Changes in human placental oxygenation during maternal hyperoxia estimated by blood oxygen level-dependent magnetic resonance imaging (BOLD MRI).

OBJECTIVES: To investigate changes in human placental oxygenation during maternal hyperoxia using non-invasive blood oxygen level-dependent (BOLD) magnetic resonance imaging (MRI). METHODS: Eight healthy pregnant women with uncomplicated singleton pregnancies at gestational weeks 28-36 were examined with BOLD MRI, over two consecutive 5-min periods of different oxygenation: first normoxia (21% O2 ) and then hyperoxia (12 L O2 /min), achieved by controlling the maternal oxygen supply with a non-rebreather facial mask. Selecting three slices showing cross-sections of the central part of the placenta, we investigated total placental oxygenation by drawing regions of interest (ROIs) covering the entire placenta, and regional placental oxygenation by drawing smaller ROIs in the darker and brighter areas of the placenta. For each ROI, the difference in BOLD signal between the two episodes was determined and the percentage increase in BOLD signal during hyperoxia (ΔBOLD) was calculated. RESULTS: In the BOLD image, the normoxic placenta appeared heterogeneous, with darker areas located to the fetal side and brighter areas to the maternal side. During hyperoxia, the placenta became brighter and the structure more homogeneous, and the BOLD signal of the total placenta increased (ΔBOLDtot , 15.2 ± 3.2% (mean ± SD), P < 0.0001). The increase was seen predominantly in the dark areas in the fetal part of the placenta (ΔBOLDfet , 32.1 ± 9.3%) compared with in the bright areas in the maternal part of the placenta (ΔBOLDmat , 5.4 ± 3.5%). CONCLUSION: During hyperoxia, placental oxygenation was increased predominantly in the darker placental areas, which, given their anatomical location, represent the fetal circulation of the placenta. To our knowledge, this is the first study to successfully visualize changes in placental oxygenation using BOLD MRI.

Late termination of pregnancy: a comparison of obstetricians' experience in eight European countries.

OBJECTIVE: To compare the experience and attitude of obstetricians in Europe towards late termination of pregnancy and the factors affecting their responses. DESIGN: Cluster sampling cross-sectional survey. All neonatal intensive care unit (NICU)-associated maternity units were recruited (census sampling) in Luxembourg, the Netherlands and Sweden. In France, Germany, Italy, Spain and the UK, units were selected at random. In every recruited unit, all obstetricians with at least 6 months' experience were invited to participate. SETTING: NICU-associated maternity units in eight European countries. POPULATION: Obstetricians with at least 6 months' clinical experience. METHODS: An anonymous, self-administered questionnaire was used. Multinomial logistic analysis was used to identify factors predicting the obstetricians' views about modifying the law governing late termination in their country. MAIN OUTCOME MEASURE: Obstetricians' experience of late termination of pregnancy and views about national policies. RESULTS: One hundred and five units and 1530 obstetricians participated (response rates 70 and 77% respectively). The most common indications for late termination were congenital anomalies and women's physical health. Feticide was not common except in France, Luxembourg and the UK. Active euthanasia of a liveborn was practiced in France and the Netherlands. Obstetricians in Germany were more likely to feel that late termination should be more severely restricted, the opposite was true in Spain and the Netherlands. In Italy, there was dissatisfaction with current status, but opinion was divided, reflecting views on both sides of the debate. CONCLUSIONS: This research outlines current practice in a difficult and sensitive area and suggests the need for more discussion and support for all those who were involved.

Transvaginal ultrasound and lactate dehydrogenase isoenzyme activity profile in uterine aspirate for diagnosis of endometrial carcinoma in women with postmenopausal bleeding.

