metacum: Cumulative meta-analysis

In guido-s/meta: General Package for Meta-Analysis

Cumulative meta-analysis

Description

Performs a cumulative meta-analysis.

Usage

## S3 method for class 'meta'
metacum(
  x,
  pooled,
  sortvar,
  prediction,
  overall = x$overall,
  text.pooled,
  no = 1,
  cid = NULL,
  cid.below.null = NULL,
  cid.above.null = NULL,
  small.values = "desirable",
  ...
)

metacum(x, ...)

## Default S3 method:
metacum(x, ...)

Arguments

x	An object of class meta.
pooled	A character string indicating whether a common effect or random effects model is used for pooling. Either missing (see Details), "common", or "random", can be abbreviated.
sortvar	An optional vector used to sort the individual studies (must be of same length as x$TE).
prediction	A logical indicating whether to report prediction intervals.
overall	A logical indicating whether overall results should be reported.
text.pooled	A character string used in printouts and forest plots to label the pooled effect estimate.
no	A numeric specifying which meta-analysis results to consider.
cid	A numeric value or vector specifying clinically important differences (CID) / decision thresholds used to calculate expected proportions of clinically important benefit or harm (see cidprop).
cid.below.null	A single numeric defining the decision threshold below the null effect to distinguish clinically important from not important effects (see cidprop).
cid.above.null	A single numeric defining the decision threshold above the null effect to distinguish clinically important from not important effects (see cidprop).
small.values	A character string specifying whether small treatment effects indicate a beneficial ("desirable") or harmful ("undesirable") effect, can be abbreviated (see cidprop).
...	Additional arguments (ignored).

Details

A cumulative meta-analysis is performed. Studies are included sequentially as defined by sortvar.

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

Value

An object of class "metacum" with dedicated print and forest functions.

The following list elements provide results from meta-analyses, each adding one study at a time (see meta-object for more information on these list elements):

studlab, TE, seTE, df.random, lower, upper, statistic, pval,

lower.predict, upper.predict, df.predict, w (sum of weights),

tau2, se.tau2, lower.tau2, upper.tau2, tau, lower.tau, upper.tau,

I2, lower.I2, upper.I2, Rb, n.harmonic.mean, t.harmonic.mean,

k, k.study, k.all, k.TE, k.MH.

The following list elements contain results of the original meta-analysis:

TE.pooled, seTE.pooled, df.random.pooled,

lower.pooled, upper.pooled, statistic.pooled, pval.pooled,

lower.predict.pooled, upper.predict.pooled,

df.predict.pooled, w.pooled,

tau2.pooled, se.tau2.pooled, lower.tau2.pooled, upper.tau2.pooled,

tau.pooled, lower.tau.pooled, upper.tau.pooled,

I2.pooled, lower.I2.pooled, upper.I2.pooled, Rb.pooled,

n.harmonic.mean.pooled, t.harmonic.mean.pooled,

k.pooled, k.study.pooled, k.all.pooled, k.TE.pooled, k.MH.pooled.

Author(s)

Guido Schwarzer [e-mail]

References

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

See Also

forest.metacum, print.metacum, cidprop

Examples

data(Fleiss1993bin)
m1 <- metabin(d.asp, n.asp, d.plac, n.plac,
  data = Fleiss1993bin, studlab = study, sm = "RR", method = "I")
m1
metacum(m1)
metacum(m1, pooled = "random")

forest(metacum(m1))
forest(metacum(m1, pooled = "random"))

metacum(m1, sortvar = study)
metacum(m1, sortvar = 7:1)

m2 <- update(m1, title = "Fleiss1993bin meta-analysis", backtransf = FALSE)
metacum(m2)

data(Fleiss1993cont)
m3 <- metacont(n.psyc, mean.psyc, sd.psyc, n.cont, mean.cont, sd.cont,
  data = Fleiss1993cont, sm = "SMD")
metacum(m3)

guido-s/meta documentation built on June 12, 2025, 11:48 p.m.