General discussion 521 13 management and potentially patient outcomes. As such, patients with compressive symptoms may benefit more from immediate thyroid surgery than from additional preoperative diagnostics and postponed surgery. This should be considered during preoperative patient counselling. Challenges in cost-utility studies The setup of a cost-utility study knows many challenges and requires highly careful methodology. There are dozens of national guidelines for economic evaluations of healthcare interventions with a reasonable degree of international consensus with regard to the type of analysis, perspective, time horizon, target population, comparator, included costs, health outcome measure (i.e., QALYs), and the requirement of an uncertainty analysis (i.e., sensitivity analysis), among others [776]. For example, it is recommended that the time horizon is based on the natural course of the disease and expected EfFECTS of the index intervention. By many national guidelines, including the Dutch one, a lifelong time horizon is preferred [477, 776, 777]. The recommended perspective for the analysis, however, may differ based on the type of health care system: in the American health care system, a payer perspective that only assesses the costs and EfFECTS on the specific patient (who is likely also the payer) may be favoured over a societal perspective which describes all costs and EfFECTS of the intervention on society as a whole (preferred in the Dutch system), the publicly funded healthcare payer perspective, or the third-party payer perspective [178]. The recommended type of costs to be included may depend on the chosen perspective. From a societal perspective, it is advised to include both direct medical costs, non-medical costs (e.g., patient and family costs), and indirect costs (e.g., productivity losses), whereas including direct medical and non-medical costs may suffice for a study from a payer perspective [477, 776, 777]. In the next paragraphs, I would like to discuss two additional considerations that I observed during the writing of our systematic review (Chapter 2) and the conduction of our cost-utility study (Chapter 6) and that are applicable to cost-effectiveness studies performed from any perspective. First, any parameter that is included in a cost-utility study (i.e., probability, price, or utility) may be of significant impact when extrapolated in a multi-year cost-utility model. As such, each model parameter should be carefully determined and well argued. As previously debated in Chapter 2 [25], the results of previous cost-effectiveness studies of one of the commercially available MD panels illustrate that (even minor) under- or overestimation of any parameter may lead to opposite and possibly inappropriate conclusions regarding cost-utility, even within the same (American) health care system and independent of the results of one-way sensitivity analyses [176, 178, 495, 579]. The validity of the results of a cost-utility study depends on the accuracy of parameter estimates, which in turn depends on the availability and quality of representative source data. When model parameters have to be estimated using expert opinion because no reference values or evidence-
RkJQdWJsaXNoZXIy MTk4NDMw