FINDING THE OPTIMAL MULTIPLE-TEST STRATEGY USING A METHOD ANALOGOUS TO LOGISTIC-REGRESSION - THE DIAGNOSIS OF HEPATOLENTICULAR DEGENERATION(WILSONS-DISEASE)

Finding the optimal strategy among a battery of tests may be cumbersom e for decision-analytic models. The authors present a method of examin ing multiple test combinations that is based on a modified Bayes' form ula analogous to logistic regression. They examined all 16 combination s of four tests used to diagnose hepatolenticular degeneration. The fo ur tests examined were: serum ceruloplasmin level, 24-hour urinary cop per excretion, free serum copper level, and liver biopsy copper concen tration. They also simulated the diagnostic workup of the disease for a hypothetical cohort of 15,000 patients at risk. Assuming the disutil ities of false positives and false negatives to be equal, and consider ing sensitivity analysis of test characteristics, the following test c ombinations were found to be optimal for making the diagnosis at a pri or probability of disease equal to 0.05: positive serum ceruloplasmin and 24-hour urinary copper excretion, combined with either positive li ver biopsy or free serum copper (or both). The strategies obtained by the modified Bayes' formula were the same as those found using the sim ulated data set with a standard logistic-regression software package. The logistic model's diagnostic accuracy is 0.98 as measured by the ar ea under the receiver operating characteristic curve. The optimal stra tegy for diagnosing hepatolenticular degeneration varies with the prio r probability of disease. For prior probabilities of 0.05, 0.25, and 0 .9, and the optimal strategy, model sensitivities are 0.801, 0.880, an d 0.997, and model specificities are 0.991, 0.985, and 0.814, respecti vely. The new method provides a convenient alternative to decision tre es when examining multiple diagnostic tests.