Transcript Related Journal Club Presentation

Journal Club
Alcohol, Other Drugs, and Health: Current Evidence
January–February 2009
www.aodhealth.org
1
Featured Article
Alcohol screening scores and
medication nonadherence
Bryson CL, et al. Ann Intern Med. 2008;149(11):795–804.
www.aodhealth.org
2
Study Objective
• To identify whether alcohol misuse is
associated with increased risk for medication
(statin, oral hypoglycemic, or
antihypertensives) nonadherence.
www.aodhealth.org
3
Study Design
• Secondary analysis of cohort data collected
prospectively as part of a randomized controlled
trial.
• Participants included:
– 5473 patients taking a statin.
– 3468 patients taking oral hypoglycemic agents.
– 13,729 patients taking antihypertensive medications.
• All participants had completed the Alcohol Use
Disorders Identification Test—Consumption
(AUDIT-C).
www.aodhealth.org
4
Study Design (cont’d)
• Based on AUDIT-C scores, patients were categorized as
–
–
–
–
–
Nondrinkers
Low-level alcohol use
Mild alcohol use
Moderate alcohol use
Severe alcohol use
(AUDIT-C
(AUDIT-C
(AUDIT-C
(AUDIT-C
(AUDIT-C
=
=
=
=
=
0)
1 to
4 to
6 to
8 to
3)
5)
7)
12)
• Adherence* was calculated over 2 observation periods:
– 90 days from the date surveys were received.
– 1 year from the date surveys were received.
*defined as having medications available for at least 80% of the observation period based on
pharmacy refill records.
www.aodhealth.org
5
Assessing Validity of an
Article About Harm
• Are the results valid?
• What are the results?
• How can I apply the results to
patient care?
www.aodhealth.org
6
Are the Results Valid?
• Did the investigators demonstrate similarity in all
known determinants of outcomes? Did they
adjust for differences in the analysis?
• Were exposed patients equally likely to be
identified in the two groups?
• Were the outcomes measured in the same way in
the groups being compared?
• Was follow-up sufficiently complete?
www.aodhealth.org
7
Did the investigators demonstrate similarity in
all known determinants of outcomes? Did they
adjust for differences in the analysis?
• Analyses were adjusted for the following potential
confounders:
–
–
–
–
–
–
–
race
gender
education
marital status
number of medications prescribed (regimen complexity)
smoking status
depression
www.aodhealth.org
8
Were exposed patients equally likely
to be identified in the groups?
 Validated screening tests for alcohol use and
depression were applied in a standard
fashion to all subjects in the study.
www.aodhealth.org
9
Were the outcomes measured in the
same way in the groups being compared?

Yes.
–
Medication adherence was measured in a
standard fashion using pharmacy refill data.
Subjects were considered to be adherent if
they had medication from the pharmacy for at
least 80% of the observation period
www.aodhealth.org
10
Was follow-up sufficiently complete?
• Yes.
– Follow-up data were obtained on all participants
at 90 days and 1 year.
www.aodhealth.org
11
What are the Results?
• How strong is the association between
exposure and outcomes?
• How precise is the estimate of the risk?
www.aodhealth.org
12
What are the Results?
• Unhealthy alcohol use was significantly
associated with lower adherence to both statin
and antihypertensive medications at 1 year.
Adherence (adjusted for confounders) was:
–
–
–
–
66%
63%
58%
55%
for
for
for
for
nondrinkers.
those with mild unhealthy use.
those with moderate unhealthy use.
those with severe unhealthy use.
• Despite a similar trend, alcohol categories were
not significantly associated with decreased
adherence to hypoglycemic medication.
www.aodhealth.org
13
How strong is the association between
exposure and outcome?
Association between AUDIT-C Scores and Proportion of Patients
Adherent to Medication at 1 Year
AUDIT-C
score
Fully Adjusted Proportion
(95% CI)
Statin cohort (n=4989)*
0
1–3
4–5
6–7
8–12
66 (64–68)
63 (60–65)
63 (59–67)
58 (52–65)†
55 (47–63)†
Oral hypoglycemic cohort
(n=3114)*
0
1–3
4–5
6–7
8–12
63
60
63
61
58
Hypertension treatment
cohort (n=12,545)*
0
1–3
4–5
6–7
8–12
64 (63–65)
62 (61–64)
61 (58–64 )†
60 (56–63 )†
56 (52–60 )†
Variable
(61–65)
(57–63)
(57–70)
(51–70)
(48–67)
*Number of patients in regression models varies because of missing covariates.
†p<0.05 compared with nondrinkers.
www.aodhealth.org
14
How precise is the estimate of the risk?
• The confidence intervals are wide, especially
for the more severe drinking categories, but
lack overlap between nondrinkers and those
with the most severe drinking categories for
adherence to statins and antihypertensive
medications.
www.aodhealth.org
15
How Can I Apply the Results to
Patient Care?
• Were the study patients similar to the patients in
my practice?
• Was the duration of follow-up adequate?
• What was the magnitude of the risk?
• Should I attempt to stop the exposure?
www.aodhealth.org
16
Were the study patients similar to the
patients in my practice?
•
•
•
•
All patients were veterans.
>95% were men.
Mean age was 64.
>75% were white.
www.aodhealth.org
17
Was the duration of follow-up adequate?
• Yes.
– One would expect to see an impact on adherence
within one year.
www.aodhealth.org
18
What was the magnitude of the risk?
• There was an approximately 10% drop in
(statin and antihypertensive) medication
adherence between nondrinkers and those with
severe alcohol use.
• There was a linear dose-response impact of
increasing alcohol problems resulting in
decreased medication adherence.
• The caveat is no data were provided on clinical
outcomes (e.g., lipid levels, glycemic control,
blood pressure).
www.aodhealth.org
19
Should I attempt to stop the exposure?
• Yes.
– These data provide evidence for the potential
adverse impact of alcohol consumption on
medication adherence. Further studies on this
association and clinical markers would strengthen
the argument to intervene to reduce harm,
especially in cases of severe alcohol misuse.
www.aodhealth.org
20