Adjusting for bias in C/E ratio estimates.

The estimator used to calculate incremental cost-effectiveness (C/E) ratios from sampled data is biased but consistent. While the bias may be negligible in studies with large sample sizes, it is potentially important in analyses based on small samples. When patient-level data on costs and effects are available, bootstrap simulation methods can be used to estimate the bias of a C/E ratio and adjust the point estimate accordingly.