Ires Ghielen

146 Chapter 8 (10MWT) [39] and the Timed One Leg Stance Test (OLST) [40] were used to assess walking speed and standing balance, respectively. Qualitative evaluation Since the experimental treatment was a new approach, the opinion and experiences of the participating patients were considered very important. Therefore, two patient researchers of the Dutch Parkinson patient association (Parkinson Vereniging) contributed to the study by providing input in the design phase and by conducting a group interview on the patient’s expectations at baseline and evaluation points at post-treatment. These interviews were in the absence of the academic employees and were documented anonymously for qualitative analysis. We thought that the participants felt more comfortable to address critical feedback points anonymously and towards the patient researchers, compared to the researchers. Data analysis The means, standard deviations, and ranges of the descriptive measures were calculated. Differences in demographic, disease related, and outcome measures were investigated using independent t-tests or Mann-Whitney U tests, and chi- square tests where appropriate. Linear mixed model analyses were used to estimate the effect of BEWARE (compared to TAU) on the outcome measures. Besides an overall intervention effect, the separated effects at both post-treatment and follow- up were estimated. For the latter, time and the interaction between condition and time were added to the model. We corrected for baseline differences in all analyses, also when baseline measures did not significantly differ between conditions. The intervention effect was estimated by the regression coefficient (β) and its related p-value. The significance level was set at a p-value of 0.05, two-tailed. In addition, we separately included covariates such as cognition, anxiety, LEDD score, and number of wearing-off symptoms at baseline as interaction-effects (with condition) in the model to investigate whether they were of significant influence on the intervention effect. These results can help to indicate for whom the BEWARE treatment is most effective. Results Participant characteristics Table 1 shows the descriptive statistics of the 38 patients that completed the study, per condition. No significant baseline differences between the two conditions were found on demographic, disease related, or outcome measures. All participants had