Research
www. AJOG.org
OBSTETRICS
Accuracy of a single fetal weight estimation at 29-34 weeks in diabetic pregnancies: can it predict large-for-gestational-age infants at term? Avi Ben-Haroush, MD; Rony Chen, MD; Eran Hadar, MD; Moshe Hod, MD; Yariv Yogev, MD OBJECTIVES: This study was undertaken to evaluate the accuracy of a single sonographic estimated fetal weight at 29-34 weeks’ gestation with respect to birthweight determination in diabetic pregnancies. STUDY DESIGN: A retrospective cohort study of 423 diabetic pregnan-
cies with detailed fetal measurements at 29-34 weeks’ gestation. Multivariate regression analysis was used to predict the birthweight. The percentiles of the estimated fetal weight and the calculated birthweight were compared with the actual birthweight percentile.
with good control. On multivariate analysis, the estimated fetal weight, interval from ultrasound to delivery, hemoglobin A1C level, gestational age at ultrasound, and classification of glycemic control were independently associated with the birthweight. Both the estimated fetal weight and the calculated birthweight had a low sensitivity and a low positive predictive value for predicting large-for-gestational-age infants.
RESULTS: The mean birthweight percentile at term was significantly
CONCLUSION: Accelerated fetal growth is evident primarily in diabetic women with poor glycemic control. These fetuses cannot be identified by a single ultrasound examination at 29-34 weeks’ gestation.
higher than the estimated fetal weight percentile at 29-34 weeks’ gestation in the women with poor glycemic control, but not the women
Key words: diabetes, estimated fetal weight, LGH, ultrasound
Cite this article as: Ben-Haroush A, Chen R, Hadar E, Hod M, Yogev Y. Accuracy of a single fetal-weight estimation at 29-34 weeks in diabetic pregnancies: can it predict large-for-gestational-age infants at term? Am J Obstet Gynecol 2007;197:497.e1-497.e6.
A
ccurate prenatal estimation of birthweight may be useful in the management of pregnancy, especially when complicated by diabetes. The mean percentage difference between the sonographically predicted and the actual birthweight ranges from 6-15%,1,2 depending on the presence of various pregnancy complications, such as fetal growth restriction and fetal macrosomia.3-8 The interval from the ultrasound examination to delivery may also influence accuracy. In most of the relevant
From the Helen Schneider Hospital for Women, Rabin Medical Center, Beilinson Hospital, Petach Tikva, and the Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel. Presented at the 27th Annual Clinical Meeting of the Society for Maternal–Fetal Medicine, Feb. 5-10, 2007, San Francisco, CA. Received Oct. 14, 2006; revised Feb. 24, 2007; accepted Apr. 17, 2007. Reprints not available from the authors. 0002-9378/$32.00 © 2007 Mosby, Inc. All rights reserved. doi: 10.1016/j.ajog.2007.04.023
studies to date, sonograms were performed in early labor or within the week preceding labor to eliminate the potential effects of fetal growth on the estimated weight.2-8 To predict fetal weight remote from term, Mongelli and Gardosi9 proposed a sonographic gestation-adjusted formula that was based on the assumption that normal fetuses do not cross their growth curve percentiles. Studies have shown that the growth acceleration in large-forgestational-age (LGA) fetuses of diabetic mothers starts in the second trimester, from as early as 18 weeks10 or 24 weeks.11 Moreover, in diabetic pregnancies, a single abdominal circumference measurement at 29-33 weeks’ gestation has been suggested for early identification of LGA fetuses and for selection of pharmacologic treatment to prevent the birth of LGA infants.12 The aim of this study was to evaluate the accuracy of a single ultrasound-based fetal weight estimation at 29-34 weeks’ gestation in diabetic pregnancies, and to determine whether fetuses of diabetic pregnancies remain within their growth curve percentiles af-
ter 34 weeks, in association with the level of glycemic control.
