24 Role of a Computerized CTG. Handbook CTG

 24

Role of a Computerized CTG

◈

Sabrina Kuah and Geoff Matthews

Handbook of CTG Interpretation: From Patterns to Physiology, ed. Edwin Chandraharan.

Published by Cambridge University Press. © Cambridge University Press 2017.

Introduction

Continuous electronic fetal monitoring (EFM) was developed in the 1960s, and the CTG

became commercially available at that time. Erich Saling developed fetal blood

sampling prior to CTG, although it subsequently found a place as an adjunct to CTG, and

its relevance is now being re-evaluated. Fetal blood sampling has not been shown to

reduce caesarean section rates or any prespecified neonatal outcomes.1

The CTG has become ubiquitous in modern labour wards, while attempts to

validate its role in improving perinatal outcomes have proved challenging.

Fetal ECG has recently become available – the technique involves computerized

analysis of the fetal ST waveform and aims, in conjunction with CTG, to provide EFM

with more robust specificity and sensitivity.

The human factor has increasingly been recognized as a potential weak link in fetal

monitoring and a variety of mitigations proposed. Even when employing standard

scoring systems, CTG suffers significantly from intra- and inter-observer variation.2

Emphasis on systematized training in fetal monitoring for all clinical staff with regular

credentialing has become a feature of many delivery suites seeking standardized care.Physiologically based CTG training has been proposed as an alternative approach to

reliance on simple pattern recognition.

Computerized decision support technology has been developed with the aim of

improving recognition of abnormal fetal heart rate (FHR) patterns and reducing times to

effective interventions. Preliminary findings appear encouraging, so clinicians eagerly

await the completion of ongoing clinical trials into the effectiveness of this technology.2

Cardiotocography

CTG is simply the fetal heart expressed over time, displayed in the patient’s room, and

in some units, it is also monitored centrally. The heart rate, its variability, the presence

or absence of accelerations and decelerations are all assessed by the human observer

and the CTG thus interpreted. Traditionally, pattern recognition of the CTG raises

suspicions of fetal metabolic acidosis, triggering an attempt to improve the fetal

environment, seek reassurance or expedite delivery.

CTG monitoring has become a routine part of clinical obstetric care, although its

sensitivity and specificity at detecting fetal metabolic acidosis is poor. The negative

predictive value of a normal CTG is in excess of 90 per cent, although the longevity of

‘reassurance’ obtained is unclear and potentially only for the period the monitoring is

ongoing. However, the positive predictive value of an abnormal CTG is quite poor and

overall CTG in labour has a false-positive rate of around 60 per cent if acidosis is

defined as pH <7.20.

The high false-positive rate of CTG has thus been implicated in the escalating rate

of caesarean section births observed in recent decades. While there has been data to

suggest a reduction in rates of neonatal seizures, there is, as yet, no level-one evidence

of reduction in rates of fetal metabolic acidosis at birth as a result of deployment of

CTGs in maternity units.

Adjuncts to CTG

Fetal Blood SamplingFetal blood is collected from the scalp and run through a blood gas analyser to obtain a

pH and more recently a lactate reading. Management is dictated by the pH result – ≤7.20

has conventionally been used as an indication for delivery.

However, correlating peripheral acidosis detected in fetal scalp blood with central

acidosis is flawed since a fetus will progressively protect the central organs (brain,

heart and adrenals) at the cost of shutting down the periphery.

Fetal ECG (STAN)

See Chapter 23.

Training and Reducing Human Error

Human error and misinterpretation of CTG data contributes to poor outcomes associated

with EFM. Reviews of case series with poor perinatal outcomes have identified

significant delay in the recognition of even severe CTG abnormalities.3 Furthermore, the

series suggests that these delays in the recognition of abnormal CTGs are associated

with outcomes including cerebral palsy and perinatal death.4 It would appear that there

is an association with clinician seniority. Rates of perinatal mortality in the United

Kingdom are significantly increased at times of lesser senior staffing such as at night and

in the summer months when many senior staff are on leave.5

Credentialing of all staff members in a recognized CTG competency-based training

program has increasingly become the norm. The content of training may influence

effectiveness, and a physiologically based training rather than pattern recognition has

been proposed as preferable.6

The Royal Colleges are increasingly standardizing CTG training. In South

Australia, all obstetric doctors and midwives must attend and achieve competency in the

Royal Australian and New Zealand College of Obstetrics and Gynaecology Fetal

Surveillance Education Program.

There is some evidence that intensive CTG training can at least temporarily reduce

the proportion of substandard CTG interpretation in cases of low apgars, but the

improvement seems to require continued intensive CTG training to be sustained over

time.7Computerized Decision Support

Human factors come into play around the issues of uniform recognition of abnormal

CTG patterns triggering appropriate decision points. Decision support technology

allows for the central alerting of an abnormal CTG trace to busy delivery suite staff and

suggests a possible interpretation (Figure 24.1). Most systems require attending staff to

acknowledge the alert, facilitating a forcing function for clinician’s review of the CTG.

Figure 24.1 Computerized CTG highlighting the onset of a ‘baseline shift’ to clinicians.

While human error can be mitigated by improved training and credentialing,

experience from a range of areas including aviation suggests that even the most highly

trained operators will potentially benefit from decision support whether it comes from a

colleague and/or computer.

