Quality counts: new parameters in blood cell counting

Authors


C. Briggs, Department of Haematology, University College London Hospital, 60 Whitfield Street, London W1T 4EU, UK. Tel.: +44 2073809882; Fax: +44 2073077329; E-mail: carolbriggs@hotmail.com

Summary

Recently several parameters have been introduced to the complete blood count such as nucleated red blood cells, immature granulocytes; immature reticulocyte fraction, immature platelet fraction and red cell fragments as well as new parameters for detection of functional iron deficiency. Leucocyte positional parameters, which may diagnose specific diseases (e.g. differentiate between abnormal lymphocytes in leukaemia and viral conditions and may also detect malarial infection) are now available. At this time they are only used for research; however, generally such parameters later become reportable. One manufacturer’s routine analyser allows measurement of cells by flow cytometry using monoclonal antibodies. Currently, there are no accredited external quality assessment schemes (EQAS) for these parameters. For a number of parameters, on some instruments, there is no internal quality control, which brings into question whether these parameters should be used for clinical decision making. Other more established parameters, such as mean platelet volume, red cell distribution width and the erythrocyte sedimentation rate do not have EQAS available. The UK National EQAS for General Haematology held a workshop earlier this year in 2008 to discuss these parameters. Participants were asked to provide a consensus opinion on which parameters are the most important for inclusion in future haematology EQAS.

Introduction

This review is based on feedback from a workshop entitled ‘Quality counts: New Parameters in Blood Cell Counting-the next steps’ run by the UK National External Quality Assessment Scheme for Haematology, UK NEQAS (H) on 14 February 2008. Invited participants, including UK representatives from each of the major haematology instrument manufacturers, discussed recently introduced new blood parameters, their clinical utility, the availability of internal and external quality control and whether laboratories actually report the results to the clinicians.

Automated blood cell counters are becoming increasingly sophisticated and generate more reportable parameters. Over the last few years several new parameters have been introduced to the routine complete blood count (CBC). Some abnormal cells that were previously only indicated to be present in the blood by the generation of an abnormal cell flag are now quantified. These include nucleated red blood cells (NRBC), immature granulocytes (IG), activated/abnormal lymphocytes and fragmented red cells (FRC).

New parameters are also available for the detection of anaemia of chronic disease or functional iron deficiency (FID); these have been shown to be useful in monitoring the availability of iron during treatment with erythropoietin (Thomas et al., 2006). The immature reticulocyte fraction (IRF) is the proportion of the earliest reticulocytes in peripheral blood and is used to demonstrate a response to treatment for anaemia and predict engraftment after stem cell transplants (Molina et al., 2007). The immature platelet fraction (IPF), available on one analyser (analogous to red cell reticulocytes) are the youngest platelets in the peripheral circulation and contain remnants of RNA, they have been reported to be useful in the differential diagnosis thrombocytopenia. The IPF is high when there is peripheral consumption/destruction of platelets in diseases such as idiopathic thrombocytopenia purpura (ITP) and thrombotic thrombocytopenia purpura (TTP) and low in cases of bone marrow failure (Briggs et al., 2006). It may also be used as a predictor of platelet recovery following haematopoietic progenitor cell transplantation, as the IPF rises before the platelet count recovers (Zucker et al., 2006). However, there are currently no accredited EQAA for these new parameters. Cells traditionally measured on a dedicated flow cytometre, such as lymphocyte subsets are now available on one manufacturer’s haematology instruments using monoclonal antibodies and flow cytometry principles. Table 1 shows the new parameters discussed in this review and their clinical utility. Additionally for some of the parameters, on a number of instruments, there is no internal quality control (IQC) available, which means that these parameters should not be used for clinical decision making.

Table 1.   Recently introduced haematology parameters and their clinical utility
Instrument and manufacturerParameterClinical utility
  1. *Nonreportable parameter, for research use only.

Abbott Sapphire
Beckman Coulter LH 750
Horiba Medical Pentra
Siemens Advia
Sysmex XE series
Nucleated red blood cell countAutomatic correction of WBC and lymphocyte counts where necessary, fewer manual microscopic counts.
Diagnosis of haematological diseases and damage to bone marrow environment
Horiba Medical Pentra*
Sysmex X series
Immature granulocyte countDiagnosis of infection and inflammatory states
Abbott Cell-Dyn & Sapphire
Beckman Coulter LH 750
Horiba Medical Pentra
Siemens Advia
Sysmex XE series
Immature reticulocyte fractionMonitoring of bone marrow regeneration post transplant or chemotherapy. Classification of anaemias and monitoring of treatment
Siemens Advia

Sysmex XE series


Abbott Sapphire


Beckman Coulter LH 750


Horiba Medical
Percentage hypochromic red cells, reticulocyte haemoglobin content
Percentage hypochromic red cells, reticulocyte haemoglobin concentration
Percentage hypochromic red cells*, reticulocyte haemoglobin content*, mean reticulocyte volume*
Mean reticulocyte volume*
Low haemoglobin density*
Red cell size factor*
Mean reticulocyte volume*
Functional iron deficiency
Assessment of the availability of iron for erythropoiesis
Sysmex XE seriesImmature platelet fractionDifferential diagnosis of thrombocytopenia, prediction of platelet recovery post transplant or chemotherapy
Siemens Advia*
Sysmex XE series*
Fragmented red cell countDiagnosis and monitoring of microangiopathies
Abbott Cell-Dyn & SapphireMonoclonal antibody applicationImmunophenotyping, replaces traditional flow cytometer for some protocols and low volume of samples
Beckman Coulter LH 750


Sysmex XE series
White cell volume, conductivity and scatter measurements*
High fluorescent lymphocytes*, NEUT-X*
Advanced flags for diagnosis of specific diseases which cause changes to white cell populations
Diagnosis and monitoring of bacterial or viral sepsis Diagnosis of Myelodysplastic Syndrome (in combination of anaemia)

It was recognized that development of an EQA material could be difficult for some of the parameters, particularly those based on white blood cell positional parameters, because different instrument methodologies for differentiating the leucocytes do not respond in the same way to fixed or partially fixed blood preparations. Short presentations were made by users of each major instrument technology about each parameter and participants divided into three groups (the instrument manufacturer’s representatives made up their own group) for discussion and to decide, which were the most important parameters for inclusion in the UK NEQAS (H) scheme. A representative of each group reported back to the workshop and a final consensus decision was agreed on the priority of parameters for the development of EQA schemes. During the discussion the erythrocyte sedimentation rate (ESR) and some more established CBC parameters currently not monitored by UK NEQAS (H) were identified, although not all laboratories reported these CBC parameters.

