Changing cost of English HIV service provision 1996-1997. NPMS Steering Group. National Prospective Monitoring System.
The objectives of this study were to provide individual and population-based unit cost estimates of HIV treatment and care by stage of HIV infection for adults in England and estimate the financial impact of the use of combination antiretroviral therapy. Individual unit cost estimates were calculated, based on 1997 activity data, and linked to the number of diagnosed HIV-infected individuals using statutory medical services by clinical stage of HIV infection in England during 1997 to obtain population-based cost estimates; these were compared with 1996 estimates. Most clinical guidelines now recommend the use of 3 antiretroviral agents, but cost estimates for mono and dual therapy were included as baseline estimates. Baseline costs for treating AIDS patients with zidovudine (AZT) monotherapy were estimated at pound sterling 16,830 (95% CI 14,633-18,985) per patient-year which was substantially lower than the 1996 estimate; costs for asymptomatic individuals and people with symptomatic non-AIDS were pound sterling 4450 (95% CI 3521-5612) and pound sterling 7289 (95% CI 6169-8386) per respective patient-year which did not differ substantially from 1996. The total annual population cost estimate for HIV service provision amounted to pound sterling 128 million (95% CI pound sterling 109m to pound sterling 147m), if all patients with HIV disease were treated with AZT monotherapy only. For all eligible patients to be treated with 2 nucleoside reverse transcriptase inhibitors (NRTI) (AZT and didanosine (ddI) or zalcitabine (ddC)), cost estimates amounted to pound sterling 161m (95% CI pound sterling 141m to pound sterling 181m), while for triple therapy, annual estimated expenditure amounted to pound sterling 185m (95% CI pound sterling 165m to pound sterling 206m) when a non-nucleoside reverse transcriptase inhibitor (NNRTI) (nevirapine) was included or pound sterling 205m (95% CI pound sterling 186m to pound sterling 235m) when a protease inhibitor was included. Compared with 1996 population-based cost estimates, the estimates for monotherapy decreased by 14%, by 11% for dual therapy, by 10% for triple therapy which included a NNRTI and by 9% if a protease inhibitor was used as part of a triple therapy regimen. Similarly, compared with 1996 estimates, the proportion of total costs attributable to treating asymptomatic individuals increased by 5% and 2-3% for people with symptomatic non-AIDS, while the proportion attributable for treating people with AIDS decreased by 8-9%.