Faculty Bibliography

Background: We use machine learning to create predictive models from preoperative data to predict the Shoulder Arthroplasty Smart (SAS) score, the American Shoulder and Elbow Surgeons (ASES) score, and the Constant score at multiple postoperative time points and compare the accuracy of each algorithm for anatomic total shoulder arthroplasty (aTSA) and reverse total shoulder arthroplasty (rTSA). Methods: Clinical data from 2270 patients who underwent aTSA and 4198 patients who underwent rTSA were analyzed using 3 supervised machine learning techniques to create predictive models for the SAS, ASES, and Constant scores at 6 different postoperative time points using a full input feature set and the 2 different minimal feature sets. Mean absolute errors (MAEs) quantified the difference between actual and predicted outcome scores for each model at each postoperative time point. The performance of each model was also quantified by its ability to predict improvement greater than the minimal clinically important difference (MCID) and the substantial clinical benefit (SCB) patient satisfaction thresholds for each outcome measure at 2-3 years after surgery. Results: All 3 machine learning techniques were more accurate at predicting aTSA and rTSA outcomes using the SAS score (aTSA: ±7.41 MAE; rTSA: ±7.79 MAE), followed by the Constant score (aTSA: ±8.32 MAE; rTSA: ±8.30 MAE) and finally the ASES score (aTSA: ±10.86 MAE; rTSA: ±10.60 MAE). These prediction accuracy trends were maintained across the 3 different model input categories for each of the SAS, ASES, and Constant models at each postoperative time point. For patients who underwent aTSA, the XGBoost predictive models achieved 94%-97% accuracy in MCID with an area under the receiver operating curve (AUROC) between 0.90-0.97 and 89%-94% accuracy in SCB with an AUROC between 0.89-0.92 for the 3 clinical scores using the full feature set of inputs. For patients who underwent rTSA, the XGBoost predictive models achieved 95%-99% accuracy in MCID with an AUROC between 0.88-0.96 and 88%-92% accuracy in SCB with an AUROC between 0.81-0.89 for the 3 clinical scores using the full feature set of inputs. Discussion: Our study demonstrated that the SAS score predictions are more accurate than the ASES and Constant predictions for multiple supervised machine learning techniques, despite requiring fewer input data for the SAS model. In addition, we predicted which patients will and will not achieve clinical improvement that exceeds the MCID and SCB thresholds for each score; this highly accurate predictive capability effectively risk-stratifies patients for a variety of outcome measures using only preoperative data. Level of evidence: Level III; Retrospective Comparative study

Introduction: The purpose of this study is to evaluate the relationship between patients"™ own health expectations and treatment outcomes following surgical repair of proximal humerus fractures. Hypothesis: Patients"™ health expectations will correlate with treatment outcomes following surgical repair of proximal humerus fractures. Material and methods: Over a 14-year period, 247 patients with a displaced proximal humerus fracture who underwent ORIF with locking compression plates were prospectively followed at one academic institution. Minimum follow-up period was 12 months. Patient-reported functional outcome data for the latest follow up visit (12 months and greater) was obtained from Disabilities of Arm, Shoulder, and Hand (DASH) questionnaires. Survey responses regarding health expectations were recorded at 3-month follow-up and converted to dichotomous variables. Two groups were identified: the high expectations and the low expectations groups. Statistical analysis comparing the two groups and their functional and clinical outcomes was performed using the independent t-test, using p < 0.05 for significance. Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were calculated to further statistically characterize the relationship between health expectations at 3 months and long-term outcomes. Results: 185 (75.0%) patients available for analysis with a mean follow-up length of 24.8 months. The cohort included 124 (67%) females and 61 (33%) males and the average age at time of injury was 59.5 years. Eighty-six (46.5%) patients had low expectations for their overall health and 99 (53.5%) patients had high expectations for their health. No significant differences were seen between groups in regards to age, gender, follow-up length, Charlson Comorbidity Index (CCI), smoking and tobacco use, fracture pattern (OTA and Neer classifications), early complications (p > 0.05), fracture healing, and avascular necrosis. The mean DASH score at the latest follow up for patients with low expectations was 31.42 ± 22.8 whereas the mean for those with high expectations was 16.76 ± 20.2 (p < 0.0005). The mean forward flexion of the shoulder for patients with low expectations was 137.8 ± 31.5 degrees as compared to 148.5 ± 26.3 degrees (p < 0.05). The positive predictive value of good expectations correlating with good outcomes was 71,7%. Discussion: Patients with high expectations for their health early following injury had better outcomes in the long term. These high expectations also appeared to have an optimal influence on range of shoulder motion. This data suggests attitudinal and psychological factors that affect patient health expectations early on in the course of treatment may also influence patients"™ functional and clinical outcomes. Level of evidence: II; Retrospective Study.

