Background: The Cascade of Care (COC) visualizes stages of HIV care progression within a population. It is predicated on a local population model and thus may not address the impact on the COC of HIV-experienced individuals diagnosed and cared for elsewhere who move into the area.
Methods: All individuals with a confirmed HIV+ test in Calgary, Canada, between January 1, 2006, and January 1, 2013 were included. Individuals were categorized as "local" if diagnosed within the area, or "transfer" if diagnosed elsewhere. Subgroups were separately placed within the COC and then aggregated.
Results: Of 1019 new cases, 47% were transfers. Transfer patients were more likely female (35% vs. 23%; P < 0.01), non-white (61% vs. 46%; P < 0.001), heterosexual (56% vs. 38%; P < 0.001), and have higher CD4 counts (400 vs. 282/mm) with undetectable viremia in 57% [63% on antiretroviral therapy (ART)] at baseline. Engagement was higher at every stage for transfer patients: 94% of transfer vs. 92% of local patients linked to HIV care, 90% vs. 76% (P < 0.001) were retained, 86% vs. 67% (P < 0.001) received ART, and at study's end, 75% vs. 58% (P < 0.001) had undetectable viremia. When patients were aggregated, linkage increased by 1%, retention by 6%, patient use of ART by 8%, and patients with viral suppression by 7%.
Conclusions: The COC of local and transfer patients differs so significantly that both need to be considered separately in measuring COC, adding a previously under-recognized level of complexity. Use of aggregate COC without considering different levels of engagement could lead to imprecise information for public health initiatives and program metrics.