Late diagnosis, delayed presentation and late presentation among persons enrolled in a clinical HIV cohort in Ontario, Canada (1999-2013).

OBJECTIVES: Timely HIV diagnosis and presentation to medical care are important for treatment and prevention. Our objective was to measure late diagnosis, delayed presentation and late presentation among individuals in the Ontario HIV Treatment Network Cohort Study (OCS) who were newly diagnosed in Ontario. METHODS: The OCS is a multi-site clinical cohort study of people living with HIV in Ontario, Canada. We measured prevalence of late diagnosis [CD4 count < 350 cells/µL or an AIDS-defining condition (ADC) within 3 months of HIV diagnosis], delayed presentation (≥ 3 months from HIV diagnosis to presentation to care), and late presentation (CD4 count < 350 cells/µL or ADC within 3 months of presentation). We identified characteristics associated with these outcomes and explored their overlap. RESULTS: A total of 1819 OCS participants were newly diagnosed in Ontario from 1999 to 2013. Late diagnosis (53.0%) and presentation (54.0%) were common, and a quarter (23.1%) of participants were delayed presenters. In multivariable models, the participants of delayed presentation decreased over calendar time, but that of late diagnosis/presentation did not. Late diagnosis contributed to the majority (> 87%) of late presentation, and the prevalence of delayed presentation was similar among those diagnosed late versus early (13.4 versus 13.4%, respectively; P = 0.99). Characteristics associated with higher odds of late diagnosis/presentation in multivariable analyses included older age at diagnosis/presentation; African, Caribbean and Black race/ethnicity; Indigenous race/ethnicity; female sex; and being a male who did not report sex with men. There were lower odds of late diagnosis/presentation among participants who had ever injected drugs. In contrast, delayed presentation risk factors included younger age at diagnosis and having ever injected drugs. CONCLUSIONS: Late presentation is common in Ontario, as it is in other high-income countries. Our findings suggest that efforts to reduce late presentation should focus on facilitating earlier diagnosis for the populations identified in this analysis.

Mortality is influenced by locality in a major HIV/AIDS epidemic.

OBJECTIVES: The aim of the study was to compare the risks of death among HIV-infected patients on highly active antiretroviral therapy (HAART) in two proximate, yet distinct neighbourhoods: a neighbourhood with a high concentration of gay men, and a neighbourhood with a high concentration of injecting drug users. METHODS: We compared the clinical and socioeconomic characteristics of HIV-infected patients from the two neighbourhoods entering the British Columbia Centre for Excellence in HIV/AIDS Drug Treatment Program from 1 September 1997 to 30 November 2005, using contingency table statistics. Cox survival models and Kaplan-Meier methods were used to estimate the cumulative mortality rates. RESULTS: We found significant differences between patients from the two neighbourhoods for all socioeconomic variables. Patients in the neighbourhood with a high concentration of injecting drug users were more likely to be female, have a history of injecting drug use, have a less HIV-experienced physician and be less adherent. Patients in the neighbourhood with a high concentration of gay men were more likely to have AIDS. Mortality was significantly higher for patients in the neighbourhood with a high concentration of injecting drug users [hazard ratio (HR) 3.01; 95% confidence interval (CI) 1.73, 5.24]. CONCLUSIONS: A threefold increase was observed in the risk of death among HIV-infected individuals on HAART in the neighbourhood with a high concentration of injecting drug users relative to the neighbourhood with a high concentration of gay men. The implications of this study should be assessed in similar HIV/AIDS epicentres.