Reference method for the enumeration of erythrocytes and leucocytes

Transcription

1 Clin. lab. Haemat , Reference method for the enumeration of erythrocytes and leucocytes INTERNATIONAL COUNCIL FOR STANDARDIZATION IN HAEMATOLOGY; PREPARED BY THE EXPERT PANEL ON CYTOMETRY Members: J.M. England (Chairman, UK), R.M. Rowan (Secretary, UK), 0. W. van Assendelft (USA), B.S. Bull (USA), W.H. Coulter (USA), K. Fujimoto (Japan), W. Groner (USA), A.R. Jones (USA), J.A. Koepke (USA), S.M. Lewis (UK), N. Tatsumi (Japan) & R.L. Verwilghen (Belgium). Consultant: C.E. McLaren (USA) Accepted for publication 14 January 1994 Introduction The International Council for Standardization in Haematology (ICSH) defines a Reference Method (1991) as a clearly and exactly described technique for a particular determination which, in the opinion of a defined authority, provides sufficiently accurate and precise laboratory data for it to be used to assess the validity of other laboratory methods for this determination. The accuracy of the reference method must be established by comparison with a definitive method where one exists. A reference method should be traceable to a primary metrological standard and the degree of inaccuracy and imprecision must be stated. Accurate measurement, traceable to primary metrological standards, can be obtained for Hb and PCV by means of specified reference procedures defined by the International Council for Standardization in Haematology (ICSH 1980; 1987). Accurate measurements for other parameters of the blood count have been less easy to achieve. The traditional methods for red and white cell counting have been based on direct haemocytometry of diluted blood in a counting chamber, however, the imprecision and inaccuracy of this technique renders it unsuitable as a reference method. ICSH has recommended that chamber counting be replaced by a single channel semi-automated electronic counter using the aperture-impedance principle (ICSH 1988) provided that a known volume of diluted blood is displaced through the orifice. However this procedure too has recognized sources of error amongst which are inaccurate displacement volume, recirculation of cells inside the orifice tube, cell adherence to the counting vial and flow pathway, inclusion of signals due to artefact and coincidence counting., Correspondence: Dr R.M. Rowan, Department of Haematology, Western Infirmary, Dumbarton Road, Glasgow, GI1 6NT, Scotland, UK. 131

2 132 ICSH Expert Panel on Cytometry Coincidence remains an important source of error unless correction is applied. Several formulae have been proposed for coincidence correction, and these have either been integrated into the instrument counting system or used to prepare correction tables. Most of these formulae assume that coincidence has a constant relationship at a particular orifice size and is unaffected by other factors such as cell volume or flexibility. Recently Lewis et al. (1989) described a method for accurate correction of red blood cell count for coincidence in aperture-impedance electronic blood cell counters which was based on extrapolation using counts obtained with several dilutions of the sample. This document extends the previous recommendations (ICSH 1988) to include more accurate coincidence corrections. BLOOD SAMPLES Fresh venous blood specimens should be obtained by syringe or evacuated container. Specimens should be rejected if there is visible haemolysis or if microclots are present. Specimens should be anticoagulated with K, EDTA (ICSH 1993) in a final concentration of pmol ( mg/ml) and placed in containers which allow sufficient air remaining to enable proper mixing. Specimens should be maintained at 20 2 C until processing. Ideally the period between venous sampling and processing should not exceed 4 h. Before testing the specimen should be well mixed by gently inverting the container. The number of inversions of a container, to achieve adequate homogeneity, will depend on the type and dimensions of the container. Standard evacuated containers, x 75 mm, containing 5 ml blood require an air bubble comprising at least 20% of the tube volume. Non-standard tubes, particularly narrower tubes require more complete inversions (NCCLS 1993). Vortex mixers should not be used. GLASSWARE Pipettes Positive displacement pipettes should be used, calibrated to an accuracy of +0.5%, validated by a process traceable to a primary metrological standard (Appendix). Volumetric flasks Grade A volumetric flasks made of borosilicate glass, should be used, each with a stated volume which has been tested by an appropriate national standards body, or which has been checked by the individual user by means of a gravimetric method using certified weights and corrections for buoyancy and temperature (Appendix). Counting vials The counting vial should have a minimum volume of 10ml. It should be of sufficient height that the entry orifice of the electronic cell counter is at approxi-

