metainf: Influence analysis in meta-analysis using leave-one-out...

Description Usage Arguments Details Value Author(s) References See Also Examples

View source: R/metainf.R

Description

Performs an influence analysis. Pooled estimates are calculated omitting one study at a time.

Usage

1
metainf(x, pooled, sortvar)

Arguments

x

An object of class meta.

pooled

A character string indicating whether a fixed effect or random effects model is used for pooling. Either missing (see Details), "fixed" or "random", can be abbreviated.

sortvar

An optional vector used to sort the individual studies (must be of same length as x$TE).

Details

Performs a influence analysis; pooled estimates are calculated omitting one study at a time. Studies are sorted according to sortvar.

Information from object x is utilised if argument pooled is missing. A fixed effect model is assumed (pooled="fixed") if argument x$comb.fixed is TRUE; a random effects model is assumed (pooled="random") if argument x$comb.random is TRUE and x$comb.fixed is FALSE.

Value

An object of class c("metainf", "meta") with corresponding print, and forest functions. The object is a list containing the following components:

TE, seTE

Estimated treatment effect and standard error of pooled estimate in influence analysis.

lower, upper

Lower and upper confidence interval limits.

studlab

Study label describing omission of studies.

p.value

P-value for test of overall effect.

w

Sum of weights from fixed effect or random effects model.

I2

Heterogeneity statistic I2.

Rb

Heterogeneity statistic Rb.

tau

Square-root of between-study variance.

df.hakn

Degrees of freedom for test of treatment effect for Hartung-Knapp method (only if hakn=TRUE).

sm

Summary measure.

method

Method used for pooling.

k

Number of studies combined in meta-analysis.

pooled

As defined above.

comb.fixed

A logical indicating whether analysis is based on fixed effect model.

comb.random

A logical indicating whether analysis is based on random effects model.

TE.fixed, seTE.fixed

Value is NA.

TE.random, seTE.random

Value is NA.

Q

Value is NA.

level.comb

The level used to calculate confidence intervals for pooled estimates.

hakn

A logical indicating whether the method by Hartung and Knapp is used to adjust test statistics and confidence intervals.

method.tau

A character string indicating which method is used to estimate the between-study variance τ^2.

tau.preset

Prespecified value for the square-root of the between-study variance τ^2.

TE.tau

Overall treatment effect used to estimate the between-study variance τ^2.

n.harmonic.mean

Harmonic mean of number of observations (for back transformation of Freeman-Tukey Double arcsine transformation).

version

Version of R package meta used to create object.

Author(s)

Guido Schwarzer [email protected]

References

Cooper H & Hedges LV (1994), The Handbook of Research Synthesis. Newbury Park, CA: Russell Sage Foundation.

See Also

metabin, metacont, print.meta

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
data(Fleiss93)
meta1 <- metabin(event.e, n.e, event.c, n.c,
                 data=Fleiss93, studlab=study,
                 sm="RR", method="I")
meta1

metainf(meta1)
metainf(meta1, pooled="random")

forest(metainf(meta1))
forest(metainf(meta1), layout="revman5")
forest(metainf(meta1, pooled="random"))

metainf(meta1, sortvar=study)
metainf(meta1, sortvar=7:1)

meta2 <- update(meta1, title="Fleiss93 meta-analysis",
                backtransf=FALSE)
metainf(meta2)


data(Fleiss93cont)
meta3 <- metacont(n.e, mean.e, sd.e, n.c, mean.c, sd.c,
                  data = Fleiss93cont, sm = "SMD")
metainf(meta3)

Example output

Loading 'meta' package (version 4.8-2).
Type 'help(meta)' for a brief overview.
           RR           95%-CI %W(fixed) %W(random)
MRC-1  0.7420 [0.5223; 1.0543]       2.4        7.8
CDP    0.6993 [0.4828; 1.0129]       2.1        7.2
MRC-2  0.8270 [0.6487; 1.0545]       4.9       13.6
GASP   0.8209 [0.5269; 1.2789]       1.5        5.3
PARIS  0.8193 [0.5927; 1.1326]       2.8        8.9
AMIS   1.1183 [0.9411; 1.3289]       9.7       20.4
ISIS-2 0.9142 [0.8596; 0.9722]      76.6       36.8

Number of studies combined: k = 7

                         RR           95%-CI     z  p-value
Fixed effect model   0.9137 [0.8658; 0.9643] -3.28   0.0010
Random effects model 0.8929 [0.8006; 0.9959] -2.03   0.0419

Quantifying heterogeneity:
 tau^2 = 0.0074; H = 1.29 [1.00; 1.98]; I^2 = 39.6% [0.0%; 74.6%]

Test of heterogeneity:
    Q d.f.  p-value
 9.93    6   0.1277

Details on meta-analytical method:
- Inverse variance method
- DerSimonian-Laird estimator for tau^2

Influential analysis (Fixed effect model)

                      RR           95%-CI  p-value    tau^2     I^2
Omitting MRC-1    0.9183 [0.8696; 0.9698]   0.0022   0.0074   41.5%
Omitting CDP      0.9190 [0.8703; 0.9705]   0.0024   0.0060   36.6%
Omitting MRC-2    0.9185 [0.8691; 0.9706]   0.0025   0.0100   45.9%
Omitting GASP     0.9152 [0.8669; 0.9663]   0.0014   0.0095   48.5%
Omitting PARIS    0.9166 [0.8679; 0.9680]   0.0018   0.0095   47.3%
Omitting AMIS     0.8940 [0.8447; 0.9462]   0.0001   0.0000    0.0%
Omitting ISIS-2   0.9124 [0.8162; 1.0199]   0.1066   0.0214   49.6%
                                                                   
