Abstract Body


PrEP users can have ambiguous HIV test results for detection of primary infection. The CDC 2021 PrEP guidelines recommend HIV RNA testing for PrEP initiation and monitoring. We evaluated occurrence of discordant HIV test results in oral and long-acting cabotegravir (CAB-LA) PrEP users in a large real-world cohort of PrEP users.


We analyzed the HealthVerity HIV cohort with linked longitudinal medical claims, antiretroviral prescriptions, and laboratory testing records. Our primary outcomes were discordant HIV Ag/Ab and RNA test results, new HIV diagnoses, and Long-acting Early Viral Inhibition (LEVI) cases among PrEP users. We used a validated algorithm and identified persons prescribed oral or injectable PrEP and extracted all their laboratory records. We defined a ‘combined test’ as both Ag/Ab and RNA testing within 7 days. We analyzed all tests from 30 days before the first PrEP prescription through 30 days after last prescription. Among persons with discordant results, we reviewed all their laboratory tests, diagnosis records, and ARV prescriptions.


Among 30 548 PrEP users, we identified 9090 combined tests in 5391 individuals; 8995 (99.0%) were same-day tests. HIV Ag/Ab and RNA results were concordant for 9070 (99.8%) tests with 18 dual positives (+) in 17 persons, and 9052 dual negatives (-) in 5374 persons. We identified 14 combined tests in 12 persons with Ag/Ab(-) and RNA(+) results, accounting for 0.15% of all combined tests. Review of records found 4 HIV diagnoses; 4 likely false positive RNA tests (with repeated dual (-) tests and follow-up >30 days); and 4 inconclusive results (insufficient follow-up) (Figure). Excluding inconclusive cases, RNA tests had a false positive rate (FPR) of 0.04% and a positive predictive value (PPV) for HIV infection of 84% among PrEP users. No LEVI cases were observed among 439 CAB-LA users. We also found 6 discordant tests in 4 patients with Ag/Ab(+) and RNA(-) results. Of these, one person had a likely false positive Ag/Ab test (FPR of 0.01%, PPV 94.4%), while 3 persons had insufficient follow-up.


Discordant combined test results among PrEP users were rare and a third of discordant tests identified a new HIV diagnosis, accounting for 19% of all diagnosed HIV infections. Combined testing resulted in similar early HIV diagnoses rates and potential false positive NAT rates. The FPR of Ag/Ab testing was low. Further assessment of predictive values and cost-effectiveness of HIV RNA testing of PrEP users is warranted.