Transcript May-Jun 2013

Update on
Alcohol, Other Drugs,
and Health
May–June 2013
www.aodhealth.org
1
Studies on
Interventions &
Assessments
www.aodhealth.org
2
Alcohol Screening and Brief
Intervention Not Effective in
Clinical Practice
Kaner E, et al. BMJ 2013;346:e8501.
Summary by Richard Saitz, MD, MPH
www.aodhealth.org
3
Objectives/Methods


Randomized controlled-efficacy trials consistently
find modest favorable effects of brief intervention
(BI) for nondependent unhealthy alcohol use
identified by screening in primary care settings.
Investigators randomly assigned 24 primary care
practices to implement screening for unhealthy
alcohol use and then either a patient information
brochure, 5 minutes of brief advice, or 20
minutes of health behavior change counseling.
www.aodhealth.org
4
Results


Practices were paid approximately $4500 plus
additional compensation for each screening and
each intervention.
Five practices did not recruit the requisite 31
patients and were replaced; 5 practices were
reassigned to one of the more intensive
interventions after completing initial recruitment
targets. Due to slow recruitment, research staff
accomplished the screening and BI in 10 of the
practices.
www.aodhealth.org
5
Results (cont’d)


Of 756 patients enrolled, 81% had 6-month
outcome data; only 57% of those assigned to
counseling received it.
The odds of having an Alcohol Use Disorders
Identification Test score of <8 (i.e., no unhealthy
use) were lower, though not significantly
different for advice (odds ratio [OR], 0.85) and
counseling (OR, 0.78), versus information only.
Intention-to-treat and per-protocol analyses
yielded similar results.
www.aodhealth.org
6
Comments


Screening and BI were both challenging to
implement and ineffective in primary care
practices.
These findings should be carefully considered
by those involved in major implementation
efforts worldwide and serve as a reality check
for clinicians working to implement alcohol
screening and BI in their practices.
www.aodhealth.org
7
Systematic Screening Is More
Effective than Targeted
Screening at Engaging Primary
Care Patients in Discussions
about Alcohol, Especially Those
with Lower Risk Use
Reinholdz H, et al. Alcohol Alcohol. 2013;48(2):172–179.
Summary by Nicolas Bertholet, MD, MSc
www.aodhealth.org
8
Objectives/Methods


Implementing systematic screening for risky
drinking can be a burden for primary care
practices (PCPs). An alternative is the use of a
targeted approach.
Both methods were compared in a randomized
trial of 16 PCPs in Sweden (N=3609 patients) to
assess their impact on alcohol discussions.
www.aodhealth.org
9
Objectives/Methods (cont’d)


In the systematic group, patients were asked to
complete the AUDIT-C* and give it to their primary
care physician before the consultation.
In the consultation-based early identification
group, clinicians were encouraged to be alert for
signs of alcohol-related issues during the
consultation but did not receive AUDIT-C screening
results; patients completed the AUDIT-C after the
consultation.
*Alcohol Use Disorders Identification Test—Consumption.
www.aodhealth.org
10
Results


Among patients who screened positive for risky drinking*
the mean age was 44 years for men and 42 years for
women. Among patients with whom the issue of alcohol
was brought up, the mean age was 54 years for men
and 44 years for women.
In the systematic group, 62.5% of patients with risky
drinking and 64.2% of patients with lower-risk drinking
had the issue of alcohol brought up. In the identification
group, the proportions were 29.7% and 30.1%,
respectively.
*Defined as AUDIT-C scores of ≥5 for men and ≥4 for women.
www.aodhealth.org
11
Results (cont’d)


Looking at patients with risky drinking only, a
higher proportion in the systematic group had
the issue brought up compared with the
identification group. This association was
significant for men, women, and both genders
combined.
Among patients with risky drinking who had the
issue of alcohol brought up, the mean AUDIT-C
score was significantly higher in the identification
group (5.8) compared with the systematic group
(5.2).
www.aodhealth.org
12
Comments


The method used by PCPs to identify at-risk
drinking dramatically impacts the proportion of
patients who discuss alcohol with their primary
care physicians.
Although systematic screening did not lead to
systematic conversations about alcohol in this
study, it did allow for a much higher proportion
of patients with risky alcohol use to engage in
such conversations.
www.aodhealth.org
13
Comments (cont’d)

