Jj. Bailey et al., Dysrhythmia hazard after hospitalization for myocardial infarction: Two ECG prognostic methods compared, J ELCARDIOL, 33, 2000, pp. 151-154
We retrieved reports of heart rate variability and signal-averaged electroc
ardiograms (SAECG) used to predict risk of a dysrhythmic event. From each r
eport the number of cases with and without events was extracted to establis
h accurate values for true positive rate (tpr = sensitivity) and false posi
tive rate (fpr = 1 minus specificity). For all the heart rate variability r
eports, these values were collected and tpr values were plotted versus fpr.
The (fpr,tpr) data were summarized by a meta ROC graph using the method of
Moses and Shapiro. A composite weighted mean value and 95% confidence inte
rval were also derived. A summary meta-ROC curve for thc SAECG reports was
similarly obtained., Meta-ROC analysis of multiple reports better summarize
s the performances of different prognostic methods and allows the effect of
combining tests for a larger population to be simulated.