Two-centre evaluation of the Abbott CD3500 blood counter

Summary The CD3500 blood counter (Abbott Laboratories) is a 33 parameter fully automated blood counter that produces a five part differential count with flagging of leucocyte abnormalities. In this evaluation excellent correlation between CD3500 and Coulter STKR blood counter was found for all red cell and platelet parameters on the 221 samples tested. Studies of carryover, mixing efficiency and precision also gave excellent results. There was a good correlation with manual 400 cell differential counts for neutrophils, lymphocytes, monocytes and eosinophils for the 468 samples compared. Correlation of CD3500 and manual basophil counts was poor. Normal samples stored at 4°C and analysed while cold showed satisfactory stability for WBC, RBC, Hb, MCV and platelets for 48 h and a stable differential for 24 h. Correlation with the differential count produced by the Coulter STKS showed good correlation for neutrophils, lymphocytes, monocytes and eosinophils; correlation with STKS basophils was poor. False positive flagging rate varied between 8.9% (Band and/or IG) and 0.9% (NRBC) depending on the nature of the flag; 5.8% of samples exhibited two or more false positive flags. No significant breakdowns were encountered during the period of the evaluation. The scatterplot displays of laser light scatter produced by the instrument provide an interesting adjunct to conventional morphology.     

Evaluation of the Roche Cobas Argos 5Diff automated haematology analyser with comparison to a Coulter STKS

Summary. The performance of the Roche Cobas Argos 5Diff (Argos) automated haematology analyser was evaluated by comparison to manual blood film examination and a Coulter STKS (STKS) analyser. The Argos demonstrated excellent between and inter-batch imprecision for all parameters, except the MCHC, and good linearity for Hb, WBC and platelet count (PLT). After an initial fall the PLT, results were stable for up to six h at 18°C in EDTA(K3) after which an increasing proportion of cells were classified as lymphocytes. Results of 239 patient samples analysed on both instruments, compared by linear regression, gave excellent correlation (r² > 0.90) for most parameters with the exceptions of the MCHC (0.317), eosinophils% (0.756), monocytes% (0.48) and basophils% (0.002).‘Flagging’of cellular abnormalities by the Argos resulted in excellent sensitivity (97.5%), specificity (93.2%) and efficiency/agreement (93.2%), with fewer false positive and negative results than the STKS, although these differences were not statistically significant. The performance characteristics of the Argos were comparable to those of the STKS with a possible improvement in its flagging abilities.     

An evaluation of leucocyte analysis on the Coulter STKS

Summary The performance of leucocyte analysis on the Coulter STKS (Coulter, Hialeah, FL, USA) was evaluated for accuracy, precision and reliability. The results were compared with those obtained from visual examination of a Romanowsky stained blood film together with the automated WBC-diff. from the Technicon H*1 (Technicon, Tarrytown, NY, USA). The relationship between the number of cells counted per WBC-diff. and the WBC count of the sample was established. Precision of the STKS WBC-diff. was acceptable on blood samples with normal and low WBC counts. Correlation with an 800 cell manual WBC-diff. (n = 104) was excellent (r = 0.97, 0.97, 0.83, 0.98 and 0.53 for neutrophils, lymphocytes, monocytes, eosinophils and basophils respectively). Blood specimens, collected into dipotassium EDTA, could be stored at 20–25°C for at least 8 h with no significant effect on the STKS WBC-diff. In a study of 513 patient samples, the BLASTS suspect flag gave 5.4% false positives and zero false negatives, the VARIANT LYMPHS flag gave 1.5% false positives and 0.4% false negatives, and the IMM GRANS/BANDS flag gave 30.8% false positives and 2.3% false negatives. Several instrument and sample related problems were encountered during this study. Despite these limitations, the STKS can provide efficient 5 part WBC-diffs. and effective screening for WBC abnormalities.     

