.- help for ^metan^, ^labbe^, ^funnel^ (STB-44: sbe24) .- Fixed and random effects meta-analysis -------------------------------------- ^metan^ varlist [^if^ exp] [^in^ range] [, ^rr or rd hedges cohen glass^ ^nosta^ndard ^fixed random peto cornfield chi2^ ^il^evel^(^#^) ol^evel^(^#^) sortby(^sort_vars^) label(^label_vars^)^ ^nokeep notable nograph xla^bel^(^#,..,#^) force t1(^#,..,#^)^ ^t2(^#,..,#^) b1(^#,..,#^) boxsha(^#^) boxsca(^#^) texts(^#^)^ ^saving(^filename^) nowt nostats noove^rall ] ^labbe^ varlist [^if^ exp] [^in^ range] [^weight^] , [ ^nowt perc^ent graph options ] ^funnel^ [varlist] [^if^ exp] [^in^ range] [, ^sa^mple ^noinv^ert ^ysq^rt ^ov^erall(#) graph options ] Description ----------- These routines provide methods for the meta-analysis of either binary (event) or continuous data with two groups. There are several methods available for pooling the study results, and performing formal tests on the data. Additionally, the results may be displayed graphically by a forest plot. ^labbe^ draws a L'Abbe plot for event data (proportion of successes in the two groups). This is an alternative to the graph produced by ^metan^ ^funnel^ may be used for producing a graph of either the study sample size, standard error or precision (inverse of s.e.) against the effect size. Options for ^metan^ ----------------- The main meta-analysis routine ^metan^ requires either four or six variables to be declared. When four variables are specified, analysis of binary data is performed. When six, the data are assumed continuous. Scaling and pooling options for ^metan^ ------------------------------------- Options for binary data ^rr^ pools risk ratios (the default). ^or^ pools odds ratios. ^rd^ pools risk differences. ^fixed^ specifies a fixed effect model using the method of Mantel and Haenszel (the default). ^fixedi^ specifies a fixed effect model using the inverse variance method. ^peto^ specifies that Peto's assumption free method is used to pool odds ratios. ^random^ specifies a random effects model using the method of DerSimonian & Laird, with the estimate of heterogeneity being taken from the Mantel-Haenszel model. ^randomi^ specifies a random effects model using the method of DerSimonian & Laird, with the estimate of heterogeneity being taken from the inverse variance fixed effect model. ^cornfield^ computes confidence intervals for odds ratios by method of Cornfield, rather than the (default) Woolf method. ^chi2^ displays the chi-squared statistic (instead of z) for the test of significance of the pooled effect size. This is available only for odds ratios pooled using Peto or Mantel-Haenszel methods. Options for continuous data ^cohen^ pools standardized mean differences by the method of Cohen (the default). ^hedges^ pools standardized mean differences by the method of Hedges. ^glass^ pools standardized mean differences by the method of Glass. ^nostandard^ pools unstandardized mean differences. ^fixed^ specifies a fixed effect model using the inverse variance method (the default). ^random^ specifies a random effects model using the DerSimonian & Laird method. General output options for ^metan^ -------------------------------- ^ilevel()^ specifies the significance level (eg 90,95,99) for the individual trial confidence intervals. ^olevel()^ specifies the significance level (eg 90,95,99) for the overall (pooled) confidence intervals. ^ilevel^ and ^olevel^ need not be the same, and by default are equal to the significance level specified using the ^set level^ command. ^sortby()^ sorts by given variable(s). ^label(^ [^namevar^=namevar] [^,yearvar^=yearvar] ^)^ labels the data by its name, year or both. However, neither option is required. For the table display the overall length of the label is restricted to 16 characters. ^nokeep^ denotes that Stata is not to retain the study parameters in permanent variables (see below). ^notable^ prevents display of table of results. ^nograph^ prevents display