The Use and Cost of HIV Service Provision in England in 1996

Objective: The aim of the study was to measure the use and estimate the cost of HIV service provision in England. Design and Setting: Standardised activity and case-severity data were collected prospectively in 10 English HIV clinics (5 London and 5 non-London sites) for the periods 1 January 1996 to 30 June 1996 and 1 July 1996 to 31 December 1996 and linked to unit cost data. In total, 5440 patients with HIV infection attended during the first 6 months and 5708 during the second 6 months in 1996. Main Outcome Measures and Results: The mean number of inpatient days per patient-year for patients with AIDS was 19.7 [95% confidence interval (CI): 13.7 to 25.7] for January to June and 20.8 (95% CI: 15.3 to 26.4) for July to December 1996. The mean number of outpatient visits for asymptomatic patients with HIV infection was 14.8 (95% CI: 11.9 to 17.6) and 13.3 (95% CI: 10.8 to 15.7) for the respective periods and 16.1 (95% CI: 13.21 to 18.97) and 15.7 (95% CI: 11.2 to 20.2), respectively, for patients with symptomatic non-AIDS (i.e. symptomatic patients with HIV infection but without AIDS-defining conditions). Substantial centre-to-centre variation was observed, suggesting that many clinics can continue the shift from an inpatient- to an outpatient-based service. Cost estimates per patient-year for HIV service provision for 1996 varied from Lstg 4695 (95% CI: Lstg 3769 to Lstg 5648) for asymptomatic patients, to Lstg 7605 (95% CI: Lstg 6273 to Lstg 8909) for symptomatic non-AIDS patients to Lstg 20 358 (95% CI: Lstg 17 681 to Lstg 23 206) for patients with AIDS. Conclusions: Different combinations of antiretroviral therapy affect the cost estimates of HIV service provision differently. Anticipated reduction in inpatient-related activity through the increased use of combination antiretroviral therapy will further shift service provision from an inpatient- to outpatient-based service and reduce costs per patient-year. The extent and duration of such effects are currently unknown. The long term effects of combination treatment on the morbidity and mortality patterns of individuals infected with HIV are also currently unknown, as are their implications on the use and cost of HIV service provision. Multicentre databases like the National Prospective Monitoring System (NPMS) will provide healthcare professionals with information to improve existing services and anticipate the impact of new developments.