Refrigerated Storage Improves the Stability of the Complete Blood Cell Count and Automated Differential

Transcription

1 H e m a t o p a t h o l o g y / REFRIGERATION IMPROVES CBC COUNT STABILITY Refrigerated Storage Improves the Stability of the Complete Blood Cell Count and Automated Differential Brent L. Wood, MD, PhD, Judy Andrews, MLM, Sharon Miller, MS, and Daniel E. Sabath, MD, PhD Key Words: Hematology; Automated; Differential; CBC count; Refrigeration A CBC count was performed on 113 random patient and 21 control specimens before and after 24hour room temperature (RT) storage; 98 random patient and control specimens also were analyzed before and after our 4 C storage. The Cell-Dyn 300 (Abbott Laboratories, Abbott Park, IL) was used for analysis. RT storage showed a decline in WBC count using the optical but not the impedance method, resulting in a large number of WBC flags. An increase in mean corpuscular volume also was seen for patient specimens. The automated WBC differential showed a decrease in the percentage of neutrophils and an increase in the percentage of lymphocytes, owing primarily to neutrophil degeneration. These changes also were seen in the manual differential to a similar degree. Storage of specimens at 4 C largely prevented all of these changes. The implementation of refrigerated specimen storage is a simple, inexpensive method to improve the accuracy of CBC results for aged specimens on automated hematology analyzers. Changes in the regulatory and reimbursement arenas of health care are resulting in a redistribution of sites at which laboratory testing is performed. The nature of contracts with managed care and preferred provider networks often necessitates transport of specimens to centralized laboratories for testing. As a result, testing often is delayed 12 to ours or more after venipuncture. Because the cellular elements in hematology specimens have limited stability in EDTA-anticoagulated blood,1 the delay in testing results in suboptimal specimen quality that requires the expenditure of additional laboratory resources to guarantee the quality and accuracy of results. This is occurring during a time of increasing pressures to reduce laboratory costs and streamline laboratory operation. Refrigerated storage of anticoagulated blood has been noted to improve the stability of some CBC count parameters, 2- but a systematic evaluation using a modern hematology analyzer has not been performed. We observed that the stability of the WBC count was dependent on whether impedance or optical methods were used. In addition, preliminary studies suggested that the WBC differential itself was stabilized by refrigerated storage. Consequently, we hypothesized that routine refrigerated storage would provide a means to improve the quality of hemogram parameters obtained on aged specimens and reduce the number of manual WBC differential counts required. Because manual differential counts are the most labor-intensive procedure performed in the hematology laboratory, a reduction in the number of manual differential counts could result in significant savings in labor and improvement in specimen turnaround time. To study this issue, we systematically performed a comprehensive evaluation of the stability of CBC parameters in EDTA-anticoagulated blood stored for ours at room temperature or 4 C. The Cell-Dyn 300 (Abbott Laboratories, Abbott Park, IL) performs WBC Aw J Clin Pathol 1999,112: Abstract

2 Wood et al / REFRIGERATION IMPROVES CBC COUNT STABILITY counts simultaneously by impedance and optical methods and provides a unique opportunity to evaluate the performance of these 2 technologies on compromised specimens. Materials and Methods approximately 2.-mL aliquots. An initial aliquot was analyzed within 2 hours of being obtained, and then refrigerated and reanalyzed at 12 hours. The remaining aliquots were refrigerated within 1 hour of being obtained, and analyzed at 24, 36, and 48 hours within to 30 minutes after removal from refrigeration. WBC Differential Hematology Analyzer Room Temperature Storage For this study, 113 randomly selected specimens from inpatient and outpatient locations were evaluated, and 21 healthy laboratory volunteers were used as a normal control group. For each specimen, 3 ml of blood was drawn into Vacutainer Hemoguard tubes containing ml of 7.