Outcomes of Fourth-Generation HIV Screening and Confirmatory Testing at a Tertiary Care Center: Evaluation of 286,425 Patient Samples

Abstract

Objective: Fourth-generation human immunodeficiency virus (HIV) antigen/antibody assays reduce the diagnostic window; however, reactive results necessitate further evaluation. The objective of this study was to characterize the distribution of samples across the stages of HIV screening and confirmatory testing in a high-volume tertiary-care laboratory and to identify confirmed HIV infections. Methods: This retrospective descriptive study analyzed 286,425 unique serum samples from patients aged 18 years or older, collected between May 2024 and January 2026. Initial screening was conducted using fourth-generation HIV antigen/antibody assays on the Atellica IM or ADVIA Centaur systems. Samples demonstrating repeated reactivity underwent further testing with the VIDAS HIV Duo Quick assay. VIDAS-reactive samples were referred to a reference laboratory for supplemental serological testing. HIV-1 RNA testing was performed using the NeuMoDx platform through June 2025 and the ELITe BeGenius platform from July 2025 onward. Data analyses were descriptive. Results: Among the 286,425 patient samples, 4480 tested repeated reactivity on the initial screening platforms (1.56%; 95% confidence interval [CI], 1.52%–1.61%). Of these, 1378 samples (30.76%) remained reactive on VIDAS, while 3102 (69.24%) were VIDAS-nonreactive and did not undergo HIV-1 RNA testing or reference laboratory confirmation. HIV infection was confirmed in 188 of the 1378 referred samples (13.64%; 95% CI, 11.87%–15.57%), including 100 newly diagnosed cases and 88 with a prior HIV diagnosis. Final results for the remaining 1190 samples were unavailable. In total, 5050 HIV-1 RNA tests were performed during the study period. Conclusion: Discordant reactivity was identified between the initial screening platforms and VIDAS. This discordance may be attributable to low pretest probability, variations in antigen/antibody capture designs and analytical thresholds, or nonspecific reactivity. Due to the absence of verification data for VIDAS-nonreactive samples and VIDAS-reactive samples without final results, sensitivity, specificity, and predictive values could not be determined.