Chapter 3 40 Our second aim was to determine if the RAdMAT-NL can be used as a single measure of adherence, so Rasch analysis was used for testing unidimensionality of the RAdMAT-NL. Rasch analysis indicates unidimensionality via Principal Component Analysis (PCA) of the residuals and local independence analysis [26]. If there are signs of other dimensions, Rasch analysis indicates items which are potentially contributing to another dimension. To perform the dichotomous Rasch analysis, the item scores were first dichotomized according to their mean (item score < mean = 0; item score > mean = 1) (to distinguish between patients scoring low and high). Visual assessment using item-characteristic curves (ICC) plots and mean square infit and outfit effect sizes were used to examine the fit of the data to the unidimensional Rasch model [27]. Mean square infit and outfit effect sizes between 0.5 and 1.5 were considered acceptable [28]. The Andersen’s likelihood ratio (LR) test and goodness of fit plots were used to test whether the Rasch model fits to the data. Unidimensionality was further examined by a parallel analysis on the residuals that remained after Rasch analysis. The internal consistency reliability of the RAdMAT-NL was examined by calculating McDonalds’s omega [29]. Internal consistency is considered acceptable if McDonald’s omega is 0.70 or higher [30]. Construct validity was examined by calculating Spearman’s correlations between the subscales of the RAdMAT-NL and the other adherence measures; SIRAS, percentage change on FTSST, self-reported home-based adherence, overall adherence and percentage attendance at scheduled rehabilitation appointments. The strength of correlations were interpreted based on the guidelines described by Cohen: small = 0.10-0.29, medium = 0.30-0.49 and large is 0.50-1.0 [31]. Results Patients from 53 different physiotherapy practices participated in the study. Out of 199 patients who gave informed consent, data from 193 patients were analyzed. The most common reasons for exclusion were; the physiotherapists did not provide any information, or the patient left the study (died or diagnosed with cancer). Table 1 summarizes the demographic and disease characteristics of the patients. Characteristics of the participating physiotherapists are summarized in Table 2.
RkJQdWJsaXNoZXIy MjY0ODMw