Factors affecting the success of influenza laboratory diagnosis

Abstract

Influenza is one of the most common human infectious diseases, and has profound health and economic consequences. The laboratory diagnosis of influenza virus infections plays an important role in the global surveillance of influenza. Therefore, there is a growing demand for highly sensitive and rapid methods for detecting influenza. The performance of particular diagnostic methods is affected by various factors. In this study, we assess the effects of patients’ age and time to diagnosis on the probability of detecting influenza using four diagnostic methods (virus isolation, rapid test, RT-PCR and real-time RT-PCR). We examined 3,546 samples from central and eastern Slovakia during the influenza seasons from 2005-2006 to 2010-2011. In general, the probability of influenza detection significantly decreased with the time from onset of illness to sample collection (T1) as well as with patients’ age (AGE). On the contrary, time from sample collection to delivery (T2) did not play a role in the probability of influenza detection. As judged by odds ratios, the virus isolation method was most sensitive to T1, followed by the rapid test and RT-PCR methods. For the effect of AGE, the rapid test and virus isolation methods were more sensitive than PCR-based methods. The effects of T1 and AGE were independent of each other. Laboratories which participate in influenza surveillance should use several methods to enable rapid and accurate influenza A and B virus detection.