Evaluation and use of the white blood cell differential provided by the Coulter® STKS in a children's hospital

Summary. The Coulter® STKS was evaluated in a children's hospital, in order to (a) compare the WBC differential given by the instrument to a 400 cell visual differential (reference method); (b) evaluate the sensitivity and specificity of the alarm system, and (c) provide data concerning the use and interpretation of results in children. 653 blood samples were collected. The Coulter® STKS results were studied in 523 patients having no morphological abnormalities in the blood smears, separated into subgroups according to the presence of STKS alarms and according to age. The results were found accurate both in STKS negative and STKS positive patients (i.e., those with alarms:‘Blasts', Imm Gran 2, Variant Lymph, NRBC, review slide). Negative STKS results had the same accuracy in all age groups, except in neonates where slide review must be systematically performed. The instrument exhibited a good sensitivity of the suspect flags studied (91.4%), with a lower specificity (72%) reflecting the number of false positive results found in our group, probably due to the cytological features particular to children. However, it was shown that the numerical results given by the Coulter® STKS in positive patients could be taken into account, provided that a scan of the blood smear was negative for morphological WBC abnormalities.     

An evaluation of leucocyte analysis on the Coulter STKS.

The performance of leucocyte analysis on the Coulter STKS (Coulter, Hialeah, FL, USA) was evaluated for accuracy, precision and reliability. The results were compared with those obtained from visual examination of a Romanowsky stained blood film together with the automated WBC-diff. from the Technicon H*1 (Technicon, Tarrytown, NY, USA). The relationship between the number of cells counted per WBC-diff. and the WBC count of the sample was established. Precision of the STKS WBC-diff. was acceptable on blood samples with normal and low WBC counts. Correlation with an 800 cell manual WBC-diff. (n = 104) was excellent (r = 0.97, 0.97, 0.83, 0.98 and 0.53 for neutrophils, lymphocytes, monocytes, eosinophils and basophils respectively). Blood specimens, collected into dipotassium EDTA, could be stored at 20-25 degrees C for at least 8 h with no significant effect on the STKS WBC-diff. In a study of 513 patient samples, the BLASTS suspect flag gave 5.4% false positives and zero false negatives, the VARIANT LYMPHS flag gave 1.5% false positives and 0.4% false negatives, and the IMM GRANS/BANDS flag gave 30.8% false positives and 2.3% false negatives. Several instrument and sample related problems were encountered during this study. Despite these limitations, the STKS can provide efficient 5 part WBC-diffs. and effective screening for WBC abnormalities.     

An evaluation of the differential from the Abbott CD 3500 in a population of patients with haematological abnormalities

Summary The Cell Dyn 3500 (CD 3500) Haematology Analyser (Abbott Diagnostics, Lane Cove, NSW, Australia) has been evaluated to determine the reliability of the differential and white cell suspect flagging in a population of patients with known haematological abnormalities. The evaluation included the assessment of white cell differential parameters in addition to sensitivity, specificity and efficiency of white cell suspect flagging. The study showed that the CD 3500 is an efficient and sensitive screening tool for detecting the presence of clinically significant white cell abnormalities. Generally, the Blast and Band flags demonstrated the highest level of false positive flagging (lowest sensitivity). The immature granulocyte flag was found to be an extremely reliable indicator of the presence of myelocytes, metamyelocytes and promyelocytes on the film. There was a 0.4% false negative rate where significant numbers of variant lymphocytes (> 10%) were not detected by the CD 3500 on flagging alone. When analyser flags and white cell scatter distribution are considered in combination with defined laboratory limits, all white cell suspect flags have acceptable sensitivity, specificity and efficiency rates.     

