DIFFERENCES IN THYROID UPTAKE PERCENTAGE BASED ON COUNTING STATISTICS

Introduction: The thyroid probe is an important piece of nuclear medicine instrumentation, which measures thyroid uptake percentage. Thyroid uptake measurements are an important diagnostic element for diagnosing hyperthyroidism. Nuclear medicine thyroid uptake measurements are a diagnostic procedure used for determination of hyperthyroidism. The interpretation of the patient’s thyroid uptake as normal or abnormal is determined by comparing the uptake value to a set normal range for the facility and country where the study is performed. These studies are commonly considered during a diagnostic work-up for hyperthyroidism. Due to the fairly significant occurrence of hyperthyroidism and its effects on patients, it is important that patients are diagnosed correctly. Additionally, there is currently not a specific guideline provided for measuring a patient’s thyroid uptake. Our objective was to determine whether there are differences in thyroid uptake percentage when taking one neck measurement as opposed to taking two neck measurements and averaging the two measurements.

Methods: After OUHSC IRB approval (#14855), we acquired data from thyroid uptake studies performed in the last 5 years. A total of 66 sets of thyroid count data were collected. All data were deidentified including the standard pill count, patient background, first neck count, second neck count, and averaged neck count. Thyroid uptake was computed two ways: (1) using the average of two thyroid counts and (2) randomly selecting and using only a single count. After testing for normality assumptions, average thyroid uptake count between the counting methods were compared using a Wilcoxon sign rank tests. Additionally, multiple Monte Carlo simulation studies were performed using average patient count data from the original patient data study, and each simulation different count variability scenarios based on Chi-square testing. 5000 Monte Carlo data sets (n=66 each) were created for each variability scenario, and uptake percentage was computed using two randomly generated count values and only one randomly generated count value. The average difference between thyroid uptake percentage was computed across all 5000 Monte Carlo data sets. This difference was compared to the null value of zero using a non-parametric Wilcoxon sign rank test. All statistical tests assumed a 5% chance of a type one error, using SAS 9.4.

Results: The patient data study showed the thyroid uptake percentage did not differ whether only one count or the average of two counts were used as is demonstrated in figure 1. The same results were demonstrated by the Monte Carlo simulations under the given count variability scenarios.

Conclusions: There was no statistically significant difference between the two counting methods. A limitation of note for this study is that there was limited count data as not all patients at this facility had two thyroid uptake measurements performed.

• Download figure
• Open in new tab
• Download powerpoint