Transcript The Impact of Alcohol Consumption on HIV Disease Progression

ISBRA 2006
HIV and Alcohol
Symposium
Jeffrey H. Samet, MD, MA, MPH, Chairman
Evgeny M. Krupitsky, MD, Phd, Co-Chairman
The Impact of Alcohol
Consumption on HIV
Disease Progression
ISBRA 2006 HIV and Alcohol Symposium
Jeffrey H. Samet, MD, MA, MPH
Chief, Section General Internal Medicine
Boston Medical Center
Professor of Medicine and Public Health
Boston University Schools of Medicine and
Public Health
Authors
Samet JH1, Cheng DM1, Libman H3, Nunes D1,
Alperen J1, Faber V6, Saitz R1
1Boston
Medical Center, Boston University School of
Medicine, United States;
2Beth Israel Deaconess Medical Center, Harvard
Medical School, United States;
3 DM-STAT, United States
Funded by the National Institute on Alcoholism and Alcohol Abuse: R01AA13216, R01-AA11785, & R01-AA10870 and USPHS M01-RR00533 (GCRC)
Background

Alcohol use is common among HIVinfected persons
– 36% of HIV-infected veterans (n=881)
were current hazardous drinkers*
– 42% of HIV-infected patients establishing
primary care (n=664) had history of
alcohol problems**
*Conigliaro, Gordon, McGinnis, Rabeneck, Justice. JAIDS. 2003;33:521-525.
**Samet, Phillips, Horton, Traphagen, Freedberg. AIDS Res Hum
Retroviruses. 2004;20:151-155.
Background

The impact of alcohol use on HIV disease
progression is unclear.
– Pre-HAART (circa 1996), no association found *
– Among persons receiving antiretroviral therapy (ART)
between 1997-2000 cross-sectional evidence found an
association of heavy alcohol use with lower CD4 and
higher HVL (HIV viral load).**
*Dingle, Oei. Psychol Bull. 1997;122:56-71.
**Samet, Horton, Traphagen, Lyon, Freedberg. Alcohol Clin Exp Res.
2003;27:862-7.
Background

Potential mechanisms of alcohol’s
impact on disease progression
– Decreased medication adherence.*
– Physiological impact is suspected from
studies in rhesus macaques.†‡
*Cook RL, et al. J Gen Intern Med. 2001;16:83-88.
† Bagby GJ, et al. Alcohol Clin Exp Res. 2003;27:495-502.
‡Stoltz
DA, et al. Am J Respir Crit Care Med. 2000;161:135-140.
Hypothesis

Alcohol consumption is associated with
more rapid HIV disease progression:
– CD4 decrease
– HVL (i.e. HIV RNA) increase
Participants & Design

Two consecutive prospective cohorts
of HIV-infected persons with current
or past alcohol problems
HIV-ALC (HIV-Alcohol Longitudinal
Cohort): 7/97-7/01
HIV-LIVE (HIV-Longitudinal
Interrelationship between Viruses
and Ethanol): 8/01-03/06
Eligibility criteria


Inclusion Criteria
– HIV infection
– Two or more positive CAGE* responses
– Fluent in English or Spanish
Exclusion criteria
– Mini Mental State Examination** score < 21
– Plans to move from area in next year
*Ewing. JAMA. 1984;252:1905-07.
**Folstein et al. J Psychiatr Res. 1975;12:189-98.
Subject Assessment

Interview, medical record, and/or
phlebotomy at 6-month intervals for up to 7
years (1997-2006) for the following:
–
–
–
–
–
CD4
HVL
ART
ART adherence
alcohol and drug use
Primary Outcome
Measures



CD4 cell count per µL
log10 HVL (HIV RNA copies per mL)
Obtained within 3 months of
assessment interview
Primary Independent
Variable

Past 30-day alcohol use:
– Heavy


> 4 drinks on any day or >14 drinks/week in men
>3 on any day or >7 drinks/week in women
– Moderate (alcohol use less than “heavy”)
– Abstinent
Other Independent
Variables









Gender
Age
Race (black, white, or other)
HIV risk factor (injection drug use, men having sex
with men, or heterosexual behavior)
Homelessness (> 1 night in past 6 months)
3-day adherence to ART (100% adherence,
[yes/no])
Time since study enrollment
Year of study entry
Cohort study participation (HIV-ALC vs. HIV-LIVE)
Analysis




