Page 78 - The value of total hip and knee arthroplasties for patients

Chapter 5
items are answered on an 8 point scale with two anchors ( 1=totally disagree and 8= totally agree).The hope scale is considered to be a unidimensional scale in which agency and pathways together represent the construct ‘trait hope’. Analysis of the Dutch version of the HS has shown good model fits for a one factor structure.3 Hence, a sumscore which ranges from 8-64 points is derived by summing the 8 items of the HS.
Treatment Credibility and Treatment Expectancy
The Credibility Expectancy Questionnaire is a self-reported six item questionnaire that aims to measure treatment credibility and expectancy for improvement. Originally it was developed by Devilly et al in 200014, and validated in several groups. The Dutch translation was done by Smeets et al in 200836. In both the original and the Dutch version three items (e.g. at this point, how successfully do you think the surgery will be in reducing your complaints) were found to load on the credibility factor and three items (e.g. at this point, how much do you really feel that the surgery will help to reduce your complaints) on the expectancy factor. Introductory instructions tell the patient that beliefs about how well the therapy might help contain both thoughts and feelings about the therapy and that these may be the same or different.36 Items 1 to 3 and 5 are answered on a scale ranging from 1 (not at all) to 9 (very much), Items 4 and 6 are answered on a 0 (not at all) to 100% (very much). In accordance with Smeets et al scores on item 4 and 6 were transformed with a minimum of 1 and a maximum of 9, and a sum score was formed for each factor ranging from 3 to 27.
Statistical analysis
Confirmatory factor analysis (CFA) for ordered categorical items was used to examine whether the constructs optimism (LOT-R), pessimism (LOT-R), hope (HS), treatment credibility (CEQ credibility subscale) and treatment expectancy (CEQ expectancy subscale), are distinguishable. Because observed variables were all answered on ordinal scales, a matrix based on polychoric correlations was used for CFA. Negatively formulated items of the LOT-R were reverse scored prior to entry into the CFA models. Analyses were conducted using the weighted least squares mean and variance adjusted estimator (WLSMV) in Mplus 6.12. For the total group of THA and TKA patients four theory-driven models with five, four, two and one latent factors, in which the factors were allowed to correlate within the CFA models, were evaluated using multiple fit indices and compared using Δχ2
76