- Email: [email protected]
349: Heads or tails: does presentation affect accuracy of ultrasound-estimated fetal weight?
Poster Session II
Diabetes, Labor, Ultrasound-Imaging
CONCLUSIONS: Our data confirm that first-trimester FGR and low PAPP-A are associated with delivery of an SGA infant. Combination of these two parameters results in a modest improvement in screening efficiency for SGA compared to either of them alone.
www.AJOG.org
Table 1: Association of isolated pyelectasis with aneuploidy Maternal age > 35 years (nⴝ18,243)
Isolated Pyelectasis nⴝ236
No Isolated Pyelectasis nⴝ18,007 OR [95% CI]
Maternal age < 35 years (nⴝ39,849)
Combined
Isolated Pyelectasis nⴝ821
OR [95% CI]
No Isolated Pyelectasis nⴝ39,028
OR [95% CI]
LR ⴙ [95% CI]
Table 1: Association between first trimester FGR and low PAPP-A with SGA at delivery Prediction Criteria
SGA n (%)
Sensitivity (%)
Specificity (%)
Area under the ROC Curve
CRL z-score ⬍ ⫺1 (n⫽590)
43(7.2)
23.6
85.6
0.55
PAPP-A ⬍ 5% (n⫽202)
22(10.9)
Either CRL z-score ⬍ ⫺1 or PAPP-A ⬍ 5% (n⫽759)
61(8.0)
Any Aneuploidy
..........................................................................................................................................................................................
Adjusted Odds Ratio [95% CI] 1.78 [1.24-2.56]
12.2
95.7
0.54
2.96 [1.82-4.83]
..........................................................................................................................................................................................
33.0
82.1
0.58
2.17 [1.57-3.00]
..........................................................................................................................................................................................
4(12.1)
3.13
99.2
0.51
4.23 [1.37-13.03]
..........................................................................................................................................................................................
348 The risk for Down Syndrome in fetuses with isolated pyelectasis Jeanine Carbone1, Methodius Tuuli1, Jeffrey Dicke1, George Macones1, Anthony Odibo1 1
Washington University in St. Louis, St. Louis, MO
OBJECTIVE: Isolated pyelectasis has been linked to an increased risk of
aneuploidy. However, studies have produced contradicting results likely due to small sample sizes. Our objective was to test the hypothesis that isolated pylectasis is associated with aneuploidy and calculate likelihood ratios using a large ultrasound database. STUDY DESIGN: We performed a retrospective cohort study of pregnancies presenting to our prenatal ultrasound unit(1996 - 2006) between 16 and 22 weeks gestation. Pyelectasis was defined as fetal renal pelvis ⬎4mm in the anterior-posterior dimension. Fetuses with coexisting major structural anomalies or markers of aneuploidy were excluded. Association between isolated pyelectasis and any aneuploidy as well as specific aneuploidy types was assessed using odds ratios. Screening efficiency was evaluated using sensitivity, specificity, negative/positive predictive values, and negative/positive likelihood ratios. Results were also stratified by maternal age ⱖ 35 and ⬍35 years. RESULTS: Among 62,111 patients who had obstetric ultrasounds during the study period, 1057(1.7%) were diagnosed with isolated pyelectasis. In the overall cohort, 0.6% were diagnosed with aneuploidy. Of fetuses diagnosed with isolated pyelectasis, 12(1.1%) were diagnosed with aneuploidy. The presence of isolated pyelectasis was significantly associated with any aneuploidy (OR 5.26[CI 2.90-9.54]), and specifically trisomy 21(T21) (OR 6.11[CI 3.17-11.76]). When stratified by maternal age, the association with any aneuploidy and T21 remained significant (Table 1). Calculation of risk for all chromosomal abnormalities and T21 indicates greater than a 4-fold increase in the presence of isolated fetal pyelectasis, (LR⫹) 4.16[CI 2.417.19] and 4.73[CI 2.61-8.59] respectively. CONCLUSIONS: Our data confirm a significant association between isolated pyelectasis and increased risk of aneuploidy, particularly T21. This increased risk appears independent of maternal age. These likelihood ratios can be used to adjust aneuploidy risk and be valuable in counseling patients regarding appropriateness of amniocentesis.