Nucleated red blood cells

Nucleated red blood cells or erythroblasts (ERB) are the precursors of peripheral red blood cells and, except during the immediate neonatal period are normally only found in bone marrow. Their presence in peripheral blood at all other times is abnormal and regarded as a reflection of extreme increases in erythropoietic activity or as a result of damage to the bone marrow environment such as myelofibrosis, leukaemia or cancer. The ability to perform precise and accurate NRBC counts over the entire concentration range in peripheral blood by automated haematology analysers offers considerable advantages to the routine laboratory and clinicians (Kratz et al., 2006). Enumeration of NRBC is not only important in identifying disease but also because their presence can have a direct effect on the accuracy of the white blood cell count (WBC) on some of the blood cell counters used. Because NRBC have nuclei they are commonly erroneously counted as white blood cells by some automated methods. The correct WBC was traditionally only obtained by the labour intensive and imprecise method of examination of a peripheral blood film. The NRBC were reported as the number per 100 white blood cells and subtraction of the number of NRBC from the total nucleated count gave the correct WBC. The morphological correction of the WBC can be inaccurate because if the nuclear size of an NRBC falls below the white blood cell threshold of the instrument, these cells are not included in the automated WBC in the first place. All sophisticated haematology analysers now provide an automated NRBC count using various methods.

The Abbott Sapphire (Abbott Park, IL, USA) instruments determine NRBC counts using both fluorescence and multi-angle polarized scatter separation. Red blood cells are rapidly lysed and the exposed NRBC nuclei are stained with propidium iodide and the resulting mixture is analysed. The 0° and 90° laser light scatter, coupled with fluorescence, allows the discrete separation of the NRBC population from the white blood cell clusters and negates the requirement for WBC correction.

The Horiba Medical (Montpellier, France) Pentra DX120 analyser offers quantification of NRBC (ERB). ERB are separated from leucocytes by volume and from the platelet background by their fluorescence using the stain Thiazole Orange. The WBC is then corrected for the presence of the ERB.

Beckman Coulter instruments (Hialeah, FL, USA) identify NRBC found in a signature position on the impedance size WBC histogram, just below the white blood cell size threshold (Figure 1). Using Volume, Conductivity and Scatter (VCS) technology they are also detected in a specified region in the differential data-plot (Figure 1). Based on the presence of these cell populations in both places the instrument will derive the number of NRBC from the lower channels of the WBC size histogram. The instrument corrects the WBC count for any interference caused by the NRBC, although not all WBC need to be corrected because if the nucleus of the NRBC is very small (<35 fL) it will not be included in the WBC. Therefore, only immature NRBC with a large nucleus are included in the WBC.

Figure 1.

 Nucleated red blood cells found in a signature position on the impedance size white blood cell count histogram and using Volume, Conductivity and Scatter technology in a specified region in the differential data-plot.

The white cell channels of the Siemens instruments (Siemens Diagnostics, Tarrytown, NY, USA) comprise the Basophil channel and the Peroxidase channel. In the Basophil channel red blood cells and platelets are lysed and the cytoplasm of all white cells is stripped, apart from basophils. NRBC are identified in the peroxidase negative area of the Peroxidase channel (Figure 2). In the Basophil channel, NRBC are identified by nuclear density and are located in the neutrophil/eosinophil area (Figure 2). The difference between the number of nuclei in the neutrophil/eosinophil area in the Basophil channel and the sum of the neutrophils and eosinophils in the Peroxidase channel represents the NRBC count. The white cell count and differential are corrected for the presence of NRBC. The NRBC counting method on the Beckman Coulter and Siemens’ instruments is not a direct measure of the cells and there is the possibility of other interfering substances in blood occupying the NRBC signature position.

Figure 2.

 The nucleated red blood cells count derived from the Peroxidase and Basophil channel on the Siemens haematology analysers.

The Sysmex XE-2100 (Kobe, Japan) analyser uses flow cytometry with a semiconductor red diode laser and a polymethine fluorescent dye. NRBC mode is selected on the analyser, mature red cells are completely lysed; white cell membranes are perforated but retain their original shape and NRBC are de-nucleated and shrunken. The dye stains intracytoplasmic organelles and the nucleus of the white cells quite strongly but staining of NRBC is comparatively weak. These different staining characteristics allow clear discrimination of the cell counts.

The automated NRBC count from all these instruments is now routinely reported to clinicians. However, manufacturer’s IQC material is not available for two of these instruments and in the UK there is no EQA scheme. An UK NEQAS (H) scheme should be developed with NRBC included in both the CBC scheme, to asses the accuracy of the corrected WBC and the automated leucocyte differential scheme to asses the accuracy of the corrected lymphocyte count. EQA for NRBC is available from the College of American Pathologists (CAP) proficiency testing programme as part of the automated differentials programme but only in two of the instrument group specific EQA materials. The use of instrument group specific material overcomes some of the problems associated with the use of stabilized blood in cellular counting but precludes any direct comparison across instrument methods, as each material type has a different matrix and usually different analyte levels.

Immature granulocytes

Quantitative analysis of IG is useful for clinical management of patients with haematological disease, malignancy or infection. Conventionally, IG are classified on the basis of cell morphology by the microscopical examination of a stained blood film as promyelocytes, myelocytes and metamyelocytes. However, the manual differential count is imprecise because of the small number of cells counted and interobserver variability (Dutcher, 1984; Cornbleet & Novak, 1994). Often small numbers of IG, particularly in leucopenic samples or when small percentages are present, may be missed in a standard 100-cell differential or routine film review. In addition, the results of manual differential counts may be influenced by differences in morphological identification and variable observer skills (Rumke,1979; Houwen, 2001). Promyelocytes, myelocytes and metamyelocytes are all included in the automated IG count but are not presently identified as separate classes of cells. The presence of IG may indicate sepsis, bacterial infection, inflammatory states, steroid therapy, cancer, trauma or myeloproliferative diseases. They are also present in the later stages of pregnancy. In these cases, there is often an increased neutrophil count; however, this myeloprolferative response varies considerably, particularly in newborn infants, the elderly and myelosuppressed patients where response is less dramatic. Therefore changes in the neutrophil count tend to be extremely variable and nonspecific and the neutrophil count can also be normal despite clinical evidence of an infection. Neutrophil morphological abnormalities and automated left shift flags are notoriously unreliable as specific diagnostic features. The presence of low numbers of IG is more reliably detected on automated haematology analysers than using manual microscopy. This is because of the high number of cells counted and an increase of IG (>2%) can be useful in identifying infection even when not suspected (Briggs et al., 2003). Some top of the range analysers do not quantitate IG and still rely on an abnormal white cell flag generated by the instrument to indicate their presence in the blood sample.

The Horiba Medical Pentra DX120 provides an IG count by utilizing the double differential matrix and the differential is produced using two channels. In one, light absorbance (after staining of the cells) and impedance are measured and in the other basophils are differentiated from other white cells by impedance measurement following cytoplasmic stripping of the leucocytes. Cells are plotted on the double differential matrix with light absorbance vs. volume. In addition to the IG other abnormal cell populations are enumerated including atypical lymphocytes, immature lymphocytes and immature monocytes. The clinical utility of these other immature cells has not been widely studied or reported.

On the Sysmex XE-2100 IG are analysed from the neutrophil population in the differential channel. Red cells are lysed and white cells ‘permeabilised’. A polymethine fluorescent dye stains nucleic acid and flow cytometry is used to detect side-scattered light and fluorescence. IG have higher fluorescence than neutrophils and can be clearly separated. On both instruments, the IG count is reported when the instrument performs a routine CBC with differential.