Postmenopausal patients with vaginal bleeding (n = 72) were evaluated with the combination of transvaginal ultrasonography (TVS) and analysis of the lactate dehydrogenase (LD) isoenzyme activity profile in uterine fluid aspirates. TVS evaluation of the endometrium was classified as <5 mm, > or =5 mm, or poorly defined. The LD isoenzyme activity profile was characterized as abnormal or normal. Pathologic findings were further evaluated with diagnostic curettage. TVS found the endometrium to be > or =5 mm or poorly defined in 44 patients (61%). Endometrial carcinoma was found in 6 of 72 patients (8%). They appeared in the TVS groups endometrium >5 mm (n = 2) and endometrium poorly defined (n = 4) but not in the endometrium <5 mm. The LD isoenzyme activity profile was abnormal in the six malignant cases and in ten benign cases. Thus, the need for further evaluation with hysteroscopy and curettage was reduced to 16 cases. Since TVS had 100% sensitivity but only 42% specificity, it is suitable for first-level examination in patients with postmenopausal bleeding. The second-level method should have similarly high sensitivity but much higher specificity. The LD isoenzyme activity profile in uterine fluid aspirates had 100% sensitivity and 85% specificity. Another important feature is that the method is not sensitive to endometrial thickness, amount of sample, sampling device, or dilution. Thus, it is more reliable than aspiration histology. For every hysteroscopy or curettage that can be replaced by LD analysis, the cost is reduced by approximately EUR 720 or 540, respectively.

Validation of fractional moving blood volume measurement with power Doppler ultrasound in an experimental sheep model.

OBJECTIVE: To compare fractional moving blood volume (FMBV) estimation using power Doppler ultrasound (PDU) with blood flow estimation using radioactive microspheres (RMS) for evaluation of fetal organ blood perfusion. METHODS: Blood flow was measured in the adrenal gland of nine exteriorized fetal lambs. Five fetal lambs underwent total umbilical cord occlusion in order to induce changes in the adrenal blood flow (asphyxia group). Four lambs were used as sham controls (control group). Three RMS injections, with coincident PDU recordings of the adrenal gland, were performed in each lamb. In the asphyxia group, measurements were taken before the cord occlusion, 5 min later and when the mean blood pressure decreased below 25 mmHg. In the control group, the measurements were done with an interval of 5 min. FMBV normalized for attenuation of PDU signals, and mean pixel intensity (MPI) were estimated offline. After completion of the study, adrenal blood perfusion was calculated according to the reference sample microsphere technique, using the isotope activity and expressed in mL/min/100 g. The correlation between RMS and FMBV and MPI, respectively, was analyzed individually for each lamb. RESULTS: In the asphyxia group, all lambs showed a marked reduction in the adrenal blood perfusion towards the third RMS injection. In the control group, the adrenal perfusion showed small variations throughout the experiment. In the total material, there was a higher correlation between FMBV and RMS (median, r = 0.90; range, 0.43-0.99) than between MPI and RMS (median, r = 0.55; range, -0.53 to 0.99). CONCLUSION: The FMBV method of quantifying PDU signals correlates highly with blood flow perfusion estimation using RMS in the fetal lamb adrenal gland.

Fractional moving blood volume estimation in the fetal lung using power Doppler ultrasound: a reproducibility study.

OBJECTIVE: To evaluate the reproducibility of fractional moving blood volume (FMBV) estimation in the fetal lung using power Doppler ultrasound (PDU). METHODS: The lung blood perfusion of 20 normally grown singleton fetuses at 32-35 weeks of gestation was evaluated by two experienced observers using PDU. Each observer recorded two consecutive sequences of images from the posterior part of the fetal lung and calculated FMBV offline. FMBV expresses the percentage of blood movement within a defined region of interest (ROI). Repeatability and agreement were evaluated by means of the intraclass (intraCC) and interclass (interCC) correlation coefficients. RESULTS: FMBV was successfully evaluated in 17/20 fetuses by both observers (kappa index 0.82; 95% CI 0.51-0.93). The intraCC for repeatability for Observer A was 0.92 (95% CI 0.78-0.96), and for Observer B 0.90 (95% CI 0.74-0.96). The mean difference between the first and the second measurement was 0.7% (SD 4.5%). The interCC for repeatability over time and between the operators was 0.70 (95% CI 0.56-0.76) and the mean difference between the observers was 0.6% (SD 4.65%). The interCC for agreement was 0.92 (95% CI 0.84-0.95) and the mean difference in FMBV results when both observers analyzed the same sequences offline was 0.6% (SD 3.85%). CONCLUSION: In the hands of experienced operators, using a well-defined ROI and standard settings, FMBV estimation is a reproducible method of quantifying power Doppler signals recorded from fetal lung blood perfusion.

Uterine artery score and perinatal outcome.