6(2.5)
101(0.56)
4.62 [1.64-10.57]
6(0.7)
47(0.1)
6.10 [2.13-14.36]
5.26 [2.90-9.54]
4.16 [2.41-7.19]
.......................................................................................................................................................................................... Trisomy 21
.......................................................................................................................................................................................... n⫽117
..........................................................................................................................................................................................
Both CRL z-score ⬍ ⫺1 and PAPPA ⬍5% (n⫽33)
n⫽160
5(2.1)
75(0.42)
5.17 [1.61-12.78]
5(0.6)
32(0.1)
7.47 [2.27-19.37]
6.11 [3.17-11.76]
4.73 [2.61-8.59]
..........................................................................................................................................................................................
349 Heads or tails: does presentation affect accuracy of ultrasound-estimated fetal weight? Jennifer McNamara1, Anthony Odibo1, George Macones1, Alison Cahill1 1
Washington University in St. Louis, St. Louis, MO
OBJECTIVE: Estimated fetal weight (EFW) is a critical factor in modern
obstetric decision-making. Despite its importance, the accuracy of EFW to predict actual birth weight (ABW) is variable, particularly at late gestation. We sought to determine if accuracy of third trimester ultrasound-EFW varies by fetal presentation. STUDY DESIGN: This is a retrospective cohort study of all viable, singleton pregnancies that underwent a growth scan within three weeks of delivery at our institution during an 18-year study period from 1990 to 2008. Breech presenting fetuses were compared to those presenting cephalic. EFW using the Hadlock formula was compared to ABW and reported as mean absolute difference and mean percentage difference based on ABW ([(EFW-ABW)/ABW] X 100). Subgroup analyses were performed of only women who delivered within 7 days of EFW scan and excluding women with any comorbidities. Logistic regression analysis was used to estimate the risk of inaccuracy by presentation while adjusting for potentially confounding effects. RESULTS: Of 4547 singleton fetuses, 4325 were cephalic and 222 were breech presenting. EFW by Hadlock estimation was slightly less accurate for women with breech compared to cephalic presenting fetuses (Table 1). Pregnancies with a breech presenting fetus were at slightly increased risk to have an EFW⬎10% different than ABW compared to those with cephalic presentation, after adjusting for scan-to-delivery interval, race, fetal gender, and parity (aOR 1.35, 95% CI 1.11-1.80, p⫽0.04). CONCLUSIONS: EFW estimation with Hadlock formula is slightly less accurate for fetuses in breech compared to cephalic presentation. This is both a reasonable consideration for clinical decision making, as well as a potential opportunity to consider other algorithms for greater accuracy by presentation. Breech
Cephalic
p-value
All (nⴝ4,547)
.................................................................................................................................................................................
Mean absolute difference ⫺ 217 ⫾ 324g ⫺ 206 ⫾ 353g 0.64 (g)* ⫹/⫺ SD ................................................................................................................................................................................. Mean Percentage Difference (%)**⫹/⫺ SD
7.4% ⫾ 11.7
6.2% ⫾ 10.8 0.09
..........................................................................................................................................................................................
No comorbidities (nⴝ2,878) ................................................................................................................................................................................. Mean absolute difference ⫺ 238 ⫾ 352g ⫺ 201 ⫾ 352g 0.21 (g)* ⫹/⫺ SD ................................................................................................................................................................................. Mean Percentage Difference (%)**⫹/⫺ SD
11.3% ⫾ 8.5
9.9% ⫾ 7.3
0.04
..........................................................................................................................................................................................
Delivery with 1 week of scan................................................................................................................................................................................. (nⴝ783) Mean absolute difference ⫺71 ⫾ 204g 20 ⫾ 279g 0.03 (g)* ⫹/⫺ SD ................................................................................................................................................................................. Mean Percentage Difference (%)**⫹/⫺ SD
7.0% ⫾ 6.9
7.6% ⫾ 6.0
0.06
..........................................................................................................................................................................................
