Author response to: Fetal monitoring—a risky business for the unborn and for clinicians

Authors


Author’s Reply

Sir,

We welcome the opportunity to further discuss the role of the STAN technology in intrapartum fetal monitoring. After the Swedish randomised study comparing cardiotocograph (CTG) + STAN with STAN,1 the technology was introduced in clinical practice in many centres in Sweden. The guidelines from the manufacturer stated that scalp blood sampling for assessment of pH or lactate was ‘of very limited value’ and could be abandoned if the STAN technology was used. The evidence for this statement is unclear, since scalp blood sampling was frequently used in the Swedish randomised study in both study arms. Further, the guidelines also stated expectancy up to 90 minutes in the second stage of labour and no time limit in the first stage of labour for abnormal CTG pattern if no STAN warnings (events) appeared. In the Lancet study the time frame for expectancy was 60 minutes. That these guidelines could be hazardous became evident when several cases were reported to the National Board of Health and Welfare and we have previously reported 12 cases in this journal.2

In their commentary on page 935, Amer-Wahlin and Dekker3 claim that the adverse outcome was caused by the staff not taking proper action upon CTG, misinterpretation of the CTG and/or incorrect use of STAN because of lack of proper training. Limitations of the STAN technology is related to ‘the same old problem: CTG’. It is striking that not a word is said about the fact that STAN is far from always warning in cases with fetal hypoxia and in some cases too late for intervention. Amer-Wahlin, heavily involved in the implementation of the technology in Scandinavia, may have a special obligation to give both pros and cons for STAN. Among the 12 cases, STAN events did not appear in six (three had a preterminal and three an ominous CTG pattern). In another three cases, STAN events appeared late in the hypoxic process. The guidelines have probably misled the obstetrician in some cases to wait with an ominous pattern, since no STAN events appeared. The advice to abandon fetal scalp blood sampling has probably led to catastrophic results in some of these cases. One cannot avoid reflecting what would have happened if conventional fetal monitoring had been used (the readers are welcome to make up their own mind by examining the recordings with clinical data in English (http://www.sfog.se/PDF/Cases%201-4.pdf). After 6 years of use in clinical practice revised guidelines have now been introduced in Sweden stressing the need for scalp blood sampling in uncertain cases, and an upper limit of expectancy of 60 minutes for abnormal CTG without STAN events.

What is the predictive value of STAN to predict cord artery metabolic acidosis or low pH? Amer-Wahlin et al. studied the 46 cases with metabolic acidosis from the Swedish randomised study.4 STAN events appeared in 63% (indication for intervention in 56%). In a recent retrospective study Vayssiere et al. demonstrated similar poor results with sensitivity to pH < 7.05 of 63% and to pH < 7.15 of 38%.5 Two prospective randomised trials on STAN versus CTG published after the original paper by Amer-Wahlin et al. have failed to demonstrate a difference in cord artery pH, rate of instrumental deliveries and in neonatal outcome. The only benefit is a significant reduction in use of fetal blood sampling.6,7 Obviously, more studies are warranted to clarify the potential benefit of the STAN technology.

In the meantime, caution should be taken when using the method bearing in mind that false-negative cases can occur. The promoters of this technology need to be clear about the limitations of the method, since in clinical practice false-positive tests can be handled, but it is difficult to accept false negatives.

Ancillary