A targeted approach allows patients with more
severe risky use to be identified, but systematic
screening helps identify patients with lower-risk
alcohol use (potentially even more than it helps
identify patients with more severe problems)
who may benefit from a single brief intervention.
www.aodhealth.org
14
Studies on
Health Outcomes
www.aodhealth.org
15
Update on the Association
between Alcohol Consumption,
Cognitive Functioning, and
Dementia
Panza F, et al. Int J Geriatr Psychiatry. 2012;27(12):1218–1238.
Summary by R. Curtis Ellison, MD
www.aodhealth.org
16
Objectives/Methods


Scientists in Italy conducted an extensive review
of the research published between 1987–2011
that examined the relationship between alcohol
consumption, cognitive function, and dementia.
For cognitive impairment, the authors reviewed
14 cross-sectional studies; for cognitive decline,
they reviewed 21 longitudinal studies; and for
dementia or predementia, they reviewed 26
longitudinal and 7 other studies.
www.aodhealth.org
17
Results


Current data support an association between lowrisk drinking* and lower risk of dementia.
Protective effects of low-risk alcohol consumption
against cognitive decline are suggested to be more
likely in the absence of the apolipoprotein E (APOE)
e4 allele associated with Alzheimer Disease and
where wine is the beverage.
*There was significant variability across studies in defining low-risk drinking in both men and women.
www.aodhealth.org
18
Results (cont’d)

Suggested mechanisms by which low-risk alcohol
intake may be neuroprotective include reduction
in vascular risk factors; fewer brain infarcts and
a U-shaped relationship with white matter
lesions; stimulation of the release of
hippocampal acetylcholine (in animal models);
and antioxidant effects of polyphenols, especially
from wine.
www.aodhealth.org
19
Comments


There are always problems in having to develop
guidelines for alcohol consumption based only on
observational data, with no large clinical trials to
judge effect.
While results in prospective studies are quite
consistent, many studies included in this review
lacked data on beverage type and pattern of
drinking, which may be important determinants of
health outcomes.
www.aodhealth.org
20
Comments (cont’d)

Despite the recognized limitations, current
observational data support that consuming lowrisk amounts of alcohol, especially of wine, is
associated with less cognitive decline and
dementia.
www.aodhealth.org
21
Distributing Naloxone to People
Who Use Heroin for Lay
Treatment of Overdose is CostEffective
Coffin PO, et al. Ann Intern Med. 2013;158(1):1–9.
Summary by Kevin L. Kraemer, MD, MSc
www.aodhealth.org
22
Objectives/Methods



Researchers used a decision analytic computer model to
estimate the cost-effectiveness (CE) of distributing naloxone
to people with heroin use to reverse overdoses.
The model compared a strategy of distributing naloxone to
20% of people who use heroin with a strategy of no
distribution and was designed to bias against the naloxone
strategy.
Values and ranges for the model parameters (proportions,
transition rates, costs, utilities) were extracted from
published literature. The model outputs were costs, qualityadjusted life-years (QALYs), and incremental CE ratios
(costs per QALY).
www.aodhealth.org
23
Results

In the base case analysis, the naloxone distribution
strategy:
 prevented 6% of overdose deaths.
 required the distribution of 227 naloxone kits to
prevent 1 overdose death.
 had an incremental CE ratio of $438 per QALY
gained.
www.aodhealth.org
24
Results (cont’d)

In sensitivity analyses, the naloxone distribution
strategy remained cost-effective (most
incremental CE ratios less than $1000 per QALY
gained) over a wide range of scenarios and
parameter values. Even the worst-case scenario
(overdoses rarely witnessed and naloxone more
expensive, less efficacious, and rarely carried) had
an incremental CE ratio of $14,000 per QALY.
www.aodhealth.org
25
Comments



This well-done analysis suggests that naloxone
distribution is highly cost-effective when compared
with no distribution.
The incremental CE ratios are far lower than those
for many common healthcare practices and well
within thresholds considered cost-effective by
policymakers.
Although controlled data were not available for
some parameters of the model, it appears robust
and lends support to the distribution of naloxone
to people who abuse heroin and prescription
opioids.
www.aodhealth.org
26
Overdose Education and Nasal
Naloxone Distribution are
Associated with a Reduction in
Overdose Fatalities
Walley AY, et al. BMJ 2013;346:f174.
Summary by Darius A. Rastegar, MD
www.aodhealth.org
27
Objectives/Methods


