Fetal-Maternal Hemorrhage in First-Trimester Intrauterine Hematoma.

OBJECTIVE: The objective of this study was to compare the frequency and percentage of fetal hemoglobin (HbF%) by flow cytometry of (1) first-trimester asymptomatic patients with intrauterine hematoma (IUH), (2) first-trimester pregnant patients with vaginal bleeding (VB), and (3) first-trimester asymptomatic pregnant women without hematoma. METHODS: Prospective study involving pregnant women in the first trimester of pregnancy. Patients with ultrasound findings of asymptomatic hematoma and with VB were paired with asymptomatic pregnant women of same gestational age without hematoma (control group [CG]). Maternal blood HbF% was evaluated by flow cytometry. The groups were compared in terms of circulating fetal hemoglobin and HbF%. RESULTS: Sixty-six patients were selected, 22 with hematoma, 17 with bleeding, and 27 in the CG. Fetal hemoglobin was detected in 15 patients with hematoma (68.2%) and 13 with bleeding (76.5%) and in 20 of the control (74.1%) (p = 0.830). The mean HbF% of each group was 0.054, 0.012, and 0.042 for hematoma, bleeding, and control, respectively, and differences were not significant (p = 0.141). There was a moderate negative correlation between the volume of hematoma and HbF% (rSpearman = -0.527; p = 0.012). CONCLUSIONS: The fetal-maternal hemorrhage expressed by Hbf% in first-trimester pregnancies did not seem to differ between patients with and without ultrasound findings of IUH.

Risk factors for gastroschisis: A case-control study in a Brazilian population.

OBJECTIVE: To evaluate risk factors associated with fetal gastroschisis. METHODS: As a secondary aim of a larger case-control study, pregnant women attending the Fetal Medicine Unit at the Department of Obstetrics and Gynecology at Hospital das Clinicas, Sao Paulo University Medical School between July 1, 2013, and July 31, 2015, were allocated into either the gastroschisis group, where the woman was carrying a fetus with gastroschisis, or the control group, where the fetus was normal. Patients in the control group were matched at study entry for maternal age, preconception body mass index and weeks of gestation. In-person interviews were conducted during pregnancy to obtain data on demographic, medical, and social characteristics; exposure to substances; pregnancy history; the presence of chronic disease, urinary tract infections (UTIs), influenza, and fever; and the occurrence of stress events between the month before the last menstrual period and the first trimester of pregnancy. RESULTS: Of 171 women included in the study, 57 were allocated to the gastroschisis group and 114 to the control group. There were significant associations between gastroschisis and maternal UTI (P=0.011), tobacco use (P=0.001), alcohol consumption (P≤0.001), and illicit drug use (P=0.012). After analysis by standard logistic regression, the remaining significant factors were UTI, tobacco use, and alcohol consumption. CONCLUSION: UTI and exposure to tobacco or alcohol just before conception and during early pregnancy were associated with an increase in the likelihood of fetal gastroschisis.

Expression of dNK cells and their cytokines in twin pregnancies with preeclampsia.

OBJECTIVES: To assess the expression of decidual natural killer (dNK) cells and their cytokines in twin pregnancies with preeclampsia. METHODS: This was a prospective case-control study. The inclusion criteria were diamniotic (monochorionic or dichorionic) twin pregnancies in the third trimester with negative serological results for infectious diseases; absence of major fetal abnormalities or twin-twin transfusion syndrome; and no history of administration of corticosteroids in this pregnancy. The control group (CG) included uncomplicated twin pregnancies, and the preeclampsia group (PEG) included twin gestations with clinical and laboratory confirmation of the disease according to well-established criteria. Samples of the decidua were obtained and analyzed by immunohistochemistry for the expression of dNK cells and interleukins (ILs) 10, 12 and 15. In addition, maternal serum samples were collected to determine the levels of these interleukins. RESULTS: Thirty twin pregnancies were selected: 20 in the control group (CG) and 10 in the preeclampsia group (PEG). The PEG showed strong placental immunostaining for IL-15 (p=0.001) and high maternal serum levels of IL-10 (22.7 vs. 11.9 pg/mL, p=0.024) and IL-15 (15.9 vs. 7.4 pg/mL, p=0.024). CONCLUSION: A higher maternal serum concentration of both pro- and anti-inflammatory factors was observed in the twin pregnancies in the PEG. However, no difference in placental expression of IL-10 was found between the groups. These findings may suggest that maternal attempts to balance these interleukins were not sufficient to cause a placental response, and this failure may contribute to the development of preeclampsia.

