Journal Club

1
Journal Club

• Alcohol and Health Current Evidence
• November-December 2005

2
Featured Article

• Combining the AUDIT questionnaire
• and biochemical markers to assess
• alcohol use and risk of alcohol withdrawal
• in medical inpatients
• Dolman JM, et al. Alcohol Alcohol.
  200540(6)515519.

3
Study Objective

• To examine whether the AUDIT and/or
• blood testing could
• predict risk of alcohol withdrawal in medical
  inpatients
• Alcohol Use Disorders Identification Test

4
Study Design

• Screening with the AUDIT and blood testing (GGT,
  AST, ALT, MCV)
• 874 medical inpatients (aged 16 or older)
  screened
• Incident alcohol withdrawal assessed
  prospectively during hospitalization in 98
  patients with a positive AUDIT score (gt8)
• Gamma glutamyltransferase, aspartate
  aminotransferase, alanine aminotransferase, and
  mean corpuscular volume

5
Assessing Validity of an Article about Prognosis

• Are the results valid?
• What are the results?
• How can I apply the results to patient care?

6
Are the Results Valid?

• Was the sample representative?
• Were the subjects sufficiently homogeneous with
  respect to prognostic risk?
• Was follow-up sufficiently complete?
• Were objective and unbiased outcome criteria used?

7
Was the sample representative?

• 267 of 1243 admissions were excluded because of
  incomplete AUDIT questionnaires and lab results.
• This likely biased the sample though it is not
  clear in what direction.
• So, whether the study sample was representative
  of all admissions is unknown.
• Those who were confused or transferred quickly
  were excluded.
• The screened sample included 874 medical
  inpatients, aged 16 or older.

8
Was the sample representative? (cont.)

• The sample that was monitored for withdrawal
  development included only the 98 subjects (11)
  with a positive AUDIT (gt8).
• Therefore, the sample does not represent people
  with an AUDIT lt8 who might develop withdrawal.

9
Were the subjects sufficiently homogeneous with
respect to prognostic risk?

• Subjects were likely sufficiently homogeneous
  with respect to withdrawal risk because...
• all were eligible for the study from the time of
  admission.
• However, time of last drink was not reported. If
  highly variable, it might have led to a
  heterogeneous sample with respect to withdrawal
  risk.

10
Was follow-up sufficiently complete?

• All subjects were followed through the period of
  risk of developing alcohol withdrawal.
• However, the authors did not provide details on
  whether any subjects were lost to follow-up.

11
Were objective and unbiased outcome criteria
used?

• Subjects with an AUDIT score of gt8 (a positive
  test for an alcohol use disorder) were monitored
  with the CIWA-Ar, an objective outcome measure.

12
What are the Results?

• How likely are the outcomes over time?
• How precise are the estimates of likelihood?

13
How likely are the outcomes over time?

• Of the 98 subjects with positive AUDITs, 17 (17)
  experienced clinically significant withdrawal
  symptoms.
• All patients with withdrawal had positive AUDITs
  (gt8 sensitivity 100).
• However, those with AUDITlt8 were not monitored
  for withdrawal using the objective outcome
  measure, raising the possibility of overestimated
  sensitivity.
• All but 1 patient with withdrawal had abnormal
  blood test results.

14
How likely are the outcomes over time? (cont.)

• Most patients without withdrawal had normal AUDIT
  scores (specificity 91).
• Although a positive AUDIT score plus any 2
  abnormal blood tests had a sensitivity of 94 and
  a specificity of 98...
• fewer than half of patients with this combination
  had withdrawal.

15
How precise are the estimates of likelihood?

• The authors did not provide measures of precision.

16
How Can I Apply the Results to Patient Care?

• Were the study patients and their management
  similar to those in my practice?
• Was the follow-up sufficiently long?
• Can I use the results in the management of
  patients in my practice?

17
Were the study patients similar to those in my
practice?

• Study patients may have been representative of
  those on medicine services in general hospitals.

18
Was the follow-up sufficiently long?

• The patients were followed until the CIWA-Ar
  score was lt11 for 12 hours.
• This is likely long enough that no cases of late
  withdrawal were missed.
• However, longer follow-up would have reduced the
  likelihood of missing any cases (e.g., 24 hours
  CIWA-Ar lt8).
• The authors do not report any loss to follow-up.

19
Can I use the results in the management of
patients in my practice?

• Limitations to this study
• Researchers monitored alcohol withdrawal only in
  subjects with AUDIT gt8 so the study cannot draw
  conclusions about those with AUDITlt8.
• A substantial number of patients were excluded
  because of incomplete AUDITs or blood tests,
  likely biasing the sample.
• Follow-up may not have been long enough.
• The number of patients with symptomatic
  withdrawal is too small to draw firm conclusions.

20
Can I use the results in the management of
patients in my practice? (cont.)

• Nonetheless, given that the AUDIT identifies
  alcohol dependence, it is not surprising that it
  can also predict who will have alcohol
  withdrawal.
• But, most patients with dependence will not have
  significant withdrawal symptoms.
• Adding blood tests improves detection of those at
  risk of withdrawal but may predict only 50, at
  best, of withdrawal cases.

21
Can I use the results in the management of
patients in my practice? (cont.)

• Therefore, alcohol screening in the hospital is
  mainly useful for
• ruling out risk of withdrawal and
• identifying patients who might be ready for
  alcohol-dependence treatment.