To address high levels of undiagnosed HIV, the CDC recommends routine HIV screening regardless of risk. While this has been shown to be cost-effective from a societal perspective, little data exists on the real-world costs of broadly implementing programs, including at sites of various sizes.
We performed a cost analysis using an ingredients-based approach for the 3-year implementation period of a multi-site, integrated routine HIV screening program on Chicago’s south side, the largest predominantly African American community in the US. Ingredients included training, labor and tests for the uninsured. Outcomes included median cost/test, cost/new diagnosis cost/patient linked to care, and cost/patient re-linked to care for those previously diagnosed but not in care, and were stratified by academic medical centers (AMC), community hospitals (CH), and community health centers / federally qualified health centers (CHC/FQHC). We performed Kruskall-Wallis rank sum with Dunn’s pairwise tests for significance. All costs are reported in 2013 USD.
Ten sites conducted 75,351 HIV tests and identified 176 new HIV cases over 3 years. Of new cases, 89% were linked to care and 63% achieved virologic suppression. Median annual program cost was $92,292 (range $83,497 (CHC/FQHC) to $119, 399 (AMC) p<0.05). The median cost/test, cost/new diagnosis, cost/linkage, and cost/re-linkage was $40, $12,250, $15,080, and $28,630, respectively. Costs/test and new diagnosis were greatest in CHC/FQHC. Costs/linkage and re-linkage did not significantly differ by site type. Case management accounted for 36% of total costs, followed by administration (35%). Testing costs for the uninsured accounted for 5.1% of costs.
In a large, urban routine HIV screening and linkage to care program, program costs were proportional to testing volume. Linkage to care costs did not differ significantly between site types, suggesting that regardless of volume, smaller sites have capacity to link those patients who are identified to care. The increased cost associated with re-linkage of previously diagnosed patients suggests additional challenges associated with caring for this population.
M. T. Pho,
R. Eavou, None
B. Livak, None
N. Benbow, None
J. Schneider, None
D. Pitrak, Gilead: Grant Investigator , Grant recipient