Maternal nutrient intake and fetal gastroschisis: A case-control study.

Fetal gastroschisis is a paraumbilical abdominal wall defect with herniation of the abdominal organs. This multifactorial malformation occurs in young pregnant women, and the underlying cause of the disease remains unknown; however, nutritional factors may play a role in its development. This case-control study explored the association of maternal nutrient intake with the occurrence of gastroschisis. The gastroschisis group (GG) comprised 57 pregnant women with fetuses with gastroschisis, and the control group (CG) comprised 114 pregnant women with normal fetuses matched for maternal age, gestational age, and preconception body mass index classification. Nutritional assessments related to the preconception period were obtained using the food consumption frequency questionnaire, and nutrient intakes were calculated using nutrition programs. The median daily calorie intake was higher (2,382.43 vs. 2,198.81; p = .041) in the GG than in the CG. The median intake of methionine (763.89 vs. 906.34; p = .036) and threonine (1,248.34 vs. 1,437.01; p = .018) was lower in the GG than in the CG. Pregnant women with fetuses with gastroschisis have a diet characterized by higher calorie intake and lower levels of essential amino acids (methionine and threonine) during the preconception period than pregnant women with normal fetuses.

The effect of prenatal vaginal progesterone on cervical length in nonselected twin pregnancies.

PURPOSE: The aim of this study was to investigate the influence of vaginal progesterone on cervical length (CL) in asymptomatic nonselected twin gestations. METHODS: This was a secondary analysis of a randomized, double-blind, placebo-controlled trial of twin pregnancies exposed to vaginal progesterone or placebo. The CL was examined at six different time periods: 18-21+6 weeks (T1), 21-23+6 weeks (T2), 24-26+6 weeks (T3), 27-29+6 weeks (T4), 30-32+6 weeks (T5) and 33-34+6 weeks (T6). The rate of cervical shortening per week and the percent cervical shortening were compared between the groups, with analyses of the entire cohort and of those who delivered spontaneously according to gestational age at birth. RESULTS: The final analysis included 184 women in the progesterone group and 188 women in the placebo group. The baseline characteristics were similar in both groups. No differences in cervical shortening in terms of absolute value or percent shortening were observed between the groups at each time period or throughout gestation. Furthermore, no difference was found in cervical shortening for those who delivered spontaneously. CONCLUSION: Cervical shortening in asymptomatic nonselected twin pregnancies occurred at a similar rate, regardless of vaginal progesterone treatment.

Expression of dNK cells and their cytokines in twin pregnancies with preeclampsia

OBJECTIVES: To assess the expression of decidual natural killer (dNK) cells and their cytokines in twin pregnancies with preeclampsia. METHODS: This was a prospective case-control study. The inclusion criteria were diamniotic (monochorionic or dichorionic) twin pregnancies in the third trimester with negative serological results for infectious diseases; absence of major fetal abnormalities or twin-twin transfusion syndrome; and no history of administration of corticosteroids in this pregnancy. The control group (CG) included uncomplicated twin pregnancies, and the preeclampsia group (PEG) included twin gestations with clinical and laboratory confirmation of the disease according to well-established criteria. Samples of the decidua were obtained and analyzed by immunohistochemistry for the expression of dNK cells and interleukins (ILs) 10, 12 and 15. In addition, maternal serum samples were collected to determine the levels of these interleukins. RESULTS: Thirty twin pregnancies were selected: 20 in the control group (CG) and 10 in the preeclampsia group (PEG). The PEG showed strong placental immunostaining for IL-15 (p=0.001) and high maternal serum levels of IL-10 (22.7 vs. 11.9 pg/mL, p=0.024) and IL-15 (15.9 vs. 7.4 pg/mL, p=0.024). CONCLUSION: A higher maternal serum concentration of both pro- and anti-inflammatory factors was observed in the twin pregnancies in the PEG. However, no difference in placental expression of IL-10 was found between the groups. These findings may suggest that maternal attempts to balance these interleukins were not sufficient to cause a placental response, and this failure may contribute to the development of preeclampsia.

Low serum fatty acid levels in pregnancies with fetal gastroschisis: A prospective study.

