Hacker & Moore's Essentials of Obstetrics and Gynecology: With STUDENT CONSULT Online Access,5th ed.

Chapter 15

Rhesus Isoimmunization

Lony C. Castro

Rhesus (Rh) isoimmunization is an immunologic disorder that occurs in a pregnant, Rh-negative patient carrying an Rh-positive fetus. The immunologic system in the mother is stimulated to produce antibodies to the Rh antigen, which then cross the placenta and destroy fetal red blood cells.

image Pathophysiology

The Rh complex is made up of a number of antigens, including C, D, E, c, e, and other variants, such as Du antigen. More than 90% of cases of Rh isoimmunization are due to antibodies to D antigens. Therefore, this chapter is mainly limited to a discussion of the D antigen, although the same principles apply to any other antigen-antibody combination. A person who lacks the D antigen on the surface of the red blood cells is called “Rh negative,” and an individual with the D antigen is considered “Rh positive.”

Among African Americans, about 8% are Rh negative, whereas among white Americans, about 15% are Rh negative. Only 1% to 2% of Asians and Native Americans are Rh negative. When Rh-negative patients are exposed to the Rh antigen, they may become sensitized. Two mechanisms are proposed for this sensitization. The most likely mechanism is the occurrence of an undetected placental leak of fetal red blood cells into the maternal circulation during pregnancy. The other proposal is the “grandmother” theory. This theory suggests that an Rh-negative woman may have been sensitized from birth by receiving enough Rh-positive cells from her mother during her own delivery to produce an antibody response.

In general, two exposures to the Rh antigen are required to produce any significant sensitization, unless the first exposure is massive. The first exposure leads to primary sensitization, whereas the second causes an anamnestic response leading to the rapid production of immunoglobulins, which can cause a “transfusion reaction” or hemolytic disease of the fetus during pregnancy.

The initial response to exposure to Rh antigen is the production of immunoglobulin M (IgM) antibodies for a short period of time, followed by the production of IgG antibodies that are capable of crossing the placenta. If the fetus has the Rh antigen, these antibodies will coat the fetal red blood cells and cause hemolysis. If the hemolysis is mild, the fetus can compensate by increasing the rate of erythropoiesis. If the hemolysis is severe, it can lead to profound anemia, resulting in hydrops fetalis from congestive cardiac failure and intrauterine fetal death. High bilirubin levels can damage the central nervous system and lead to neonatal kernicterus. Before the widespread use of Rh immune globulin for prevention of Rhesus isoimmunization, neonatal kernicterus was one of the leading causes of cerebral palsy.

The fetal and maternal circulations are normally separated by the placental “barrier.” Small hemorrhages occur in either direction across the intact placenta throughout pregnancy. With advancing gestational age, the incidence and size of these transplacental hemorrhages increase, with the largest hemorrhages usually occurring at delivery. Most immunizations occur at the time of delivery, and antibodies appear either during the postpartum period or following exposure to the antigen in the next pregnancy.

If a pattern of mild, moderate, or severe disease has been established with two or more previous pregnancies, the disease tends either to be of the same severity or to become progressively more severe with subsequent pregnancies. If a woman has a history of fetal hydrops with a previous pregnancy, the risk for hydrops with a subsequent pregnancy is about 90%. Hydrops usually develops at the same time as, or earlier than, in the previous pregnancy.

image Incidence

Although transplacental hemorrhage is very common, the incidence of Rh immunization within 6 months of the delivery of the first Rh-positive, ABO-compatible infant is only about 8%. In addition, the incidence of sensitization with the development of a secondary immune response before the next Rh-positive pregnancy is 8%. Therefore, the overall risk of immunization for the second full-term, Rh-positive, ABO-compatible pregnancy is about 1 in 6 pregnancies. The risk for Rh sensitization following an ABO-incompatible, Rh-positive pregnancy is only about 2%. The protection against immunization in ABO-incompatible pregnancies is due to the destruction of the ABO-incompatible cells in the maternal circulation and the removal of the red blood cell debris by the liver.

