meta (version 6.2-1)

meta-object: Description of R object of class "meta"

Description

Detailed description of R objects of class "meta".

Author

Guido Schwarzer

Details

The following R functions create an object of class "meta":

• metabin, metacont, metacor, metagen, metainc, metamean, metaprop, metarate, metacr, metamerge, trimfill

The following generic functions are available for an object of class "meta":

• as.data.frame.meta, labels.meta, print.meta, print.summary.meta, summary.meta, update.meta, weights.meta

An object of class "meta" is a list containing the following components.

studlab	Study labels
sm	Effect measure
null.effect	Effect under the null hypothesis
TE	Effect estimates (individual studies)
seTE	Standard error of effect estimates (individual studies)
statistic	Statistics for test of effect (individual studies)
pval	P-values for test of effect (individual studies)
df	Degrees of freedom (individual studies)
level	Level of confidence intervals for individual studies
lower	Lower confidence limits (individual studies)
upper	Upper confidence limits (individual studies)
three.level	Indicator variable for three-level meta-analysis model
cluster	Cluster variable (three-level meta-analysis model)
k	Number of estimates combined in meta-analysis
k.study	Number of studies combined in meta-analysis
k.all	Number of all studies
k.TE	Number of studies with estimable effects
overall	Print meta-analysis results
overall.hetstat	Print overall heterogeneity statistics
common	Print results for common effect meta-analysis
random	Print results for random effects meta-analysis
prediction	Print prediction interval
backtransf	Back transform results in printouts and plots
method	Meta-analysis method
w.common	Weights for common effect model (individual studies)
TE.common	Estimated overall effect (common effect model)
seTE.common	Standard error of overall effect (common effect model)
statistic.common	Statistic for test of overall effect (common effect model)
pval.common	P-value for test of overall effect (common effect model)
level.ma	Level of confidence interval for meta-analysis estimates
lower.common	Lower confidence limit (common effect model)
upper.common	Upper confidence limit (common effect model)
w.random	Weight for random effects model (individual studies)
TE.random	Estimated overall effect (random effects model)
seTE.random	Standard error of overall effect (random effects model)
statistic.random	Statistic for test of overall effect (random effects model)
pval.random	P-value for test of overall effect (random effects model)
method.random.ci	Confidence interval method (random effects model)
df.random	Degrees of freedom (random effects model)
lower.random	Lower confidence limit (random effects model)
upper.random	Upper confidence limit (random effects model)
seTE.classic	Standard error (classic random effects method)
adhoc.hakn.ci	Ad hoc correction for Hartung-Knapp method (confidence interval)
df.hakn.ci	Degrees of freedom for Hartung-Knapp method
	(if used in meta-analysis)
seTE.hakn.ci	Standard error for Hartung-Knapp method
	(not taking ad hoc variance correction into account)
seTE.hakn.adhoc.ci	Standard error for Hartung-Knapp method
	(taking ad hoc variance correction into account)
df.kero	Degrees of freedom for Kenward-Roger method
	(if used in meta-analysis)
seTE.kero	Standard error for Kenward-Roger method
method.predict	Method to calculate prediction interval
adhoc.hakn.pi	Ad hoc correction for Hartung-Knapp method (prediction interval)
df.hakn.ci	Degrees of freedom for Hartung-Knapp method
	(prediction interval)
seTE.predict	Standard error used to calculate prediction interval
df.predict	Degrees of freedom for prediction interval
level.predict	Level of prediction interval
lower.predict	Lower limit of prediction interval
upper.predict	Upper limit of prediction interval
seTE.hakn.pi	Standard error for Hartung-Knapp method
	(not taking ad hoc variance correction into account)
seTE.hakn.adhoc.pi	Standard error for Hartung-Knapp method
	(taking ad hoc variance correction into account)
Q	Heterogeneity statistic
df.Q	Degrees of freedom for heterogeneity statistic Q
pval.Q	P-value of heterogeneity test
method.tau	Method to estimate between-study variance \(\tau^2\)
control	Additional arguments for iterative estimation of \(\tau^2\)
method.tau.ci	Method for confidence interval of \(\tau^2\)
tau2	Between-study variance \(\tau^2\)
se.tau2	Standard error of \(\tau^2\)
lower.tau2	Lower confidence limit (\(\tau^2\))
upper.tau2	Upper confidence limit (\(\tau^2\))
tau	Square-root of between-study variance \(\tau\)
lower.tau	Lower confidence limit (\(\tau\))
upper.tau	Upper confidence limit (\(\tau\))
tau.preset	Prespecified value for \(\tau\)
TE.tau	Effect estimate used to estimate \(\tau^2\)
detail.tau	Detail on between-study variance estimate
H	Heterogeneity statistic H
lower.H	Lower confidence limit (heterogeneity statistic H)
upper.H	Upper confidence limit (heterogeneity statistic H)
I2	Heterogeneity statistic I\(^2\)
lower.I2	Lower confidence limit (heterogeneity statistic I\(^2\))
upper.I2	Upper confidence limit (heterogeneity statistic I\(^2\))
Rb	Heterogeneity statistic R\(_b\)
lower.Rb	Lower confidence limit (heterogeneity statistic R\(_b\))
upper.Rb	Upper confidence limit (heterogeneity statistic R\(_b\))
method.bias	Method to test for funnel plot asymmetry
text.common	Label for common effect model
text.random	Label for random effects model
text.predict	Label for prediction interval
text.w.common	Label for weights (common effect model)
text.w.random	Label for weights (random effects model)
title	Title of meta-analysis / systematic review
complab	Comparison label
outclab	Outcome label
label.e	Label for experimental group
label.c	Label for control group
label.left	Graph label on left side of forest plot
label.right	Graph label on right side of forest plot
keepdata	Keep original data
data	Original data (set) used in function call (if keepdata = TRUE)
subset	Information on subset of original data used in meta-analysis
	(if keepdata = TRUE)
exclude	Studies excluded from meta-analysis
warn	Print warnings
call	Function call
version	Version of R package meta used to create object

