Main Findings 
The cardiotocograph was developed in the 1960s to improve fetal surveillance. But its generalization has led to an increase in the rate of cesarean section and extractions for FHR abnormalities and without significant reduction in neonatal risk (16,17). A recent Cochrane review also found no differences in cerebral palsy, infant mortality, or other standard measures of neonatal well-being between intermittent or continuous FHR auscultation (17). The only difference found is the reduction of neonatal convulsion rates during continuous auscultation. These findings, including the increase in caesarean section rate, are related to the high sensitivity of the FHR and its low specificity, but not only. Misinterpretations or erroneous decisions are involved with a failure to take into account a pathological pattern in 20% of cases in newborns with metabolic acidosis (18). It therefore appears essential to improve our analysis of FHR and we wanted to evaluate the interest of training on fetal physiology in obstetric gynecology residents regardless of their initial level. We find a better global knowledge in fetal physiology with a reduction of ”extreme” classification during the analysis of FHR, resulting in a decrease in the number of FBS, a better estimation of these and a homogenization of practices in clinical cases performed.
The FHR interpretation training starts at the residency but the training time is variable and generally limited. The training is not systematic in France and varies according to the enrollments in workshops during congresses and the courses planned in the university. Conversely, in other countries such as England, the regulation has imposed the need for continuing education on the interpretation of the FHR every 6 months for midwives (19). Therefore, the formation of residents on the FHR and fetal physiology is insufficient in France and all of our residents interviewed were applicants for such training.
The FHR analysis makes it possible to detect situations at risk of fetal acidosis when its interpretation is correctly performed and thus to provide indications of second-line examinations or fetal extraction. Metabolic acidosis and associated neonatal morbidity could potentially be prevented in 40-50% of cases (18). Indeed, the most found errors are a misinterpretation of the RCF, an imprudent use of oxytocin and a failure to recognize at-risk pregnancies (20–22). Training development could reduce the consequences of inadequate monitoring. Indeed, Draycott and al conducted a retrospective study evaluating the value of training in obstetric emergencies. They were interested in the Apgar scores at 5 minutes of all the liveborn singletons with vaginal deliveries at term, between 1998 and 2003. They also identified hypoxic and ischemic encephalopathies. All the medical staff (midwife, gynecologist, anesthesiologist …) benefited from a training day during the year 2000. Draycott and al did not analyze the year 2000 and compared two periods: a pre-training period (from 1998 to 1999) and a post-training period (from 2001 to 2003). They found a significant reduction in low Apgar scores (<6) and in the incidence of hypoxic and ischemic encephalopathy (23). Thellesen and al, for their part, found a 14% decrease in fetal extraction, without increased risk of fetal hypoxia, after a training program for midwives and gynecologists from a Danish maternity hospital (24). Their training consisted of e-learning sessions and a day of theoretical courses.
The interpretation of the FHR is subject to intra- and inter-observer variability well studied now (25–27). This variability persists despite the existence of classification and this is more important when it comes to FHR classified as intermediate or pathological according to the FIGO classification (28). Regular training on FHR could reduce this inter-observer variability. Pehrson and al interviewed the Medline database to study and evaluate FHR training programs (29). Of the 409 citations they found, 20 studies included and analyzed. They report a better inter-observer agreement after training on the physiology and interpretation of FHR. We had the same findings by studying the reproducibility of the responses and therefore the inter-observer variability via the Krippendorff’s alpha index (0.60 before training vs. 0.72 after). Therefore, training in fetal physiology promotes the homogenization of answers. Thus, an education in the interpretation of the FHR would allow a reduction of its variability inter and intra observer and therefore standardize our practices.
Finally, we wanted to evaluate the impact on the use of a second-line examination. Indeed, when the FHR analysis is considered non-reassuring, there are different second-line exams to better characterize the fetal state such as the FBS with pH measurement to study the acidobasic state of the fetus (3,4,30). The interest of FBS is currently debated. It may not be representative of fetal acid-base status as it is derived from peripheral tissue or because of the compression of the fetal scalp during labor for example. This could lead to unnecessary interventions in fetuses that are not really hypoxic, which means that the debate on this subject persists (9,31). The purpose of this study was not to discuss the interest of FBS but to evaluate the decision to perform this second-line exam in our residents. We note a decrease in the number of FBS requested so second-line examination, currently being discussed, through a better interpretation.