Non-invasive intrapartum fetal ECG: preliminary report

BJOG: an International Journal of Obstetrics and Gynaecology August 2005, Vol. 112, pp. 1016 –1021

DOI: 10.1111 /j .1471-0528 .2005.0 0643.x

Non-invasive intrapartum fetal ECG: preliminary report Myles J.O. Taylor,a Matthew J. Thomas,b Mark J. Smith,b Salome Oseku-Afful,a Nicholas M. Fisk,a Andrew R. Green,b Sara Paterson-Brown,a Helena M. Gardinera Objectives To obtain fetal heart rate, detailed fetal electrocardiography (f ECG) signals and uterine contractions during labour using a single device. Design Prospective observational study. Setting Delivery suite at a tertiary referral hospital, London, UK. Population Fifteen patients at median gestation of 39 weeks (range 24– 41) were recruited at median cervical dilatation of 4.0 cm (range 0 –10) of whom 8/15 (53%) had intact amniotic membranes. Methods Using 12 abdominally sited electrodes, we recorded the composite abdominal signal in pregnancies intrapartum. The recorded data were analysed off-line using a blind signal separation technique. Main outcome measures Success of signal separation and f ECG time intervals. Results Successful f ECG signal acquisition was achieved in 12/15 (80%) patients and an averaged fECG waveform acquired. In these patients, P and QRS waves were seen in all cases, and T waves in 11/12 (92%). True beat-to-beat heart rate (HR) was displayed and measures of its variability obtained. The mother’s ECG and uterine electrical activity, shown to match tocographically recorded uterine contractions, were also separated and displayed. Failure to acquire fECG in three cases was attributed to excessive abdominal muscular activity and electrical interference. Conclusions This study demonstrates a non-invasive technique that displays detailed intrapartum fECG waveforms, HR variability, maternal ECG and uterine contractions simultaneously, all in a single device and which avoids the potential risks of invasive monitoring with a fetal scalp electrode.

INTRODUCTION Birth is a hazardous process. Intrapartum mortality in the UK is 1 in 1600 live births1 and the risk of newborn encephalopathy secondary to intrapartum hypoxia, with its attendant risk of long term disability,2 is greater at 1 in 260 births.3 In the UK each damaged baby costs the Clinical Negligence Scheme for Trusts (CNST) approximately £3.5 million, and the National Health Service currently has potential liabilities for clinical negligence of £6 billion (http://www.nhsla.com), two-thirds of which is in obstetrics. The widely held concept that intrapartum events account for a minority of brain injury at term has been challenged by a recent population-based study4 where birth asphyxia was thought the likely cause of cerebral palsy in 28% and further supported by brain MRI and postmortem examinations5 where evidence of acute brain insult

a

Queen Charlotte’s and Chelsea Hospital and Division of Paediatrics, Obstetrics and Gynaecology, Faculty of Medicine, Imperial College, London, UK b Advanced Signal and Information Processing Group, QinetiQ, Malvern, Worcestershire, UK Correspondence: Mr M. J. O. Taylor, Royal Devon and Exeter, Hospital (Heavitree), Gladstone Road, Exeter, EX1 2ED, UK. D RCOG 2005 BJOG: an International Journal of Obstetrics and Gynaecology

was seen in 77% of term infants with neonatal encephalopathy or seizures. Only 1% showed evidence of lesions predating the onset of labour thus highlighting the potential impact of effective intrapartum monitoring. Current intrapartum monitoring uses either intermittent auscultation of the fetal heart or continuous electronic monitoring by cardiotocography (CTG). CTG uses Doppler ultrasound to demonstrate a fetal heart rate which is not a true beat-to-beat heart rate but an average over three neighbouring beats, analysed to give baseline rate, baseline variability and periodic changes. Abnormal heart rate patterns indicative of fetal hypoxaemia prompt obstetric interventions such as caesarean section or assisted vaginal delivery to prevent birth asphyxia. The expectation that CTG monitoring would reduce the incidence of deaths or brain damage in labour has not been realised.6,7 Problems include the low reproducibility of CTG interpretation8 and inadequate staff response to changes.1 As CTG is recorded from the labouring mother by cumbersome belts, technical problems occur frequently and resultant inadequate quality traces lead to confusion of fetal with maternal heart rate and halving or doubling of fetal heart rates. The low predictive value of abnormal CTG traces for fetal acidemia is explained in part by the lack of a genuine beat-to-beat recording and any fetal electrocardiographic (f ECG) waveform detail. Inadequate monitoring has been partly responsible for the rise in overall caesarean section rates www.blackwellpublishing.com/bjog