M ATERIALS AND M ETHODS A retrospective cohort study was conducted. We reviewed the charts of all pregnant women diagnosed with type 1, type 2, or gestational diabetes mellitus (GDM) who attended our maternal–fetal outpatient clinic between 1993-2004. For the current study, the inclusion criteria were a singleton pregnancy, normal anatomy scanning at 14-16 or 20-22 weeks’ gestation, and ultrasound documentation of the fetal biparietal diameter, head circumference, abdominal circumference, and femur length at 29-34 weeks’ gestation. All ultrasound examinations were routinely performed (1 examination for each patient) at our ultrasound unit by experienced ultrasound technicians or physicians. Gestational age was determined by the last menstrual period and by ultrasonographic measurements of the crown-rump length before 12 weeks’ gestation. The local ethics committee approved the study.
NOVEMBER 2007 American Journal of Obstetrics & Gynecology
497.e1
Research
Obstetrics
www.AJOG.org
TABLE 1
Patient characteristics
Selected characteristics Hemoglobin A1C (%)
Overall (n ⴝ 423) 5.9 ⫾ 0.9 (4-10)
Patients who achieved glycemic goals (n ⴝ 289) 5.8 ⫾ 0.8
Patients who did not achieve glycemic goals (n ⴝ 134) 6.3 ⫾ 1.0
P value ⬍.001
................................................................................................................................................................................................................................................................................................................................................................................ 2
Prepregnancy maternal BMI (kg/m )
30.9 ⫾ 5.8 (19-38)
31.0 ⫾ 5.8
30.9 ⫾ 5.9
.862
Maternal weight gain in pregnancy (kg)
10.8 ⫾ 6.0 (–8 to ⫹34)
10.5 ⫾ 5.9
11.6 ⫾ 6.2
.111
Gestational age of ultrasound (wks)
31.3 ⫾ 1.1 (29-34)
31.3 ⫾ 1.1
31.4 ⫾ 1.1
.412
Gestational age at delivery (wks)
37.8 ⫾ 1.7 (31-42)
37.9 ⫾ 1.6
37.7 ⫾ 1.7
.268
................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................
Interval between EFW and delivery (wks)
6.5 ⫾ 1.9 (0.2-12.7)
Macrosomia (⬎4000 g)
47 (11.1%)
6.2 ⫾ 1.7
.096
25 (8.6%)
6.6 ⫾ 2.0
22 (16.4%)
.039
................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................
LGA infants
127 (30%)
72 (24.9%)
55 (41.0%)
.016
Male fetus
49.7%
48.1%
53.6%
.328
................................................................................................................................................................................................................................................................................................................................................................................
Type 1 or type 2 diabetes mellitus was diagnosed before the onset of the current pregnancy. GDM was diagnosed according to the criteria of Carpenter and Coustan13 after a 100-g oral glucose tolerance test. All patients were under the care of a registered dietician for individualized counseling and instructions. The patients used portable glucose meters and were instructed to check their morning fasting blood glucose and the glucose levels before and after the 3 major meals. At the time of the study, we did not use downloaded data from the meters, and we had to rely on patients’ hand-recorded results. Insulin was added to the therapeutic regimen of patients with GDM after diet therapy alone failed to maintain the fasting plasma glucose below 95 mg/dL or 2-hour values below 120 mg/dL or both. On each prenatal visit, on a weekly basis, patients were classified by the treating diabetes specialist according to these fasting and/or postprandial glucose levels as having or not having achieved their glycemic goals (“good” or “poor” glycemic control). For the current study, overall “good” glycemic control was defined if the desired level of glycemic control was recorded in at least 75% of the prenatal visits throughout the third trimester. Fetal weight at 29-34 weeks’ gestation was estimated on the basis of the biometry data with 2 Hadlock’s formulas that use either the abdominal circumference 497.e2