Only one of these instrument methods (Sysmex) has manufacturer’s IQC material available, but again there is no current EQA scheme is available. The CAP scheme offers this parameter as part of the automated differentials proficiency testing programme but only for the Sysmex instrument group.

Immature reticulocyte fraction

All sophisticated haematology analysers provide a reticulocyte count with the IRF. Few modern laboratories continue to count reticulocytes manually as it is imprecise and coefficients of variation between operators have been reported to be as high as 50% (Peebles, 1981; Schimenti et al., 1992). Automated counts are more precise because of the high number of cells counted and have improved the quality of the results allowing reliable flagging of reticulocytopenia. An assessment of reticulocyte maturation is important for diagnosing the cause of anaemia and assessing the degree of effective erythropoiesis.

The term IRF was introduced to indicate the least mature erythrocytes which contain the most RNA. The IRF can be defined as the ratio of young, immature reticulocytes to the total number of reticulocytes. Immature reticulocytes are released into the peripheral circulation during periods of intense erythropoietic stimulation such as haemorrhage, certain anaemias or in response to therapy to stimulate bone marrow production. The IRF has numerous clinical applications as in many clinical situations the IRF increases before the total reticulocyte count and it can be used to monitor bone marrow or stem cell regeneration post-transplant or chemotherapy. The first response after bone marrow ablation is an increase in the IRF, which precedes the increase in the reticulocyte count by several days (Davies, 1996;Grotto et al.,1999). In states of dyserythropoiesis, such as acute myeloid leukaemia or myelodysplastic syndromes (MDS) and megaloblastic anaemia the IRF may be raised with a normal or reduced reticulocyte count (Daliphard et al., 1993; Watanabe et al.,1994). In other anaemias where there is reduced erythropoiesis, such as iron deficiency or anaemia of chronic disease, the total reticulocyte count is reduced but the IRF is normal. After treatment for a nutritional anaemia (B12, folate or iron deficiency) the increase in IRF occurs several days before an increase in the reticulocyte count.

Most automated reticulocyte count methods depend on the ability of stains to bind RNA in the reticulocyte; however, different automated methods have varying sensitivities to the dyes used and use different methods to identify positive cells.

The stained cells are separated from nonstained cells (mature red cells) and cell populations where size or level of staining, place them outside the reticulocyte area (i.e. platelets, NRBC and white cells). The fluorescent intensity of each reticulocyte is directly proportional to the quantity of RNA inside the cell, allowing the analyser to differentiate reticulocyte subpopulations. Light absorbance, light scatter or fluorescence is used to analyse the cells.

Abbott analysers use a patented fluorescent reagent to stain reticulocyte RNA and a combination of fluorescence and narrow angle laser light scatter to detect the cells and provide a reticulocyte and IRF count (Kim et al., 1997).

The Horiba Medical Pentra also employs fluorescent technology for the measurement of reticulocytes, using Thiazole Orange as the RNA stain. Fluorescence is measured in a flow cell, giving three types of information on the size of the cell size, measured by resistivity and cell content measured by forward scattered light and the fluorescence signal. Mature red blood cells, without RNA, show little or no fluorescent signal and so are clearly separated from reticulocytes.

Beckman Coulter instruments use new methylene blue stain which is a nonfluorescent stain, to precipitate the reticulocyte RNA. After the cells have been stained, VCS technology is used to analyse and count the cells. Volume measures the size of the cell, conductivity provides information about the internal structure of the cell and scatter provides information on the cell surface characteristics and cell granularity. Reticulocytes, containing RNA, scatter more light than mature red cells and the more immature reticulocytes are identified as being larger and have the greatest light scatter properties. The VCS data are plotted three-dimensionally on a scatter plot.

On Siemens instruments, red blood cells are sphered, fixed then analysed using a laser. Monochromatic light scatter measures the size and the refractive index of the cells. Reticulocyte analysis is undertaken in the same manner as red cells with the addition of Oxazine 750 (a nucleic acid dye) which allows differentiation of the reticulocytes from mature red cells. RNA staining intensity is measured to identify the youngest reticulocytes.

Sysmex instruments also use a fluorescent polymethine dye and a semi-conductor laser to identify reticulocytes. The dye stains DNA and RNA simultaneously, but as the concentration of DNA in white blood cells far exceeds that of RNA in reticulocytes, the cells are separated by their differing fluorescence. Reticulocytes are separated into three categories, the most immature, moderately immature and mature reticulocytes. The most immature and moderately immature categories comprise the IRF.

Different automated methods use different dyes with varying sensitivities to identify reticulocytes therefore reference ranges for reticulocytes and IRF should be instrument specific, as they vary considerably between different makes of instruments reference (Briggs et al., 2001; Buttarello et al., 2002). All instruments counting reticulocytes and reporting the IRF have IQC available for the counts.

Despite the reported clinical utility of the IRF many laboratories still do not report the parameter to the clinician. This is probably because the automated methods are not standardized and reference ranges are method/instrument specific. However, as some laboratories do report the IRF and it may be being used for clinical decision making, there is a need for the IRF to be included in the existing reticulocyte EQA scheme; however, the use of partially fixed or fully fixed blood necessary for EQA material to be stable, rather than fresh blood, causes problems on some instruments when measuring reticulocytes or reticulocyte parameters.

Functional iron deficiency

Because of the long life span of erythrocytes measurements of traditional red cell indices are not sensitive indicators of early iron deficient erythropoiesis. The reticulocyte count provides a quantitative measure of erythropoiesis but no information on the quality of erythropoiesis. There are now several modern markers available which can help to identify varying degrees of iron deficiency and monitor treatment response. The condition of ‘FID’ has attracted attention recently. FID occurs when reticuloendothelial iron stores are normal, or even high, but the iron is not delivered to the bone marrow and so is unavailable for erythropoiesis. This condition is most often seen in patients with chronic inflammation, cancer or those undergoing long-term renal dialysis. The mechanism of FID is not fully understood but it may be due to the inappropriate production of hepcidin, an iron regulatory hormone produced in the liver (Roy et al., 2007). This is an acute phase reactant which prevents iron absorption from the gastrointestinal tract, its release from hepatic stores and its recycling within the reticuloendothelial system (Nemeth & Ganz, 2006). It also known that in this state raised cytokines disturb the expression of transferrin receptor and causes an increase in plasma ferritin (Fitzsimons & Brock, 2001) which makes the interpretation of the biochemical markers of iron difficult. The ideal test for assessing iron available for erythropoiesis would be one that measures the availability of iron at the point of haemoglobin synthesis in the red cell precursors or reticulocytes.

