Comparing the accuracy of maternal, clinical, and ultrasound estimations to predict birthweight: a meta-analysis

Abstract

Introduction: Accurate prediction of birthweight prior to delivery is necessary to establish periconceptional strategies to reduce the rates of adverse pregnancy outcomes. This study was performed to determine whether maternal, clinical, or ultrasound estimation provides the best predictor of birthweight. Material and methods: Meta-analysis was performed to summarize the relative ratios of the rate of one of maternal, clinical, and ultrasound estimation within 10% of actual birthweight to that of each of the others within the same range and the difference in correlation coefficient of actual birthweight between clinical and ultrasound estimations. Study quality was assessed with regard to the representativeness of participants, verification bias, and review bias. Publication bias was assessed using Egger's test. Results: Thirty-seven relatively good quality studies were included in the analysis. There was no significant difference in the rates of estimates within 10% of actual birthweight between clinical and maternal estimations (n = 8 and p = 0.491) or between ultrasound and maternal estimations (n = 6 and p = 0.568), or in the correlation coefficient of actual birthweight with ultrasound estimation compared with clinical estimation (n = 5 and p = 0.621). However, the rate of ultrasound estimations within 10% of actual birthweight was significantly higher than that of clinical estimations (n = 18 and p = 0.033). The data showed no publication bias (p = 0.058–0.854). Conclusions: Ultrasound estimation may be the best predictor of birthweight.