The effect of churn on "community viral load" in a well-defined regional population.

Background: The concept of community viral load (CVL) was introduced to quantify the pool of transmissible HIV within a community and to monitor the potential impact of highly active antiretroviral therapy (HAART) on reducing new infections. The implications of churn (patient movement in/out of care in a community) on CVL have not been studied.
Methods: The annual CVL was determined in the entire geographic HIV population receiving care in southern Alberta from 2001 to 2010; the CVL for specific subpopulations was analyzed for 2009. CVL was determined for patients under continuous care, newly diagnosed, new to the region, moved away, returned, and lost to follow-up (LTFU). Viral loads (VLs) <50 or <200 copies per milliliter were deemed undetectable and suppressed, respectively. The mean VL per patient and total VL were used to determine CVL.
Results: From 2001 to 2010, the HAART uptake for all patients increased from 62% to 81%, undetectability from 32% to 66%, and suppression from 49% to 72%. The annual total CVL however did not vary significantly after 2003. Incidence rates for new locally diagnosed infections increased from 4.4 to 5.8/100,000 per year. In 2009, newly diagnosed HIV patients (6.6%) contributed 37.5% to the CVL, whereas patients transferring in/out of the region or lost to follow-up contributed 33% to the CVL. Patients in continuous care (79% of all patients) contributed 29.5% to the total CVL.
Conclusions: Increasing HAART coverage did not reduce the CVL or reduce new HIV diagnoses in our population. The effect of churn significantly limited CVL use as a measure for evaluating the impact of HAART in reducing HIV transmissions in our population.