PurposeThe goal of 90-90-90 first requires the expansion of access to HIV testing. Our aim was to record frequencies of HIV indicator conditions (ICs) and identify missed opportunities for an early HIV diagnosis.MethodsWe retrospectively identified ICs in a population of 231 people living with HIV with known infection dates who attended our clinic. The study population was divided into four groups: (1) those self-tested pre-emptively (47/231, 20.3%), (2) those offered targeted testing based on risk factors (67/231, 29%), (3) those tested after an IC (73/231, 31.6%) and (4) those who were not offered testing after an IC (44/231, 19%). HIV acquisition dates were estimated by molecular clock analysis.ResultsA total of 169 healthcare contacts (HCCs) were recorded. The most frequent HCC was mononucleosis-like syndrome (20.1%), unexplained weight loss (10.7%) and STIs (10.1%). AIDS-defining conditions were detected in 11.8%. Only 62.4% (73/117) of those with an IC were offered testing after their first HCC. Patients in group 4 had statistically significant delay in diagnosis compared with group 3 (109.1 weeks (IQR 56.4–238.6) vs 71.6 weeks (IQR 32.3–124.6)). The proportion of patients diagnosed as late presenters in each group was: (1) 16/47 (34%), (2) 37/67 (55.2%), (3) 43/73 (58.9%) and (4) 27/44 (61.4%) (p=0.027).ConclusionsOur study uses a combination of molecular and clinical data and shows evidence that late presentation occurs in a high proportion of patients even in the presence of an IC. Given that risk-based targeted testing has low coverage, IC-guided testing provides a reasonable alternative to facilitate earlier HIV diagnosis and to improve late diagnosis across Europe and globally.