Return to CAP Home
Printable Version

  Q and A





October 2008

Following are two answers to a reader’s question about Kleihauer-Betke stain testing, the first by Robert H. Davis, MD, PhD, of Mary Rutan Hospital, Bellefontaine, Ohio, and the second by Bruce H. Davis, MD, of Trillium Diagnostics, Bangor, Me. Please note that Trillium provides some of the products that Dr. Bruce Davis, the company’s president, discusses in his answer.

Question Q. What is the protocol for Kleihauer-Betke stain testing? We are seeing an increase in testing on positive moms, premature labor, repeat KB stain within 24 hours, and trauma cases. I am concerned because the test is so dependent on technique and interpretation.

A. Currently, fetal-maternal hemorrhage can be documented by such methods as flow cytometry, gel agglutination, and Kleihauer-Betke (acid elution). The most commonly used method is Kleihauer-Betke, or KB. While this procedure is used to determine the number of vials of Rh immune globulin to be administered to an Rh-negative mother, we are finding our obstetricians-gynecologists are using it more and more for assessing placental injury in cases of trauma and premature labor, as was stated in the question (along with fetal monitoring).

We have performed the KB test using the Sure-Tech Kit for more than 15 years. We have found this method to be reliable and to offer a good turnaround time. We have also performed this procedure on a stat basis. Our proficiency tests, which can be purchased from the CAP, have been good. As indicated in the question, the test must be performed as indicated on the instruction sheet. I reviewed our 2007 CAP Survey; the expected results are in the first set of parentheses and our results are in the second parentheses: 2007 HBF-1 (0.5 percent) (0.3 percent), HBF-2 (2.0 percent) (1.6 percent), HBF-3 (1.0 percent) (0.8 percent), HBF-4 (0 percent) (0 percent). Our 2008 results were HBF-1 (2.0 percent) (1.9 percent) and HBF-2 (0 percent) (0 percent). These results are in good agreement with the Surveys.

Though immunophenotyping has a greater accuracy, the KB test can give reliable results at a lower cost and with a faster turnaround time. I recently called five national laboratories that perform flow on leukemia/lymphoma samples. None of the labs performs determination on HbF. The sixth laboratory I called does do flow on HbF with a 48-hour turnaround time at a cost of $467. In addition, I checked with our larger regional medical center, which does flow on leukemia/lymphoma samples; it does not perform flow on HbF and, instead, uses the KB test.

If performed correctly, the KB test can give reliable results with a good turnaround time and at reasonable cost. A possible source of error is in the interpretation. Nucleated cells such as granulocytes and lymphocytes must not be counted. (I wrote a reference manual for Sure-Tech that addresses this issue.) Additionally, our blood bank specialist uses cord blood along with male blood at various dilutions for quality control. Because interpretation is critical, only a few of our medical technologists are trained in this procedure. This way, the technologists are more familiar and more comfortable with the procedure.


  1. Brecher ME. AABB Technical Manual, 14th ed. 2002: 510.
  2. Davis RH. Kleihauer-Betke Fetal Hemoglobin Reference Manual. Sure-Tech Diagnostics Assocs. Inc., 1996.

Robert H. Davis, MD, PhD
Department of Pathology
Mary Rutan Hospital
Bellefontaine, Ohio

Question A. Laboratory assessment of fetomaternal hemorrhage is one of the first examples of therapeutic monitoring, or where a laboratory test is used to determine the appropriate dosage of a drug. Despite the availability of FDA-cleared methods for fetal red cell detection using flow cytometry with excellent sensitivity and low imprecision for almost a decade, American laboratories have held firmly in the use of the microscopic Kleihauer-Betke acid elution method. As repeatedly published in the literature and consistently documented by the CAP HBF, Sure-Tech Diagnostics (no longer offered), and UK NEQAS proficiency tests, the precision of the KB test typically has a CV of 30 percent to 100 percent. Reported experiences of imprecision with the KB assay have called into question the clinical validity of this manual method.1-3

