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

36 Abstracts WePeB5817-WePeB5820

XIV International AIDS Conference

WePeB581 7 Crack cocaine use and adherence to antiretroviral
treatment among HIV-infected black women
T.T. Sharpe', L.M. Lee2, A.K. Nakashima3, L.D. Elam-Evans2, P.L. Fleming4.
' Epidemic Intelligence Service Officer, NCHSTR DHAR SB, Centers for Disease
Control and Prevention, Atlanta, GA, United States; 2Epidemiologist, NCHSTR
DHAR SB, Centers for Disease Control and Prevention, Atlanta, GA, United
States; 3Epidemiologist, NCHST DHAR SB, Centers for Disease Control and
Prevention, Atlanta, GA, United States; 4Branch Chief, NCHSTR DHAR SB,
Centers for Disease Control and Prevention, Atlanta, Georgia, United States
Background: We sought to determine whether HIV-infected Black women reported crack use more than other HIV-infected women and to examine the relationship between crack use and antiretroviral treatment (ART) adherence among
Black women.
Methods: We analyzed data from interviews of HIV-infected women 18 years of
age reported to 12 state and local health departments between July 1997 and
December 2000. We stratified our sample into nonusers of drugs, users of drugs
other than crack and crack users. Because the majority of eligible women were
Black, we selected only Black women to examine the relationship between crack
cocaine use and ART adherence. We used logistic regression to examine the
effect of crack use on adherence to ART, controlling for age and education.
Results: Of 1655 HIV-infected women, 585 (35%) were nonusers of drugs, 694
(42%) were users of other drugs and 376 (23%) were crack users. Of the 1196
(72%) women who were Black, 306 (26%) were crack users. In multivariate analysis, both crack users and users of other drugs were less likely than non-drug
users to take their ART medicines exactly as prescribed (odds ratio [OR]= 0.37;
95% confidence interval [CI]= 0.24-0.56), OR= 0.47; 95% (CI) = 0.36-0.68), respectively.
Conclusions: HIV-infected Black women who use illegal drugs, especially crack
cocaine, may require sustained treatment and counseling to help them reduce
substance use and adhere to ART
Key Words: Crack cocaine, Black women, HIV/AIDS, antiretroviral treatment
Presenting author: Tanya Sharpe, Centers for Disease Control and Prevention,
1600 Clifton Road Mailstop E-47, Atlanta, Georgia 30333, United States, Tel.:
[phone], Fax: [phone], E-mail: [email protected]
WePeB5818 Prospective predictors of medication adherence:
A study of the first six months of highly active
antiretroviral therapy (HAART) using electronic
monitoring
L.J. Koenig', T. Ellerbrock1, M. Pratt-Palmore2, D. Stratford1, T. Bush1,
E. Malatino2, M. Todd-Turner2, M. Golde2. 1Centers for Disease Control and
Prevention, Atlanta, GA, United States; 2Emory University Atlanta, United States
Background: Determine predictors of antiretroviral adherence among indigent,
HAART-naive patients receiving care at a U.S. inner city clinic.
Methods: During 1999-2001, 231 (82%) of 283 consecutive HIV-infected patients who registered for care and were prescribed HAART for the first time were
enrolled in a randomized controlled adherence intervention trial. Patients selfreported demographic, behavioral, and psychosocial characteristics. Clinical data
were abstracted from medical records. Patients received either adherence support and individualized problem-solving or enhanced standard of care. Medication
Event Monitoring System [MEMS] caps placed on one pill vial (85% on a protease
inhibitor; 92% BID dosing) assessed adherence, calculated as the proportion of
pills taken (~3 hours of dosing time)/pills prescribed.
Results: 206 patients (age 22-69; 62% male, 83% black/non-Hispanic) have now
been enrolled > 6 months. Of these, 125 (61%) completed the 6-month assessment; MEMS data are currently available for 100 of these patients. With the exception of older age, patients with MEMS data did not differ from all 206 patients
on demographics or health. Among the 100, adherence was 74% (74% intervention, 73% control) at 0-3 months, 71% (73% intervention, 68% control) at 3-6
months, and did not differ by intervention arm (p's > 0.38). In multivariate analysis, better adherence was predicted by: an "autonomy-supportive health care
context", defined as the perception that providers/partners offered choice, minimized pressure, and acknowledged patients' perspectives; no intravenous drug
use or opportunistic infection; and less HIV-related physical impairment.
Conclusions: Many patients (39%) discontinued therapy by 6 months. Even
those still on therapy missed over I of their prescribed doses, regardless of intervention. Provider/partner support for patient autonomy in managing their own
health may be critical to adherence and should be provided early in treatment.
Presenting author: linda koenig, Centers for Disease Control and Prevention,
1600 Clifton Rd, Mailstop E45, Atlanta, GA 30333, United States, Tel.: [phone], Fax: [phone], E-mail: [email protected]

