Abstracts Vol. 1 [International Conference on AIDS (11th: 1996: Vancouver, Canada)]

Track B: Clinical Science
them. Providers stated that the main benefit of the NPMS would lead to establishing
national standards of HIV service provision. Based on this feedback the Developmental
Phase of the NPMS was commenced. A number of multidisciplinary meetings were held to
develop the datasets.The Activity and Case-Severity datasets are currently being piloted in
12 HIV units across England.The Acceptability dataset has been finalised and will be piloted
in these sites. A Working Party is being established to standardise costing of service provi
sion across the pilot sites.The organisational structure of the NPMS and how it operates
will be presented.
Lessons Learned: Despite the introduction of increased competition in the NHS as a
whole both Purchasers and Providers of HIV services in England have welcomed and supported the establishment of the NPMS.The NPMS is considered to be beneficial by the different organisations involved with HIV-related health care. Implementation of the NPMS
beyond the current I 2 pilot sites is anticipated upon the successful completion of the
Development Phase of the NPMS.
M. Egan, Department of Epidemiology & Public Health, St Mary 's Hospital Medical School,
Norfolk Place, London W2 I PG.Telephone: ++017 1-594-38 I 3, Fax: ++017 1-402-2150,
email: [email protected]
Tu.B.2159
A SUCCESSFUL STRATEGY FOR DECREASING COSTS ASSOCIATED WITH THE USE
OF G-CSF IN AIDS PATIENTS
Ball, Susan C, Huang K, O'Brien L, Jacobs J. Center for Special Studies, New York HospitalCornell Medical Center, New York, NY USA
Objective: To establish dosing parameters for Granulocyte Colony Stimulating Factor (GCSF) in AIDS patients. Our hospital adopted guidelines for the use of G CSF and assessed
rates of breakthrough bacteremia and cost.
Methods: We reviewed the charts of all in-patients on the HIV service who received GCSF during three-month periods before and after the inception of these guidelines in April,
1995. (Guidelines: Initiation of G CSF when ANC<500; dosing began at 150 mcg three
times weekly G-CSF was discontinued when ANC> 1000.)
Results: Frorm January through March, 26 patients who received G-CSF were admitted 31
tirmes.There were 3 positive blood cultures including cultures of methicillin-resistant Staph
aureus, E coli and Streptococcus pneumoniae.Two of the positive cultures were in patients
who were not neutropenic. Only ten episodes of neutropenia (ANC<500) were documented. A total of 65,065 mcg of G-CSF was given over 228 patient days.The average dose
of G-CSF given was 322. I mcg making an average daily dose of 285 mcg.
From June through August, 1995, 23 patients received G-CSF at some point during 25
admissions.There were 2 positive blood cultures (Staph saprophyticus and Enterococcus
faecium) and each occurred in a neutropenic patient. Fourteen episodes of neutropenia
were documented. A total of 18,320 mcg was given over 123 patient days.The average
dose of G-CSF was 220.7 mcg and the average dose per day was 149 mcg.
At a cost per unit dose, the expenditure for G-CSF fell from $25,722 to $7,242 after the
establishment of guidelines.
Conclusion: This preliminary study suggests that many AIDS patients receive G-CSF in the
absence of neutropenia.The establishment of guidelines can restrict the use of G-CSF without an adverse affect on the incidence of bacteremia and can substantially reduce cost.
Further evaluation of these guidelines is ongoing.
Susan C. Ball, F 24 Center for Special Studies, NewYork Hospital, 525 E. 68th Street, New
York, NY 1002I USATel: [phone] Fax: [phone]
Tu.B.2160
A CLINICAL DATABASE FOR ESTIMATING COSTS OF PRIMARY HIV CARE
Cohn J, Debra Kosko, Purdue B, Weidle P The Evelyn Jordan Center, University of Maryland
Medical Center Baltimore, Maryland, U.S.A.
Issues: To compete in the managed care marketplace while maintaining our goal of providing comprehensive and state of the art care, a method was needed to measure actual costs
of comprehensive outpatient HIV care in a University based clinic.The bill generated by our'
parent institution for a clinic visit does not reflect the actual costs of care; certain institutional charges are inflated whereas other institutional charges are less than the costs of actually
providing a service. A different methodology was required to quantify actual costs.
Project: A relational database (MS ACCESS) was programmed to record resource utilization of individual patients. Related tables stored basic clinical information and services provided by the primary care provider (physician or nurse-practitioner), case manage pharmacist, respiratory therapist, psychiatrist, nurse, procedures performed in clinic, outpatient
referrals, diagnostic tests ordered and medications prescribed.
Results: Average resource utilization for patients at different stages of HIV disease can be
calculated from these data. Different prices can then be attached to these services, including
the actual costs of these services in our clinic or costs typical of other community
providers, to generate average costs over a specified time period for patients in different
clinical categories.This system provides rea time data on outpatient costs and can be
updated to include new standards for monitoring patients (such as quantitative plasma
RNA measurements), new interventions (such as newly licensed antiretroviral medications
or combinations of medications), and new standards for when to initiate or alter therapies.
Lessons Learned: The ability to assess changing costs in real time will be crucial to negotiate contracts for comprehensive HIV care with managed care organizations.
Debra Kosko, University of Maryland. Adult HIV Program 22 S. Greene Street, Box 174
Baltimore, Maryland 2120 I Telephone [phone] I Fax [phone]
email:[email protected]
Tu.B.216 I
THE DEVELOPMENT OF HIV/AIDS CLINICAL PATHWAYS
Frazie, *, Gathe, Jr ]*, Gustafson, P *, Anthony K *, Brown, N*, David, J], Dodds. Wt,
Goetz, A, Goldstein, J *, Nelso, S, Pendarvis, K*, Benz, P**, Esle R *. *Park Plaza Hospital
** Hennepin County Medical Center