Performance of the SE-9000 automated haematology analyser in a laboratory serving a haematological oncology unit

The performance of the SE-9000 automated haematology analyser in a laboratory receiving a high number of abnormal specimens from haematological oncology patients was assessed according to formal protocols for the evaluation of blood cell counters. Linearity over a useful working range, precision in clinically important ranges and negligible carry-over were demonstrated in this group of patient samples confirming the results of previous investigators. The comparability of instrument derived differential leucocyte counts from both normal and distributionally abnormal samples with those obtained by visual microscopy using the NCCLS H-20 A protocol was very good. The sensitivity of flags for the detection of immature granulocytes and myeloid blast cells was high and this can be attributed to the incorporation of a new measuring channel (Immature Myeloid Information or IMI channel). The number of unrecognized abnormalities was low and when compared with the poor sensitivity of the routine 100-cell visual differential leucocyte count, the analyser was judged suitable for monitoring patients with haematological malignancies. The performance of flags such as ‘left shift’ and ‘atypical lymphocytes’ can be improved by taking into consideration distributional abnormalities such as neutrophilia and lymphocytosis. The trigger level for these flags should be adapted to the clinical need particularly in cases of neutropenia following chemotherapy, and in lymphoproliferative disorders and infection.     

Performance of the XE‐2100 leucocyte differential

The XE‐2100? was evaluated in a multicentre study following a previously established protocol. In this paper, we demonstrate the results of analytical performance studies, including comparison of the leucocyte differential with the NCCLS H20‐A method and evaluation of flagging sensitivity. Linearity of the leucocyte count over a wide clinical range, low imprecision in clinically important ranges and no measurable carry over were confirmed. For comparability studies, 4 × 200 cell microscopic differential leucocyte counts were correlated with the automated five‐part‐differential counts. No significant differences were detected in (1) a group without morphological abnormality and in (2) a leukopenic group. The sensitivity of flags for the detection of immature granulocytes and myeloid blasts was very good. Only few samples containing blast cells remained unrecognized but these would have been examined microscopically in any event because of other abnormalities indicated by the instrument. Atypical/abnormal lymphocytes/and lymphoblasts were detected very reliably when the total lymphocyte count and the flags were evaluated in combination. A similiar procedure is recommended for the detection of left shift. When the neutrophil count is elevated, the sensitivity of the left shift flag is improved. The absolute immature granulocyte (IG) count by the instrument correlates well with that of myeloid precursor cells by microscopy.     

An evaluation of the CELL-DYN 1700® haematology analyser: automated cell counting and three-part leucocyte differentiation

The performance of the CELL-DYN 1700® (Abbott Diagnostics, Abbott Park, IL, USA) was evaluated in a tertiary care hospital laboratory using the guidelines proposed by the German Society of Clinical Chemistry. Precision, accuracy, linearity, background counts, and carry-over were satisfactory for all measured standard parameters including haemoglobin concentration, haematocrit, red blood cell count, mean corpuscular volume (MCV), red cell distribution width (RDW), white blood cell count and platelet count. With 259 selected normal and abnormal blood samples the results of the CELL-DYN 1700® (CD1700) compared very well (r > 0.96 for all parameters with exception of RDW) with those obtained with the Bayer Diagnostic H-1 and the Hoffmann-La Roche Cobas Argos systems. This study considered in particular the performance of the CD1700 three-part leucocyte differential. For those samples without instrument-generated suspect flags, the neutrophil and lymphocyte percentages were highly correlated with the results of the H-1 blood cell counter (r = 0.97 and 0.98, respectively) and with manual 400-cell differentials (r = 0.91 and 0.88, respectively). In contrast, the CD1700 mid-fraction which comprised the composite total of mono cytes, eosinophils, basophils and precursor white cells (when present) could not be directly compared to the differentials from the H-1 system or from manual microscopy. For those samples with CD1700 instrument suspect flags, the neutrophil and lymphocyte differential results also compared well with both the H-1 (r = 0.93 and 0.93, respectively) and manual estimates (r = 0.89 and 0.87, respectively). In conclusion, the CD1700 is an accurate haematology analyser for cellular blood counts and three-part leucocyte differentiation.     