Generalized linear mixed effects models
Stratified by ART use (on/off) to account for
possible effect modification
The data were restricted to observations
beginning at baseline until a change in ART
usage occurred (i.e., went on or off ART)
Regression analyses controlled for baseline
CD4 counts
Results: Cohort (N=595)
Only in
HIV ALC
N=195
In HIV
ALC &
HIV
LIVE
N=154
Only in
HIV LIVE
N=246
Baseline Characteristics
(N=595)
Characteristic
%
Male
75
Race
Black
White
Other
41
34
25
HIV risk group
Hetero/Blood
Inject Drug
Men Sex Men
24
54
21
Currently receiving ART
60
Baseline Characteristics
(N=595)
Characteristic
CD4
HVL (n=557)
Log 10 HVL
Age, years
Mean (SD)
421 (287.4)
153,655 (1,113,963)
3.3 (1.2)
41 (7.4)
Baseline Characteristics
(N=595)
Baseline drinking
status
60%
30%
10%
Abstinent
Moderate
Heavy
Results: Observations
(N=595)

CD4 analyses observations = 1495

HVL analyses observations = 2031
Results: Multivariable Analyses
Adjusted mean differences in CD4 and Log10 HVL associated
with alcohol use
ART Status
Alcohol
consumption
CD4
cell count (SE)
Log10 HVL (SE)
On ART
(n=355)
Abstinent
--
--
Moderate
12.31 (13.8)
0.03 (0.08)
Heavy
-1.46 (10.9)
0.13 (0.07)†
--
--
-27.03 (18.3)
-0.11 (0.08)
Not on ART Abstinent
(n=240)
Moderate
Heavy
†p=0.09
*p=0.02
-53.4 (22)
*
0.0003 (0.08)
Limitations



Participants in the no ART group may have
been exposed to these medications in the
past but were no longer receiving them at
the time of study entry.
Observational cohort: possible uncontrolled
confounding
Inconsistent time frames: alcohol assessed
30 days prior to the interview; CD4 & HIV
RNA within 3 months
Conclusion


In those on ART, heavy drinking was
possibly associated with higher HVL.
In those not on ART, heavy drinking
was associated with lower CD4 cell
counts.
Implications


Avoiding alcohol consumption at heavy
levels may have a beneficial effect on
HIV disease progression.
Determining the behavioral and/or
biological basis for these effects and
addressing alcohol use in HIV-infected
patients are important research and
clinical issues.
Reduction of risky sexual
behavior among hospitalized
Russian substance dependent
patients
The Russian Partnership to Reduce the
Epidemic Via Engagement in Narcology
Treatment (Russian PREVENT) Study
ISBRA-2006 Symposium "Alcohol and HIV"
Supported by National Institute on Alcohol Abuse and Alcoholism (NIAAA), NIH:
R21-AA014821
Krupitsky E.1, Cheng D.M.2, Raj
A.2, Egorova V.1, Levenson S.2,
Bridden C.3, Zvartau E,1 Samet
J.H.2,3
1St.
Petersburg State Pavlov Medical University,
Russian Federation; 2Boston University School of
Public Health, United States;
3 Boston University School of Medicine, Boston
Medical Center, United States
Prevalence of
addictions
Prevalence of drug addictions in the Leningrad Region (number of
subjects per 100.000 of general population)
250
200
150
100
50
0
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005 Years
Prevalence of registered HIV positive individuals in the
Leningrad Region
8000
7234
7000
6241
6000
5304
5000
4258
4000
2755
3000
2000
843
1000
6
12
63
0
1997
1998
1999
2000
2001
2002
2003
2004
2005
HIV positive individuals
i. v. drug users
alcoholics
Leningrad Regional HIV/AIDS Center Data
100%
100%
6
100%
100%
6
51
100%
100%
1912
780
94.2%
100%
1503
100%
1046
100%
937
100%
993
91.2%
90%
82.2%
80%
70%
66.7%
66%
66.7%
60%
60%
47.4%
50%
45%
40%
30%
20%
10%
0%
0%
3.8%
3.6%
0.6%
0%
4.5%
4.6%
4%
0%
1997
1998
1999
2000
2001
2002
2003
2004
2005
Background



Russia has one of the highest per capita
alcohol consumption rates in the world.
The Russian HIV epidemic is propelled by
injection drug use.
Alcohol use may increase high-risk sexual
behaviors and promote spreading of HIV from
IDUs into the general population.
Background


Regional narcology hospitals play a central
role in Russia’s efforts to address alcohol and
drug dependence but have not aggressively
addressed HIV.
Risky sexual behaviors need to be addressed
with effective and feasible interventions
among Russian substance dependent
persons.
Purpose

