When a physician requests hematological screening of the critically ill, the implicit requirement is evidence of any unexpected hematological diagnosis or complication, including possible drug toxicity; 'full blood count and film (smear)' allows this evaluation.
There are two reasons for measuring a full blood count on a venous (or arterial) sample which are particularly relevant to the critically ill. The first is simply to measure the hemoglobin concentration (or hematocrit) as an indicator, albeit imperfect (see below), of the total red cell mass in the body. The second is to monitor changes in the red cell, white cell, and platelet populations. Because marrow cell production is a highly dynamic process (new red cells are formed at a rate of 2 * 10 6/s), it is a sensitive biological indicator of adverse changes at the cellular level.
There is only one way to obtain the correct hemoglobin concentration, which is to take a venous (or arterial) blood sample and measure the hemoglobin concentration with a standardized and properly controlled instrument (Thom 1990). Hemoglobin concentration in a capillary sample is entirely dependent upon the uncontrollable degree of hemoconcentration involved in the process. Most blood gas analyzers do not measure hemoglobin; rather, they infer it. They will be fooled by denatured or aberrant forms of hemoglobin and thus usually use diluted samples.
How the laboratory makes its measurements is important and relevant to the way in which results are viewed. The basic analysis is the full blood count (hemoglobin, red blood cells, packed cell volume, mean red cell volume and hemoglobin content, neutrophils, lymphocytes, monocytes, eosinophils, basophils, and platelets) which is produced as a package by automated cell counters in about 50 s. Reticulocyte counts may become increasingly useful. These results must be available on a time scale which meets intensive care needs. Near-patient testing is only one option, and the benefits and penalties of this must be fully considered. When it is necessary, it should be delivered on site by the laboratory.
Counting blood cells appears to be very easy, but it is not. The technologies that are used to overcome the problems involved can differ substantially between different counters, and it is important to appreciate the strengths and limitations of each type. No result is better than the technique and quality control that has gone into producing it. It is even more important for the clinician to appreciate that the major cause of wrong results is misidentification of the sample. A total quality strategy begins with, and is entirely dependent on, the correct sample from the correct patient going into a bottle that is correctly identified.
It is no longer appropriate to refer to the laboratory and its 'Coulter counter'. Cell counters are not all the same and they have differing solutions to the problem of counting cells. These flexible entities come in a wide range of shapes and sizes. The first way in which technologies differ relates to the way in which they stabilize the cells before they can be counted. The second difference is in the way in which the separate blood cell populations are distinguished from one another.
Some technologies rely heavily on an assessment of size to identify blood cells. This is rarely clear cut, and these systems must employ sophisticated software to try and set cut-off points between the different populations. The problem is particularly difficult when the red cells are microcytic. Although the mean volume of the red cell population is clearly distinct from that of the platelets, the wide distribution of cell sizes contained with the red cell population means that a significant number may be of very similar size to the larger platelets. At the other end of the platelet distribution, i.e. for the smallest fragments in the population, the problem is to distinguish these from other small particles (dust) and the electronic noise in the sensing systems.
The problems involved in counting and classifying white cells ( Bentley.1990) are even more complex and revolve around how to define the white cell populations in terms which can be appreciated by a machine rather than the human eye. Traditionally, division of the total white cell population into its constituent elements has been based on visual assessment of a blood smear treated with appropriate dyes. This was never a reliable quantitative measure. Older machines attempt to distinguish white cell populations on the basis of an analysis of the size distribution alone. More recent versions may use other characteristics related to cell conductivity as well as size. The best of the modern machines use size and cytochemical characteristics to define the white cells. While each of the technologies will give similar results in normal samples, they can differ considerably when pathological samples are analysed. This means that analyses carried out on one machine cannot always be translated to another clinical laboratory setting. Reference ranges for any given laboratory must reflect the technology that is in use.
It has been historical practice to express the white cell differential in terms of the percentage of the total white cell count. This has probably contributed substantially to misunderstandings of the significance of the differential count data. It should have been clear that when the total white cell count either increases or decreases, the absolute numbers of each of the white cell populations will change significantly even with a normal percentage differential. All white cell differential count data must be expressed and understood as an absolute count (in units of 10 9/l) for each of the populations. If an automated differential count is not available, a total white blood cell count of less than 1 * 109/l represents severe neutropenia.
In the measurement of red cells there are significant differences in the analyses carried out by different automated cell counters. While the red cell count will often be similar, the red cell indices, particularly the mean cell volume, may vary substantially. As a measure of cell size, the mean cell volume has considerable drawbacks. These arise from the essentially elastic nature of the red cell. In vivo, red cells do not have a fixed volume. Ex vivo, their volume will vary in relation to storage conditions and oxygenation. A highly agitated and oxygenated sample will have a lower mean cell volume than an unoxygenated sample, but as time passes the cells in an EDTA anticoagulated sample will swell as they take up water. In addition, the red cell can function as an osmometer and will reflect, for example, changes in D-glucose plasma levels. In contrast, the mean cell hemoglobin is unaffected by different technologies and totally unaffected by the storage changes which afflict the mean cell volume.
Changes in erythropoietic activity can be monitored by a quantitative reticulocyte count. Because reticulocytes only exist in the circulation for about 12 to 60 h, changes in the reticulocyte count will reflect current erythropoietic output. Inflammatory cytokines will suppress erythropoiesis, and the earliest indication of this may be obtained by quantitative reticulocyte counting. This should not be confused with the semiquantitative reticulocyte estimation produced as a result of a visual microscope assessment of the reticulocytes on a spread blood film. The latter has a coefficient of variation of 30 to 50 per cent in normal subjects, and may be even higher in cases where the reticulocyte count is diminished. The newest counters now have automated reticulocyte counting built in as part of the regular full blood count.
Iron supply to the erythropoietic tissue has an immediate effect on the adequacy of hemoglobin synthesis. When the rate of iron supply is insufficient to meet the demands of the erythroid tissue, hypochromic red cells will be produced. This is functional, as opposed to storage, iron deficiency. The measurement of the percentage of cells which are hypochromic, i.e. that have an individual red cell hemoglobin concentration below 28 g/dl, is the single most effective variable for monitoring this condition. This measurement should not be confused with the mean corpuscular hemoglobin concentration which reflects the mean hemoglobin concentration in all the cells and is relatively stable and unchanging. Functional iron deficiency is a major limiting factor in the response to erythropoietin therapy. It is most common in patients with chronic renal failure treated with recombinant erythropoietin, but may be equally significant in patients with decreased oxygen supply which stimulates erythropoietin production.
At present only one family of instruments (Technicon H°, Bayer Diagnostics) is capable of making all these analyses. These counters are widely available in hematology laboratories in the United Kingdom, but it is important to ascertain from the laboratory which technologies are being used. When quantitative reticulocyte counting and quantitative percentage hypochromic estimates are required, these needs should be made known to the laboratory. However, interpretation and applications of these indices have yet to be established in intensive care patients.
Was this article helpful?