Parallel processing and maintaining adequate alignment between instruments and methods

Abstract

Parallel processing of laboratory tests across more than one instrument platform: permits dealing with increasing workloads; but broadens uncertainty of measurement; minimising measurement uncertainty means keeping assay performances continuously aligned. Important questions are: Why is there the need to demonstrate "acceptable alignment" between methods/instruments? What methods/tools can be used to test method/instrument alignment and how adjustments can be made? What is an "acceptable" alignment? How often should alignments be checked and what is the reasoning for this?

Figure 1

Parallel analysis of calcium across three Roche Modular Systems shown in Levey-Jennings plots. The same targets (fixed mean calcium: 2.09 mmol/L) and standard deviation limits (SD: 0.05 mmol/L) apply. The cumulative mean calcium varies between 2.05 and 2.06 mmol/L across the three analysers.

Figure 2

A sample report from an Internal Quality Assessment Program. The performance of a Regional Laboratory (Roche Integra) is shown versus the Central Laboratory (Roche Modular analyser). All analyte deviations are normalised in terms of the RCPA QAP Allowable Limits of Performance (ALP’s) with acceptable performances being within ±1 ALP (eg. For chloride, 1 ALP = 3mmol/L and the Regional labs deviation from target is 4 mmol/L. This normalises to 4/3 = 1.33 ALP’s, and flags the analyte as being outlying). Note two outliers are highlighted: bicarbonate and chloride.

Figure 3

Internal Quality Assessment Program and the long term performance at one Regional Laboratory for alkaline phosphatise. In the Levey Jennings type-plot, the Regional Lab’s ‘deviations from target’ for each sample have been normalised in terms of RCPA-QAP Allowable Limits of Performance (ALP). Acceptable performances are within ±1 ALP.

Figure 4

Royal College of Pathologists of Australasia Chemical Pathology Quality Assurance Programs Pty. Ltd. Subgroup report. Results for all the participating laboratories are plotted against sample concentration and Allowable Limits of Performance. Laboratories in the subgroup (your network of laboratories) are highlighted in bold.

Figure 5

Assessment of the alignment of two Architect i4000 analysers measuring luteinising hormone using EQA material. A, Comparison of methods - Regression analysis. B, % Difference Plot, highlighting an average bias of ~ 2.9%.