To determine whether providing opioid overdose
education with naloxone distribution (OEND) is an
effective strategy for reducing overdose fatality,
the researchers evaluated the impact of an OEND
program in Massachusetts.
The program provided education to people at risk
for overdose and bystanders on minimizing the risk
of overdose and recognizing and responding to
overdose by providing rescue breathing,
administering nasal naloxone, and staying with the
person until medical personnel arrived.
www.aodhealth.org
28
Objectives/Methods (cont’d)


Enrollees received a rescue kit with two doses of
naloxone.
The program included 19 communities that
accounted for 30% of the state population and
used interrupted times series analysis to compare
community-year strata that have high and low
rates of OEND implementation with those with no
implementation.
www.aodhealth.org
29
Results



Over a 3-year period in the 19 communities studied,
2912 individuals were enrolled in OEND and 327 rescue
attempts were reported.
Compared with no implementation, both low and high
implementation of OEND were associated with lower
rates of opioid overdose deaths, with an adjusted rate
ratio of 0.73 in low-implementer community-year strata
and 0.54 in the high-implementer community-year
strata.
There was no significant difference in opioid overdoserelated acute care hospital utilization.
www.aodhealth.org
30
Comments


Opioid overdose education and naloxone
distribution are two tools to help reduce
unintentional opioid overdose deaths. Other
measures that have been shown to reduce
overdose fatalities are opioid agonist treatment
and supervised injection facilities.
While acute care hospital utilization was not
reduced, the authors point out that this could be
due to the fact that the training encouraged
bystanders to engage the emergency medical
system.
www.aodhealth.org
31
Comments (cont’d)

This program targeted people who use illicit
drugs; another population at risk that may
benefit from OEND is individuals who are
prescribed opioids for chronic pain.
www.aodhealth.org
32
Controlled Trials Support the
Use of Benzodiazepines for
Prevention and Treatment of
Alcohol Withdrawal Syndrome
in the Intensive Care Unit
Unger LA, et al. Alcohol Clin Exp Res. 2013;37(4):675–686.
Summary by Kevin L. Kraemer, MD, MSc
www.aodhealth.org
33
Objectives/Methods



Two independent working groups conducted a
systematic review of controlled trials published
1971–2011 on prevention and treatment of alcohol
withdrawal syndrome (AWS) in the intensive care
unit (ICU).
Six prevention trials (4 randomized, 2
nonrandomized) and 8 treatment trials (4
randomized, 4 nonrandomized) were identified.
Reasons for ICU admission were a mix of surgery,
trauma, medical illness, and severe AWS. Definition
and scoring of AWS and study outcome were
highly heterogeneous across trials.
www.aodhealth.org
34
Objectives/Methods (cont’d)


Drugs assessed for AWS prevention were
ethanol infusion, clonidine, flunitrazepam,
clomethiazol + haloperidol, flunitrazepam +
haloperidol, and midazolam.
Drugs assessed for AWS treatment included
those listed above as well as gammahydroxybutyric acid, phenobarbital, lorazepam,
and flunitrazepam + clonidine.
www.aodhealth.org
35
Results


All AWS prevention and treatment regimens were
effective compared with control.
Benzodiazepines were effective and safe for AWS
prevention and treatment, whereas ethanol infusion
was effective and safe for AWS prevention.
www.aodhealth.org
36
Comments


The trials in this systematic review tended to be
small and not high-quality, and heterogeneity
prevented pooling of results. Nonetheless,
benzodiazepines can be regarded as the strategy
with the most evidence-based support for AWS
prevention and treatment in the ICU.
Although ethanol infusion was effective for AWS
prevention, it not should supplant
benzodiazepines because the level of evidence is
weaker, it is difficult to titrate, and has no
apparent benefit over benzodiazepines.
www.aodhealth.org
37
Office-based Buprenorphine
Treatment is Feasible in
Federally Qualified Health
Centers
Haddad MS, et al. Drug Alcohol Depend. January 6, 2013 [Epub
ahead of print]. doi: 10.1016/j.drugalcdep.2012.12.008.
Summary by Alexander Y. Walley, MD, MSc
www.aodhealth.org
38
Objectives/Methods


