Abstract Book Vol. 2 [International Conference on AIDS (14th: 2002: Barcelona, Spain)]

XIV International AIDS Conference

Abstracts WePeB5813-WePeB5816 35

Abstract WePeB5815 - Table
Month -6 -3 0 +3 +6 +9
Adherent patients (%) 32 (65) 29 (59) 33 (67) 37 (75) 37 (75) 36 (74)
CD4 cells, /mm3, mean ~ SD 508 ~ 285 462 ~ 268 545 ~ 317 546 ~ 303 506 ~ 253 483 ~ 268
HIV RNA, copies/ml, log10, mean ~ SD 2.16 ~ 1.58 2.19 ~ 1.62 1.94 ~ 1.56 1.81 ~ 1.47 1.61 ~ 1.40 1.68 ~ 1.47

Presenting author: patrick french, office of program evaluation and research,
aids institute, 150 broadway, 5th floor, menands, new york, 12204, United States,
Tel.: +[phone], Fax: +[phone], E-mail: ptf01 @health.state.ny.us
[WePeB5813 IAssessing the usual care control in adherence
intervention trials
G.J. Wagner, D. Kanouse. RAND, Santa Monica, California, United States
Issues: Most adherence intervention trials include a "usual care" control group.
However, usual care practices aimed at promoting adherence have not been measured in these studies, even though usual care can vary widely across clinics.
Description: This presentation provides the rationale for measuring usual care
in adherence intervention trials and presents findings from a pre-enrollment assessment of usual care in a trial currently underway. Qualitative interviews and
quantitative measures of strategies used by clinicians in general and in specific
patients visits were administered to clinicians at 5 university-affiliated HIV outpatient clinics.
Lessons learned: Methods used in this study provide examples of how usual
care can be measured and show how the use of multiple methods can reveal
different aspects of the content, frequency and intensity of usual care practices.
The qualitative and quantitative measures consistently identified a common set of
core adherence strategies used at each clinic; however, the qualitative interviews
revealed differences in the extent to which adherence training was reinforced,
whereas quantitative measures were more sensitive to differential use of specific
strategies before and after onset of treatment.
Recommendations: Measuring the components of usual care needs to be a
fixture in controlled trials of adherence and other behavioral interventions. In controlled adherence intervention trials, measurement of usual care practices is essential to: 1) assess the intervention's potential utility in settings with different or
similar usual care practices; 2) determine whether variation in usual care across
sites moderates site differences in the effectiveness of the experimental intervention; and 3) assess for contamination of the control group. Measurement of usual
care will enable researchers to calibrate intervention effect sizes and better judge
the likely effects of implementing interventions in various community settings.
Presenting author: Glenn Wagner, RAND, 1700 Main St., Santa Monica, California, 90407, United States, Tel.: [phone], Fax: [phone], E-mail:
[email protected]
WePeB5814 Using practice trials as a tool to assess adherence
readiness among drug users
G.J. Wagner. RAND, Santa Monica, California, United States
Background: With highly active antiretroviral therapy (HAART) now seen as less
urgent, there is increased emphasis on assessing adherence readiness prior to
prescribing treatment, particularly among high risk populations such as active
drug users. However, there are no established methods to assist providers in accurately assessing adherence readiness.
Methods: To evaluate the utility of a placebo practice trial in assessing adherence
readiness among drug users, we administered a 2-week practice trial that mimicked HAART (Phase 1) to HIV+ individuals not currently on antiretroviral therapy
and who had a history of drug dependency (past or current). When and if participants started HAART, adherence to a selected antiretroviral was measured
for two weeks (Phase 2). Adherence was measured using electronic monitoring
caps.
Results: To date, 193 participants have enrolled; 76% are male, 68% are nonwhite, and at baseline 35% had CD4 counts < 350 and 50% were active drug
users (cocaine, heroin or methamphetamine). Among the 179 (93%) Phase 1
completers, mean adherence to the 2-week practice trial was 67%. Sixty-five participants have started HAART and completed Phase 2. Variables associated with
initiation of treatment included discontinuation of drug use, lower CD4 count, having an AIDS diagnosis, stable housing, greater perceived treatment efficacy, and
having a longer relationship with one's primary care provider (p<0.05). Mean adherence during Phase 2 was 75%, which was not significantly different from the
70% adherence achieved by this subgroup in Phase 1; adherence during Phase 1
was significantly correlated (r = 0.50) with adherence during Phase 2. Active drug
users did not differ significantly from those in recovery with regard to adherence
to both the practice trial and HAART.
Conclusions: These results suggest that adherence to a practice trial can predict how a patient will adhere to HAART and may be an effective tool to assess
adherence readiness.
Presenting author: Glenn Wagner, Rand, 1700 Main St., Santa Monica, California, 90407, United States, Tel.: [phone], Fax: [phone], E-mail:
[email protected]