Objectives: To illustrate the development of Clinical Pathways (CPs) in a dedicated,
HIV/AIDS hospital unit designed to correct diagnostic and treatment systems, for the pur
pose of improving patient outcome, quality of care and cost-effectiveness.

Tu.B.2159 - Tu.B.2163
Method: A multidisciplinary physician-led team was formed to include hospital
administrators from nursing, ancillary services (lab and respiratory) and quality management.
Initial development focused on HIV related pneumonias, specifically PCPTasks included an
assessment of current treatment patterns, the identification of treatment and system bottlenecks, and the creation of new pathways based on a similar model developed by the
Hennepin County Medical Center (MN).
Results: Seven HIV related pneumonia CPs were created: two initial pneumonia diagnostic
CPs (where diagnostics are performed either entirely out-patient or in-patient), that flow
into five specific treatment pathways (PCP CMV, bacterial, fungal, AFB). Initially significant
bottlenecks in the diagnostic CPs were identified prior to addressing the treatment CPs:
inefficient induced sputum procedures; inaccuracy and delay of sputum results; indications
for bronchoscopy, gallium lung scan (GLS) utilization; and choice of empiric medications. A
90-day pilot program on the induced sputum issues resulted specifically in an improvement.n the identification of PCP from I% to 58%, a reduction in turnaround time for results to
be received (48 hours or less), and avoidance of bronchoscopies in 13% of patients in the
sample. Specific indicators for GLS utilization were identified, and appropriateness and
length of use of empiric drugs were analyzed.
Conclusion: I.)CPs provide the occasion to take an analytical look at practice patterns and
resource utilization. 2.)The composition and dynamics of the team are critical to the devel
opment of CPs. 3.)The success of the pilot program on the induced sputum issues suggests
the usefulness of CPs for enhancing the efficiency of HIV/AIDS diagnostic processes. 4.)CPs
produce better communication among health care team members: identify and alleviate
bottlenecks in systems to facilitate improvement in services, promote more timely treat
ment and less invasive procedures: and enhance evaluation by requiring systematic analysis
of outcomes. 5.)Data is being prospectively analyzed regarding each specific treatment CP in
regard to patient outcomes and cost effectiveness. 6.)The specific challenge of creating CPs
for HIIV/AIDS lies in the multiplicity of simultaneous opportunistic infections, the varied clinical presentations and the resultant need for multidimensional CPs.
Roderick Frazie: M.D. 4101 Greenbriar #200 Houston,Texas 770098 Telephone:(7 13)520
5537 Fax: [phone]
Tu.B.2162
EVALUATION OF TWO DIFFERENT HIV-I CONFIRMATORY METHODS ALONG A
YEAR
Martinez Cerana Pablo H, de la Torre K, De Vinzenzi V, Martinez T Pasteran F, Hermes R,
Peikert R, Draz L, Canete L, Cando O, Valdez I. "Juan A. Fernandez Hospital" (Laboratory
Division), Buenos Aires, Argentina.
Objective: Compare the methods used daily in our service in relation with their specificity
and sensibility according to the screening and diagnostic in a select population with an HIV
type I infection risk.
Methods: We studied 2934 patients that assisted to the Infectology Service of the hospital
during 1993 with HIV probable infection.The methods used were: recombinant EIA for
HIV I/HIV-2 (2~ generation), Abbot Diagnostics; the criteria used was: positive relation > I. 1
reactive, < 0.9 nonreactive and 0.9 <RP< i. 1 indeterminate. Line Immunoassay (LIA) for
proteins and peptides HIVI +2, LiaTek, Organon Teknika and Western Blot, Biorad HIVIblot.The results were evaluated as the manufacturers criteria, and were reevaluated with
the CDC and FDA criteria.
Results: From the total of patients, 663 (22.6 %) were reactive with ELISA, 20 (0.7 %) indeterminate and 2251 (76.7 %) nonreactive.To 557 reactive samples LIA were made and to
106 reactive samples WB, getting the following results; when the results were reanalized
using other criteria, as CDC and FDA, significantly differences were observed.

