Factors contributing to clinical misdiagnosis of fetal head position: an ultrasound based cohort study from Tanzania.

OBJECTIVES

Accurate determination of fetal head position is essential for managing labor, particularly in cases of slow progress or before operative vaginal interventions. Several studies have shown that ultrasound examinations are more accurate than clinical examinations, but few ultrasound studies are done in sub-Saharan Africa. Clinical vaginal assessment of fetal position remains the standard in most African settings, but its accuracy is limited by examiner skills and labor conditions. This study aimed to compare clinical and ultrasound assessments of fetal position during active labor, to identify factors contributing to clinical misdiagnosis, and to evaluate the impact of misdiagnosis on delivery mode.

METHODOLOGY

An observational cohort study was conducted at Kilimanjaro Christian Medical Centre in Moshi, Tanzania, from 19 November 2023 to 13 April 2024. Fetal position was categorized as a clock; 10 to 2 o'clock as occiput anterior (OA) position, 3 o'clock as left occiput transverse (LOT) position, 4 to 8 o'clock as occiput posterior (OP) position, and 9 o'clock as right occiput transverse (ROT) position. Three trained doctors conducted ultrasound examinations, while midwives performed clinical assessments recording position, station, cervical dilatation, caput succedaneum, and molding. Fetal station ranged from -5 to +5. Caput succedaneum was first graded as 0, +1, +2, or +3 and recategorized as none (0), slight (grade +1), and huge (grade +2 and +3). Molding was categorized as absent or present. Clinical misdiagnosis of position was defined as any discrepancy between clinical and ultrasound assessments using ultrasound examinations as the gold standard. Misdiagnosis included cases where the fetal head position could not be determined clinically, as well as those that were incorrectly assessed through vaginal examination. Blinded comparisons evaluated agreement and clinical misdiagnosis. Factors contributing to clinical misdiagnosis were assessed using logistic regression and potential confounders included maternal age, gestational age, body mass index (BMI), station during the second stage, caput succedaneum, and molding.

RESULTS

The final study population comprised 215 women, 204 were examined in the active first labor stage and 210 in the second stage. Fetal position could not be determined clinically in 40/204 (19.6%) women in the active first stage, but ultrasound successfully determined position in all cases. The overall agreement rate in the active first stage was 101/164 (61.6%), with 60/76 (78.9%) agreement in OA position and 41/88 (46.6%) agreement in non-OA positions. Agreement in classifying position into four categories (OA, LOT, OP, and ROT) was moderate (Cohen's kappa,  = 0.41, 95% CI 0.31-0.52). In the second stage, clinical assessment failed in 11/210 (5.2%) cases, while ultrasound failed in 5/210 (2.4%). The overall agreement rate was 155/194 (79.9%), with 133/152 (87.5%) agreement in OA position and 22/42 (52.4%) agreement in non-OA positions. Cohen's kappa in classifying position into four categories showed moderate agreement ( = 0.46, 95% CI 0.32-0.59). Presence of molding showed a strong association with clinical misdiagnosis, adjusted OR 5.81 (95% CI 1.95-17.30). Slight caput succedaneum was not associated; however, a huge caput succedaneum was significantly associated with clinical misdiagnosis, adjusted OR 5.95 (95% CI 1.85-19.13). Fetal station showed a significant inverse association in unadjusted analysis, where a lower fetal station reduced the likelihood of clinical misdiagnosis, unadjusted OR 0.63 (95% CI 0.41-0.96), but this association was not significant in the adjusted analysis, OR 0.64 (95% CI 0.39-1.06). Maternal age, gestational age, and BMI were not associated with clinical misdiagnosis in either model. The cesarean section (CS) rate among parturients with misdiagnosed positions was 19/50 (38.0%), compared to 14/155 (9.0%) in women with correctly assessed positions ( < .001).

CONCLUSIONS

Clinical and ultrasound assessments showed moderate agreement. Caput succedaneum and molding influenced clinical misdiagnosis, and misdiagnosed positions were associated with higher CS rates.