OBJECTIVE: To evaluate a modified uterine artery score based on the pulsatility index and presence or absence of notching in the Doppler velocity waveform recorded from both uterine arteries in relation to the perinatal outcome. METHODS: A retrospective analysis was performed in 741 third-trimester high-risk pregnancies. The uterine artery score was constructed assigning one point to each abnormal parameter-high pulsatility index and presence of notch-thus ranging from 0 (normal findings in both uterine arteries) to 4 (notch and high pulsatility index in both uterine arteries). In a subgroup with lateral placenta (n = 359), two definitions of abnormal pulsatility index were compared. In the uniform uterine artery score, a pulsatility index > 1.20 in both uterine arteries was considered abnormal, disregarding the placental location; in the subgroup with lateral placenta, the high pulsatility index was defined as > 1.00 on the placental side and > 1.40 on the non-placental side. RESULTS: Receiver-operating characteristic curves did not reveal any difference in the diagnostic capacity between the group with a uniform uterine artery score and the subgroup with lateral placenta (P = 0.54). In the total material, the odds ratios and linear regression analysis showed an increased risk for an adverse perinatal outcome with increasing uterine artery score (P < 0.01). At a uterine artery score > 2, there was a significantly increased risk for operative delivery for fetal distress, neonatal intensive care unit admission, 5-min Apgar score < 7, preterm delivery and delivery of a small-for-gestational age fetus. CONCLUSION: In high-risk third-trimester pregnancies, Doppler velocity waveforms of the uterine arteries can be evaluated using the uterine artery score disregarding the placental location. The uterine artery score possesses a high predictive value regarding adverse perinatal outcome.

Reduced postnatal cerebral glucose metabolism measured by PET after asphyxia in near term fetal lambs.

The effects of fetal asphyxia on cerebral function and development, involve the transition from fetal to neonatal life. Changes in cerebral glucose metabolism may be an early postnatal indicator of fetal asphyxia. The objective is to develop an experimental lamb model involving the transition from fetal to neonatal life and to examine the effect of fetal asphyxia with cerebral hypoxic ischemia on early postnatal cerebral glucose metabolism. Fetal asphyxia was induced by total umbilical cord occlusion in eight near-term fetal lambs (134-138 days) with the ewe under isoflurane-opiate anesthesia. The mean occlusion time until cardiac arrest was 14.5 (4.2) min (SD). Lambs were immediately delivered and standardized resuscitation was instituted after 2 min asystole. At 4 hr postnatal age, [18-F]Fluoro-2-deoxy-glucose (18-FDG) was injected intravenously in eight asphyxiated lambs and in eight controls. Cerebral glucose metabolism was examined by positron emission tomography (PET). As a result the mean arterial blood pressure, acid-base values, blood glucose and serum lactate at 4 hr postnatal age did not differ significantly between lambs subjected to umbilical cord occlusion and controls. EEG was abnormal in all lambs subjected to cord occlusion and normal in the controls at 4 hr postnatal age. Global cerebral metabolic rate (CMRgl) as determined by PET was significantly lower in lambs subjected to cord occlusion mean/median (SD) 22.2/19.6 (8.4) micromol/min/100 g) than in controls mean/median (SD) 37.8/35.9 (6.1); P < 0.01). Global CMRgl is significantly reduced in newborn lambs 4 hr after fetal asphyxia induced by umbilical cord occlusion. A reduction in CMRgl is an early indicator of global hypoxic cerebral ischemia.

A critical appraisal of the use of umbilical artery Doppler ultrasound in high-risk pregnancies: use of meta-analyses in evidence-based obstetrics.

OBJECTIVES: To reanalyze randomized controlled trials on the use of umbilical artery Doppler velocimetry in high-risk pregnancies and determine which high-risk pregnancies benefit from the use of Doppler velocimetry. METHODS: Searching Medline, the Cochrane Library and Embase we found 13 randomized controlled trials on the use of Doppler velocimetry in high-risk pregnancies. Of these, six included pregnancies with strictly defined suspected intrauterine growth restriction and/or hypertensive disease of pregnancy ('well-defined studies'); the rest included a great variety of high-risk pregnancies ('general risk studies'). The studies were analyzed with particular regard to the heterogeneity and to outcome. Audits of the perinatal deaths reported in the randomized controlled trials were performed by a panel of 32 international experts. RESULTS: The 'well-defined studies' had a more uniform study design as compared to the 'general risk studies' and they showed a significant reduction in antenatal admissions (odds ratio, 0.56; 95% confidence interval, 0.43-0.72), inductions of labor (0.78; 0.63-0.96), elective deliveries (inductions of labor and elective Cesarean sections) (0.73; 0.61-0.88) and Cesarean sections (0.78; 0.65-0.94). By perinatal audit it was found that more perinatal deaths in the 'well-defined studies' were potentially avoidable by use of Doppler velocimetry ( P < 0.0005) and the rate of avoidable perinatal deaths was higher among controls (50%) than cases (20%) in this group. CONCLUSION: The randomized controlled trials on umbilical artery Doppler velocimetry show major differences regarding study design and technical and clinical issues and, therefore, they should not be pooled in a simple meta-analysis. By stratification it was found that only in pregnancies with suspected intrauterine growth restriction and/or hypertensive disease of pregnancy will the use of umbilical artery Doppler velocimetry reduce the number of perinatal deaths and unnecessary obstetric interventions.

