Title | Meta-analysis in Stata | |

Authors |
Jonathan A. C. Sterne, University of Bristol Ross J. Harris, University of Bristol Roger M. Harbord, University of Bristol Thomas J. Steichen |

As of Stata 16, Stata has an official suite of meta-analysis commands. See Stata's full list of official meta-analysis features.

Stata users have also developed numerous excellent commands for performing meta-analyses.

In 2016, Stata published *Meta-Analysis in Stata: An Updated Collection
from the Stata Journal, Second Edition*, which brought together all the *Stata
Journal* articles about the community-contributed meta-analysis commands for Stata. This book is available for
purchase at
stata-press.com/books/meta-analysis-in-stata/.

The following meta-analysis commands are all described in *Meta-Analysis in
Stata: An Updated Collection from the Stata Journal, Second Edition*. All of
them are community-contributed except 40. **sem** and **gsem**.

**metan** is the main community-contributed Stata meta-analysis
command; see **meta** for
official features. Its latest version allows the user to input the cell
frequencies from the 2 × 2 table for each study (for binary
outcomes), the mean and standard deviation in each group (for numerical
outcomes), or the effect estimate and standard error from each study. It
provides a comprehensive range of methods for meta-analysis,
including inverse-variance–weighted meta-analysis, and creates new
variables containing the treatment effect estimate and its standard error
for each study. These variables can then be used as input to other
Stata meta-analysis commands. Meta-analyses may be conducted in subgroups
by using the **by()** option.

All the meta-analysis calculations available in **metan** are based on
standard methods, an overview of which may be found in chapter 15 of
Deeks, Altman, and Bradburn (2001).

The version of the **metan** command that used Stata 7
graphics has been renamed **metan7** and is downloaded as part of the
**metan** package currently available on the SSC archive.

The most recent help file for **metan** provides several clickable
examples of using the command.

**labbe** draws a L’Abbe plot for event data (proportions of
successes in the two groups).

**metaan** performs meta-analysis on effect estimates and standard
errors. Included are profile likelihood and permutation estimation, two
algorithms not available in **metan**.

**metacum** performs cumulative meta-analyses and graphs the results.

**metap** combines *p*-values by using Fisher’s method,
Edgington’s additive method, or Edgington’s normal curve method.
It was released in 1999 as a version 6 command (no graphics) and was last
updated in 2000. It requires the user to input a *p*-value for each
study.

**metareg** does meta-regression. It was first released in 1998 and has been updated to take account of improvements in Stata estimation facilities and recent methodological developments. It requires the user to
input the treatment effect estimate and its standard error for each study.

**metafunnel** plots funnel plots. It was released in 2004 and uses Stata 8
graphics. It requires the user to input the treatment effect estimate and
its standard error for each study.

**confunnel** plots contour-enhanced funnel plots. The command has been designed to be flexible, allowing the user to add extra features to the funnel plot.

**metabias** provides statistical tests for funnel plot asymmetry. It was
first released in 1997, but it has been updated to provide recently proposed
tests that maintain better control of the false-positive rate than those
available in the original command.

**metatrim** implements the “trim and fill” method to adjust
for publication bias in funnel plots. It requires the user to input the
treatment effect estimate and its standard error for each study.

**extfunnel** implements a new range of overlay augmentations to the funnel plot
to assess the impact of a new study on an existing meta-analysis.

**metandi** facilitates the fitting of hierarchical logistic regression
models for meta-analysis of diagnostic test accuracy studies.
**metandiplot** produces a graph of the model fit by **metandi**,
which must be the last estimation-class command executed.

**mvmeta** performs maximum likelihood, restricted maximum likelihood, or
method-of-moments estimation of random-effects multivariate meta-analysis
models. **mvmeta_make** facilitates the preparation of summary datasets
from more detailed data.

**ipdforest** is a postestimation command that uses the stored
estimates of an **xtmixed** or **xtmelogit** command for
multilevel linear or logistic regression, respectively.

**ipdmetan** performs two-stage individual participant data meta-analysis
using the inverse-variance method.

**indirect** performs pairwise indirect treatment comparisons.

**network setup** imports data from a set of studies reporting count data
(events, total number) or quantitative data (mean, standard deviation, total
number) for two or more treatments.

**network import** imports a dataset already formatted for network meta-analysis.

**network table** tabulates network meta-analysis data.

**network pattern** shows which treatments are used in which studies.

**network map** draws a map of a network; that is, it shows which treatments are
directly compared against which other treatments and roughly how much
information is available for each treatment and for each treatment comparison.