With the development of flow cell haematology analysers it is now possible to measure the volume and haemoglobin content of reticulocytes. This parameter is available on the Siemens and Sysmex instruments. The parameter from Siemens is termed CHr (mean reticulocyte haemoglobin content) and from Sysmex the Ret-He (the reticulocyte haemoglobin equivalent). CHr is measured in the stained reticulocytes using two angle light scatter and Ret-He, is a measure of the forward scatter of stained reticulocytes and has a curve linear relationship with CHr (Thomas et al., 2005). The reference values in health for both parameters are the same, with the mean value being 30.8 pg, the same value in males and females and the lower limit of normal is 28 pg (Mast et al., 2002; Thomas et al., 2005). The reticulocyte haemoglobin content provides an indirect measure of the functional iron available for new red blood cell production over the previous 3–4 days and it also provides an early measure of the response to iron therapy, increasing within 2–4 days of the initiation of intravenous iron therapy (Brugnara et al., 1994; Fishbane et al., 1997). It is an early indicator of iron-restricted erythropoiesis in patients receiving erythropoietin therapy; these patients may have FID and respond to iron therapy even with very high serum ferritin values (Kopelman et al., 2007). Studies examining the use of reticulocyte haemoglobin to manage intravenous iron therapy in this group of patients have demonstrated that a reticulocyte haemoglobin of <28 pg more accurately predicts FID when compared with ferritin and transferrin saturation as well as it reduces the intravenous iron exposure in these patients (Mittman et al., 1997; Fishbane et al., 2001;Chuang et al., 2003; Tsuchiya et al., 2003).

More recently other instruments have developed parameters supposedly equivalent to reticulocyte haemoglobin. However, these are still research parameters and IQC is not available. Abbott instruments have a measured parameter for reticulocyte haemoglobin as well as a measure of reticulocyte volume (MRV) and the Horiba Medical Pentra 120 and Beckman Coulter LH 750 both report MRV. Few studies are available on the clinical utility of the MRV; however, it has been reported that it increases after treatment with iron in patients with iron deficiency anaemia and decreases with the development of iron deficient erythropoiesis (Brugnara, 1998). The MRV decreases and reticulocytes are smaller than mature red cells following treatment with vitamin B12 or folate (D’Onofrio et al., 1995). Reticulocyte haemoglobin and reticulocyte volume may have similar clinical utility, but the MRV produced by different instruments lacks standardization, which means numeric results from different manufacturers are not comparable and there is no quality control material available. The reticulocyte haemoglobin has limited clinical use as it is only available on two instruments as a reportable parameter with IQC available and there is no EQA available for any of these parameters.

An earlier parameter used in the assessment of FID is the percentage hypochromic red cells, defined as cells with intracellular haemoglobin of <28 g/dl (Horl et al., 1996). In the healthy population, the percentage of hypochromic red cells does not exceed 2.5% and values greater than this are indicative of iron deficient erythropoiesis (Schaefer & Schaefer, 1995). Until very recently this parameter was only available on Siemens haematology instruments. The instrument assesses erythrocytes on a cell by cell basis so can therefore detect and quantify subpopulations of cells. Percentage hypochromic red cells is the concentration of haemoglobin in individual cells rather than the mean such as mean cell haemoglobin (MCH) or mean cell haemoglobin concentration (MCHC). It is a more sensitive marker because small changes in the number of red cells with inadequate haemoglobin can be measured before there is any change in the MCHC.

Some sophisticated Sysmex and Abbott instruments can report this parameter but it has been argued that, as mature red cells have a longer lifespan, the percentage hypochromic red cells integrates information from over too long a period and may be less sensitive than reticulocyte haemoglobin in diagnosing FID or monitoring anaemia treatment (Briggs et al., 2001). Reticulocyte haemoglobin may be superior to percentage hypochromic red cells in detecting iron deficiency in haemodialysis patients, but reticulocyte haemoglobin does not change during treatment with erythropoietin and iron as expected (Cullen et al., 1999). Some studies report that the percentage of hypochromic red cells is more sensitive for the measurement of FID and that reticulocyte haemoglobin provides better evidence of a response to iron therapy (Chuang et al., 2003).

Other manufacturer’s instruments report parameters claimed to be equivalent to percentage hypochromic red cells, Abbott have an equivalent parameter and Beckman Coulter have their own indices, low haemoglobin density (LHD%) and red blood cell size factor (RSf). These are presently not reportable and are for research use only. LHD% is derived from the sigmoid transformation of the MCHC from the LH series instruments. LHD% has been proposed as a parameter to assess the available iron stores for erythropoiesis. The clinical utility of this parameter is in the diagnosis of FID and the assessment of an individual’s response to erythropoietin therapy. Reported correlation with the established parameters (Siemen’s CHr and percentage hypochromic red cells) supports this (Zini et al., 2006). Rsf is a parameter derived from the combination of the volume of the red blood cell and reticulocyte and good correlation to CHr was found in health and conditions of restrictive erythropoiesis (Urrechaga, 2008). There is currently no IQC available for these two parameters.

The availability of reportable, rather than research, parameters in use for the diagnosis of FID is limited to only two manufacturer’s instruments; however, as these parameters are used for clinical purposes there should be EQA material available. The development of any EQA material will be difficult because of the many different parameters used to detect FID and the different methods of measurement. Those instruments using red cell parameters for detecting FID would need EQA using the CBC scheme and those using parameters derived form reticulocytes would use material for the reticulocyte scheme. The use of partially fixed stabilized blood affects the way some instruments identify reticulocytes and associated indices and providing fresh blood has inherent operational difficulties. Ideally the EQA scheme should provide survey samples that simulate, as closely as possible, the relevant properties of the samples on which the diagnostic procedures are intended to be used (BS EN 14136, 2004).

The immature platelet fraction

Newly released platelets are larger and more reactive than mature platelets. They contain RNA and are suggested to be the platelet analogue of the red cell reticulocyte and are termed ‘reticulated platelets’ (Ingram & Coopersmith, 1969). The number of reticulated platelets reflects the rate of thrombopoiesis, increasing when platelet production rises and decreasing when production falls, thus mirroring red cell reticulocytes and the erythropoietic response. The platelet RNA can be measured by a variety of dyes and so reticulated platelets can, therefore, potentially, be quantified by flow cytometry using any fluorescent dye that binds RNA; however, dye uptake can be slow and there maybe nonspecific binding to platelet internal granules. The reticulated platelet is distinguished from the mature platelet that has not taken up the dye. There is significant variation in the published reference ranges in health for this parameter using different flow cytometric methods and even between laboratories using the same methodology (Harrison, 1997; Robinson et al., 1998).

A number of clinical papers on reticulated platelet analysis have been published previously (Ault et al., 1992; Rinder et al., 1993; Richards & Baglin, 1995; O’Malley et al., 1996) and they have clearly shown that under conditions of thrombocytopenia, platelet RNA content correlates directly with megakaryocyte activity. Patients with low megakaryocyte activity have no RNA elevation in their platelets while those with increased megakaryocyte activity have significantly elevated RNA-stained platelets. This offers the ability to determine whether thrombocytopenia is due to marrow failure or to increased peripheral platelet destruction/loss, thus avoiding the need for bone marrow examination.