Fetal gastroschisis is a paraumbilical abdominal wall defect with herniation of abdominal organs. The underlying cause of the disease remains unknown; however, studies suggest that nutritional factors may play a role in its development. This prospective case-control study explored the association of serum fatty acid levels of pregnant women and occurrence of gastroschisis. Gastroschisis group comprised 57 pregnant women with fetuses with gastroschisis, and the control group comprised 114 pregnant women with normal fetuses. Serum fatty acids levels were compared between the groups for the overall pregnancy at <34 weeks; ≤25 weeks, and >25 and <34 weeks; and at delivery. Total fatty acids (p = .008), unsaturated fatty acids (p = .002), and the C18:1n9/C18:00 ratio (p = .021) were lower in the gastroschisis group than in the control group during the overall pregnancy; however, the C16:00/C18:2n6 ratio (p = .018) was higher in the gastroschisis group than in the control group during the same period. Total fatty acids (p = .044) and unsaturated fatty acids (p = .024) were lower in the gastroschisis group at ≤25 weeks, and unsaturated fatty acid (p = .025) and the C18:1n9/C18:00 ratio (p = .013) were lower in the gastroschisis group than in the control group at >25 and <34 weeks. Pregnant women with fetuses with gastroschisis have low serum fatty acids levels during pregnancy. These findings suggest that fatty acids levels may be involved in the pathogenesis of fetal gastroschisis.

Sonographic Markers in the Prediction of Fetal Complex Gastroschisis.

OBJECTIVE: To investigate the ultrasound (US) markers predictive of complex gastroschisis (CG), mortality, and morbidity in fetuses with gastroschisis. MATERIALS AND METHODS: This was a retrospective cohort study of 186 pregnancies with isolated fetal gastroschisis. Eight US markers were analyzed. The predictions and associations of US markers with CG, mortality, and morbidity were assessed. Combinations of US markers predictive of CG were investigated. RESULTS: Extra-abdominal bowel dilatation (EABD), intra-abdominal bowel dilatation (IABD), and polyhydramnios were predictive of CG. EABD between 25 and 28 weeks had a sensitivity of 64%, a specificity of 89%, a positive predictive value (PPV) of 56.2%, and negative predictive value (NPV) of 91.8%. The predictions of IABD were sensitivity = 26.7%, specificity = 96.7%, PPV = 61.5%, and NPV = 86.8%. The odds ratios for CG in the presence of 1 and 2 US markers, compared with the absence of a US marker, were 18.3 (95% CI, 3.83-87.64) and 73.3 (95% CI, 6.14-876), respectively. CONCLUSION: US markers predictive of CG were established. The combination of these markers increases the probability of CG.

Effect of vaginal progesterone in tocolytic therapy during preterm labor in twin pregnancies: Secondary analysis of a placebo-controlled randomized trial.

AIM: Our aim was to investigate the effect of the prophylactic use of vaginal progesterone on the latency period from the initiation of tocolytic therapy to delivery in twin pregnancies with preterm labor. METHODS: This study was a secondary analysis of a randomized, double-blind, placebo-controlled trial of twin pregnancies in mothers who were exposed to a 200 mg vaginal progesterone ovule or a placebo ovule daily from 18 to 34 weeks gestation. Patients who were administered tocolysis with Atosiban because of preterm labor were included. The latency from tocolysis to delivery, mean gestational age at delivery and the rates of delivery within 48 h and within seven days were compared between progesterone and placebo groups. RESULTS: The analysis included 27 women in the progesterone group and 30 in the placebo group. The baseline characteristics were similar between the groups. Overall, there were no differences in the latency period to delivery (17.54 ± 13.54 days and 21.58 ± 13.52 days; P = 0.289), rates of delivery within 48 h (14.8% and 6.7%; P = 0.40) or within seven days (29.64% and 23.3%; P = 0.76) or mean gestational age at delivery (32.53 ± 3.33 and 34.13 ± 2.87; P = 0.08) between the progesterone and placebo groups, respectively. CONCLUSIONS: Prophylactic use of 200 mg of vaginal progesterone does not influence the latency to delivery in women with twin pregnancies treated with tocolysis because of preterm labor.

The effect of prenatally administered vaginal progesterone on uterine artery Doppler in asymptomatic twin pregnancies.