WePeB5819 A multi-method approach to measuring
antiretroviral therapy (ART) adherence from the
cross-site SPNS adherence collaboration
L.R. Hirschhorn1, R. Mukherjee2, S. Manheimer3, K. Graham4, L. Mundy5,
J. Dougherty6, S. Celano7, J. Mantell2. 1Dimock Community Health Center,
Roxbury, MA United States; 2New York Academy of Medicine, New York, United
States; 3Harlem Hospital Center, New York, United States; 4North Broward
Hospital District, Fort Lauderdale, United States; 5 Washington University School
of Medicine, St. Louis, United States; 6 Multnomah County Health Department,
Portland, United States; 'Johns Hopkins University School of Medicine,
Baltimore, United States
Background: High rates of adherence to ART is critical to successful treatment,
however the optimal measures of adherence for guiding clinical practice and identifying clients needing adherence support are still unknown.
Methods: Health Resources and Service Administration, Special Projects of National Significance, funded 14 sites to evaluate adherence support programs.
Data were collected using through standardized interviews. Adherence was measured using 3-day self-report of missed doses, last time missed and number of
days/week with a missed dose. Adherence rates of 312 patients on ART at baseline in 8 of the sites were used to describe the relationships between different
adherence measures.
Results: The patients were 39% female, 68% minority, with 47% history of
mental illness and 16% active drug use; 29% had viral load <400 at baseline.
Non-adherence (>1 dose missed) by 3-day recall correlated with frequency of
days/week with a missed dose (-.56, p=.001), difficulty taking medications on time
(.29, p=.001), and last time missed dose (-.62, p=.001). No significant correlations
were found between non-adherence on 3-day recall and the following baseline
measures: difficulty following special instructions, health status, number and type
of ART medication, ART medication class, viral load and CD4 count.
Baseline Measure Missed dose No missed doses P value
on 3-day recall on 3-day recall
Days/week with missed dose=0/<1 per week 28% 81%.001
Last time missed> 1mo or never 3% 66%.001
Difficulty taking on time=never/rarely 58% 88%.001
ART-related side effect 61% 28%.001
"A lot" of difficulty taking ART due to side effects 23% 3%.001
Any side effects 61% 28%.001
Conclusions: Correlations between a number of self-report adherence measures were found in this multi-site, diverse patient population. As more data are
collected, we will be able to develop a composite adherence measure to improve
the accuracy and reliability of adherence assessment in clinical practice.
Presenting author: Lisa Hirschhorn, 55 Dimock St, Roxbury,MA 02119, United
States, Tel.: [phone] x1255, Fax: [phone], E-mail: hirschhorn
@mediaone.net
SWePeB5820 IA comparative follow-up of six different
antiretroviral combinations in a clinical setting
J. Benetucci1, A. Montero2, L. Redini 1, D.D.V. Pugliese1, L. Puga1,
A. Maranzana1, A. Thwaites1, R. Carlevaro1, D. Garone1, A.G. Moscatello1,
V. Galache', G. Ortega', S.M. Oliva1. 'Infectious Diseases Department -
Hospital Muniz, Buenos Aires, Argentina; 2Council of Scientific Research,
Natt.University of Rosario (CIUNR), Dorrego 156 50, 2000-Rosario, Argentina
Purpose: To compare six combinations of highly active antiretroviral therapy
(HAART) and to assess the conditions leading to treatment discontinuation.
Methods: 348 antiretroviral-naive HIV-1-infected patients with CD4+ cell counts
< 500 x 106/I or plasma HIV-1 RNA > 10,000 copies/ml were enrolled and randomized to one of the following combinations: 1) zidovudine (AZT) + lamivudine
(3TC) + indinavir (IDV); 2) stavudine (d4T) + didanosine (ddl) + nelfinavir (NFV);
3) d4T + 3TC + IDV; 4) AZT + 3TC + NFV; 5) d4T + 3TC + NFV or 6) d4T + 3TC
+ saquinavir (SQV). Patients were followed for 48 weeks or until therapy discontinuation. Follow-up included clinical visits and plasma HIV-RNA and CD4+ cell
measurements at baseline and at weeks 16, 32 and 48.
Results: 113 patients were lost to follow-up or excluded. The remaining 235 completed 48 weeks on treatment but 87 of them interrupted HAART.
Percentages of patients having < 500 copies/mI at weeks 16, 32 and 48 were:
d4T+3TC+SQV: 60.71%, 60.87% and 47.83%; d4T+3TC+NFV: 65.38%, 65.22%
and 55.56%; d4T+3TC+IDV: 73.33%, 75.86% and 68.75%; d4T+ddl+NFV:
76.92%, 78.95% and 76.47%; AZT+3TC+IDV: 76.47%, 92.86% and 76,92%;
AZT+3TC+NFV: 67.86%, 77.14% and 87,50%. (p=0.6281, p=0.0446 and
p=0.0180 respectively, chi-square).

Abstract WePeB5820 - Table 1. Determinants of HAART withdrawal
Causes of HAART interruption. d4T3TCIDVn= 40 d4TddlNFV n=44 AZT3TCIDV n=56 AZT3TCNFV n=40 d4T3TCNFV n=26 d4T3TCSQV n=29 p=
Failure n=44 7(17.50%) 4(9.09%) 7(12.50%) 6 (15.00%) 7(26.92%) 13 (44.83%) 0.0022*
Toxicity n=15 6(15.00%) 4(9.09%) 3 (5.36%) 0 2(7.69%) 0 0.0649
Voluntary abandon n=1 9 1 (2.50%) 5 (11.36%) 6 (10.71%) 3 (7.50%) 2 (7.69%) 2 (6.90%) 0.7130
Intolerance n=6 0 3 (6.82%) 2 (3.57%) 1 (2.50%) 0 0 0.3121
Poor-adherence n=3 1 (2.50%) 1 (2.27%) 0 0 1 (3.85%) 0 0.5785
Totaln =87 15(37.50%) 17(38.64%) 18(32.14%) 10(25.00%) 12(46.15%) 15(51.72%) 0.2421