and the femur length (EFW1)14 or the fetal biparietal diameter, head circumference, abdominal circumference, and femur length (EFW2).15 The data were then converted to percentiles according to locally developed growth charts16 for comparison with the birthweight percentiles. The birthweight percentiles were derived by 1 of the authors (E.H.). A previous comparative study of the accuracy of 7 sonographic models in predicting fetal weight in diabetic pregnancies17 showed that formulas that used the abdominal circumference and femur length are as accurate as more complicated models. Therefore, for further analysis, the EFW1 was used. A multivariate linear regression model was fitted to the data to predict the birthweight and the birthweight percentile. The resulting equation (projectile formula) of the stepwise analysis, which included the significant variables, was used to calculate the projected birthweight. The calculated birthweight was then compared with the actual birthweight at delivery. Fetuses or infants with an estimated fetal weight (EFW) or birthweight of 10th or less and 90th or more percentile were categorized as small-for-gestational-age (SGA) and LGA, respectively. Hemoglobin A1C level obtained at 29-34 weeks’ gestation was used for the calculation. For statistical analysis, we used the SPSS statistical package, version 10 for
American Journal of Obstetrics & Gynecology NOVEMBER 2007
Windows (SPSS, Inc, Chicago, IL). Analyses included paired Student t test and multivariate linear regression and stepwise analysis. The Bland and Altman plot and Passing and Bablok regression18,19 were used to compare the calculated birthweight and the actual birthweight. Correlations and differences were considered significant at P ⬍ .05.
R ESULTS A total of 423 women were included in the study; 233 (55.1%) with GDM (32.2% treated with insulin), 132 (31.2%) with type 1 diabetes, and 58 (13.7%) with type 2 diabetes. The selected patient characteristics are shown in Table 1. Multivariate linear regression analysis was used to control for gravidity, parity, fetal sex, gestational age at ultrasound performance, interval from ultrasound assessment to delivery, prepregnancy maternal body mass index (BMI), weight gain during pregnancy, EFW1, level of hemoglobin A1C, and level of glycemic control as predictors of the birthweight (R ⫽ 0.591; R2 ⫽ 0.349; P ⬍ .001). On stepwise analysis, EFW1 (P ⬍ .001), interval from ultrasound assessment to delivery (P ⬍ .001), hemoglobin A1C levels (P ⬍ .001), gestational age at ultrasound examination (P ⫽ .004), and level of glycemic control (P ⫽ .033) were found to
Obstetrics
www.AJOG.org
Research
TABLE 2
Comparison of the EFW* at 29-34 weeks, and the calculated birthweight and their corresponding percentiles with the actual birthweight and its percentile for the entire group of diabetic women (n ⴝ 423) EFW1
EFW2
Calculated birthweight
Birthweight
P value
1809 ⫾ 369
1788 ⫾ 355
3357 ⫾ 328
3314 ⫾ 583
.62† ⬍.01‡
................................................................................................................................................................................................................................................................................................................................................................................
EFW1 percentile
EFW2 percentile
Calculated birthweight percentile
Birthweight percentile
45.5 ⫾ 22.7
44.5 ⫾ 22.7
54.1 ⫾ 25.9
54.0 ⫾ 33.1
................................................................................................................................................................................................................................................................................................................................................................................ §
.96 ⬍.01储 ⬍.01¶
................................................................................................................................................................................................................................................................................................................................................................................
* Hadlock’s formulas: EFW1, abdominal circumference, and the femur length14; EFW2, fetal biparietal diameter, head circumference, abdominal circumference, and femur length.15 For birthweight calculation only EFW1 was used. †
For the difference between the calculated birthweight and actual birthweight.
‡
For the difference between EFW1 and EFW2.
§
For the difference between the calculated birthweight and birthweight percentiles.
储
For the difference between the EFW1 and birthweight percentiles.