OBJECTIVES: This study investigated the influence of vaginal progesterone on uterine circulation in asymptomatic twin gestations. STUDY DESIGN: This study was a secondary analysis of a randomized, double-blind, placebo-controlled trial of twin pregnancies exposed to vaginal progesterone or placebo. We included all trial participants who had undergone uterine artery pulsatility index evaluation at the time of randomization. During each ultrasound examination, the uterine artery pulsatility index was evaluated transabdominally. The mean uterine artery pulsatility index between the progesterone and placebo groups were compared for each gestational age, starting between 18 to 34 weeks and 6days and were analyzed at three (Time 1), six (Time 2) and nine (Time 3) weeks after randomization. RESULTS: The final analysis included 128 women in the progesterone group and 122 women in the placebo group. The baseline characteristics were similar in both groups. No difference in the mean uterine artery pulsatility index was observed between the progesterone and placebo groups at each week of gestation or throughout gestation. CONCLUSIONS: In twin pregnancies, the use of vaginal progesterone in the second half of pregnancy does not influence uterine circulation.

Prenatal administration of vaginal progesterone and frequency of uterine contractions in asymptomatic twin pregnancies.

INTRODUCTION: A previous study indicated that progesterone reduces the mean uterine contraction frequency in singleton pregnancy at high risk for preterm birth. The aim of this study was to investigate the effect of vaginal progesterone on the frequency of uterine contractions in twin pregnancies. MATERIAL AND METHODS: This was a secondary analysis of a randomized, double-blind, placebo-controlled trial of twin pregnancies exposed to vaginal progesterone or placebo. Naturally conceived twin pregnancies with no prior history of preterm delivery, asymptomatic regarding preterm labor, who had undergone uterine contraction frequency monitoring from 24 to 34 weeks and 6 days were included in the study. Comparison of the mean frequency of uterine contractions between the treatment groups was performed. We also examined the influence of cervical length and chorionicity on the mean frequency of uterine contractions according to the group. RESULTS: The final analysis included 166 women in the progesterone and 170 in the placebo group. The baseline characteristics were similar in the two groups. Overall, no difference in the mean frequency of uterine contractions (p = 0.91) was observed between the progesterone (2.54 ± 3.19) and placebo (2.56 ± 3.59) groups. Also, no difference in the mean frequency of uterine contractions was observed between the groups in each week between 24 and 34 weeks and 6 days of gestation. Cervical length and chorionicity did not influence the frequency of contractions according to the progesterone or placebo treatment. CONCLUSIONS: Overall, progesterone does not influence the frequency of uterine contractions in twin pregnancies.

Prediction of perinatal mortality in triplet pregnancies.

PURPOSE: To investigate predictors of perinatal mortality in triplet pregnancies. METHODS: Retrospective cohort managed in a tertiary teaching hospital (1998-2012) including all pregnancies with tree live fetuses at the first ultrasound examination, performed after 11 weeks of gestation. Primary end-point was defined as the number of children alive at hospital discharge. Ordinal stepwise regression analysis examined the association with maternal age, parity, pregnancy chorionicity, gestational age at our first ultrasound evaluation, presence of maternal clinical, obstetrical and fetal complications and gestational age at delivery. RESULTS: Sixty-seven triplet pregnancies were first seen at 18.5 ± 6.8 weeks, 33 (49.3 %) were trichorionic, obstetric complications occurred in 34 (50.7 %) and fetal complications were diagnosed in 17 (25.4 %). Perinatal mortality rate was 249 ‰ (95 % CI 189-317) and 138 (73 %) children were discharged alive from hospital (11 pregnancies with no survivors; single and double survival in ten cases each; all children alive in 36). Regression analysis showed that presence of fetal complications (OR 0.10, 95 % CI 0.03-0.36) and gestational age at delivery (OR 1.55, 95 % CI 1.31-1.85) are significant predictors of outcome (p < 0.001). CONCLUSIONS: Perinatal mortality in non-selected triplet pregnancies is high and is related to the presence of fetal complications and gestational age at delivery.

Prediction of intrauterine death and severe preterm delivery in twin pregnancies discordant for major fetal abnormality.