% K^EDTA anticoagulant (Becton Dickinson, Franklin Lakes, NJ). The specimens were analyzed within 2 to 3 hours from when they were obtained. The specimens were kept at room temperature and reanalyzed after ours. Refrigerated Storage Ninety-eight randomly selected patient specimens were evaluated, and healthy laboratory volunteers were used as a normal control group. The specimens were obtained and analyzed as for the room temperature storage group. After initial analysis, the specimens were stored at 2 C to 7 C. After ours, the specimens were allowed to equilibrate at room temperature for 30 minutes and analyzed. Although rewarming of aged refrigerated specimens has been noted to result in degenerative changes,2,3 we found no difference in results between the method used in this study and more immediate analysis without rewarming for a preliminary group of 4 patient specimens (data not shown). We elected to use brief rewarming for this study, as it is difficult to maintain specimens at 4 C when using automated sample loading, and we believed these conditions were more consistent with routine operation in a highvolume clinical laboratory. Extended Refrigerated Storage Ten milliliters of blood was drawn from healthy laboratory volunteers as for the other groups and divided into 688 Am J Clin Pathol 1999;112: We performed 0-cell differential counts on all specimens in the refrigerated and room temperature control groups. A 0-cell manual differential count is standard practice in our laboratory, and we believe is the best compromise between the demands on a busy laboratory and the ideal 400-cell differential proposed in the H-A guideline of the National Committee for Clinical Laboratory Standards.7 The precision of the WBC differential was determined as follows. Samples were drawn from 2 healthy volunteers, and the samples from each individual were pooled. Following mixing, the pools were divided into aliquots, and each aliquot was analyzed twice on the Cell-Dyn 300, fresh and following storage at room temperature for 24 hours. This yielded a total of analyses for each volunteer, fresh and following storage. The same procedure was repeated for separate volunteers for storage at 4 C. The theoretical minimum precision (%CV [coefficient of variation]) for the WBC count and each component of the differential was calculated as vfn)/n x 0, where N was the number of events counted. Results One unique feature of the Cell-Dyn 300 is that the leukocyte count is measured by optical (WOC) and impedance (WIC) methods, with both counts performed on every specimen. This provided a unique opportunity to evaluate the performance of these 2 technologies on compromised specimens. The analyzer calculates the difference between WOC and WIC, and a delta (DLTA) flag is generated if the difference exceeds a preset limit, which is approximately ± % for leukocyte counts in the normal reference range. If the preset limit is not exceeded, the WOC count is selected for the WBC result. When a DLTA flag is present, ie, the preset limit has been exceeded, a proprietary algorithm is applied to the data to determine the appropriate WBC value, and either the WIC or WOC may be selected. Room Temperature Storage CBC results for the normal and patient groups at 0 and ours' storage are summarized in liable II and shown in IFigure II and IFigure 21. As expected, the patient specimens The Cell-Dyn 300 is a hematology analyzer that provides an 8-parameter hemogram including WBC count, RBC count, hemoglobin, hematocrit, RBC indices, and platelet count, and a -part WBC differential.6 The instrument also provides a -part WBC differential based on light-scattering properties combined with a proprietary technology for eosinophil detection using depolarized light. Software revision F, version 4, was used for this evaluation.

3 Hematopathology / ORIGINAL ARTICLE specimens, respectively, a change that was highly statistically significant (P < "). A comparison of WIC and WOC with the WBC is given in ITable 21. The instrument selected WOC for all WBC results on the initial and 24-hour-old specimens, whether unflagged or flagged, as demonstrated by the concordance of the WBC and WOC. WIC and WOC were equivalent initially, but after ours of storage, WIC remained constant, while WOC showed the 8.% decrease (Figure 2). Consequently, WIC was the better measure of the true WBC count on the aged specimens. While no ITable II Effect of Storage on CBC Count Storage Room Temperature Control Subjects (n = 21) 9 f WBC count, x /L RBC count, x l O 1 2 / ^ Hemoglobin, g/dl Hematocrit, % MCV, cu urn MCH, pg MCHC, g/dl Platelet count, x 9/Lt 7.2 ± ± ± ± ± ± ± ± ± ± 48 4 C Patients 1n = 113).6 ± ± ± ± ± ± ± 0.6 3±14 Patients n = 98) Control Subjects (n = ) 9.7 ± ± ± ± ± ± ± ± 1.3 ± 0.7 ± 17 ±4.7 ± 7.0 ±2.9 ± 1.0 ± ± 0.7 ± 1.7 ±4.6 ±7.1 ±2.9 ± ± ±4.0 ± 0.7 ± 1.7 ±4.9 ±.0 ± 1.9 ± 0.6 ± ±4.0 ± 0.8 ±1.7 ±.0 ± 4.9 ± 1.9 ± 0.6 ± 87 * All values are reported as mean ± SD. "Given in Systeme International (SI) units. To convert to traditional units, divide by the following factors: WBC count (/ il), 0.001; RBC count (x 6/nL), 1.0; platelet count (x lovul) to convert other reported values to SI units, multiply by the following factors: hemoglobin (g/l), ; hematocrit (proportion of 1.0), 0.01; mean corpuscular volume (MCV; fl), 1.0: mean corpuscular hemoglobin (MCH; pg). 