Maternal Estimates of Neonatal Birth Weight in Diabetic Patients.

ABSTRACT

Background. No consensus exists on estimation of birth weight or prediction of fetal macrosomia in the diabetic population. We compared intrapartum clinical, maternal, and ultrasonographic estimates of birth weight in diabetic patients.

Methods. Diabetic patients at term had clinical, maternal, and ultrasonographic estimations of fetal weight. The mean absolute error, standardized error, and percentage of estimates within 10% of actual birth weight were determined for the three estimates. Effects of maternal weight, parity, and recent ultrasonography were assessed, and statistical analysis was done.

Results. With 32 women enrolled, no statistical difference was seen among clinical (11%), maternal (8.8%), and ultrasonographic (8.0%) birth weight estimates. No difference was seen in accuracy of the three estimates. Estimates were within 10% of actual birth weight in 69% of clinical and maternal estimations and 75% of ultrasonographic estimations. Maternal weight, parity, and recent ultrasonographic evaluation did not affect accuracy of predictions.

Conclusions. Intrapartum maternal estimation of fetal weight in diabetic patients is as accurate as clinical and ultrasonographic predictions.

ACCURATE ANTENATAL ASSESSMENT of fetal birth weight is essential, especially in diabetic mothers who are at increased risk for complications of labor and delivery due to fetal macrosomia. Maternal and fetal outcomes such as prolonged labor with increased use of oxytocin, genital laceration, cesarean section, postpartum hemorrhage, shoulder dystocia, and neonatal asphyxia are more commonly encountered with macrosomic infants (>4,000 g) (1,2) Accurate prediction of macrosomia may alert the physician and staff and enable them to prepare for shoulder dystocia. The two methods most often used for the prediction of fetal weight are clinical assessment (Leopold's maneuvers) and ultrasonographic examination. Using either of these methods, accurate estimation (within 10% of actual birth weight) is made 30% to 70% of the time. (2) Moreover, neither method is accurate in predicting fetal macrosomia. If the clinical or ultrasonographic estimate of fetal weight is >4,500 g, the incidence of actual birth weight >4,000 g is approximately 50%. (2,3) Studies have shown that maternal intrapartum estimation of fetal weight is as accurate as either clinical or ultrasonographic prediction of fetal weight in normal and postterm pregnancies. (3-5) The purpose of our study was to compare the accuracy of maternal, clinical, and ultrasonographic assessment of fetal weight in diabetic women.

MATERIAL AND METHODS

Our study was approved by the Committee for the Protection of Human Subjects at our institution. From May 1999 to April 2000, 32 diabetic patients who were admitted to Memorial Hermann Hospital, Labor and Delivery (Houston, Tex) consented to participate in the study. They were instructed that the purpose of the study was to determine the accuracy of several different ways of predicting neonatal birth weight--clinical, maternal, and ultrasonographic measurements. Inclusion was limited to any diabetic patient in labor with a single fetus of 36 to 42 weeks' gestation. Subjects were included independent of age, race, parity, obstetric history, maternal weight, prenatal care, amniotic fluid index, or previous birth weights.

Once enrolled, the admitting physician first made a clinical estimate of birth weight using Leopold's maneuvers. The patient was then asked, "How much do you think your baby weighs?" Finally, the physician used bedside ultrasonography, estimating fetal weight using the model proposed by Hadlock et al. (6) The three estimates of fetal weight, patient demographic data, and actual birth weight were recorded on data sheets that were kept separate from the patient's chart. Each method of estimation was compared as mean absolute error (estimate minus actual birth weight), standardized error (absolute error/actual birth weight), and percentage of estimates within 10% of actual birth weight. Assuming ultrasonographic or clinical estimation of fetal weight would be accurate (within 10% of actual birth weight)3 70% of the time, we calculated that 32 subjects would be needed to show a difference of at least 35% in accuracy by maternal estimation with 80% power and a of 0.5. Statistical analysis was done using the [X.su p.2], student t test, or Pearson correlation when appropriate. A P value of < .05 was considered significant.

RESULTS

Thirty-two women were enrolled during the study period. The mean age of the study group was 28[+ or -]5.9 years, and the mean gestational age at enrollment was 38.4[+ or -]1.4 weeks. Mean maternal weight of the patients was 202[+ or -]30.7 pounds. Ten of the women were nulliparous. Twenty-seven women had gestational diabetes--3 had type 1 diabetes and 2 had type 2 diabetes. Nineteen women had documented sonograms, and of these, 6 had ultrasonography within 6 weeks of delivery.

Estimated fetal weight was 3,583[+ or -]380 g by clinical (Leopold's) assessment, 3,558[+ or -]497 g by maternal report, and 3,536[+ or -]507 g by ultrasonography (not significant). There was no significant difference in mean absolute error or mean standardized error for the clinical, maternal, and ultrasonographic estimations of fetal weight (Table). Using clinical or maternal estimation, fetal weight was within 10% of actual birth weight 69% of the time versus 75% of the time for ultrasonographically derived estimates (not significant).

No significant correlation was found between maternal weight and absolute or standardized error of clinical (r= 0.27), maternal (r= 0.07), or ultrasonographic (r= 0.13) estimates of birth weight. There was also no correlation between parity and accuracy of maternal estimation of fetal weight (r= 0.00). In the 19 women with a previous sonogram, there was no correlation between interval since last ultrasonography and absolute or standardized error of clinical, maternal, or ultrasonographic estimates.