Buprenorphine/naloxone treatment for opioid
dependence has been demonstrated to be feasible
in multiple primary care settings, including
academic medical centers, private practices, and
HIV clinics.
Researchers conducted a retrospective review of
electronic medical records in 2 federally qualified
health centers (FQHCs) in Connecticut to identify
clinical factors associated with retention in
buprenorphine (BUP) treatment and opioid
abstinence.
www.aodhealth.org
39
Objectives/Methods (cont’d)

Records were reviewed for 266 patients who
received at least 1 BUP prescription from 1 of 4
prescribers between July 1, 2007, and December
1, 2008, and who were observed for 6 to 21.5
months.
www.aodhealth.org
40
Results


The mean age of patients was 40 years; 70% were
male, 70% were prescribed BUP by family practitioners,
30% were treated by a psychiatrist, 72% were
diagnosed with a mood or anxiety disorder, 65% were
treated for a mood or anxiety disorder, and 80%
initiated care at the FQHC to receive BUP. The mean
stabilization dose was 17.8 mg.
Retention in treatment at 6 months was 57%. Patients
with cocaine use at treatment initiation (adjusted
hazard ratio [AHR], 2.18) were more likely to drop out
of BUP treatment, whereas patients who were older
(AHR, 0.96), female (AHR, 0.59), receiving psychiatric
medication (AHR, 0.69), HCV-infected (AHR, 0.56), or
receiving on-site substance abuse counseling (AHR,
0.54) were less likely to drop out.
www.aodhealth.org
41
Results (cont’d)

Urine toxicology results were opioid-abstinent for
30% of the patients at every test and 72% of
patients at their last observed test. Patients with
cocaine use at treatment initiation were less likely
to have an opioid-abstinent last urine test
(adjusted odds ratio [AOR], 0.43), whereas
patients prescribed psychiatric medications were
more likely to have an opioid-abstinent last urine
test (AOR, 1.66).
www.aodhealth.org
42
Comments


Whereas comorbid psychiatric illness typically is
associated with worse addiction treatment
outcomes, this study found that patients with
psychiatric comorbidity treated with medication
had improved retention and were more likely to
be abstinent from opioids.
The prospect of integrated care models that
deliver high-quality medical, psychiatric, and
addiction treatment that produces improved
outcomes warrants prospective implementation
studies.
www.aodhealth.org
43
Studies on
HIV and HCV
www.aodhealth.org
44
Drug-related Causes the Primary
Reasons for Death among People
with Injection Drug Use who are
Seropositive for HCV
Kielland KB, et al. J Hepatol. 2013;58(1):31–37.
Summary by Judith Tsui, MD, MPH
www.aodhealth.org
45
Objectives/Methods

To determine whether hepatitis C virus (HCV) is a
major cause of mortality among people who use
injection drugs, this observational cohort study
included 523 HCV seropositive patients admitted to
residential substance use treatment in Norway
followed using the national mortality register.
www.aodhealth.org
46
Results



Of the 523 HCV seropositive people with injection
drug use, 63% were HCV RNA+ (i.e., had chronic
infection).
At the end of follow-up (mean 33 years), 42%
died.
The mortality rate for HCV RNA+ people was 1.75
per 100 person-years, versus 2.05 in HCV- people
(i.e., presumed to have spontaneous
clearance)(P=0.25).
www.aodhealth.org
47
Results (cont’d)



Opioid overdose was the most frequent cause of
death, followed by violent death and suicide.
Twelve patients (7.5%) died of liver disease; 10
were HCV RNA+ and the remaining 2 had chronic
hepatitis B.
Restricting analyses to those ≥50 years of age,
the liver-related mortality rate was 0.94 per 100
person-years among HCV RNA+ people, and was
0 (no deaths) among those who were HCV RNA-.
www.aodhealth.org
48
Comments


