COST ANALYSIS OF TESTING ALGORITHMS FOR ROCHE AMPLICOR HIV-1 MONITOR STANDARD AND ULTRASENSITIVE: ACHIEVING ACCURACY AT LOWEST COST

CH Sherlock^1,3, D Guh², JSG Montaner^2,3, MV O’Shaughnessy^1,2,3, AH Anis^2,3
Department of Laboratory Medicine, Canadian HIV Trials Network and BC Centre for Excellence in HIV/AIDS, Providence Health Care, Vancouver, British Columbia

Background: The Roche Amplicor HIV-1 Monitor assays for plasma viral load (pvl) are limited by their linear ranges to 400-750,000 (Standard [SD]) or 50-100,000 (Ultrasensitive [US]) copies/ml. Providing an accurate number often necessitates using both tests for single samples. We modelled different potential testing algorithms using our test population to determine costs of different test strategies.
Methods: 16,095 pvl samples from 3300 patients were analyzed (1 year’s testing). Strategies evaluated were: 1. Use SD, re-test with US if <400 copies/ml; 2. Use US, re-test with SD if >100,000 copies/ml; 3. Use SD if not on ART and re-test with US if <400 copies/ml; use US if on ART and re-test with SD if >100,000 copies/ml; 4. Use SD if the previous test was >100,000 and re-test with US if <400 copies/ml; use US if previous test result was <100,000 and re-test with SD if >100,000 copies/ml; 5. Use US only and report results >100,000 as such. Costs of each strategy were calculated per final test result and for total annual cost of 17000 tests at $87.83 per test.
Results: Repeat tests (%), cost per final result and total annual cost for 5 strategies were: 1. 49%, $130.87, $2,224,790; 2. 12.2%, $98.55, $1,675,350; 3. 6.8%, $93.81, $1,594,770; 4. 5.2%, $92.35, $1,569,950; 0%, $87.83, $1,493,110.
Conclusions: Strategy 5 (test only with US) was found to be the least costly option. Although with varying numbers and mixes of patients, the cost per strategy may change somewhat, our overall conclusions are robust given the testing strategy. Appropriate modeling will be critical to providing the correct answer for each testing program.