Pooled estimate   0.9137 [0.8658; 0.9643]   0.0010   0.0074   39.6%

Details on meta-analytical method:
- Inverse variance method

Influential analysis (Random effects model)

                      RR           95%-CI  p-value    tau^2     I^2
Omitting MRC-1    0.9071 [0.8098; 1.0162]   0.0924   0.0074   41.5%
Omitting CDP      0.9124 [0.8201; 1.0151]   0.0921   0.0060   36.6%
Omitting MRC-2    0.8983 [0.7907; 1.0204]   0.0991   0.0100   45.9%
Omitting GASP     0.8931 [0.7922; 1.0068]   0.0643   0.0095   48.5%
Omitting PARIS    0.8963 [0.7929; 1.0131]   0.0799   0.0095   47.3%
Omitting AMIS     0.8940 [0.8447; 0.9462]   0.0001   0.0000    0.0%
Omitting ISIS-2   0.8596 [0.7249; 1.0194]   0.0820   0.0214   49.6%
                                                                   
Pooled estimate   0.8929 [0.8006; 0.9959]   0.0419   0.0074   39.6%

Details on meta-analytical method:
- Inverse variance method
- DerSimonian-Laird estimator for tau^2

Influential analysis (Fixed effect model)

                      RR           95%-CI  p-value    tau^2     I^2
Omitting AMIS     0.8940 [0.8447; 0.9462]   0.0001   0.0000    0.0%
Omitting CDP      0.9190 [0.8703; 0.9705]   0.0024   0.0060   36.6%
Omitting GASP     0.9152 [0.8669; 0.9663]   0.0014   0.0095   48.5%
Omitting ISIS-2   0.9124 [0.8162; 1.0199]   0.1066   0.0214   49.6%
Omitting MRC-1    0.9183 [0.8696; 0.9698]   0.0022   0.0074   41.5%
Omitting MRC-2    0.9185 [0.8691; 0.9706]   0.0025   0.0100   45.9%
Omitting PARIS    0.9166 [0.8679; 0.9680]   0.0018   0.0095   47.3%
                                                                   
Pooled estimate   0.9137 [0.8658; 0.9643]   0.0010   0.0074   39.6%

Details on meta-analytical method:
- Inverse variance method

Influential analysis (Fixed effect model)

                      RR           95%-CI  p-value    tau^2     I^2
Omitting ISIS-2   0.9124 [0.8162; 1.0199]   0.1066   0.0214   49.6%
Omitting AMIS     0.8940 [0.8447; 0.9462]   0.0001   0.0000    0.0%
Omitting PARIS    0.9166 [0.8679; 0.9680]   0.0018   0.0095   47.3%
Omitting GASP     0.9152 [0.8669; 0.9663]   0.0014   0.0095   48.5%
Omitting MRC-2    0.9185 [0.8691; 0.9706]   0.0025   0.0100   45.9%
Omitting CDP      0.9190 [0.8703; 0.9705]   0.0024   0.0060   36.6%
Omitting MRC-1    0.9183 [0.8696; 0.9698]   0.0022   0.0074   41.5%
                                                                   
Pooled estimate   0.9137 [0.8658; 0.9643]   0.0010   0.0074   39.6%

Details on meta-analytical method:
- Inverse variance method
Review:     Fleiss93 meta-analysis

Influential analysis (Fixed effect model)

                    logRR             95%-CI  p-value    tau^2     I^2
Omitting MRC-1    -0.0852 [-0.1397; -0.0307]   0.0022   0.0074   41.5%
Omitting CDP      -0.0844 [-0.1389; -0.0300]   0.0024   0.0060   36.6%
Omitting MRC-2    -0.0850 [-0.1403; -0.0298]   0.0025   0.0100   45.9%
Omitting GASP     -0.0886 [-0.1429; -0.0343]   0.0014   0.0095   48.5%
Omitting PARIS    -0.0871 [-0.1417; -0.0325]   0.0018   0.0095   47.3%
Omitting AMIS     -0.1120 [-0.1687; -0.0553]   0.0001   0.0000    0.0%
Omitting ISIS-2   -0.0917 [-0.2031;  0.0197]   0.1066   0.0214   49.6%
                                                                      
Pooled estimate   -0.0902 [-0.1441; -0.0363]   0.0010   0.0074   39.6%

Details on meta-analytical method:
- Inverse variance method

Influential analysis (Fixed effect model)

                      SMD             95%-CI  p-value    tau^2     I^2
Omitting 1        -0.3439 [-0.6239; -0.0638]   0.0161   0.0200   18.4%
Omitting 2        -0.2359 [-0.5567;  0.0848]   0.1494   0.0000    0.0%
Omitting 3        -0.3632 [-0.6808; -0.0456]   0.0250   0.0247   17.3%
Omitting 4        -0.4464 [-0.7373; -0.1555]   0.0026   0.0000    0.0%
Omitting 5        -0.3156 [-0.5882; -0.0431]   0.0232   0.0020    2.5%
                                                                      
Pooled estimate   -0.3434 [-0.6068; -0.0800]   0.0106   0.0000    0.0%

Details on meta-analytical method:
- Inverse variance method

meta documentation built on Dec. 6, 2017, 1:06 a.m.