Transcript Document

The Measurement of
Antiretroviral Adherence in HIV
Sharon Mannheimer, MD
Harlem Hospital / Columbia University
Treatment Adherence Services Quality
Learning Network meeting
May 3, 2007
Overview
•
•
•
•
Background on Adherence in HIV
Adherence Measurement
CASE Adherence Index
Other Self-Report Measures
Importance of Adherence
Nonadherence associated with:
• Virologic failure
• Worse immunologic (CD4) outcomes
• Higher Hospitalization rates
• OIs / HIV disease progression
• Increased Mortality
• Resistance (at some adherence levels)
Survival vs. Adherence
Progression to death among 847 initially ART-naïve HIV+ subjects with
>12 mos. follow-up; adherence > 75% (circles) vs. adherence <75%
(squares). Hogg, et al. AIDS 2002,16:1051-8.
How much adherence is enough?
Adherence to Antiretroviral Therapy
and Virologic Failure
100
Paterson, Annals of Internal Medicine, 2000
82.1
71.4
80
66.7
54.6
60
% with
virologic
failure 40
21.7
20
0
>95
90-94.9
80-89.9
70-79.9
<70
Adherence, %
Adherence (by MEMS) significantly associated with virologic failure
(P<0.001)
Adherence Measurement
– No “gold standard”
– Many methods are impractical in clinical
settings
– Simple measures predictive of HIV
outcomes would be valuable
Classification of
Adherence Measures:
• Direct
or
• Indirect
Direct Measures
of Adherence:
• direct observation
• measuring levels of the drug in body
fluids (“Therapeutic Drug Monitoring”)
• biologic markers
• monitoring clinic attendance
Indirect Measures
of Adherence:
•
•
•
•
•
•
self-report
provider assessment
electronic monitoring devices (MEMS)
pill count
medication refill rate (pharmacy)
monitoring for an expected therapeutic
outcome
Direct vs. Indirect Measures:
• In general, direct measures are more
objective and yield more reliable
assessments of adherence
• each method has limitations
Problems with direct measures:
• Direct observation: usually not practical
• Therapeutic drug monitoring: costly,
inconvenient, not widely available, reflects
recent adherence only
• Biologic marker (e.g. HIV viral load):
may not correlate 100% with adherence,
factors other than adherence could effect
marker
• Clinic attendance: does not necessarily
correlate with medication adherence
Problems with indirect measures:
• Self report: can overestimate adherence
• Provider assessment: physicians poor at
predicting adherent behavior
• Electronic monitoring devices (e.g. MEMS
caps): costly, bulky, for only 1 drug, measures
only opening, interferes with pillbox use,
inaccuracies can occur with improper use
(e.g., pocketing doses)
• Pill count: ”pill dumping,” patient can forget to
bring bottles, does not assess timing
• Refill rate: only practical if patients use same
pharmacy, not 100% correlation
Benefits of Self-Report
•
•
•
•
Easy to administer
Inexpensive
May reveal reasons for missed doses
Self-report of nonadherence very
reliable
• Adherence measured by self-report
correlates with HIV laboratory and
clinical outcomes
Self-Report Methods
• No gold standard
• AACTG 3- or 4-day self-report format widely
used
– Day-by-day, dose-by-dose recall of each ART
med. Over prior 3 or 4 days
• Other simpler formats available:
– CPCRA 7-day self-report
– Visual Analog Scale
– CASE Adherence Index
The Case Adherence Index Questionnaire
Please ask each question and circle the corresponding number
next to the answer, then add up the numbers circled to calculate
Index score.
1. How often do you feel that you have difficulty taking your HIV medications
on time? By “on time” we mean no more than two hours before or two hours
after the time your doctor told you to take it.
4
3
2
1
Never
Rarely
Most of the time
All of the time
2. On average, how many days PER WEEK would you say that you missed
at least one dose of your HIV medications?
1
2
3
4
5
6
Everyday
4-6 days/week
2-3 days/week
Once a week
Less than once a week
Never
3.When was the last time you missed at least one dose of you HIV
medications?
