Value of a single early third trimester fetal biometry for the prediction of birth weight deviations in a low risk population

Objective: To analyze the value of a single ultrasound biometry examination at the onset of the third trimester of pregnancy for the detection of small-for-gestational-age (SGA) and large-for-gestational-age (LGA) at birth in a low risk population. The aim of this study was to develop a simple and useful method for the detection of growth deviations during pregnancy in primary care (midwife or general practitioner) practices. Setting: A Dutch primary care midwifery practice. Study design: In an earlier study, we developed parity and sex specific fetal growth charts of abdominal circumference (AC) and head circumference (HC) on the basis of ultrasound data of a low-risk midwifery population in the Netherlands. In the present study, we calculated sensitivity, specificity and predictive values at different cut-off points of AC and HC for the prediction of growth deviations at birth. Patients booked for perinatal care between 1 January 1993 and 31 December 2003 (n=3449) were used for the identification of cut-off points (derivation cohort) and those admitted between 1 January 2004 and 31 December 2005 (n=725) were used to evaluate the performance of these cut-offs in an independent population (validation cohort). For the determination of SGA and macrosomia at birth, we used the recently published Dutch birth weight percentiles. Results: Most promising cut-offs were AC ≤25th percentile for the prediction of SGA (birth weight ≤10th percentile) and AC ≥75th percentile for the prediction of macrosomia (birth weight ≥90th percentile). Within the validation cohort these cut-offs performed slightly better than in the derivation cohort. For the prediction of SGA, sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were 53% (95% CI 49–58%), 81% (95% CI 80–83%), 26% (95% CI 23–29%), and 93% (95% CI 93–94%), respectively. The false positive rate was 74%. For the prediction of macrosomia, the values of these parameters were 64% (95% CI 59–69%), 80% (95% CI 78–81%), 23% (95% CI 20–26%), and 96% (95% CI 95–97%), respectively. Here, false positive rate was 77%. No cut-offs were found that predicted extreme birth weight deviations (≤2.3 percentile; ≥97.7 percentile) sufficiently well. Conclusions: In a low risk population, we could predict future growth deviations with a higher sensitivity and in a significant earlier stage (at the onset of the third trimester of pregnancy) than with the use of conventional screening methods (i.e., pa