|
|
Background: The application of traditional internal quality control processes in molecular testing for personalized medicine often proves difficult due to limited control materials. We established a quality control algorithm independent of control materials to monitor routine clinical test results.
Methods: A quality control algorithm was constructed based on the parameters p, Ppos, Pneg, and Pcon calculated by an elementary probability theory. This quality control algorithm was applied in CYP2C19 genotyping and EGFR mutation testing in two molecular diagnostics laboratories in order to illustrate how probabilities are calculated and test results are analyzed.
Results: For CYP2C19 genotyping, when the p-values based on 151 specimens from 20 runs were used, three runs were out of control within 20 runs, while two of the three runs became acceptable with p-values calculated with 362 samples in 40 runs. For EGFR mutation testing, one run was rejected because of the low Ppos-values of deletion in exon 19 with the p-value based on 421 specimens from 40 runs.
Conclusions: The quality control algorithm can be employed to monitor possible false-positives and false-negatives according to the number and distribution of positive results. Noticeably, the evaluation and interpretation of the control results is essential for the decision by a test laboratory to reject a run.
DOI: 10.7754/Clin.Lab.2015.150428
|
