Inconsistent results in real-time PCR in analyzing samples from patients with suspected bacterial meningitis

Abstract

The basic principle for obtaining a reliable result is the consistency found among the replicas and its reproducibility. In a diagnostic routine by using real-time PCR (RT-PCR), when hundreds samples are processed, and results with late Cts or replicas that differ by more than three units are inevitable. Of 3,000 samples processed in 2010, in duplicate, in the diagnostic routine of bacterial meningitis by RT-PCR for detecting N. meningitidis, S. pneumoniae and H. influenzae, 157 (5.2 %) samples showed inconsistency among replicas (difference between Cts higher than three units) and/or high Cts values; and the samples were retested. This study assessed these results, the benefits of its repetitions and probable reasons for the occurrence of these discrepant results. Among the retested samples, 18 (11 %) only showed positive results. Human errors inherent to pipetting, use of non-calibrated pipettes, low concentration of target DNA in the analyzed samples, probe degradation or even random contamination are factors which contribute to induce the discrepant results. Performing RT-PCR assay with samples in duplicate and retesting the samples showing discordant results constitute a device for efficiently evaluating and defining these results.