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

596 Abstracts 32371-32375

12th World AIDS Conference

Results: Less then half of the patients used medication regularly. The most
frequent remembered reasons for irregular use are associated to forgetfulness
(20%), lack of trust in the treatment or in the need for regular use of medication
(15%) and stopping for "weekend drinking" (12%). Less then 20% refer irregular
use due to side effects. Mean CD4 counts at entry and the last count during
treatment and pre and post treatment viral load are showed below. Although the
desirable non-detectibility of viremia was not achieved in most patients, decreases
in viral load are significantly associated with the use of regular therapy.

Adherence
Regular
Irregular
p 0.032

CD4 entry
378 ~ 218
408 1 388

VL pretreatment CD4 post treatment
4.77 ~ 0.88 288 ~ 178
4.69 +0.68 254 177

VL post treatment
3.41 ~ 1.3
4.10 ~ 1.2

Conclusion: The availability of antiretroviral drugs and laboratory tests had
made a major impact in the resources for HIV infected individual, but for maximizing its effectiveness we must address other issues related to the care of this
patient. To determine the factors that may improve compliance/adherence and to
implement strategies in this direction are urgent priorities.
32371 Compliance with antiretroviral multidrug therapy in
HIV-infected patients and reasons for non-compliance
Thore Lorenzen1, A. Stoehr1, L. Weitner2, A. Adam2, J. Jarke3, A. Plettenberg1.
1AK St. Georg, Lohmuihlenstrasse 5 20099 Hamburg; 2Brennerstrasse 71
Hamburg; 3Beh6rde Fur Arbeit, Gesundheit und Soziales, AIDS-Beratung
Hamburg, Germany
Objective: To analyse the compliance with antiretroviral multidrug therapy in
HIV-infected patients corresponding to number of pills, recommended frequency
to take the drugs and the reasons for noncompliance.
Methods: In an open prospective study HIV-infected patients were asked
anonymously for prescribed antiretroviral therapy, daily time schedule and number of pills to take. In addition they were asked how often they missed some
pills or deviated from the recommended schedule in the past week. Furthermore
the reasons for missing the drugs should be specified. For all patients basal
demographic and immunological data were collected.
Results: 214 patients were included. The median age was 40 years, median
CD4-cells were 260c//1l, the median viral load was 2500 copies/mm3. The median
duration on any antiretroviral therapy was 23 months and 13 months for protease
inhibitors. 55 out of 185 patients (30%) did not take all prescribed antiretroviral pills
for at least one day and 100 out of 177 (56%) deviated from the recommended
schedule for at least one day in the week before. By analysing subgroups no
differences in compliance could be found regarding to number of pills, number of
daily doses, duration of antiretroviral therapy, number of CD4-cells and viral load.
41 of 55 patients (75%) not taking all prescribed pills and 59 of 100 patients (59%)
who deviated from the recommended schedule declared that they just forgot their
medication. 33% of the patients (18/55) respectively 19% (19/100) announced an
acute episode of illness, heavy side effects of the drugs or not enough information
about possible side effects.
Conclusion: Patient compliance is an important factor to achieve best results
in treating HIV-infection. Our study demonstrates that about 50% of patients
did not take all prescribed antiretroviral drugs in the previous week or deviated
from the recommended schedule. The most frequent reason was forgetfulness
of the patients. Unexpectedly we were unable to find differences in compliance
by analysing subgroups in regard to number of antiretroviral pills, number of daily
doses or duration of therapy.
32372 High risk behaviors are not predictive of anti-retroviral
non-adherence in HIV+ youth
Marvin Belzer', D. Fuchs2, D. Tucker2, G. Slonimsky2. 1University of Southern
California, PO Box 54700 MS#0700 Los Angeles CA 90054-0700; 2Childrens
Hospital Los Angeles, CA, USA
Background: Many clinicians feel that high risk characteristics such as substance
abuse, homelessness, lack of education, and mental illness are predictors of noncompliance with complex anti-retroviral drug regimens in HIV-infected youth.
This study explores whether these factors predict adolescent adherence with
anti-retroviral medications. Methods: Youth attending a multi-disciplinary clinic
for HIV-infected adolescents and prescribed medication for at least 3 months
were offered enrollment. A piloted questionnaire was verbally administered which
addressed self report of adherence (at 1, 7, 30, and 90 days). A chart abstract
was performed to collect data on demographics, education level, mental health,
substance abuse, amount of time in care, anti-retroviral use history, as well as,
CD4 counts and viral loads since initiating treatment.
Results: Thirty-one (of 35 potential subjects) were enrolled. Findings: 1).
Gender-61% male. 2). Ethnicity-42% Hispanic, 39% African American, 13% Caucasian. 3). Age-mean 20.25 yrs. 4). Time in care-mean 25.7 months. 5). Homeless
ever -55%. 6). Substance abuse ever-52%. 7). 29% were on 2 anti-retrovirals,
58% were on a triple regimen, and 13% were taking 4 medications (the mean
duration of treatment was 16.13 months). 8). In regard to the previous 3 months,
61% reported taking medications at least 90% of the time; 87% reported at least
75%. 9). Viral loads were reduced to undetectable in 39% of the population, while
another 26% had at least a one log drop. 10). In our sample, homelessness,
current living situation, years of education, clinical depression, substance abuse,
and even recent substance abuse did not predict non-adherence.