A new automated method to quantify reticulated platelets when the instrument is run in the red cell reticulocyte mode, expressed as the IPF, has been developed on the Sysmex XE-2100 and XE-5000 blood cell counters. The normal range has been quoted as 1.1–6.1% (Briggs et al., 2004) with other workers finding similar ranges (Zucker et al., 2006; Cho et al., 2007). The flow cytometric IPF determination uses a proprietary fluorescent dye containing polymethine and oxazine. These two dyes penetrate the cell membrane staining the RNA in the red cell and platelet reticulocytes and the stained cells are then passed through a semiconductor diode laser beam. The resulting forward scatter light (cell volume) and fluorescence intensity (RNA content) are measured. Figure 3 illustrates optical (fluorescence) platelet scattergrams with forward scattered light on the y-axis and fluorescence on the x-axis. A computer algorithm discriminates between the mature and IPF by intensity of forward scattered light and fluorescence. Mature platelets appear as blue dots and the immature platelets are displayed as green dots, the latter constituting the IPF parameter. Figure 3 illustrates the optical platelet scattergram from a healthy individual and one from a patient with an exceptionally high count.

Figure 3.

 Sysmex XE optical platelet scattergrams from a healthy individual with a normal immature platelet fraction (IPF) and a patient with a high IPF. Mature platelets appear as blue dots, green dots represent the IPF with increased cell volume and higher fluorescence intensity compared with mature platelets.

The IPF is raised in patients with peripheral consumption/destruction of platelets (ITP and TTP) and is normal or low in patients with marrow failure (Briggs et al., 2004). Following peripheral blood stem cell transplant the IPF rises 1–2 days prior to the platelet count increasing (Briggs et al., 2006; Zucker et al., 2006). The rise in IPF occurs 4–4.5 days before the rise in platelet count following allogenic bone marrow transplant (Richards et al., 1996; Chaoui et al., 2005; Zucker et al., 2006). With the ability to predict platelet regeneration within a few days of an increase in the IPF, it should be possible to reduce prophylactic platelet transfusions in patients undergoing peripheral blood stem cell transplants (Briggs et al., 2006; Zucker et al., 2006; Takami et al., 2007).

The IPF is only available on one manufacturer’s instruments and there is IQC available but no EQA scheme. Ideally all results reported should be subject to EQA (WHO/LAB., 1998) an EQA scheme for a parameter that has such limited availability should be a low priority.

Fragmented red cells (Schistocytes)

The extent of red blood cell fragmentation in peripheral blood is useful in the diagnosis and follow-up in certain disorders. FRC are formed as a consequence of mechanical damage and are associated with heart and large vessel disorders such as synthetic heart valves, endocarditis and aortic aneurysm.

They are also found in microangiopathies such as haemolytic ureamic syndrome, thrombotic thrombocytpenia purpura (TTP), disseminated carcinoma, transplant associated thrombotic microangiopathy (TMA), cytotoxic chemotherapy, disseminated intravascular coagulation (DIC), immune disorders and infection. Red cell fragmentation is also found in march heamoglobinuria. Diagnoses of microangiopathies are of extreme importance as they are life threatening diseases and identification and quantification of FRCs is an important diagnostic criterion (Burns, Lou & Pathak, 2004). Manual microscopical counting of FRCs is both imprecise and subjective. There are differences in the definition of schistocytes between different laboratories and between individuals. Schistocytes have been defined as red blood cells with the shape of crescents, helmets, triangles and/or microspherocytes (Foerster, 1999). The Sysmex XE and Siemens Advia instruments both provide an automated FRC count currently used as a research parameter, which at present is only intended for laboratory use. Both instruments use cellular size as the detection method, cell shape is not a factor but despite this, significant correlation between automated schistocyte counting, by both instruments, and the manual method has been reported. There is, however, a tendency for overestimation by the automated methods (Jiang et al., 2001; Lesesve et al., 2004). Advia instruments use the red cell/platelet channel to quantify schistocytes with them being identified as particles with a volume smaller than 30 fL and with a refractive index <1.40 (Lesesve et al., 2004) to differentiate them from large platelets (Figure 4). Sysmex instruments use the reticulocyte channel to quantify schistocytes with them being the smallest events in the red cell area with a low RNA content (Jiang et al., 2001). Due to the greater number of cells counted by automated methods compared with manual counting the precision should be better (Banno et al., 2002). The automated methods demonstrate 100% sensitivity but the specificity is low at 20% due to false positives in the presence of anisopoikilocytosis and nonschistocyte fragments (Lesesve et al., 2004). In the laboratory, this should mean any positive counts should have a microscopical examination of a blood film to confirm the presence of schistocytes.

Figure 4.

 Platelet scattergrams from the Siemens Advia showing the position of red cell fragments.

A normal reference range for the schistocyte count is as 0.03–0.58% (Jiang et al., 2001) and TMA are defined as having a schistocyte count of >4% (Ruutu et al., 2007). The use of the IPF and the FRC count together may be helpful for the early differential diagnosis of ITP and microangiopathies. Serial counting in TTP may correlate with changes in serum lactate dehydrogenase levels, when the red cell isoenzyme is assessed and acts as a marker for disease activity.

There are neither IQC nor EQA available for either of the two automated counts. Because the FRC count is a research parameter only, with limited availability there is no current requirement for the development of an EQA scheme.

The use of monoclonal antibodies in automated haematology cell counters

The combination of laser light scatter and fluorescence technologies has allowed the development of monoclonal antibody (mAb) applications on some haematology instruments manufactured by Abbott. The Cell Dyn 4000 and Sapphire instruments have dedicated processing options for CD61 immunoplatelet counts (Gill et al., 2000) and CD3/4/8 lymphocyte subsets (Marshall et al., 2000). The automated immunoplatelet count uses reagent tubes coated with anti-CD61, the same antibody used in the international reference method for platelet counting (International Council for Standardization in Haematology (ICSH), 2001). This immunological method is not designed to be used routinely, but only for patients with abnormal platelets or on some patients with very low counts. The Abbott immunological method was found to be the most accurate on samples with a platelet count of 20 × 109/l or less, when compared with the ICSH reference method and results better than for any other routine optical or impedance counting methods and indeed the traditional counting methods on the same Cell Dyn analysers (Segal et al. 2005).

For the immunological CD3/4, and CD3/8 T-cell count the test is pre selected at the analyser data management station and is performed as part of the overall CBC. The same sample is used to perform both the blood cell count and immunophenotyping. The procedure uses sequential rack positions with the first occupied by the ethylenediaminetetraacetic acid (EDTA) patient blood sample, the second and third by tubes containing CD3/4 and CD3/8 antibody mixtures. The analyser takes fixed volumes of blood from the sample tube and injects them into the reaction tubes. After mixing and a brief incubation period an aliquot of the mixture is diluted and analysed in the flow cell. Simultaneous measurements of optical scatter and fluorescence are made and the results automatically reported. Using the mAb application causes the throughput for CBC on the instrument to slow down considerably.

The process can be adapted for many different leucocyte antigens by using empty reaction tubes and adding the appropriate antibodies e.g. other lymphoid cells, T cells, B cells, natural killer cells and myeloid markers (Molero et al. 2005, Johannessen et al. 2006). The instrument can be used to measure foeto-maternal haemorrhage, a rapid and cost effective alternative to the traditional Kleihauer–Betke technique which has been traditionally the only option available to most laboratories. The method requires a preincubation of prepared supernatant with FITC conjugated BRAD – three monoclonal anti-D for 30 min and it is then processed on the Cell-Dyn Sapphire (Little et al., 2005). RhD positive cell populations are isolated using detection using fluorescence and 7° laser light scatter, Figure 5.