Criteria

WB (%)
Positive Indeterminate Negative

LIA (%)
Positive Indeterminate Negative

Manufacturer
CDC
FDA

76.4 17.0 6.6 97.3
859 7.5 6.6 84.3

43.4

75,4

14.5 1.2
23.3 1.2

O
O,D
0
o
u
C
cn
0
C
0
U
C
a)
cc
0
2O
C
0
t--
a)
292

Conclusions: We have found that the criteria which we obtain the less differences in the
incidence of positive results, between both methods analized, is the CDC.This increase, for
LIA, if we use the manufacturers criteria, is because they consider the high sensibility of the
technique according to the origin of the antigen used. In conclusion, we consider that ELISA,
in two consecutive samples, is a good screening method and LIA, evaluated with the criteria
of CDC for WB, gives an acceptable degree of confirmation of the results, according the
population studied.
P Martinez Cerana, Melo 1194, (C.P I 638) Vicente Lopez, Buenos Aires, Argentina.Tel.: 79 I07 16 Fax: [phone]
Tu.B.2163
EVALUATION OF FIVE HIV-I/HIV-2 ENZYME IMMUNOASSAYS
Best Susan, Stephenson M, Dax EM, and the Australian HIVTesting Laboratories. National
HIV Reference Laboratory Fairfield Hospital, Fairfield,Victoria, Australia.
Aim:To evaluate five HIV- I/HIV-2 EIAs for their suitability as screening assays in Australian
laboratories.
Methods:The EIAs under evaluation included HIV-I/HIV-2 Enhanced ELISA (Ortho
Diagnostic Systems), Syva Microtrak II HIV- I/HIV 2 EIA (Syva Company), Recombigen HIVI /HIV-2 EIA (Cambridge Biotech Corporation), Innotest HIV- I/HIV 2 Ab (Innogenetics) and
Novapath HIV I/2 EIA (Bio-Rad).To determine specificity approximately 5000 blood donations were tested in parallel with the routinely used assay in I - 3 blood banks. Sensitivity
was determined in approximately 500 anti-HIV-I positive sera, 25 anti-HIV 2 positive sera,
150 anti-HIV negative sera which showed reactivity on other EIAs ("problem" samples),
I 36 samples f-om 16 individuals taken serially during seroconversion and I0 anti HIV I positive samples in 10 doubling dilutions (1:1 I 6-1:81 I92). Sensitivity testing was performed in four
public health laboratories. Anti HIV- I positive, problem and titrated samples were tested in
duplicate.