1.0; mean corpuscular hemoglobin concentration (MCHC; g/l).. I fl 1 1 / 0 0 \r± E x J <* > O 90 u 80 8 # = MCV (fl) jf 70 y = x y = 882x fl2 = 908 / s 1 60 ' i MCV (fl) Figure I I Effect of room temperature (A) and 4 C (B) storage on the mean corpuscular volume (MCV) for patient specimens. Note the constant bias toward higher values seen with storage at room temperature. Values given are Systeme International; the conversion factor for traditional units (cu urn) is 1.0. Am J Clin Pathol 1999:112: exhibited higher mean WBC counts and lower mean hemoglobin and hematocrit measurements than those of the control group. Of the directly measured parameters, the RBC count, hemoglobin level, and platelet count were stable. In contrast, a constant 2.6% increase in mean corpuscular volume (MCV) was seen in the patient specimens (P <.000; see Figure 1A) and also was reflected as a decrease in the dependent parameter, the mean corpuscular hemoglobin concentration. The MCV was essentially unchanged in the control group. Proportional 8.3% and 8.% decreases in the WBC count also were seen in control and patient

4 Wood e t al / REFRIGERATION IMP-ROVES CBC COUNT STABILITY Table 21 Effect of Storage on WBC Count* 4 C(n = 98) Room Temperature (n = 113) WBC count, x 9 /L WOC, x 9/L WIC, x 9 /L Delta flags (%) h.6 ±.1.6 ±.1.4 ± ± ± ± ± ± ± ±4.0.1 ±4.1 WBC, WBC optical count (WOC), and WBC impedance count (WIC) are reported as mean ± SD. Values given are in Systeme International (SI) units. To convert to traditional units (/ il), divide by fl 3 B : * o O g y = 377x y = 944X fl2 = 929 Ft = 83 0 ~~ I WOC (K/^tL) 30 3 e WIC (K/jiL) 30.E WOC O y = 886X y = x R2 = 882 fl = 873 WOC (K/ il) WIC (K/ al) Figure 21 Effect of room temperature (A and B) and 4 C (C and D) storage on the WBC count for patient specimens. Note the proportional bias toward lower values and increased scatter with the optical method (WOC) after storage at room temperature (A). Values are given in traditional units; to convert to Systeme International units (x9/l), multiply by WIC = WBC count by impedance method. 690 Am J Clin Pathol 1999;112:687-69

5 Hematopathology / ORIGINAL ARTICLE fresh specimens triggered a DLTA flag, 40% of the patient specimens stored for ours triggered a DLTA flag as a result of the declining WOC (Table 2). As expected, the specimens triggering a DLTA flag had the largest discrepancies between WIC and WOC. (Figure 2). DLTA flags were generated on none of the fresh specimens and on only % (n = ) of patient specimens stored at 4 C for ours, a greater than 80% decrease from the number seen for storage at room temperature. Extended Refrigerated Storage Refrigerated Storage WBC Differential Evaluation The results of the automated differential count for specimens stored at room temperature or 4 C is summarized in ITable 41. Storage of control and patient specimens at room temperature resulted in a decrease in the percentage of ITable 31 Effect of 4 C Storage on CBC Count for Samples* WBC count, x9/l RBC count, x l2 /L Hemoglobin, g/dl Hematocrit, % MCV, cu im MCH, pg MCHC, g/dl Platelet count, x9/l 1 12 h 36 h 48 h All values are reported as the mean. For abbreviations and Systeme International unit conversion factors, see Table I. Table 41 Change in WBC Automated Differential Count With Storage* Specimen Neutrophils Room temperatiire Control 6.1 (n = 21) Patient 76.3 (n= 113) 4 C Control 60. (n = ) Patient 74. (n = 98) Lymphocytes Monocytes Eosinophils Change Change ':' ' * ' : * ' :.6 : Change Change 3.0 Change Basophils NS = not statistically significant. * To obtain the proportion of 1.0 for Systeme International units, divide by Counts are given as percentages; the change is the percentage of change. * Statistically significant (P <.0). Am J Clin Pathol 1999;112: To further evaluate the effect of storage at 4 C, normal control specimens were assayed at 0, 12, 24, 36, and 48 hours after collection I Table 31. The means for RBC count, hemoglobin level, and platelet level remained stable throughout the 48-hour period, while the MCV showed a 1 % increase that was most pronounced after 36 hours. The mean for the WBC count exhibited a steady decline, resulting in a % decrease after 48 hours. The decline in WBC count was accompanied by a steady increase in DLTA flags after 12 hours of storage, such that at 48 hours, all specimens triggered DLTA flags. Results for specimens analyzed fresh and after storage at 4 C for ours also are summarized in Table 1. As expected, the patient specimens exhibited higher mean WBC counts and lower mean hemoglobin and hematocrit measurements than the control group. As with storage at room temperature, the 3 directly measured parameters, RBC, hemoglobin, and platelet values, were essentially unchanged after storage at 4 C. In contrast, while a statistically significant decrease (P <.0008) in the WBC count was observed with storage at 4 C (Table 2), it was clearly reduced from that seen with storage at room temperature (Figure 2). Storage at 4 C stabilized the MCV for patient specimens (Figure 1). Examination of the WBC count revealed that WOC was chosen for all WBC. Following storage at 4 C, WIC and WOC compared favorably with their initial values

