Abstract
Accurate antenatal diagnosis of the macrosomic infant of a diabetic mother should lead to a decrease in perinatal morbidity. Several formulas for weight estimation based on sonographic measurements are in common use, but none of them has been devised specifically for the infant of a diabetic mother. We assessed the accuracy of conventional formulas at estimating fetal weight in 160 of these infants and found their relative errors (error as a percentage of birth weight) had standard deviations of 12.2% to 13.1% (greater than has been reported in the general population) with no statistically significant difference among formulas. We also devised customized formulas generated from a "training set" of 80 patients and tested on a second set of 80 patients. The best formula yielded a standard deviation of 11.0%. The difference between customized and conventional formulas was not statistically significant. With use of one conventional formula, the likelihood of macrosomia was found to be 77% when the estimated weight was >4000 gm and 86% when the estimated weight was >4500 gm. We conclude that fetal weight prediction is even less accurate in the infant of a diabetic mother than in the general population and that "customized" formulas fare no better than conventional formulas in diabetics. Despite this, estimated weights can offer guidance in predicting fetal macrosomia and therefore aid in obstetric management.
Original language | English |
---|---|
Pages (from-to) | 441-444 |
Number of pages | 4 |
Journal | American Journal of Obstetrics and Gynecology |
Volume | 156 |
Issue number | 2 |
DOIs | |
State | Published - Feb 1987 |
Externally published | Yes |
Keywords
- Diabetes
- pregnancy
- ultrasound