Pb2486: an effective and practical cbc delta check method and criteria for the quality control of automatic hematology analyzers

Topic: 23. Hematopoiesis, stem cells and microenvironment Background: Delta check is an important auto-verification procedure that provides high-quality test results in clinical hematology laboratories; however, no standard guideline for efficient delta check methods and algorithms have yet to be proposed. Aims: To evaluate delta check methods for complete blood count (CBC) parameters and to suggest an effective algorithm for validation of check methods in clinical hematology laboratories. Methods: Consecutive 219,804 CBC samples from 151,120 inpatients and 68,684 outpatients collected from May 2019 to Jan 2020 were included. The 5 delta-check methods include (i) absolute delta difference (ADD): the absolute difference between the current and previous results; (ii) delta percent change (DPC) (%): ADD divided by the previous result; (iii) DPC rate (%/day): DPC/delta interval; (iv) DPC/reference range (DPC/RR) (%): DPC/reference range; and (v) DPC/RR rate (%/day): DPC/RR/delta interval. Delta check was performed for 9 CBC items, including white blood cells (WBC) count, neutrophil%, lymphocyte%, monocyte%, eosinophil%, basophil %, hemoglobin (Hb), mean corpuscular volume (MCV), and platelet count. Measures including sample check, medical record review, peripheral blood smear review, and resampling were taken for the samples with delta check values > 99.5% according to our algorithm (Fig.1). Results: The proportions of samples with delta check values > 99.5% were 0.9%–1.1% across the 9 items in 219,804 samples from 109,902 patients, regardless of the delta check parameters and the patient types (inpatients and outpatients). As it has been suggested that incorporation of the reference range in delta check would better explain changes in the results by reflecting both the biological variations of the test items and the clinical characteristics of patients, DPC/RR rate was further evaluated for 42,652 samples (294,588 tests) from 31,425 inpatients and 11,227 outpatients collected in 2 weeks (from March 25, 2020 to April 7, 2020) for Hb and platelet counts. Among them, 5,008 tests (1.7%) showed DPC/RR rate > 99.5%: 295 (0.10%), 199 (0.07%), 198 (0.07%), 147 (0.05%), 1,163 (0.39%), 899 (0.31%), 976 (0.33%), 327 (0.11%) and 804 (0.27%) samples showed delta check values > 99.5% in WBC count, neutrophil%, lymphocyte%, monocyte%, eosinophil%, basophil %, Hb, MCV and platelet count, respectively, in any of 5 delta check parameters. Retests and resampling were performed in 1,318 and 4 samples, respectively. The reasons for delta check values > 99.5% in 1,780 samples figured out by our algorithm include: transfusion, 1,070 (60.1%); pre-analytical errors, 113 (6.3%); in vitro hemolysis, 1 (0.1%); clot, 3 (0.2%); platelet aggregation, 2 (0.1%); changes in patient’s clinical status (operation, 64 [3.6%]; disease progression 38 [2.1%]); and unknown causes 89 (27.5%). Summary/Conclusion: We developed an effective and practical delta check method incorporating reference range and time interval (DPC/RR rate). Our algorithm would serve as a representative method to investigate the causes of exceeding the delta check criteria and to report correct test results.Fig. 1. Workflow algorithm for the investigation of the specimens exceeding the delta check criteria of hemoglobin and platelet counts Keywords: Automation, Peripheral blood