To assess the effectiveness of a sexual risk
reduction intervention in the Russian
narcology hospital setting
Hypothesis

Subjects receiving the intervention will report
fewer risky sexual behaviors.
Design and Setting

Randomized controlled trial (RCT)
•
Recruited 10/04 through 4/05

Two narcology hospitals in the
Leningrad Region of Russia
– Leningrad Region Center for Addictions
(LRCA)
– Medical Narcology Rehabilitation Center
(MNRC)
– Narcology hospitals provide 1 week
detoxification and 2-3 weeks stabilization.
Participants


181 subjects with alcohol and/or heroin
dependence
Eligibility criteria:
– age > 18
– unprotected sex in last 6 months
– willing to undergo HIV testing
– abstinent from substances for > 48 hours
Behavioral Intervention



Culturally and contextually adapted CDCendorsed RESPECT brief counseling designed
to reduce sex-risk behaviors*
Received 30 condoms at baseline
2 intervention sessions at medical center plus
3 monthly booster sessions via telephone
Behavioral Intervention


Session 1 (30-40 min)
– Personal assessment of HIV risk
– Increasing HIV risk perceptions
– Negotiating a personalized risk reduction plan
Session 2 (60 min)
– Provision and discussion of HIV test results; review plan
– Promotion of safer sex: condom skills; self-efficacy;
emphasizing relationship between alcohol and sexual
risk
– For HIV-infected subjects: skills building to reduce
violence and stigmatization when disclosing to partners
– For injection drug users: education and skills building to
promote new needle usage and cleaning of
needles/works
Behavioral Intervention

3 monthly booster sessions (10-20
min)
– Provision of ongoing case management
tailored to the individual’s stage of
readiness to engage in sexual or drug use
risk reduction
Control Program



Usual addiction treatment, which includes
no sexual behavior counseling
Received 30 condoms at baseline
HIV-infected controls received brief post-test
counseling
– Provision and discussion of HIV test results;
creation of risk reduction goals
– Referral to an HIV care program

3 monthly study check-in phone calls (3
min)
Subject Assessment


Assessed at baseline, 3, and 6 months
Baseline and 6 month assessment via
face-to-face interviews with the Risk
Assessment Battery and Time Line
Follow Back survey.
– HIV risk behaviors additionally assessed by
Audio Computer-Assisted Self Interviewing
(ACASI) System (to promote truth telling)

3-month assessment intervieweradministered via the telephone
Outcomes


*In
Assessed at 6 months*:
– Primary:
 percentage of safe sex episodes
 consistent safe sex (yes/no)
– Secondary:
 any condom use
 number of unsafe sex episodes
Assessed at 3 months (all secondary)†:
– percentage of safe sex episodes
– consistent safe sex (yes/no)
– any condom use
the past 3 months, ACASI
†In
the past 3 months, telephone interview
Primary Outcomes
(assessed at 6-month follow-up visit by ACASI)

Percentage of safe sex episodes
(continuous variable)
– number of times condoms were used out
of the number of sexual episodes (anal
and vaginal intercourse)

Consistent safe sex (yes/no)
– 100% condom use during anal and
vaginal intercourse or abstinence from
sex
Secondary Outcomes

6 months:
– Any condom use (yes/no)
– Number of unsafe sex episodes


no condom use during anal or vaginal sex
3 months:
– percentage of safe sex episodes
– consistent safe sex (yes/no)
– any condom use (yes/no)
Analysis




