The cardiotocograph was invented in early 1960’s by
Hammacher, working closely with Hewlett Packard. Its
introduction into the clinical setting was surrounded by great
fanfare and before the end of the decade continuous electronic
fetal monitoring (EFM) was in widespread clinical use 1-4
.
The general expectation was that within a short time cerebral
palsy (CP) would be virtually eliminated. At the time
obstetricians felt that the fetal monitor had given them, for
the first time, a window through which they could monitor
the wellbeing of the fetus antenatally and during labor. Since
cesarean section (CS) had become a relatively safe procedure
by this time they thought that they had the power to save
babies from the potentially catastrophic effects of hypoxia
during labor. Sadly, as we now know, this expected fall in
the incidence of CP has failed to materialize in spite of the
introduction of continuous EFM and its role has been
questioned 5
. The aim of this review is to try to understand
why and to ask whether continuous fetal monitoring has
any role to play in modern obstetrics.
With the benefit of hindsight, perhaps it is not too surprising that continuous intrapartum fetal heart rate (FHR) monitoring has failed to reduce significantly the incidence of CP. After all, the FHR is only an indirect measure of fetal hypoxia. While it has certainly been demonstrated that fetal hypoxia can affect FHR, far more valuable information would be gained if we could measure fetal blood pressure and cerebral flow or cerebral oxygen saturation. Unfortunately, such measurements are technically impossible at the moment, at least in human fetuses, and hence all we have to rely on is the FHR.
In any event, it is now generally accepted that a maximum of only 10% of all cases of CP can be explained by perinatal events 6,7. The other 90% are due to antepartum events such as intrauterine infection, antepartum hemorrhage, etc. This means that even if we had the perfect tool to identify those fetuses at risk of hypoxic brain damage during labor, we could still hope to prevent, at best, only 10% of cases of CP and the cardiotocograph (CTG) is certainly not the instrument. Several studies of CTG interpretation have shown significant interobserver variation. Even worse, there is also significant intraobserver variation. In other words, when the same CTG is shown to the same expert a few months later, there is a high chance that the interpretation will be different 8-17. In fact, in the United Kingdom the Confidential Enquiries into Stillbirths and Deaths in Infancy (CESDI) in 1997, 1998 and 2000 stated categorically that poor interpretation of the CTG was a major contributor to intrapartum stillbirths 18-20 .
Perhaps the most well known trial of continuous FHR monitoring is the Dublin randomized trial from Ireland in 1985 21. This showed that despite increased intervention, there was no improvement in neonatal outcome when low risk women were continuously monitoried. Even more worrying were the findings of Nelson et al 22 in 1996. They reviewed women with a highly abnormal CTG in labor i.e. fetal tachycardia with reduced variability and late decelerations. These were the very fetuses traditionally considered to be at the greatest risk of CP and for whom urgent delivery by CS or instrumental delivery was considered mandatory. They found that only 58% of these fetuses with a highly abnormal CTG were acidotic at birth as judged by umbilical artery pH. Moreover, only 0.2% went on to develop CP. In other words, continuous FHR monitoring, when used as a predictor of CP, has a 99.8% false positive rate! It is difficult to imagine any other test in any branch of medicine which has such a high false positive rate, particularly when the intervention is as major a procedure as CS.
Against this background, the Royal College of Obstetricians and Gynecologists (RCOG) and the National Institute for Clinical Excellence (NICE) in the United Kingdom decided to review the whole issue of FHR monitoring 23. They convened a large team of interested parties including obstetricians, midwives, statisticians, neonatologists, epidemiologists, politicians, and consumer groups such as the National Childbirth Trust. This expert committee carried out a systematic review of the relevant literature, attempting to address a series of clinical questions. For each clinical question, they made a recommendation based on the available good quality evidence. The strength of the recommendation was categorized as A, B, C or Good Practice Points, depending on the quality of the available evidence (Table 1).
Most of the evidence analyzed by the RCOG/NICE group comes from nine large randomized controlled trials (Table 2).However the Dublin RCT dominates because it includes a large number – almost 13,000 21. The conclusions of the review group are summarized below.