Transplacental hemorrhage may also occur before delivery. Establishment of the fetal circulation occurs at about 4 weeks’ gestation, and the presence of the Rh D antigen has been demonstrated as early as 38 days after conception. Consequently, Rh isoimmunization can occur at any time during pregnancy, from the early first trimester on. In the first trimester, the most common causes of transplacental hemorrhage are spontaneous or induced abortions. The incidence of immunization following spontaneous abortion is 3.5%, whereas that following induced abortion is 5.5%. The risk is low in the first 8 weeks, but it rises to significant levels by 12 weeks’ gestation. The risk for immunization following ectopic pregnancy is about 1%. Transplacental hemorrhage can also occur in the setting of second- or third-trimester vaginal bleeding, after invasive procedures such as amniocentesis or chorionic villus sampling, after abdominal trauma, or after external cephalic version. If necessary the amount of fetal blood entering the maternal circulation after an episode associated with transplacental hemorrhage can be estimated using the Kleihauer-Betke test (described later). All pregnant Rh-negative women who are not sensitized to the D antigen should routinely receive prophylactic Rh immune globulin (RhOGAM) at 28 weeks of gestation, within 72 hours of delivery of an Rh-positive fetus, and at the time of recognition of any of the problems cited previously that are associated with transplacental hemorrhage.

image Detecting Fetomaternal or Transplacental Hemorrhage

The Kleihauer-Betke test is dependent on the fact that adult hemoglobin is more readily eluted through the cell membrane in the presence of acid than is fetal hemoglobin (HbF). The maternal blood is fixed on a slide with ethanol (80%) and treated with a citrate phosphate buffer to remove the adult hemoglobin. After staining with hematoxylin and eosin, the fetal cells can readily be distinguished from the empty maternal cells. All cells are then counted, and an estimate of the extent of the fetal to maternal hemorrhage (measured in milliliters) is made on the basis of the following equation:

image

image Recognition of the Pregnancy at Risk

A blood sample from every pregnant woman should be sent at the first prenatal visit for determination of the blood group and Rh type and for antibody screening. In Rh-negative patientswhose anti-D antibody titers are positive (i.e., those who are Rh sensitized), the blood group and Rh status of the father of the baby should be determined. If the father is Rh negative, the fetus will be Rh negative, and hemolytic disease will not occur. If the father is Rh positive, his Rh genotype and ABO status should be determined. This may be done by testing the father’s red blood cells with the reagents available for the antigens D, E, C, e, and c. Newer molecular techniques are now available to assess fetal Rh genotype. If he is homozygous for the D antigen, every fetus he fathers will be Rh positive and could potentially be affected. If he is heterozygous, only half of his children will be affected. Information regarding the zygosity of the father is of value in absolutely predicting the presence or absence of the Rh antigen in the fetus if the father is homozygous and in signaling the potential need for fetal antigen testing if the father is heterozygous. About 56% of Rh-positive whites are heterozygous for the Rh D antigen. If it is not possible to test the antigen status and zygosity of the father, it must be assumed that he is antigen positive.

MATERNAL RH-ANTIBODY TITER

Anti-D antibody titers generally provide limited information regarding the severity of fetal hemolysis in Rh disease. However, many centers continue to use anti-D antibody titers to help guide their decision making regarding the initiation of testing procedures (e.g., amniocentesis, middle cerebral artery Doppler studies, and percutaneous umbilical blood sampling). The American College of Obstetricians and Gynecologists (ACOG) and other independent researchers have recommended that, in an initially immunized pregnancy, the fetus is not in serious jeopardy if the titer remains below 1:16. In patients with a positive titer less than 1:16, repeat titers should be obtained every 2 to 4 weeks. If the titer rises to 1:16 or greater, a more detailed assessment and determination of fetal Rh status is indicated. The timing and methods of invasive testing will depend on the current clinical status of the fetus, the gestational age, and the patient’s obstetric history. Titers are not generally useful for following a patient with a history of a previous fetus or neonate with hemolytic disease. In this setting, even if the titers are below the critical threshold, the patient should be followed and evaluated as if her titers were high.

TECHNIQUES FOR EVALUATING FETAL RH STATUS

In the United States, amniocentesis is the most commonly employed method to test fetal blood type in cases of a heterozygous paternal genotype. Most laboratories offering fetal red cell antigen typing on amniotic cells require an accompanying paternal blood sample, and with the discovery of an Rh D pseudogene in 21% of African Americans, a maternal blood sample should also be provided. Chorionic villus sampling has been used to determine fetal blood type but is discouraged because of the higher potential for transplacental hemorrhage and worsening fetal disease if the fetus is Rh D positive.