Figure 5.

 Example of RhD+ in cells foeto-maternal haemorrhage measured by the Abbott Cell Dyn instruments. Reproduced courtesy of Clinical and Laboratory Haematology (Little et al. 2005).

UK NEQAS provides EQA for leucocyte immunophenotyping, CD4/8 counting and foeto-maternal haemorrhage via the UK NEQAS for Leucocyte Immunophenotyping and UK NEQAS for Feto-Maternal Haemorrhage schemes. Laboratories using Cell Dyn instruments for immunological cell typing should, as for users of traditional flow cytometers, subscribe to these schemes where appropriate.

White blood cell positional parameters

Recently, VCS and light scatter white cell positional parameters have been reported to demonstrate clinical utility in the diagnosis of some diseases. Abnormal cell populations that have previously only triggered an abnormal flag on some instruments can now be quantified.

Numerical data (coordinates) generated by the analyser to produce the VCS 3D cube are now available as research parameters on the LH Series Beckman Coulter instruments. These coordinates for neutrophils, lymphocytes, monocytes and eosinophils provide 24 new parameters (mean position and standard deviation, SD, for each cell type). Normal cells of each white cell type will have characteristic volume (mean and SD), conductivity (mean and SD) and light scatter (mean and SD). Any deviations away from normal values are thought to reflect differences in cellular size and complexity for a particular cell type and may be indicative of a potential disease process or specific morphological characteristics. It may be possible to utilize these parameters as advanced flags for specific diseases or conditions in certain clinical circumstances.

It has been reported that positional parameters can be used to distinguish between different lymphoproliferative disorders and viral infections (Silva et al., 2006). The mean lymphocyte volume has been found to be low in cases of classic chronic lymphocytic leukaemia and increased in the cases with abnormal or reactive lymphocytes (Figure 6). This may help to guide and speed up diagnosis and possible follow-up investigations, for example in referring the sample for flow cytometry analysis. Mean neutrophil conductivity and mean neutrophil scatter can be used in the detection of dysplastic neutrophils (Miguel et al., 2007) with low values for these parameters correlating with neutrophil hypogranulation.

Figure 6.

 Characterization of the white cell populations with three-dimensional analysis using Volume, Conductivity and Scatter. The screen shows the different values in the volume and SD volume for the lymphocytes for a normal sample, chronic lymphatic leukaemia (CLL) and viral disease.

Sepsis and associated neutrophil left shift can be identified by a change from normal values of the mean and SD neutrophil volume and these parameters could be used as an indicator of acute bacterial infections (Chaves, Tierno & Xu, 2005).

Briggs et al. (2006) suggested the use of the SD of the volume for lymphocytes and monocytes to flag for the possible presence of malarial parasites. Due to the presence of reactive lymphocytes and histiocytic monocytes in infected patients these cells are increased in size and therefore have increased volumes and SD of the volumes. The combination of these changes has allowed the development of an algorithm for this (Malaria Factor). Using a cut-off value for the Malaria Factor of greater than 3.7 as an indicator of malaria infection the specificity was 94% and sensitivity 98%.

Several groups have reported using depolarized laser light on Abbott instruments for the detection of malaria (Scott et al., 2002Wever et al., 2002). The Haemozoin produced by malarial parasites as a result of haemoglobin breakdown is ingested by neutrophils and causes them to depolarize laser light when normal neutrophils do not. Under these circumstances an atypical depolarization flag is generated which may indicate the presence of malaria infection. Sysmex has also developed two new parameters on the basis of their abnormal positions in the differential scattergram.

The first is the high fluorescent lymphocytes (HFL). These cells appear in a high fluorescence area of the lymphocytes in the differential channel and are flagged as atypical lymphocytes by the instruments. They are detected in 10% of all differential counts. Identification of the cells in the HFL population was carried out using immunophenotyping flow cytometry methods and they were found to be activated B-lymphocytes or plasma cells (Linssen et al., 2007). They are now quantified on the latest Sysmex analysers and may have clinical utility in the diagnosis and monitoring of sepsis because of bacterial or viral infection.

Sysmex NEUT-X is the mean value for side scatter diffraction of the neutrophil population; it represents the internal structure of the neutrophils. It was found that this value was low in cases of MDS and may be used as a potential indicator of the presence of the disease. A low NEUT-X strongly correlates with hypogranularity in the neutrophils and when considered in conjunction with anaemia is highly suggestive of MDS. For many patients with MDS the only abnormal finding in the blood count is anaemia and in many laboratories this would not trigger a blood film review and the diagnosis would be missed. The use of anaemia in association with a low NEUT-X increased the number of patients with MDS triggering a blood film review from 67% to 96% with only 2% false positive results resulting in unnecessary film review (Cymbalista, 2007).

These new positional parameters provide numerical values to the changes which can be seen in the instruments scatterplots by the experienced analyser operator and this may allow for further development of specific disease flags and new abnormal cell counts.

Reference ranges need to be established locally and sites wishing to utilize these parameters need to be aware that their analysers must be fully optimized and standardized before use and that certain instrument component changes may mean that subsequent optimization checks are needed. Laboratories with more than one instrument of the same type need to standardize settings between instruments as well on a regular basis.

At the present time, there is no IQC or EQA scheme for these new positional parameters and as they are currently only for research use there is no urgent need for EQA. However, it would be useful if IQC was available for those laboratories wishing to utilize what may be potentially very useful new parameters.

Established laboratory parameters without EQA schemes

Mean platelet volume

Circulating platelets vary in both size and functional activity. Larger platelets are probably younger, more reactive and produce more thrombogenic factors (Thompson et al.,1984). Automated blood cell counters provide a platelet count and derived indices relating to the size of platelets. These parameters produce clinically useful information when methodological problems involved in obtaining the results are taken into consideration. Size-related parameters are derived from the impedance platelet size distribution curve (Figure 7). Mean platelet volume (MPV) is calculated by dividing the platelet-crit (PCT), by the number of platelets (this is the same calculation as for the mean red cell volume (MCV) namely dividing haematocrit by the red cell count) and therefore PCT is analogous to the red cell haematocrit. On instruments that count platelets using optical light scatter, the MPV is derived from the modal platelet size. Nearly all analysers report MPV and some also report the platelet size distribution width (PDW). The PDW is the width of the size distribution curve in femtolitre (fL) at the 20% level of the peak (Figure 7). The platelet large cell ratio (P-LCR), reported by some Sysmex analysers, is the number of cells falling above the 12 fL threshold divided by the total number of platelets (Figure 7).

Figure 7.

 Typical impedance platelet size distribution from an automated haematology analyser. PLT, platelets; PL, lower discrimination for platelet size distribution; PDW, platelet distribution width; P-LCR, platelets-large cell ratio; PU, upper discrimination for platelet size distribution; fL, femtolitre.