The CAP HBF Survey, in which stabilized mixtures of adult and fetal red cells are sent to more than 1,000 participants, has on several occasions examined the issue of laboratory performance around the levels of fetal cells that would affect therapeutic decisions of Rh immune globulin therapy. Surveys in 1999 and 2001 sent samples of 0.4 percent and 0.8 percent fetal cells, values significantly above and below the 30 mL or 0.6 percent fetal red cell threshold used in North America for additional RhIG dosage. Despite a two-year interval, the labs performing the KB test yielded a similar performance in that with the 0.4 percent samples, about 50 percent of the labs reported values of =0.6 percent, which in clinical practice would have resulted in the unnecessary administration of a second dose of Rh immune globulin. The 0.8 percent samples resulted in nearly 12 percent of the labs reporting results below the therapeutic threshold of 0.6 percent, which in practice would put the women at risk for Rh hemolytic disease of newborn in subsequent pregnancies; these Surveys reported all the labs using anti-HbF flow cytometry correctly reported values above 0.6 percent. Again in 2003, this CAP Survey examined the relationship of fetomaternal hemorrhage assessment and therapeutic action by sending samples with 0.2 percent and 0.6 percent fetal cells. This Survey asked labs not only to report their testing result but also whether their test value should result in any additional vials of RhIG. Here again more than 35 percent of labs would have directed unnecessary therapy for the 0.2 percent sample, some by as much as three to 10 vials in excess. Again, nearly 10 percent of the labs would have directed suboptimal dosing in the case of the 0.6 percent sample, which in the real clinical world could have disastrous consequences.

Many reasons are given for the reticence of American laboratories to embrace the more sensitive and precise flow cytometric methods.4 The arguments for retaining use of the KB assay include the lack of flow cytometric instruments in many hospitals, the limited number of personnel skilled in the use of flow cytometry, the location of flow cytometry in hematology or immunology labs and not in transfusion medicine labs, and the possible delay in the results from a reference lab using flow cytometry. Interestingly, none of these reasons seem valid enough to withhold flow cytometric assessment of samples from patients with leukemias and lymphomas.

Another reason commonly given for not adopting flow cytometric testing is that clinicians expect the test to be available as a rapid turnaround test and flow cytometry labs are typically not operational 24/7. As the questioner indicates, this is largely driven by the increasing need to deal with pregnant trauma patients, obstetric surgeries, and prenatal procedures. My personal recommendation is to use the KB assay as a semiquantitative assay for such emergent cases, as the clinical question is usually whether a significant fetomaternal hemorrhage occurred rather than the amount of the bleed. Then the flow cytometric assay can be used on positive samples to quantify more accurately the amount of the hemorrhage when the flow cytometry lab is open for any “positive” sample. This would parallel the practice of many blood banks that use a screening method for maternal samples and reflex the positive samples for enumeration of fetal cells. In practice, many hospital laboratory staff during the night shift often hold the fetomaternal hemorrhage specimen for more experienced day-shift technologists, as the KB assay is so subjective and laborious. So the resistance to flow cytometric assessment of fetomaternal hemorrhage based on a need for rapid turnaround time is a matter of personal perspective.

Currently there are several commercial assays and reagents for performing flow cytometric fetomaternal hemorrhage assessment. In several European countries, Britain, Australia, and New Zealand, flow cytometric assays are more commonly used, employing either monoclonal antibodies to hemoglobin F or the D(Rh) antigen.5-7 Their practice may have evolved to flow cytometry sooner as the unit dosage of RhIG is smaller, requiring detection of fetomaternal hemorrhage greater than 10 to 12 mL or 0.2 percent fetal red cells. The anti-HbF method is most prevalent in the U.S., as the Caltag fetal hemoglobin assay (distributed by Invitrogen, Carlsbad, Calif.) was cleared by the FDA many years before the Quant-Rho method of Alba Biosciences (distributed by Trillium Diagnostics, Bangor, Me.). The anti-D method is readily performed, though it is applicable only in D(Rh) antigen-negative women with a D(Rh)-positive fetus. The anti-HbF method is more broadly applicable to all pregnancies, but interpretation can occasionally be problematic in women with marked elevation in F cells, which can occur with hemoglobinopathies or hereditary persistence of HbF.

