Intrapartum fetal head position II: comparison between transvaginal digital examination and transabdominal ultrasound assessment during the second stage of labor

Objective To test the null hypothesis that no correlation exists between transvaginal digital examination compared with the gold standard technique of transabdominal suprapubic ultrasound assessment of fetal head position during the second stage of labor. A secondary objective was to compare the performance of attending physicians vs. senior residents in depicting fetal bead position by transvaginal digital examination in comparison with ultrasound assessment. Methods Consecutive patients in the second-stage of labor at term with normal singleton cephalic-presenting fetuses and ruptured membranes were included. Transvaginal digital examinations were performed by either attending physicians or senior residents and were followed immediately by transverse suprapubic transabdominal sonographic assessments performed by a single sonographer. Examiners were blinded to each other's findings. Power analysis dictated sample size. Exact binomial confidence intervals around observed rates were compared with chi(2) and Cohen's kappa-tests. Logistic regression was applied. P < 0.05 was considered significant throughout. Results One hundred and twelve patients were studied. The absolute error of transvaginal digital examinations was recorded in 65% of patients (95% confidence interval, 56-74%). Parity, pelvic station, combined spinal epidural anesthesia, length of first or second stages of labor, use of oxytocin augmentation, gestational age, mode of delivery, birth weight, and examiner experience did not significantly affect examination accuracy. Stratification, when the transvaginal digital examination was recorded as correct if occurring within +/- 45degrees of the ultrasound assessment, reduced the error of the transvaginal digital examinations to 39% (95% confidence interval, 30-49%). Independent variables again did not affect examination accuracy in this assessment modality. Rates of agreement between the two methods for attending physicians compared with residents were not significantly different. The overall degrees of agreement were 40% (95% confidence interval, 26-55%) and 68% (95% confidence interval, 53-80%) (kappa = 0.25 and 0.30) for the absolute agreement and +/- 45degrees assessment modalities respectively, for attending physicians, and 31% (95% confidence interval, 20-44%) and 55% (95% confidence interval, 42-68%) (kappa = 0.14 and 0.12) for senior residents. Conclusion Using ultrasound assessment as the gold standard, our data demonstrate a high rate of error (65%) in transvaginal digital determination of fetal head position during the second stage of labor. The performance of senior residents in transvaginal digital examinations did not differ significantly, from that of attending physicians. Intrapartum ultrasound increases the accuracy of fetal head position assessment during the second stage of labor.