Objectives: To compare the conventional virus isolation method for diagnosis of influenza infection with reverse-transcription polymerase chain reaction (RT-PCR) in prospectively collected nose and throat swabs from elderly patients during the winter influenza season. The use of a denaturing buffer as an alternative to viral transport medium (VTM) for submission of combined nose and throat swabs to the laboratory for PCR was then investigated in a second study.

Methods: Virus was cultured in microtitre plates using two different cell lines and detected using monoclonal antibody staining. A multiplex, matrix gene PCR assay was optimized to increase the sensitivity and specificity of detection of influenza A (H3 and H1) and B nucleic acid.

Results: The multiplex assay detected all viruses with equal sensitivity to individual assays. In a large, multicentre field study PCR detected twice as many influenza infections compared with virus isolation. No positive culture was missed. PCR has a rapid turn around time (< 36 h) vs. a minimum of 7 days for virus isolation. Greater sensitivity and specificity in the PCR were achieved using a ‘hot-start’ method. Although the numbers were small, the detection rate using PCR was greater for swabs submitted in denaturing buffer than in VTM.

Conclusions: PCR significantly increased the sensitivity and clinical utility of influenza A (H3 and H1) and B diagnosis. there were a number of advantages in using denaturing buffer for submission of samples, including high sensitivity, rapidity, ease of use and no requirement for the virus to be viable on arrival at the laboratory. Therefore, PCR is a rapid, sensitive and user-friendly alternative for influenza diagnosis. Virus isolation technology should be limited to referral centres for further epidemiological characterization.