This study confirms high mortality rates for
people who use injection drugs and reinforces
that drug-related causes remain the primary risks
for death in this population. Thus, some patients
with HCV may not benefit from HCV treatment
because of competing causes of death.
However, among people with injection drug use
who survive beyond age 50, liver-related
mortality emerges as a cause of premature
death, confirming prior modeling studies.
www.aodhealth.org
49
Comments (cont’d)

The study’s main finding should be cautiously
interpreted given that there was some potential
for misclassification bias (cases of unidentified
reinfection or clearance through treatment).
www.aodhealth.org
50
Risky Alcohol Use Associated
with Elevated Serum
Aminotransferase Levels in
HIV/HCV Coinfected Patients
Tsui JI, et al. J Stud Alcohol and Drugs. 2013;74(2):266–270.
Summary by Jeanette M. Tetrault, MD
www.aodhealth.org
51
Objectives/Methods


Using data from the prospective observational cohort
HIV-Longitudinal Interrelationships of Viruses and
Ethanol (HIV-LIVE) study, the authors assessed the
association between National Institute on Alcohol
Abuse and Alcoholism (NIAAA)-defined risky alcohol
consumption* and serum aminotransferase levels in
HIV-infected patients with and without HCV infection.
Regression models that can account for repeated
measures were used to assess the association between
risky alcohol use and AST and ALT levels over time.
* >14 standard drinks per week or >4 drinks per day for men, or >7 drinks per week or >3
drinks per day for women.
www.aodhealth.org
52
Results


Of 397 HIV-infected subjects in the HIV-LIVE study
with HCV serologic testing, 200 (50.4%) had
detectable HCV viral load and were considered
HIV/HCV coinfected.
Among coinfected subjects, risky alcohol use was
associated with higher mean AST (adjusted means
62.2 versus 51.4 U/L; adjusted ratio of means 1.2)
and ALT (adjusted means 51.3 versus 41.6 U/L.;
adjusted ratio of means 1.2), compared with those
without risky alcohol use.
www.aodhealth.org
53
Results (cont’d)

Among HIV monoinfected subjects, the authors
found no differences in AST or ALT among
patients with or without risky alcohol use.
www.aodhealth.org
54
Comments


Among HIV/HCV coinfected subjects, mean AST
and ALT levels were approximately 20% higher in
those with risky alcohol use compared with those
without risky alcohol use, while no such
association was noted in HIV monoinfected
subjects.
These data suggest that alcohol use may be
particularly concerning in HIV/HCV coinfected
patients and that physicians should consider
tailoring counseling messages regarding the
harmful effects of alcohol to this population.
www.aodhealth.org
55
Adding Enhanced Counseling to
Standard Medical Management
Does Not Improve Outcomes for
HIV-Positive Individuals with
Opioid Dependence Who are
Prescribed Buprenorphine
Tetrault JM, et al. J Subst Abuse Treat. 2012;43(4):433–439.
Summary by Darius A. Rastegar, MD
www.aodhealth.org
56
Objectives/Methods


The provision of buprenorphine with medical
management in an office-based setting has been
shown to be an effective treatment for opioid
dependence. The addition of further counseling,
however, has not been shown to improve
outcomes.
For HIV-positive individuals, providing
buprenorphine treatment may improve their care,
but the impact of further counseling has not been
studied in this population.
www.aodhealth.org
57
Objectives/Methods (cont’d)

The 47 HIV-positive opioid-dependent subjects in
this 12-week study were all prescribed
buprenorphine and received standard bi-weekly
physician visits; 22 were randomly assigned to
receive additional enhanced counseling (drug
counseling and medication adherence
management), delivered by a trained nurse.
www.aodhealth.org
58
Results


There was no significant difference in the
percentage of opioid-negative urine drug tests
between the standard and enhanced treatment
arms (64% versus 69%, respectively).
There was no significant difference in the mean
number of continuous weeks of opioid abstinence
(4.9 weeks for standard treatment versus 5.2
weeks for enhanced).
www.aodhealth.org
59
Results (cont’d)

There was likewise no significant difference in the
following secondary endpoints:





percent cocaine-negative urine-drug tests
study completion rate
buprenorphine adherence
antiretroviral adherence
HIV clinical outcomes
www.aodhealth.org
60
Comments


