Diagnostic accuracy and cost analysis of the Alere™ i Influenza A&B near-patient test using throat swabs

Clinical diagnostic sensitivity alone is inadequate in the diagnosis of influenza. Polymerase chain reaction (PCR) testing is sensitive but the inherent delays in result availability potentially prolong time to isolation and treatment. Until recently no near-patient test (NPT) has demonstrated adequate sensitivity for routine clinical use.

To evaluate diagnostic accuracy, time to result availability, clinical impact, and cost consequences of Alere™ i Influenza A&B NPT (Alere Inc., Waltham, MA, USA) using off-label throat swabs.

Prospective, multi-centre [four UK National Health Service (NHS) hospitals], diagnostic accuracy cohort study with cost modelling. Throat swab samples from suspected influenza patients were tested for influenza using the reference standard of PCR; a second throat swab was tested using NPT.

A total of 827 participants were recruited; 589 were suitable for analysis: sensitivity was 75.8% [95% confidence interval (CI): 67.0-84.6]; specificity was 96.8% (95% CI: 95.2-98.3). Sensitivity varied between Sheffield (Northern General Hospital: 82.1%; Royal Hallamshire Hospital: 83.3%) and other sites (Doncaster Royal Infirmary: 71.4%; Newcastle's Royal Victoria Infirmary: 50.0%) whereas specificity was high (92-100%). Positive predictive value (PPV) was 81.2% (95% CI: 72.9-89.5) with negative predictive value 95.6% (95% CI: 93.9-97.4) with observed prevalence of 15.4%. Median time to result for PCR was 1.1 days (on-site laboratories) and 5.2 days (remote laboratories). Isolation findings: 75% influenza positive not isolated; 69% of isolated participants did not have influenza. For a cohort of 1000 participants, annual estimated non-diagnostic cost savings with NPT are £215,040.

This first prospective study of the Alere i NPT using throat swabs demonstrates high specificity, high PPV during seasonal epidemics, and rapid result availability which could lead to substantial cost savings.