Flow cytometry has been successfully reported in sorting fetal cells from maternal blood. DNA amplification using a single fetal nucleated erythrocyte can be used to determine fetal Rh D blood type. Determination of free fetal DNA in maternal plasma or serum is another noninvasive test that is being increasingly used to detect fetal Rh D status.

ULTRASONIC DETECTION OF RH SENSITIZATION

Serial ultrasonic examinations of a woman with a fetus at risk for hemolytic disease can be a useful adjunct to amniocentesis in confirming fetal well-being and determining the advent of fetal hydrops. The examination should include a routine fetal assessment plus a determination of placental size and thickness and hepatic size. Both the placenta and the fetal liver are enlarged with hydrops. Fetal hydrops is easily diagnosed by the characteristic appearance of one or more of the following: ascites, pleural effusion, pericardial effusion, or skin edema. Appearance of any of these factors during an ultrasonic examination eliminates the need for diagnostic amniocentesis and necessitates therapeutic intervention based on fetal gestational age.

Doppler assessment of peak velocity in the fetal middle cerebral artery (MCA) in cm/sec has proved to be the most valuable tool for detecting fetal anemia. At-risk pregnancies should have this test performed every 1 to 2 weeks from 18 to 35 weeks of gestation. A fetal MCA peak systolic velocity (PSV) value above 1.5 multiples of the median for gestational age is predictive of moderate to severe fetal anemia and is an indication for percutaneous umbilical blood sampling for precise determination of fetal hemoglobin concentration. Intrauterine fetal transfusion should follow if indicated. After 35 weeks’ gestation, this test may produce a higher false-positive rate (Figure 15-1), and amniotic fluid spectrophotometry may be indicated.

image

FIGURE 15-1 Middle cerebral artery (MCA) Doppler peak velocities based on gestational age. MoM, multiples of the median.

(Data from Moise KJ Jr: Management of rhesus alloisoimmunization. Obstet Gynecol 100:600-611, 2002.)

AMNIOTIC FLUID SPECTROPHOTOMETRY

Before widespread use of MCA Doppler studies, analysis of amniotic fluid was the most frequently used method of gauging the severity of fetal hemolysis. It is still used in settings in which expertise in the performance of an MCA Doppler is not available, or occasionally, after 35 weeks of gestation. There is an excellent correlation between the amount of biliary pigment in the amniotic fluid and the fetal hematocrit, beginning at 27 weeks’ gestation.

The most likely source of bilirubin in the amniotic fluid is tracheal and pulmonary efflux with some transudate from the umbilical and placental vessels. Because of the small concentrations found in the amniotic fluid, spectrophotometric analysis is the most widely used technique for estimating amniotic fluid bilirubin concentration.

The optical density deviation (ΔOD) at 450 μ from a baseline drawn between the optical density values at 365 and 550 μ forms a peak that can be used to calculate the change in ΔOD in nanomoles (nm) at 450 μ measuring the amniotic fluid unconjugated bilirubin, which correlates with the cord blood hemoglobin of the newborn at birth.

Bilirubin is oxidized to colorless pigments when it is exposed to light; therefore, the fluid should be protected from light. Heme pigments and meconium may cause falsely high spectrophotometric values.

Bilirubin is normally found in amniotic fluid in a concentration that gradually diminishes toward term. For predictive interpretation, Liley devised a spectrophotometric graph based on the correlation of cord blood hemoglobin concentrations at birth and the amniotic fluid change in optical density at 450 μ. Using this method, he was able to establish predictive zones for mild, moderate, and severe disease. The Liley chart (Figure 15-2can be used to determine, with accuracy, the severity of the disease and the appropriate management, beginning at 27 weeks’ gestation. The Queenan curve, a modified Liley curve with four zones instead of three, is used as a predictive tool in some centers from 14 to 40 weeks’ gestation (Figure 15-3). Because single ΔOD 450 values are helpful only if they are very high (zone III) or very low (zone I), serial sampling of amniotic fluid is generally indicated.

image

FIGURE 15-2 Modified Liley chart used to determine the appropriate management of the patient with isoimmunization. The optical density at a wavelength of 450-nm (ΔOD 450-nm) level in the amniotic fluid at a given weeks’ gestation determines whether fetal transfusion or delivery is advisable.

image

FIGURE 15-3 Queenan curve for optical density at a wavelength of 450-nm (ΔOD 450-nm) values for the management of the patient with isoimmunization. Rh, rhesus.

