Use of Patient Samples for Quality Control in Hemogram

Abstract

Background: Quality control (QC) is an integral part of hematology practice. Various commercial controls are available and routinely used. A complementary method is to use patient sample in various ways especially in a resource limited setup. Methods: This study was performed in a tertiary care hospital in southern India with 12 laboratories, four using hematology cell counters and eight other smaller side-labs performing Hemoglobin (Hb) estimation by Sahli's or Drabkin's method. The QC tests performed using patient samples were: a. Daily - Average of numbers (AON) of RBC indices of 250-270 samples. These were compared with Levey-Jennings (LJ) charts of commercial controls b. Weekly - Replicate Test (RT) using a single sample for various levels of Hb, WBC and platelet counts; Duplicate Test (DT) using 10 samples; Correlation Check (CC) of 10 samples and Weekly AON c. Monthly - Deviation Index (DI) using one sample for inter-method/ instrument comparability among participating labs. Result: In daily AON, the graphs of RBC indices did not show significant difference with LJ charts of commercial controls. The co-efficient of variation in RT for various levels of Hb, WBC and platelet counts were in the acceptable range. The precision was higher for Hb and least for low platelet counts. In DT, an occasional random error was detected. In CC, monocytosis flag did not correlate with peripheral smear findings 66.6% times. The DI using z-score showed satisfactory performance among various labs. Conclusion: Use of patients' blood sample was helpful for internal QC and inter-method/ instrument comparability.