S142
American Journal of Obstetrics & Gynecology Supplement to JANUARY 2011
Diabetes, Labor, Ultrasound-Imaging
CONCLUSIONS: Our data confirm that first-trimester FGR and low PAPP-A are associated with delivery of an SGA infant. Combination of these two parameters results in a modest improvement in screening efficiency for SGA compared to either of them alone.
www.AJOG.org
Table 1: Association of isolated pyelectasis with aneuploidy Maternal age > 35 years (nⴝ18,243)
Isolated Pyelectasis nⴝ236
No Isolated Pyelectasis nⴝ18,007 OR [95% CI]
Maternal age < 35 years (nⴝ39,849)
Combined
Isolated Pyelectasis nⴝ821
OR [95% CI]
No Isolated Pyelectasis nⴝ39,028
OR [95% CI]
LR ⴙ [95% CI]
Table 1: Association between first trimester FGR and low PAPP-A with SGA at delivery Prediction Criteria
SGA n (%)
Sensitivity (%)
Specificity (%)
Area under the ROC Curve
CRL z-score ⬍ ⫺1 (n⫽590)
43(7.2)
23.6
85.6
0.55
PAPP-A ⬍ 5% (n⫽202)
22(10.9)
Either CRL z-score ⬍ ⫺1 or PAPP-A ⬍ 5% (n⫽759)
61(8.0)
Any Aneuploidy
..........................................................................................................................................................................................
Adjusted Odds Ratio [95% CI] 1.78 [1.24-2.56]
12.2
95.7
0.54
2.96 [1.82-4.83]
..........................................................................................................................................................................................
33.0
82.1
0.58
2.17 [1.57-3.00]
..........................................................................................................................................................................................
4(12.1)
3.13
99.2
0.51
4.23 [1.37-13.03]
..........................................................................................................................................................................................
348 The risk for Down Syndrome in fetuses with isolated pyelectasis Jeanine Carbone1, Methodius Tuuli1, Jeffrey Dicke1, George Macones1, Anthony Odibo1 1
Washington University in St. Louis, St. Louis, MO
OBJECTIVE: Isolated pyelectasis has been linked to an increased risk of
aneuploidy. However, studies have produced contradicting results likely due to small sample sizes. Our objective was to test the hypothesis that isolated pylectasis is associated with aneuploidy and calculate likelihood ratios using a large ultrasound database. STUDY DESIGN: We performed a retrospective cohort study of pregnancies presenting to our prenatal ultrasound unit(1996 - 2006) between 16 and 22 weeks gestation. Pyelectasis was defined as fetal renal pelvis ⬎4mm in the anterior-posterior dimension. Fetuses with coexisting major structural anomalies or markers of aneuploidy were excluded. Association between isolated pyelectasis and any aneuploidy as well as specific aneuploidy types was assessed using odds ratios. Screening efficiency was evaluated using sensitivity, specificity, negative/positive predictive values, and negative/positive likelihood ratios. Results were also stratified by maternal age ⱖ 35 and ⬍35 years. RESULTS: Among 62,111 patients who had obstetric ultrasounds during the study period, 1057(1.7%) were diagnosed with isolated pyelectasis. In the overall cohort, 0.6% were diagnosed with aneuploidy. Of fetuses diagnosed with isolated pyelectasis, 12(1.1%) were diagnosed with aneuploidy. The presence of isolated pyelectasis was significantly associated with any aneuploidy (OR 5.26[CI 2.90-9.54]), and specifically trisomy 21(T21) (OR 6.11[CI 3.17-11.76]). When stratified by maternal age, the association with any aneuploidy and T21 remained significant (Table 1). Calculation of risk for all chromosomal abnormalities and T21 indicates greater than a 4-fold increase in the presence of isolated fetal pyelectasis, (LR⫹) 4.16[CI 2.417.19] and 4.73[CI 2.61-8.59] respectively. CONCLUSIONS: Our data confirm a significant association between isolated pyelectasis and increased risk of aneuploidy, particularly T21. This increased risk appears independent of maternal age. These likelihood ratios can be used to adjust aneuploidy risk and be valuable in counseling patients regarding appropriateness of amniocentesis.
