Page 111 - The value of total hip and knee arthroplasties for patients

Patients’ pre-operative general and specific outcome expectations predict postoperative pain and function after total knee and total hip arthroplasties
study showed that expectations consistently are part of the set of variables that
together best predict the outcomes function and pain 1 year after TKA and THA.
Post-hoc we assessed for each model what the amount of explained variance that
could be attributed to the expectations measure by running the final prediction
models again without the expectations variable and subtracting the R2 of these
models from the R2 of the final models described in the results section. The
amount of variance explained by expectations alone ranged from 1.3 to 6.5%. We
suggest that in planning surgical treatment the orthopedic surgeon should take
into account not only relative objective measures like age, degree of osteoarthritis
and comorbidity but also what the patient thinks to achieve with this THA or TKA
surgery. Although these factors seem important they only account for a limited
amount of the variance in outcomes. Still, we think that routinely assessing patients’
expectations in clinical practice is advisable because besides this predictive role
discussing patients’ expectations for TKA and THA has more functions in treatment
setting. Assessing and discussing patient’s expectations is also valuable for patient- 6 practitioner communication and shared decision making[45]. It is further suggested
that patients’ expectations may be a factor that is causally related to treatment outcome[46]. This could imply that through altering expectations one would be able to achieve better treatment outcomes. Although experimental research with healthy volunteers seems to point in this direction [47;48], clinical research has not confirmed this as RCT’s are scarce and observational studies have found mixed results and cannot fully establish causality[14]. Furthermore, it is still unclear what the most optimal expectation is in clinical situations. Should an expectation be high in order for the non-specific or placebo effects of the intervention to be optimal, or should high expectations for instance be tempered to prevent disappointment?
The second research question of this study examined whether the measurement approach used to measure expectations influenced the predictive value of expectations. Results showed that expectations that were measured with the more general CEQ expectancy subscale predicted most outcomes slightly better than the more outcome specific HSS expectation surveys, specifically for the outcome functioning. However, the differences in predictive value between the CEQ and HSS expectations survey are too small to give a definite answer to the question which one better predicts outcomes. Results do not correspond to those of[49] Iles et al who found that that the more specific the items of a questionnaire were, the better the predictive value for that outcome. A recent systematic review [32]
109