¶
For the difference between the EFW1 and EFW2 percentiles.
be independently and significantly associated with the birthweight, as follows: Calculated weight ⫽ 3398.95 ⫹ 0.806 * EFW1 ⫹ 213.37 * interval ⫹ 123 * HbA1C – 93.86 * gestational age – 121.34 * good control. Overall, the mean birthweight percentile (54.0 ⫾ 33.1) was significantly higher than the mean EFW1 percentile (45.5 ⫾ 22.7; P ⬍ .001; Table 2). There was no significant difference between the calculated and the actual birthweight, or between the calculated birthweight and the birthweight percentile (Table 2).
A separate analysis by level of glycemic control (Table 3) showed that in women with good glycemic control, the actual birthweight percentile was similar to both the calculated birthweight percentile and the EFW percentile. However, in women with poor glycemic control, the actual birthweight percentile was similar only to the calculated birthweight percentile, and significantly higher than the EFW percentile. The EFW percentile was within 10% of the actual birthweight percentile in 51.2% and 45.4% of the women with good or poor glycemic control (P ⫽ .311).
In correlation studies, 37.2% of the variation in birthweight could be explained by the calculated birthweight (Figure 1). A separate correlation analysis for women with good (n ⫽ 289, 68%) or poor glycemic control did not improve the results (R2 ⫽ 0.348 and R2 ⫽ 0.368, respectively; P ⬍ .001 for both). Table 4 presents the sensitivity, specificity, and positive and negative predictive values of the EFW1 and the calculated birthweight in predicting LGA infants for all women, and separately for women who achieved or did not achieve
TABLE 3
Comparison of EFW* at 29-34 weeks, and calculated birthweight and their corresponding percentiles with actual birthweight and its percentile for diabetic women with good (n ⴝ 289) or poor (n ⴝ134) glycemic control Patient group
Achieved glycemic goals (n ⴝ 289)
Did not achieve glycemic goals (n ⴝ 134)
P value†
Birthweight (g)
3255 ⫾ 555
3471 ⫾ 603
⬍.001
EFW1 percentile
45.5 ⫾ 22.5
45.6 ⫾ 21.7
NS
EFW2 percentile
44.8 ⫾ 22.6
44.0 ⫾ 22.7
NS
Calculated birthweight percentile
50.3 ⫾ 25.8
62.3 ⫾ 24.2
⬍.001
Birthweight percentile
48.4 ⫾ 33.4
63.5 ⫾ 31.2
⬍.001
P value
.148 .094§ .768储
⬍.001 ⬍.001§ .673储
................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................ ................................................................................................................................................................................................................................................................................................................................................................................ ‡ ‡
................................................................................................................................................................................................................................................................................................................................................................................
* Hadlock’s formulas: EFW1, abdominal circumference, and the femur length14; EFW2, fetal biparietal diameter, head circumference, abdominal circumference, and femur length.15 †
For each category between women with good or poor glycemic control.
‡
For the difference between EFW1 percentile and birthweight percentile.
§
For the difference between EFW2 percentile and birthweight percentile.
储
For the difference between calculated birthweight percentile and birthweight percentile.
NOVEMBER 2007 American Journal of Obstetrics & Gynecology
497.e3
Obstetrics
www.AJOG.org
FIGURE 1
Calculated and actual birthweight CALCULATED BIRTH WEIGHT (gr)
5000
4000
3000
2000 1000
2000
3000
4000
5000
6000
BIRTH WEIGHT (gr)
Correlation between the calculated and actual birthweight (R2 ⫽ 0.372; P ⬍ .001). Ben-Haroush. Predicting LGA infants in diabetic pregnancy. AJOG 2007.
their glycemic goals. Because only 14 infants (3.3%) were diagnosed as SGA, they were excluded from this analysis, which focused on LGA infants. On Bland and Altman plots (Figure 2), the mean difference between the calculated and the actual birthweight was -77.6 g. On Passing and Bablok regression, the difference between the calculated and the actual birthweight showed a significant deviation from linearity (birthweight ⫽ -3445.9981 ⫹ 2.0129 [calculated birthweight]; P ⬍ .01).