(Adapted from Queenan JT, Tomai TP, Ural SH, et al: Deviations in amniotic fluid optical density at a wavelength of 450-nm in Rh-immunized pregnancies from 14 to 40 weeks’ gestation: A proposal for clinical management. Am J Obstet Gynecol 168:1370-1376, 1993.)

The severity of hemolytic disease in the prior pregnancy provides an approximate index for the timing of the first amniocentesis. This may range from 22 to 30 weeks with prior severe disease indicating the initial procedure as early as 22 weeks. With repeat sampling one of three trends will emerge: (1) Falling ΔOD 450 values are indicative of a fetus that is either unaffected (e.g., Rh negative) or very mildly affected. No intervention is indicated in these patients (see Figures 15-1 and 15-2). (2) If the ΔOD 450 is either stable or rising, frequent ΔOD 450 determinations are necessary to determine the timing of delivery. (3) If the ΔOD 450 enters zone III (refer to zone levels on the right side ofFigure 15-2) before 34 weeks, percutaneous umbilical blood sampling is performed for determination of fetal hemoglobin followed by intrauterine transfusion if indicated.

TECHNIQUE FOR AMNIOCENTESIS

An ultrasonic examination is performed to localize a pocket of amniotic fluid far enough away from the fetus and placenta to obtain a sample safely. A 22-gauge spinal needle is inserted, and 10 mL of fluid is aspirated using a sterile technique. The fluid is transferred to a dark or foil-wrapped tube to prevent deterioration due to light exposure and is sent for assessment of the ΔOD 450. The incidence of fetal mortality from amniocentesis for hemolytic disease is reported to be less than 1 in 900 in experienced centers. Of potential concern is the procedure-related risk for fetomaternal hemorrhage, which may worsen the severity of the sensitization. The incidence of fetomaternal hemorrhage is reported to be 8.4% to 11% per procedure.

PERCUTANEOUS UMBILICAL BLOOD SAMPLING

Advances in fetal interventional techniques and high-resolution ultrasonography have made direct fetal blood sampling the most accurate method for the diagnosis of fetal hemolytic disease. Percutaneous umbilical blood sampling (PUBS) can allow measurement of fetal hemoglobin, hematocrit, blood gases, pH, and bilirubin levels. The hemoglobin values for normal fetuses from 18 to 30 weeks’ gestation range from about 11.5 to 13.4 ± 1 g/dL. The technique for fetal blood sampling is similar to that described for fetal intravenous transfusion. One drawback is that it requires expertise above and beyond that required for amniocentesis. The major risk is fetal exsanguination from tears in placental vessels. If performed by an experienced practitioner, the risk for this complication and fetal death is 1% to 2% or less. However, there is a greater risk for fetomaternal hemorrhage,reported to be as high as 40%. Percutaneous umbilical blood sampling should not be a first-line method of assessing fetal status unless clearly indicated.

INTRAUTERINE TRANSFUSION

Intrauterine transfusion, initially introduced in 1963 as an intraperitoneal transfusion and presently usually administered as an intravascular transfusion has markedly changed the prognosis for severely affected fetuses. The goal is to transfuse fresh group O, Rh-negative packed red blood cells. In addition to routine blood screening, the blood for transfusion is irradiated, washed, processed through a leukocyte-poor filter, and screened for cytomegalovirus. Curare is usually injected directly into the fetal thigh with a 22-gauge spinal needle before transfusion, regardless of method, to immobilize the fetus during the procedure. Repeat transfusions are generally scheduled at 1- to 3-week intervals. The final transfusion is typically performed at 32 to 34 weeks’ gestation. In general, the fetus is delivered when the lungs are mature.

The overall survival rate following intrauterine transfusion is about 85%. In fetuses with no evidence of hydrops, the survival rate is about 90%, and for fetuses with hydrops before the transfusion, the survival rate is about 75%.

Fetal Intraperitoneal Transfusion

Red blood cells are absorbed through the subdiaphragmatic lymphatics and proceed through the right lymphatic duct into the fetal intravascular compartment. After transfusion, the absorption of blood may be monitored with serial transverse ultrasonic scans of the fetal abdomen. In nonhydroptic fetuses, the blood should be absorbed within 7 to 9 days. In the presence of hydrops, absorption is variable and may necessitate removal of ascitic fluid at the time of transfusion.