NON-INVASIVE INTRAPARTUM FETAL ECG

1017

Table 1. Obstetric and fetal f ECG data. Case

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Gestation (weeks)

Cervix (cm)

39 38 40 38 39 39 38 41 41 41 40 40 42 24 26

8 4 4 6 2 5 4 6 5 3 8 2 10 0 0

SROM

Analgesia

Mode of delivery

BL HR (bpm)

CoV

RV

PR (ms)

QRS (ms)

QT (ms)

QTc (ms)

T/QRS ratio

No No No Yes Yes Yes Yes Yes Yes No No No Yes No No

Ep Ep Ep Ep Ep Ep Ent Ep Ep None Ep Ent Ep None None

SVD SVD SVD LSCS Forceps Forceps SVD SVD SVD SVD SVD Forceps SVD n/a n/a

94 – 147 144 162 124 154 – 143 138 134 – 129 158 147

0.35 – 0.03 0.06 0.02 0.19 0.02 – 0.05 0.01 0.02 – 0.01 0.03 0.02

0.74 – 0.15 0.22 0.1 1.8 0.12 – 0.17 0.04 0.1 – 0.07 0.17 0.1

114 – 105 104 86 97 100 – 93 102 96 – 141 98 87

61 – 47 53 47 69 60 – 51 63 55 – 53 39 85

234 – 254 233 238 261 245 – 249 274 329 – 245 – 255

292 – 397 360 391 375 392 – 384 415 491 – 359 – 399

0.03 – 0.06 0.1 0.06 0.12 0.04 – 0.10 0.14 0.65 – 0.86 – 0.07

SROM ¼ spontaneous rupture of membranes; BLHR ¼ baseline heart rate; CoV ¼ coefficient of variation; RV ¼ relative variation; QTc ¼ corrected QT interval; Ep ¼ epidural; SVD ¼ spontaneous vaginal delivery; LSCS ¼ lower segment caesarean section; Ent ¼ entonox; n/a ¼ not applicable.

Fig. 1. Summary data obtained from a f ECG recording.

D RCOG 2005 BJOG: an International Journal of Obstetrics and Gynaecology 112, pp. 1016 – 1021

1018

M.J.O. TAYLOR ET AL.

from 4.3% in 1970 to 21.3% in 2000 in England9 and prompted the development of intrapartum f ECG recordings10,11 allowing assessment of true heart rate variability (HRV; from the R-R interval), and additional physiological information from the PR interval and ST segment. Intrapartum ST analysis in conjunction with standard CTG monitoring is associated with a 44% reduction in birth asphyxia12 and significantly reduced rates of operative delivery for suspected fetal distress. The major shortcoming of f ECG techniques in labour is their invasiveness, relying on a direct contact needle-like fetal scalp electrode (FSE) to obtain the f ECG signal.13 Fetal scalp injury and perinatal infection from membrane rupture required to apply the electrode may occur particularly in regions of high HIV and hepatitis B and C prevalence. Indeed, infection and maternal fever in labour are associated with cerebral

palsy.14 Widespread use of fetal scalp electrodes may not be acceptable to obstetricians or to mothers; it is not possible in early labour, as there has to be adequate cervical dilatation to permit FSE application and recent evidence suggests that emergency caesarean section rates may be lower if the membranes remain intact during labour.15 Non-invasive f ECG recording should enable monitoring during early and late labour, as well as being more acceptable to women, and additionally avoid the risks of infection or trauma to the fetal scalp. Our group has recently demonstrated the feasibility of antepartum f ECG recording in over 600 singleton and multiple pregnancies from 15 to 40 weeks of gestation and derived normal ranges.16 We now demonstrate non-invasive acquisition of both the f ECG and uterine contractions during early and late labour using abdominally sited electrodes.