Evaluation of performance for automated differential leucocyte counting on Sysmex NE-8000 by NCCLS recommended protocol,H20-T

Summary In July and August 1989, the automated differential cell count (ADCC) capability of the Sysmex NE-8000 was evaluated by a modified H20-T protocol, the standard for differential leucocyte counting (DLC) method published by the National Committee for Clinical Laboratory Standards (NCCLS), USA. The ADCC of NE-8000 was compared with the manual reference method, using the Central Hematology Laboratory of the National Taiwan University Hospital as a venue. Of the total 493 blood samples analysed, 246 were from inpatients and 247, from outpatients. The NE-8000 system was found to be 4 to 4.5 times more precise than the 200-cell manual differential count in neutrophil, lymphocyte and eosinophil counts, and 1.5 to 2.5 times more precise in monocyte and basophil counts. According to our laboratory-defined criteria, the clinical sensitivity of NE-8000 was satisfactory with a sensitivity of 95.22%, a specificity of 83.71%, a false positive rate of 16.29%, a false negative rate of 4.78% and an efficacy of 95.10%. Correlation of the NE-8000 with the manual reference method was acceptable with correlation coefficients of 0.85 for neutrophils, 0.88 for lymphocytes, 0.97 for eosinophils, 0.62 for monocytes and 0.40 for basophils. The suspected interpretive messages given by the NE-8000 were good in flagging blasts and immature granulocytes only. In leucopenic specimens, the NE-8000 failed to flag morphological abnormalities in half of the cases. Traditional manual differential count of 100 or 200 leucocytes on stained smears also gave 2.3% false negative results on morphological anomalies. Generally speaking, it is practical to apply analysers which combine complete blood cell count and ADCC to deal with specimens without DLC abnormalities, and saving the manpower for those specimens with morphological abnormalities.     

Performance of the XE-2100 leucocyte differential

The XE-2100 trade mark was evaluated in a multicentre study following a previously established protocol. In this paper, we demonstrate the results of analytical performance studies, including comparison of the leucocyte differential with the NCCLS H20-A method and evaluation of flagging sensitivity. Linearity of the leucocyte count over a wide clinical range, low imprecision in clinically important ranges and no measurable carry over were confirmed. For comparability studies, 4 x 200 cell microscopic differential leucocyte counts were correlated with the automated five-part-differential counts. No significant differences were detected in (1) a group without morphological abnormality and in (2) a leukopenic group. The sensitivity of flags for the detection of immature granulocytes and myeloid blasts was very good. Only few samples containing blast cells remained unrecognized but these would have been examined microscopically in any event because of other abnormalities indicated by the instrument. Atypical/abnormal lymphocytes/and lymphoblasts were detected very reliably when the total lymphocyte count and the flags were evaluated in combination. A similar procedure is recommended for the detection of left shift. When the neutrophil count is elevated, the sensitivity of the left shift flag is improved. The absolute immature granulocyte (IG) count by the instrument correlates well with that of myeloid precursor cells by microscopy.     

Evaluation of the Vega haematology analyser in a university hospital setting

The performance of the ABX Vega haematology analyser was compared with that of the Sysmex NE-8000, with specific attention to flagging performance and ergonomics. Eight hundred routine samples underwent precision and interinstrument variability studies and 168 samples corresponding to various blood disorders were studied meanwhile. Results from the two instruments gave excellent correlation (r > 0.900) for most parameters except MCHC (r = 0.114), basophil and monocyte percentages (r = 0.617 and 0.552, respectively). The reproducibility, repeatability, linearity, carry-over and stability of the Vega were satisfactory; 'flagging' occurred in 31% of routine samples with sensitivity 88.8%, specificity 41.3% and positive predictive value 85.7%. Various flags appeared in 91% (42/46) of cases where blast cells were microscopically identified. In the four remaining cases, CBC anomalies would themselves have justified microscopic examination of a smear. On 'CBC only' mode reagent consumption was significantly reduced. In the laboratory the analyser was best appreciated for its user-friendliness.     