Under real-time ultrasonic guidance, a 20-gauge spinal needle is inserted through the mother’s abdomen into the fetal peritoneal cavity. The correct positioning of the needle is determined by injection of a small amount of normal saline and carbon dioxide, which can be easily visualized with ultrasonography. The red blood cells are slowly injected manually in 10-mL aliquots through an extension catheter attached to the spinal needle. If fetal bradycardia occurs at any time during the procedure, the transfusion is terminated.

For intraperitoneal transfusions, the volume to be infused is based on the following formula:

image

For example, a 30-week fetus would require a 100-mL transfusion (30 weeks − 20 × 10 = 100 mL).

Intravascular Transfusion

Because many fetuses are not subjected to transfusion until ascites is present, intravascular fetal transfusion has become increasingly popular. In addition, transfusion into the peritoneal cavity can result in fetal bradycardia or a pseudosinusoidal fetal heart rate pattern following the procedure because of compression at the site of insertion of the umbilical cord.

Under ultrasonic guidance, a 22-gauge spinal needle is inserted into the umbilical vein or the hepatic part of the umbilical portal venous system at the level of the umbilical cord on the fetal abdomen. If the umbilical vein is used, the preferred sites are either at the placental cord insertion or into a loop of umbilical cord. The volume of blood to be transfused is based on the fetal body weight, as determined by ultrasonography. One maternal side effect of intrauterine transfusion is the development of alloantibodies.

OTHER MODES OF THERAPY

Maternal plasmapheresis may be helpful in severe erythroblastosis when intrauterine transfusions are not successful, but perinatal outcome with this technique has not been impressive. Phenobarbital has been used to induce fetal hepatic enzyme maturation, thereby increasing uptake and excretion of bilirubin by the liver. Treatment with phenobarbital is initiated 2 to 3 weeks before delivery.

image Timing of Delivery in the Rh-Sensitized Fetus

In addition to serial ultrasounds for MCA Doppler studies and detection of hydrops, (or alternatively, serial amniocenteses for ΔOD 450 readings), these fetuses are evaluated at least twice weekly from 24 to 28 weeks for fetal well-being (NST, modified biophysical profile) and fetal growth. While the goal is a term neonate, the risks for intrauterine demise, including that from procedure-related losses, must be balanced against the risks for prematurity. There is no absolute gestational age cutoff for intrauterine transfusion, but after 34 weeks, the risk for an intrauterine loss in this setting may be greater than the risk for a neonatal death, and it may be prudent to deliver the fetus. If delivery is expected to occur before 34 weeks’ gestation (or if amniocentesis suggests an immature lung profile), betamethasone should be given at least 48 hours before delivery to enhance fetal pulmonary maturation.

image Prevention of Rhesus Isoimmunization

Because Rh isoimmunization occurs in response to exposure of an Rh-negative mother to the Rh antigen, the mainstay for prevention is the avoidance of maternal exposure to the antigen. Rh immune globulin (RhO-GAM) diminishes the availability of the Rh antigen to the maternal immune system, although the exact mechanism by which it prevents Rh isoimmunization is not well understood.

RhO-GAM is prepared from fractionated human plasma obtained from hyperreactive sensitized donors. The plasma is screened for hepatitis B surface antigen and anti-HIV-1. The globulin is available in several dosages for intramuscular injection. Since the advent of its use in 1967, Rh immune globulin has dramatically reduced the incidence of Rh isoimmunization.

Because the greatest risk for fetal-to-maternal hemorrhage occurs during labor and delivery, Rh immune globulin was initially administered only during the immediate postpartum period. This resulted in a 1% to 2% failure rate, thought to be due to exposure of the mother to fetal red blood cells during the antepartum period. The indications for the use of Rh immune globulin have therefore been broadened to include any antepartum event (such as amniocentesis) that may increase the risk for transplacental hemorrhage. The routine prophylactic administration of Rh immune globulin at 28 weeks’ gestation is now the standard of care. Despite adherence to this suggested Rh immune globulin protocol, 0.27% of primiparous Rh-negative patients still become sensitized.

