Skip to main content

Advertisement

Table 5 Results of multiple regression analysis explaining variance in satisfaction at 2 years

From: The role of patient expectations in predicting outcome after total knee arthroplasty

Step Step change in R2 Pvalue for step change in R2 β for final model (only significant predictor variables shown) Pvalue
First 0.095 0.040 0.194 (other joint problems) 0.042
Second 0.231 < 0.0001 0.517 (pain at 2 years) < 0.0001
Adjusted R2 for model 0.293    
  1. Results of the multiple regression analysis showing the factors that made a unique significant contribution to explaining the variance in satisfaction at 2 years (1 = very satisfied, 4 = very dissatisfied). In the final model, the significant predictors of a poorer outcome were: other joint problems, more pain at 2 years post operation, and greater functional limitation at 2 years post operation. n = 87 patients (listwise exclusion of missing data, and excluding patients (n = 13) that underwent further surgery on the index knee). Apart from the demographic variables, predictor variables were entered on the basis of the significance of their bivariate correlation with the dependent variable: step 1, simultaneous entry for age, gender, other joint problems (yes/no); step 2, forward conditional entry for preoperative expectations (about pain and about functional limitations), knee status at 2-year follow-up (in terms of pain and functional limitations), change in knee status from pre surgery to 2 years (in terms of pain and functional limitations), expectations - actuality scores for knee status (that is, expected status minus actual status at 2 years) (in terms of pain and functional limitations). Step change in R2, increase in explained variance at the given step; adjusted R2, R2 - (k - 1)/(n - k) × (1 - R2), where n is the number of observations and k is the number of independent variables; β for final model, β value after all variables have been entered; P value, significance of final β value for the stated variable.