OBJECTIVE: To investigate predictors of spontaneous fetal death and preterm delivery in twin pregnancies with one fetus affected by a major structural malformation. STUDY DESIGN: Retrospective study (1999-2012) conducted at a tertiary teaching hospital involving 51 twin pregnancies (dichorionic=31, monochorionic diamniotic=15, monochorionic monoamniotic=4, not established=1) with a major fetal abnormality, enrolled before 26 weeks and managed expectantly. PRIMARY OUTCOMES: spontaneous fetal death, and/or delivery before 32 weeks. Prediction was examined with stepwise logistic regression analysis, and independent variables included: maternal age, gestational age at diagnosis, chorionicity, fetal gender, number and type of fetal abnormalities. Significance level was set at 0.15. RESULTS: Fetal abnormalities were diagnosed at a mean gestation of 21.5±3.7 weeks: cardiac abnormalities were observed in 31.4% of abnormal fetuses, abdominal wall defects in 29.4%, central nervous system 21.5%, spine 17.6%, effusions 17.6%, noncardiac thoracic abnormalities 15.7%, genital and urinary system 13.7%, limbs and soft tissue 3.9%, intestinal 1.9% and facial defects 1.9%. Fetal death occurred in 15 (29.4%) abnormal fetuses and was significantly correlated with the number of fetal malformations (p=0.02, OR=2.54, 95% CI=1.14-5.62), presence of effusion/hydrops (p=0.06, OR=4.7, 95% CI=0.95-24) and monochorionic placenta (p=0.11, OR=2.8, 95% CI=0.78-9.8). Normal co-twin fetal death occurred in four cases (7.8%) and was related to monochorionic pregnancies (p=0.14, OR=5.8, 95% CI=0.56-61). Delivery before 32 weeks was observed in 14 (27.5%) pregnancies and was related to presence of effusion/hydrops (p=0.04, OR=5.5, 95% CI=1.07-28). CONCLUSION: Spontaneous fetal death and/or delivery before 32 weeks in twin pregnancies with one fetus affected by a major structural malformation are related to the number of abnormalities diagnosed and presence of fetal effusion or hydrops.

Should we measure fetal omphalocele diameter for prediction of perinatal outcome?

OBJECTIVE: To evaluate the fetal omphalocele diameter/abdominal circumference ratio (OD/AC) as a predictor of adverse perinatal outcome. METHODS: Analysis involving 47 singleton pregnancies with fetal omphalocele, normal karyotype and absence of other major abnormalities. The OD/AC ratio was determined antenatally by ultrasound and the best cutoff for the prediction of neonatal death was determined by receiver operating characteristic curve analysis. Additional secondary outcomes included need for oral intubation in the first 24 h of life, two-step surgery or use of synthetic mesh, reoperation, parenteral feeding and need for respiratory assistance >21 days, time to first oral feed, and time to hospital discharge. RESULTS: Fetal OD/AC did not change significantly with gestational age. Postnatal death occurred in 10 (21.3%) cases and the best cutoff for prediction was an OD/AC ratio ≥0.26. In pregnancies with the first ultrasound evaluation performed before 31 weeks' gestation and an OD/AC ≥0.26, the likelihood ratio for needing intubation in the first 24 h of life was 2.6 (95% CI: 1.2-5.7), needing two-step surgery or use of mesh was 4.9 (95% CI: 1.9-14.4), and postnatal death was 4 (95% CI: 1.9-7.5). CONCLUSION: A fetal ultrasound OD/AC ratio ≥0.26 is associated with increased postnatal morbidity and mortality.

Twenty-five years of fetal echocardiography in conjoined twins: lessons learned.

BACKGROUND: The aim of this study was to determine the accuracy of prenatal echocardiography in the diagnosis of intracardiac malformations and the degree of cardiac fusion in conjoined twins presenting to a single center over a 25-year period. METHODS: The study group included 53 sets of conjoined twins from 1987 to 2012, including 38 thoracopagus, six parapagus, six omphalo-ischiopagus, two omphalopagus, and one cephalopagus. Twins were classified according to the degree of cardiac fusion: separate hearts and pericardium (group A, n = 10), separated hearts and common pericardium (group B, n = 2), fused atria and separated ventricles (group C, n = 2), and fused atria and ventricles (group D, n = 39). Postmortem examination was possible in 68 individual cases (98 deaths [69.3%]). RESULTS: Cardiac defects were diagnosed in 47 sets of twins (88.6%). In 10 (18.8%), only one fetus was affected, and in 37 (69.8%), both fetuses were affected (n = 84/106 [79.2%]). There was a high predominance of right-sided lesions (63.0% [53 fetuses in 84 affected]) including pulmonary atresia or stenosis (35.7%), tricuspid atresia (11.9%), and hypoplastic or small right ventricle (21.4%). Autopsy findings added information to fetal echocardiographic findings in nine sets of twins (25.7%). Three pairs classified antenatally in groups A, B, and D were confirmed by autopsy in groups B, C, and C, respectively. CONCLUSIONS: This study demonstrates that specialized fetal echocardiography is not a perfect diagnostic tool but is sensitive enough to establish prognosis in the counseling process. Because of complexity, such evaluations should be performed only at tertiary centers by specialists who are familiar with the peculiarities of this rare malformation. The predominance of right-sided lesions is not only an interesting finding, but this information has essential importance in terms of shortening examination times, allowing a more focused analysis of the fetal heart.

