[Development of a questionnaire to assess treatment representations in rehabilitation patients]

The purpose of the study was to develop a generic questionnaire for assessing the patients' rehabilitation treatment representations on the theoretical basis of the Common Sense Model of self-regulation and analogous with established assessment instruments for illness representations and medication-related treatment representations.

Using theory-based and empirical patient and provider input, a pool of 115 items on the patients' rehabilitation treatment representations was generated. The items were first pre-tested on n=182 rehabilitation patients in nine rehabilitation centres for the indications musculoskeletal diseases and psychosomatic illnesses. The next step was the factor analytical reduction of the item pool. The resulting factors were analyzed for unidimensionality, and the remaining items were analyzed for distribution of values. Internal consistency and the "known groups validity" were used to determine reliability and validity, respectively. Finally, patient acceptance of the questionnaire was assessed.

The item pool was considerably reduced by factor analysis and the remaining items were grouped under four factors. After eliminating other single items on the basis of modification indices and indicator reliability, the unidimensionality of these factors was assessed as good. The majority of the remaining 23 items showed an asymmetrical distribution, which was also reflected by low item difficulty. The contents of the scales can be interpreted as "somatic outcome expectation", "psychological outcome expectation", "process expectations", and "concerns". The internal consistency of the scales (Cronbach's alpha) lies between 0.80 and 0.84 for the entire sample. The treatment representations for psychosomatic patients differ in the anticipated direction from those of orthopaedic rehabilitation patients. Patient acceptance of the questionnaire was very positive.

The questionnaire developed allows aspects of the patients' rehabilitation treatment representations to be measured, the features of which are highly relevant for providers as a basis for providing individualized patient information. However, the results of the pretest have to be tested confirmatory and also indicate that there is still potential for improving the instrument and making it more precise.