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

Mo.C.1594 - Mo.C.1598

Monday, July 8, 1996

Methods: A cohort of 202 (I 18 males, 84 females) HIV seronegative STD clinic attenders
who spontaneously returned for post test HIV counselling, was assembled between June
and Dec 1993. Baseline information was collected on HIV risk behaviors and vaccine attitudes and beliefs. Follow up questionnaire assessed new partners, new STDs and condom
use. Hepatitis B vaccine was offered in the middle of recruitment to 109 of 202 consecutive
(unrandomised) subjects and follow up rates were compared at the 6 months visit in the
vaccinated and urn vaiccinated groups. Hepatitis B vaccine is not routinely given in Uganda.
Results: The median age was 24 years (range 17-57), 87 of 202 subjects had documented STDs, the rest were SID cont acts or sought FIIV blood tests from the STD clinic. 12%
reported consistent confons use and 20% reported new partners in the previous 6
months. 95% said they would try a new HIV accine anrid 72% were willing to be among the
first. Among 109 subjects offered hepatitis B vaccine, 89 (82%) consented to be vaccinated.
82 out 89 (92%) vacc inated subjects returned for routine follow up at 6 months compared
to 59 of I 3 (52%) un vaccinated subjects (p " 0.001).
Conclusions: We confi rmed that subjects in our clinic had positive attitudes to new vaccines including HIV vacrcines ( substantial number were w ling to take a new vaccine).
Those who accepted the vaccine were significantly more likely to return for follow-up. In
our setting, vaccine offer- ray be an appropriate incentive to retain cohorts in a longitudinal
study, at least in the short term.
Moses R. Kamya, Department of Medicine, Makerere Medical School, PO.Box 705 I Kampala
Uganda.Tel [phone] Fax 0 I I25641 2456413 email: [email protected].
Mo.C.1594
A COMPARISON OF METHODS OF ASSESSING QUALITY OF LIFE OUTCOMES IN A
RANDOMIZED TRIAL WITH MISSING DATA DUE TO DROPOUTS AND DEATH
Raboud M, Singer J, Thorne A, Fanning M, forna EF,Turgeon F, Gill J, Schlech WF, Cameron
DW, Smaill FM, Lemieux C, Phillips P Rachlis A, Salit I, Fong I, Aoki FWalmsley SL, Shafran S,
the Canadian MAC Study Group. Canradian HIV Trials Network,Vancouver, Canada.
Objective: To assess analytic methods of estimating the effect of treatment on quality of life
(QOL) in a clinical trial comparing two mrultidrug regimens for the treatment of mycobacterium avium complex (MAC) in the presence of missing data due to dropouts and death.
Methods: The MOS-HIV instrument adapted for MAC was measured at 0, 2, 4, 8, 12 and
16 weeks of follow up. The Q)OL of each patient at each follow-up time was assigned to
one of 5 categories according to whether the total score from 8 parts of the Symptom
subscale each with a response from I (best) to 5 (worst) was < 16 (Level I), 16-24 (Level
2), >24 (Level 3), the patient had missing data because of drop out (Level 4) or death
(Level 5). A generalized estoimatoing equation (GEE) model for categorical data assuming a
cumulative logit link was used to model the effect of time, treatment group, level of bacteremia and the presence of other uncontrolled disease at enrolment on QOL.
Results: 93 patients on the 3-drug arrn and 82 on the 4-drug arm were included in this
analysis.The proportions of missing data due to dropout in the 2 treatment groups were
16% and 28% and due to death were 5% and 10%. Naive comparisons of the uncategorized QOL data between treatnment groups ignoring missing data showed no treatment
effects at any of the follow -up times. Comparisons of the uncategorized data between
treatment groups after assigning the worst possi ble score to patients with missing data due
to death and the 2nd worst possible scoe to patients with missing data due to dropout
detected treatment effects (p <.01) it weeks 2 through 16. From the GEE model, treatment
with the 3 drug regimer was associated with being in a better QOL category (RR-2.32,
p<.0001) while later follow up tines and the presence of another uncontrolled disease at
baseline were associated with worse QOI. (RR.60, p=.01).
Conclusions: Analyses of QOL data ern be misleading if missing data is not handled appropriatelyThe GEE model for categor ical data provides a mechanism for modelling missing
data, a summary measure of the treatwnst effect and can model the effects of covariates
while controlling for correlated data.
Janet M Raboud, Canadian HIV Trials Network, 200-1033 Davie St,Vancouver BC,V6E IM7
Canada.Tel [phone] Fax [phone] email: jraboud(@hivnet.ubc.ca
Mo.C.1595
ISSUES IN THE DESIGN AND IMPLEMENTATION OF LARGE SCALE,
COMMUNITY-BASED CLINICAL TRIALS IN NUTRITION AND AIDS
Gibert Cynthia L, Wheeler DA, I auner C, Muurahainen N, Raghavan S, Madans M, Higgs E,
Elion R, Bartsch G, and the Community Programs for Clinical Research on AIDS (CPCRA)..
NIAID, NIH, Bethesda, MD, USA
Issues: Weight loss in HIV infection s associated with disease progression and mortalityTo
date, most intervenrtional studies in nutrition/ 1IV have involved short term evaluations of
small numbers of patients. In order to evaluate the clinical benefit of nutritional interventions, larger numbers of participants need to be enrolled in studies conducted in the community practice setting.
Project: In designing and planning the implementation of a large community based clinical
trial of nutritional interverntion for HIV patients, the CPCRA investigators have addessed
the unique demands oad constrinsts f a rucnutron al interventron study employing the usual
clinical trial cosiponsents.
Results: Clinical trial cosnponnents rind issues for-nutsritional studies: I ) participant selection -
healthy, 10% loss of body weigtt, atr rsk for-/os misting 2)stratiication -physical activity at
baseline: 3) outcome mseasures -disceise progressons/surrvval or changes in weight/body
comsposition;, 4) samnple size/study duran -is estiatt-s uf occurrenroot of endpoints and of
drop out rate of psi 150rpants, 5) quality control -starndardization of hseight, weight, and
bioimpedance assay measurenrts; evaluation of total nutritional intake using dietary recall
and of compliance with study thor s6 election of therapeutic agents taste and composition of nutritional supplements (caloric density, proportion of medium chain tonglycerides (MCT), peptides, and micronutsrents): use of combinations and/or placebos: use of
appetite stirnulants, anaboio agents: 7) selection of study personnel --staff with/without formal training in nutrition.
Lessons Learned: It wiibe 'mportant to address each of these ssues when designng large
scale nutritional studi esf tie investigton s believe it an early intervention study of an
MCT supplemnent for patients at risk ion* warsting can roost benefit the field.The investigators
concluded that to increase tire likelihood of success for such a trial a palatable nutritional
supplement with crodest MCI and olive ongoing dieltary counseling is necessary.

