Application of Prospect Theory in Obstetrics by Evaluating Mode of Delivery and Outcomes in Neonates Born Small or Appropriate for Gestational Age.

IMPORTANCE

Antenatal diagnosis of fetal weight is challenging, and the detection rate of fetal growth restriction (FGR) is low. Neonates with FGR are known to have an increased rate of obstetric intervention during labor, but the association of antenatal fetal weight estimation with mode of delivery and neonatal outcomes among neonates who are small and appropriate for gestational age (SGA and AGA) has not been reported.

OBJECTIVE

To evaluate the association of antenatal fetal weight estimation with mode of delivery and neonatal outcomes among neonates who are SGA and AGA, applying psychological concepts of cognitive bias and prospect theory to a model of clinical behavior.

DESIGN, SETTING, AND PARTICIPANTS: This cohort study was conducted between 2019 and 2020 using data from 2006 to 2018 at a tertiary care center in Jerusalem, Israel. Participants were 100 198 term singleton neonates without anomalies who were categorized into 4 groups according to the presence of an antenatal suspicion of FGR and final birth weight. Neonates with false positives (FPs; ie, group 1-FP: those with suspected FGR who were AGA) and neonates with true positives (TPs; ie, group 2-TP: those with suspected FGR who were SGA) were compared with neonates with AGA antenatal fetal weight estimation, including neonates with false negatives (FNs; ie, group 3-FN: those not suspected to have FGR who were SGA) and neonates with true negatives (TNs; ie, group 4-TN: those not suspected to have FGR who were AGA). Data were analyzed from July 2019 to July 2020.

EXPOSURES

Fetal weight estimation was performed according to sonographic and clinical evaluation at admission to labor, with FGR defined as a birth weight less than the 10th percentile for gestational age. Sonographic fetal weight estimation was performed according to Hadlock formula. Clinical weight estimation was performed by trained obstetricians.

MAIN OUTCOMES AND MEASURES

The primary outcomes were obstetric intervention and mode of delivery; the secondary outcomes were neonatal Apgar score (with low Apgar score defined as <7) and neonatal intensive care unit (NICU) admission rates.

RESULTS

Among 100 198 neonates eligible for the study (50941 [50.8%] male neonates), there were 5671 neonates in group 1-FP, 3040 neonates in group 2-TP, 8508 neonates in group 3-FN, and 82 979 neonates in group 4-TN. Mean (SD) maternal age was 28.6 (5.7) years. Among 8711 neonates with suspected FGR, 34.9% were below the 10th percentile at birth, while 65.1% were AGA. Neonates with suspected FGR had a significantly increased rate of induction of labor (group 1-FP: 649 neonates [11.4%] and group 2-TP: 969 neonates [31.9%]) compared with neonates in group 3-FN (1055 neonates [12.4%]) and group 4-TN (7136 neonates [8.6%]) (P < .001) and a significantly increased rate of cesarean delivery (group 1-FP: 915 neonates [16.1%] and group 2-TP: 556 neonates [18.3%] vs group 3-FN: 1106 neonates [13.0%] and group 4-TN: 6588 neonates [7.9%]; P < .001). Increased NICU admission was found for neonates who were SGA compared with neonates who were AGA (group 2-TP: 182 neonates [6.0%] and group 3-FN: 328 neonates [3.9%] vs group 1-FP: 51 neonates [0.9%] and group 4-TN: 704 neonates [0.8%]; P <.001), as was increased rate of low Apgar score (eg, at 1 minute: group 2-TP: 149 neonates [4.9%] and group 3-FN: 384 neonates [4.5%] vs group 1-FP: 124 neonates [2.2%] and group 4-TN: 1595 neonates [1.9%]; P < .001). In a multivariable model comparing group 1-FP, group 2-TP, and group 3-FN with group 4-TN, suspicion of FGR was independently associated with increased risk of caesarean delivery among neonates in group 1-FP (odds ratio, 1.72; 95% CI, 1.56-1.88; P < .001).

CONCLUSIONS AND RELEVANCE

This study found that antenatal diagnosis of FGR was independently associated with an increase in risk of caesarean delivery by 70% in neonates who were AGA without improvement in neonatal outcomes. These findings suggest that such outcomes may be explained by application of prospect theory and may be associated with cognitive bias in clinical decision-making.