383 [18F]FDG-PET/CT: cost-utility analysis alongside an RCT 6 (e.g., utilities concerning surgery for benign disease and concurrent complications); others could be disadvantageous to the [18F]FDG-PET/CT-driven group (e.g., utilities concerning active surveillance of [18F]FDG-negative nodules). For the probabilities, the higher means were typically disadvantageous to the [18F]FDG-PET/CT-driven group (e.g., the probability of surgery for benign lesion after continued surveillance for [18F]FDG-negative nodules or the probabilities of complications due to thyroid surgery beyond the first year). Altogether, we believe the use of asymmetric triangular distributions was likely disadvantageous to the [18F]FDG-PET/CT-driven group and may have underestimated its cost-effectiveness, which was nevertheless more favorable than in the diagnostic surgery group. As a Markov model remains a simplified reflection of the real situation, this is a limitation of any model-based cost-utility analysis and thus also applicable to the current study. The accuracy of the estimated probabilities, costs, and utilities are dependent on the availability and quality of representative source data. Although we performed a comprehensive literature search to ensure a careful, evidence-based determination of all model parameters, the best fitting literature for some variables was only moderately related. In these cases, an expert panel was additionally consulted. This included all parameters concerning the active surveillance of [18F]FDG-negative indeterminate thyroid nodules, for which we had to rely on literature about benign nodules and expert opinion. For example, a disutility of 0.02 was assigned to the active surveillance health state. We chose a limited but conservative disutility as compared to the disutility of observation after an uncomplicated HT for a benign nodule (0.01) to prevent overestimation of HRQoL in favour of an [18F]FDG-PET/CT-driven workup and to account for any suspense of not knowing a definite histopathological diagnosis. Patients under surveillance may experience some degree of cyclic psychological distress centring around their yearly follow-up visits, although evidence supporting that assumption is currently lacking and we have not observed it in the EfFECTS trial [495]. A recent study with a limited median 15-month follow-up found no evidence of such effects and showed sustained HRQoL in patients under surveillance following a negative molecular test [580]. We included the disutilities of both observation after HT and observation after a negative [18F]FDG-PET/CT scan in our univariate sensitivity analysis. Although these disutilities did affect the incremental net benefit, [18F]FDG-PET/ CT remained the cost-effective strategy across the tested ranges. In conclusion, the current cost-utility study showed that an [18F]FDG-PET/CT-driven diagnostic workup reduced the 1-year thyroid nodule-related and societal costs while sustaining quality of life. Following the observed reduction in diagnostic surgery for benign nodules, an [18F]FDG-PET/ CT-driven workup is very likely cost-effective from a Dutch societal perspective as compared to diagnostic surgery for Bethesda III/IV nodules.
RkJQdWJsaXNoZXIy MTk4NDMw