The use and cost of HIV services and changing mortality patterns for AIDS patients treated at St. Mary's Hospital, London 1982-1991
File(s)
Author(s)
Beck, Eduard Jan
Type
Thesis
Abstract
Aim of this thesis was primarily to describe the changes in the use and cost of HIV service provision within the context of changing mortality and morbidity patterns for an open cohort of HIV infected individuals managed at St.Mary's Hospital, London, between 1982 and 1991. Background to these investigations was the improvement in survival time from diagnosis of AIDS observed at St.Mary's, England and in other industrialized nations since 1987, and associated with a reduced mortality in patients with Pneumocystic carinii Pneumonia (PCP). The latter has often been attributed to the introduction of zidovudine (AZT) or PCP prophylaxis into routine clinical practice. The background studies to this thesis themselves pointed to a contribution made by AZT but also suggested the importance of the severity of PCP among patients diagnosed with their first episode of PCP. Methods involved the analyses of observational data on the use, cost and outcome for an open cohort of HIV infected individuals managed at St.Mary's Hospital, London, 1982 1991. All activity and case-severity data were collected retrospectively from case-notes; cost data were obtained through a costing study of the 37 departments involved with HIV service provision. The Use and Cost Studies were based on 459 HIV infected individuals generating 429 asymptomatic patient-years, 132 symptomatic non-AIDS patient-years and 388 AIDS patient-years of follow-up respectively; the two PCP studies were based on 211 and 159 subjects respectively. Results Patients with AIDS used more hospital services compared with those with asymptomatic HIV infection or symptomatic non-AIDS disease and average costs per patient-year reflected similar differences: £14,000 per AIDS patient-year, which ranged from £6,000 to £32,000 per patient-year for different AIDS defining conditions {1989/1990 prices). A shift from a predominant inpatient to an outpatient-based service for HIV infected individuals was observed especially among AIDS patients. Annual costs per patient-year either decreased or remained stable despite the introduction and increased use of drugs, including AZT. Drug-costs increased from 9% to 29% of total costs per patient-year for people with symptomatic HIV disease, but these were off-set by reduced inpatient cost due to a reduction in care costs. Increased survival from time of AIDS was associated with earlier patient presentation. Total lifetime treatment costs were estimated at £80,000 (1992/93 prices) , 60% directly attributable to hospital care. Patients diagnosed with their since 1987 presented with less severe PCP compared with those before 1987. The number of individuals who survived their first episode of PCP increased during the study period. Younger patients who presented with an index diagnosis of PCP, had no other concurrent AIDS defining conditions and whose HIV status had been known for more than two years were more likely to survive. This was especially so if their haemoglobin levels were greater than 12 gm/dl and A-a gradient less than 40 mmHg of their Survival after first episode of PCP was more likely in younger patients with fewer pre-existing AIDS diagnoses, who were regularly seen in outpatients after their first episode of PCP, had less severe immunosuppression and had been treated with PCP prophylaxis or AZT since their first episode of PCP. Early access to medical services and diagnosis of HIV-1 infection with regular medical follow-up is associated with reduced likelihood of dying during and after the in survival from diagnosis of first episode of PCP. The observed improvement AIDS at St. Mary's has been associated with a variety of changes that attributed to the effect of any one single intervention or drug. Conclusion: to assess the use. cost and outcome of health service interventions, ultimately one has to rely on information derived from a variety of studies which need to be assessed together but which all will have their own intrinsic level of uncertainty. This includes the use of information obtained through observational data on the use of services, case-severity at the time of their use, and economic data, ideally complemented by information obtained through randomized trials.
Version
Open Access
Date Awarded
1998
Advisor
Miller, David
Elliott, Paul
Publisher Department
School of Public Health
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)