Successful implementation and completion of such studies will contribute to the development of a standard of care for the prevention and tretmeit of HIV-asocilted nalnutritron.
Cynthia L. Gibert, M.D.,VA Medical Center, Dept. of Ihelctioui D,rcses, 50 I I Sireet,
NW Washington, DC 20422 Ofc.Tel: [phone] IFax (212 3 3 -1'
Mo.C.1596
HIV/AIDS COMMUNITY-BASED CLINICAL TRIALS ACHIEVE LOW LOST-TO-FOLLOW
UP RATES
Caldwell, Reginald A., Cox, L.E., Morse, E.V, Pieper. M.,- lurner, MV lier,VW., Simns, RM,
Launer, C., the Terry Beirn Community Pyr o ris for Clinr cal P"',r tcIt /) ( I RA).
NIAID, NIH, USA.
Objective: Community-based HIV/AIDS clinical trials pror s,,wn ui n,r, ' r r enrted populations, African-Americans, Latinos, wonien and injci,dri'i 'rt n,, 5, it OI In o
lost-to-follow up rate.
Method: In 1989, the NIH established the CPCRA which pr ovidcd access to clinic l IriIls
within the context of primary health care delivery sites. Significant umbers) ofc underrepresented populations were enrolled which altered the sociodesroen phic characti tstics of study
participants.The CPCRA units successfully complete scientifically rionots AIDS cnical trials in
primary care settings with socially and economically marginalized populatrons. Ifhese target
populations, including African-American, Latino. worries, and injectirn 'drug users, oltens face
barriers that impede their enrollment and retention in AIDS clinical5 s e, Inirch. te ICPRA has
a structure to address these barriers which include: access to rese irch in the patients' preferred and trusted primary care setting; study designs which easily it into primary care: long
term relationship with health care staff; constituent input; and careful site selection.
Results: Among the 3110 patients enrolled in six completed randornized treatnr t studies.
there were 42% Africa-American/L atino, 23% history of injecting drug use, and 16%
women.The six completed studies'"vital status unknrown" are lddl/ddC (N=467) 0.9%,
Fungal (N=323)-1.2%, PCP-INTI (N=72) -I.4%, oxo (N- 396) -07%, and (:F'IMV (N -99.1)2.2%, NuCombo (N= I I 13)- I.9% with an average of 2.2%. By con pariison, in published anti
retroviral studies of other research groups. lost to-follow up rates range nfrom 4.'1% to
34.8%.The CPCRA's comparatively low rates are in part itstnbutable to thie its comnmunitybased structure.
Conclusion:While disenfranchised HIV infected populations may have bar iers thit make
participation in AIDS clinical trials difficult, the CPSCRA's conrnunity bbased Isiructure as core
pared to other research groups successfully mnairtains the partic p ton of the ptient popu
lation with low lost-to-follow up rates.
Reginald A. Caldwell, Denver Public Health, 605 Bannock Street, Denver,; Colorado 8020t.
USA 303-436-7195 (phone) / 303.436-7 190 (fix) / '106- 3112 (e,al).
Mo.C.1597
BENEFITS OF A COMPLETE LABORATORY MANAGEMENT SOFTWARE SYSTEM IN
HIV CLINICAL TRIALS
Michels, Cheryl*, Ladd E*, Michels R*, Br eror JW.". 1 Dataworks Dvl,'Iln erit, Is.,
Seattle, WA, USA; " Rush Medical College., Chicago, IL. USA
Objective: To automate patient specimen mranagernent procedu r s a d irpE cve the quality
of the resulting assay data in HIV ciinicalr trials.
Methods: In January 1988, paper forms and hand calcular ions of datsa were eplaied by an