Intent-to-treat
Descriptive statistics (e.g., medians,
interquartile ranges [IQR], and proportions)
were used to characterize the sample by
treatment group.
Chi-square, Fisher’s exact, or Wilcoxon rank
sum tests used as appropriate for dichotomous
and continuous variables.
Additional analyses using logistic regression
and median regression models to adjust for
possible group differences at baseline
Results
Baseline Demographic Characteristics
Characteristic
Control
n=87
Intervention
n=94
P-value
Age, Median
(IQR)
31
(26-40)
30
(25-39)
0.36
Male
68 (78%)
67 (71%)
0.31
Employed full time
42 (48%)
47 (50%)
0.88
Heterosexual
84 (97%)
90 (96%)
0.45
Diagnosis
Alcohol
Heroin
Dual
55 (63%)
27 (31%)
5 (6%)
53 (57%)
31 (33%)
10 (11%)
0.42
HIV infected
11 (13%)
16 (17%)
0.53
Follow-Up
Follow-up was 90% (162/181) at
3 months and 80% (144/181)
at 6 months, with no differential
follow-up between intervention
groups
Results - Primary
Percentage of Safe Sex
% of Safe Sex Episodes, Past 3 Months,
Median (IQR)
Baseline
3 Months†
6 Months *
Control
Intervention
P-value
*In
8
(0-25.0)
0
(0-21.4)
0.09
the past 3 months, ACASI
50
37
(12.5-100.0)
(0-100.0)
67
80
(14.1-100.0) (25.9-100.0)
0.98
0.02
†In
the past 3 months, telephone interview
The Effect of the PREVENT Intervention on
Median Percentage of Safe Sex Episodes
100
Median percentage of safe sex
episodes
90
80
70
60
Intervention
50
Control
40
30
20
10
0
0
3
Months
6
Results - Primary
Consistent Safe Sex
Subjects Reporting Consistent Safe Sex,
Past 3 Months
Baseline
3 Months
6 Months
Control
Intervention
P-value
4
(5%)
2
(2%)
0.37
28
(36%)
30
(36%)
0.98
19
(29%)
29
(40%)
0.18
Results – Secondary
Any Condom Use
Subjects Reporting Any Condom Use,
Past 3 Months,
Control
Intervention
P-value
Baseline
3-Months
6Months
50
(57%)
41
(44%)
67
(86%)
67
(80%)
46
(69%)
65
(84%)
0.06
0.30
0.02
Results- Secondary
Unsafe Sex Episodes
# Unsafe Sex Episodes, Past 3 Months,
Median (IQR) [n]
Control
Intervention
P-value
Baseline
3 Months
6 Months
20 (645)
[n=85]
12 (536)
[n=89]
6 (0-16)
[n=78]
6 (0-30)
[n=66]
5 (0-12.5)
[n=84]
3 (0-15)
[n=73]
0.50
0.045
0.23
ResultsDependence Category
% Safe Sex Episodes, Past 3 months,
Median (IQR) [n]
Alcohol
Control
30.0 (0-78.6)
[n=41]
Interventio 90.0 (23.8-100.0)
n
[n=39]
P-value
0.007
Heroin
59.1 (0-100)
[n=22]
74.2 (39.1-100)
[n=24]
0.49
ResultsDependence Category
Percentage with Consistent Safe Sex,
Past 3 months, [n]
Alcohol
Heroin
Control
Intervention
P-value
22%
[n=41]
48%
[n=39]
0.01
32%
[n=22]
29%
[n=24]
0.85
Limitations
•
Use of self-reported instruments/assessments
•
Possibility of social desirability bias
•
No objective biological outcomes (e.g. STDs or new
HIV infection) assessed
Conclusions


Adaptation of a pragmatic sexual risk
reduction intervention in two Russian
narcology hospitals reduced risky sexual
behaviors in substance dependent persons.
Dissemination of this effective intervention
in comparable settings could be one
component of a broad strategy needed to
reduce the risk of HIV infection in Eastern
Europe and other settings.
Current Alcohol Consumption and
Cardiovascular Disease among Men
Infected with HIV
Matthew Freiberg, MD, MSc
University of Pittsburgh, USA
Alcohol and HIV Symposium
ISBRA 2006 World Congress on Alcohol Research
Sydney, Australia
September 11th, 2006
Alcohol Consumption

In the general population
– 17.6 million adults abuse alcohol or are
alcohol dependent1

Among those with HIV
– 40-50% have a history of alcohol abuse or
dependence2
Grant BF,: The 12-month prevalence and trends in DSM-IV alcohol abuse and
dependence: United States, 1991-1992 and 2001-2002. Drug Alcohol Depend
74:223-234, 2004.
2 Lefevre F et al. Alcohol consumption among HIV-infected patients. J Gen Intern
Med 10:458-460, 1995
1
Cardiovascular Disease

In the general population
– Cardiovascular disease (CVD) is the leading
cause of death in the United States1

Among those with HIV
– Combined Antiretroviral Therapy (ART) is
associated with an increased risk of myocardial
infarction2
– ART is associated with increased insulin
resistance and dyslipidemia
Mokdad AH et al: Actual causes of death in the United States, 2000.
JAMA 291:1238-1245, 2004.
2 Friis-Moller N et al.: Combination antiretroviral therapy and the risk of
myocardial infarction. N Engl J Med 349:1993-2003, 2003.
1
Alcohol Consumption and
CVD