WePeB5815 Adherence and efficacy of HAART with bid or qd
didanosine
B. Roca, C. Lapuebla, J. Madero, M. Beneyto, A. Lillo. Hospital General,
Catalunya, 33-A, 4, 12004 Castellon, Spain
Background: Adherence to treatment of chronic disease is difficult for patients. In
HIV infection, cost, side effects, and complexities of HAART further increase the
problem, and alarming rates of non-adherence have been reported, which leads
to decreased efficacy and viral resistance. Easier to take modalities of HAART
may help to improve adherence and efficacy
Methods: In a cohort study, we assess adherence to HAART that includes didanosine, and evaluate CD4 cell count and HIV RNA. We compare results of the
three visits before with results of the three visits after switching from bid tablets to
qd enteric-coated tablets of didanosine. In every visit, patients are considered adherent when: 1) they keep the appointment, 2) they report they have taken more
than 80% of prescribed doses, and 3) their HIV RNA level is at least 1.5 log10
below baseline.
Results: We study 49 patients. Mean and SD of age is 37 ~7 years, 42 (86%)
are male, and 36 (73%) are drug users. All subjects have been taking one of
the modalities of HAART recommended by guidelines for at least one year; 32 of
them (65%) are on zidovudina, and 35 (71%) are on one protease inhibitor. Table
shows data of the three visits before switching from bid to qd didanosine (months
-6, -3 and 0), and the three next visits (months +3, +6 and +9).
Pooled data of the first three visits compared with pooled data of the last three
visits shows improved adherence in the last three visits (P = 0.03), no difference
in CD4 cell count, and lower HIV RNA in the last three visits (P = 0.05).
Conclusion: Adherence and virological efficacy of HAART with didanosine qd is
superior to identical regimens with didanosine bid.
Presenting author: Bernardino Roca, Catalunya, 33-A, 4, 12004 Castellon,
Spain, Tel.: [phone], Fax: [phone], E-mail: [email protected]
WePeB58166 Lower socioeconomic status does not impact on
adherence or outcomes in an African cohort on
HAART
C.J. Orrell, M. Badri, R. Wood. Infectious Diseases Unit, University of Cape
Town, PO Box 50309, Waterfront, Cape Town, 8005, South Africa
Background: There is a prejudicial view that, due to poverty and poor education, individuals in Africa may be less likely to be adherent to antiretroviral therapy
(ART) than HIV-positive people in the developed world.
This study examines factors which may impact on virological success in an
African cohort.
Methods: The socio-economic status of a public-sector, outpatient cohort of 242
HIV-positive individuals on HAART was recorded using a composite index, determined by area of residence.
This index combined education, welfare, unemployment and overcrowding status
with annual income.
An index of >28.5 indicated poor social circumstances (shack dwelling, outside
tap). Home language was recorded (English, Afrikaans or local African language,
Xhosa). Adherence of the cohort was assessed using tablet returns. Returns were
correlated with virological outcome. Viral loads and CD4 counts were recorded at
baseline and 48 weeks. Logistic regression analysis models were used to determine predictors of virologic failure (>400 copies/ml). Factors fitted to the model
were low socio-economic status, language spoken at home, <80% adherence,
viral load at baseline and CD4 count <200.
Results: 49% of the cohort are Xhosa-speaking. 43% live in poor social circumstances. 74.5% of the
Xhosa-speaking group live in poor social circumstances. The median adherence of the cohort was 94.5%(mean 88.5%). Adherence by tablet return correlated significantly with viral load reduction (p<0.001). Factors predicting virologic
failure were three times daily medication dosing (HR=5.92), adherence <80%
(HR=3.29) and baseline viral load (HR=1.93). Socio-economic status, Xhosa and
low CD4 count were not associated with failure.
Conclusion: Treatment success depends on baseline viral load and reduced frequency of treatment dosing.
In this study, low socio-economic status and home language were not barriers to
successful antiretroviral therapy.
Presenting author: Catherine Orrell, PO Box 50309, Waterfront, Cape Town,
8005, South Africa, Tel.: [phone], Fax: [phone], E-mail: correll
@uctgshl.uct.ac.za