Cost-effectiveness in nuclear medicine

There are many pressures on nuclear medicine physicians to keep costs down and to produce the most efficacious results for the money spent. A procedure is cost-effective if its benefits (correct diagnosis, effective treatment, etc.) are worth the additional costs (money, radiation dose, etc.). The benefits of diagnostic procedures are measured by test performance (diagnostic efficacy [efficacy-D]), which in turn depends on a specific task: detection, classification, localization, or quantitation. For the detection task, measures of test performance include sensitivity, specificity, predictive value, accuracy, diagnostic utility, posttest versus pretest disease probability, area under a receiver operating characteristic (ROC) curve, and others. Standard measures of test performance in the classification, localization, and quantitation tasks have not yet been developed. Efficacy-M (management efficacy) is a measure of the extent to which a procedure influences patient management appropriately. Efficacy-O (outcome efficacy) is a measure of the success of therapeutic procedures. Some publications illustrating these points are presented. Costs associated with nuclear medicine procedures include money, time, discomfort, possible drug reactions, radiation dose, and the hypothetical risk of radiation-associated cancer. The public's perception of these latter risks, or costs, is much higher than our knowledge would justify. Decision trees are useful for laying out all possible strategies, outcomes, costs, and benefits, and estimating the cost-effectiveness of each strategy. Some articles estimating cost-effectiveness in nuclear medicine are reviewed, although few articles in the literature make any detailed use of decision analysis. Much work remains to be done in forming a coherent, consistent procedure for assessing cost-effectiveness in nuclear medicine.