Fig. 2. Comparison of heart rate and uterine activity obtained by CTG and non-invasive f ECG. Line A was obtained using f ECG monitoring and the resulting heart rate averaged over three heart beats for comparison with heart rate recorded by conventional CTG (line B). Line C shows the electrical activity of the uterus with tocographically recorded contractions shown in line D.

D RCOG 2005 BJOG: an International Journal of Obstetrics and Gynaecology 112, pp. 1016 – 1021

NON-INVASIVE INTRAPARTUM FETAL ECG

1019

Fig. 3. The 10-s rhythm strip from case 1. Normal QRS complexes (open arrows) are consistently followed by a ventricular ectopics (closed arrows) (ratio 2:1).

METHODS With ethics committee approval from the Hammersmith Hospitals NHS Trust and written consent, 15 women underwent f ECG recording in threatened (cases 13 and 15) or established labour (median gestation 39 weeks, range 24 – 41), at a median cervical dilatation of 4.0 cm (range 0 – 10) of whom 8/15 (53%) had intact amniotic membranes (see Table 1). Ten of the 15 (67%) had epidural anaesthesia. With the mother reclined, 12 electrodes were placed evenly over the anterior abdominal wall and two reference electrodes sited near the umbilicus. The skin was prepared for low impedance by gentle excoriation of surface skin cells (3M Skinprep 2236). Electrodes were connected to a 12-bit multichannel digital recorder with sampling rate of 512 Hz. The noise level of the amplifier was 0.2 AV rms (Rin 10 kV). Electrode impedances were checked to ensure that they were less than 5 kV. Raw data were observed on a laptop computer screen. High-pass and low-pass filtering was employed at 1 and 150 Hz, respectively. To demonstrate feasibility about 5– 15 minutes of data were recorded in each case. Generation of the parameterised fECG for the patient report (Fig. 1) took 5 –8 minutes. Successful acquisition of the fECG signal was defined as the continuous detection of R waves in a 60-s data sample. Measures of HRV over 60 s were calculated as follows: coefficient of variation ¼ standard deviation of heart rate / mean of heart rate; relative variation ¼ (maximum heart rate  minimum heart rate) / mean of heart rate. The T/QRS ratio was also calculated as a measure of ST segment elevation.13 The processing technique for the trans-maternal abdominal acquisition of the fECG has been published16 and the recording and processing system described in more detail in International Patent Application numbers GB2002/ 004410 and WO 2004/084087 A1.

RESULTS A continuous recording of f ECG was successfully obtained within 15 minutes in 12/15 (80%) patients with discreet P, QRS waves in all, and T waves in 11/12 (92%) of successful recordings (Fig. 1) from which PR, QRS, QT and other parameters were derived (Table 1). Reference

ranges for all fetal f ECG time intervals derived from an antenatal population have been published elsewhere.16 True beat-to-beat heart rate (HR) and measures of HRV were displayed. Uterine electrical activity, shown to match tocographically recorded uterine contractions, was also separated and displayed (Fig. 2). Failure to acquire the fECG in three cases was attributed to the presence of excessive abdominal muscular activity in two and electrical interference and/or maternal retching in the other. In case 1, an inadequate CTG was recorded because of ectopic cardiac activity. Fetal heart rate was determined by the larger ectopic beats alone and gave an incorrectly low FHR, which could have resulted in an unnecessary caesarean section. However, fECG (Fig. 3) demonstrated ventricular trigeminy (a single ventricular ectopic beat occurring after two normal heart beats). The detail provided allowed a more informed clinical judgement and a normal delivery ensued with normal postnatal echocardiogram and neonatal ECG showing occasional isolated ventricular ectopic beats of no clinical significance. All term fetuses, seven male and six female delivered in good condition (Apgar scores 8– 10 at 1 minute and 8– 10 at 5 minutes) with normal birthweights (median 3268 g, range 2614 –4026 g). Two cases with threatened preterm labour (cases 14 and 15) did not deliver within 24 hours of obtaining fECG recordings.