Among those without HIV
– “J”-shaped relation between alcohol and
CHD risk1

Mechanism of action
– Increased insulin sensitivity
– Increased HDL cholesterol

Among those with HIV, however…
– data are sparse
Corrao G, et al.: Alcohol and coronary heart disease: a meta-analysis.
Addiction 95:1505-1523, 2000
1
The Present Study

Specific Aims
–
–
To evaluate the cross-sectional association
between current alcohol consumption and
prevalent CVD among male veterans infected
with HIV using multivariable logistic regression
To determine if the relationship between
current alcohol consumption and prevalent
CVD is the same for male veterans with HIV as
compared with male veterans without HIV
The Present Study

Hypotheses
– The relationship between current
alcohol consumption and prevalent CVD
will be “J” shaped for male veterans
with and without HIV but….
– The observed benefit of current
moderate alcohol consumption will be
less in HIV infected male veterans
Research Design

Veterans Aging Cohort Study (VACS)
– Observational longitudinal cohort of U.S.
veterans
– 2979 HIV+ and 3019 HIV- age, race/ethnicity,
site matched comparison participants
– Uses data from provider surveys and electronic
medical record reviews (including laboratory and
pharmacy data) from 8 Veteran Affairs Medical
Center GIM and ID clinics
Subjects, Eligibility, Data

Subjects were eligible for the present study
if
– They were a male VACS participant
– Responded to provider surveys and answered
questions regarding alcohol consumption,
covariates, and prevalent CVD outcomes
– Were current alcohol consumers


All data for the present study are from the
baseline examination
The present study contains 2028 HIV+ and
1927 HIV- participants
Dependent variable

Total cardiovascular disease (CVD):
defined as a yes response to one of
the following questions, “Has a doctor
ever told you that you had…
(1)
(2)
(3)
(4)
“…angina or CHD,”
“…a myocardial infarction,”
“…congestive heart failure,” OR
“…a stroke or TIA.”
Independent Variable
(Alcohol)

Number of drinks per week
– Constructed from the Alcohol Use Disorders
Identification Test (AUDIT)
– Using quantity and frequency questions:


When you are drinking how often do you have a drink
containing alcohol? Never, monthly or less, 2-4 x per
month, 2-3 x per week, 4+ x per week
How many drinks containing alcohol do you have on a
typical day when you are drinking? 1-2, 3-4, 5-6, 7-9,
10 or more
Independent Variable
(Current Alcohol
Consumption)



Hazardous drinking: > 14 drinks a
week or 6 or more drinks on one
occasion
Moderate drinking: 1-14 drinks a week
and no binge drinking
Infrequent drinking: <1 drink per
week (referent)
Covariates




Age
Race (White, Black, Other)
Height
Weight
Covariates

Self-reported
–
–
–
–
High cholesterol, lipids, or triglycerides
Diabetes or high blood sugar
Hypertension or high blood pressure
Current smoking: defined as “Do you now smoke
cigarettes?” (i.e. within the last week)
– Liver disease or (bad liver) or Cirrhosis
– Kidney failure or (bad kidneys)
– Regular exercise: defined as engaging in regular
activities (e.g., brisk walking, jogging) long
enough to work up a sweat at least 3 times a
week
Covariates



Hepatitis C virus (HCV) status: defined as a
positive Hepatitis C antibody test or HCV
RNA test
CD4 count: closest lab value -180 days to
+7 days from the time of enrollment
Current antiretroviral therapy (ART) use:
defined as any ART use -90 days to +7 days
from the time of enrollment based on
survey and pharmacy data
Analysis