C OMMENT In this study, we evaluated the accuracy of a single sonographic estimation of fe-
tal weight at 29-34 weeks’ gestation in predicting LGA infants at term in 423 diabetic pregnancies. The key findings in our study were as follows: (1) the mean actual birthweight percentile at term was significantly higher than the EFW percentile at 29-34 weeks’ gestation in women who did not achieve their glycemic goals, and not significantly different in women with good glycemic control. (2) The multivariate correlation between the calculated birthweight and the actual birthweight (R2 ⫽ 0.379; P ⬍ .001), although statistically significant, showed a significant deviation from linearity on method comparison tests. (3) Despite the significant linear correlation, both the EFW and the calculated birthweight were characterized by low sensitivity in predicting LGA infants and low positive predictive values. Because of the small number of SGA infants, we were unable to perform an appropriate analysis of the accuracy of ultrasound in predicting SGA infants. The higher birth percentile at delivery could be explained by a systematic error in the EFW itself or by an actual increase in fetal weight and weight percentile between the early third trimester to birth. The similarity of the early EFW percentile and the birthweight percentile in women with who achieve and maintain desired glycemic goals supports the second explanation. Therefore, it could be suggested that the fetuses of women who achieve desired glycemic goals do not
FIGURE 2
Difference between calculated and actual birthweight 2000 1500 1000 BW - Cal_BW
Research
+1.96 SD 864.9
500 Mean
0
-11.9 -500 -1.96 SD -1000
-888.7
-1500 1500
2000
2500
3000
3500
4000
4500
5000
AVERAGE of BW and Cal_BW
Bland and Altman plot showing the difference between the calculated birthweight (CA-BW), and the actual birthweight (BW) in 398 women. The differences between the 2 techniques are plotted against the averages of the 2 techniques. Ben-Haroush. Predicting LGA infants in diabetic pregnancy. AJOG 2007.
significantly cross their own growth curves percentiles, at least not after 29-34 weeks’ gestation. The reason for the use of ultrasound in pregnancy in diabetic women, besides early identification of congenital malformations, is recognition of deviant fetal growth.19 Earlier detection of LGA fetuses may enable earlier intervention and improve pregnancy outcome. More than 20 years ago, Ogata et al,21 by serial ultrasound examinations, reported an association in diabetic mothers of an abnormally large fetal abdominal circumference between 28-32 weeks’
TABLE 4
Sensitivity, specificity, positive and negative predictive values (PPV and NPV), and likelihood ratio of a positive test (LRⴙ) of EFW1 percentile at 29-34 weeks and the calculated birthweight percentile for diagnosing LGA (EFW >90th percentile) infants, for entire sample by glycemic control Patient groups
Sensitivity
Specificity
PPV
NPV
LRⴙ
All women
.......................................................................................................................................................................................................................................................................................................................................................................
EFW1 percentile
24.2%
96.3%
76.9%
71.6%
6.54
Calculated birthweight percentile
42.2%
83.7%
57.6%
73.5%
2.58
....................................................................................................................................................................................................................................................................................................................................................................... ................................................................................................................................................................................................................................................................................................................................................................................
Achieved glycemic goals (n ⫽ 289)
.......................................................................................................................................................................................................................................................................................................................................................................
EFW1 percentile
24.2%
94.9%
65.3%
76.1%
4.74
Calculated birthweight percentile
34.3%
85.3%
48.9%
76.0%
2.33
....................................................................................................................................................................................................................................................................................................................................................................... ................................................................................................................................................................................................................................................................................................................................................................................
Did not achieve glycemic goals (n ⫽ 134)
.......................................................................................................................................................................................................................................................................................................................................................................
EFW1 percentile
20.4%
100%
100%
60.2%
NA
66.6%
2.56
................................................................................................................................................................................................................................................................................................................................................................................
