|Abstract / Summary
The overall purpose of this dissertation was to compare two types of osteosynthesis for distal radius fractures, using two questionnaires relevant for wrist fractures, with both questionnaires being part of the outcome measures used. Firstly, the Disabilities of Arm, Shoulder and Hand (DASH) questionnaire is a 30-item region specific questionnaire used to measure the effect of clinical treatment of the upper extremity. It exists in a Danish version, but reports about its reliability and validity were not available. Secondly, the Patient Rated Wrist Evaluation (PRWE), a wrist specific questionnaire that does not exist in Danish, but it would be of value if it did, since its high sensitivity towards measuring patients with distal radius fractures, has been established in other languages. As part of the validation process, the questionnaires would be tested for unidimensionality by using Rasch analysis and fitting from ordinal scale to interval scale. The dissertation encompasses four articles:
- A reliability and validity study of the DASH.
- A translation and cross cultural adaption of the PRWE, followed by a reliability and validity study.
- Rasch analysis of the DASH and PRWE investigating the questionnaires unidimensionality and conversion from an ordinal scoring scale to an interval scoring scale.
- A randomised prospective clinical trial comparing internal fixation (Micronail) and external fixation (Hoffmann II Compact non-bridging) for unstable distal radius fractures using the DASH and PRWE as outcome measures.
Material and methods:
Studies I and II. Sixty patients with wrist fractures were included. The patients answered the DASH, PRWE and Nottingham Health Profile (NHP) questionnaires after 1 week and 6 weeks after start of treatment. We measured the time-to-complete the questionnaires and missing items. We investigated internal consistency and test-retest reliability. Furthermore we investigated convergent and divergent validity, content validity, concurrent validity, construct validity and responsiveness. The translations component of the PRWE consisted of translation by a group of experts that included cross-cultural adaptations with a feedback phase by a group of uninjured laymen and a group of patients.
Study III. Data drawn from studies I and II resulting in 120 questionnaires for both the DASH and PRWE. Rasch analysis was conducted using RUMM2030 software to assess person separation index, unidimensionality, misfit of items, differential item functioning (DIF) and a fitted model in order to convert ordinal scores to interval scores.
Study IV. Patients were included from the accident and emergency departments of the three participating hospitals. Inclusion criteria were Older type 2 and 3 fractures. External fixation was managed with Hoffmann II compact non-bridging. Internal fixation was managed with Micronail. Patients were followed for 12 weeks. Primary outcome was the DASH. Secondary outcomes were PRWE, grip strength, satisfaction, radial length and volar tilt. Furthermore, complications were registered and an activity-based costing analysis was done.
Study I. Time-to-complete the questionnaire was 11 minutes. Cronbach´s alpha was 0.96 and intraclass correlation coefficient 0.89. Difference of Mean was 4.60(CI: 0.477-8.720, P=0.030). Convergent validity at first and last control was high for pain, 0.40 and 0.45, respectively, and for physical mobility, 0.60 and 0.67, respectively. Construct validity was significant. No floor or ceiling effect was seen. Effect size was 0.53.
Study II. Translation was done by an expert panel followed by evaluation by a lay panel and a field test on 10 patients. Both the lay panel comments and the field test revealed issues not dealt with by the expert panel and resulted in a Danish questionnaire that included technically and culturally adapted changes. Time-to-complete the questionnaire was 7 minutes. Cronbach´s alpha: 0.94. Intraclass correlation coefficient: 0.88. Difference of Mean: 5.70 (CI: 1.12-10.37, P=0.017). Convergent validity at first and last control was for pain, 0.44 and 0.46, and physical mobility, 0.51 and 0.64, respectively. No significant floor or ceiling effect was seen. Concurrent validity was 0.84. Construct validity was significant. Effect size: 0.62.
Study III. Rasch analysis showed good person separation index for the DASH and PRWE, and both showed unidimensionality. There was no DIF for the time interval. There were indications of misfit of items, but these items displayed good content validity and were within acceptable parameters. A nomogram for conversion from ordinal score to interval score could be made for both questionnaires.
Study IV. Thirty patients were randomized to external fixation and 31 patients to internal fixation. Internal fixation showed significantly better grip strength at 5 and 12 weeks. Apart from a shorter surgical time and minor differences in radiologic follow-up no other clinically relevant difference could be found. No significant result was found with the DASH and PRWE on either the ordinal and interval scale. An activity-based costing analysis showed a three times higher overall cost when using external fixation.
Though concerns about reliability, studies I, II and III showed acceptable reliability and good validity and unidimensionality, and both questionnaires can be used in clinical and research settings. When comparing internal fixation and external fixation, in study IV, internal fixation had significantly greater grip strength, and a lower overall cost when compared to external fixation.