Possible bias estimate outcomes of Clinical and Laboratory Standards Institute guideline EP09-A3. Outcome A: no significant bias; outcome B: the bias is acceptable; outcome C: the bias is acceptable, but an alternate conclusion that an inappropriate percentage of sample results will have bias outside the limits is also possible; outcome D: the bias is acceptable, but there is less confidence in this conclusion than for outcome C; outcome E: the performance of the candidate measurement procedure is not acceptable for the defined application. Abbreviations: MD, medical decision level; CI, confidence interval.