College of American Pathologists
Printable Version

  Q and A





cap today



December 2006

Richard A. Savage

Q: What values for hematology should have delta checks to prevent preanalytical errors?

A: Preanalytical errors that might affect hematology testing include specimen mix-ups, improper specimen acquisition, and compromise of specimen integrity between when the specimen is acquired and when it is received in the laboratory. Delta checks, the process of flagging differences in specific analytes between consecutive analyses, are one way to detect such problems. Delta checks are effective in detecting some forms of preanalytical errors.

To detect specimen mix-ups in hematology, it is best to perform delta checks on parameters that show the least short-term biologic variability, such that deviations are highly unlikely to be caused by changes in patient status. Among commonly assessed hematology parameters, mean cell volume, or MCV, and mean cell hemoglobin concentration, or MCHC, serve this purpose best. Both are extremely stable in a patient for the short term—that is 24 hours. For example, the diurnal biological coefficient of variation in MCV in healthy individuals is only 0.5 percent. Even in situations of rapid change in other red cell parameters, such as in acute hemorrhage, the MCV and MCHC will not change significantly in a 24-hour period since the reticulocyte response to acute blood loss does not begin for two to three days. With the exception of red blood cell transfusion and (rarely) acute intravascular hemolysis, there are no acute patient events that will change these indices appreciably in the short term. In the case of acute hemolysis with hemoglobinemia, the MCHC may be affected, but the MCV will not be.

Of the other preanalytical specimen issues, specimen hemolysis is also amenable to detection by delta checks applied to the MCHC. In this scenario, the red blood cell, or RBC, count decreases, with no change in hemoglobin and MCV, resulting in an increase in MCHC. Intravenous fluid contamination of samples and changes in cell counts due to specimen clotting theoretically may be detected by delta checks of hemoglobin; hematocrit; RBC count; white blood cell, or WBC, count; or platelet count. However, it is inadvisable to use delta checks on any one of these parameters to detect such specimen compromise, as acute changes in any of these are common in hospitalized patients and the false-positive rate would be unacceptably high. Yet IV fluid contamination produces a proportionate decrease in all of these values, so a composite delta check parameter triggered by decreases in all of the formed elements of the blood might be useful.

It should also be noted that performing delta checks on the MCHC has the added benefit of detecting analyzer malfunction, as the MCHC is calculated from the three RBC parameters—hemoglobin, MCV, and RBC counts—that are directly measured on most hematology analyzers. Therefore, problems in any of these measurements will affect the MCHC. The choice of triggering limits for delta checks depends on a number of factors, including the desired balance of sensitivity and specificity and the patient population being evaluated. A statistical approach to delta checks can be achieved by accruing pairs of patient data points from a representative population. Limits encompassing the desired proportion of the population (for example, 95 percent or 99 percent limits) can then be easily selected. Alternatively, the expected day-to-day, within-patient variability can be calculated from published data on biologic diurnal variability (SBiol) combined with the analytic imprecision (SAnal), derived from replicate analyses on the instrument in use, according to the following formula:STotal2 = v(SBiol2 + SAnal2)

Finally, one can simply choose limits based on clinical experience. It likely will be necessary to modify limits set by any of these methods after a period of use, based on the laboratory’s accumulated experience with the effectiveness of the procedure.


  1. Cembrowski GS, Carey RN. Quality control procedures employing patient data. Laboratory Quality Management. Chicago, Ill.: ASCP Press;1989:133–174.
  2. Jones AR, Twedt D, Swaim W, et al. Diurnal change of blood count analytes in normal subjects. Am J Clin Pathol. 1996;106:723–727.

Steven H. Kroft, MD
Associate Professor of Pathology
Medical College of Wisconsin
Director of Hematopathology
Dynacare Laboratories/
Froedtert Hospital

Q: What career opportunities—pathology related or other—are possible for the pathologist who no longer wants to deal with the medicolegal consequences of diagnostic work?

A: Pathologists seeking relief from the medical malpractice environment, financial premiums, as well as the emotional toll of their profession can continue to ply their diagnostic/analytical skills and play a significant role in medicine.

Federal, military, state, and, in selective situations, municipal agencies provide opportunities in the pathology/medical laboratory arena. State health laboratories provide environmental and public health testing. Large municipalities fund a medical examiner's office where a special interest in forensics is obviously a prerequisite.

A lesser move might be to contract for locum tenens pathology services with a company that pays your malpractice premium. Two other venues that come to mind are biotechnology and pharmaceutical companies engaged in investigational therapies and clinical trials. Employment in these fields usually requires the appropriate academic credentials.

Another avenue to explore is teaching. Health-related science programs at many levels are in need of faculty with real-world experience.

Thomas A. Gaffey, MD
Medical Laboratory
Mayo Clinic
Rochester, Minn.
Member, CAP Education Committee