Optimal threshold for birth weight discordance: Does knowledge of chorionicity matter?

OBJECTIVE: To establish the optimal threshold of birth weight discordance (BWD) for prediction of stillbirth, perinatal mortality and morbidity in twins born in British Columbia with or without chorionicity information. STUDY DESIGN: This is a retrospective population-based cohort study of twins born in British Columbia from 2000 to 2010. Data from one hospital was used to adjust for chorionicity. Multivariate generalized estimating equation and receiver operating characteristic curve analyses were performed to evaluate the predictability of BWD in comparison with other fetal anthropometric measurements. Positive likelihood ratio is used to estimate test accuracy. Survival analysis was conducted to take gestational age and other confounders into account. RESULTS: We analyzed two cohorts, with (pairs = 1493) and without (pairs = 6328) chorionicity information, of which 1.5% experienced stillbirth, 2.9% suffered perinatal mortality and 22.6% identified with perinatal morbidities. BWD was a significant predictor of stillbirth. Standard receiver operating characteristic curve analysis and survival analysis suggested that BWD of >= 30% is the optimal thresholds for stillbirth and perinatal mortality irrespective of chorionicity. However, the P-value for predictive accuracy of BWD was nonsignificant for perinatal morbidity, after adjusting for Confounding variables engaging multivariate analysis. Sex discordance can be used as a proxy for chorionicity. CONCLUSION: BWD is a good predictor for stillbirth. A BWD cutoff limit of 30% and higher has optimal accuracy for detecting perinatal mortality.