simul_confusion/Modules/ado/plus/m/metafunnel.hlp

{smcl}
{* 19aug2003}{...}
{hline}
help for {hi:metafunnel} {right:dialog:  {dialog metafunnel}{space 20}}
{hline}

{title:Funnel plots for meta-analysis}

{p 8 15 2}
{cmd:metafunnel}
{it:{ theta { se | var } | exp(theta) ll ul [cl] }}
[{cmd:if} {it:exp}]
[{cmd:in} {it:range}]
[{cmd:,}
{cmdab:ef:orm}
{cmdab:nol:ine}
{cmdab:rev:erse}
{cmdab:forc:enull}
{cmdab:graph_options}
]


{title:Description}

{p 4 4 2}
{cmd:metafunnel} plots funnel plots. These graphical displays are used to
examine whether the results of a meta-analysis may have been affected by
publication or other types of bias.


{title:Options}

{p 4 8 2}
{cmd:by(}{it:byvarname}{cmd:)} displays subgroups according to the value
of {it: byvarname}. The legend
displays the value labels for the levels of {it: byvarname} if these are 
present; otherwise it displays the value of each level of{it: byvarname}. 

{p 4 8 2}
{cmd:eform} exponentiates the treatment effect theta and displays the
horizontal axis (treatment effect) on a log scale. This is useful for
displaying ratio measures such as odds ratios and risk ratios.

{p 4 8 2}
{cmd:reverse} inverts the funnel plot so that larger studies are displayed
at the bottom of the plot with smaller studies at the top.
This may also be achieved by specifying {cmd:noreverse} as part of the
{cmd: yscale(}{it:axis_description}{cmd:)} graphics option.

{p 4 8 2}
{cmd:noline} specifies that pseudo 95% confidence interval lines
should not be included in the plot.  The default is to include them.

{p 4 8 2}
{cmd:forcenull} forces the vertical line at the centre of the funnel
to be plotted at the null treatment effect of zero (1 when the treatment
effect is exponentiated). The default is for the line to be plotted at
the value of the fixed effect summary estimate.

{title:Remarks}

{p 4 4 2}
Funnel plots are simple graphical displays of a measure of study size
on the vertical axis against intervention or treatment effect on the 
horizontal axis. The name "funnel plot" is based on the fact that the
precision in the estimation of the underlying intervention or treatment
effect will increase as the size of component studies increases. Results
from small studies will therefore scatter more widely, with the spread
narrowing among larger studies. In the absence of bias the plot will
resemble a symmetrical inverted funnel.

{p 4 4 2}
If there is bias, for example because smaller studies showing no
statistically significant effects remain unpublished, then such
publication bias will lead to an asymmetrical appearance of the funnel
plot. It should be noted that although funnel plots have traditionally
been used to examine evidence for publication bias, funnel plot asymmetry
may reflect other types of bias, or even result from the true intervention
or treatment effect differing between small and large studies. They should
thus be seen as displaying the evidence for "small study effects" in general
rather than publication bias in particular. These issues are discussed by
Egger {it: et al.} (1997) and Sterne {it:et al.} (2001a).

{p 4 4 2}
{cmd:metafunnel} uses the same syntax as other meta-analysis commands
such as {help meta}, {help metabias}, {help metainf} and {help metatrim}.
The user provides an estimate of the treatment or intervention effect,
{it: theta}, together with its associated standard error {it: se_theta}
(the default) or variance {it: var_theta}, (in which case the {cmd: var}
option should be specified). Alternatively, the user provides {it:exp(theta)}
(i.e., a risk ratio or odds ratio) and its confidence interval,
{it: (ll, ul)}).

{p 4 4 2}
The funnel plots are displayed in line with meta-analytic convention and
the recommendations of Sterne et al. (2001b). The effect of the 
treatment or intervention in each study (horizontal axis) is plotted
against the study size, as measured by the standard error of the treatment
or intervention effect.
(vertical axis). The vertical axis is reversed so that larger studies are
displayed towards the top of the graph (this behaviour may be changed using
the {cmd: reverse} option). Users who wish to plot the treatment effect on the vertical axis should use the {cmd:graph(begg)} option of
the {help metabias} command. The {help funnel} command, which is part
of the {help metan} package, also provides an alternative way to draw
funnel plots.

{p 4 4 2}
The plots include pseudo-95% confidence interval lines, which are
drawn around the summary fixed-effect estimate of the intervention or
treatment effect. The lines may be omitted using the {cmd:nolines} option.
The user may also specify that the pseudo confidence limits are centred
around a zero intervention effect using the {cmd:forcenull} option.

{p 4 4 2}
When the {cmd:eform} option is used, the label of the horizontal axis
(treatment effect, {it:theta}) is 
changed accordingly, unless there is a variable label for {cmd: theta} or
the {cmd: xtitle(}{it:axis_title}{cmd:)} graphics option is used.

{p 4 4 2}
By default, the subtitle "Funnel plot with pseudo 95% confidence limits"
is displayed. This may be changed using the graphics option
{cmd:subtitle(}{it:tinfo}{cmd:)}.


{title:Examples}

{p 4 8 2}{cmd:. metafunnel meandiff semeandiff}

{p 4 8 2}{cmd:. metafunnel logor selogor, eform xtitle("Odds ratio (log scale)")}

{p 4 8 2}{cmd:. metafunnel sttd stderr, by(dose) subtitle(Funnel plot with subgroups) forcenull}

{p 4 8 2}{cmd:. metafunnel logor varlogor, var reverse nolines xtitle(log odds ratio)}



{title:Acknowledgements}

{p 4 4 2}
{cmd:metafunnel} was written by Jonathan Sterne, University of Bristol.
Portions of the code were originally written by Tom Steichen, who also
gave helpful comments on an early version of the command and provided
the dialog.


{title:References}

{p 4 8 2}Egger, M., Davey Smith, G., Schneider, M., & Minder, C. (1997).
Bias in meta-analysis detected by a simple, graphical test.
{it:British Medical Journal} 315: 629-634

{p 4 8 2}Sterne, J.A.C., Egger, M. and Davey Smith, G. (2001a).
Investigating and dealing with publication and other biases
in meta-analysis. {it: British Medical Journal} 323: 101-105

{p 4 8 2}Sterne, J.A.C. & Egger, M. (2001b). Funnel plots for detecting
bias in meta-analysis: guidelines on choice of axis.
{it: Journal of Clinical Epidemiology} 54: 1046-1055


{title:Also see}

{p 4 13 2}
Online:  help for {help meta}, {help metabias},
{help metainf}, {help metatrim}, {help metan}, {help funnel} (if installed)
{p_end}