Evidence that large platelets are haemostatically more active than smaller platelets is derived from in vitro studies (Thompson et al., 1982) and suggests that large platelets may be more important functionally than smaller platelets. An increase in MPV has been observed in patients at risk of, and following, myocardial infarction (Khandekar et al., 2006) and cerebral infarction (D’Erasmo et al., 1990). Some inherited congenital macrothrombocytopenias are readily diagnosed by the measurement of MPV such as Bernard-Soulier syndrome. A low MPV has been reported in thrombocytopenic patients with marrow disease (Bowles et al., 2005) and a high P-LCR or PDW may indicate peripheral immune destruction of platelets (Kaito et al., 2005). Although the derived platelet parameters must be interpreted carefully, an inverse relationship normally exists between MPV and the platelet count, which contributes to the maintenance of haemostatic function.

Derived platelet parameters are highly specific to the individual technologies, with different analysers having different normal ranges for the MPV and are influenced by external factors such as the anticoagulant used and delay time from sampling to analysis (e.g. EDTA-induced swelling). If MPV is to be reliably measured, then the potential influence of EDTA anticoagulant on the MPV must be controlled, either by using an alternative anticoagulant or standardizing the time delay between sampling and analysis. With impedance counting the MPV increases over time as platelets swell in EDTA, with increases of 7.9% within 30 min having been reported and an overall increase of 13.4% over 24 h but with the majority of this increase occurring in the first 6 h (Bowles et al., 2005). In severely thrombocytopenic samples it may not be possible to collect sufficient data (platelet pulses) to calculate platelet indices and therefore they cannot be reported. In addition, for platelet counts below 50 × 109/l, repeated measurements on the same sample show the CV% of platelet indices may be more than three times greater than normal (Ogura et al., 1995). When MPV is measured by optical light scatter systems, derived from the modal platelet size, the MPV decreases over time, possibly due to the dilution of cytoplasmic contents leading to a decrease in light scattering abilities (Patterson, 1997). An equation allowing for the correction of MPV over time from venepuncture has been published (Trowbridge et al., 1985) but this is not widely used due to unpredictable behaviour in individual samples in the time taken to equilibrate.

Many laboratories do not report the MPV to clinicians and this is probably due to the lack of standardization and the dependency of the results on the age of the sample and on the measurement method. All manufacturers provide IQC material with assigned values for the MPV but there is currently no EQA scheme available. Results are currently generated by instruments when the CBC EQA is analysed so it should be quite simple to start to collect the result from the current EQA scheme.

Red cell distribution width

The red cell distribution width (RDW) is derived from pulse height analysis and is the width of the red cell size distribution curve in fL at the 20% level of the peak (the same calculation as for PDW). The RDW can also be expressed as the CV % of the measurements of the red cell volume. It is a quantitative measurement of variation in red cell size and is equivalent to anisocytosis seen on the examination of a stained blood film. Widely different normal reference ranges have been quoted for the RDW-CV (Rowan, 1983; Roberts & Badawi, 1985) and reference ranges should be instrument specific. The RDW can be used as a guide for the differential diagnosis of anaemia (Lin et al., 1992); it is usually normal in thalassaemia trait and increased in iron deficiency anaemia. In megaloblastic anaemia RDW is increased but is usually normal in macrocytosis due to other causes. Where microcytosis and macrocytosis exist within the same sample the two abnormalities may cancel each other out and cause a normal MCV, however, the resulting high RDW will identify the error.

Not all laboratories report the RDW to the clinicians but there are manufacturers IQC materials with assigned values available for all instruments, though no EQA scheme. Like the MPV results will be produced by instruments when the CBC EQA is analysed so it should be possible to collect the results form the current EQA material. This material would need to be validated for the performance assessment of the RDW (and the MPV) before these analytes are added to the CBC EQA profile.

Erythrocyte sedimentation rate

The ESR is a nonspecific screening test used to detect the acute phase inflammatory response. Some laboratories now prefer a sensitive C reactive protein test as a more useful clinical alternative. An increase in the ESR reflects a rise in plasma fibrinogen (Bain, 1983) and other plasma proteins including immunoglobulins. There is often confusion over the normal reference range, which is affected by sex and age, this is normally quoted as up to 10 mm/h in males and up to 12 mm/h in females under the age of 50 years, thereafter it gradually increases up to 30 and 35 mm/h in both men and women (Lewis, Bain & Bates, 2006). The ESR is strongly influenced by the presence of anaemia which causes a falsely high reading and there is evidence that in elderly anaemic patients an apparent ‘normal’ ESR may in fact be influenced by incipient cardiovascular disease and increased fibrinogen levels (Bain, 1983) thus masking a high ESR. The International Council for Standardization in Haematology published a guideline document in which it is stated that measuring ESR is of little value for monitoring the onset or resolution of the acute phase response, it is only of use for measuring protein changes that occur in chronic disease (International Council for Standardization in Haematology, 1988), this document is currently being updated. This advice from ICSH leaves the clinical utility of the ESR unclear.

The ESR is a very frequently requested test but is probably the worst controlled and it is only in recent years that a whole blood control has become available. Most automated ESR methods are closed systems and allow measurement using the primary blood collection tube (either sodium citrate or EDTA). New methods are being introduced for the measurement of ESR that are based on <1 h. Optical sensors are used to determine the sedimentation of the erythrocytes and the information is then extrapolated. Some optical sensors are capable of reading through identification and bar-code labels and therefore it is unnecessary to aspirate any blood from the tube.

Laboratories are reporting large numbers ESR results every day and at present there is no ESR EQA scheme available in the UK, although EQA for this investigation is provided by both the CAP proficiency testing programme and the Royal College of Pathologists of Australasia Quality Assurance Programme. Setting up a scheme should be a priority for UK NEQAS (H), particularly as there are new methods available for the measurement of ESR which are theoretically correlated to the ICSH standardized method (ICSH, 1993).

Discussion

Automated blood cell counters are becoming more sophisticated and the range of reportable parameters available is ever increasing. There are increasing amounts of data provided, which require specialist knowledge to interpret as well as understand the limitations in the measurement of the parameter. Both laboratory scientists and clinicians need to keep up to date with new parameters and methods in haematology; often it is the laboratory scientists that need to introduce the new parameters and their clinical utility to the medical staff. Good laboratory practice ensures that reliable results of clinically relevant laboratory tests are reported to the clinician. Ideally these parameters should be standardized between different manufacturer’s instruments as numerical results provided for some parameters (MPV, IRF and RDW) are still very different. Various different parameters are used on different instruments for the diagnosis of FID. The rapid expansion of the range of parameters generated on automated haematology counters is partly driven by technological advances from instrument manufacturers, and is partly driven by limitations in existing technology, where an existing methodology is not sufficiently accurate, sensitive or reproducible. Sometimes a parameter can be enumerated but at the time has unknown clinical application, for example Sysmex Ret-Y, originally a service parameter, it is now the reported parameter Ret-He used in the diagnosis and monitoring of FID. This expanding range of parameters does allow for novel applications and introduces an element of research and development into routine laboratory haematology practice. Generally, these new parameters are generated when using the instrument in the routine CBC and differential mode or for IRF, some parameters for FID, IPF and one method for FRC in reticulocyte mode. On some instruments there is an additional cost for the stain used to identify NRBC. IQC is intended to ensure that measurements are sufficiently precise and within established limits. Ideally results on patient samples should not be reported unless it can be demonstrated through the use of IQC that the analytical procedure is valid and free from problems. Some instruments have reportable parameters available in the absence of manufacturers IQC material, these results should be reported with caution. The British Committee for Standardisation in Haematology (BCSH) introduced interlaboratory quality control into the United Kingdom in 1968 using both stabilized red blood cells and fresh blood samples. Initially they focused on the red blood cell count, MCV, haemoglobin and packed cell volume (Lewis & Burgess, 1969).