3 Reference method for enumeration of erythrocytes and leucocytes 133 mately half the depth of the fluid before counting commences and there should be more than 1 cm of fluid above that orifice after the completion of counting. Before use, the vial should be cleaned free of chemical contaminants and adventitious particles. It is necessary to confirm that cells do not adhere to the counting vial and cause a progressive fall in the count. This is ascertained by testing a number of vials from each batch. Counts are made at measured intervals after the dilutions have been placed in the counting vials (see below). The dilution should be mixed in the vial before counting. INSTRUMENT SPECIFICATIONS The design of the electronic cell counting instrument should be such that each cell is counted once and only once. The probability of re-circulation of particles through the sensing zone must be low and any impulses so created should be below the appropriate lower threshold. The detection electronics should have the ability to discriminate, with a low probability of error, by pulse amplitude between the cells of interest, other cells and electrical noise. Although in principle these criteria may be met with optical sensing counters, at present, only semi-automated single channel aperture-impedance particle counters are available. The instrument should have an orifice diameter pm and length % of the diameter. The volume displaced during the counting period must be known to within an accuracy of 1 % traceable to a national or international metrological standard. This displacement can be achieved by the motion of a mercury column or other suitable means. The differential displacement due to thermal effects should not be greater than 0.1% per "C. To avoid artefactual signals, ICSH recommends threshold settings which ensure that the signal to noise ratio is greater than 100: 1. REAGENTS Diluent The diluent must be a sterile, non-toxic, buffered salt solution suitable for the measurement system. Its tonicity should be stated and should fall in the range 280f 15 mosms. The constituents and their concentrations must be given on the label of the container. When counted in the vial it must contain less than 1 x lo5 particles per litre measurable at the threshold setting for the cells which are to be counted. The diluent must not cause cell lysis nor alter cell volume by > 2 fl over a 20 min period. Lytic agent When white blood cells are to be counted in a whole blood sample the red cells must first be lysed. The lytic agent must not affect the ability of the white cells to

4 134 ICSH Expert Panel on Cytometry be counted whilst the red cell stroma must be reduced to a residue which does not contribute to the white cell count. Flushing solution The flushing solution should consist of diluent degassed by vacuum or heating to 90 C. METHOD FOR RED CELL COUNT I. Flushing Immediately before carrying out each individual count, flush the instrument through with flushing solution. 2. Dilution Coincidence correction should be performed for every blood specimen which is to be counted. Make two primary dilutions (each of 0.1 ml blood plus 20 ml diluent). Take 0.02, 0.04 and 0.08ml of the primary dilutions and add these to 20ml aliquots of diluent to make two sets of four secondary dilutions, the most concentrated of these, i.e., ml, having a dilution factor of 1/ Each secondary dilution is then counted as follows: Secondary dilutions No. of counts Transfer to counting vial The diluted sample must be transferred to counting vials and gently mixed for about 30 s to get rid of bubbles. The count must be completed within 5 min of performing the dilution. 4. Counting procedure It is expected that thresholding will be verified as described below on a sample-bysample basis. Ideally 0.5ml should be drawn by displacement through the sampling aperture within s. The time required should be in conformity with that recommended by the manufacturer. This check is designed to recognize partial blockage of the aperture. The counts should be replicated as explained above and the replicate counts in each secondary dilution summated. 5. Validating thresholds Using a pulse height analyser the lower counting threshold is set in the valley between platelets and red cells. The count per channel measured at this position can be compared with the count per channel at a position corresponding to the red cell modal volume. Provided the count channel in the latter position is at least 50 times that of the former, the lower threshold is deemed to be satisfactory. The

5 Reference method for enumeration of erythrocytes and leucocytes 135 reference method is inapplicable if this criterion does not apply. The upper threshold is then set to the maximum possible value (it should be remembered that this will include leucocytes as well as red cells). 6. Correction for coincidence A correction for coincidence is made by using regression analysis with analysis of variance to check for non-linearity of the regression; the data points for the analysis are derived from the two sets of four secondary dilutions (Figure 1). The y values are summated counts at each level of secondary dilution and the x values are given in an arbitrary scale on which the most concentrated secondary dilution ( ml) is given the value of 1.0. Secondary dilution (ml) x value oooo If the anaiysis of variance fails to show evidence of non-linearity, the intercept of the regression line represents the coincidence corrected count (Lewis et al. 1989). This count gives the number of red cells aspirated through the orifice tube at the dilution level of the most concentrated secondary dilution ( ml). Divide this result by three because of the three replicate counts. Express the results as counts per ml of secondary dilution knowing the volume of secondary dilution aspirated through the orifice. Finally, multiply this count by to obtain the RBC. Summated count i Coincidence corrected count I I I I o Arbitrary scale Figure 1. Principle of coincidence correction using summation of replicated counts on two sets of four secondary dilutions. -X

