College of American Pathologists
Printable Version

  Q & A





cap today

August 2003

Richard A. Savage, MD

Q. What product can be used to perform calibration verification for methemoglobin?

A. Methemoglobin is usually measured with carboxyhemoglobin and oxyhemoglobin using a multiwavelength oximeter. Absorbance readings at several wavelengths are combined with preset absorptivities of each hemoglobin species at each wavelength to calculate the concentration of each species.

Total hemoglobin is reported as the sum of oxy-, deoxy-, and carboxyhemoglobin, and methemoglobin, and the individual species are reported as a percentage of total hemoglobin. Calibration is possible only for the total hemoglobin reading because the factors used to calculate concentrations of individual hemoglobin species from absorbance readings are set by the manufacturer and cannot be adjusted.

Calibration verification needs to be done if there is a calibration for the analyte in the first place. Since oximeters can be calibrated only for total hemoglobin, this is the only analyte that requires calibration verification.

A manual method for methemoglobin uses absorbance differences at 630 nm before and after conversion to cyanmethemoglobin. Again, no calibration is involved, and calibration verification is not necessary because results are expressed as a percentage of total hemoglobin without determining the hemoglobin concentration. Accurate results can be obtained as long as the spectrophotometer exhibits linear absorbance readings over the range used.

Robert W. Burnett, PhD
Department of Pathology and Laboratory Medicine
Hartford (Conn.) Hospital
Consultant, CAP Chemistry Resource Committee

Q. Are there established criteria to help laboratories decide which complete blood cell count results require review? How do laboratories decide when a manual smear review is necessary? Who should perform manual reviews?

A. The issue of how best to handle the review of CBC results can be problematic. It’s not easy to use technology and personnel cost effectively while achieving a reasonable turnaround time and simultaneously maximizing clinical utility.

Automated blood cell analyzers can generate results with quantitative or qualitative abnormalities, or both. They can also generate numerically normal results, which can reflect a false-positive or false-negative result. Laboratories have written procedures for such situations that delineate which abnormalities warrant microscopic review and by whom. Ideally, these procedures reflect the level of training and experience of the laboratory personnel, sophistication of the automated blood cell analyzer, and incidence of disorders in the patient population. This means that no guideline can be universal or economically feasible. Identifying the criteria that should be adopted in a given setting requires the expert judgment of an experienced laboratory director using input from the clinical physicians and other qualified laboratory personnel.

So when should a laboratory perform a manual peripheral smear review?1,2 PSR might be warranted for assessing the accuracy of a platelet count, enumerating or confirming leukocyte populations if automated blood cell results are unavailable or invalid, and verifying the accuracy of these results if spurious results are suspected. PSR might also be warranted for diagnosing specific qualitative abnormalities, including such hematological malignancies as leukemias, hematological stem-cell disorders, such as myelodysplasias, and hereditary leukocyte disorders. And PSR might be warranted for evaluating specific quantitative and qualitative abnormalities, including cytopenias, hereditary hemolytic disorders, and for the presence of infectious agents, as well as for classifying lymphoproliferative disorders.

A related but secondary issue is who should perform the review.3 In some instances a bench technologist may suffice, such as when verifying the platelet count. In other situations, a technologist specialist or other experienced technologist may be acceptable. Still other cases may require the laboratory physician’s opinion. This physician should correlate the numerical and morphological findings with other relevant information in a particular clinical context, such as when classifying or distinguishing between infectious agents, such as malaria and babesiosis, or when initially classifying a newly identified acute leukemia.

Each laboratory must decide who should perform PSR based on its resources and patient population. In some cases, outside consultations may be required.


  1. Peterson P, Blomberg DJ, Rabinovitch A, et al. Physician review of the peripheral blood smear: when and why. An opinion. Lab Hematol. 2001;7:175–179.
  2. Peterson P. Standard criteria for smear review [letter]. Lab Med. 2002;33:671.
  3. Ault K. Peripheral smear review: now by whom? Lab Hematol. 2001;7:173–174.

Powers Peterson, MD Department of Pathology
Sentara Virginia Beach (Va.) General Hospital
Member, CAP Hematology/Clinical Microscopy Resource Committee

Q. Some merged hospitals in our group do automated and manual reticulocyte counting while others perform only manual counting. Since the results of manual and automated methods often yield a somewhat different result, do we need to distinguish between these methods with separate test identification and reference ranges?

A. The advent of automated reticulocyte counting using various fluorescent dye/laser excitation combinations has created a group of tests that have criteria for identifying a red cell as a reticulocyte, which clearly differs from the criteria applied when identifying reticulocytes visually in new methylene blood-stained smears. Therefore, it is not surprising that manual and automated methods do not agree entirely.

The clinical literature and CAP reticulocyte counting Surveys have demonstrated that automated methods using fluorescent dye/ laser excitation yield higher values than manual methods. Consequently, manual reticulocyte counts and automated reticulocyte counts should be treated as different tests and reported with a reference range derived specifically for that test.

In reviewing this issue with members of the CAP Hematology/Clinical Microscopy Resource Committee, I learned that all of the members who perform automated and manual testing maintain separate tests and reference intervals.

Robert Novak,
MD Department of Pathology
Children's Hospital Medical Center of Akron (Ohio) Chair,
CAP Hematology/Clinical Microscopy Resource Committee