6 Wood et al / REFRIGERATION IMPROVES CBC COUNT STABILITY Prolonged storage of control specimens for up to 48 hours at 4 C revealed stability of the WBC count, percentage of neutrophils, and percentage of monocytes to within % of their original values IFigure 31. The percentage of lymphocytes showed a progressive decline that reached roughly % of the original value at 48 hours. This decline was largely due to a loss of light scatter with the lymphocyte population slipping below the discriminator, as described. A mild increase in the percentage of eosinophils was present beyond ours, but a marked progressive increase in the percentage of basophils was observed that reached more than twice the original value at 48 hours. Comparison of the automated with the manual differential count after storage at room temperature revealed that the changes described for the automated percentage of neutrophils and percentage of lymphocytes also are seen with the manual differential count to a similar degree ITable 1. This is consistent with neutrophil degeneration morphologically and by light scatter as the primary cause of the observed changes. With storage at 4 C, the decrease in the percentage of lymphocytes seen with the automated differential was not seen with the manual differential, consistent with a mild loss in lymphocyte light scatter as described. However, if one compares the theoretical minimum 9% confidence intervals 692 Am J Clin Pathol 1999; 112: a> >" - -WBC - * - Neutrophils - * - Lymphocytes - - Monocytes -"- Eosinophils - - Basophils % Line 1% Line 2 0 o <v J c a> o i_ <D Q Hours IFigure 31 Effect of 4 C storage on the automated WBC differential count. Values are given in percentages. Note the marked percentage increase in basophils, the decline in the percentage of lymphocytes, and the gradual mild increase in the percentage of eosinophils. for the 2 differential methods based on the number of events counted (± 2 x vfn)/n), the decrease in the percentage of lymphocytes for the automated method is well within the error of the manual differential count. This would not significantly improve by switching to a 400-cell manual differential. The apparent increase in statistically significant differences for many of the populations in the automated differential is largely due to the better precision of the automated differential and is a reflection of the poorer precision of the manual method. Consequently, while the automated differential after storage at room temperature for ours shows significant changes in the percentage of neutrophils and the percentage of lymphocytes, the manual differential shows similar changes and offers no significant improvement. Similarly, given the error in the manual differential count, there is no convincing difference between the automated and manual differential counts after storage for 24 hours at 4 C. The precision of the WBC count and differential was determined for specimens stored at room temperature or 4 C and compared with the precisions obtained before storage ITable 61. This revealed that the percentage of neutrophils and the percentage of eosinophils were near the theoretical minimum CVs obtained from consideration of the number of events counted (see the "Materials and Methods" section). The CVs of both measurements increased with storage at room temperature but were stabilized by storage at 4 C. The precision of the percentage of lymphocytes was also similar to the theoretical neutrophils with a concurrent increase in the percentage of lymphocytes, a finding most marked for the patient specimens. Examination of the scattergrams revealed that this was largely due to a variable decrease in neutrophil light scatter with expansion and loss of definition of the neutrophil cluster, blurring the boundary with the lymphocyte cluster and resulting in inclusion of neutrophils with the lymphocyte population. In contrast, storage at 4 C resulted in a slight increase in the percentage of neutrophils and a mild decrease in the percentage of lymphocytes. Examination of these scattergrams revealed that the light scatter characteristics and definition of the neutrophil cluster were well maintained at ours, but the lymphocyte cluster exhibited a slight decrease in light scatter such that a portion of the cluster fell below the discriminator and was not included in the differential. This loss of lymphocytes appeared largely responsible for the small rise in the percentage of neutrophils. The percentage of monocytes showed a slight decrease that did not reach statistical significance in the larger patient groups, but seemed stabilized by storage at 4 C. A small decrease in the percentage of eosinophils with storage at room temperature was largely corrected by storage at 4 C. However, a consistent increase in the percentage of basophils was present with storage at room temperature and was not corrected by storage at 4 C in control and patient samples. Separation of the patient specimens into those with and without DLTA flags confirmed that the previously described changes were most marked for those with DLTA flags.

