Table 1.
Incremental cost-effectiveness of PrEP strategies over a 5-year time horizon.
| number of new infections | % of new infections averted | number of PrEP usage (person-year) | Discounted QALYG(1) | Plan A | Plan B | Plan C | ||||
|---|---|---|---|---|---|---|---|---|---|---|
| Discounted incremental cost, USD (2) | (2)/(1) | (2) | (2)/(1) | (2) | (2)/(1) | |||||
| basecase | 3450 | / | 0 | |||||||
| non-targeting, 10% | 2590 | 8% | 17959 | 67 | 123458936 | 1842204 | 17294670 | 258064 | 9806914 | 146335 |
| non-targeting, 30% | 2048 | 23% | 53910 | 212 | 370266861 | 1745524 | 51648582 | 243483 | 29176464 | 137545 |
| non-targeting, 90% | 942 | 55% | 161936 | 526 | 1113780354 | 2115619 | 157156635 | 298518 | 89686050 | 170358 |
| targeting, 10% | 2754 | 3% | 7629 | 24 | 52571166 | 2162072 | 7459389 | 306779 | 4277659 | 175926 |
| targeting, 30% | 2470 | 11% | 22896 | 99 | 157200505 | 1583136 | 21831597 | 219862 | 12284040 | 123710 |
| targeting, 90% | 1780 | 31% | 68752 | 287 | 472011282 | 1642874 | 65661580 | 228540 | 37001772 | 128788 |
| test-and-treat | 2075 | 23% | 0 | 98 | 39055533 | 396874 | 39055533 | 396874 | 39055533 | 396874 |
| non-targeting, 10% | 1849 | 29% | 17987 | 170 | 158411503 | 929215 | 52137608 | 305830 | 44642119 | 261863 |
| non-targeting, 30% | 1463 | 40% | 53985 | 296 | 398568822 | 1345390 | 79665459 | 268915 | 57173233 | 192991 |
| non-targeting, 90% | 673 | 63% | 162094 | 568 | 1127434311 | 1985645 | 170226003 | 299803 | 102714188 | 180901 |
| targeting, 10% | 1964 | 26% | 7643 | 134 | 89432361 | 668940 | 44267840 | 331116 | 41082391 | 307290 |
| targeting, 30% | 1760 | 32% | 22934 | 199 | 190572365 | 956132 | 55056540 | 276227 | 45498622 | 228274 |
| targeting, 90% | 1266 | 46% | 68847 | 363 | 496573340 | 1366821 | 89865774 | 247356 | 61180726 | 168400 |
Plan A – market price for PrEP drug (annual cost of USD7880 at the end of 2017); Plan B – generic price for PrEP drug (annual cost of USD519); Plan C – zero cost for PrEP drug. All PrEP users are assumed to be in high adherence with an average of 87.5% usage per year. As we assume that 20% of high adherence users would change to low adherence users, whereas 10% of low adherence users would change to high adherence users in a year, a proportion of PrEP users would be in low adherence, with an average of 38% usage per year.
(2)/(1) = Discounted incremental cost-effectiveness (incremental $/QALYG).
Non-targeting – low-threshold approach with PrEP for both low- and high-risk MSM; targeting – PrEP for high-risk MSM only.
QALYG – quality-adjusted life-years gained.