The UK NEQAS (H) has evolved greatly from the early beginnings in 1969 with parameters being added to keep abreast of the technological advancements on modern analysers. The white blood count (with five population leucocyte differential), platelet and reticulocyte counts have already been added to the assessment repertoire. EQA allows for comparison of results from different laboratories, establishing differences in between laboratories and between method performances. Using EQA results, it is possible to identify the best methods for performing a particular test and identify unreliable methods. They also have an essential role in maintaining and improving analytical quality and medical appropriateness of clinical laboratory data.

The relatively recent availability of the NRBC, IG, IRF, IPF counts and parameters indicating FID presents a challenge to EQA. Partially fixed stabilized blood causes problems for some analysers, particularly in relation to the separation of leucocytes for the automated differential and reticulocyte counts; this is also likely to be the case for the new parameters described, as they are mostly measured in the differential and reticulocyte channels. The European standard that covers the use of EQA schemes for the assessment of in vitro diagnostic examination procedures states ‘The EQAS organization shall provide survey samples that simulate as closely as possible the relevant properties of the samples on which the diagnostic procedures are intended to be used. EQAS organizations should not select survey materials which disadvantage an individual in vitro diagnostic medical device’ (BS EN 14136, 2004). This very important point provides a challenge for those developing the EQA scheme, particularly in the field of cellular counting. The provision of EQA material for cellular counting that has the same performance characteristics as fresh human whole blood across a range of analyser technologies and is sufficiently stable to be distributed for large scale EQA trials remains impossible. Similarly, the provision of EQA material with pathologically abnormal values at the level of clinical decision making is a challenge.

Medical laboratories are required to participate in an accredited EQA scheme wherever it is available. This is necessary to meet the standards for the medical laboratory under which Clinical Pathology Accreditation (UK) Ltd (CPA) assess and award accreditation status. Where no such EQA scheme is available then participation should be in a professionally directed scheme if possible (Clinical Pathology Accreditation Standards, 2008). The EQA schemes themselves are assessed under their own CPA standards and are required to demonstrate quality improvement, so there is a mutual need for EQA participant and provider collaboration in developing schemes as well as anticipating and responding to changes in technology and reporting of new clinically significant parameters.

The general consensus from UK NEQAS (H) users at the workshop and the instrument manufacturers’ representatives was that UK NEQAS (H) should most urgently address the provision of EQA for parameters with the widest availability. The greatest priority should be the NRBC count and samples should cover the entire clinical range. As NRBC affect the white cell count and lymphocyte count on many instruments, it should be included in the CBC and the automated differential schemes. Although IGs are also widely available, many laboratories mainly use this parameter as an indication of whether a blood film review is needed. It was thought that EQA was needed, but was less urgent. The IRF is not reported by all laboratories and there are still real difficulties in counting total reticulocytes in laboratories. EQA results show a high CV% not only between different instruments but also between different laboratories using the same instrument. IPF, red cell fragment count and leucocyte positional parameters have very limited availability at the current time so should not be a priority for inclusion into an EQA scheme.

The parameters used for the detection of FID vary significantly between instruments; however, the results can affect patient treatment so an EQA scheme is desirable. The group thought that guidelines for the diagnosis of FID would be useful and a request should be made to the BCSH general haematology taskforce to produce a peer reviewed guideline. It would be useful to have samples prepared by UK NEQAS (H) for analysis on all instruments using their individual parameters, but in addition to this an interpretive answer could be reported on whether the results indicate FID or not.

The introduction of EQA for MPV and RDW should be very straight forward as the laboratories will already have data produced from their instruments from the EQA CBC schemes. The CBC material would need to be validated for the performance assessment of the RDW and the MPV before these parameters are added to the profile but should be achievable in a relatively short time frame.

For laboratories using their routine haematology analysers to report parameters traditionally measured by flow cytometry it is desirable that they subscribe to an appropriate EQA scheme.

Despite the limited clinical utility of the ESR it remains a very popular test with the requesting clinicians, but it is poorly controlled. The workshop participants decided that an ESR scheme should be set up by UK NEQAS (H) and treated as a high priority.

For the new parameters with limited availability, where it may not be cost effective to provide a national EQA scheme, it may be possible to devise some form of interlaboratory comparison, e.g. using the manufacturer’s control material or local laboratories with same instrumentation sharing blood samples. Whilst this may not be an ideal scenario, it would provide laboratories with some reassurance that the results they were reporting were within consensus of other laboratories using the same technology and ensure that results reported on new parameters are within the same acceptable range.

Results on patient samples should not be reported unless they are supported by the use of IQC to show there is no problem with the analytical procedure. It should be the manufacturer’s responsibility to ensure that there is IQC material available with assigned values for all reportable parameters. Ideally there should also be EQA schemes for all parameters reported to the clinicians, however, with the rate of introduction of new tests, and the limited availability of some of the tests, this is an unrealistic expectation.

UK NEQAS (H) should, as soon as possible, introduce schemes for new reportable parameters as well as established parameters not currently included in the scheme. Both laboratories and clinicians should understand the limitations of reporting parameters which influence clinical decisions when there is no EQA scheme available.

Appendix

Other Contributors

I. Mellors, A. Roderick, A. Ward, C. O'Malley, J. Barker, B. De La Salle, P. McTaggart, K. Hyde, S. J. Machin, on behalf of the UK NEQAS General Haematology Scientific Advisory Group.

Other members of the UK NEQAS (H) General Scientific Advisory Group and invited participants in the UK NEQAS (H) workshop

John Ardern, Manchester Royal Infirmary and UK NEQAS (H) GSAG; Michelle Brereton, Manchester Royal Infirmary; Caroline Dore, Medical Research Council Statistical Unit and UK NEQAS (H) GSAG; David Guthrie, Manchester Royal Infirmary; Rod Hinchliffe, Sheffield Children’s Hospital and UK NEQAS (H) GSAG; Anne Mahon, UK NEQAS (H); Julie Oakes, Royal Preston Hospital; Neil Porter, Sheffield Teaching Hospitals NHS Foundation Trust; Albert Quick, St James’ University Hospital, Leeds.Tracey Smith Straney, University Hospital Aintree; Wayne Thomas, Derriford Hospital and UK NEQAS (H) GSAG.

Haematology instrument manufacturers’ representatives participating in the UK NEQAS (H) workshop

Abbott Alastair Manning; Beckman Coulter Corin Evans, Sandra Walch; Horiba Medical Mandy Campbell; Siemens Dave Barmby, Sue Mead; Sysmex Milly Mitchell.

Ancillary