Main findings
In this quasi-experimental, pre-post intervention, diagnostic trial, the use of routinized fetal scalp stimulation testing appeared to improve the performance of intermittent auscultation for detecting clinically important severe acidemia (pH < 7.0) from 27 to 70% (p = 0.032). The negative predictive value of the fetal scalp stimulation test ranged from 88 to 99% for mild (pH < 7.2) to severe fetal acidemia, but the positive predictive value remained low.
Both cohorts had a significant proportion of women that received oxytocin augmentation in labor, 61 and 55% respectively for the IA and IA + FSST groups. Labor management was left to the discretion of the obstetric providers in charge for each shift and followed the WHO partogram. It was common that oxytocin would be provided intravenously if a woman crossed the action line on the partogram. Augmenting labor is a fairly advanced obstetric practice that necessitates increased fetal monitoring beyond intermittent auscultation [
10]. Access to oxytocin combined with limited access to cesarean delivery has created a situation around the world that practice has progressed beyond the ability to safely monitor fetuses. It may be that adding fetal scalp stimulation to IA protocols would help make augmented labor safer for mothers and babies in areas where CTG is not available.
In 2012, researchers at KCMC reported a perinatal death rate of 12.5/1000 (1.3%) live births and identified intrapartum fetal hypoxia as the leading cause of in-hospital neonatal mortality [
22]. Subsequently in 2017, Simiyu et al., showed that 5.6% of newborns were diagnosed with hypoxic-ischemic encephalopathy secondary to intrapartum fetal hypoxia [
23]. In our study, 15–20% of fetuses experienced hypoxia during labor, putting them at ten times the risk for such morbidity compared with fetuses in US labor wards [
13]. If, as a diagnostic tool, FSST-enhanced intermittent auscultation is definitively proven to improve the diagnosis of fetal hypoxia, accelerated and targeted intervention could be implemented to improve outcomes.
FSST has been tested previously in a low-income country. Rathore and co-authors used FSST to enhance intermittent auscultation with a Pinard stethoscope only in fetuses that were already diagnosed with NRFS based on the presence of tachycardia, bradycardia, or meconium-stained amniotic fluid. Adding FSST to auscultation with a Pinard stethoscope demonstrated a 41% sensitivity and 81% specificity to detect a pH < 7.20 following an absent FSST [
18]. This study was an important first step for advancing fetal monitoring in low-resource settings, but it begged the question of how a Doppler would perform in a similar environment. More importantly, our study highlights a high false-negative rate with 78% (40/51) of the babies born with acidemia not being identified with routine intermittent auscultation alone. Waiting to perform the FSST only when NRFS has been identified would not lead to improved outcomes.
Recently, a study by Clark and co-authors raised suspicion about the limits of fetal monitoring accuracy in high-income countries. Using a standardized algorithm for the management of concerning CTG tracings, they found that retrospective, expert review only had 46% sensitivity and 82% specificity to predict fetal acidemia. This study did not include mention of the FSST, which may or may not have influenced the results [
19]. The baseline clinical performance of CTG monitoring only had a sensitivity of 30% and specificity of 81%. Our study shows that midwives using FSST-enhanced intermittent auscultation can achieve diagnostic accuracy on par with American obstetricians using CTG.
We demonstrate, thanks to the efforts of Tanzanian midwives, that it is feasible to introduce the FSST using locally-sourced fetal monitors in Sub-Saharan Africa. The study used broad inclusion criteria such that nearly all women undergoing fetal monitoring were invited to participate. The marginal costs of adding 3–4 handheld Doppler units to a labour ward and performing an extended vaginal exam 1–3 times on each woman is negligible compared to the costs that would be incurred introducing continuous CTG. Because the majority of labor wards in Tanzania perform fewer than one delivery per day [
24], simple, low-cost solutions, such as FSST may be more feasible. It is possible that performing additional examinations may lead to an increased risk of intraamniotic infection, but this would need to be monitored in future studies. The FSST was planned to be completed at times of routine vaginal exams when possible in order to minimize this risk. Investors may want to consider other methods such as vibroacoustic stimulation or Pawlik’s grip if they are concerned.
Our study was a proof-of-concept study that was limited by a relatively small sample size and lack of randomization. As such it does not definitively establish improved fetal outcomes with the use of FSST. It does, however, suggest that a definitive clinical trial would be a logical next step, and we offer meaningful estimates of effects sizes and have probed important feasibility issues. Such a clinical trial would necessitate strict adherence to an FSST-based protocol. It is possible that implementing FSST into labor protocols may lead to under-diagnosis or over-diagnosis of non-reassuring fetal status, but our preliminary data indicate that an IA + FSST protocol would likely improve diagnosis. We believe that the addition of FSST to a protocol may help prevent cesarean deliveries because it gives providers a physiologically-based assessment of the long-term oxygenation status for the fetus. There are two types of continuous Doppler fetal heart monitors that are being developed for use in low-resource settings [
25,
26]. It is imperative that we do not rush to adopt continuous fetal monitoring in Sub-Saharan Africa without fully understanding the diagnostic utility of our devices. Future studies should focus on neonatal outcomes rather than surrogate markers such as pH at birth. The four clinical champion midwives that participated in the study quickly felt comfortable performing and interpreting the FSST, but other midwives in other facilities may not perform as well. Successful implementation of the FSST in other settings will depend on education in order to utilize this technique, thus ensuring correct interpretation and intervention when indicated. Ultimately changing global policy related to fetal monitoring would require in-service education distributed through the WHO and pre-service education in medical schools and midwifery schools in relevant areas around the world.