A priori considerations when using laboratory determinations in cost-effectiveness and clinical decision analyses.

The incorporation of laboratory determinations into cost-effectiveness and clinical decision analyses can be weakened by failure to consider various sources of error in the generation and utilization of laboratory data. To improve these analyses, three phases in the process that converts a patient specimen into patient treatment are identified. Phase I examines quality of methodology and laboratory performance and identifies possible errors in generation of the analyte value. Phase II considers the process by which this value is compared to normal values; possible errors involved in normal value selection and the problem of false positives and false negatives are discussed. Phase III involves the interpretation of the values; possible errors in this phase are discussed. Recommendations are made to explicitly identify sources of error in all phases and thus strengthen decision analysis involving laboratory data.