6(2.5)
101(0.56)
4.62 [1.64-10.57]
6(0.7)
47(0.1)
6.10 [2.13-14.36]
5.26 [2.90-9.54]
4.16 [2.41-7.19]
.......................................................................................................................................................................................... Trisomy 21
.......................................................................................................................................................................................... n⫽117
..........................................................................................................................................................................................
Both CRL z-score ⬍ ⫺1 and PAPPA ⬍5% (n⫽33)
n⫽160
5(2.1)
75(0.42)
5.17 [1.61-12.78]
5(0.6)
32(0.1)
7.47 [2.27-19.37]
6.11 [3.17-11.76]
4.73 [2.61-8.59]
..........................................................................................................................................................................................
349 Heads or tails: does presentation affect accuracy of ultrasound-estimated fetal weight? Jennifer McNamara1, Anthony Odibo1, George Macones1, Alison Cahill1 1
Washington University in St. Louis, St. Louis, MO
OBJECTIVE: Estimated fetal weight (EFW) is a critical factor in modern
obstetric decision-making. Despite its importance, the accuracy of EFW to predict actual birth weight (ABW) is variable, particularly at late gestation. We sought to determine if accuracy of third trimester ultrasound-EFW varies by fetal presentation. STUDY DESIGN: This is a retrospective cohort study of all viable, singleton pregnancies that underwent a growth scan within three weeks of delivery at our institution during an 18-year study period from 1990 to 2008. Breech presenting fetuses were compared to those presenting cephalic. EFW using the Hadlock formula was compared to ABW and reported as mean absolute difference and mean percentage difference based on ABW ([(EFW-ABW)/ABW] X 100). Subgroup analyses were performed of only women who delivered within 7 days of EFW scan and excluding women with any comorbidities. Logistic regression analysis was used to estimate the risk of inaccuracy by presentation while adjusting for potentially confounding effects. RESULTS: Of 4547 singleton fetuses, 4325 were cephalic and 222 were breech presenting. EFW by Hadlock estimation was slightly less accurate for women with breech compared to cephalic presenting fetuses (Table 1). Pregnancies with a breech presenting fetus were at slightly increased risk to have an EFW⬎10% different than ABW compared to those with cephalic presentation, after adjusting for scan-to-delivery interval, race, fetal gender, and parity (aOR 1.35, 95% CI 1.11-1.80, p⫽0.04). CONCLUSIONS: EFW estimation with Hadlock formula is slightly less accurate for fetuses in breech compared to cephalic presentation. This is both a reasonable consideration for clinical decision making, as well as a potential opportunity to consider other algorithms for greater accuracy by presentation. Breech
Cephalic
p-value
All (nⴝ4,547)
.................................................................................................................................................................................
Mean absolute difference ⫺ 217 ⫾ 324g ⫺ 206 ⫾ 353g 0.64 (g)* ⫹/⫺ SD ................................................................................................................................................................................. Mean Percentage Difference (%)**⫹/⫺ SD
7.4% ⫾ 11.7
6.2% ⫾ 10.8 0.09
..........................................................................................................................................................................................
No comorbidities (nⴝ2,878) ................................................................................................................................................................................. Mean absolute difference ⫺ 238 ⫾ 352g ⫺ 201 ⫾ 352g 0.21 (g)* ⫹/⫺ SD ................................................................................................................................................................................. Mean Percentage Difference (%)**⫹/⫺ SD
11.3% ⫾ 8.5
9.9% ⫾ 7.3
0.04
..........................................................................................................................................................................................
Delivery with 1 week of scan................................................................................................................................................................................. (nⴝ783) Mean absolute difference ⫺71 ⫾ 204g 20 ⫾ 279g 0.03 (g)* ⫹/⫺ SD ................................................................................................................................................................................. Mean Percentage Difference (%)**⫹/⫺ SD
7.0% ⫾ 6.9
7.6% ⫾ 6.0
0.06
..........................................................................................................................................................................................
S142
American Journal of Obstetrics & Gynecology Supplement to JANUARY 2011