Calculated birthweight percentile
497.e4
53.0%
American Journal of Obstetrics & Gynecology NOVEMBER 2007
79.3%
68.4%
Obstetrics
www.AJOG.org gestation with accelerated fetal somatic growth. Since then, despite extensive research, little information has been reported on the type and timing of accelerated fetal growth in this patient population.21,22 The abnormal fetal growth pattern becomes apparent after 26-28 weeks’ gestation,20 when it is confined mostly to the abdominal circumference and subcutaneous fat.19 According to different researchers, onset of the fetal growth acceleration in future LGA infants ranges from as early as 18 weeks10 to about 24 weeks11 or later, between 28-32 weeks20 to week 32.21 Therefore, some authors have suggested the existence of both early and late accelerated growth patterns in fetuses of gestational and pregestational diabetic women.19 Several investigators have proposed that an abdominal circumference beyond the 75th percentile at 28-33 weeks’ gestation could serve as the criterion for fetal risk of macrosomia and an indication for the initiation of pharmacologic therapy.12,23-25 This assumption was based on the finding of a lower rate of delivery of LGA infants in diabetic pregnancies treated with insulin (13%) rather than with diet only (45%). However, others found that when targeted levels of glycemic control were not achieved, the rate of macrosomia was high in fetuses with an abdominal circumference above and below the 75th percentile, regardless of the treatment modality.26,27 Thus, the predictive value of a single, third-trimester abdominal circumference measurement is questionable, further supporting the use of serial ultrasound examinations for the assessment of fetal growth. In our study, the accelerated fetal growth was evident primarily in women who did not achieve their glycemic goals. Importantly, these fetuses were not identified at the single ultrasound examination at 29-34 weeks’ gestation. It is likely that serial ultrasonographic measurements would perform better in the early detection of those fetuses that cross their own growth percentile. This was demonstrated in normal pregnancy by Hedriana and Moore,28 who showed that multiple abdominal circumference percentile measurements resulted in im-
proved predictions for SGA (sensitivity 100%, specificity 88%) and LGA (sensitivity 84%, specificity 100% ) infants. In addition, in their analysis, at 32-36 weeks’ gestation, either the single or multiple EFW examinations were within 10% of the actual birthweight percentile approximately 50% of the time, the same as our results. The assessment of fetal growth and clinical decision making for management should also include other measurements, such as fetal heart size, body composition, and liver size, all of which assist the clinician in differentiating the constitutionally normal fetus from the abnormally large one.19 In addition, clinical factors, such as glycemic profile and obesity, should be incorporated. The relatively high proportion of LGA infants in our study, even in women who achieved the desired glycemic goal, could be explained by misclassification of level of glycemic control by the treating physician. Alternatively, it is possible that the desired goals of fasting plasma glucose below 95 mg/dL and/or 2-hour values below 120 mg/dL or both (yielding a mean HbA1c level of 5.8 ⫾ 0.6%), are not sufficient to prevent excessive fetal growth. In previous studies, we showed that continuous glucose monitoring may identify high blood glucose levels that are unrecognized by intermittent blood glucose monitoring,29 and that in normal pregnancies, the mean fasting and postprandial glucose levels are at least 1 SD lower than the recommended glycemic goals in the diabetic pregnancies.30 We also found that in diabetic pregnancies, the postprandial