Computerized CTG has been available in various formats since the 1990s and

incorporates a decision support capability. Automated analysis of CTG tracings requires

processing uterine contraction signals, short-term variability, estimation of FHR

baseline, detection of accelerations, abnormal and mean long-term variability and

detection and classification of decelerations. Algorithms calculate these variables based

on preprogrammed system-specific criteria. Omniview Sis-Porto and the K2 Medical

Systems Data Collection System (Guardian)8 are two currently available commercial

systems.In 2010, a Cochrane Review of antenatal CTG noted that computerized CTG in two

studies (469 patients) significantly reduced the relative risk of perinatal mortality (RR

0.2; 95% CI 0.04–0.88).2 Similarly, intrapartum computerized CTG studies from the

1990s suggest that the software being assessed performed as well as expert

obstetricians in interpreting CTGs and predicted poor outcome sooner than the experts.9

There are currently two large clinical trials underway assessing two systems of

computerized clinical support, the INFANT study (Intelligent Fetal Assessment

Monitoring) and the Omniview Sis-Porto trial.

The INFANT trial10 is randomizing approximately 46,000 patients in units

throughout the United Kingdom and is powered to detect a 50 per cent relative risk

reduction (5 per cent significance) in a composite score of ‘poor perinatal outcome’

between those randomized to decision support and those receiving standard care. This

trial is utilizing the ‘Guardian’ system developed by the Plymouth Group.

The Omniview Sis-Porto system also has decision support software and provides

visual and auditory alerts based on the interpretation of CTG but also incorporates fetal

ECG (STAN) data (Figure 24.2). The software analyses the CTG–fetal ST and produces

a colour-coded alert notifying the operator when there are characteristics that may

increase the likelihood of fetal hypoxia. The colour-coded alerts produce management

advice (e.g. consider discontinuing/reducing oxytocin infusion or acute tocolysis)

depending on the severity of abnormality.

Figure 24.2 The Omniview Sis-Porto decision support software. Note that the computer

is flagging up ‘very repetitive decelerations’ in amber. More severe abnormalities will behighlighted in red.

The Omniview Sis-Porto trial has randomized 8,133 women and is powered (5 per

cent significance) to detect a 1 per cent reduction in the overall rate of fetal metabolic

acidosis from 2.8 per cent to 1.8 per cent. The secondary outcome measures include rate

of caesarean section for nonreassuring fetal state; fetal blood sampling rates; operative

vaginal delivery rates; apgar scores <7; and admission to neonatal intensive care.

Conclusion

Improved perinatal outcomes may be achieved with intensive physiology-based CTG

education for all staff provided there is systematic and sustained effort at credentialing

participants.

Fetal ECG ST analysis adds fetal cardiac oxygen metabolism as a second

parameter to the existing CTG data, potentially allowing for greater specificity and

sensitivity in fetal monitoring.

The incorporation of computerized decision support to the CTG ± fetal ECG

mitigates human factors and may provide an opportunity to realize greater dividends

from EFM. However, preliminary data from both the Sis-Porto Trial and the Infant Trial

which have been presented at scientific meetings (i.e. pending publications) suggest that

the use of ‘decision-support’ computerised systems have not resulted in any significant

improvements in perinatal outcomes. This illustrates the importance of using the

knowledge of fetal physiology to interpret CTG Traces.

References

1. Alfirevic Z, Devane D, Gyte GML. Continuous cardiotocography (CTG) as a form of

electronic fetal monitoring (EFM) for fetal assessment during labour. Cochrane Database

Syst Rev 2013;5:CD006066.

2. Grivell RM, Alfirevic Z, Gyte GML, Devane D. Antenatal cardiotocography for fetal

assessment. Cochrane Database Syst Rev 2010;1:CD007863.3. Ennis M, Vincent CA. Obstetric accidents: a review of 64 cases. BMJ 1990;300:1365–7.

4. Gaffney G, Sellers S, Flavell V, Squire M, Johnson A. Case-control study of intrapartum

care, cerebral palsy, and perinatal death. BMJ 1994;308:743–50.

5. Stewart JH, Andrews J, Cartlidge PH. Numbers of deaths related to intrapartum asphyxia

and timing of birth in all Wales Perinatal Survey, 1993–5. BMJ 1998;316:657–60.

6. Khangura T, Chandraharan E. Electronic fetal heart rate monitoring: the future. Curr

Women’s Health Rev 2013;9:169–74.

7. Young P, Hamilton R, Hodgett S, Moss M, Rigby C, Jones P, et al. Reducing risk by

imp

roving standards of intrapartum fetal care. J R Soc Med 2001;94:226–31.

8. Ayres-de-Campos D, Bernardes J, Garrido A, Marques-de-Sá J, Pereira-Leite L. SisPorto

2.0: a program for automated analysis of cardiotocograms. J Matern Fetal Med

2000;9:311–18.

9. Keith RDF, Beckley S, Garibaldi JM, Westgate JA, Ifeachor EC, Greene KR. A multicenter

comparative study of 17 experts and an intelligent computer system for managing labour

using the cardiotocogram. Br J Obstet Gynaecol 1995;102:688–700.

10. Brocklehurst P. A multicentre randomized controlled trial of an intelligent system to

support decision making in the management of labour using the cardiotocogram (INFANT).

Protocol 2014. Version 12. www.ucl.ac.uk/ctu/infant