Bridging the Gap: Conference Record [Abstract book, International Conference on AIDS (12th: 1998: Geneva, Switzerland)]

466 Abstracts 24126-24130

12th World AIDS Conference

24126 Comparative health care costs in a regional HIVIAIDS
population
M. John Gill1, M. Webek2, W.E. Davidson2. 1213-906 8th Avenue, Calgary
Alberta; 2 University of Calgary, Calgary, Canada
Objectives: To determine and compare the costs of treating HIV infection in a stable regional population receiving care through a centralized outpatient program,
for the periods preceding and following the introduction of protease inhibitors.
Methods: Protease inhibitors became widely available in Alberta, Canada in
November 1996. We have calculated average monthly health care costs for all
patients seen at the Southern Alberta Clinic during two 10-month periods, January-October 1996, and November-August, 1997. Costs per patient have been
disaggregated by disease severity, as determined by CD4 count.
Results:
1996 1997
>500 200-500 75-200 <75 >500 200-500 75-200 <75
Pharmaceuticals 229.32 388.78 528.23 931.36 524.12 672.79 852.52 1331.74
Hospitalization 23.65 34.31 52.83 160.51 4.65 31.45 61.39 87.76
Lab Tests & Diag.
Imaging 61.56 82.78 87.22 59.60 58.96 61.12 61.19 58.58
Clinical Care 53.98 75.60 81.86 137.5 86.70 85.58 108.92 145.34
Homecare 0.00 18.41 38.39 68.56 0.00 20.17 37.51 126.96
Totals 368.51 599.88 788.53 1357.53 674.43 871.11 1121.53 1750.38
Conclusions: Since the introduction of protease inhibitors, average monthly
pharmaceutical costs have increased 43% to 129% depending on disease severity. The subtotal of all other health care costs remained relatively unchanged
during the same time period. However, for those with CD4 counts <75, homecare
costs increased 85% while hospitalization costs decreased 45%, suggesting that
the trend away from inpatient to community-based care is continuing.
477*/24127 Costing HIV using an observation database
Dale Mc Murchy, M. Millson, K. Leeb, RWH. Palmer, G. Robinson, A. Rachlis,
R. Wall. HIV Health Evaluation Unit G314 2075 Bayview Ave Toronto ON M4N
3M5, Canada
Background: The rapid advances in the treatment of HIV have made it increasingly important - yet difficult - to ascertain the costs of HIV so that policy-makers
and health providers can provide care cost-effectively. The HIV Ontario Observational Database (HOOD) contains demographic and clinical data on over 3,000
people living with HIV in Ontario, and presents a unique opportunity to lay the
groundwork for determining the cost of the various aspects of HIV treatment and
conducting economic analyses. This study aims to 1) develop extensive economic
and health service utilisation data and 2) estimate a wide range of costs related
to HIV by linking HOOD to numerous data sources.
Methods: The HOOD data set was expanded through the linkage with administrative data sets in Ontario such as the Ontario Health Insurance Plan, Ontario
HIV Drug Distribution and Monitoring Database, and hospital-based databases
in order to assemble a comprehensive account of patient-specific costs. Reliability studies were conducted to determine the best data sources. Additionally,
a self-administered, cross-sectional economics survey was distributed to HOOD
enrolees to capture detailed information on their service utilisation and costs
related to treatment and care for HIV. Costs were then evaluated according to
individual patient's characteristics and clinical history.
Results: The best and most reliable data sources for service utilisation and
costs were established and used to estimate a variety costs. Costs were estimated for formal and informal care, each accounting for a quarter of HIV-related
costs. Medication accounted for about half of costs. The public-private share of
HIV-related costs is about 50%-50%.
Conclusion: The ability to link patient cost and service utilisation information
with demographic, clinical and community-based data creates a more complete
picture of the complexity of HIV care. This study has laid the groundwork for
building a reliable and sustainable set of variables to capture costs for formal and
informal HIV-related health care.
124128 Pharmaco-economics analysis on a subset of French
patients included in the Caesar trial
Bruno Fantino1, W. Rozenbaum2, A. Beresniak3. 1UMR5823 Du CN RS; Las
S-Universite Lyon 1 43 BLD Du 11 Nov 1945 69622 Villeurbanne; 2Hopital
Rotschild, Paris; 3Glaxowellcome Paris, France
Objectives: Caesar was a randomised double-blinded placebo trial conducted in
Canada, Australia, Europe and South Africa.. The objective of our analysis was
to compare the effects of 3TC vs placebo on the resource utilisation of a subset
of 283 French patients as measured by the care giver support questionnaire.
Methods: HIV-1 positive patients with 25 < CD4 cell count <250 cells/mm3
were allocated to receive either placebo or lamivudine'" (150 mg bid) in addition

