Reducing noise in radioactive iodine activity selection: the utility of an online clinical calculator

Noise, an unwanted variability in judgment, is ubiquitous in medicine, including in the prescription of radioactive iodine (RAI). Building upon our recently developed predictive risk model, we created an online clinical support tool to facilitate the translation of our model into clinical practice. The aim of this study is to assess the utility of an online clinical support tool to reduce noise in the treatment for patients with differentiated thyroid cancer (DTC).

The tool was accessible via weblink or a QR code. Activity recommendations were applied to the calculator's four risk categories: 0 GBq for very low risk, 1 GBq for low risk, 4 GBq for intermediate risk, and 6 GBq for high risk. The tool was applied prospectively to 103 patients who received RAI at Royal North Shore Hospital between 2021 and 2022 and retrospectively to 393 patients treated with RAI between 2017 and 2021.

A significant difference was observed in administered activity between the 2021-2022 and 2017-2021 cohorts in patients stratified as intermediate risk (median activity 3.95 GBq, interquartile range 2.03-4.04 vs 4 GBq, 4-4) and high risk (4.07 GBq, 3.95-5.7 vs 6 GBq, 6-6) with P-values of 0.01 and <0.01, respectively. No difference was seen in low-risk patients (2.01 GBq, 1.03-3.98 vs 1 GBq, 1-4, P = 0.30). Additionally, no clinically significant recurrence was observed between the two cohorts (6.6% vs 4.5%; P = 0.628).

Optimal risk classification and activity recommendation continue to be established. Our data suggest that providing risk stratification and activity recommendation in an easy-to-access online tool can reduce noise and variability in activity prescription for patients with DTC.