USING BEST-WORST SCALING TO ELICIT PATIENT PREFERENCES FOR SYMPTOM CONTROL, SIDE-EFFECTS AND PROCESS CHARACTERISTICS OF TREATMENTS IN PARKINSON'S DISEASE

Purpose:

   The objective was to use Best-Worst Scaling (BWS) to elicit treatment preferences for symptom control, side-effects and process characteristics of various treatments used in the management of motor symptoms in Parkinson’s disease.

Method:

   From interviews with patients and neurologists seven attributes were selected which were most representative and distinctive between current treatments (treatment modality, tremors, slowness of movement, posture and balance problems, fatigue, dizziness, and dyskinesia). A BWS case 2 experiment was set up,  in which patients had to indicate which aspects they perceived as the best and the worst in nine hypothetical treatment alternatives. Additionally, a BWS case 1 experiment elicited patient preferences over the best and worst features of nine sets of process characteristics of neurosurgery, pump infusion, and oral intake of mediation.  In analysis, the questionnaires with responses errors were excluded and subsequently Best-minus-Worst scores, conditional logit analysis, attribute importance (BWS case 2), and subgroup analysis were calculated or executed.

Result:

   In total, 255 questionnaires were submitted:  224 could be included in BWS case 1 analysis and 192 in BWS case 2 analysis. Older and lower educated patients had significantly more response errors. According to patients, treatment modality had the greatest impact on the value of treatment (25% attribute importance), followed by two attributes related to treatment efficacy: posture and balance problems (16%) and slowness of movement (14%). Side-effects of treatment were considered to be the least important. Subgroup analysis showed that for patients who receive or have received advanced treatments, treatment modality loses its negative impact. BWS case 1 revealed that all patients valued the characteristics of oral intake of medication as least burdensome compared to all process characteristics of pump infusion and brain surgery.

Conclusion:

   This was the first study in the field of Parkinson’s disease which informs on the relative desirability of treatment outcomes with respect to each other. For most patients their favored treatment consists of oral medication. Any treatment requiring neurosurgery has a considerably lower value for which a higher effectiveness does not compensate in most cases. Our results raise the question whether patients who undergo brain surgery have different preferences than the average patient before choosing this treatment, or whether experience with a treatment increases its perceived value (or reduces the disutility of negative aspects).