Quality control of multichannel haematology analysers: critique of current methods and the need for performance goals.

Performance goals for internal quality control of multichannel haematology analysers have been developed by consensus in the Subcommittee on Quantitative Cellular Haematology of the National Committee of Clinical Laboratory Standards (NCCLS). Conventional methods of quality control (i.e. use of standards for machine calibration and control samples to detect drift) are not feasible for the automated complete blood-cell count because standards and generally accepted reference methods for cell enumeration are not currently available. In the absence of an ideal system, a host of different quality control strategies are utilized with fresh whole blood, stabilized blood preparations, commercial surrogate materials, and statistical functions of measurements on patient specimens. All have their theoretical strengths and weaknesses which, up to now, have been difficult to quantify and to compare experimentally between methods and between laboratories. Performance goals, or what have been judged by a consensus of experts to be achievable tolerance limits for the various sources of analytical variations, were determined for the following measurements: haemoglobin concentration, erythrocyte count, leukocyte count, platelet count, and mean cell volume. These performance goals can be used to compare the effectiveness of the various quality control strategies now used and to determine which strategy would perform best and be most cost-effective in a particular laboratory situation.