1
2
3
4
5
6
Within the past week
1-2 weeks ago
3-4 weeks ago
Between 1 and 3 months ago
More than 3 months ago
Never
INDEX SCORE: ______ (> 10 = good adherence, < 10 = poor adherence)
Development of CASE Index
• Developed during a large Health Resources
and Services Administration (HRSA)-funded
evaluation study (1999-2003) of 12 US
adherence support programs
– Special Projects of National Significance (SPNS)
• Cross-site evaluation coordinated by the New
York Academy of Medicine’s (NYAM) Center
for Adherence Support Evaluation (CASE)
CASE Cross-Site Evaluation
• CASE insured uniform data collection:
–
–
–
–
Standardized core data elements
Uniform Instrument
Central interviewer and chart abstractor training
Uniform measurement periods
• Adherence questions:
– Individual questions about adherence behavior
– AACTG 3-day self-report
Adherence Intervention & Evaluation Sites
Health Services Center, Inc., Hobson City, AL
Chase Brexton Health Services, Inc., Baltimore, MD
Dimock Community Health Center, Roxbury, MA
Harlem Hospital Center, New York, NY
Helena Hatch Special Care Center, Washington University,
St. Louis, MO
Johns Hopkins University School of Medicine, Baltimore,
MD
Mission Neighborhood Health Center, San Francisco, CA
Multnomah County Health Department, Portland, OR
SUNY Downstate Medical Center, Brooklyn, NY
St. Luke's Roosevelt Hospital Center, New York, NY
North Broward Hospital District, Ft. Lauderdale, FL
Urban Health Study, San Francisco, CA
Participants in Adherence Analysis
• 1,154 participants in HRSA/SPNS crosssite study:
– enrolled between July 1, 2000 and July 1, 2003
• 524 cases included in adherence
analyses:
– Had at least 1 follow-up
– On ART at baseline and follow-up
– had corresponding CD4 and HIV RNA data at the
first 3-month follow-up
Participant Characteristics
%
Gender
Male
65
Female
34
Transgender
1
Race/Ethnicity
African American
66
Latino
5
White
26
Education
Not high school graduate
41
High school graduate or GED
36
Some college/technical school
23
Mean age, years
40.13
(SD = 8.6)
Participant Characteristics - 2
%
Self-reported HIV risk behavior
Men reporting sex with men (MSM)
29
Injecting drug use
16
MSM and injected drugs
2
Heterosexual contact
45
Heterosexual contact and injected
drugs
3
Blood transfusion, blood
components, or tissue
3
Other
2
Mean CD4 count, cells/mm3(SD)
256 (251)
Median CD4 count, cells/mm3
193
Mean log10 HIV RNA level (SD)
3.99 (1.35)
Analysis of HRSA/SPNS Cross-Site
Adherence Data:
Development of the CASE Index
• Principal component analysis performed:
– Responses to 3 adherence questions
explained 69% of total variation in
adherence, higher than any other
combination of questions
– Responses to each of the 3 CASE
questions carried approximately equal
importance
Mannheimer, et al. AIDS Care 2006;18:853-861.
The Case Adherence Index
3 adherence questions:
– 1. Frequency of “difficulty taking HIV medication on
time (no more than two hours before or two hours after
the time your doctor told you to take it)” –
• Response options: Never, Rarely, Most of the time, or All of the time
– 2. frequency of “average number of days per week at
least one dose of HIV medications was missed”
• Response options: Everyday, 4-6 days per week, 2-3 days per week,
Once a week, Less than once a week, or Never
– 3. “Last time missed at least one dose of HIV
medications”
• Response options: Within the past week, 1-2 weeks ago, 3-4 weeks
ago, 1 to 3 months ago, More than 3 months ago, or Never
The Case Adherence Index –
Statistics / Scoring
• Responses coded:
– For #1 (reverse coded) - possible range of 1 to 4
points
– For #2 and #3 - possible range of 1 to 6 points
• Composite score obtained by adding
responses:
– Range 3 to 16
– Higher scores indicate better adherence
The Case Adherence Index Questionnaire
Please ask each question and circle the corresponding number
next to the answer, then add up the numbers circled to calculate
Index score.
1. How often do you feel that you have difficulty taking your HIV medications
on time? By “on time” we mean no more than two hours before or two hours
after the time your doctor told you to take it.
4
3
2
1
Never
Rarely
Most of the time
All of the time
2. On average, how many days PER WEEK would you say that you missed
at least one dose of your HIV medications?
1
2
3
4
5
6
Everyday
4-6 days/week
2-3 days/week
Once a week
Less than once a week
Never
3.When was the last time you missed at least one dose of you HIV
medications?
