What Factors are Predictive of Patient-reported Outcomes? A Prospective Study of 337 Shoulder Arthroplasties

Although shoulder arthroplasties generally are effective in improving patients’ comfort and function, the results are variable for reasons that are not well understood.

We posed two questions: (1) What factors are associated with better 2-year outcomes after shoulder arthroplasty? (2) What are the sensitivities, specificities, and positive and negative predictive values of a multivariate predictive model for better outcome?

Three hundred thirty-nine patients having a shoulder arthroplasty (hemiarthroplasty, arthroplasty for cuff tear arthropathy, ream and run arthroplasty, total shoulder or reverse total shoulder arthroplasty) between August 24, 2010 and December 31, 2012 consented to participate in this prospective study. Two patients were excluded because they were missing baseline variables. Forty-three patients were missing 2-year data. Univariate and multivariate analyses determined the relationship of baseline patient, shoulder, and surgical characteristics to a “better” outcome, defined as an improvement of at least 30% of the maximal possible improvement in the Simple Shoulder Test. The results were used to develop a predictive model, the accuracy of which was tested using a 10-fold cross-validation.

After controlling for potentially relevant confounding variables, the multivariate analysis showed that the factors significantly associated with better outcomes were American Society of Anesthesiologists Class I (odds ratio [OR], 1.94; 95% CI, 1.03–3.65; p = 0.041), shoulder problem not related to work (OR, 5.36; 95% CI, 2.15–13.37; p < 0.001), lower baseline Simple Shoulder Test score (OR, 1.32; 95% CI, 1.23–1.42; p < 0.001), no prior shoulder surgery (OR, 1.79; 95% CI, 1.18–2.70; p = 0.006), humeral head not superiorly displaced on the AP radiograph (OR, 2.14; 95% CI, 1.15–4.02; p = 0.017), and glenoid type other than A1 (OR, 4.47; 95% CI, 2.24–8.94; p < 0.001). Neither preoperative glenoid version nor posterior decentering of the humeral head on the glenoid were associated with the outcomes. The model predictive of a better result was driven mainly by the six factors listed above. The area under the receiver operating characteristic curve generated from the cross-validated enhanced predictive model was 0.79 (generally values of 0.7 to 0.8 are considered fair and values of 0.8 to 0.9 are considered good). The false-positive fraction and the true-positive fraction depended on the cutoff probability selected (ie, the selected probability above which the prediction would be classified as a better outcome). A cutoff probability of 0.68 yielded the best performance of the model with cross-validation predictions of better outcomes for 236 patients (80%) and worse outcomes for 58 patients (20%); sensitivity of 91% (95% CI, 88%–95%); specificity of 65% (95% CI, 53%–77%); positive predictive value of 92% (95% CI, 88%–95%); and negative predictive value of 64% (95% CI, 51%–76%).

We found six easy-to-determine preoperative patient and shoulder factors that were significantly associated with better outcomes of shoulder arthroplasty. A model based on these characteristics had good predictive properties for identifying patients likely to have a better outcome from shoulder arthroplasty. Future research could refine this model with larger patient populations from multiple practices.

Level II, therapeutic study.