glucose peak occurs approximately 90 minutes after meals throughout the day and is not affected by the level of glycemic control.31 Nevertheless, it is not yet clear if changing the clinical management according to these findings will improve the obstetric outcome. Another potential explanation for LGA despite “good” glycemic control is that maternal glycemia explains only a small fraction of neonatal weight.32 Recently, we reported that in normal pregnancies, fetal weight estimation in the early third trimester poorly predicts the birthweight percentile at term.33 The mean birthweight percentile at term was
Research
significantly lower than the mean EFW percentile at 28-34 weeks’ gestation in low-risk pregnancies, and the EFW percentile had a low sensitivity and a low positive predictive value for diagnosing SGA or LGA infants. However, the multivariate projectile formula had a high negative predictive value for diagnosing SGA or LGA infants. Because the prevalence of SGA and LGA infants is relatively low in pregnancies uncomplicated by maternal disease (7.3% and 8.8% in the mentioned study, respectively), we speculated that in pregnancies at higher risk for LGA infants, the test would yield higher positive predictive values. Nevertheless, in the current study, the higher prevalence of LGA infants (34.3%) did not improve the predictive accuracy of early ultrasound measurement. In conclusion, the EFW at 29-34 weeks’ gestation in diabetic pregnancies is characterized by a low sensitivity for diagnosing LGA infants at term and a low positive predictive value. Accelerated fetal growth is evident primarily in diabetic women with poor glycemic control. These fetuses are not identified by a single ultrasound examination at 29-34 f weeks’ gestation. REFERENCES 1. Pressman EK, Bienstock JL, Blakemore KJ, Martin SA, Callan NA. Prediction of birth weight by ultrasound in the third trimester. Obstet Gynecol 2000;95:502-6. 2. Ben-Haroush A, Yogev Y, Hod M. Fetal weight estimation in diabetic pregnancies and suspected fetal macrosomia. J Perinat Med 2004;32:113-21. 3. Guidetti DA, Divon MY, Bravermann JJ, Langer O, Merkatz IR. Sonographic estimation of fetal weight in the intrauterine growth retardation population. Am J Perinatol 1990;7:5-7. 4. O’Reilly-Green CP, Divon MY. Receiver operating characteristics curves of sonographic estimated fetal weight for prediction of macrosomia in prolonged pregnancies. Ultrasound Obstet Gynecol 1997;9:403-8. 5. Benacerraf BR, Gelman R, Frigoletto FD. Sonographically estimated fetal weights: accuracy and limitation. Am J Obstet Gynecol 1988;159:1118-21. 6. Pollack RN, Hauer-Pollack G, Divon MY. Macrosomia in postdates pregnancies: the accuracy of routine ultrasonographic screening. Am J Obstet Gynecol 1992;167:7-11. 7. Ben-Haroush A, Yogev Y, Bar J, et al. Accuracy of sonographic estimated fetal weight in 840 women with different pregnancy complica-
NOVEMBER 2007 American Journal of Obstetrics & Gynecology
497.e5
Research
Obstetrics
tions prior to induction of labor. Ultrasound Obstet Gynecol 2004;23:172-6. 8. Alsulyman OM, Ouzounian JG, Kjos SL. The accuracy of intrapartum ultrasonographic fetal weight estimation in diabetic pregnancies. Am J Obstet Gynecol 1997;177:503-6. 9. Mongelli M, Gardosi J. Gestation-adjusted projection of estimated fetal weight. Acta Obstet Gynecol Scand 1996;75:28-31. 10. Wong SF, Chan FY, Oats JJ, McIntyre DH. Fetal growth spurt and pregestational diabetic pregnancy. Diabetes Care 2002;25:1681-4. 11. Greco P, Vimercati A, Scioscia M, Rossi AC, Giorgino F, Selvaggi L. Timing of fetal growth acceleration in women with insulin-dependent diabetes. Fetal Diagn Ther 2003;18:437-41. 12. Buchanan TA, Kjos SL, Montoro MN, et al. Use of fetal ultrasound to select metabolic therapy for pregnancies complicated by mild gestational diabetes. Diabetes Care 1994;17: 275-83. 