Challenging the high volume of smears
The advances in automated blood cell counter successfully provide the precise data of WBC, RBC, and platelet counts with leukocyte differentials. However, the differential information by blood cell counters is limited for cell morphologic features especially for abnormal and immature cells, and the morphological smear reviews triggered by specific flagging of automated analyzer are necessary. The criteria for manual action is depending on local guidelines, and review rates have been reported to vary from 9.9% – 50% in different laboratories 1.
At Juntendo University Hospital, Japan (1,020 beds, 4,000 outpatients/day, 1,500 CBCs/day), manual screens of blood smears were performed on 29.2% of CBC determinations. Under the requirements for faster turnaround times with reduced staff number, guaranty of consistency and quality of manual reviews became difficult. Whereas utilization of the automated image analysis system CellaVision DM96 in the routine workflow is an attractive resolution, the large number of blood smears (approximately 500/day) as well as requirement of fast turnaround time prevents DM96 introduction, which performs only 35 slides/hour analysis.
We therefore challenged to reduce the number of blood smear examination as the preliminary step to introduce DM96 to the routine work.
Material and methods
For this study, 4,319 patient residual K2EDTA samples were randomly selected from Juntendo University Hospital Hematology laboratory (Tokyo, Japan), which receives a high proportion of abnormal samples, after all routine testing had been completed. Samples were selected on different days of the week to mimic averaged 1 day’s workload.
Samples were analyzed by the automated blood cell counter Sysmex XE-5000 (Sysmex) in CBC and DIFF mode and reviewed by direct microscopic observation. Blood smears were made by Sysmex SP100 slide maker, and stained with May–Grunwald-Giemsa.
This study used the anatomized data, and have been approved by the Juntendo Clinic Institutional Review Board, and performed in accordance with institutional guidelines.
Result and discussion
Manual reviews of blood morphology and leukocyte differential counts are time-consuming, require highly trained and experienced technical staff, furthermore, may not always be necessary. For example, we have demonstrated that manual slide observation is less accurate to detect the RBC size abnormalities compared to automated blood cell counters in the previous study.2 The main changes of triggering values in this study were (1) not perform slide review for RBC size abnormalities (2) use company recommended cut-off Q-FLAG values for Blasts and Atypical lymphocytes (for which we have used the lower cut-off value to avoid false negative) (3) not perform slide review for Left shift flagging (4) use absolute numbers for cut-off criteria of mature white cell differential (for which we have utilized both absolute numbers and relative percentages).
With the new criteria, the number of blood smear examination has been significantly reduced (20% decrease of review rates compared to previous criteria) without missing important diagnostic information such as blasts, abnormal/atypical lymphocytes or nucleated red cells. The sensitivity and specificity of the new triggering values referencing to manual reviewed results were 78.9% and 52.6%, respectively. All false negative cases were associated to left shift (The cut-off values of reference; > 0 % for blasts, > 2 % for immature granulocytes, > 20 % for band neutrophils, > 2 % for atypical lymphocytes, > 1 % for nucleated red cells).
The laboratory productivity has been reported to be related with the number of manual smear reviews 1. In Japan, blood smear review rate is generally high, which became a burden under the pressure of efficiency enhancement. Our results indicate that reducing the manual review rates by utilizing CBC results efficiently should be the first step to introduce the automated morphological review system, which must promote the efficiency of hematological examination.
1. Novis DA, Walsh M, Wilkinson D, St Louis M, Ben-Ezra J. Laboratory productivity and the rate of manual peripheral blood smear review: a College of American Pathologists Q-Probes study of 95,141 complete blood count determinations performed in 263 institutions. Arch Pathol Lab Med. 2006;130:596-601.
2. Honma H, Ishii K, Tabe Y, et al. Evaluation of the basic performance and red cell size information of Sysmex XE5000. J Jpn Soc Lab Hematol. 2009; 10: 367-374 (Japanese)