Descriptive statistics
Multivariable logistic regression models
– Model 1: Age adjusted
– Model 2: Model 1 + demographics +
traditional CHD risk factors
– Model 3: Model 2 + ART + CD4 +HCV
– Model 4: Model 3 + remaining covariates
Demographics
Demographics
Median age (yr)
HIV+
N=2028
49
HIVN=1927
50
Race (% black)
68
63
Median height
(inches)
Median weight (lbs)
70
70
175
195
Traditional CHD Risk
Factors
Traditional CHD
Risk Factors
HIV+
HIVN=2028 N=1927
Hypercholesterolemia (%)
28.3
36.7
Diabetes (% )
15.1
24.9
Hypertension (%)
32.7
46.8
Current smoking (%)
53.8
46.1
Non-Traditional CHD Risk
Factors
Non-Traditional CHD
Risk Factors
HIV+
HIVN=2028 N=1927
Hepatitis C (%)
33.6
16.2
Median CD4 count cells/mm3*
367
--
Current antiretroviral use (%)**
81.7
--
*Data available for n=1995
**Data available for n=1771
Other Covariates
Other Covariates
HIV+
HIVN=202 N=192
8
7
Liver disease (%)
16.9
9.9
Kidney disease (%)
4.6
3.5
Regular exercise (%)
54.9
55.3
Prevalent Cardiovascular
Disease
Prevalent Cardiovascular Disease
HIV+
N=2028
5.2
HIVN=1927
10.0
Heart attack or MI (%)
4.2
7.7
Congestive heart failure (%)
3.9
5.4
Stroke (%)
5.0
5.5
Total CVD (%)
12.2
17.3
Angina or CHD (%)
Current Drinkers
HIV+
N=2028
HIVN=1927
Infrequent
Drinker
% (n)
23.9
(484)
Moderate
Drinker
% (n)
Hazardous
Drinker
% (n)
25.1
(508)
51.1
(1036)
23.6
(454)
22.9
(441)
53.4
(1032)
Prevalent CVD among
Current Drinkers
HIV+
N=2028
HIVN=1927
Infrequent
Drinker
%
11.6
Moderate
Drinker
%
Hazardous
Drinker
%
10.0
13.6
18.3
17.0
17.0
Prevalent Total CVD among
Current Drinkers with HIV*
Moderate Drinker
Model
Model
Model
Model
1
2
3
4
0.86
0.93
1.06
1.09
(0.57-1.29)
(0.61-1.42)
(0.68-1.67)
(0.69-1.72)
Hazardous Drinker
1.27
1.36
1.52
1.63
(0.91-1.77)
(0.96-1.93)
(1.04-2.24)
(1.10-2.41)
*Infrequent drinkers were the referent group
Model
Model
Model
Model
1=Adjusted for age
2=Model 1 + demographics and traditional CHD risk factors
3=Model 2 + Non-traditional CHD risk factors (ART, CD4, and HCV)
4=Model 3 + remaining covariates
Prevalent Total CVD among
Current Drinkers without
HIV*
Model
Model
Model
Model
1
2
3
4
Moderate Drinker
Hazardous Drinker
0.72 (0.50-1.04)
0.77 (0.53-1.12)
-0.74 (0.51-1.09)
0.92 (0.68-1.24)
0.92 (0.68-1.26)
-0.92 (0.67-1.27)
*Infrequent drinkers were the referent group
Model
Model
Model
Model
1=Adjusted for age
2=Model 1 + demographics and traditional CHD risk factors
3=Model 2 + Non-traditional CHD risk factors (ART, CD4, and HCV)
4=Model 3 + remaining covariates
Limitations





Possible non-differential misclassification
associated with self-reported outcomes
Possible differential misclassification
associated with HIV and frequent health
care visits
Hepatitis C laboratory screening was
provider dependent
Cannot comment on cause and effect
Veterans may not be generalizable to other
populations
Conclusions

A “J” shaped relationship between alcohol
and prevalent CVD was observed for HIV+
and HIV- veterans
– For HIV+ veterans the J shaped relationship was
not present after adjusting for confounders
– For HIV- veterans, the J shaped relationship
remained after adjusting for confounders but
was not statistically significant
Acknowledgements


Funding: National Institute of Alcohol Abuse and
Alcoholism (NIAAA) Grants 5U01AA013566 and
7K23AA015914
Co-Investigators:
–
–
–
–
–
–
–
–
–
Amy Justice and the VACS Project Team
Jeffrey Samet
Lewis Kuller
Kevin Kraemer
R. Curtis Ellison
Richard Saitz
Arlene Ash
R.S. Vasan
Lewis Kazis
Missing Data




VACS cohort has 5988
1454 participants were former or
never consumers of alcohol or did not
respond to one of the quantity
frequency questions
301 participants were women
208 participants were missing data on
covariate data
Prevalent Total CVD among
Current Drinkers with HIV*
Moderate Drinker
Model
Model
Model
Model
1
2
3
4
0.98
1.08
1.06
1.09
(0.63-1.52)
(0.69-1.69)
(0.68-1.67)
(0.69-1.72)
Hazardous Drinker
1.40
1.54
1.52
1.63
(0.97-2.03)
(1.05-2.26)
(1.04-2.24)
(1.10-2.41)
*Infrequent drinkers were the referent group and sample restricted to those
with data for ART and CD4 (n=1742)
Model 1=Adjusted for age
Model 2=Model 1 + demographics and traditional CHD risk factors
Model 3=Model 2 + Non-traditional CHD risk factors (ART, CD4, and HCV)
Model 4=Model 3 + remaining covariates
Alcohol & HIV:
Developing Interactive
Computerized Brief
Interventions
Joseph Conigliaro, MD, MPH
Center for Enterprise Quality and
Safety
University of Kentucky
Alcohol Use and Abuse