13. Carpenter MW, Coustan DR. Criteria for screening tests for gestational diabetes. Am J Obstet Gynecol 1982;144:768-73. 14. Hadlock FP, Harrist RB, Carpenter Rj, Deter RL, Park SK. Sonographic estimation of fetal weight. Radiology 1984;150:535-40. 15. Hadlock FP, Harrist RB, Sharman RS, Deter RL, Rossavik UK. Estimation of fetal weight with the use of head, body, and femur measurements: a prospective study. Am J Obstet Gynecol 1985;151:333-7. 16. Leiberman JR, Fraser D, Weitzman S, Glezerman M. Birthweight curves in southern Israel populations. Isr J Med Sci 1993;29:198-203. 17. McLaren RA, Puckett JL, Chauhan SP. Estimators of birth weight in pregnant women re-
497.e6
www.AJOG.org quiring insulin: a comparison of seven sonographic models. Obstet Gynecol 1995;85: 565-9. 18. Bland JM, Altman DG. Measuring agreement in method comparison studies. Stat Methods Med Res 1999;8:135-60. 19. Passing H, Bablok W. A new biometrical procedure for testing the equality of measurements from two different analytical methods. Application of linear regression procedures for method comparison studies in Clinical Chemistry, Part I. J Clin Chem Clin Biochem 1983;21:709-20. 20. Langer O. Ultrasound biometry evolves in the management of diabetes in pregnancy. Ultrasound Obstet Gynecol 2005;26:585-95. 21. Ogata ES, Sabbagha R, Metzger BE, Phelps RL, Depp R, Freinkel N. Serial ultrasonography to assess evolving fetal macrosomia. JAMA 1980;243:2405-8. 22. Landon MB, Mintz MC, Gabbe SG. Sonographic evaluation of fetal abdominal growth: predictor of the large-for-gestational age infant in pregnancies complicated by diabetes mellitus. Am J Obstet Gynecol 1989;160:115-21. 23. Basel D, Lederer R, Diamant YZ. Longitudinal ultrasonic biometry of various parameters in fetuses with abnormal growth rate. Acta Obstet Gynecol Scand 1987;66:143-9. 24. Rossi G, Somigliana E, Moschetta M, Bottani B, Barbieri M, Vignal M. Adequate timing of fetal ultrasound to guide metabolic therapy in mild gestational diabetes mellitus. Acta Obstet Gynecol Scand 2000;79:649-54. 25. Schaefer-Graf UM, Kjos SL, Fauzan OH, et al. A randomized trial evaluating a predominately fetal growth-based strategy to guide management of gestational diabetes in
American Journal of Obstetrics & Gynecology NOVEMBER 2007
Caucasian women. Diabetes Care 2004; 27:297-302. 26. Kjos SL, Schaeffer-Graf U, Sardesi S, et al. A randomized controlled trial using glycemic plus fetal ultrasound parameters versus glycemic parameters to determine insulin therapy in gestational diabetes with fasting hyperglycemia. Diabetes Care 2001;24:1904-10. 27. Rosenn B, Langer O, Brustman L, Skorupski J, Yogev Y. It’s not just the abdominal circumference: glycemic control still matters. Am J Obstet Gynecol 2004;189: 165. 28. Hedriana HL, Moore TR. A comparison of single versus multiple growth ultrasonographic examinations in predicting birth weight. Am J Obstet Gynecol 2004;70:1600-6. 29. Rosenn B, Langer O, Brustman L, Yogev Y. Should insulin therapy in gestational diabetes be guided by fetal or maternal criteria? Am J Obstet Gynecol 2004;189:66. 30. Yogev Y, Ben-Haroush A, Chen R, Rosenn B, Hod M, Langer O. Diurnal glycemic profile in obese and normal weight nondiabetic pregnant women. Am J Obstet Gynecol 2004;191: 949-53. 31. Ben-Haroush A, Yogev Y, Chen R, Rosenn B, Hod M, Langer O. The postprandial glucose profile in the diabetic pregnancy. Am J Obstet Gynecol 2004;191:576-81. 32. Sacks DA, Liu AI, Tsadik GW, Amini SB, Presley LH, Catalano PM. What proportion of birth weight is attributable to maternal glucose among infants of diabetic women? Am J Obstet Gynecol 2006;194:501-7. 33. Ben-Haroush A, Yogev Y, Hod M, Bar J. Predictive value of a single early fetal weight estimate in normal pregnancies. Eur J Obstet Gynecol Reprod Biol 2007;130:187-92.