Before an analysis can be performed, you must have completed these steps:
To add an analyses
Right-click Analyses under Data and Analyses in the outline and choose Add Analysis.
The New Analysis Wizard opens.
Enter a Name for the analysis.
Click Next.
Select the analysis type and any tests to include.
Click Finish.
After you have specified the name and basic properties of an analysis, you can provide details of the analysis using the Analysis properties.
To further specify or modify an analysis
Right-click on a specific analysis under the Analyses heading in the outline and choose Properties
The Analysis properties window opens.
Edit the information in the four tabs:
General for editing the name of analyses, the type of test analyses, the data that will be used in the analyses, and the confidence interval used for individual study estimates of sensitivity and specificity.
SROC plot for editing the feature of SROC plot, e.g. to define a scale of size of points in SROC plot, weights of analysis, and plotting options..
Forest plot for editing the feature of the forest plot. You can add risk of bias and applicability items and covariates on the forest plot and sort studies within the forest plot.
Sources of heterogeneity for specifying subgroup analyses.
Click OK.
Single test analysis is for analyzing the accuracy of a single index test.
Multiple test analysis is for analyzing and visualizing the accuracy of two or more index tests.
The Analyze Paired Data Only option can only be used for multiple test analyses. This option allows you to restrict the analyses to those studies which have reported data on a pair of index tests.
Select Investigate Sources of Heterogeneity if you want to examine accuracy by subgroups of studies.
Under Tests, you can select which data tables to include in the analysis.
You can specify the characteristics of the SROC analysis.
Display SROC curves determines whether or not to draw a SROC curve.
Axis Off means that the normal axes are removed, the scale markers are moved a little away from the graph and the background of the graph area is colored very light grey.
Display Study Points controls whether the pairs of sensitivity and specificity of individual studies will be displayed in the ROC plot.
Display CI on Study Points controls whether confidence intervals will be displayed as lines through each study point in the ROC plot.
Symmetry determines whether to draw a SROC curve that is symmetrical around the line running from the upper left to lower right corner of the SROC graph (=constant odds ratio model) or asymmetric (=odds ratio is allowed to vary with threshold).
Use Weight for Analysis to specify the type of weight to be used in the SROC analysis.
Use Scale for Size of Points to specify whether individual study points are displayed with equal size markers or markers where their sizes reflect either differences in sample size, inverse standard error. You can also have different markers according to the value of a covariate. If you have many overlapping points, you can use Percentage Scaling for All Points to reduce all point sizes by the same percentage.
Display Paired Data Lines controls whether lines are drawn between paired data points when Analyze Paired Data Only is selected. If enabled, you can also choose a line style and colour.
A forest plot provides a visual overview of the results of individual studies that will go into the meta-analyses. It shows estimates of sensitivity or specificity of individual studies together with their confidence intervals. It provides a simple representation of the amount of variation among results of different studies.
The following characteristics of a forest plot can be changed:
Quality Items Displayed on Forest plot – To add information about specific quality items of individual studies to the forest plot.
Covariates Displayed on Forest plot – To add information about certain covariates of individual studies to the forest plot.
Sort Study by – You can determine the ranking of individual studies within the forest plot. You can rank studies according to various factors, including the height of sensitivity or specificity, year of publication, quality item or any other covariate you have specified.
You can examine whether the accuracy of studies differ by certain characteristics of studies. A stratified analysis of studies will be performed showing whether accuracy of studies differ by subgroups of studies. No formal testing such as p-values will be calculated in this version of RevMan to determine whether results are significantly different. RevMan will show separate ROC curves per subgroup.
Subgroups of studies can be based on quality items or based on any other specified covariate (using all, or only a subset of the covariate values).
To create an analysis showing subgroups in ROC plane
Add a number of covariates using categorical data .
See Covariates.
Open the data table for the test.
See Entering data.
Assign the categories.
There is a column for each covariate.
Create a new analysis, set to use the Investigate Sources of Heterogeneity option.
Open the Properties for the analysis and go to the Sources of Heterogeneity tab.
Select the Covariate option, and use the pull down list to choose a covariate.
If you only want to use some of the covariate values, deselect the ones to exclude.
Click Apply.
The unfilled symbols represent individual study estimates, solid circles are the summary points (meta-analytical estimates of sensitivity and specificity), solid lines are the summary HSROC curves, dotted lines surrounding the summary points are x% confidence regions, and the dashed lines around these points are the x% prediction regions.
RevMan automatically chooses a colour and symbol to use for each test included in an analysis. But you can also choose you own custom options for each test in the Analysis content pane.
You can use a custom Specificity Range to control the specificities over which the SROC curve will be drawn.
In the Analysis content pane there is also the option to make additional figures based on the results of more complex models, like the HSROC and Bivariate model (see the Cochrane Handbook for Systematic Reviews of Diagnostic Test Accuracy). These more complex models can not be fitted in RevMan, but you can perform these analyses in an external statistical package and then import the results into RevMan.
You choose a model for creating a corresponding confidence ellipse around summary point. There are two models:
1. HSROC Model
Here you can enter the five parameters of the HSROC model:
Lambda – accuracy parameter.
Theta – cut-point parameter.
Beta– shape parameter.
Var(accuracy) – variance of accuracy parameter.
Var(threshold) – variance of threshold parameter.
These parameters can be calculated using BUGS software, or in SAS using Proc NLMIXED. Software code for BUGS to fit this model can be found in Rutter et al (2001), and SAS code for fitting the model is available in the appendix of Chapter 10 of the Handbook. Based on these parameters a (new) summary ROC curve and a summary point will be drawn.
2. Bivariate Model
You can enter the five parameters of the bivariate model:
E(logitSe) – expected mean value of logit transformed sensitivity.
E(logitSp) – expected mean value of logit transformed specificity.
Var(logitSe) – between-study variance of logit transformed sensitivity.
Var(logitSp) – between study variance of logit transformed specificity.
One of the following:
Cov(logits) – covariance between logit transformed sensitivity and specificity, or
Corr(logits) –correlation between logit transformed of sensitivity and specificity
These parameters can be calculated using SAS or STATA software. Details of the bivariate model can be found in Reitsma et al. (2005) and in Chapter 10 of the Handbook.
References:
1. Reitsma JB, Glas AS, Rutjes AWS, Scholten RJPM, Bossuyt PM, Zwinderman AH. Bivariate analysis of sensitivity and specificity produces informative summary measures in diagnostic reviews. Journal of Clinical Epidemiology. 2005;58(10):982-90.
2. Rutter CM, Gatsonis CA. A hierarchical regression approach to meta-analysis of diagnostic test accuracy evaluations. Stat Med. 2001;20(19):2865-84.
Creating confidence and prediction regions
Based on the results of the bivariate model, you can also create a confidence region around the summary sensitivity and specificity as well as a prediction region (the region likely to contain the sensitivity and specificity of a new individual study). The following parameters need to be imported into RevMan:
SE(E(logitSe)) – standard error of the expected mean logit transformed sensitivity.
SE(E(logitSp)) – standard error of the expected mean logit transformed of specificity.
Cov(Es) – covariance between expected mean logit sensitivity and specificity
Studies– the number of studies in the analysis.
2. Click one or more of the following check boxes:
To obtain a summary curve, click check box in font of “Display summary curve”.
To obtain a summary point, click check box in font of “Display summary point”.
To obtain a X% confidence region (options are 90%, 95% or 99%, the default is 95%), click the check box in front of “Display X% confidence region”.
To obtain a X% prediction region (options are 50%, 90% or 95%, the default is 95%), click the check box in front of “Display X% prediction region”.