90% currently use or have used
alcohol
14% report abuse or dependence
Major factor in hospital, emergency
visits, sick days & accidents
Economic burden
Alcohol & HIV
Veterans Aging Cohort Study
 914 HIV (+) patients
– 15% hazardous drinkers (AUDIT)
– 13% drank more than 30 drinks per month

Hazardous drinkers
– More often had detectable VL [> 500 copies/ml] (70% vs. 55%;
P= .001) compared to non-hazardous drinkers
– Higher AST and ALT levels

Multivariate analysis (antiretroviral therapy, age, drug use,
and HIV risk factor)
– Hazardous drinkers were 1.8 (95% CI 1.16-2.64) times as likely
as non-hazardous drinkers to have a detectable VL
– Consumption above 30 drinks/month associated with increased
odds of detectable VL (OR=1.82; 1.17-2.86)
Alcohol & HIV


Significant implications for clinical
management and outcomes
Associated with increased morbidity &
mortality, rapid disease progression,
poorer adherence to antiretroviral
regimens, and viral resistance
Institute of Medicine

Providers should be able to:
– identify
– treat alcohol problems
– refer for specialist treatment

Unique position for early detection & Rx
–
–
–
–


prevalence
patient access
Linkage of medical problems
Rapport with patient
Lack expertise and capability
Limited time and resources
Brief Interventions




Reduce alcohol consumption
Decrease alcohol related complications
Reduce alcohol related health care costs
Not routine practice
Interactive Computer
Programs & BIs









Assess drinking status & readiness to change
Initiate provider delivered BIs
Prepare patient & provider for targeted session
Saves time
Reduce time lag between assessment and feedback.
Facilitate individualized feedback immediately upon
submission of data
Provide lower-cost and customized intervention to more
drinkers
Provide anonymity, convenience –can be done anytime, day or
night
Feedback objective and not influenced by counselor bias
Internet to Reduce Problem
Drinking




Computers, and the internet, have become
integral part of life
Approx 80 % of internet "surfers" in the US
have reportedly used it to access health
information
In-person brief motivational interventions
are currently offered via the internet
Drinkers may prefer this format
– way to save face
– can begin to look at their drinking in private and
nonjudgmental way
AlcoholScreening.org



Anonymous
Free online service
Offers visitors
– self-screening of drinking behaviors
– individualized feedback
– when appropriate, information about
treatment
AlcoholScreening.org


Examined whether the site reached potential hazardous
drinkers.
14-months
– over 66,000 visitors
– nearly 40,000 > age 18 completed screen about drinking
habits



90% of all visitors who completed screen were hazardous
drinkers (by AUDIT, and 2 quantity and frequency questions
- >14 drinks per week or >4 drinks per occasion for men,
and >7 drinks per week or >3 drinks per occasion for
women)
65% had possible alcohol abuse or dependence
After receiving results, 19% chose “Learn More” or “Get
Help” options
The Drinker's Check-up


internet equivalent of 2-3 face-face sessions with
counselor
same elements of original DCU
– drinker's risk factors, family history, alcohol & drug use,
alcohol-related problems, symptoms of dependence, &
motivation for change
– objective feedback based on answers;
– module to resolve ambivalence about whether to change



helps users decide to change their drinking
goals of change –stopping or cutting back
heavy drinkers increased internal motivation for
change and reduced drinking, alcohol-related
problems and symptoms of dependence by 50 % at
12-months
The e-CHUG




web-based version of the Check-Up to Go
(CHUG) mailed feedback instrument
proven effective in college trials
favorable to more lengthy prevention
programs and may increase the impact of
educational or skill-based prevention efforts
provides information about personal
consumption, potential risk factors, and
comparison to campus norms
Current Internet
Programs






Accessible to those with
computer/internet
Geared toward younger persons
Not specific to HIV
Not linked to EMR
Not linked to provider
Not linked to clinic visit
Questions


Can it be done in the clinic?
What about older veterans?
Long Term Goal

To identify and treat hazardous
drinking among HIV infected veterans
through the use of BIs and to identify
and refer alcohol use disorders among
veterans using brief interventions
Specific Aims



