Return to CAP Home
Printable Version

  Q & A





August 2010

Fredrick L. Kiechle, MD, PhD

Question Q. Our laboratory had a polycythemic patient come in for a CBC. We ran this sample on our Coulter LH750 and recorded a HCT value of 50.3. We then performed a manual-spun HCT and recorded a value of 56.0. Relying on the spun microhematocrit as the gold standard, we reported a value of 56.0 for this patient. Before this sample, the patient’s HCT values had been reported within the range of 45.0 and 50.0 and all had been instrument reported. After we reported the microhematocrit value of 56.0, the physicians wondered why the patient’s HCT had jumped all of a sudden. This patient had the following values reported from the Coulter LH750 instrument: RBC, 7.96; Hgb, 14.5; HCT, 50.3; MCV, 63.1; MCH, 18.2; MCHC, 28.9; and RDW, 24.7 (a review of the slide shows moderate poikilocytosis). What is the correct way to handle polycythemic patients who have such high hematocrits? Two different ideas are being proposed at our institution. One is that the instrument HCT should be reported because the spun hematocrit is falsely elevated owing to the polycythemia and poikilocytosis reported (which in turn causes an excess in trapped plasma). Poikilocytosis or anisocytosis, or both, can cause an elevated RDW, but it is suggested that only excessive poikilocytosis will falsely elevate the spun hematocrit because of the inability of the cells to pack tightly based on the different shapes encountered. The second proposed idea is that the spun hematocrit should be reported because it is the gold standard.

A. The CBC results from this polycythemic patient raise a number of issues, including linearity of results, the rule of three, and the spun hematocrit (that is, packed cell volume) versus the instrument hematocrit. First, is the RBC linearity exceeded on this sample? If the RBC linearity is exceeded, then the results are suspect. The reported RBC of 7.96 M/µL exceeds the linearity of the instrument in my laboratory, and, thus, a dilution of the blood would be prepared and run.

Second, the rule of three (3 x RBC = Hgb, 3 x Hgb = HCT) is violated. This is not surprising because the rule of three is true only for samples that have normal erythrocytes. The blood smear should be examined. In this polycythemic patient, we notice that the red cells are microcytic. I suspect that examination of the blood smear may reveal hypochromic and microcytic red blood cells. Polycythemic patients are known to become microcytic once their iron stores are depleted. Hypochromic and microcytic red blood cells will have a lower hemoglobin content than normochromic normocytic red blood cells. The hemoglobin result here, if questioned, can be rechecked by a second method.

Third, what is the true hematocrit in this patient? A spun hematocrit (or packed cell volume) is easy to perform, but sources of error are inherent in this technique. The packed cell volume measures the red cell concentration, not red cell mass, and patients with volume depletion may have normal or high PCV values owing to hemoconcentration. Insufficient centrifugation, poor mixing, and inappropriate anticoagulant concentration can also contribute to technical errors. In spun hematocrits, a number of disorders that have abnormal red cells may cause plasma to be trapped in the erythrocytes. The trapping of the plasma causes the microhematocrit to be one percent to three percent higher than the hematocrit reported on automated instruments, and with abnormal cells, up to six percent of the PCV can be increased.

The hematocrit is calculated on the LH750, and thus this value is unaffected by the trapped plasma. Errors of automated hematocrit are more common in patients with polycythemia, and manual methods for determining hematocrit may be preferred. Given the inherent difficulties associated with both the microhematocrit (or packed cell volume) and the automated hematocrit, switching hematocrit methods on this patient creates a great deal of confusion for the clinician and explains the large discrepancy in this patient’s hematocrit.


1. Fairbanks VF. Nonequivalence of automated and manual hematocrit and erythrocytic indices. Am J Clin Pathol. 1980;73:55–62.

2. Pearson TC, Guthrie DL. Trapped plasma in the microhematocrit. Am J Clin Pathol. 1982;78:770–772.

3. Guthrie DL, Pearson TC. PCV measurement in the management of polycythaemic patients. Clin Lab Haematol. 1982;4:257–265.

