Accuracy considerations when using early (four- or six-hour) radioactive iodine uptake to predict twenty-four-hour values for radioactive iodine dosage in the treatment of Graves' disease.

Although literature has offered methods to predict 24-hour radioactive iodine uptake values from early (4- to 6-hour) measurements, the resultant dosage errors have not been examined. Potential errors include underdosage, overdosage, and a failure to recognize rapid turnover patients (early-to-late uptake ratios > or = 1) who are at high risk for treatment failure and full-body radiation exposure. We developed and tested a novel method for minimizing error involved in using a single early uptake measurement to derive late uptake. From a retrospective analysis of 203 Graves' disease patients, receiver operating characteristic (ROC) curve analysis enabled us to identify patients likely to experience rapid turnover and therefore should receive 24-hour studies. Twenty-four-hour uptake measurements are necessary with 77% or more 4-hour uptake values and 80% or more 6-hour values. After eliminating these patients, we developed linear regression equations to predict the 24-hour uptake from 4-hour (n = 61) and 6-hour (n = 22) rule groups, testing their efficacy on separate 4-hour (n = 61) and 6-hour (n = 21) patient groups. We also used our test population to measure error in four early-to-late uptake conversion formulas presented in the literature. Error involved in these predictions ranged from a 10.6% overestimate for 4-hour calculations to a 5.9% underestimate for 6-hour calculations. When applied to two dosage formulas incorporating gland size, absorbed dose, and 24-hour uptake, average dosage error was 7%. In comparison to the other sources of error radioactive iodine (131I) dosimetry, potential error in predicting 24-hour uptake from 4- or 6-hour uptake values is low.