Internal quality control and external quality management of data in practice.

In this chapter, the use of control charts to both continuously evaluate testing in individual laboratories as well as provide data for external monitoring is examined in detail. The data is based on the publication by D. E. Rebeski' et al., "Charting methods to monitor the operational performance of ELISA method for the detection of antibodies against trypanosomes" in Veterinary Parasitology, 2001, 96, 11-50, and is a detailed example of the investigation of the performance of four indirect ELISAs for the detection of antibodies against trypanosomes using Trypanosoma congolense and T. vivax antigen-precoated plates in 15 veterinary diagnostic laboratories in Africa and Europe. The study shows the practical use of charting methods with respect to assessing the operational performance of each ELISA. Data from standardized internal quality control (IQC) samples were used to assess ELISA performance indicators with reference to expected upper and lower control limits, as determined from studies by the kit producer (tentative values). Based on unprocessed (optical density) and normalized absorbance values (calculated as a percentage positivity of a control), dispersion of values from the expected data range were estimated though plotting the location and deviation of the values. In addition, assay precision was estimated by plotting the distribution of coefficients of variation <10% of the IQCs. Binding ratios of various controls were calculated to estimate the assay proficiency with respect to the accuracy of assessing whether the IQC samples tested positive or negative in the test proper. The graphical analysis of dispersion of absorbance values in combination with assay precision and proficiency criteria were considered satisfactory to allow the evaluation of the operational performances of the ELISAs, and provide useful decision-making criteria for plate acceptance and rejection. The establishment of standardized and transparent IQC data charting methods for the indirect ELISAs provided an increased measure of confidence to national laboratories with respect to their reports on disease occurrence. Moreover, the relative assay performances among all laboratories were examined, using summary data charts, with reference to the performance criteria described. The IQC data were also examined using modified Youden plot analysis, demonstrating that indirect ELISA methods can be successfully applied at diagnostic laboratories in the tropics for monitoring trypanosomosis control programs.