automated laboratory data management system (the Retrovirus Lboiratory i vternentProgram~) for the collection, editing, storage, and retrieval of p24 antig s ren nid,rlit ativ cultune data in the AIDS Clinical Trials Group (ACTG). As technio ogyue p deouce n ssaysi, the
system design allowed new modules to be easily integrated. Additional mo rdules were rincorporated for the collection and reporting of data to the Virology Quaity Assurancite Program.
The quality of the data was evaluated by the data managem ent centers, nd te usabiity of
the software was evaluated through questionnaires, inter views, arid users group in otugs.
Results: Anecdotally, reporting of patient data dramatically i smproved with the iniplet entation of the automated system. Furthermore, the quality of sthe data s er enu e d by elimie
nating transcription errors. Although the number of tests aind volume of patient specimnens
increased, the resources necessary to perform specimenn Terrnageent proc edures dd not
increase proportionally.Turnaround time for the receipt of complete ndus cnurate quality
assurance data from the laboratories and its subsequenrt eport to the VIselegsyQuality
Assurance Program decreased from 10 days to 24 hours.
Conclusion:The Retrovirus Laboratory Management Program05 sus trot.Iy rimproved the
day-to-day activities in virology laboratories supporting HIV cisictl trinls, resulting n both a
higher quality of patient data and a significantly faster turnar ntin.e.(onsequer tly iti s
now used in both the pediatric arid adult ACcG as yell as nuLerous other nulti site studies.
Cheryl Michels, 6920 220th St. S.W, Suite 107, Mountlake lerrace, VVA 980113 US/A
Telephone: [phone] Fax: [phone] email:73164.3002@co nps itve.com
Mo.C.1598
PERINATAL AIDS CLINICAL TRIALS INTEGRATING RESEARCH INTO PRIMARY
CARE "THE MIAMI EXPERIENCE"
F. Doyle R.N.C., C. Alfonso R.N., A. tHerrnun R.N., M. Ago to M.S.,. Air tiro OJ.,
D'Suisvan, M.D. University of Miami School of Medic uit' Depin tnnt of Gil/GYN
The University of Miami/Jackson Memorial Hiospaital us a irgo r~rmopoit u tom tury iare
center serving Dade County and its surrounding conun ties.Trioe Unists 1f iir
Department of OB/GYN Research works within tthus tysteni, h rg tr ri i-r ity cl so ninth
treatment, as well as opportunities to participte in clirstal ts ml tt IIV It In sun
and their families.
Perinatal ACTG was initiated with Protocol 082, ar Purse I Study urn 1)8') At that Insae, we
had one patient and one protocol. As of today, we hirve par ticrited in seen protocols,
four in which we are now actively recrutirng. Duo total nuonsber of mounits ms ninoty tinseno
OB/GYN is often the primary access route for wonton seeking iron th cuts' tools> thsough
our pre-natal and GYN Special Imnunology clinics, we offer tslr i r us Igovicm
for the whole family referral to specialized sonhvices is imnd, coordrhs itspint inn larrmos rId
appointments, and other services required to meet tire cieouts roods A ruultrdrsonpl nour
team of physiciann, nurse practitioners, nurses, social work-ss nitrm ase o.innayens nrc
responsible for ensuring that the vacuous needs of woolen lIrving wills the sius aro intel,
while maintaining the standards of research required by the AL f ja0 poIos.

157