College of American Pathologists
Printable Version

  Q & A





April 2009

Fredrick L. Kiechle, MD, PhD

Question Q. The neonatologist at our hospital insisted O-negative platelets be given to a female infant who was O-negative. We had O-positive platelets but had to make a special courier run to another facility an hour away. What quantity of O-positive platelets can be given to a female infant before we must worry about an antibody (D) response?

A. Whenever an Rh-negative female of childbearing potential is transfused, she should be transfused with Rh-negative products to prevent the formation of anti-D, which can cause hemolytic disease of the fetus and newborn if the patient becomes pregnant. Since less than 1 mL of Rh-positive RBCs can cause alloimmunization, plate­lets should also be Rh-negative. The risk of an infant under four months old forming anti-D is very low because of the immune-naive/tolerant state that occurs in utero and after birth, but the risk is not zero. Therefore, the requirements for compatibility testing in children under four months old are abbreviated. Having said that, the low risk of forming anti-D is probably not acceptable to most patients and practitioners, so avoiding the use of Rh-positive platelets is recommended when there is time to obtain Rh-negative platelets. If the baby has to receive Rh-positive platelets because of a life-threatening bleed in the setting of thrombocytopenia, administering RhIG can be considered. Unfortunately, IM and IV RhIG are not recommended for infants. IV RhIG is recommended for Rh-negative “female children and female adults” (package insert for WinRho SDF, Baxter Healthcare). Some practitioners will ask the pharmacy to use the package insert to calculate a dose based on the amount of RBCs in the transfused platelet, but this is not a standard practice. In addition, RhIG is manufactured from human plasma and, according to the package insert, may carry the same infectious disease transmission risks as any blood product that has undergone viral inactivation. If Rh-positive platelets are given in an emergency, the use of RhIG will have to be discussed with the patient’s doctors so that the risks and benefits of administration can be weighed. Though not required, ABO-identical platelets are used by most prac­titioners as a first choice for their patients. Second choice is to choose a plate­let product so the isoagglutin­ins in the plasma are compatible with the antigens on the RBCs of the recipient. This is especially important for patients who are transfused with large volumes of platelet products (those who are rendered cyto­penic after human progenitor cell transplants or undergoing high-dose chemotherapy for hematologic ma­lignancies). In addition, infants, because of their small blood volume, are also preferentially transfused with ABO-identical or ABO-compatible plate­lets. This is to prevent the coating of RBCs with isoagglutinins. For an occasional transfusion, though, using ABO­in­com­patible platelets is acceptable. If there is still concern about the risk, the unit can be washed or volume-reduced. Of note, there are case reports of A patients who have received O platelets developing severe hemolysis because of high titer anti-A in the transfused product. Therefore, this should be avoided. In an O patient, if there are no O-negative platelets available, ABO-compatible products (A, B, AB, or O) can all be used. In conclusion, for an O baby I would first recommend using Rh-negative platelets of a different ABO type. If this isn’t an option and the baby isn’t actively bleeding, I would wait for O-negative platelets to be shipped.

Susan D. Roseff, MD
Medical Director,
Transfusion Medicine
Virginia Commonwealth University
Medical Center

Member, CAP Transfusion Medicine Resource Committee

Question Q. With our new blood gas analyzers, we sometimes have problems reporting values when we have patients with high methemoglobins. The manufacturer assures us that, because our machine is so sensitive and reads at 520 wavelengths, the analyzer doesn’t know if it is detecting methemoglobin or something else. So we don’t get a result we can report. How do we know whether to treat the patient for high methemoglobin? What can we do?

A. It is critically important for laboratories to be able to detect significant concentrations of methemoglobin and carboxyhemoglobin in blood. Because neither species can be oxygenated, these compounds seriously impair the oxygen-carrying capacity of blood. Each five percent methemoglobin or carboxyhemoglobin represents five percent of hemoglobin that cannot be oxygenated. Thus, a patient with 10 percent methemoglobin or carboxyhemoglobin is equivalent to a patient who has had a 10 percent bleed.

It is relatively well known that pulse oximeters, because they use just two wavelengths, cannot detect either methemoglobin or carboxyhemoglobin.1,2 In contrast, laboratory-based co-oximeters, which use many wavelengths, can distinguish among reduced hemoglobin, oxyhemoglobin, methemoglobin, and carboxyhemoglobin. In theory at least, the more wavelengths at which absor­bance readings are taken, the greater the number of different species one can identify and quantitate.

With all these additional wavelengths, many manufacturers are embedding in their instruments algorithms to check for “sample integrity.” Does the sample seem to be real human blood? If a sample fails this check, one gets an error message, and the instrument may provide no results or results with disclaimers.

In my opinion, this is a case where the cure (checks for sample integrity) may be worse than the disease (the possibility that something is interfering with the measurement of the various hemoglobin species). In our experience, these cases, fortunately quite uncommon, are almost always false alerts.

As to troubleshooting, I’d recommend the following:

  1. Rerun the sample, on another co-oximeter if available. Different manufacturers use different algorithms for these sample integrity checks, so you may be able to get a result on another co-oximeter.
  2. Request a new sample. Remember, for oxyhemoglobin, methemoglobin, and carboxyhemoglobin, the specimen need not be an arterial sample; it can be venous (properly anticoagulated, usually heparin).
  3. Call the ordering physician. Ask whether the situation is consistent with a toxic exposure (smoke inhalation, nitrite exposure, oxidant drug). If not, it may be acceptable to report a calculated O2 saturation based on the PO2, with an appropriate disclaimer that this cannot rule out a toxic exposure.
  4. Check the sample visually. A sample with methemoglobin will appear chocolate brown; one with carboxyhemoglobin will appear “cherry red.” Even if you can’t provide a number, in a dangerously toxic situation knowing just this color difference can be quite helpful.
  5. If the clinical situation warrants, advise the clinicians to treat the patient presumptively. Having a dialogue with the ordering physician in this situation is critical. In an urgent clinical situation, it is not acceptable to say “unable to measure.”


1. Barker SJ, Tremper KK, Hyatt J. Effects of methemoglobinemia on pulse oximetry and mixed venous oximetry. Anesthesiology. 1989;70:112–117.

2. Vergilio J. Carbon monoxide poisoning. In: Shaw LM, ed. The Clinical Toxicology Laboratory. Washington, DC: AACC Press, 2001:350–352.

Gary L. Horowitz, MD
Associate Professor of Pathology, Harvard Medical School

Chair, CAP Chemistry
Resource Committee

Dr. Kiechle is medical director of clinical pathology, Memorial Healthcare, Hollywood, Fla.