Evaluating the CELL-DYN® 3500 Haematology Analyser in an acute general hospital

The CELL-DYN® (CD3500) 3500 (Abbott Diagnostics Division, Santa Clara, CA, USA) is a multiparameter, automated haematology analyser system capable of producing 22 haematological parameters including a screening five-part differential. The evaluation was performed in a recently expanded acute care general hospital. Studies of linearity, carryover and precision were acceptable and within manufacturer's stated limits. Correlation of CD3500 with manual differential counts were good except for the basophil count. The white cell flagging system had a sensitivity of 83.6% and specificity of 88.9% with 85.6% agreement. The platelet flagging system also had comparable sensitivity and specificity rate of 88.4% and 94.5% and agreement rate of 92.2%. The overall false positive rate and false negative rate for both the white cell and platelet flags combined were 8.6% and 10.3%. All samples provided a differential with no rejection experienced. The MAPSS technology, together with the extended lyse mode and the investigational nine-part differential provides potential for future development in WBC differential.     

Clinical Evaluation of an in Vitro Test for Radiosensitivity of Leukemic Lymphocytes

A recently developed slide-chamber method was used to test the radiosensitivity of blood lymphocytes from 80 patients with chronic lymphocyticor lymphosarcoma-cell leukemia. The objective of this study was to determinewhether these in vitro tests on sensitivity to x-rays had any clinical significance.

Two objective criteria were used to measure the clinical reactions of theleukemic patients. The first was the duration of survival of patients followingthe in vitro test. The second was the minimal leukocyte count of a patientfollowing x-ray therapy; the minimal count was expressed as a percentage ofthe count before therapy.

The in vitro radiosensitivity was measured by the 10 per cent survival timeof lymphocytes irradiated with 1000 r. Blood lymphocytes from non-leukemicindividuals were highly radiosensitive with indices of 1.1 to 2.2 days. Ininitial tests, the lymphocytes of 61 leukemic patients had the same high sensitivity to x-rays as lymphocytes from non-leukemic individuals. In contrast, thelymphocytes of 19 leukemic patients were radioresistant to irradiation withindices of 2.5 to 11 days.

The 61 patients with radiosensitive lymphocytes had a median survivaltime of 22 months after the in vitro test. In contrast, the 19 patients with radioresistant lymphocytes had a median survival time of only 4 months. Clinicalx-ray therapy caused a greater decline in leukocyte counts in patients withradiosensitive lymphocytes than in those with radioresistant cells.

A significant index of 0.60 was obtained for the correlation of in vitro radiosensitivity of lymphocytes and the in vivo decrease in leukocyte counts ofpatients after x-ray therapy.

It is concluded that an in vitro finding of radioresistant lymphocytes iscorrelated with a poor response of the leukocyte count to x-ray therapy anda short survival time of the patient.

Submitted on January 15, 1962 Accepted on June 20, 1962     

Patterns and Prognostic Value of Total and Differential Leukocyte Count in Chronic Kidney Disease

Summary

Background and objectives

The purpose of this study was to evaluate the levels and patterns of total and differential leukocyte counts and their prognostic importance in a cohort of people with and without chronic kidney disease (CKD).

Design, setting, participants, & measurements

Among 153 veterans without CKD and 267 with, blood leukocyte count was measured at baseline and then repeatedly over a decade. The patterns of change in leukocyte count between the two groups were compared. In the CKD cohort, the spikes in leukocyte counts were compared to the combined endpoint of ESRD and death.

Results

Patients with CKD had more granulocytes and eosinophils and fewer lymphocytes. Over time, granulocytes increased and lymphocytes decreased in those with and without CKD. In addition, in those with CKD, over time eosinophils fell and monocytes increased. Compared with their non-CKD counterparts, patients with CKD had between 1.5- and 3.0-fold more spikes in leukocyte counts. Independent risk factors for the combined endpoint were associated with spikes in the leukocyte counts of absolute and percent eosinophil count, percent granulocyte, and percent monocyte counts. In a multivariate adjusted joint model, both granulocyte and monocyte spikes were independently associated with ESRD and death (hazard ratio 1.67 and 1.52 respectively, P < 0.05).