Changes in amniotic fluid index after maternal oral hydration in pregnancies with fetal gastroschisis: initial observations.

OBJECTIVE: To evaluate the effect of maternal oral hydration on amniotic fluid index (AFI) in pregnancies with fetal gastroschisis. METHODS: AFI was evaluated at 24-hour intervals, during 4 consecutive days, under a continuous maternal oral water hydration regimen, in singleton pregnancies with isolated fetal gastroschisis. RESULTS: Nine pregnancies were examined at a mean gestational age of 31.6 weeks (+/-1.4) and mean maternal daily oral water intake was 3,437 (+/-810) ml. Mean AFI on days 0-3 were 13.2 (+/-2.9), 14.8 (+/-3.3), 14.5 (+/-3.1) and 14.8 (+/-2.6), respectively. AFI on day 0 was significantly lower compared to all the other 3 days (p = 0.01 and 0.02). Significant correlation was found in relative difference in AFI between day 0 and day 1 and gestational age (r = -0.67, p = 0.05) and the amount of water intake in the previous 24 h (r = 0.76, p = 0.02). CONCLUSION: Maternal oral water hydration significantly increases AFI in pregnancies with isolated fetal gastroschisis.

Placental lakes on sonographic examination: correlation with obstetric outcome and pathologic findings.

PURPOSE: The aim of this study was to compare pregnancy outcome and pathologic findings in cases of placental lake visualized on sonography in pregnancies without concurrent fetal or maternal disease. METHODS: Placental aspect was evaluated sonographically in 4,106 singleton pregnancies without maternal or fetal disease at between 15 and 34 weeks3 gestation. Fifty-nine cases presenting with placental lakes were followed through the end of pregnancy. Obstetric outcome was compared with that of cases with sonographically normal placental aspects (n=37). Macroscopic and microscopic examinations for pathology were also performed in a few cases in both groups. RESULTS: Placental lakes were present in 2.2% (92/4,106) of the cases. No statistically significant difference was observed in birth weight, gestational age at delivery, adverse obstetric outcome, and macroscopic or microscopic results between the control and study groups. CONCLUSION: Our findings do not seem to indicate an increase in the risk for an adverse pregnancy outcome in cases presenting with placental lakes.

A comparison between middle cerebral artery peak systolic velocity and amniotic fluid optical density at 450 nm in the prediction of fetal anemia.

OBJECTIVE: The purpose of this study was to compare fetal middle cerebral artery peak systolic velocity with amniotic fluid delta optical density at 450 nm in the prediction of fetal anemia. STUDY DESIGN: A prospective study that involved 28 singleton pregnancies that were at-risk for fetal anemia was carried out in a tertiary teaching hospital. Middle cerebral artery peak systolic velocity was measured immediately before the determination of deltaoptical density at 450 nm and fetal hemoglobin concentration. Sensitivities and predictive values for fetal anemia were examined. RESULTS: Fetal hemoglobin concentrations correlated significantly with middle cerebral artery peak systolic velocity (correlation coefficient, -0.77; P <.0001) and deltaoptical density at 450 nm zones (correlation coefficient, -0.56; P =.0025). Middle cerebral artery peak systolic velocity was >1.5 multiples of the median in 15 of 28 cases (54%); for this cutoff value, the sensitivity and positive-predictive values for a hemoglobin deficit of >-3SD were 75% and 60% and for a hemoglobin deficit of >-5SD were 100% and 47%, respectively. The corresponding values for deltaoptical density at 450-nm zone III (6/28 cases, 21%) were 0% (hemoglobin deficit, <-3SD) and 86% and 100% (hemoglobin deficit, <-5SD). CONCLUSION: Middle cerebral artery peak systolic velocity and amniotic fluid optical density at 450 nm are both useful in the prediction of fetal anemia. However, Doppler examination has the advantage of being a noninvasive method that can help reduce the number of invasive procedures in pregnancies that are at-risk for fetal anemia.