7 Hematopathology /ORIGINAL ARTICLE Discussion Of the directly measured hemogram parameters, the RBC count, hemoglobin level, and platelet count largely were unaffected by storage at room temperature or 4 C. In contrast, for patient specimens, the MCV showed a significant increase with storage at room temperature that was prevented by refrigeration. This phenomenon has been described by others and is partly a consequence of the anticoagulant used.1-8-" An increase in platelet count with room temperature storage has been reported in older studies,,12 but such an increase was not seen in the present study or in other more recent studies.3"11 The WBC count showed an 8.% decrease with storage at room temperature that was prevented by refrigeration. This has been noted by others using the Cell-Dyn but not by those using instruments relying only on an impedance method.9-" As the Cell-Dyn 300 uses optical and impedance methods to generate the WBC count, further investigation revealed that WOC was responsible for the decrease. In the WOC, light scatter measurements are analyzed from cells diluted in a reagent that is formulated to maintain the cells in their near native state. When leukocytes age, their light scatter properties change, and these cells fall below the lower threshold on the 0 / scattergram. Events below the Table 1 Automated n Manual WBC Differential Count for Control Samples* Automated Oh Oh Room temperature (n = 21) Neutrophils Lymphocytes Monocytes Eosinophils Basophils 4 C (n = ) Neutrophils Lymphocytes Monocytes Eosinophils Basophils Manual * 41.* 9.8* * * 43.0* 7.0* * 24.4* 7.; 1.* NS = not statistically significant. * Values are given as percentages. To convert to Systeme International units {proportion of 1.0), multiply by ' Statistically significant (P <.0). Table 61 Precision of the WBC Differential Count With Storage 4 C Room Temperature WOC WIC Neutrophils (%) Lymphocytes (%) Monocytes (%) Eosinophils (%) Basophils (%) Theoretical Theoretical Cell Differential WOC = WBC optical count: WIC = WBC impedance count. Am J Clin Pathol 1999; 112: %CV but seemed to worsen somewhat more on storage at 4 C than at room temperature. The monocyte %CV was more variable between specimens but generally seemed to be less precise than the aforementioned measurements, increased with storage at room temperature, and was stabilized by storage at 4 C. The basophils were noticeably less precise than all previous measurements and largely were unaffected by storage. The %CV of the fresh specimens was similar to 3 randomly selected precision runs performed during the previous year (data not shown). These data indicate that difficulties remain in properly categorizing monocytes and basophils on the Cell-Dyn 300 but that neutrophils and eosinophils generally are categorized quite well. In addition, storage at 4 C is capable of stabilizing the increases in %CV for neutrophils, eosinophils, and monocytes seen during storage at room temperature.