**network convert** converts between the three formats described in the help file
for **network**.

**network query** displays the current network settings.

**network unset** deletes the current network settings.

**network meta** defines a model to be fit: either the consistency model or the
design-by-treatment interaction inconsistency model.

**network rank** ranks treatments after a network meta-analysis has been fit.

**network sidesplit** fits the node-splitting model of Dias et al. (2010).

**network forest** draws a forest plot of network meta-analysis data.

**networkplot** plots a network of interventions using nodes and edges.

**netweight** calculates all direct pairwise summary effect sizes
with their variances, creates the design matrix, and estimates the
percentage contribution of each direct comparison to the network summary
estimates and in the entire network.

**ifplot** identifies all triangular and quadratic loops in a network of
interventions and estimates the respective inconsistency factors and their
uncertainties.

**netfunnel** plots a comparison-adjusted funnel plot for assessing
small-study effects within a network of interventions.

**intervalplot** plots the estimated effect sizes and their
uncertainties for all pairwise comparisons in a network meta-analysis.

**netleague** creates a "league table" showing in the off-diagonal cells
the relative treatment effects for all possible pairwise comparisons estimated
in a network meta-analysis.

**sucra** gives the surface under the cumulative ranking curves percentages and mean ranks, and produces
rankograms (line plots of the probabilities versus ranks) and cumulative ranking
plots (line plots of the cumulative probabilities versus ranks) for all
treatments in a network of interventions.

**mdsrank** creates the squared matrix containing the pairwise relative
effect sizes and plots the resulting values of the unique dimension for each
treatment.

**clusterank** performs hierarchical cluster analysis to group the
competing treatments into meaningful groups.

**glst** calculates a log-linear dose–response
regression model using generalized least squares for trend estimation of
single or multiple summarized dose–response epidemiological studies.
Output from this command may be useful in deriving summary effects and their
standard errors for inclusion in meta-analyses of such studies.

**metamiss** performs meta-analysis with binary outcomes when some or all studies have missing data.

Describes how to fit fixed- and random-effects meta-analysis
models using the **sem** and **gsem** commands, introduced in Stata 12 and 13 respectively,
for structural equation modeling.

**metacumbounds** provides *z*-values, *p*-values, and
Lan-DeMets bounds obtained from fixed- or random-effects meta-analysis.
It plots the boundaries and *z*-values through a process.

**metasim** simulates a specified number of new studies based on the estimates
obtained from a preexisting meta-analysis.

**metapow** implements an approach to estimating the power of a newly simulated
study generated by using the program **metasim**.

**metapowplot** estimates the power of an updated meta-analysis including a new
study and plots each value against a range of sample sizes.

The following commands are documented in the Appendix:

**metacurve** models a response as a function of a continuous covariate, optionally adjusting for other variable(s) specified by **adjust()**.

**metannt** is intended to aid interpretation of meta-analyses of binary
data by presenting intervention effect sizes in absolute terms, as the
number needed to treat (NNT) and the number of events avoided (or added) per
1,000. The user inputs design parameters, and **metannt** uses the
**metan** command to calculate the required statistics. This command is
available as part of the **metan** package.

**metaninf** is a port of the **metainf** command to use **metan** as
its analysis engine rather than **meta**. It was released in 2001 as a
version 6 command using version 6 graphics and was last updated in 2004. It
requires the user to provide input in the form needed by **metan**.

**midas** provides statistical and graphical routines for undertaking
meta-analysis of diagnostic test performance in Stata.

**meta_lr** graphs positive and negative likelihood ratios in diagnostic
tests. It can do stratified meta-analysis of individual estimates. The user
must provide the effect estimates (log positive likelihood ratio and log
negative likelihood ratio) and their standard errors. Commands **meta**
and **metareg** are used for internal calculations. This is a version 8
command released in 2004.

**metaparm** performs meta-analyses and calculates confidence
intervals and *p*-values for differences or ratios between parameters for
different subpopulations for data stored in the **parmest** format.

**metaeff** is a pre-processing command for meta-analysis and a
companion to **metaan** which calculates effect sizes and their
standard errors.

- Deeks, J. J., D. G. Altman, and M. J. Bradburn. 2001.
- Statistical methods for examining heterogeneity and combining results
from several studies in meta-analysis. In
*Systematic Reviews in Health Care: Meta-Analysis in Context, 2nd Edition*, ed. M. Egger, G. Davey Smith, and D. G. Altman. London: BMJ.

- Dias, S., N. J. Welton, D. M. Caldwell, and A. E. Ades. 2010.
- Checking
consistency in mixed treatment comparison meta-analysis.
*Statistics in Medicine*29: 932–944.