Hyrtl's anastomosis, the only connection between the two umbilical arteries. A study in full term placentas from AGA infants with normal umbilical artery blood flow.

BACKGROUND: The aim of this study was to describe the variable anatomy in the anastomosis between the umbilical arteries for better understanding of the physical characteristics of the umbilical flow velocity waveform (FVW). METHODS: The arterial system of 67 placentas from pregnancies with normal umbilical FVW and resulting in a full-term AGA infant was visualized by angiography. The method allowed study of the anatomy of the anastomosis between the umbilical arteries and calculation of the relative placental area supplied by each umbilical artery. RESULTS: In 60 cases there was one anastomosis between the umbilical arteries, in one case there were two, in four the anastomosis was absent, and another two cases had a single umbilical artery. The anastomosis was represented by a vessel, a fenestration or coalescence of the umbilical arteries. In case the anastomosis diameter was of at least that of the umbilical arteries, they supplied in mean 26% and 74% (+/- 8.2%) of the placental area respectively. When the anastomosis diameter was smaller than that of the umbilical arteries their supply areas were in mean 41% and 59% (+/- 6.0%) respectively. In placentas lacking anastomosis the two umbilical arteries supplied 45% and 55% (+/- 2.6%) respectively, indicating a higher degree of symmetry. CONCLUSIONS: By using angiography we were able to demonstrate the variable anatomy of the anastomosis of Hyrtl. We found the occurrence and width of the anastomosis was correlated to the symmetry in size between the supply areas of each umbilical artery.

Three methods for hysterectomy: a randomised, prospective study of short term outcome.

OBJECTIVE: To detect differences in clinical short term outcome between total abdominal hysterectomy, vaginal hysterectomy and laparoscopic assisted vaginal hysterectomy. DESIGN: Randomised controlled trial. SETTING: Department of Obstetrics and Gynaecology, Hospital of Helsingborg, Sweden. SAMPLE: One hundred-twenty women scheduled for hysterectomy for various indications. METHODS: Randomisation into three treatment arms: total abdominal hysterectomy (n = 40); vaginal hysterectomy (n = 40) and laparoscopic assisted vaginal hysterectomy (n = 40). During traditional abdominal and vaginal surgery, laparoscopic assistance was kept to a minimum. Substantial number of cases needed volume-reducing manoeuvres due to uterine size. MAIN OUTCOME MEASURES: Duration of surgery, anaesthesia, time in hospital and recovery time. RESULTS: Mean duration (range) of surgery was significantly longer for laparoscopic assisted vaginal hysterectomy compared with vaginal hysterectomy and total abdominal hysterectomy, 102 min (50-175), 81 min (35-135) and 68 min (28-125), respectively. Mean stay in hospital and mean time to recovery was significantly longer for total abdominal hysterectomy compared with vaginal hysterectomy and laparoscopic assisted vaginal hysterectomy. The difference between vaginal hysterectomy and laparoscopic assisted vaginal hysterectomy was not significant. It was possible to remove uteri under 600 g with all three methods. Four laparoscopic assisted vaginal hysterectomies and one vaginal hysterectomy were converted to open surgery. Reoperation and blood transfusion were required after two vaginal hysterectomies and one laparoscopic assisted vaginal hysterectomy. One woman needed blood transfusion after total abdominal hysterectomy. CONCLUSIONS: Traditional vaginal hysterectomy proved to be feasible and the faster operative technique compared with vaginal hysterectomy with laparoscopic assistance. The abdominal technique was somewhat faster, but time spent in theatre was not significantly shorter. Abdominal hysterectomy required on average a longer hospital stay of one day and one additional week of convalescence compared with traditional vaginal hysterectomy. Vaginal hysterectomy should be a primary method for uterine removal.