8 Wood et al / REFRIGERATION IMPROVES CBC COUNT STABILITY We found that the decreased WBC count was largely due to a subset of degenerate specimens that triggered a DLTA flag, indicating a roughly % difference between WOC and WIC. If WIC was substituted for the reported WBC (WOC) on these specimens, the resulting WBC count was nearly corrected for both control and patient specimens stored at room temperature (data not shown). These results suggest that for specimens without a DLTA flag, the WBC count may be reported directly off the instrument. For specimens having a DLTA flag, further evaluation will be required, since a variety of conditions may give rise to a difference between WOC and WIC large enough to generate a DLTA flag, including fragile lymphocytes, lyse-resistant RBCs, and nucleated RBCs.6 If WOC is greater than WIC, lyse-resistant RBCs should be suspected and confirmed by disappearance of the DLTA flag by reanalysis in extended lyse mode. Otherwise, a manual smear evaluation will be required, and WIC should be reported. We find that the presence of 2 counting technologies on the same instrument is an extremely useful feature for quality control of the WBC count, since each method is subject to different artifacts, and, hence, they are complementary. The automated differential of the Cell-Dyn 300 has been evaluated extensively.3-413"' However, the stability of the WBC differential on automated hematology analyzers has been evaluated by relatively few studies. On the Coulter STKS (Coulter, Hialeah, FL), an instrument that uses light scatter, cell volume, and conductivity to perform a -part WBC differential, a decrease in the percentage of neutrophils with a compensatory increase in the percentage of lymphocytes was noted after storage at room temperature for more than 18 hours.2 Storage at 4 C stabilized this portion of the differential for up to 72 hours. This finding is similar to those of the present study and the study of Al-Ismail et al,3 but different from that of Vives-Corrons et al4 who showed no significant change over 48 hours storage at 18 C to C. In our study, an increase in the percentage of basophils with a decrease in the percentage of eosinophils was seen on storage at room temperature, with stabilization of only the percentage of eosinophils by storage at 4 C, a finding also 694 AmJCIinPathol 1999;112: seen in the data of Al-Ismail et al.3 However, these findings contradict those of Vives-Corrons et al,4 who showed a marked decrease in the percentage of basophils and an increase in the percentage of eosinophils with storage at 18 C to C and stabilization of both by storage at 4 C. The explanation for the differences between the findings of Vives-Corrons et al4 and those of our study and that of AlIsmail et al3 is unclear. The improved precision of the automated differential over the manual differential for fresh specimens is well recognized.16 However, in aged specimens, little literature exists comparing the manual and automated WBC differential counts. The default practice of most laboratories is to perform manual differentials on specimens that do not meet the laboratory's criteria for reporting automated differentials, often aged specimens. The present study demonstrates that the changes in WBCs that result in a suboptimal automated differential with specimen aging at room temperature, ie, light-scatter properties, also result in morphologic degeneration primarily of the neutrophil population. Consequently, the use of the manual differential on aged specimens stored at room temperature provides no significant improvement in accuracy over the automated differential but does degrade the precision of the reported differential owing to the smaller number of cells counted. In contrast, refrigerated specimen storage stabilizes the light-scatter properties and morphologic features of the WBCs, resulting in markedly improved automated and manual differential counts and stabilization of the %CV for neutrophils, eosinophils, and monocytes. The conclusion we draw from these data is that there is no justification for performing manual differential counts on aged specimens, regardless of storage conditions. Because the Cell-Dyn 300 automated differential also has been validated for fresh specimens in multiple studies, we suggest that the automated differential from the Cell-Dyn 300 is the preferred reportable result over the manual differential for virtually all specimens, regardless of age. Our study shows that refrigerated storage of specimens rather than manual differential counts is the better way to improve the accuracy of the WBC differential in aged specimens. In our opinion, the primary purpose of a manual smear examination should be the detection of morphologic abnormalities, not cell enumeration, and it is necessary only when instrument flags are present. We have shown that sample age alone may be responsible for generating a significant decrease in the WBC count and may result in increasing numbers of DLTA flags on the Cell-Dyn 300, and that both phenomena may be significantly reduced by refrigerated storage of the specimens. Consequently, for laboratories receiving larage numbers of aged specimens, such as reference laboratories, specimen refrigeration would markedly reduce the labor required for threshold are excluded from the count, resulting in WOC at ours being less than the fresh WOC. In contrast, WIC showed no significant change with storage at room temperature or 4 C. This is largely due to the lytic reagent system used for impedance counting, which strips away the WBC cytoplasm, leaving only nuclei for counting; hence, the method is insensitive to cytoplasmic changes seen in early WBC degeneration. In general, WIC is clearly superior to WOC in generating an accurate WBC count for aged specimens. Unfortunately, the Cell-Dyn 300 almost exclusively reports WOC for the WBC count on all specimens regardless of age or presence of DLTA flags.

