Comparative Evaluation of Three Rapid Influenza Diagnostic Tests for Detection of Influenza A and B Viruses Using RT-PCR as Reference Method.

Abstract

To evaluate the performance of three rapid influenza diagnostic tests (RIDTs) for detecting influenza A and B viruses compared to RT-PCR. A total of 291 subjects with acute respiratory infections were enrolled. Respiratory specimens were collected and tested for influenza A and B viruses using three RIDTs. The results were compared with those obtained using an RT-PCR assay from Shanghai Berger Medical Technology Co. Ltd. as the reference method. Among the 291 subjects, 119 (40.9%) tested positive for influenza A virus and 38 (13.1%) for influenza B virus by RT-PCR. The sensitivities of the three RIDTs for influenza A virus were 92.4%, 89.1%, and 79.8%, respectively, while their specificities were 98.8%, 98.8%, and 100%, respectively. For influenza B virus, the sensitivities of the three RIDTs were 92.1%, 92.1%, and 73.7%, respectively, and their specificities were 100%, 100%, and 100%, respectively. The positive predictive values (PPVs) for influenza A virus were 98.2%, 98.1%, and 100%, respectively, while the negative predictive values (NPVs) were 94.5%, 92.4%, and 86.9%, respectively. For influenza B virus, the PPVs were all 100%, and the NPVs were 99.2%, 99.2%, and 97.7%, respectively. The three evaluated RIDTs demonstrated high specificity but varied sensitivity for detecting influenza A and B viruses. Negative results from RIDTs should be confirmed by RT-PCR, especially during peak influenza seasons. The high PPVs suggest that positive RIDT results are reliable for influenza diagnosis, while the high NPVs indicate that negative results are more likely to be true negatives.