Acceptability Curves and Confidence Ellipses

Acceptability curves and confidence ellipses are both methods for graphically presenting the uncertainty surrounding the estimate of cost-effectiveness. A confidence ellipse provides a visual representation of the region containing x% (where x is usually 95) of the uncertainty. An acceptability curve provides a graphical representation of the probability that an intervention is cost-effective compared with the alternative(s), given the data. Confidence ellipses can only be used for comparisons between two interventions, whereas acceptability curves can be produced for decisions involving multiple interventions. Confidence ellipses are determined parametrically from information about the distribution of costs and effects (mean, variance, and covariance). The acceptability curve can be determined from the confidence ellipse or direct from the data following an assessment of uncertainty through bootstrapping (for trial data) ...