9 Hematopathology / ORIGINAL ARTICLE From the Department of Laboratory Medicine, University of Washington, Seattle, WA. Address reprint requests to Dr Wood: Department of Laboratory Medicine #371, 199 NE Pacific St, Seattle, WA Acknowledgments: We thank Courtney Ota-Bishop, MT(ASCP), and Jeanne Meade, MT, for organizing the specimen and data collection, and Lisa McDonnel, MT(ASCP), Kristen Smith, MT(ASCP), Kara MacGowan, MT(ASCP), and Tan Ong, MT(ASCP), for assistance compiling the computer database. References 1. Hamilton PJ, Davidson RL. Recommendations of the International Council for Standardization in Haematology for Ethylenediaminetetraacetic Acid Anticoagulation of Blood for Blood Cell Counting and Sizing, International Council for Standardization in Haematology: Expert Panel on Cytometry. The interrelationships and stability of Coulter S-determined blood indices. AmJ Clin Pathol. 1993;0: Al-Ismail SA, Bond K, Carter AB, et al. Two-centre evaluation of the Abbott CD300 blood counter. Clin Lab Haematol. 199;17: Vives-Corrons JL, Besson 1, Jou JM, et al. Evaluation of the Abbott Cell-DYN 300 hematology analyzer in a university hospital. AmJ Clin Pathol. 1996;:3-9.. Ward SL, Castleberry R, Smalley DL, et al. Use of the CellDyn 300 in a reference laboratory. Lab Med. 1996;27: Principles of operation. In: Cell-D^n 300 Operator's Manual. Abbott Park, IL: Abbott Laboratories; 1993:chap National Committee for Clinical Laboratory Standards. Reference leukocyte differential count (proportional) and evaluation of instrumental methods; approved standard. Villanova, PA: National Comittee for Clinical Laboratory Standards;1992.NCCLS document H-A. 8. Lawrence AC, Bevington JM, Young M. Storage of blood and the mean corpuscular volume. J Clin Pathol. 197;28: Brittin GM, Brecher G, Johnson CA, et al. Stability of blood in commonly used anticoagulants: use of refrigerated blood for quality control of the Coulter Counter Model S. Am J Clin Pathol. 1969;2: Cohle SD, Saleem A, Makkaoui DE. Effects of storage of blood on stability of hematologic parameters. AmJ Clin Pathol. 1981;76: Payne BA, Pierre RV, Lee WK. Evaluation of the TOA E-000 automated hematology analyzer. AmJ Clin Pathol. 1987;88: Lampasso JA. Changes in hematologic values induced by storage of ethylenediaminetetraacetate human blood for varying periods of time. AmJ Clin Pathol. 1968;49:443^ Fournier M, Gireau A, Chretien MC, et al. Laboratory evaluation of the Abbott Cell DYN 300 -part differential. AmJ Clin Pamol. 1996;: Burchert-Graeve M, Kock R. Automated leucocyte differentials in 292 patients with leucopenia: an evaluation of the Abbott CELL-DYN 300 (CD300) haematology analyser. Clin Lab Haematol. 1996;18:3-9.. Dorner K, Schulze S, Reinhardt M, et al. Improved automated leucocyte counting and differential in newborns achieved by the haematology analyser CELL-DYN 300. Clin Lab Haematol. 199;17: Simson E, Groner W. The state of the art for the automated WBC differential, part 1: analytic performance. Lab Hematol. 199;1: Warner BA, Reardon DM. A field evaluation of the Coulter STKS. Am J Clin Pathol. 1991;9: Am J Clin Pathol 1999;112: additional manual smear evaluation. We also have shown that WIC may be substituted for the instrument WBC count (usually WOC) in most cases if specimen refrigeration is not possible and other causes of a DLTA flag have been excluded. The implementation of refrigerated specimen storage for CBC counts and differentials is a simple, inexpensive method to improve the accuracy of reported results for aged specimens. However, this method is not without limitations, including a decrease in the lymphocyte percentage that is mild for specimens stored for less than 24 hours, no improvement in the elevations seen in basophil percentages, and progressive degenerative changes for specimens stored more than ours. Despite these limitations, the reduced number of instrument flags and substantial reduction in manual differential counts performed will result in improved specimen turnaround time and significant savings in technologist time.