The female pelvic floor: a dome--not a basin.

BACKGROUND: Classically and based on necropsy observations, the muscles of the pelvic floor (the levator ani and coccygeus muscles) have been described as having the shape of a basin. In a previous study by Hugosson et al. (1991), magnetic resonance imaging revealed the dome shape of the pelvic floor in live subjects when muscular tonus is present. METHODS: The normal anatomy and the dynamic movements of the female pelvis were examined with magnetic resonance imaging including conventional and fast sequences. Six nulliparous and six parous healthy women without signs or symptoms of pelvic relaxation were studied in the supine position at rest, during voluntary pelvic contractions and during bearing down. RESULTS: The levator ani muscle was dome-shaped at rest. During voluntary pelvic contractions the muscle straightened, becoming more horizontal and during bearing down it descended, becoming basin-shaped. The width of the genital hiatus was the same in both groups at rest and it widened during bearing down in all but two 3-para women. The bladder base was lifted upward and forward during voluntary pelvic contractions but during bearing down it descended. The rectum at rest had a posterior angle which decreased during voluntary pelvic contractions and increased during bearing down. CONCLUSIONS: MRI of the female pelvis is valuable in anatomic and dynamic analysis in healthy women and offers new information about the female pelvic floor. The female pelvic floor is shaped like a dome-not like a basin.

A comparative study of breast feeding after traditional postnatal hospital care and early discharge.

OBJECTIVE: To provide quality assurance for the care plan and working structure within the early discharge unit at the Women's Clinic, Central Hospital, Helsingborg, Sweden. DESIGN: Survey, using postal questionnaire. SETTING: The Women's Clinic, Central Hospital, Helsingborg, Sweden. PARTICIPANTS: 304 women with babies of six months of age, delivered at the Central Hospital, Helsingborg, between September and December 1993 and who, together with their baby, met the criteria for early discharge. Early discharge is generally defined in Sweden as discharge before 72 hours postpartum. MEASUREMENTS AND FINDINGS: Of the participants 41% chose early discharge (ED) and 59% chose traditional hospital care (THC). Four groups were studied for breast feeding frequency - THC primiparae, ED primiparae, THC multiparae and ED multiparae. Further division was made for breast feeding at 2, 4 and 6 months of age. The four main groups were examined for demographic differences. A difference was found in education level; early discharge mothers had a lower level of education than THC mothers. No significant difference was found for the frequency or duration of breast feeding between the early discharge and the traditional hospital care groups, despite the higher education level in the traditional hospital care group. IMPLICATIONS FOR PRACTICE: A possible explanation for this finding is that a care plan aimed at supporting the individual's responsibility and participation, providing relevant knowledge and a subliminal communication of trust in the competence of parent and child, is of particular significance for women with less education. The presence and participation of the baby's father at an early stage may also be a factor.

Cordocentesis.

During the last decade cordocentesis has become common procedure in perinatal medicine. Although the method represents potential possibilities, it is also afflicted with several limitations. The present review focuses on technique, indications, and complications.

Oxygenation of the human fetus as a function of hemoglobulin concentration.

We studied the relationships between fetal hemoglobin concentration, acid base status, and lactate concentration in umbilical venous and fetal heart blood in 157 fetuses affected by blood group incompatibility who had been exposed for 214 fetal blood sampling procedures (cordocentesis in 153 and intracardiac puncture in 61 cases). All blood samplings were obtained before fetal blood transfusions were administered. The results indicate that the human fetus can maintain a normal acid-base status until a 50% reduction of the hemoglobin concentration. A further reduction of hemoglobulin is associated with an accumulation of lactate both in umbilical venous and fetal heart blood. The partial oxygen tension and the oxygen saturation in umbilical venous blood remained virtually unchanged with decreasing hemoglobin concentration (r = -0.11, P = 0.21; r = 0.09, P = 0.31, respectively), whereas these parameters decreased significantly (r = 0.33, P = 0.02; r = 34, P = 0.02) in blood obtained from the fetal heart. The partial carbon dioxide tension of umbilical vein blood decreases significantly with a reduced hemoglobin concentration (r = 0.25 P = 0.008). We speculate that these alterations in acid-base status in umbilical vein and fetal heart blood reflect a circulatory transition from a high to a low cardiac output as the hemoglobin concentration decreases.