DISCUSSION This study demonstrates a technique for obtaining intrapartum f ECG waveforms non-invasively, avoiding the risks of invasive monitoring via a fetal scalp electrode, which can be used regardless of the state of labour or cervical dilatation. Our technique displays similar basic information on HR and uterine contraction patterns as conventional CTG, together with the fECG parameters and HR variability. The limitations of previous non-invasive f ECG techniques include failure to obtain complete fECG waveforms (Pieri et al.17 failed to obtain T waves in any of 400 antenatal singleton recordings) and restriction to the antenatal period. The chief advantage of our technique is that signal separation does not rely critically upon the number or the location or configuration of the electrodes and a reduction to eight or less is expected with further

D RCOG 2005 BJOG: an International Journal of Obstetrics and Gynaecology 112, pp. 1016 – 1021

1020

M.J.O. TAYLOR ET AL.

refinement and experience of the technique. Compared with conventional CTG, the advantage of this system tested is the provision of complete fECG signals. ST segment analysis has already been shown to reduce the incidence of neonatal acidemia and of operative delivery.12 In addition, fECG measures true heart rate and HRV.18 Reduced HRV has been shown to be the most significant intrapartum fetal heart rate parameter to predict the development of neonatal acidemia19 and potentially provides a powerful index of fetal wellbeing. Mode of analgesia may influence fECG detection rates. In this study, 10/15 subjects had an epidural anaesthesia, currently used in about 1/3 of deliveries in the UK. We speculate that epidurals assist in detection by making the mother less restless, thereby reducing muscular electrical interference. Conventional CTG records uterine contractions using a separate transducer held onto the abdomen by a tight-fitting belt, which mothers find uncomfortable and often requires resiting during labour. This is an indirect measure that uses the tension recorded at the abdominal surface to infer the level of uterine muscle activity. In contrast, the present system obtains information about uterine contractions via the same electrodes, which is a more direct indicator of uterine muscle activity and hence potentially more useful clinically. Excessive maternal abdominal muscular activity resulted in failure to detect f ECG in 3/15 cases. Our technique is currently a two-stage process. Firstly, raw data are inspected visually to exclude undue electrical interference. Next, the combined data are stored and off-line signal acquisition attempted. On-line adaptation of our technique and recording for longer than 5 – 15 minutes as occurred in the present study would in future enable the operator to recognise unsuccessful separation immediately and prompt measures to improve signal detection such as repositioning of the mother. Thus, detection rates are likely to improve with an on-line version of our f ECG system. In this study, values of HRV and/or time intervals lying outside the 95% CI of the non-labouring population were seen in five cases. In case one, ventricular trigeminy would explain the abnormalities of HRV, but the effects of labour such as fetal head compression and use of analgesia on human fetal physiology are poorly understood. Furthermore, the recordings were made over a relatively short period in this feasibility study and longer recordings would be required to assess the significance of these observations.

CONCLUSIONS The aim of this study was to establish the feasibility of obtaining the detailed fECG during labour. This has been demonstrated successfully in the majority of cases together with the additional and unexpected finding of a useful electrical signal indicating uterine activity. In this preliminary

study the system has already shown advantages over conventional CTG including the acquisition of true heart rate and HRV. The development of an on-line version of this fECG system will enable full fECG recordings throughout labour, currently only available invasively late in labour through the use of a scalp electrode.

Acknowledgments The authors would like to thank QinetiQ for development and supply of recording equipment, performing of off-line analysis of the intrapartum fECG recordings and funding of the midwife SOA. Conflict of interest statement MJS, MJT and ARG are employees of QinetiQ. MJOT received funding to attend meetings from QinetiQ. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.