to their current treatment regimen, i.e. zidovudine"T monotherapy or zidovudine"T
+ didanosine'" or zalcitabine"T combination therapy at standard doses. Cost-effectiveness (CEA) and cost-benefit analyses (CBA) were carried on. Effectiveness
was measured by death rates, and progression rates towards a new defining
AIDS event. Costs were identified as drugs prescribed, X-rays, lab. tests, hospitalisations and outpatients visits. Benefits were defined as saved costs concerning
drug prescriptions, radiological and biological investigations, hospitalisations, un
scheduled outpatients visits, for the 3TC group compared to the placebo roup.
group.
Results: Progression rates towards a new defining AIDS event are 6.4% in the
3TCTM group and 13% in the placebo group. Death rates are respectively 1.4%,
and 7.2%. CEA demonstrates in the 3TC'M group: 1) a first cost-effectiveness
ratio of 83309 FF indicating the average amount of money spent per patient
without progression towards a new defining AIDS events (vs 93588 FF in the
placebo group); 2) a second cost-effectiveness ratio of 79085 FF indicating the
average amount of money spent per patient surviving after a one-year follow-up
(vs 87738 FF in the placebo group). CBA demonstrates a ratio of 0.769 in the
3TC T group, indicating the average amount of money spent under lamivudineT", in
order to obtain 1 FF benefit. Extrapolation to the study population seems relevant
This hypothesis is justified, in the absence of significant statistical difference in
progression and death rates between the analysed sample and the whole trial
and even after CD4 cell count adjustment.
Conclusion: Pharmaco-economics analysis demonstrates that savings can be
expected with lamivudine prescribed in the treatment of HIV-1 infected persons
with concurrent zidovudine -containing regimens. Cost effectiveness ratios indicate the average cost of additional clinical effectiveness involved by the treatment
with lamivudine'T vs placebo in terms of death rate and progression rate.
24129 An economic evaluation of changing HIV treatment
patterns in Ontario, Canada
Dale McMurchy. HIV Health Evaluation Unit G314, 2075 Bayview Ave, Toronto,
ON M4N3M5, Canada
Background: During 1996/97 protease inhibitors and the viral load testing became widely available in Ontario, Canada. As treatment options for HIV expand
and these lead to prolonged survival, there is a need for policy-makers and medical practitioners to consider cost-effective treatment strategies. The long term
impact of the new therapies, such as years of life gained, remains unknown. Thus,
advances in HIV have made it increasingly important, yet difficult, to conduct economic evaluations. The aim of this study is to investigate the impact of the recent
changes in treatment strategies on the cost HIV. The objectives of the study are:
1) to estimate the costs of formal HIV services and antiretroviral therapies (ART)
for the years 1995/96 and 1996/97 for a sample of Ontario HIV patients; 2) to
compare costs between these two years (pre and post new treatments); and 3) to
estimate the cost per life year gained based on hypothesised outcome scenarios.
Methods: Inpatient costs were calculated based on administrative data provided by Ontario hospitals. A micro-costing was conducted at the two HIV
outpatient clinics. Physician costs are based on billing rates for Ontario Health
Insurance Plan. ART utilisation is based on the Ontario Drug Distribution and
Monitoring Database, and a study by the HIV Ontario Observational Database.
Costs reflect wholesale prices for ART. The cost of per life year gained is estimated
based on hypothesised outcomes. Costs are in 1996/97 prices and discounted at
5%.
Results: A net decrease in inpatient costs was found and the cost of outpatient
visits increased slightly. Utilisation of ART increased in the population, and the
average cost of ART increased by $6,000 per person. The estimated cost per life
year gained prior to the wide-scale introduction of protease inhibitors is $6,207;
after the introduction it is $9,178-$10,292. The incremental cost per life year
gained if the new therapies increase life by five years over the previous therapy
is $600, and $200 if an additional ten years were gained.
Conclusion: This study represents an initial attempt to investigate the cost
implications of the recent advances in HIV treatment. These advances have been
shown to have had dramatic effects on the cost of the HIV care. However, if the
life span of PHAs on these new therapies increases substantially the incremental
cost per life year gained would be modest.
24130 1 The cost of HIV treatment and care at the community level
Blaize Mumford1,2, D. McMurchy3. 144 Shearer Crs, Kanata, ON K2L3N2;
2Communuty, Ottawa, ON; 3HIV Health Evaluation Unit, Toronto, ON, Canada
Background: Recent advances in the treatment of HIV have resulted in improved
outcomes for many people living with HIV/AIDS (PHAs). PHAs are not just surviving longer with the disease, many are managing their health care needs in
community settings. However, there is little documentation of the cost of the delivery of community care to PHAs and their caregivers. Research which enumerates
community resources devoted to HIV care could assist in the development and
planning of future support, and could increase the effectiveness of the community-based response to the disease. This study aims to examine health-related
costs of living with HIV in a community setting. Specifically, we 1) identify the
types of products and support required for community living, and 2) determine
out of pocket expenditure associated with the informal care of PHAs.
Methods: Fifty-three patients with a Karnofsky score of 80 or less were randomly chosen from two care sites to participate in a diary study. The diary,
developed through consultation with the community, contained a comprehensive
list of items and services used by PHAs for health-related reasons. Participants

recorded their expenditure for 4 weeks. Out of pocket expenditure was aggregated into monthly averages. Subjects were then stratified by income to determine
whether there were differences in expenditure by income level. Reported unmet
was also evaluated.
Results: Average annual HIV-related out of pocket expenditure is $4,885 per
PHA. The largest component of out of pocket expenditure is extra food and
nutrition ($2,266 per year). Nonetheless, unmet need for support in the home,