Complete blood count

A complete blood count (CBC) is a set of measurements made on a sample of blood anticoagulated with EDTA: the number of red cells, white cells and platelets, the hemoglobin, and a handful of quantities derived from them. An automated CBC typically includes nine parameters: the white cell count, the red cell count, hemoglobin (Hb), hematocrit (Hct), mean red cell volume (MCV), mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), red cell distribution width (RDW) and the platelet count. A differential, which divides the white cells into their types, may be added.^[1]

Counting was once a technician's task with a microscope and a hemocytometer, and an accurate count of the few hundred cells in its grid was good to only about ±16% for red cells and ±21% for white cells.^[2] It became a machine's through the Coulter principle, and a hematology analyzer now does it automatically; see Coulter counter for the history.

Hematocrit is the fraction of the blood's volume that is red cells. It is the red cell count times the mean volume of a red cell: 4.8 × 10¹² cells per liter at 90 fL each (1 fL is 1 µm³) fill 0.43 of the volume, a hematocrit of 43%. The other two indices of the red cell come from the same quantities. MCV is the hematocrit divided by the red cell count, MCH the hemoglobin divided by the red cell count, and MCHC the hemoglobin divided by the hematocrit.

The differential gives five kinds of white cell, neutrophils, lymphocytes, monocytes, eosinophils and basophils, as counts and percentages.^[3]

The numbers have to agree with each other, and laboratories use that. For red cells that are of normal size and color, the "three rules of three" hold: Hct is Hb × 3 ± 3, Hb is RBC × 3 ± 0.3, and RBC is Hb ÷ 3 ± 0.3, with Hb in g/dL, Hct in percent and RBC in 10¹²/L. They work less well when the cells are small or large. A result that breaks them marks a specimen to check before it is reported.^[1]

The Coulter Model S of late 1968 produced seven parameters from one sample. It diluted the blood for a white cell count and split it. One part was lysed and sent to a bath for white cells and hemoglobin, and the other was diluted again for red cells. Each bath was drawn through three apertures, 100 µm wide for white cells and 70 µm wide for red cells, and the counts were corrected for coincidence and averaged. If the stream from one aperture differed from the other two, for example because of a partial blockage, that aperture was voted out and the technologist was warned. The red cell count and sizing gave the MCV, and with the hemoglobin measured in the white cell bath they gave the hematocrit, MCH and MCHC. All seven were printed within a minute. The first advertisement claimed that "One operator in one hour can make more seven-parameter blood analyses more accurately than three overworked technologists can in eight."^[4] It did not automate the platelet count or tell the five types of white cell apart.^[4]

Normal blood holds about 5,000,000 red cells and 7,000 white cells in each microliter, according to the account of Coulter's motive by his historian.^[2] That is a ratio of roughly 700 to 1.

Reference intervals belong to a laboratory, an analyzer and a population. A study of 280 healthy adults at one hospital in Seoul, run on a Sysmex XN-2000 analyzer, took the central 95% of the results as the reference interval. For the white cell count it was 3.89 to 9.23 × 10⁹/L for all subjects. Across the six groups of age and sex the platelet count ran from 141 to 421 × 10⁹/L, the MCV from 81.3 to 104.1 fL and the hemoglobin from 11.5 to 16.9 g/dL. The authors found hemoglobin and hematocrit lower in women, MCV higher with age and platelets lower with age.^[3]

The World Health Organization defines anemia by hemoglobin thresholds, 130 g/L for males of 15 and older among them. They are fifth centiles of a reference population, set more than 50 years ago. A pooled analysis of healthy adults of 18 to 65 years from international data sources found fifth centiles of 134.9 g/L in men and 119.7 g/L in non-pregnant women.^[5]

An analyzer counts what it is shown. Platelets in blood anticoagulated with EDTA sometimes clump, and the clumps are not counted as single platelets, which gives a falsely low count called pseudothrombocytopenia. It is not uncommon in laboratories that process EDTA samples. Cold-reactive IgM antibodies agglutinate red cells, which gives a falsely low red cell count and hematocrit and a falsely high MCV, MCH and MCHC. Red cells that are extremely small, with an MCV below 60 fL, may fall under the red cell threshold and above the platelet threshold, and impedance analyzers then count them as platelets. Lipemia, hemolysis, hyperbilirubinemia, a glucose concentration of 600 mg/dL or more and dilution by an intravenous infusion are on the list of interferents too.^[1] Analyzers raise flags and display histograms and scattergrams for review.^[1]

The volume thresholds are a consequence of the principle. In a Coulter counter the height of a pulse is proportional to the volume of the cell, and sorting the pulses by height is how red cells and platelets are counted from the same stream. A cell that is too small to clear the threshold for one class lands in another. Analyzers measure by impedance, by optical analysis or by both. The Sysmex XN-2000 takes a platelet count from a dedicated channel, PLT-F, and reads the white cell types from a scattergram, with data items such as the fluorescent light intensity of the neutrophil area.^[3]

Blood was the first sample that Molecular Reality Corporation put through one of its pores. In late 2018 its founder bled into the flow cell of an adjustable pore, an episode told in Rain of blood (demonpore). A red cell in that account has a volume of roughly 90 µm³, which is the 90 fL of the example above, and the records list the gating and oversampling of single cells among the results of January 2019.^[6]

• Flow cytometry
• Coulter principle
• Wallace H. Coulter