Objectives: The aim of the study was to examine the validity of a translated and culturally adapted version of the Arthritis Impact Measurement Scales 2, Short Form (AIMS2-SF) in patients suffering from osteoarthritis (OA) in primary care.
Methods: A structured procedure was used for the translation and cultural adaptation of the AIMS2-SF into German. The questionnaire was administered to 220 primary care patients with OA of the knee or hip. Test-retest reliability was tested in 35 randomly selected patients, who received the questionnaire a second time after 1 week. The physical scale of the original AIMS2-SF was divided into an 'upper body limitations' scale and a 'lower body limitations' scale.
Results: With values ranging from 0.52 to 0.97 for Pearson's r, item-scale correlations were reasonably good. The discriminative power of separate scales was also good, reflected in low values for correlation between different scales, indicating little redundancy. Only two items (13 and 15) referring to the symptom scale showed item-scale correlation of r = 0.72 and r = 0.67, respectively with the lower body limitation scale. The assessment of internal consistency reliability also revealed satisfactory values: Cronbach's alpha was > or =0.83 for all scales, except for the social interaction scale (0.66). The test-retest reliability, estimated as the intraclass correlation coefficient (ICC), exceeded 0.85 except for the affect scale (0.72). Substantial floor effects occurred in the upper limb scale (33.8%). Principal factor analysis confirmed the postulated three-factor structure with physical, physiological and social dimensions, explaining 49.8, 14.1 and 6.4% of the variation, respectively. The assessment of external validity revealed satisfactory correlations with the corresponding WOMAC (Western Ontario and McMaster Universities Arthrosis Index) scales. As expected, correlations with radiological grading were moderate to low. The correlation with the physician's assessment was high in the scales that were dominated by physical factors, but rather low in the areas of health, which were found to be dominated by psychological or social factors.
Conclusion: The German AIMS2-SF is a reliable and valid instrument to assess the quality of life in primary care patients suffering from OA. When addressing the different impacts of OA, the physical scale should be divided into an upper body scale and a lower body scale. The floor and ceiling effects revealed are in accordance with the disease characteristics of the study sample and do not limit the significance of the questionnaire.