DISCUSSION

The rate of fetal macrosomia is increased in women with diabetes, thus increasing both fetal and maternal risks. (2,7) The incidence of macrosomia in pregnancies complicated by diabetes is 15% to 45 %. (7) Fetal macrosomia may predispose patients to the increased use of oxytocin, prolonged labor, cesarean section, genital lacerations, postpartum hemorrhage, traumatic injury and asphyxia. (1,2) Furthermore, because of the disproportionate size of the trunk and shoulders of diabetic fetuses, there is an increased risk of shoulder dystocia. (7) While the incidence of shoulder dystocia in diabetics is 10%, the risk of traumatic injury is 25% to 50%, (2,7) Based on risk factors alone, the diagnosis of macrosomia would be made in 40% of cases. (2) An accurate diagnosis of macrosomia can lead to a decrease in perinatal morbidity. (8) Its prediction may enable the physician and staff to prepare for shoulder dystocia or prevent a traumatic injury.

A simple, reliable, and inexpensive method of estimating fetal weight in diabetics, such as maternal estimation, would prove invaluable. Clinical estimation using Leopold's maneuvers has been shown to accurately predict birth weight (within 10% of actual birth weight) 54% to 70% of the time. (2) This technique can prove challenging, depending on maternal body habitus, uterine anomalies, or increase in amniotic fluid index. (9) In the 1970s, the use of ultrasonography to estimate fetal weight gained popularity because of the perceived ability to standardize and reproduce measurements. (3) However, ultrasonography has not proven to be more accurate than clinical estimation, with only 30% to 68% of ultrasonographic measurements within 10% of actual birth weight. (2) In diabetic subjects, ultrasonographic estimates of fetal weight were within 10% of estimates in 63% of patients using the formula of Hadlock et al. (9) Intrapartum maternal estimates of fetal weight have proven to be comparable to either clinical or ultrasonographic predictions in both term and postdate gestations. (4,5) Our study shows similar results in a population of diabetic women in labor at term.

Our results suggest that maternal estimation, of birth weight in diabetic pregnancies is as accurate as concurrent clinical or ultrasonographic predictions. Our study had 80% power to detect a 35% difference between accuracy of maternal estimation and established methods (ultrasonography or Leopold's maneuvers) of estimation of fetal weight. Chauhan et al (5) previously reported that age, parity, and maternal weight did not affect the accuracy of maternal estimates. Our study suggests that in diabetic women, neither parity nor maternal weight affect the accuracy of either of the birth weight estimates. Previous investigations of maternal estimation of birth weight have not studied the effect of recent ultrasonography on error. We did not find any correlation between the interval since the most recent antepartum sonogram and the accuracy of any intrapartum method of estimation of fetal weight. However, only six women had a documented sonogram within 6 weeks of delivery.

CONCLUSION

Since diabetic mothers are at increased risk for delivery of a macrosomic fetus, they often have fetal weight estimated at term. Accurate prediction of fetal weight may help physicians and staff prepare for and prevent complications of labor and delivery. Maternal, clinical, and ultrasonographic estimations of fetal weight appear to be comparable in this population. Until better means of assessing fetal weight are established, clinical, maternal, and ultrasonographic evaluations should be used.

References

(1.) Hirata GI, Medearis AL, Horenstein J, et al: Ultrasonographic estimation of fetal weight in the clinically macrosomic fetus. A Gynecol 1990; 162:238-242

(2.) Chauhan SP, Hendrix N: Fetal macrosomia. Obstet Gynecol Manage 1998; 8:75-83

(3.) Herrero RL, Fitzsimmons J: Estimated fetal weight. maternal vs physician estimate. J Reprod Med 1999; 44:674-678

(4.) Chauhan SP, Lutton PM, Bailey KJ, et al: Intrapartum clinical, onographic and parous patients' estimates of newborn birth weight. Obstet Gynecol 1992; 79:956-958

(5.) Chauhan SP, Sullivan CA, Lutton TC, et al: Parous patients' estimate of birth weight in postterm pregnancy. J Perinatol 1995; 15:192-194

(6.) Halock FP: computer-assisted, multiple parameter assessment of fetal age and growth. Semin Ultrasound CT MR 1989; 10:383-395

(7.) Mondalou HD, Sorchester WL, Thorsion A, et al: Macrosomia-maternal, fetal and neonatal implications. Obstet Gynecol 1980; 55:420-424

(8.) Chauhan SP, Hendrix NW, Magann EF, et al: Limitations of clinical and sonographic estimates of birth weight: experience with 1,034 parturients. Obstet Gynecol 1998; 91:72-77

(9.) Alsulyman OM, Ouzounian JG, Kjos SL: The accuracy of intrapartum ultrasonographic fetal weight estimation in diabetic pregnancies. Amf J Obstet Gynecol 1997; 177:503-506

TABLE.

Clinical, Maternal, and Sonographic Estimates of Fetal Weight

                                                 Estimates Within
                 Mean Absolute    Standarized     10 % of Birth
Method             Error (g)       Error (%)        Weight (%)

Clinical         268[+ or -]261   11[+ or -]22          69
Maternal         305[+ or -]294  8.8[+ or -]9.9         69
Ultrasonography  270[+ or -]177  8.0[+ or -]5.7         75

RELATED ARTICLE: KEY POINTS

* Fetal macrosomia is reported to occur in 15% to 45% of diabetic pregnancies.

* After shoulder dystocia, the risk of traumatic injury in diabetics is reported to be 25% to 50%.

* Clinical (Leopold's maneuvers) and ultrasound estimates have a reported positive predictive value of 50% for fetal macrosomia.

* Our results indicate that maternal estimation of fetal weight in labor is as accurate as clinical or ultrasound estimation of fetal weight in diabetic patients.

* Maternal, clinical, or ultrasound estimation of fetal weight in diabetic women in labor is not influenced by maternal weight, parity, or recent previous ultrasound.