Results
A total of 34 residents participated in the 3 sessions, with seniority ranging from the 1st to the 5th year (Table 1). Only 11.8% of the residents surveyed felt that they had sufficient training on FHR analysis and 2.9% on fetal physiology during labor. Of these, 14.7% felt that they had sufficient training in the indication of second-line surveillance examinations. They were 81.8% have received specific training on FHR during their residency, mainly during a planned course in their university studies. Yet 90.9% believed that their training was insufficient on the use of the FIGO and CNGOF classifications. As for FBS, 66.7% had already laid an indication and 48.4% had never achieved a FBS.
The median MCQ score before teaching was 1.5 [1.0 to 2.0] with scores ranging from 0.0 to 4.5. After our training, the median was 4.0 [3.0 to 4.5] with scores ranging from 1.0 to 6.5 (p <0.0001). Of all the FHR analyzed during the defined periods, 49.2% of the rhythms were considered normal or low risk before our training against 54.7% after (Table 2). The distribution of FHR grades found a less pejorative evaluation after training (p<0.001). In fact, 17.4% of the FHR analyzed by the residents were considered to be at significant risk of acidosis and 2.7% at major risk in pre-training, compared with respectively 11.2% and 0.1%. The reproducibility of residents’ responses to FHR evaluation was improved after training with Krippendorff’s alpha index pre-training of 0.60 [IC95, 0.55 to 0.65] and post training of 0.72 [IC 9, 0.67 to 0.76]. Before our formation, 327 FBS were requested on all clinical cases against 269 after our training, a decrease of 6.9% (p=0.002). The evaluation of the pH estimate relative to the actual value was significantly different between the two sessions (p = 0.020) with a higher good estimate rate after training (47.2% vs 40.0%) although not significant with p = 0.28, a lower underestimation rate after training (26.4% vs. 38.2%) and a higher overestimate rate after training (26.4% vs. 21.8%). The rate of underestimation and overestimation was identical after training.