Conclusions

Compared with those without CKD, patients with CKD have more eosinophils and granulocytes and fewer lymphocytes. Greater variation in leukocytes is seen. Spikes in granulocyte and monocyte percentages among patients with CKD are of independent prognostic importance.     

Blood Leukocyte Count on Admission Predicts Cardiovascular Events in Patients with Acute Non-ST Elevation Myocardial Infarction

We aim to test the hypothesis that blood leukocyte count adds prognostic information in patients with acute non–ST-elevation myocardial infarction (non-STEMI).A total of 585 patients with acute non-STEMI (thrombolysis in myocardial infarction risk score ≥ 3) were enrolled in this cohort retrospective study. Blood leukocyte count was measured immediately after admission in the emergency department. The composite of death, reinfarction, urgent revascularization, and stroke during hospitalization were defined as the primary end point of the study.The mean age of the patients was 61 ± 9.6 years and most of them were male (79%). Using multivariate Cox regression analysis involving seven variables (history of smoking, hypertension, heart rate > 100 beats/minute, serum creatinine level > 1.5 mg/dL, blood leukocyte count > 11,000/µL, use of β-blocker, and use of angiotensin-converting enzyme inhibitor), leukocyte count > 11,000/µL demonstrated to be a strong predictor of the primary end point (hazard ratio = 3.028; 95% confidence interval = 1.69–5.40, p < 0.001).The high blood leukocyte count on admission is an independent predictor of cardiovascular events in patients with acute non-STEMI.     

Elevated cerebrospinal fluid leukocyte count and protein concentration at diagnosis: independent risk factors in children with acute lymphoblastic leukemia

Summary The aim of this study was to investigate whether determination of the initial cerebrospinal fluid (CSF) protein concentration and leukocyte count in children with acute lymphoblastic leukemia (ALL) could yield useful information about the patient's central nervous system status and prognosis. The population-based unselected series comprised 160 children. The mean follow-up time was 72 months (range 25–143 months). Both the CSF protein concentration and the leukocyte count, if elevated, were significantly, although not independently, associated with diminished probability of event-free survival. The patients were divided into three groups for the final analyses: those without any abnormalities in the CSF (n=133), those with elevated protein concentration and/or elevated leukocyte count, but with no malignant lymphoblasts in the CSF (n=21)), and those with malignant lymphoblasts in the CSF (n=6). The probabilities of 5-year event-free survival for the first and second group were 65% and 15%; the probability of 2-year event-free survival for the third group was 17%. These differences were statistically significant (p<0.001). In multivariate analysis the relative risks of death or relapse for these groups were 1, 2.8 (95% confidence limits 1.5–4.9), and 7.6 (2.4–24.3), respectively (p<0.001). The inclusion of an elevated CSF protein concentration or leukocyte count in the risk group criteria of further trials should be considered.     

全自动仪器法与人工镜检法测定尿中白细胞结果的比较分析

目的：比较全自动尿液分析仪（干化学法）、全自动尿沉渣分析仪（尿沉渣法）与人工镜检法检测尿中白细胞的结果。方法分别采用干化学法、尿沉渣法和人工镜检法对采集的845份尿标本进行白细胞测定,并对结果进行比较分析。结果以人工显微镜检测结果作为金标准,845份受检尿液中,干化学法分析尿液白细胞阳性符合率为89.69％,阴性符合率为66.82％,假阳性率为33.12％,假阴性率为10.31％,阳性似然比为2.71,阴性似然比为0.15,总符合率为72.07％,两者结果比较差异有统计学意义（ P＜0.01）;尿沉渣法分析尿液白细胞阳性符合率为95.87％,阴性符合率为83.56％,假阳性率为16.44％,假阴性率为4.13％,阳性似然比为5.80,阴性似然比为0.05,总符合率为86.39％,两者结果比较差异有统计学意义（ P＜0.01）;干化学法与尿沉渣法比较阳性符合率为89.41％,阴性符合率为76.81％,假阳性率为23.19％,假阴性率为10.59％,总符合率为81.18％,阳性似然比为3.85,阴性似然比为0.13,两者结果比较差异有统计学意义（ P＜0.01）。结论尿液中有形成分复杂,全自动仪器法不能完全取代显微镜镜检,只是大批量标本分析时的一种过筛手段,对检测结果阳性者仍需用显微镜复检,以提高报告结果的准确性。三种方法联合应用,可以提高检测结果的准确度,具有较高的临床应用价值。     