References 1. CESDI. Confidential Enquiry into Still-Births and Deaths in Infancy 8th Annual Report. London: Maternal and Child Health Research Consortium, 2001. 2. Ellenberg JH, Nelson KB. Cluster of perinatal events identifying infants at high risk for death or disability. J Pediatr 1988;113(3):546 – 552. 3. Badawi N, Kurinczuk JJ, Keogh JM, et al. Antepartum risk factors for newborn encephalopathy: the Western Australian case – control study. BMJ 1998;317(7172):1549 – 1553. 4. Hagberg B, Hagberg G, Beckung E, Uvebrant P. Changing panorama of cerebral palsy in Sweden. VIII. Prevalence and origin in the birth year period 1991 – 94. Acta Paediatr 2001;90(3):271 – 277. 5. Cowan F, Rutherford M, Groenendaal F, et al. Origin and timing of brain lesions in term infants with neonatal encephalopathy. Lancet 2003;361(9359):736 – 742. 6. Prentice A, Lind T. Fetal heart rate monitoring during labour—too frequent intervention too little benefit? Lancet 1987;2(8572):1375 – 1377. 7. Grant A, O’Brien N, Joy MT, Hennessy E, MacDonald D. Cerebral palsy among children born during the Dublin randomised trial of intrapartum monitoring. Lancet 1989;2(8674):1233 – 1236. 8. Barrett JF, Jarvis GJ, Macdonald HN, Buchan PC, Tyrrell SN, Lilford RJ. Inconsistencies in clinical decisions in obstetrics. Lancet 1990; 336(8714):549 – 551. 9. Thomas J, Paranjothy S. National Sentinel Caesarean Section Audit Report. London: Clinical Effectiveness Support Unit, RCOG, 2001. 10. Westgate J, Harris M, Curnow JS, Greene KR. Randomised trial of cardiotocography alone or with ST waveform analysis for intrapartum monitoring. Lancet 1992;340(8813):194 – 198. 11. Strachan BK, van Wijngaarden WJ, Sahota D, Chang A, James DK. Cardiotocography only versus cardiotocography plus PR-interval analysis in intrapartum surveillance: a randomised, multicentre trial. FECG Study Group. Lancet 2000;355(9202):456 – 459. 12. Neilson JP. Fetal electrocardiogram (ECG) for fetal monitoring during labour. Cochrane Database Syst Rev 2003(2):CD000116. 13. Amer-Wahlin I, Hellsten C, Noren H, et al. Cardiotocography only versus cardiotocography plus ST analysis of fetal electrocardiogram

D RCOG 2005 BJOG: an International Journal of Obstetrics and Gynaecology 112, pp. 1016 – 1021

NON-INVASIVE INTRAPARTUM FETAL ECG for intrapartum fetal monitoring: a Swedish randomised controlled trial. Lancet 2001;358(9281):534 – 538. 14. Grether JK, Nelson KB. Maternal infection and cerebral palsy in infants of normal birth weight. JAMA 1997;278(3):207 – 211. 15. Levy R, Ferber A, Ben-Arie A, et al. A randomised comparison of early versus late amniotomy following cervical ripening with a Foley catheter. Br J Obstet Gynaecol 2002;109(2):168 – 172. 16. Taylor MJ, Smith MJ, Thomas M, et al. Non-invasive fetal electrocardiography in singleton and multiple pregnancies. Br J Obstet Gynaecol 2003;110(7):668 – 678.

1021

17. Pieri JF, Crowe JA, Hayes-Gill BR, Spencer CJ, Bhogal K, James DK. Compact long-term recorder for the transabdominal foetal and maternal electrocardiogram. Med Biol Eng Comput 2001;39(1):118 – 125. 18. Peters M, Crowe J, Pieri JF, et al. Monitoring the fetal heart noninvasively: a review of methods. J Perinat Med 2001;29(5):408 – 416. 19. Williams KP, Galerneau F. Intrapartum fetal heart rate patterns in the prediction of neonatal acidemia. Am J Obstet Gynecol 2003;188(3): 820 – 823. Accepted 5 January 2005

D RCOG 2005 BJOG: an International Journal of Obstetrics and Gynaecology 112, pp. 1016 – 1021