meta: metamerge – R documentation

Become an expert in R — Interactive courses, Cheat Sheets, certificates and more!

metamerge

Merge pooled results of two meta-analyses

Description

This function can be used to merge pooled results of two meta-analyses into a single meta-analysis object. This is, for example, useful to produce a forest plot of a random-effects meta-analysis with and without using the Hartung-Knapp method.

Usage

metamerge(
  meta1,
  meta2,
  pooled1,
  pooled2,
  text.pooled1,
  text.pooled2,
  text.w.pooled1,
  text.w.pooled2,
  detail.tau1,
  detail.tau2,
  backtransf
)

Arguments

`meta1`	First meta-analysis object (of class `"meta"`).
`meta2`	Second meta-analysis object (see Details).
`pooled1`	A character string indicating whether results of fixed effect or random effects model should be considered for first meta-analysis. Either `"fixed"` or `"random"`, can be abbreviated.
`pooled2`	A character string indicating whether results of fixed effect or random effects model should be considered for second meta-analysis. Either `"fixed"` or `"random"`, can be abbreviated.
`text.pooled1`	A character string used in printouts and forest plot to label the estimate from the first meta-analysis.
`text.pooled2`	A character string used in printouts and forest plot to label the estimate from the second meta-analysis.
`text.w.pooled1`	A character string used to label weights of the first meta-analysis.
`text.w.pooled2`	A character string used to label weights of the second meta-analysis.
`detail.tau1`	A character string used to label estimate of between-study variance of the first meta-analysis.
`detail.tau2`	A character string used to label estimate of between-study variance of the second meta-analysis.
`backtransf`	A logical indicating whether results should be back transformed in printouts and plots. If `backtransf=TRUE` (default), results for `sm="OR"` are printed as odds ratios rather than log odds ratios, for example.

Details

In R package meta, objects of class "meta" contain results of both a fixed effect and random effects meta-analysis. This function enables the user to keep the results of one of these models and to add results from a second meta-analysis or a sensitivity analysis.

Applications of this function include printing and plotting results of the fixed effect or random effects meta-analysis and the

Hartung-Knapp method (see argument hakn in metagen),
trim-and-fill method (trimfill),
limit meta-analyis (limitmeta from R package metasens),
Copas selection model (copas from R package metasens),
robust variance meta-analysis model (robu from R package robumeta).

The first argument must be an object created by a meta-analysis function, e.g., metagen or metabin. The second meta-analysis could also be an object created with trimfill, limitmeta, copas, or robu.

The created meta-analysis object only contains the study results from the first meta-analysis which are shown in printouts and forest plots. This only makes a difference for meta-analysis methods where individual study results differ, e.g., Mantel-Haenszel and Peto method for binary outcomes (see metabin).

R function metabind can be used to print and plot the results of more than two meta-analyses, however, without showing individual study results.

Value

An object of class "meta" and "metamerge" with corresponding print, summary, and forest functions. The following list elements have a different meaning:

`TE, seTE, studlab`	Treatment estimate, standard error, and study labels (first meta-analyis).
`lower, upper`	Lower and upper confidence interval limits for individual studies (first meta-analysis).
`statistic, pval`	Statistic and p-value for test of treatment effect for individual studies (first meta-analysis.
`w.fixed`	Weight of individual studies (first meta-analysis).
`w.random`	Weight of individual studies (second meta-analysis).
`TE.fixed, seTE.fixed`	Estimated overall treatment effect and standard error (first meta-analysis).
`lower.fixed, upper.fixed`	Lower and upper confidence interval limits (first meta-analysis).
`statistic.fixed, pval.fixed`	Statistic and p-value for test of overall treatment effect (first meta-analysis).
`TE.random, seTE.random`	Estimated overall treatment effect and standard error (second meta-analysis).
`lower.random, upper.random`	Lower and upper confidence interval limits (second meta-analysis).
`statistic.random, pval.random`	Statistic and p-value for test of overall treatment effect (second meta-analysis).
`lower.predict, upper.predict`	Lower and upper limits of prediction interval (related to first meta-analysis).
`k`	Number of studies combined in first meta-analysis.
`Q`	Heterogeneity statistic (first meta-analysis).
`df.Q`	Degrees of freedom for heterogeneity statistic (first meta-analysis).
`pval.Q`	P-value of heterogeneity test (first meta-analysis).
`tau2`	Between-study variance(s) τ^2 (first and second meta-analysis).
`lower.tau2, upper.tau2`	Lower and upper limit of confidence interval(s) for τ^2 (first and second meta-analysis).
`tau`	Square-root of between-study variance(s) τ (first and second meta-analysis).
`lower.tau, upper.tau`	Lower and upper limit of confidence interval(s) for τ (first and second meta-analysis).
`text.fixed`	Label for the first meta-analysis.
`text.random`	Label for the second meta-analysis.

See metagen for information on other list elements.

Author(s)

Guido Schwarzer sc@imbi.uni-freiburg.de

Examples

data(Fleiss1993cont)
#
m1 <- metacont(n.psyc, mean.psyc, sd.psyc, n.cont, mean.cont, sd.cont,
               data = Fleiss1993cont, sm = "MD",
               comb.fixed = FALSE,
               text.random = "Classic random effects",
               text.w.random = "RE")
#
# Use Hartung-Knapp method
#
m2 <- update(m1, hakn = TRUE,
             text.random = "Hartung-Knapp method",
             text.w.random = "HK")
#
# Merge results of the two meta-analyses
#
m12 <- metamerge(m1, m2)
m12
forest(m12, rightcols = c("effect", "ci", "w.fixed"))

# Show results for DerSimonian-Laird and REML estimate of
# between-study variance
#
m3 <- update(m1,
             text.random = "Random effects moded (DL)",
             text.w.random = "DL")
m4 <- update(m1, method.tau = "REML",
             text.random = "Random effects moded (REML)",
             text.w.random = "REML")
#
m34 <- metamerge(m3, m4)
m34

data(Fleiss1993bin)
#
# Mantel-Haenszel method
#
m5 <- metabin(d.asp, n.asp, d.plac, n.plac, data = Fleiss1993bin,
              studlab = paste(study, year),
              sm = "OR", comb.random = FALSE,
              text.fixed = "MH method", text.w.fixed = "MH")
#
# Peto method
#
m6 <- update(m5, method = "Peto", text.fixed = "Peto method",
             text.w.fixed = "Peto")
#
# Merge results (show individual results for MH method)
#
m56 <- metamerge(m5, m6)
m56
forest(m56, digits = 4)
#
# Merge results (show individual results for Peto method)
#
m65 <- metamerge(m6, m5)
m65