G. Son

Random urine protein/creatinine ratio readily predicts proteinuria in preeclampsia.

Objective To assess the diagnostic accuracy of random urine protein-creatinine (P/C) ratio for prediction of significant proteinuria in preeclampsia as an alternative to the time-consuming 24-hour urine protein collection. Methods Retrospective record analysis was performed on 140 pregnant women who were admitted with suspicion for preeclampsia from January 2006 to June 2011. Random urine protein and/or 24-hour urine protein levels were assessed and their correlation to random urine P/C ratio and 24-hour urine protein excretion was evaluated. Results Out of 140 patients, random urine P/C ratio or/and 24-hour urine protein was performed in 79 patients to evaluate significant proteinuria. Of 79 patients, 46 (58%) underwent both tests whereas in 33 women (42%) 24-hour urine collection was not available due to urgent delivery. In 39 cases (85%), significant proteinuria (≥300 mg/24 hr) was detected with 6 cases (13%) having values over 5,000 mg/24 hr, corresponding to the diagnosis of severe preeclampsia. Random urine P/C ratio highly correlated with 24-hour urine protein excretion (r=0.823, P<0.01). The optimal random urine P/C ratio cutoff points were 0.63 and 4.68 for 300 mg/24 hr and 5,000 mg/24 hr of protein excretion, respectively. with each sensitivity, specificity, and positive and negative predictive values of 87.1%, 100%, 100%, and 58.3%; and 100%, 85%, 50%, and 100%, for significant and severe preeclampsia, respectively. Conclusion Random urine P/C ratio is a reliable indicator of significant proteinuria in preeclampsia and may be better at providing earlier diagnostic information than the 24-hour urine protein excretion with more accuracy than the urinary dipstick test.

OP11.03: Prediction of preeclampsia by second trimester serum screening and uterine artery Doppler velocimetry

Objective: To determine the performance of screening for preeclampsia by combining second trimester maternal serum screening and uterine artery Doppler ultrasound. Methods: A total of 876 women underwent both maternal serum screening (using α-fetoprotein (AFP), uE3, and human chorionic gonadotropin (hCG)), measurement of uric acid, and second trimester uterine artery Doppler ultrasound. The sensitivity, specificity and predictive value of different combinations of tests were compared. Results: Of 876 pregnancies originally participating, 38 (4.3%) women developed preeclampsia. The mean values for hCG and AFP were significantly higher in women with subsequent preeclampsia (P=0.001 and P<0.0001, respectively). Taking into account obstetrical history, hCG, AFP, and uric acid levels, abnormal uterine artery Doppler and parity, a high level of AFP (> 1.5 MoM (multiples of the median)) and the presence of abnormal uterine artery Doppler were significantly and independently associated with a preeclampsia. (adjusted odds ratios: 2.60 95% CI:0.78–8.66, adjusted odds ratios: 5.00 95% CI:2.03–12.32, respectively). The positive predictive value (PPV) for preeclampsia of the uterine artery Doppler ultrasound is 14%, and the combination of an elevated serum AFP level and the presence of abnormal uterine artery Doppler velocimetry had a PPV for preeclampsia of 21%. The sensitivity of the different tests ranged from 2–40%. Conclusions: The combination of serum markers and abnormal uterine artery Doppler ultrasound may help identify women at risk for subsequent preeclampsia. However, the sensitivity of tests is too low to provide an efficient screening.

OP26.10: Effect of deviation from the mid‐sagittal plane on the measurement of fetal nuchal translucency (NT) using Volume NTTM program

in each image plane. Planes with scores higher than an empirically determined threshold value were selected as standard planes by the AM. Two evaluation criteria were used to evaluate the performance of the AM: Positive Predictive Value = True Positive Planes/(True Positive Planes + False Positive Planes), and Sensitivity = True Positive Planes/(True Positive Planes + False Negative Planes). Results: The automated method showed accurate prediction of the manually labelled planes with an average sensitivity of 91.29% and a positive predictive value of 76.29%. Through closer inspection, all false positive planes selected by the automated method contained both the SB and UV in the correct position but had not been selected by the expert because of stricter criteria imposed. Conclusions: Our results show that machine learning can be used to automatically select a correct AC plane from a fetal 3D volume, and is an early first step in the direction of automated fetalsonography. Such an automated method has the potential for reducing the time needed to complete an ultrasound examination or to perform offline examinations. Further work will be needed to evaluate this and to extend it to other areas of fetal biometry.

P31.12: Feasibility of three-dimensional reconstruction and automated measurement of fetal long bones using auto Long Bone ™

Introduction: survival of a normal fetus to term in the presence of hydatidiform mole remains an extremely uncommon occurrence. Prenatal diagnosis is possible by ultrasonography showing molar invasion of the normal placenta without clear margin. Antenatal detection of mole co existing with a viable fetus should warrant genetic analysis and search for gross malformation of the fetus. Case report: we report a rare case exhibiting molar pregnancy and a coexisting normal fetus. A 29 year, primigravida had an episode of vaginal bleeding at 18 weeks which suggested avascular lower placenta with multiple cystic area measuring 12 × 5 × 9 cm with normal fetus. She had two weekly growth scan and Doppler as diagnosis was not clear. It was thought to be multiple placental lakes. Growth and Doppler remains normal and had term SVD. Placenta was sent for histopathology which suggested normal placenta with an area of complete hydatidiform mole. It was difficult to differentiate from mosaic CHM, twin with normal co-twin and CHM. Discussion: coexistent live fetus and molar placenta has never been reported. Complications including pre-eclampsia, hyperthyroidism, vaginal bleeding, persistent gestational trophoblastic disease, preterm labour, miscarriage and anaemia increases. Several factors may influence the outcome in partial mole such as karyotype of the fetus, size of the molar placenta, the speed of molar degeneration and fetal anaemia.

OP11.05: Does fetal image size affect the measurement of nuchal translucency thickness using Volume NT™?

Objectives: During the first trimester screening, NT measurement is one of the main components in calculating the risk for trisomies. Systematic error is common among accredited sonographers in audit trail. According to the Fetal Medicine Foundation criteria for NT measurement, the calipers should be placed at the thickest area of the subcutaneous translucency with the calipers’ arm flush with the line. The objective of this study is to observe the effectiveness of an alignment exercise in improving the accuracy of individual sonographers in caliper placement. Methods: Twenty mid sagittal pictures suitable for NT measurement were selected. Each picture is duplicated into a set of 4 pictures with different caliper placements. (A total of 20 sets = 80 pictures). Fifteen FMF accredited sonographers from two of the teaching hospitals in Singapore took part in the exercise. Twenty MCQ questions were created with these 20 sets of pictures. One out of each set of 4 pictures has the ideal caliper placement. The results were assessed anonymously using an electronic voting system. The aggregate scores were disclosed after each question. The same procedure was repeated again after a 20 minute short revision lecture on criteria of NT measurement by a FMF accredited trainer. The results were saved and statistically analysed for the individual and group accuracy of measurement before and after the lecture. Results: The results are shown in Table 1. Conclusions: There is a statistically significant improvement in caliper placement for FMF accredited sonographers after a caliper placement exercise and lecture. The improvement happened for the sonographers in both the hospital. Continuing education and refresher courses for caliper placement should be conducted from time to time to standardize the caliper placement for FMF accredited sonographers.