Leukocyte Cell Population Data for Hematology Analyzer-Based Distinction of Clonal-versus-Non-Clonal Lymphocytosis: A Real-World Testing Experience

Automated blood counts revealing lymphocytosis necessitate smear reviews. Even expert morphological evaluation may however, fail to differentiate a benign-versus-malignant etiology without further testing. Automated analyser-derived quantitative data on leukocyte cell populations remain undertested for distinguishing such etiologies. Instrument manufacturers claim that if successful, they may be used to generate software flags that help under-resourced laboratories better triage hemogram specimens requiring further testing. We tested the diagnostic accuracy of volume-conductivity-scatter (VCS) indices together with complete blood count (CBC) parameters in such scenarios. We compared LH780-derived (Beckman Coulter, FL, USA) CBC + VCS parameters from patients with clonal lymphoproliferations (n = 42, including 30 chronic lymphocytic leukemia cases) versus 83 controls with absolute or relative lymphocytosis (derivation cohort). Diagnostic performances of 11 logistic regression equations derived were subsequently evaluated on two specific validation cohorts (n = 130 and n = 1465). Clonal lymphocytoses showed significantly lower hemoglobin and higher leukocyte counts but similar lymphocyte percentages (LY %) vis-à-vis controls. The most significant, albeit overlapping predictor of clonality was the absolute lymphocyte count, LY# (47.8 ± 48.4 × 10⁹/L vs. 2.9 ± 1.4 × 10⁹/L in clonal vs. benign cases). In eleven logistic regression equations constructed using four combinatorial approaches, only the models with LY# (highest sensitivity/specificity of 99.3%/100%) and the lymphocytic VCS parameters alone (highest sensitivity/specificity of 76.2%/90.2%) performed consistently in both validation cohorts. Lymphocytic VCS parameters were moderately successful in distinguishing benign-versus-malignant lymphocytes. Other approaches of CBC-plus-VCS parameters did not sustain their initial excellent performances in the validation cohorts, highlighting a need for careful appraisal and better standardization of automated cellular analysis technologies.     

Comparative evaluation of differential leukocyte counts by Coulter VCS cytometer and direct microscopic observation

Comparative analyses of the leukocyte differential counting were performed using a Coulter VCS Hematology Flow Cytometer and direct microscopic observation on 547 unselected individuals analyzed at the outpatient clinic of the Institute of Hematology "L. e A. Seràgnoli" of Bologna. The Coulter VCS is able to provide leukocyte differential counting by measuring cell volume, conductivity and laser light scatter. Negative (true negative: 50.8%) results were observed in 278 subjects by both automated and direct observation methods. Abnormal cells in the blood samples (true positive results) were detected both by VCS and by direct observation in 199 cases (36.4%). Regional distribution of the cells in specific "flags" on the scatterplot was often associated with specific cell types. In 66 cases (12.1%) no abnormal cell was detected by direct observation, while Coulter VCS gave an abnormal pattern, even if only to a slight extent (false positive cases). In 4 cases only (0.7%), false negative results were given by the VCS system. The correlation of the results given by the VCS system with those given by direct microscopic analysis was very high; however careful control by the operator was essential in evaluating the data given by the automated system and in identifying the type of abnormal cells detected.