Table 1.
Base case model input values, ranges used in sensitivity analyses and data sources. HAART drug costs reflect generic prices, rather than the higher procurement costs borne by the HBAC program.
| Base case | Range | Source | |
|---|---|---|---|
|
General | |||
| Number of (standardized) clients | 1000 | - | - |
| Population access to HAART | 100% | 50%1 | Assumption |
| Discount rate | 3% | 2% – 4% | World Bank, 1993 |
| Annual rate of regimen change (1st line to 2nd line) | 1.7% | 0% – 15.0% | HBAC program records |
|
Health-related quality of life (“utility”) by WHO stage | |||
| Disablity weight: WHO stage 1 | 0.06 | 95% – 105%2 | Mathers, 2004; Jamison, 2006; Authors’ construction |
| Disablity weight: WHO stage 2 | 0.14 | ||
| Disablity weight: WHO stage 3 | 0.25 | ||
| Disablity weight: WHO stage 4 | 0.51 | ||
|
Program cost per 1000 individuals receiving HAART (Two years) | |||
| Personnel | $545,882 | 50% – 150% | HBAC program records |
| ARV drugs | $306,469 | - | |
| Annual cost of 1st-line regimens at treatment start; per client | $125 | 50% – 150% | MSF, 2007 |
| Annual cost of 2nd-line regimen per client | $1,119 | 50% – 150% | |
| Other recurrent goods | $341,125 | ||
| Recurrent services | $45,640 | - | HBAC program records |
| Equipment | $143,485 | - | |
| Buildings | $4,068 | - | |
|
Cost of care for illness episodes prevented | |||
| Death | $400 | 50% – 150% | Assumption |
| TB episode | $30 | 50% – 150% | HBAC program records |
| Malaria episode | $7.19 | 50% – 150% | HBAC program records |
| Hospitalization (survive) | $31.36 | 50% – 150% | Pitter, 2007 |
|
Health Benefits | |||
| Illness episodes and care averted | |||
| Deaths averted in adults (per 1000 individuals) | 606 | HBAC program records | |
| HAART efficacy: mortality reduction in HIV-neg. children | 79% | 75% – 125% | HBAC program records |
| DALYs averted per childhood death averted (discounted) | 33.9 | Authors’ construction; Murray, 1994 | |
| Life-years added per death averted (undiscounted) | 9.5 | Morgan, 1997; Marseille, 2006 | |
| DALYs averted per death averted (discounted) | 7.6 | 50% – 150% | Calculated from life years |
| TB new active cases (per 1000 individuals) | 52.2 | 50% – 150% | HBAC program records |
| Malaria episodes (per 1000 individuals) | 110 | 50% – 150% | HBAC program records |
| Hospitalizations (survive; per 1000 individuals) | 478.2 | 50% – 150% | HBAC program records |
| Non-mortality related health benefits | |||
| Cases that would not have been fatal absent HAART (per 1000 individuals) | 906 | 50% – 150% | HBAC program records |
| DALYs averted per case | 0.13 | 50% – 150% | Authors’ construction |
| Cost of care for HIV infections prevented | |||
| Adult’s lifetime cost of treatment (discounted) | $7,326 | 50% – 150% |
Morgan, 1997; Marseille, 2006; MSF, 2007 Authors’ construction, HBAC records, MSF, 2007 |
| Infant’s lifetime cost of treatment (discounted) | $6,526 | 50% – 150% | |
| New adult HIV infections prevented | |||
| Infections expected per 1000 HIV- infected individuals per year absent HAART | 45.7 | 50% – 150% | HBAC program records |
| Two-year number expected from 1000 HIV-infected individuals per year | 1.3 | Authors’ construction | |
| Multiplier to reflect additional cases prevented after year 2 | 2 | 50% – 150% | Authors’ construction |
| Year when the non-year 1 HIV infections averted would have occurred | 2 | 50% – 150% | Assumption |
| DALYs averted per HIV infection prevented | 7.2 | 50% – 150% | Authors’ construction |
| Mother-to-child-transmission prevention | |||
| Live births per 1000 female HBAC clients | 32.6 | 50% – 150% | HBAC program records |
| Late background MTCT rate | 0.37 | 50% – 150% | Nduati, 2000 |
| HAART effectiveness for MTCT | 0.85 | 50% – 150% | Palombi, 2006 |
| Infant cases averted | 20.3 | 50% – 150% | Authors’ construction |
| DALYs averted per averted case of infant AIDS (discounted) | 3.4 | 50% – 150% | Authors’ construction |
1
50% population access to HAART used in one-way sensitivity analysis only.
2
This and all remaining ranges below refer to relative percent variation of base case values.