Skip to main content

Advertisement

Table 1 Definition of the different methodological choices and number of possible analytical scenarios

From: Vibration of effects from diverse inclusion/exclusion criteria and analytical choices: 9216 different ways to perform an indirect comparison meta-analysis

Category Criteria Number of possibilities
Medical condition Inclusion of all studies (AUDs and/or AD)
Exclusion of studies including patients with AUDs
2
Abstinence Inclusion of all studies (abstinent or non-abstinent patients)
Exclusion of studies requiring a minimum period of abstinence of 5 days or more before the beginning of the study
2
Gender Inclusion of all studies (mixed gender, males only or females only)
Exclusion of studies with males or females only
2
Somatic comorbidity Inclusion of all studies (patients with or without systematic somatic comorbidities)
Exclusion of studies on patients with systematic somatic comorbidities (e.g. studies on patients with HIV)
2
Psychiatric comorbidity Inclusion of all studies (patients with or without systematic psychiatric comorbidities)
Exclusion of studies on patients with systematic psychiatric comorbidities (e.g. studies on depressed patients)
2
Psychological support Inclusion of all studies (with or without psychological intervention)
Exclusion of studies with no psychological intervention
2
Treatment and dose Only approved dose and route of administration
Approved dose and route of administration OR closest dose to the approved dose
Maximum dose tested
3
Treatment duration Inclusion of all studies, regardless of treatment duration
Exclusion of studies with a treatment duration of less than 12 weeks
2
Outcome Quantity of alcohol consumed*
Frequency of drinking**
Abstinence***
3
Publication Published and unpublished studies (e.g. study reports, ClinicalTrials.gov)
Exclusion of unpublished studies
2
Risk of bias Inclusion of all studies, regardless of the risk of selective outcome reporting
Exclusion of studies with a high risk of selective outcome reporting
2
Analysis Fixed effect model
Random effect model
2
Total of possible combinations 9216
  1. The choice of the 5-day cut-off was based on our previous meta-analysis [25]
  2. If there were several outcomes for quantity consumed, frequency of drinking, or abstinence reported in the same study, only one criterion of each type was collected
  3. *Outcomes for the quantity of alcohol consumed were extracted in this order of preference: (1) total alcohol consumption, (2) number of drinks per day, (3) number of drinks per drinking day, and (4) alcohol consumption per drinking day
  4. **Frequency of drinking outcomes was extracted in this order of preference: (1) number of heavy drinking days, (2) percentage of heavy drinking days, and (3) percentage of drinking days
  5. ***Abstinence outcomes were extracted in this order of preference: (1) number of abstinent days, (2) percentage of abstinent days, (3) percentage of abstinent subjects, and (4) percentage of relapsing subjects