For subgroup analysis (argument subgroup), the following additional components are added to the list.

subgroup	Subgroup information (for individual studies)
subgroup.name	Name of subgroup variable
print.subgroup.name	Print name of subgroup variable
sep.subgroup	Separator between name of subgroup variable and value
test.subgroup	Print test for subgroup differences
prediction.subgroup	Print prediction interval for subgroup(s)
tau.common	Assumption of common between-study variance in subgroups
subgroup.levels	Levels of grouping variable
k.w	Number of estimates combined in subgroups
k.study.w	Number of studies combined in subgroups
k.all.w	Number of studies in subgroups
k.TE.w	Number of studies with estimable effects in subgroups
TE.common.w	Estimated effect in subgroups (common effect model)
seTE.common.w	Standard error in subgroups (common effect model)
statistic.common.w	Statistic for test of effect in subgroups (common effect model)
pval.common.w	P-value for test of effect in subgroups (common effect model)
lower.common.w	Lower confidence limit in subgroups (common effect model)
upper.common.w	Upper confidence limit in subgroups (common effect model)
w.common.w	Total weight in subgroups (common effect model)
TE.random.w	Estimated effect in subgroups (random effect model)
seTE.random.w	Standard error in subgroups (random effects model)
statistic.random.w	Statistic for test of effect in subgroups (random effects model)
pval.random.w	P-value for test of effect in subgroups (random effects model)
df.random.w	Degrees of freedom in subgroups (random effects model)
lower.random.w	Lower confidence limit in subgroups (random effects model)
upper.random.w	Upper confidence limit in subgroups (random effects model)
w.random.w	Total weight in subgroups (random effects model)
seTE.classic.w	Standard error (classic random effects method)
df.hakn.ci.w	Degrees of freedom for Hartung-Knapp method in subgroups
seTE.hakn.ci.w	Standard error for Hartung-Knapp method in subgroups
	(not taking ad hoc variance correction into account)
seTE.hakn.adhoc.ci.w	Standard error for Hartung-Knapp method in subgroups
df.kero.w	Degrees of freedom for Kenward-Roger method in subgroups
seTE.kero.w	Standard error for Kenward-Roger method in subgroups
seTE.predict.w	Standard error for prediction interval in subgroups
df.predict.w	Degrees of freedom for prediction interval in subgroups
lower.predict.w	Lower limit of prediction interval in subgroups
upper.predict.w	Upper limit of prediction interval in subgroups
seTE.hakn.pi.w	Standard error for Hartung-Knapp method in subgroups (prediction intervals)
	(not taking ad hoc variance correction into account)
seTE.hakn.adhoc.pi.w	Standard error for Hartung-Knapp method in subgroups (prediction intervals)
Q.w	Heterogeneity statistic Q in subgroups
pval.Q.w	P-value for test of heterogeneity in subgroups
tau2.w	Between-study variance \(\tau^2\) in subgroups
tau.w	Square-root of between-study variance \(\tau\) in subgroups
H.w	Heterogeneity statistic H in subgroups
lower.H.w	Lower confidence limit for H in subgroups
upper.H.w	Upper confidence limit for H in subgroups
I2.w	Heterogeneity statistic I\(^2\) in subgroups
lower.I2.w	Lower confidence limit for I\(^2\) in subgroups
upper.I2.w	Upper confidence limit for I\(^2\) in subgroups
Rb.w	Heterogeneity statistic R\(_b\) in subgroups
lower.Rb.w	Lower confidence limit for R\(_b\) in subgroups
upper.Rb.w	Upper confidence limit for R\(_b\) in subgroups
Q.w.common	Within-group heterogeneity statistic Q (common effect model)
Q.w.random	Within-group heterogeneity statistic Q (random effects model)
	(only calculated if argument tau.common = TRUE)
df.Q.w	Degrees of freedom for Q.w.common and Q.w.random
pval.Q.w.common	P-value of test for residual heterogeneity (common effect model)
pval.Q.w.random	P-value of test for residual heterogeneity (random effects model)
Q.b.common	Between-groups heterogeneity statistic Q (common effect model)
df.Q.b.common	Degrees of freedom for Q.b.common
pval.Q.b.common	P-value of test for subgroup differences (common effect model)
Q.b.random	Between-groups heterogeneity statistic Q (random effects model)
df.Q.b.random	Degrees of freedom for Q.b.random
pval.Q.b.random	P-value of test for subgroup differences (random effects model)

