Dechartres_Comparison_of_Estimates_between

This excel file includes
all relevant trial-level variables to replicate the meta-epidemiological
analysis for the article " Comparison of Estimates between Cohort and Case-control studies in
Meta-analyses of Therapeutic Interventions: A Meta-epidemiological Study"

Variable coding

nMA = meta-analysis ID

studyID = number of study ID

author = first author last name

date = year of publication

design =

[0] prospective nonrandomized study (all subjects are recruited and evaluated
prospectively, but the control arm is not created through randomization)

[1] retrospective cohort study (subjects are evaluated retrospectively and
study arms are concurrent [without matching])

[2] historical control study (uses retrospective,
nonconcurrent controls)

[3] case-control study (compared groups are defined on the basis of the
outcome and/or matching is used)

[4] other or not specified design

funding = [0] public, [1] private, [2]
both, [3] unreported, [4] unclear

center = [0] single center, [1]  multicenter, [2] unreported

propens = propensity score analysis
conducted [0] with matching, [1] with adjustment, [2] with weighting, [3]
unreported

adjust = confounding variables were
adjusted for in analysis [0] yes, [1] no, [2] unreported

ana_exp = number of participants analyzed
in experimental group 

ana_ctl = number of participants analyzed
in control group 

ana_both = number of participants analyzed
in both groups

out_exp = number of participants with
outcome in experimental group

out_ctl = number of participants with
outcome in control group

out_both = number of participants with
outcome in both groups

result = were the results reported as
[OR], [RR], [HR], or unreported [.]

unad_CI = unadjusted 95% confidence interval
for results

adj_CI = adjusted 95% confidence interval
for results

unad_val = unadjusted value for results

adj_val = adjusted value for results

logestcrude=log estimate for the crude analysis

selogestcrude=standard error of the log estimate
for the crude analysis

logestadj= log estimate for the adjusted
analysis

selogestadj=standard error of the log estimate
for the adjusted analysis

logest=log estimate (adjusted if adjusted
analyses, crude if no adjusted analysis)

selogest= corresponding standard error

casecontrol= 1=case control study; 0=cohort
study

pharmaco= 1=pharmacological intervention, 0=non-pharmacological
intervention