Influenza remains a global health threat with an estimated 1 billion cases, 3–5 million severe cases, and 290,000–650,000 influenza-related respiratory deaths worldwide. New quadrivalent SARS-CoV-2 and influenza A/B antigen combination rapid test kits have recently become available in Australia. Limited data exist for real-world rapid antigen test (RAT) use, with data from a 2024 South Korean study comparing diagnostic accuracy of RAT to polymerase chain reaction (PCR) demonstrating sensitivities of 88%, 92%, and 100% for SARS-CoV-2, Flu A, and Flu B, respectively, and universal specificities of 100%. However, a study in Finland in 2024 raised concerns about low sensitivity for RATs, especially at high cycle thresholds (CTs; 25–30) where sensitivity was as low as 68%. Concerningly, data from a United Kingdom primary-care-based study from 2025 demonstrated RAT sensitivities as low as 29% for influenza A and 22% for influenza B. To our knowledge, no data exist regarding RAT performance in the Southern Hemisphere during the 2025 influenza season. We aimed to assess the real-world performance of influenza RATs compared to PCR testing during the 2025 influenza season in Central Australia.
A retrospective analysis was conducted at the only referral centre in Central Australia during the 2025 influenza season, where RATs [TouchBio RSV (respiratory syncytial virus), Flu A/B & COVID-19 Rapid Antigen Test] were deployed to guide early infection prevention and control measures prior to PCR result availability. Education sessions were provided to sample collectors by our infection prevention and control unit. Multiple sessions were held across shifts to ensure the program was delivered to all collectors. Patient episodes with suspected influenza who had received a RAT and a subsequent PCR test (Abbott Alinity m, performed on site) were identified through hospital records, with 683 episodes assessed from 10 June to 10 August 2025. Cases with incomplete documentation were excluded. Given a low number of influenza B (six), COVID-19 (11), and RSV (17) cases across the period, the analysis was restricted to only influenza A. The real-life performance of RATs for influenza A was assessed using PCR results as the standard reference. Sensitivity, specificity, and Student t tests were conducted using R (version 4.5.2; https://www.r-project.org/), as appropriate. Proportions were expressed as percentages.
A total of 683 episodes of acute respiratory illnesses were screened during the study period; of these, 48 episodes had either not documented or not performed RAT or PCR testing and were excluded from assessment. A total of 635 patient presentations where both influenza RAT and PCR were performed were included in this study. A total of 209 of 635 (32.9%) episodes were confirmed PCR positive for influenza A (Table 1), where 59.8% (125/209) were tested RAT negative. In RAT-positive cases, only three of 87 were PCR negative, giving a RAT accuracy, sensitivity, and specificity of 79.8%, 40.2%, and 99.3%, respectively. When comparing CTs of true-positive (RAT+ and PCR+) and false negative (RAT– and PCR+) cases, using the Student t test, the mean (±standard deviation) CT of the true-positive cases was significantly lower (18.7±3.3 vs 26.7±5.5, p<0.001).