Evaluation of the average of patients: application to endocrine assays

Assessment of average of normals (AON) procedure for outlier-free datasets including qualitative values below limit of detection (LoD): an application within tumor markers such as CA 15-3, CA 125, and CA 19-9

Average of normals (AON) is a quality control procedure that is sensitive only to systematic errors that can occur in an analytical process in which patient test results are used. The aim of this study was to develop an alternative model in order to apply the AON quality control procedure to datasets that include qualitative values below limit of detection (LoD). The reported patient test results for tumor markers, such as CA 15-3, CA 125, and CA 19-9, analyzed by two instruments, were retrieved from the information system over a period of 5 months, using the calibrator and control materials with the same lot numbers. The median as a measure of central tendency and the median absolute deviation (MAD) as a measure of dispersion were used for the complementary model of AON quality control procedure. The u_bias values, which were determined for the bias component of the measurement uncertainty, were partially linked to the percentages of the daily median values of the test results that fall within the control limits. The results for these tumor markers, in which lower limits of reference intervals are not medically important for clinical diagnosis and management, showed that the AON quality control procedure, using the MAD around the median, can be applied for datasets including qualitative values below LoD.

Quality control of automated cell counters

The hematology analyzers of today provide more reproducible analyses compared with those of a few decades ago, necessitating an evolution in hematology quality practices. The improved performance allows use of simple quality control rules. This improved performance also renders the repeat analysis of critical value specimens non value added. Some of the patient averaging techniques have become outmoded and need to be replaced by less complicated calculations but with truncation of predictable outlying populations. The ready availability of comparative peer quality control values helps to improve analyzer precision and accuracy, and simplifies instrument validation and calibration.

Laboratory quality control: using patient data to assess analytical performance

Quality control plays a vital role helping to ensure the reliability of laboratory test results. The application of statistical quality control has been a component of laboratory medicine for approximately 50 years. Many of the control rules based on the early applications of statistical quality control have remained essentially unchanged since their initial introduction. Optimization of quality control rules can vary depending on the application for which a test is to be used. This review explores the various applications of laboratory quality control procedures and their role in identifying laboratory error. The ubiquitous use of computers in today's laboratories has enabled the development of more sophisticated means of assessing laboratory quality. The use of the Six Sigma technique and its adoption by the laboratory community is one example. Other examples include the use of patient-derived quality control procedures as a means of assessing laboratory performance. Early examples of these types of applications include use of Bull's algorithm, anion gap measurements, and delta checking. More recent applications include the correlation of laboratory test results, the average of normals procedure, and the Bhattacharya method.

Thoughts on quality-control systems: a laboratorian’s perspective

State-of-the-art prospective quality-control systems entail the use of medically relevant, analyte-specific quality control limits. With analyte-specific limits broader than those generally used in the clinical laboratory, there will be fewer false rejections, fewer unnecessary reanalyses, and shorter delays in run reporting. If the analyte-specific limits are narrower than those used in the laboratory, more errors will be detected, but the user is at risk of identifying errors over which s/he and the manufacturer have little control. The use of various patient data quality-control algorithms is described. Conservatism is stressed in adopting manufacturers’ guidelines for surrogate, nondestructive quality-control testing. A simple, optimized approach is suggested for the systematic retrospective review of proficiency data. Finally, an approach is presented for converting from older, previously accepted quality control procedures to more efficient analyte-specific quality control.