Tracy I. George, MD
Director, Hematology Laboratory
Stanford University Medical Center
Stanford, Calif.

Chair, CAP Hematology/Clinical
Microscopy Resource Committee

Question Q. How useful are immature granulocytes in an automated differential cell count?

A. Elevation of immature granulocytes (IG), bands, and neutrophils within the complete blood count is not necessarily indicative of infection and may be seen in a variety of other conditions. Such disorders include chronic inflammatory diseases, neoplasia, hemolysis, tissue damage, seizures, metabolic abnormalities, and myeloproliferative neoplasms. Elevation may be seen also with the use of certain medications such as steroids. Howev-er, the CBC has been used for early detection of bacteremia and infection for many years, with the objective of initiating adequate treatment in a timely fashion. The ratio of IG to the total leukocyte count and IG to the total neutrophil count has been described in the literature to predict bacteremia, particularly in febrile children, thus providing a confusing body of data. In these studies, the immature cell counts (designated as bands and less mature granulocytes) have been derived from manual differential cell counts of 100 to 200 cells. Due to differences in the definitions of a band neutrophil and to interobserver variability, the band count is difficult to measure accurately and precisely using manual methods. Many studies of suspected bacterial infection in adults have demonstrated poor performance of the band count as a diagnostic test, with unacceptably low sensitivity and specificity. For an extensive review of the clinical utility of the band count, see an article on this topic by P. Joanne Cornbleet, MD, PhD.1

Immature granulocytes (for example, promyelocytes, myelocytes, and metamyelocytes) can be enumerated as part of the leukocyte differential cell count on some hematology analyzers, such as the Sysmex XE-2100 and XE-5000, using analysis of nucleic acid fluorescence and side scatter. Normal IG reference range is less than 0.030 × 103/µL or less than 0.5 percent for older children and adults. Normal values in infants and newborns are not available, but clinical studies indicate that IG values above 0.3 × 103/µL or IG values greater than three percent correlate with serious infections and sepsis. Hospitalized premature infants who are older than seven days are more than three times as likely to have a positive blood culture result if any IG are present in the manual differential count or if the Sysmex IG count exceeds 0.5 percent, according to one study.2

Ansari-Lari, et al., evaluated 102 infected blood samples from adult patients and 69 noninfected blood samples and found that the automated IG count was comparable to the absolute neutrophil count as a predictor of infection. Receiver operating characteristic analysis demonstrated a sensitivity of about 35 percent and a specificity of 90 percent for the IG count.3

A study performed in an ER/ICU in England compared the IG ratio (IG/total white blood cell count) with a calculated I/T ratio (IG/total neutrophil count) and to C-reactive protein to determine their utility as early indicators of sepsis for a pediatric population.4 The study demonstrated that the I/T ratio was a more sensitive and specific indicator of sepsis than the IG ratio or CRP. The immature granulocyte fraction must stand on its own and be correlated with clinical conditions. Therefore, additional studies are needed to demonstrate utility in clinical practice.


1. Cornbleet PJ. Clinical utility of the band count. Clin Lab Med. 2002;22:101–136.

2. Nigro KG, O’Riordan M, Malloy EJ, et al. Performance of an automated immature granulocyte count as a predictor of neonatal sepsis. Am J Clin Pathol. 2005;123:618–624.

Ansari-Lari M, Kickler TS, Borowitz MJ. Immature granulocyte measurement using Sysmex XE-2100. Relationship to infection and sepsis. Am J Clin Pathol. 2003;120: 795–799.

Iddles C, Taylor J, Cole R, et al. Evaluation of immature granulocyte count in the diagnosis of sepsis using the Sysmex XE-2100 analyser. Sysmex J Int. 2007;17:20–29.

Alice L. Werner, MD
Children’s Hospital of the King’s Daughters
Norfolk, Va.

Member, CAP Hematology/ Clinical Microscopy Resource Committee

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