(1107) What to expect when you’re expecting: impact of physician and patient factors on predicted delivery

The cesarean delivery rate continues to rise, and it is critical to understand how physician perceptions of likely mode of delivery influence the delivery outcome. The aim of this study was to perform an early, intrapartum assessment of physician’s predictions of their patient’s likelihood of vaginal delivery (VD) and to identify which physician and patient factors affect these predictions.

Study Design:

We prospectively assessed physician prediction of anticipated mode of delivery for patients presenting to labor and delivery for planned vaginal delivery. At 4 hours, physicians were sent a single-item REDCap survey, “On a scale of 1-10 (where 10 = the highest), what is your patient’s likelihood of a vaginal delivery?” and the value was converted to a probability (0-1.0).

We chart abstracted patient demographics and outcomes and obtained physician demographics through publicly available sources. We used linear mixed models to assess differences in predictions by patient and physician characteristics, and performed subgroup analysis for nulliparous, term, singleton vertex (NTSV) patients.

Results:

We received 203 physician responses from 46 physicians. Table 1 lists physician and patient demographics.  The median prediction was 0.9. Physicians predicted a lower likelihood of VD for patients with obesity and a higher likelihood of VD for multiparas. (Table 2)

The NTSV subgroup was 111 patients delivered by 39 physicians. Physician and patient characteristics were similar to the overall population. The median prediction was 0.8. Physicians predicted a lower likelihood of VD for patients age > 35 and for BMI >30 and a higher likelihood of VD for Asian patients. There were no differences in predictions by physician demographics (age, gender, language, practice duration).

Conclusion:

Numerous patient factors influence physician’s perception of the probability of a VD. Given that we anticipate the physician’s early perception impacts the eventual outcome (anchoring bias), future research will focus on the importance of early recognition to determine strategies to mitigate those without causal association.