Background: Glenoid wear is a common challenge in patients undergoing reverse total shoulder arthroplasty (rTSA). Augmented baseplates have recently been designed to address this. The purpose of this study is to determine the clinical outcomes, complications, and revision rates of patients undergoing rTSA with an augmented baseplate compared to those that received a standard, non-augmented baseplate. Methods: Preoperative and postoperative data were analyzed for 414 patients with glenoid bone loss who underwent primary rTSA with either an 8° posterior augmented glenoid baseplate, a 10° superior augmented baseplate (SAB), or an 8° posterior/10° superior augmented baseplate (P/SAB). These patients were compared to 1547 primary rTSA patients who received a standard baseplate. Clinical outcomes were scored using the Simple Shoulder Test (SST), UCLA, ASES, Constant and SPADI clinical outcome scoring metrics. Range of motion for active abduction, forward flexion, internal rotation and external rotation were used to quantify function. Complication and revision rates were also documented, and postoperative radiographs were analyzed for scapular notching. A two-tailed, unpaired t-test was used to identify differences between continuous parameters and a Chi Square test was used for categorical parameters, with P < .05 denoting a significant difference. Results: At an average follow-up of 47.1 ± 23.1 months, the augmented baseplates performed as well, or better, than standard baseplates, with a similar complication rate, revision rate, aseptic glenoid loosening rate, and scapular notching rate. Discussion: Augmented baseplates are a safe and effective option for patients with glenoid bone loss in the setting of rTSA at mid-term follow-up. For each baseplate type, the outcomes were similar, and the complication/revision and scapular notching rates were low and comparable to that of the standard baseplate cohort. Level of evidence: Level III

Background: An important psychometric parameter of validity that is rarely assessed is predictive value. In this study we utilize machine learning to analyze the predictive value of 3 commonly used clinical measures to assess 2-year outcomes after total shoulder arthroplasty (TSA). Methods: XGBoost was used to analyze data from 2790 TSA patients and create predictive algorithms for the American Shoulder and Elbow Surgeons (ASES), Constant, and the University of California Los Angeles (UCLA) scores and also quantify the most meaningful predictive features utilized by these measures and for all questions comprising each measure to rank and compare their value to predict 2-year outcomes after TSA. Results: Our results demonstrate that the ASES, Constant, and UCLA measures rarely considered the most-predictive features relevant to 2-year TSA outcomes and that each outcome measure was composed of questions with different distributions of predictive value. Specifically, the questions composing the UCLA score were of greater predictive value than the Constant questions, and the questions composing the Constant score were of greater predictive value than the ASES questions. We also found the preoperative Shoulder Pain and Disability Index (SPADI) score to be of greater predictive value than the preoperative ASES, Constant, and UCLA scores. Finally, we identified the types of preoperative input questions that were most-predictive (subjective self-assessments of pain and objective measurements of active range of motion and strength) and also those that were least-predictive of 2-year TSA outcomes (subjective task-specific activities of daily living questions). Discussion: Machine learning can quantify the predictive value of the ASES, Constant, and UCLA scores after TSA. Future work should utilize this and related techniques to construct a more efficient and effective clinical outcome measure that incorporates subjective and objective input questions to better account for the preoperative factors that influence postoperative outcomes after TSA. Level of Evidence: Level III; Retrospective Comparative Study

Background: The aim of this study was to compare outcomes of anatomic total shoulder arthroplasty (aTSA) and reverse total shoulder arthroplasty (rTSA) after prior shoulder stabilization versus matched cohorts without previous stabilization surgery. Hypotheses were as follows: (1) patients undergoing aTSA or rTSA after stabilization procedures would have worse outcomes than matched cohorts and (2) patients undergoing TSA would have better outcomes after soft-tissue stabilization procedures (aTSAST or rTSAST) than after bony stabilization procedures (aTSAB or rTSAB). Methods: Retrospective cohort study was performed comparing (1) 36 patients who underwent aTSA and (2) 32 patients who underwent rTSA with prior shoulder stabilization with 3-to-1 matched cohorts (based on age, gender, and follow-up length) with no prior shoulder instability or surgery. Baseline demographics, perioperative data, adverse events (AEs), radiographic outcomes, functional outcome scores, range of motion (ROM), and patient satisfaction were analyzed. Subgroup analyses compared patients who underwent aTSAST or rTSAST with patients who underwent aTSAB or rTSAB. Results: The postoperative AE rate was 8.3% and 4.6% in the aTSA group and matched cohort, respectively (P =.404), with a trend toward a significantly higher incidence of aseptic glenoid loosening in the aTSA group (8.3% vs. 1.9%, P =.067). Functional outcomes, ROM, and patient satisfaction did not differ at follow-up >4 years. In the subgroup analysis, two AEs required reoperation among 25 patients who underwent aTSAST versus one among 11 patients who underwent aTSAB, all related to aseptic loosening. There was a trend toward greater functional outcomes and satisfaction among patients who underwent aTSAST. There was a trend toward a clinically significant difference in active abduction at final follow-up favoring aTSAST (128 vs. 108 degrees, P =.096). The postoperative AE rate was 6.3% and 4.2% among the rTSA group and matched cohort, respectively (P =.632). Functional outcomes, ROM, and patient satisfaction did not differ at 4-year follow-up. In the subgroup analysis, no AEs were reported among 18 patients who underwent rTSAST and 14 patients who underwent rTSAB. A trend toward greater functional outcomes and patient satisfaction favored patients who underwent rTSAB, who achieved greater improvements in ROM from baseline and greater ROM in all planes at the final follow-up. Conclusion: Both aTSA and rTSA are reliable options for the treatment of dislocation arthropathy in appropriately selected patients. aTSA and rTSA after prior shoulder stabilization procedures have nearly equal rates of AEs and yield similar clinical and functional outcomes as matched cohorts. There may be an increased risk of glenoid aseptic loosening in aTSA after prior shoulder stabilization. Functional outcomes tend to be greater for patients who underwent aTSAST than those for patients who underwent aTSAB. On the contrary, rTSA may optimize postoperative function when performed for dislocation arthropathy after bony rather than soft-tissue stabilization procedures. Level of evidence: Level III; Retrospective Cohort Design; Treatment Study