The only advantage associated with continuous EFM is a significant decrease in the incidence of neonatal seizures (RR 0.51, 95% CI 0.32 – 0.82). But this was true only in trials which included fetal blood sampling and those with a high quality score (meaning a well-conducted trial). The RCOG/ NICE committee also noted that only the Dublin study gave a definition of neonatal seizures.
However, when this group analyzed the three studies with longer follow up (Dublin, 1985 21; Denver, 1979 27; and Seattle; 1987 30) they found that the use of continuous EFM was not associated with a reduction in the incidence of CP 32,33. Indeed, one of these studies found an increased frequency among the EFM group 30 !
There was no other beneficial impact of continuous EFM on perinatal outcome. The incidence of 1 minute apgar scores of less than 4 or less than 7, and of admission to the neonatal intensive care unit were unchanged, as was the overall perinatal mortality rate.
The group found that continuous EFM led to a significant increase in the rate of emergency CS. Overall, the relative risk was 1.41 (95% CI 1.23-1.61), compared to that with intermittent auscultation of the fetal heart. If analysis is limited to include only those trials which included fetal scalp blood sampling (FSBS) 34, the increased risk of CS was not as great but was still significant (RR 1.24, 95% CI 1.05 – 1.48).
Continuous EFM also increases the rate of operative vaginal delivery the overall relative risk being 1.20 (95% CI 1.11 – 1.30). This increase was most marked in trials which included FSBS.
It is clear therefore that continuous EFM leads to significantly more intervention in the form of CS and operative vaginal delivery, but without any benefit to the baby 34,35. In addition, it leads to considerable parental anxiety as well as clinician stress 36-39. There is also an increase in other invasive procedures such as the application of fetal scalp electrodes and FSBS; these of course carry a risk of maternal and fetal infection, trauma and hemorrhage 40. In the era of HIV, this consequence must be taken into account.
The equipment itself–fetal monitors with consumables like paper and contact gel – is expensive. However, these costs are relatively small when compared to the cost of an increase in the CS rate. A calculation in the United Kingdom in 2000 suggests that in obstetric units which don’t have access to FSBS the cost of the increased CS rate (RR 1.24) is just over £ 100,000 extra per 1,000 births 23,29,41 . In units which use FSBS (RR of CS 1.41), the additional cost is only £ 54,000 per 1,000 births. In an average unit in the United Kingdom which delivers 4,000 babies a year, this considerable expense would be enough to fund the salaries of 20 extra midwives.
Shocking as it seems, however, all of this is dwarfed by
the cost of the increased litigation associated with the use
of continuous EFM. In a medico-legal case where a baby
has ended up with CP, the finding of any abnormality in
the CTG recording is usually enough to lead to a verdict
against the hospital/doctor. The courts rarely take into
consideration the fact that we now know that at least
90% of these babies will have sustained cerebral damage
even before the labor started. Awards in the United
Kingdom for babies affected by CP are counted in millions
of pounds and a figure of five or seven million pounds is
not uncommon.
The RCOG/NICE committee issued guidelines, where appropriate, based on their review of the evidence.
Low risk pregnancies
Intermittent auscultation (IA) should be offered and recommended (Level A).
Because the use of continuous EFM in low risk
women increased intervention rates without offering
any advantage to the fetus IA should not just be
offered, but it should be actively recommended. In
other words, we should tell low risk women who
request continuous EFM that it is an option, but it
will not improve the outcome for their baby while it
will increase their own risk of operative delivery.
High risk pregnancies (Table 3)
There is some evidence that continuous EFM
improves the outcome in high risk pregnancies, and
should be recommended 24-28,30. Pregnancy can be
classified as high risk according to the presence of
maternal fetal or intrapartum risk factors, as depicted
in Table 3.
RCOG/NICE also recommended that in units using
EFM, there should be ready access to FSBS
facilities. This was a level A recommendation because
of the good evidence suggesting that FSBS helps to
limit the increase in CS rates associated with
continuous EFM.
The NICE guidelines of 2001 concluded that there was no
evidence that an admission CTG for low risk women improves
perinatal outcome and therefore advised against its use. This
recommendation is supported by a subsequent trial which
randomized over 8,000 low risk women to have an admission
CTG or not to have 42. They found no difference in perinatal
mortality or morbidity. This of course does not change the
recommendation that high risk women should be offered
not just an admission CTG, but continuous fetal monitoring
throughout labor.
