Abstract
Significance
In spite of evidence that total knee replacement (TKR) surgery is effective, numerous studies have demonstrated that approximately 20 percent of patients who have undergone TKR surgery are not satisfied. This relatively high rate of patients who are not satisfied is the result of unmet patient expectations. The strongest predictor of dissatisfaction after TKR is unmet expectations (RR = 10.7, Bourne, Chesworth, et al, 2010). This is confirmed by Dunbar, Richardson, and Robertsson (2013): “Unmet expectation seems to be a major cause of unsatisfactory outcomes and satisfaction is most strongly correlated with relief of pain, followed by improvement in physical function.” Hypothesis: One year post-operative pain relief and activity level expectations can be measured pre-operatively and used for shared decision making.
Methods
A web–based system for prospectively collecting patient reported outcomes (PROs) has been developed. The data set for total hip/knee replacement surgery includes: (1) European quality of life, EQ-5D; Oxford Hip Score/Oxford Knee Score; (3) Lower Extremity Activity Scale (LEAS); and (4) Pain Likert Scale (PLS). The EQ-5D was selected as the health related quality of life (HRQL) general outcome measure because it has been adopted by multiple international joint replacement registries (Swedish Hip Arthroplasty Register, Norwegian Arthroplasty Register, United Kingdom National Joint Registry). The EQ-5D can be used to calculate quality adjusted life years (QALYs) for economic and/or comparative effectiveness analyses. The OHS/OKS questionnaires are used by the United Kingdom National Joint Registry and the New Zealand Joint Registry. The LEAS and PLS are used to measure patient's expectations for pain relief and functional improvement by asking patients to report their pre-operative pain and activity level before surgery and asking patients to report their pain and activity level expectations one year after surgery.
Results
The LEAS change (one-year post-operative activity level minus pre-operative activity level) mean and standard deviation are 1.45 and 3.18. The LEAS minimum clinically important difference (MCID) is 1.59. The mean LEAS change is 0.91 MCIDs. The PLS change (one-year post-operative pain score minus pre-operative pain score) mean and standard deviation are −4.57 and 2.85. The PLS MCID is 1.43. The mean PLS change is −3.2 MCIDs. The z statistic for expected change is z = (Δexpect – μΔ)/σΔ. The probability of realizing an expected activity level change greater than or equal to 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10 is 68%, 56%, 43%, 31%, 21%, 13%, 7.6%, 4.0%, 2.0%, 0.9%, and 0.4%, respectively. The probability of realizing an expected pain change greater than or equal to 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10 is 95%, 90%, 82%, 71%, 58%, 44%, 31%, 20%, 11%, 6.0%, and 2.8%, respectively.
Conclusions
Dramatic improvement in activity level is unlikely. However, 86 percent of patients can expect clinically significant pain relief defined by pain relief greater than the MCID. Shared decision making should discuss unrealistic expectations prior to proceeding with surgery.