INDICATIONS FOR ADMINISTRATION OF RhO-GAM

It is the responsibility of every health care practitioner who is involved in the care of pregnant women to prevent Rh isoimmunization. This is done by the timely administration of RhO-GAM to pregnant Rh-negative women, whose anti D titers are negative, at about the time of events associated with fetal-maternal hemorrhage. The following provides a practical approach to the administration of Rh immune globulin to an Rh-negative unsensitized patient.

All pregnant women should have a blood type and antibody screen on their first prenatal visit. During an uncomplicated pregnancy, the Rh-negative woman whose initial antibody screen is negative should have a repeat antibody titer at 28 weeks’ gestation. If the antibody screen is still negative, she should routinely receive an intramuscular injection of 300 μg of RhO-GAM prophylactically. A positive antibody screen does not necessarily mean the woman is not a candidate for RhO-GAM. In this case, an antibody identification and titer must be requested. If the antibody is not an anti-D antibody, then she is still a candidate for RhO-GAM.

RhO-GAM should also be administered during the antepartum period at any gestational age to an Rh-negative unsensitized (anti-D–negative) woman at the time of spontaneous or induced abortion, treatment of an ectopic pregnancy, significant vaginal bleeding, performance of an amniocentesis, abdominal trauma, or external cephalic version. Before 12 weeks of gestation, 50 to 100 μg of RhO-GAM should be sufficient to prevent isoimmunization.

Because chorionic villi in gestational trophoblastic disease are avascular and are devoid of fetal erythrocytes, RhO-GAM is probably not necessary following termination of a “complete” molar pregnancy. A “partial” molar pregnancy may have fetal tissue, and theoretically fetal cells could enter the maternal circulation. At least one case of sensitization following a molar pregnancy has been reported.

The risk for transplacental hemorrhage increases at the time of delivery, especially with cesarean birth or manual removal of the placenta. At delivery, cord blood must be sent for determination of the fetal blood group, Rh type, and for a direct Coombs’ test. RhO-GAM (300 μg) is routinely given to all Rh-negative, anti-D–negative women who deliver an Rh-positive infant within 72 hours of delivery. If a transplacental hemorrhage of greater than 30 mL of fetal blood is suspected (as might occur in the setting of an abruption, manual removal of the placenta, or severe maternal abdominal trauma), a Kleihauer-Betke test is helpful in determining the volume of the hemorrhage. Additional RhO-GAM may then be given at a dose of 10 μg of RhO-GAM per 1 ml of fetal blood that entered the maternal circulation.

Irregular Antibodies

Although Rh isoimmunization is the most common cause of hemolytic disease in the newborn, other blood group systems may be involved, such as Kell, Duffy, or Kidd. Kell antigen can elicit a strong IgG response similar to Rh isoimmunization. For this reason, any positive antibody screen in pregnancy, even in an Rh-positive woman, should be followed up with an antibody identification and titer. If the antibody screen is positive for one or more antibodies associated with hemolytic disease of the newborn, the pregnancy should be followed in a fashion similar to that advised for the Rh-sensitized pregnancy. A potential exception to this is Kell sensitization. Antibody titers and amniotic fluid spectrophotometry are not as reliable for detecting fetal anemia in this situation, probably because the anemia is due more to suppression of hematopoiesis than to hemolysis. The MCA PSV remains an excellent predictor of anemia in this setting. It is extremely important to test the Kell antigen status of the father of the fetus before any invasive testing because 90% of the population is Kell negative.

SUGGESTED READING

American College of Obstetricians and Gynecologists. Management of alloimmunization during pregnancy. ACOG Practice Bulletin No. 75. Washington DC: ACOG; 2006.

Brown S., Kellner L.H., Munson M., et al. Non-invasive prenatal testing: Technical strategies to achieve testing of cell free fetal DNA (cffDNA) RHD genotype in a clinical lab. Am J Obstet Gynecol. 2007;197:S173.

Howe D.T., Michailidis G.D. Intraperitoneal transfusion in severe early-onset Rh isoimmunization. Obstet Gynecol. 2007;110:880-884.

Mari G., Deter R.L., Carpenter R.L., et al. Noninvasive diagnosis by Doppler ultrasonography of fetal anemia due to maternal red-cell alloimmunization. Collaborative Group for Doppler Assessment of the Blood Velocity of Anemic Fetuses. N Engl J Med. 2000;342:9-14.

Mari G., Deti L., Oz U., et al. Accurate prediction of fetal hemoglobin by Doppler ultrasonography: Obstet Gynecol. 2002;99:589-593.