of graph. Graphical display options for forest plot in ^metan^ -------------------------------------------------- ^xlabel()^ defines x-axis labels. ^force()^ forces the x-axis scale to be in the range specified in ^xlabel()^. ^boxsha()^ controls box shading intensity, between 0 and 4. The default is 4, which produces a filled box. ^boxsca()^ controls box scaling, which by default is 1. ^texts()^ specifies font size for text display on graph. The default size is 1. ^saving(^filename^)^ saves the forest plot to the specified file. ^nowt^ prevents display of study weight on graph. ^nostats^ prevents display of study statistics on graph. ^nooverall^ prevents display of overall effect size on graph (automatically enforces the ^nowt^ option). ^t1(..)^, ^t2(..)^, ^b1(..)^ adds titles to the graph in the usual manner. Note that for graphs on the log scale (that is, ORs or RRs), values outside the range [10e-8,10e8] are not displayed. A confidence interval which extends beyond this will have an arrow added at the end of the range; should the effect size and confidence interval be completely off this scale, they will be represented as an arrow. By default, ^metan^ adds the following new variables to the data set: ^_ES^ Effect size (ES) ^_seES^ Standard error of ES ^_LCI^ Lower confidence limit for ES ^_UCI^ Upper confidence limit for ES ^_WT^ Study weight ^_SS^ Study sample size Options for ^funnel^ ------------------ If the ^funnel^ command is invoked following ^metan^ with no parameters specified it will produce a standard funnel plot of precision (1/SE) against treatment effect. Addition of the ^noinvert^ option will produce a plot of standard error against treatment effect. The alternative sample size version of the funnel plot can be obtained by using the ^sample^ option (this automatically selects the ^noinvert^ option). Alternative plots can be created by specifying ^precision_var^ and ^effect_size^. If the effect size is a relative risk or odds ratio, then the ^xlog^ option should be used to create a symmetrical plot. All options for graph are valid. Additionally, ^sample^ denotes that the y-axis is the sample size and not a standard error. ^noinvert^ prevents the values of the precision variable from being inverted. ^ysqrt^ represent y-axis on square root scale. ^overall(^x^)^ draw a dashed vertical line at overall effect size given by x. Options for ^labbe^ ----------------- By default the size of the plotting symbol is proportional to the sample size of the study. If weight is specified the plotting size will be proportional to the weight variable. All options for graph are valid. Additionally, the following two options may be used. ^nowt^ declares that the plotted data points are to be the same size. ^percent^ display the event rates as percentages rather than proportions. One note of caution: depending on the size of the studies, you may need to rescale the graph (by means of the ^psize()^ option) Examples -------- . ^metan tdeath tnodeath cdeath cnodeath, or chi2 label(namevar=trialid)^ . ^metan tdeath tnodeath cdeath cnodeath, rd random^ . ^metan tdeath tnodeath cdeath cnodeath, olevel(99) ilevel(95) sortby(year)^ ^label(namevar=trialid, yearid=year) nostats nowt^ . ^local ovratio=$S_1^ . ^funnel , sample ysqrt xlabel(0.1,0.5,1,5,10) ylabel(0,500,1000)^ ^overall(`ovratio')^ . ^labbe tdeath tnodeath cdeath cnodeath, xlabel(0,0.25,0.5,0.75,1)^ ^ylabel(0,0.25,0.5,0.75,1)^ . ^metan n1 m1 sd1 n2 m2 sd2, nostandard^ . ^metan n1 m1 sd1 n2 m2 sd2, random xla(-2,-1,0,1,2) force saving(metagrph)^ Authors ------- Michael J Bradburn Institute of Health Sciences, Oxford m.bradburn@@icrf.icnet.uk Jonathan J Deeks Institute of Health Sciences, Oxford j.deeks@@icrf.icnet.uk Douglas G Altman Institute of Health Sciences, Oxford d.altman@@icrf.icnet.uk Also see -------- STB: STB-44 sbe24 On-line: help for @meta@ (if installed), @metacum@ (if installed), @metareg@ (if installed), @metabias@ (if installed)