The objective of this dissertation was to investigate the influence of process-utility (which is the value attached to the process and convenience of care without reference to the outcome) on the relative value of treatments in Parkinson’s Disease (PD). The main treatment modalities (processes) in PD are oral intake of medication, continuous pump infusion of medication or brain surgery. This thesis clearly shows that the process of care is an important driver of preferences, in both patients with PD and in the general public. Where oral intake medication has a positive effect on the value of treatment, pump infusion and brain surgery have a detrimental effect. Furthermore, the occurrence of side effects (dizziness, drowsiness, and dyskinesia) had less impact on the perceived desirability of treatment than the treatment’s effect on motor symptoms (slow movement, posture and balance problems and tremor). In this thesis, process-utility was estimated using different techniques such as best-worst scaling, time trade-off and visual analogue scales. Results have shown all methods perform equally well with regard to distinguishing treatment alternatives on the basis of aggregated process-utility scores. However, on an individual patient-level, such as in shared decision making, the composed nature of Best-Worst Scaling allows for better prioritization of treatment alternatives. In contrast to its use in societal decision making, shared decision making requires that process characteristics are explicitly discussed with patients. What, however, is at issue here is how preferences (process-utility) should be measured. The use of stated preference methods with a strong theoretical foundation is not common in clinical decision making. Therefore the second part of this thesis focused on the use of best-worst scaling case 2 as a value clarification method to help individual patients clarify their values in treatment decisions. This thesis has found that although best-worst scaling has potential in shared decision making, methodological issues (e.g. reliability of individual preference estimations) should first be investigated before widespread implementation can take place.
|Qualification||Master of Science|
|Award date||1 Jun 2017|
|Place of Publication||Enschede|
|Publication status||Published - 1 Jun 2017|