The two-color Fetal Cell Count assay (IQ Products, Groningen, Netherlands) uses both anti-HbF and anti-carbonic anhydrase monoclonals to separate F cells and fetal red cells, as carbonic anhydrase expression on red cells is induced only after birth.8 However, the Fetal Cell Count assay is not FDA cleared and not available in the United States. Trillium Diagnostics recently developed a more rapid, no-wash fetomaternal hemorrhage QuikQuant assay for use both on commercial flow cytometers and the Abbott Sapphire blood counter, but this, too, is not FDA cleared and is available now for research use only. However, the availability of a fetomaternal hemorrhage assay on a blood cell counter instrument would certainly address the perceived requirement for rapid turnaround 24/7 availability.

As is the case with all clinical assays, the use of controls in an assay is important. Controls can be used not only to assess the quality of KB staining but also to better define the region of analysis positive for fetal cells in flow cytometry. Controls for fetomaternal hemorrhage assays can be homebrew artificial mixtures of fresh cord blood and adult blood samples or the commercial pre-assayed three-level stabilized control product FetalTrol (Trillium). The ­Fetal­Trol product is composed of Rh or D antigen-positive fetal cells and Rh or D antigen-negative adult cells and can be used for anti-D antigen, anti-HbF, or KB assays.

Probably the best single source of information for the quality perfor­mance of the fetomaternal hemorrhage or fetal red cell detection assays is the CLSI H52-A document,9 which describes all the potential pitfalls of the KB and flow cytometry methods.


  1. Bromilow IM, Duguid JK. Measurement of fetomaternal haemorrhage: a comparative study of three Kleihauer techniques and two flow cytometry methods. Clin Lab Haematol. 1997;19:137–142.
  2. Duckett JR, Constantine G. The Kleihauer technique: an accurate method of quantifying fetomaternal haemorrhage? Br J Obstet Gynaecol. 1997;104:845–846.
  3. Emery CL, Morway LF, Chung-Park M, et al. The Kleihauer-Betke test. Clinical utility, indication, and correlation in patients with placental abruption and cocaine use. Arch Pathol Lab Med. 1995;119:1032–1037.
  4. Davis BH, Davis KT. Laboratory assessment of fetomaternal hemorrhage is improved using flow cytometry. Laboratory Medicine. 2007;38:365–373.
  5. Davis BH, Olsen S, Bigelow NC, Chen JC. Detection of fetal red cells in fetomaternal hemorrhage using a fetal hemoglobin monoclonal antibody by flow cytometry. Transfusion. 1998;38:749–756.
  6. Kumpel BM. Quantification of anti-D and fetomaternal hemorrhage by flow cytometry. Transfusion. 2000;40:6–9.
  7. Nelson M, Zarkos K, Popp H, et al. A flow-cytometric equivalent of the Kleihauer test. Vox Sang. 1998; 75: 234– 241.
  8. Porra V, Bernaud J, Gueret P, et al. Identification and quantification of fetal red blood cells in maternal blood by a dual-color flow cytometric method: evaluation of the Fetal Cell Count kit. Transfusion. 2007;47:1281–1289.
  9. Clinical and Laboratory Standards Institute. 2001. Fetal Red Cell Detection; Approved Guideline. CLSI document H52-A [ISBN 1-56238-452-X]. Wayne, Pa.

Bruce H. Davis, MD
President, Trillium Diagnostics
Bangor, Me.

Question Q. The technologists in our laboratory think that patients should be instructed to wait 24 hours before having their blood drawn or urine collected after a procedure in which they receive radiology contrast media. Do you have information about this?