1
2
3
4
5
6
Within the past week
1-2 weeks ago
3-4 weeks ago
Between 1 and 3 months ago
More than 3 months ago
Never
INDEX SCORE: ______ (> 10 = good adherence, < 10 = poor adherence)
CASE Index Compared to
AACTG 3-day Self Report
• CASE Index’s sensitivity and specificity
relative to 3-day self-report at cut-off of ≥ 95%
was calculated
• Based on the analysis, CASE Index was
recoded as a dichotomy where:
– CASE Index scores > 10 indicated high adherence
– CASE Index scores ≤ 10 indicated low adherence
Sensitivity and Specificity
of CASE Adherence Index
vs. 3-day recall adherence self-report
CASE Index Score
Sensitivity (%)
Specificity (%)
5
99.28
8.00
6
98.32
18.00
7
96.64
30.00
8
92.57
52.00
9
84.65
77.00
10
74.10
99.00
11
62.83
99.00
12
55.64
100.00
13
48.20
100.00
14
38.61
100.00
15
25.42
100.00
CASE Index Concurrent Validity
with 3-day Self Report
• Logistic regression showed strong correlation:
– Odds of 3-day self-report > 95% was at least 60 times more
for patients with a CASE Index score > 10 compared to
those with a CASE Index score ≤ 10 (p < 0.001) across four
serial cross-section follow-up periods (3, 6, 9 and 12 months
after enrollment)
• Receiver Operating Characteristic Curves (ROC)
showed a very strong association between 3-day
self-report at 95% and CASE Adherence Index
Scores (>10 vs. < 10) across the four measurement
quarters
Mannheimer, et al. AIDS Care 2006;18:853-861.
Relationships between
Self-reported Adherence Measures and HIV RNA
• CASE Adherence Index was strongly
associated with:
– a 1 log decrease in HIV RNA levels (p ≤ 0.05)
– achieving HIV RNA < 400 copies/ml (p ≤ 0.05)
• Association between 3-day self-report with
HIV RNA was not as strong:
– significance for a 1-log decrease from baseline
HIV only at 6-month follow-up
– significance for HIV < 400 copies/ml also only at
6-month follow-up
Relationship between Self-reported Adherence
and HIV RNA
Odds ratio estimates
3 Months 6 Months 9 Months 12 Months
Adherence
Measure
Comparison
3-Day SelfReport
> 95%
vs. ≤
95%
HIV RNA
Measure
1-log
Decrease
HIV RNA
< 400
CASE
Adherence
Index
> 10
vs. ≤ 10
1-log
Decrease
HIV RNA
< 400
Note: * p < 0.10
** p < 0.05
1.23
2.26**
1.53
1.16
0.97
2.30**
1.66*
1.23
1.52*
1.90**
1.76**
2.13**
1.60**
1.68**
1.87**
1.60*
Relationships between
Self-reported Adherence Measures
and CD4 Lymphocyte Counts
• A significant relationship between CASE Adherence
Index and changes in CD4 lymphocyte counts from
baseline only at 12 months
• There were no observed relationships between
changes in CD4 and 3-day self-report
Limitations of HRSA/SPNS data
• Only 524 of 1,154 individuals in local sites’
adherence programs were included in
adherence analyses
• ART Naive and ART experienced but not
currently receiving ART were excluded from the
analyses
• High attrition rates
• Social desirability of self-report
• Adherence instruments administered in same
interview
• Only self-reported adherence
CASE Index Summary
• a new measure of self-reported ART
adherence
• easy to administer and score
• high degree of sensitivity and specificity with
the 3-day self-report (concurrent validity)
• a better predictor of HIV RNA changes over
time than 3-day self-report
Other Self-Report Methods
• AACTG 3- or 4- day recall
• CPCRA 7-day recall
• Visual Analog Scale
AACTG 3-day recall
CPCRA 7-day recall
Virologic Outcome by Adherence*
in two CPCRA Antiretroviral Trials
Mannheimer, et al. CID 2002
80
70
60
% HIV
RNA
<50
copies
/ml
50
100%
80-99%
0-79%
40
30
20
10
0
Month 1
Month 4
Month 8
)
(n=1074)
(n=922)
(n=699)
Month 12
(n=531)
P < 0.005 for difference between categories at months 4,8,12
*by adherence self-report
C•P•C•R•A
Immunologic Outcome by Adherence*
in two CPCRA Antiretroviral Trials
Mannheimer, et al. CID 2002
190
170
150
Change in
CD4
(cells/mm3)
from
baseline
130
100%
110
80-99%
90
0-79%
70
50
30
10
Month 1
Month 4
Month 8
(n=1074)
(n=922)
(n=699)
Month 12
(n=531)
P < 0.05 for difference between categories at months 4,8,12
*by adherence self-report
C•P•C•R•A
Visual Analog Scale
“Put a cross on the line below at the point showing your
best guess about how much medication you have
taken in the last month.
We would be surprised if this was 100% for most
people, e.g. 0% means you have taken no
medication; 50% means you have taken half your
medication; 100% means you have taken every
single dose of medication.”
____________________________________________
___
0
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Summary
• Adherence critical for successful HIV
treatment
• Many methods available for measuring
adherence
• CASE Index
– easy to administer and score
– correlates with HIV RNA outcomes