OBJECTIVES: To develop and test a novel approach to estimate the ICER via a new bootstrap approach based upon the sample coefficient of variance and optimized via computational intelligence. METHODS: A novel bootstrap ICER estimation approach was developed that incorporated the sample coefficient of variance to better capture information within cost-effectiveness data.  In this derivation, an optimal design value parameter was also obtained via computational intelligence.  Across illustrative cost and outcome correlation structures and sample sizes, a simulation study of 1111 replications with 999 bootstrap resamples each was conducted utilizing MatLab R2012b.  Comparative results of point estimates versus the existing bootstrap method were presented as relative efficiencies, with 95% confidence intervals (CI) presented as coverage probability, coverage error, length, left and right bias, and relative bias. RESULTS: The proposed ICER yielded less statistical estimation error than the typical bootstrap approach across all cases, with the relative efficiency of point estimates ranging from +106.03% to +113.35%.  An equal or improved coverage error for the CI was also consistently achieved, deviating from the population value by zero (i.e., perfect coverage) to 0.0200 versus from 0.0060 to 0.0210.  Subsequently, an improved shortening of the CI length was noted.  The relative bias suggested slightly more left bias and less right bias across both positive and negative cost and outcome correlation structures, reaching a maximum of 0.5238 for the proposed ICER versus 0.2222 for the usual bootstrap. CONCLUSIONS: This novel method to estimate the ICER via the sample coefficient of variation found improvements in the relative efficiency of point estimators and in the coverage error and length of the 95% CI across all simulated cases.  Irrespective of cost and outcome correlation structure, the relative bias of this ICER suggested a slight increase in potential left-sided bias and decrease in right-sided bias versus the usual bootstrap.