In answer to the above question, Roger Calam, PhD, associate director of clinical chemistry, St. John Hospital and Medical Center, Detroit, provided us with a statement from the Guidelines on Contrast Media prepared by the Contrast Media Safety Committee of the European Society of Urogenital Radiology. It says, “Do not perform non-emergency biochemical analysis of blood and urine collected within 24 hours of contrast medium injection” (Thomsen HS, ed. Contrast Media—Safety Issues and ESUR Guidelines. Springer; 2006. ESUR Guidelines published 2007). Dennis Ernst, MT(ASCP), of the Center for Phlebotomy Education Inc., provided the following additional information regarding blood draws only:

A. The more time that passes between the infusion of contrast media and the drawing of blood, the better, but there really isn’t a set standard. Waiting 24 to 48 hours would be optimum but is not always practical. Fluorescein dye has been reported to interfere with creatinine, cortisol, and digoxin measurements.1 Iodine as a contrast media has been shown to cause spurious bicarbonate and chloride results owing to analytical interference of the absorbed iodine during a cystogram.2 Another study showed iodine in the iodinated contrast agent Ioxaglate causes a dose-dependent prolongation of thrombin time, thrombin coagulase time, partial thromboplastin time, and calcium thromboplastin time.3 Finally, two more studies indicate that iodinated contrast media alters RBC morphology.4,5

At the very least, specimens should be labeled as having been drawn after the infusion of radiologic dyes and a notation should accompany the test result. Stating the time of the infusion and the name of the dye or contrast media would be of help in interpreting test results properly.


  1. Garza D, Becan-McBride K. Phlebotomy Handbook. 7th ed. Upper Saddle River, NJ: Prentice Hall; 2005.
  2. Abdel-Wareth LO, Lirenman DS, Halstead AC, et al. Spurious rise in total carbon dioxide and chloride with negative anion gap after cystogram. Pediatr Nep­hrol. 1995;9(3):348–350.
  3. Schulze B, Beyer HK, Hanstein U. The in-vitro effects of Ioxaglate (Hexabrix) on coagulation, fibrinolysis and complement system (author’s transl) (article in German) Rofo. 1981;134(5):566–570.
  4. Aspelin P, Stöhr-Liessen M, Almén T. Effect of iohexol on human erythrocytes. I. Changes of red cell morphology in vi­tro. Acta Radiol Suppl. 1980;362:117–122.
  5. Man EB. A note about iodine-containing contrast media and the interference of teridax for cholangiograms in evaluation of thyroid function by measurement of serum iodine. Am J Roentgenol Radium Ther Nucl Med. 1960;83:497–500.

Dennis J. Ernst, MT(ASCP)
Center for Phlebotomy Education Inc.
Ramsey, Ind.

Question Q. As part of the standard processing for many small biopsy specimens, such as thin-needle biopsies, we prospectively prepare unstained slides for potential immunohisto­chemistry or other special studies. Is there a recommen­dation or requirement for archiving this material if the slides are not used, or a suggested time frame after which it may be discarded?

A. There is no specific requirement for retaining unstained slides intended for immunohistochemistry or other special studies. The CAP’s 10-year slide retention requirement does not state specifically which slides must be retained, but the assumption one would logically make is that the retention policy applies only to stained slides reviewed in the diagnostic evaluation.

From a practical viewpoint, it makes sense to retain unstained slides for a set period to accommodate requests from oncologists for additional markers or requests from other treating institutions for unstained slides after cases have been signed out. However, to avoid confusion, these slides should not be placed in the main slide file. Many institutions keep a separate slide box containing all unstained slides in numerical order and retain the slides for a predetermined period, typically four to six weeks, before discarding them. The medical director should establish the appropriate guidelines for retention and ensure that they are codified in the policies and procedures of the histology laboratory.

Richard W. Brown, MD
Medical Director
System Laboratory Services
Memorial Hermann Healthcare System
Chair, CAP Surgical Pathology
Resource Committee