6 136 ICSH Expert Panel on Cytometry METHOD FOR WHITE CELL COUNT The principles described above also apply for white cell counting. Make two dilutions, each of 0.08 ml blood plus 20 ml diluent. Prior to counting add the lytic agent. Count the white cells as soon as the red cell stroma has been reduced to a residue which does not contribute to the white cell count but before the ability to count the white cells has been affected. The same orifice can be used as for red cell counting. Validation of thresholds This follows the same principle as for RBC with the lower counting threshold placed between the noise from red cell stroma and leucocyte signals. Coincidence correction This is performed as for red cell counting save that four primary dilutions of whole blood in diluent are required: ml ml ml ml Analyse the data as explained above; the intercept of the regression line represents the coincidence corrected count for the most concentrated primary dilution ( ml). Correct for replication and for volume aspirated through the orifice as explained above. Also correct for dilution by the lytic agent. Finally, multiply this count by 251 to obtain the WBC. ANALYSIS OF ERROR The maximum permissible bias is 2.0% for red cell counts and 4.0% for white cell counts. Potential sources of error are listed below and apply to both red and white cell counting. Sampling error This error arises if the measured sample is not representative of the original whole blood specimen. Factors to be considered include poor mixing, haemorheological effects in extracting the sample and the inaccuracy in dilution. Transport error Errors due to transport occur if each and every cell is not counted in a known volume of diluted blood. Such errors are caused by sedimentation, cell loss and/or recirculation, as well as the inaccuracy of the displaced volume. Counting error Errors can arise if the final count is distorted because of improper discrimination, spurious counts or inaccurate correction for coincidence loss. The imprecision of

7 Reference method for enumeration of erythrocytes and leucocytes 137 the reference cell count should be no greater than 1%. Potential sources of error include extracting the sample, dilution, the imprecision of the displaced volume and the statistical limitation of the number of cells counted. References ICSH (1980) Recommendation for reference method for determination by centrifugation of packed cell volume of blood. J. Clzn. Pathol. 33, 1-2 ICSH (1987) Recommendation for reference method for haemoglobinometry in human blood (ICSH) Standard (1986) and specification for international haemiglobincyanide reference preparation (Third Edition). Clin. lab. Haemat. 9, ICSH (1988) The assignment of values to fresh blood used for calibrating automated blood cell counters. Clin. lab. Haemat. 10, ICSH (1991) Rules and Operating Procedures, p. 8, ICSH Secretariat, Leuven, Belgium. ICSH (1993) Recommendations of the International Council for Standardization in Haematology for Ethylenediaminetetraacetic Acid Anticoagulation of Blood for Blood Cell Counting and Sizing. Am. J. Clin. Pathol. 100, LEWIS S.M., ENGLAND J.M. & KUBOTA F. (1989) Coincidence correction in red blood cell counting. Phys. Med. Bid. 34, NCCLS (1993) Methods for the Erythrocyte Sedimentation Rate Test. Third edition. Candidate Standard for Approval. NCCLS Document H2-A3

8 138 ICSH Expert Panel on Cytometry Appendix 1. CALIBRATION OF VOLUMETRIC PIPETTES The pipette is filled to the calibration mark with distilled water, which is then transferred to a weighed beaker in accordancc with the normal usage of the pipette. Care must be taken to avoid evaporation. The beaker is reweighed. To achieve imprecision < 1% on a 10 ml pipette requires weighing which is reliable to 0.1 mg. Thus, the analytic balance should be stable for measurement at 0.01 mg, i.e. at the 1 pg level. If this is too stringent, then at least a 5 pg level is required. The ambient temperature is noted. The volume of the pipette (in ml) is calculated by dividing the weight of the water (in g) by one of the following factors depending on temperature: Temperature ("C) Factor The calibration must be performed in duplicate for each pipette. In calibrating a 20 pl pipette by weighing, imprecision should not exceed 1 %. 2. CALIBRATION OF VOLUMETRIC FLASKS The pre-weighed flask is filled to the calibration mark with distilled water and re-weighed. The ambient temperature is noted and the calculation carried out as in section 1.