This study supports the findings of previous studies
and extends them to individuals with opioid
dependence who are HIV-positive. The results of
these studies have been consistent and indicate that
for patients who are prescribed buprenorphine,
routinely adding counseling beyond that which is
delivered by physicians in 15-minute visits does not
offer additional benefit.
This study was small and somewhat underpowered
and could not exclude a modest benefit; moreover,
there may always be select individuals who benefit
from additional counseling.
www.aodhealth.org
61
Concurrent Addiction and HCV
Treatment Improves HCV
Treatment Completion
Dimova RB, et al. Clin Infect Dis. 2013;56(6):806–816.
Summary by Jeanette M. Tetrault, MD
www.aodhealth.org
62
Objectives/Methods


The purpose of this meta-analysis was to
determine the impact of support services on HCV
treatment completion and success rates among
HCV-infected people with drug use, measured by
sustained virologic response (SVR), which is
defined as undetectable viral load at the end of
treatment and 24 weeks subsequent.
The meta-analysis included 36 studies—published
between 2004 and 2011—that collectively
reported data from 2866 patients.
www.aodhealth.org
63
Results



The HCV treatment completion rate was 83.4% (95%
confidence interval [CI]: 77.1–88.9%).
Among 25 studies including patients undergoing
concurrent addiction treatment, the higher the
proportion of addiction-treated patients, the greater
the likelihood of treatment completion.
After removing outliers, the pooled SVR rate was
55.5% (95% CI: 50.6–60.3%). Genotype 1 and 4
infection and HIV coinfection were associated with a
decreased SVR rate; after adjusting for these factors,
SVR was associated with treatment performed by a
multidisciplinary team.
www.aodhealth.org
64
Comments

Despite some heterogeneity of studies, the
results of this meta-analysis suggest that overall
treatment completion SVR rates among HCVinfected drug users undergoing treatment with
pegylated interferon and ribavirin are
comparable to previous trials that excluded
people with drug use, particularly when patients
are concurrently treated for addiction.
www.aodhealth.org
65
Self-reported Unhealthy
Alcohol Use Predicts Virological
Rebound in HIV/HCV
Coinfected Patients
Marcellin F, et al. Addiction. February 20, 2013 [Epub ahead of
print]. doi: 10.1111/add.12149
Summary by David Fuster, MD, PhD & Alexander Y. Walley, MD, MSc
www.aodhealth.org
66
Objectives/Methods


Unhealthy alcohol use is common in HIV-infected
patients and is believed to have a negative impact
on the course of HIV infection, in part because it
is associated with poor adherence to antiretroviral
therapy (ART).
Unhealthy use is even more harmful in people
coinfected with HCV because of the increased risk
of progression to end-stage liver disease.
www.aodhealth.org
67
Objectives/Methods (cont’d)



Investigators in France performed a prospective
study to evaluate the impact of heavy alcohol
use,* measured either by physician’s perceptions
or patient self-administered interview, on time to
detectable viral load.
The sample included 512 patients coinfected with
HIV/HCV who had achieved viral suppression**
through ART.
Median follow-up was 21 months.
*Defined as consuming ≥4 drinks (11-14 g alcohol) per day for men or ≥3 drinks per day for
women “on a regular basis” in the past 6 months.
**HIV RNA below the limit of detection.
68
Results


At study enrollment, 9% of participants selfreported heavy alcohol use, and 15% were
considered to have heavy alcohol use by their
physicians.
Physician report agreed with self-report in 90% of
patients not reporting heavy alcohol use and in
77% of those who did, but 61% of those classified
by physicians as having heavy alcohol use did not
report heavy alcohol use, and 2% of those
classified as abstainers or low-risk drinkers
reported heavy alcohol use.
www.aodhealth.org
69
Results (cont’d)

During follow-up, 101 participants (20%) had a
detectable viral load. Self-reported heavy alcohol
use was independently associated with detectable
viral load (adjusted hazard ratio 2), while
physician-reported heavy alcohol use was not.
www.aodhealth.org
70
Comments


This study found that heavy alcohol use is
associated with detectable HIV viral load,
probably through its effect on ART adherence.
Furthermore, the results stress the importance
of patient self-report (in contrast to physician’s
impressions).
These results add to mounting evidence that
heavy alcohol use should be avoided in HIV/HCV
co-infected patients.
www.aodhealth.org
71