Test and adapt an alcohol screening and
interactive computer prototype using
iterative process of user testing, focus
groups and face-to-face interviews with
provider & patients
Test feasibility of implementing prototype in
two VA HIV clinics
Gather information on effect size of
intervention to reduce consumption, and
HIV relevant consequences (sexual risk
behavior & antiretroviral medication
adherence)
Computer Assisted
Lifestyle Management
(CALM)

Identifies hazardous drinkers
– Alcohol Use Disorders Identification Test
(AUDIT)
– Quantity and frequency of consumption
– Alcohol related consequences

Readiness to change
CALM

Delivers Brief Intervention
– Patients & providers explore ETOH
severity, consequences, goals & Rx
barriers
– Brief negotiation using FRAMES & Stages
of Change
– Computer intervention pulls from
electronic medical record
CALM
FRAMES
Feedback
Responsibility
Advice
Menu of options
Empathy
Self-efficacy
FRAMES
Feedback
– Specific and relative to mental, physical & psychosocial health
Responsibility
– Stated explicitly by CALM
Advice
– Simple and explicit; given as a prescription
Menu of options
– Patient chooses goal that matches needs & situation
– Increases perceived personal choice and control
Empathy
– Acknowledge difficulty of change
– By health care provider
Self efficacy
– Statements of hope and optimism
– By health care provider
Pilot Study
Specific Aims
 To assess ease of use and acceptability
of CALM among veterans in primary
care clinic
 To assess patient knowledge and
attitudes regarding computers
 To assess provider attitudes regarding
use of CALM in clinic
Methods






Veterans approached in PC waiting area
Completed self administered computer
program
Touch screen tablet computer
Patient print out – summary & change plan
Provider print out - patient responses &
change plan
Providers surveyed after patient visit
Methods

Measures
– assessment of ease of use and
acceptability of CALM
– knowledge and attitudes regarding
computer
Methods
Subjects
 67 of 80 VA patients surveyed after using
CALM
–
–
–
–
–

92% male
25% non-white
mean age 62 years
81% graduated high school
11% hazardous drinking (AUDIT > 8 or  16
drinks/week)
9/15 (60%) VA primary care providers
returned surveys (Physicians and Nurse
Practitioners)
Results




60% of patients reported having used a
computer
97% felt “at ease” with the computer
77% would be as honest or more honest
71% more private way to collect information
Results





76%
75%
71%
87%
after
64%
CALM easy to use
interesting
liked it or liked it very much
would heed provider’s advice
CALM
more likely to ask questions
Results
Providers
– 78% CALM provides reliable information
& influence interactions with patients
– 66% patients more honest with computer
– 78% would use program
– 55% program would make them more
effective
Conclusions

Delivering a computerized BI in
primary care
– acceptable to providers and patients
– viewed as facilitating dialogue about
drinking
– may enhance patient receptiveness to
provider advice
Methodological Issues
1.
2.
3.
4.
What is “hazardous drinking” in the
HIV Population?
Is the clinic an appropriate venue to
administer CALM? Role of Internet?
What is the best way to deliver info
to providers?
Timing of intervention with respect to
provider visit?
Future Directions


Refinement
Customization in Subspecialty Clinics
– HIV Clinics

Linkage to CPRS
– Wireless “print out” to provider

Timing of CALM delivery
– Before visit at home? waiting room? after
clinic?
Tailoring Computerized
BIs

BIs
– Need to be tailored to individual patients
– Need to be tailored to individual
conditions
– Varying age, health problem


Link to clinical care, provider
Tailor any BI
CALM SPECS



Java/J2EE application that runs on a Tomcat
5.5.7 application server and Appache 2.2
web server
Database is MySql 5.0
Follows MVC (Model-View-Controller) object
oriented design pattern
– Java servlets used to implement Controller
– JSPs used for presentation of data (the View)
– Java classes are the Model
CALM
Can be used as:
1. authoring of BIs
2. presentation of BIs
3. Reporting/statistical tool (all data can be
exported into Excel, CSV or HTML format)
Provides application level security, where
Administrators (aka super-users) can
manage access privileges of other users
Future enhancements:



Question Library.
Allow multimedia to be inserted into
Brief Intervention text.
HTML toolbox to allow nicer formatting
of text questions.
CALM





NIAAA
VACS
University of Kentucky
Baltimore VA
Pittsburgh VA
Manage Brief
Interventions
Multiple Types of BIs
Reports