstet Gynecol 2001;184:724–30.
5. Murphy KW, Johnson P, Moorcraft J, Pattinson R, Russell V, Turnbull A. Birth asphyxia and the intrapartum cardiotocograph. Br J Obstet Gynaecol 1990;97:470–9.
6. Grant A. Monitoring of the fetus during labour. In: Chalmers I, Enkin M, Keirse MJNC, editors. Effective care in pregnancy and childbirth. Oxford (UK): Oxford University Press; 1991:846–82.
7. Nelson KB, Dambrosia JM, Ting TY, Grether JK. Uncertain value of electronic fetal monitoring in predicting cerebral palsy. N Engl J Med 1996;334:613–8.
8. Alfirevic Z, Devane D, Gyte GM. Continuous cardiotocography (CTG) as a form of electronic fetal monitoring (EFM) for fetal assessment during labor. Cochrane Database Syst Rev 2006;3:CD006066.
9. Westgate J, Greene K. How well is fetal blood sampling used in clinical practice? Br J Obstet Gynaecol 1994;101:250–1.
10. Bloom SL, Spong CY, Thom E, Varner MW, Rouse DJ, Weininger S, et al. Fetal pulse oximetry and cesarean delivery. N Engl J Med 2006;355:2195–202.
11. Rosen KG, Luzietti R. Intrapartum fetal monitoring: its basis and current developments. Prenatal and Neonatal Medicine 2000;5:155–68.
12. Rosen KG, Dagbjartsson A, Henriksson BA, Lagercrantz H, Kjellmer I. The relationship between circulating catecholamines and ST waveform in the fetal lamb electrocardiogram during hypoxia. Am J Obstet Gynecol 1984;149:190–5.
13. Westgate JA, Bennet L, Brabyn C, Williams CE, Gunn AJ. ST waveform changes during repeated umbilical cord occlusions in near-term fetal sheep. Am J Obstet Gynecol 2001;184:743–51.
14. Westgate J, Harris M, Curnow JS, Greene KR. Plymouth randomized trial of cardiotocogram only versus ST waveform plus cardiotocogram for intrapartum monitoring in 2400 cases. Am J Obstet Gynecol 1993;169:1151–60.
15. Amer-Wåhlin I, Hellsten C, Norén H, Hagberg H, Herbst A, Kjellmer I, et al. Cardiotocography only versus cardiotocography plus ST analysis of fetal electrocardiogram for intrapartum fetal monitoring: a Swedish randomized controlled trial. Lancet 2001;358:534–8.
16. Norén H, Amer-Wåhlin I, Hagberg H, Herbst A, Kjellmer I, Marşál K, et al. Fetal electrocardiography in labor and neonatal outcome: data from the Swedish randomized controlled trial on intrapartum fetal monitoring. Am J Obstet Gynecol 2003;188:183–92.
17. Ojala K, Vaarasmaki M, Makikallio K, Valkama M, Tekay A. A comparison of intrapartum automated fetal electrocardiography and conventional cardiotocography—a randomized controlled study. BJOG 2006;113:419–23.
18. Westerhuis ME, Moons KG, van Beek E, Bijvoet SM, Drogtrop AP, van Geijn HP, et al. A randomized clinical trial on cardiotocography plus fetal blood sampling versus cardiotocography plus ST-analysis of the fetal electrocardiogram (STAN) for intrapartum monitoring. BMC Pregnancy Childbirth 2007;7:13.
19. FIGO guidelines for the use of fetal monitoring. Int J Gynaecol Obstet 1987;25:159–67.
20. Siggaard-Andersen O. An acid-base chart for arterial blood with normal and pathophysiological reference areas. Scand J Clin Lab Invest 1971;27:239–45.
21. Wiberg N, Källén K, Olofsson P. Base deficit estimation in umbilical cord blood is influenced by gestational age, choice of fetal fluid compartment, and algorithm for calculation. Am J Obstet Gynecol 2006;195:1651–6.
22. Low JA, Panagiotopoulos C, Derrick EJ. Newborn complications after intrapartum asphyxia with metabolic acidosis in the term fetus. Am J Obstet Gynecol 1994;170:1081–7.
23. MacLennan A. A template for defining a causal relation between acute intrapartum events and cerebral palsy: international consensus statement. BMJ 1999;319:1054–9.
24. Sarnat HB, Sarnat MS. Neonatal encephalopathy following fetal distress. A clinical and electroencephalographic study. Arch Neurol 1976;33:696–705.
25. van Handel M, Swaab H, de Vries LS, Jongmans MJ. Long-term cognitive and behavioral consequences of neonatal encephalopathy following perinatal asphyxia: a review. Eur J Pediatr 2007;166:645–54.
26. Kwee A, van der Hoorn-van den Beld CW, Veerman J, Dekkers AH, Visser GHA. STAN S21 fetal heart monitor for fetal surveillance during labor: an observational study in 637 patients. J Matern Fetal Neonatal Med 2004;15:400–7.
27. van der Heijden GJ, Donders AR, Stijnen T, Moons KG. Imputation of missing values is superior to complete case analysis and the missing-indicator method in multivariable diagnostic research: a clinical example. J Clin Epidemiol 2006;59:1102–9.
28. Donders AR, van der Heijden GJ, Stijnen T, Moons KG. Review: a gentle introduction to imputation of missing values. J Clin Epidemiol 2006;59:1087–91.
29. Greenland S, Finkle WD. A critical look at methods for handling missing covariates in epidemiologic regression analyses. Am J Epidemiol 1995;142:1255–64.
30. Little RJA, Rubin DB. Statistical analysis with missing data. New York (NY): Wiley; 1987.
31. Vayssière C, David E, Meyer N, Haberstitch R, Sebahoun V, Roth E, et al. A French randomized controlled trial of ST-segment analysis in a population with abnormal cardiotocograms during labor. Am J Obstet Gynecol 2007;197:299.e1–6.
32. Neilson JP. Fetal electrocardiogram (ECG) for fetal monitoring during labour. Cochrane Database of Syst Rev 2006;3:CD000116..
33. Kirkendall C, Phelan JP. Severe acidosis at birth and normal neurologic outcome. Prenat Neonat Med 2001;6:267–70.
34. van de Riet JE, Vandenbussche FP, Le Cessie S, Keirse MJ. Newborn assessment and long-term adverse outcome: a systematic review. Am J Obstet Gynecol 1999;180:1024–9.
35. Goldaber KG, Gilstrap LC, Leveno KJ, Dax JS, McIntire DD. Pathologic fetal acidemia. Obstet Gynecol 1991;78:1103–7.
36. van den Berg PP, Nelen WL, Jongsma HW, Nijland R, Kollée LA, Nijhuis JG, Eskes TK. Neonatal complications in newborns with an umbilical artery pH < 7.00. Am J Obstet Gynecol 1996;175:1152–7.
37. Dennis J, Chalmers I. Very early neonatal seizure rate: a possible epidemiological indicator of the quality of perinatal care. Br J Obstet Gynaecol 1982;89:418–26.
38. Lavrijsen SW, Uiterwaal CS, Stigter RH, de Vries LS, Visser GH, Groenendaal F. Severe umbilical cord acidemia and neurological outcome in preterm and full-term neonates. Biol Neonate 2005;88:27–34.
39. Low JA, Lindsay BG, Derrick EJ. Threshold of metabolic acidosis associated with newborn complications. Am J Obstet Gynecol 1997;177:1391–4.