Umbilical artery Doppler flow-velocity waveforms and fetal acid-base balance in Rhesus-isoimmunized pregnancies.

Pulsed Doppler blood-flow velocity waveforms in the umbilical arteries, as well as blood gases, hematocrit, and lactate concentration in umbilical venous blood, were examined in 21 patients undergoing 49 cordocentesis, 34 of which were followed by fetal blood transfusion into the umbilical vein. The aim of the study was to evaluate the correlations, if any, between the Doppler indices from the umbilical artery (pulsatility index, resistance index, systolic/diastolic ratio) and the blood gas values (pO2, pCO2, O2 content, pH) and lactate content in the umbilical vein. The only correlation confirmed in this study was in the subgroup of anemic fetuses undergoing fetal blood transfusion, where correlation existed between A/B and the initial O2 content (r = -0.41, p < 0.02). We conclude that, in Rhesus-isoimmunized pregnancies, in contrast to other pregnancies, a close correlation does not exist between the Doppler indices in the umbilical artery and the fetal blood gas values.

Umbilical artery Doppler flow-velocity waveforms in Rhesus-isoimmunized fetuses before and after fetal blood sampling or transfusion.

Pulsed-Doppler examinations of blood-flow velocities in the umbilical artery were carried out before and after 15 diagnostic cordocenteses and 34 fetal blood transfusions into the umbilical vein. There were decreases in the systolic/diastolic ratio (A/B) (p < 0.01), the pulsatility index (PI) (p < 0.05), and the resistance index (RI) (p < 0.01) after cordocentesis but not after fetal blood transfusion. There were no correlations between the initial hematocrit and the umbilical artery Doppler indices in the sample nor in the fetal blood sampling group. In the fetal blood transfusion group, on the other hand, there was a negative correlation between the initial hematocrit and A/B (r = -0.44; p < 0.01) and the RI (r = -0.35; p < 0.05). The umbilical artery Doppler flow-velocity indices did not predict the fetal hematocrit.

Hypoxanthine in fetal umbilical venous blood and amniotic fluid from pregnancies complicated by rhesus isoimmunization.

Concentrations of hypoxanthine (HX) was determined in umbilical venous blood and amniotic fluid obtained at 74 instances in 36 rhesus immunized patients before the onset of labor. HX concentrations were related to gestational age, concentrations of hemoglobin and lactate, pH, and partial oxygen pressure in umbilical venous blood. Multiple regression analysis revealed hemoglobin concentration to be the only variable that had any explanatory power to HX in amniotic fluid. No one of the studied variables gave any significant contribution to a regression model to explain HX in umbilical venous blood. We conclude that HX levels in umbilical venous blood and in amniotic fluid from rhesus immunized patients were not associated with fetal blood gases before the onset of labor.

Erythropoietin concentrations in amniotic fluid and umbilical venous blood from Rh-immunized pregnancies.

We set out to investigate prospectively the levels of erythropoietin in amniotic fluid and umbilical venous blood, and to attempt to relate these to fetal haemoglobin and lactate concentrations and to pCO2 and PO2 in Rh immunised patients studied before the onset of labor. Fetal blood was obtained by cordocentesis, and amniotic fluid by amniocentesis from a consecutive series of 36 Rh immunized patients at the time of fetal blood sampling. There was a close correlation (tau = 0.357, P = 0.0001) between the concentrations of erythropoietin in umbilical venous blood and those in amniotic fluid. Erythropoietin in umbilical venous blood correlated inversely with hemoglobin (tau = 0.453, P = 0.0001), and directly with lactate concentrations (tau = 0.450, P = 0.0005). When all other variables were considered, multiple regression analysis demonstrated hemoglobin concentration to be the only variable to be related to the level of erythropoietin in umbilical venous blood taken before transfusion. When the same analysis was performed on the same variables, adding erythropoietin concentration in amniotic fluid as the dependent variable, only erythropoietin in umbilical venous blood was found to be related to the level of erythropoietin in amniotic fluid. We conclude that the erythropoietin concentration in umbilical venous blood from Rh-immunized patients before the onset of labor, is related to fetal anemia. We also conclude that erythropoietin concentration in amniotic fluid is related to that in fetal blood, thereby indicating that the fetus is an important source of amniotic fluid erythropoietin in non laboring patients.