Conclusion: These data do not support the routine withholding of anti-retroviral
medications from high risk youth. Possible explanations include: 1). Our comprehensive clinic which provides close supervision/support may improve medication
adherence in high risk youth. 2). Patients most likely not to adhere may refuse
treatment or discontinue use shortly after starting. Longitudinal studies are needed
to substantiate these hypotheses and address long term anti-retroviral adherence
in HIV+ youth.
32373 Validity of patient questioning versus pill count as an
assessment of compliance
Gillian Hales1, J. Mitchell2, D.E. Smith3, S. Kippex4. CHRN, Level 1, 376
Victoria St. Darlinghurst, NSW; 2Community HIV Research Network, Melbourne,
Vic; 3University of NSW, Sydney; 4 Macquarie University Sydney, Australia
Background: Compliance is a critical issue in ensuring maximal effectiveness of
combination antiretroviral therapy. This is often assessed by directly questioning
patients.
Method: HIV positive participants in a primary care based phase III antiretroviral, clinical trial were invited to participate in a social research study. The trial
involved the addition of treatment A (one pill twice a day) and treatment B (one
pill three times a day) to background antivirals. Halfway through the clinical trial,
at week 28, 33 patients completed a questionnaire in which they were asked
to assess what percentage of their trial medication they believed that they had
missed. A pill count of returned trial medication was performed by the trial nurses.
Results: The median percentage that the patients believed they had omitted
to take was 5%, however the median percentage the patients had actually not
taken was 16% for treatment A and 17% for treatment B. Patient reports of their
levels of compliance was found to be a significant overestimation when tested
by Wilcoxin signed rank test (p = 0.0099 for treatment A and p = 0.0006 for pill
B). The main reasons the patient's cited for not taking their pills were simply
forgetting, not having their tablets with them at the necessary time, or getting out
of their normal rouine.
Conclusions: Patients self reports of compliance are not a reliable method
of assessing their level of pill taking. Missed doses was more common with tid
therapy compared to bid therapy.
32374 Adherence to antiretroviral therapy among HIV-positive
women
Ann Williams1, H. Wolf2, C. Yu2, M. Singh2. Yale School of Nursing 100 Church
St S., Box 9740 New Haven CT 06536-0740; 2 Yale University New Haven, CT,
USA
Objective: To describe patterns of adherence to antiretroviral therapy (ART)
among HIV infected women and to identify characteristics associated with adherence.
Design: Cross-sectional survey.
Methods: Women enrolled in an on-going clinical trail responded to interviewer
administered questionnaires between April-July, 1997. Adherence was measured
by calculating the proportion of prescribed doses reported as taken in the week
prior to interview. Median proportions of adherence were compared by Wilcoxon
rank sum test.
Results: Of 103 women completing the survey, 53 (51%) were African-American, 34 (32%) white, 16 (16%) Latina. Twenty-two (21%) reported use of heroin,
cocaine or marijuna in the past month. Medications were prescribed for 101
(99%); ART was prescribed for 89 (86%). Among the 89 women for whom ART
was prescribed, self-reported adherence ranged from 0-100% of doses in the
previous week (Mean = 86%; Median = 98%). Fifty-four percent (48/89) reporting
taking at least 97% of prescribed doses. Positive predictors of adherence included
no current illicit drug use (p = 0.02), no medication related side effects (p = 0.03),
a stable living situation (p = 0.008), feeling that the physician spent adequate time
during patient visits (p = 0.03) and an accurate understanding of the purpose of
ART (p = 0.003).
Conclusion: Medication adherence is associated with patient, provider and
medication factors. Most of these factors are amenable to interventions which
could enhance adherence.
S32375 Predictors of non adherence to HIV combination
therapies
Lee E. Klosinski1, R.N.A. Brooks2. 11313 North Vine Street, Los Angeles,
California; 2AIDS Project Los Angeles, Los Angeles CA, USA
Objectives: To identify factors that can predict non-adherence to HIV anti-viral
combination therapies.
Design: Cross-sectional, mail administered survey of clients of a large nongovernmental AIDS service organization.
Methods: We analyzed self-reported adherence information from 1,322 persons living with HIV/AIDS. The outcome variable for the study was non-adherence

to prescribed combination therapies; non-adherence was assigned if a subject
reported not taking their medications exactly as prescribed in the past 7 days.
Multiple logistic regression (backward elimination method) was used to identify
which social demographic, economic, and quality of life factors were predictive of
non-adherence among the sample population at a significance level of p < 0.05.
Results: Findings from our study indicate that certain characteristics (younger
age, government subsidized health insurance, extreme levels of anxiety, extreme