Page 89 - The value of total hip and knee arthroplasties for patients

Distinctiveness of Psychological Constructs
missing data and the fact that that participants with and without missing data (on item 6 of the HS) did not significantly differ on baseline characteristics we believe that the MCAR assumption may hold for our data.
Furthermore, we used Δχ2-test to compare models, which is controversial as the χ2 is influenced by sample size; therefore we also included other model fit statistics and based our conclusions on a combination of a-priori defined cut-points. CFA is a test of acceptance of a-priori defined models that are not data driven. A limitation of this method is however, that besides the theoretically plausible models tested in this study, there might be other models that show an even better fit to the data. Our results have furthermore not been validated in an external dataset we therefore encourage future research in TKA and THA and also in other patient groups.
Conclusions and Implications 5 Based on the results of the current study and previous work we suggest that
the constructs treatment expectancy, treatment credibility, hope, optimism and
pessimism are distinguishable in THA and TKA patients. Posthoc, a bifactor model in
which (besides the 5 separate factors) a general factor is hypothesized accounting for the commonality of the items showed a significantly better fit than the five factor model. All specific factors, except for the hope factor, showed to explain a substantial amount of variance beyond the general factor. Future studies should investigate the factorial structure of the CEQ. Our results may be valuable for the design of clinical studies aiming to measure one or more of these constructs as well as for the evaluation of interventions focussed on altering treatment expectancy which have been initiated lately by several groups 23;24;27. As optimism and hope have been hypothesized to be relatively stable traits, it is necessary for researchers evaluating interventions aimed at altering treatment expectancy, to measure the possibly alterable treatment expectancy distinct from optimism and hope.
A next step in making these constructs of benefit for the patient undergoing THA and TKA is to investigate the relationships between these factors (e.g. Does optimism influence treatment specific expectancies?) and to find out which one or which combination of constructs predicts with more accuracy treatment outcomes after THA and TKA like pain, quality of life and physical well-being the best. In the future clinicians may use these constructs in addition to other tools, in order to identify patients with a high-risk for poor outcome in their decision for the type of intervention, either surgical or conservative.
87