Objective: To describe the derivation, accuracy, and limitations of a new formula for estimating birth-weight using symphysiofundal height measurement in centimetres. Methods: The derivation of the formula was described. The formula was then applied to 2646 consecutive parturients with sure or ultrasonic dates and no clinical or sonographic evidence of uterine fibroids, polyhydramnios, fetal anomalies or death. The symphysiofundal height was measured with the patient supine, the bladder empty and the uterus relaxed. The accuracy of the method was assessed by the mean simple error, the mean absolute error, the mean simple percentage error, the mean absolute percentage error and the percentage of estimates within ± 10% of the actual birth-weights. Possible factors which may affect the accuracy of the formula were assessed using the Student's t-test at the 95% confidence level. Results: The mean simple error was +44.9 ±377.2g. The mean absolute error was 254.4 ±342g. The simple percentage error [mean ±SD] was 1.7 ±11.6%. The absolute percentage error [mean ±SD] was 7.7 ±8.9%. 76.5% of estimates were within ±10% of the actual birth-weights. There was a slightly significant tendency of the formula to overestimate birth- weights [p=0.05, one-tailed t-test]. The formula was inaccurate in the estimation of the weights of low birth-weight and extreme premature fetuses. Conclusion: The symphysiofundal height [SFH] in centimetres inserted into the formula [SFH]3/ 16 estimates the birth-weight to ±10% of the actual birth-weight in grams in 76.5% of parturients. The limitations of the formula are discussed.