An object created with metabin has the additional class "metabin" and the following components.

event.e	Events in experimental group (individual studies)
n.e	Sample size in experimental group (individual studies)
event.e	Events in control group (individual studies)
n.e	Sample size in control group (individual studies)
incr	Increment added to zero cells
method.incr	Continuity correction method
sparse	Continuity correction applied
allstudies	Include studies with double zeros
doublezeros	Indicator for studies with double zeros
MH.exact	Exact Mantel-Haenszel method
RR.Cochrane	Cochrane method to calculate risk ratio
Q.Cochrane	Cochrane method to calculate \(\tau^2\)
Q.CMH	Cochran-Mantel-Haenszel statistic
df.Q.CMH	Degrees of freedom for Q.CMH
pval.Q.CMH	P-value of Cochran-Mantel-Haenszel test
print.CMH	Print results for Cochran-Mantel-Haenszel statistic
incr.e	Continuity correction in experimental group (individual studies)
incr.c	Continuity correction in control group (individual studies)
k.MH	Number of studies (Mantel-Haenszel method)

An object created with metacont has the additional class "metacont" and the following components.

n.e	Sample size in experimental group (individual studies)
mean.e	Estimated mean in experimental group (individual studies)
sd.e	Standard deviation in experimental group (individual studies)
n.c	Sample size in control group (individual studies)
mean.c	Estimated mean in control group (individual studies)
sd.c	Standard deviation in control group (individual studies)
pooledvar	Use pooled variance for mean difference
method.smd	Method for standardised mean difference (SMD)
sd.glass	Denominator in Glass' method
exact.smd	Use exact formulae for SMD
method.ci	Method to calculate confidence limits
method.mean	Method to approximate mean
method.sd	Method to approximate standard deviation

An object created with metacor has the additional class "metacor" and the following components.

cor	Correlation (individual studies)
n	Sample size (individual studies)

An object created with metainc has the additional class "metainc" and the following components.

event.e	Events in experimental group (individual studies)
time.e	Person time in experimental group (individual studies)
n.e	Sample size in experimental group (individual studies)
event.c	Events in control group (individual studies)
time.c	Person time in control group (individual studies)
n.c	Sample size in control group (individual studies)
incr	Increment added to zero cells
method.incr	Continuity correction method
sparse	Continuity correction applied
incr.event	Continuity correction (individual studies)
k.MH	Number of studies (Mantel-Haenszel method)

An object created with metamean has the additional class "metamean" and the following components.

n	Sample size (individual studies)
mean	Estimated mean (individual studies)
sd	Standard deviation (individual studies)
method.ci	Method to calculate confidence limits
method.mean	Method to approximate mean
method.sd	Method to approximate standard deviation

An object created with metaprop has the additional class "metaprop" and the following components.

event	Events (individual studies)
n	Sample size (individual studies)
incr	Increment added to zero cells
method.incr	Continuity correction method
sparse	Continuity correction applied
method.ci	Method to calculate confidence limits
incr.event	Continuity correction (individual studies)

An object created with metarate has the additional class "metarate" and the following components.

event	Events (individual studies)
time	Person time (individual studies)
n	Sample size (individual studies)
incr	Increment added to zero cells
method.incr	Continuity correction method
sparse	Continuity correction applied
method.ci	Method to calculate confidence limits
incr.event	Continuity correction (individual studies)

An object created with trimfill has the additional classes "trimfill" and "metagen" and the following components.

k0	Number of added studies
left	Studies missing on left side
ma.common	Use common effect or random effects model to estimate
	number of missing studies
type	Method to estimate missing studies
n.iter.max	Maximum number of iterations
n.iter	Number of iterations
trimfill	Filled studies (individual studies)
class.x	Primary class of meta-analysis object

See Also

meta-package, meta-sm, print.meta, summary.meta, forest.meta