This function calculates the probabilities of each treatment being at each possible rank and the SUCRAs (Surface Under the Cumulative RAnking curve) in frequentist network meta-analysis.
# S3 method for netmeta
rankogram(
x,
nsim = gs("nsim"),
common = x$common,
random = x$random,
small.values = x$small.values,
cumulative.rankprob = FALSE,
keep.samples = FALSE,
nchar.trts = x$nchar.trts,
warn.deprecated = gs("warn.deprecated"),
...
)An object of class rankogram with corresponding print
and plot function. The object is a list containing the
following components:
As defined above.
As defined above.
As defined above.
As defined above.
As defined above.
Function call.
Version of R package netmeta used to create object.
The following components are available if argument common = TRUE:
SUCRA values for the common effects model.
Numeric matrix giving the probability of each treatment being at each possible rank for the common effects model.
Numeric matrix giving the cumulative ranking probabilities for the common effects model.
Mean ranks for the common effects model.
Median ranks for the common effects model.
Generated samples for the common effects model
(if keep.samples = TRUE).
The following components are available if argument random = TRUE:
SUCRA values for the random effects model.
Numeric matrix giving the probability of each treatment being at each possible rank for the random effects model.
Numeric matrix giving the cumulative ranking probabilities for the random effects model.
Mean ranks for the random effects model.
Median ranks for the random effects model.
Generated samples for the random effects model
(if keep.samples = TRUE).
An object of class netmeta.
Number of repetitions.
A logical indicating to compute ranking probabilities and SUCRAs for the common effects model.
A logical indicating to compute ranking probabilities and SUCRAs for the random effects model.
A character string specifying whether small
treatment effects indicate a beneficial ("desirable") or
harmful ("undesirable") effect, can be abbreviated.
A logical indicating whether cumulative ranking probabilities should be printed.
A logical indicating whether to keep the generated samples.
A numeric defining the minimum number of characters used to create unique treatment names.
A logical indicating whether warnings should be printed if deprecated arguments are used.
Additional arguments (ignored).
Theodoros Papakonstantinou [email protected], Guido Schwarzer [email protected]
We derive a matrix showing the probability of each treatment being at each possible rank. To this aim, we use resampling from a multivariate normal distribution with estimated network effects as means and corresponding estimated variance covariance matrix. We then summarise them using the ranking metric SUCRAs (Surface Under the Cumulative RAnking curve).
Salanti G, Ades AE, Ioannidis JP (2011): Graphical methods and numerical summaries for presenting results from multiple-treatment meta-analysis: an overview and tutorial. Journal of Clinical Epidemiology, 64, 163--71
netmeta, netrank,
plot.rankogram, Woods2010
pw <- pairwise(treatment, event = r, n = N, studlab = author,
data = Woods2010, sm = "OR")
nma <- netmeta(pw, small.values = "desirable")
set.seed(1909) # get reproducible results
rnk <- rankogram(nma, nsim = 100)
rnk
print(rnk, cumulative.rankprob = TRUE)
# Also print mean ranks
summary(rnk)
plot(rnk)
Run the code above in your browser using DataLab