To estimate time from HIV infection to linkage-to-care and its determinants. Linkage-to-care is usually assessed using the date of HIV diagnosis as the starting point for exposure time. However, timing of diagnosis is likely endogenous to linkage, leading to bias in linkage estimation.
We used longitudinal HIV serosurvey data from a large population-based HIV incidence cohort in KwaZulu-Natal (2004–2013) to estimate time of HIV infection. We linked these data to patient records from a public-sector HIV treatment and care program to determine time from infection to linkage (defined using the date of the first CD4+ cell count).
We used Cox proportional hazards models to estimate time from infection to linkage and the effects of the following covariates on this time: sex, age, education, food security, socioeconomic status, area of residence, distance to clinics, knowledge of HIV status, and whether other household members have initiated antiretroviral therapy.
We estimated that it would take an average of 4.9 years for 50% of HIV seroconverters to be linked to care (95% confidence intervals: 4.2–5.7). Among all cohort members who were linked to care, the median CD4+ cell count at linkage was 350 cells/μl (95% confidence interval: 330–380). Men and participants aged less than 30 years were found to have the slowest rates of linkage-to-care. Time to linkage became shorter over calendar time.
Average time from HIV infection to linkage-to-care is long and needs to be reduced to ensure that HIV treatment-as-prevention policies are effective. Targeted interventions for men and young individuals have the largest potential to improve linkage rates.
Supplemental Digital Content is available in the text
aDepartment of Epidemiology, Biostatistics, and Occupational Health, McGill University, Montréal, Québec, Canada
bAfrica Health Research Institute (AHRI), Somkhele
cSchool of Nursing and Public Health
dCentre for the AIDS Programme Research in South Africa – CAPRISA, University of KwaZulu-Natal, Durban, South Africa
eDepartment of Infectious Disease Epidemiology, Imperial College London, St Mary's Hospital
fDivision of Infection and Immunity, University College London, London, UK
gDepartment of Epidemiology and Public Health, Swiss Tropical and Public Health Institute
hUniversity of Basel, Basel, Switzerland
iDepartment of Global Health & Population, Harvard T.H. Chan School of Public Health, Boston, Massachusetts, USA
jInstitute of Public Health, University of Heidelberg, Heidelberg, Germany.
Correspondence to Mathieu Maheu-Giroux, ScD, Department of Epidemiology, Biostatistics, and Occupational Health, McGill University, Montréal, Canada. E-mail: firstname.lastname@example.org
Received 20 September, 2016
Revised 17 January, 2017
Accepted 23 January, 2017
Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Website (http://www.AIDSonline.com).