Background: Massive irreparable rotator cuff tears (MIRCT) are challenging problems for both patients and surgeons. Reverse total shoulder arthroplasty (RTSA) is a treatment option for patients with MIRCTs. However, previous reports have shown inconsistent results, varying patient satisfaction, and higher complication rates. Methods: This is a retrospective multi-institutional study (22 institutions, 24 surgeons) of 203 patients (average age, 71 years) who underwent RTSA for MIRCT without glenohumeral arthritis with a mean follow-up of 50 months. Patients were divided into 4 groups based on preoperative shoulder active forward elevation (aFE) (<60°, <90°, ≥90°, >120°). Clinical outcomes were assessed using multiple patient-reported outcome measures (PROs), postoperative range of motion (ROM), patient satisfaction, and complication rate. Radiographic outcomes assessment included evaluation of postoperative scapular notching and humeral radiolucent lines. Results: Patients in each group had significant (P≤ 0.02) improvements in PROs and ROM postoperatively. Patient satisfaction was highest in the group with >120° preoperative aFE (44/44, 100%). Scapular notching and humeral radiolucency were noted in 6% and 7% of patients, respectively. There were only 3 complications that required 2 revision surgeries. Overall, the complication rate (1.6%) and reoperation rate (1.1%) were considerably lower than previously reported. Conclusion: RTSA is a reliable treatment for MIRCTs without glenohumeral arthritis that results in significant improvements in PROs and shoulder ROM. Compared to previous studies, we report a substantially higher satisfaction rates in all patients, especially in those with better preoperative ROM (aFE >120°), and a lower overall complication rate. Level of evidence: Level IV; Therapeutic Study

Background: The purpose of this study was to determine if addition of CT to axillary radiographs (AXR) alters preoperative decision making for shoulder arthroplasty. Methods: Preoperative deidentified images (XR alone and XR with CT) of 50 patients with glenohumeral arthritis were reviewed independently by 3 reviewers in a blinded fashion. Each reviewer graded images for glenoid wear pattern as simple (Walch A1 or B1) or advanced [A2, B2, C]), adequacy of AXR and need for advanced imaging. The reviewers determined a preoperative plan for all patients based on XR alone vs. XR and CT including the arthroplasty type (anatomic or reverse total shoulder) and their plan for treating glenoid wear (eccentric or standard reaming vs. bone graft or augment). Kappa values (κ) were calculated to determine inter-rater agreement and consistency among multiple reviewers. Fisher's exact test was used to assess any difference in preoperative plan once the shoulders were separated into simple and advanced glenoid wear patterns. Results: The 3 reviewers agreed that quality of AXRs was significantly inadequate (P <.001) for assessing glenoid wear in advanced glenoid wear patterns compared to simple wear patterns. Following evaluation on AXRs alone, the need for CT imaging was significantly higher in advanced glenoid wear patterns compare to simple ones (81% vs. 31%; P <.001). The addition of CT images did not result in a significant change to the preoperative plan in simple glenoid wear patterns but in advanced glenoid wear, addition of CT can change the preoperative plan with respect to arthroplasty type and/or strategy for addressing glenoid wear. Conclusion: Axillary radiographs are often inadequate for preoperative planning in shoulder arthritis with advanced glenoid wear patterns (Walch A2, B2, C types). Addition of CT imaging to radiographs in shoulder arthritis with advanced glenoid wear can affect the preoperative decision with respect to type of shoulder arthroplasty and/or plan for addressing glenoid wear (reaming, bone graft or augmented glenoids). Level of evidence: Level IV

BACKGROUND:The number of reverse total shoulder arthroplasties (RTSAs) performed annually has increased, and the indications for RTSA have expanded beyond rotator cuff arthropathy to include treatment of complex proximal humeral fractures. No studies exist comparing clinical, functional, and radiographic outcomes in patients receiving RTSA for the treatment of acute fracture vs. those undergoing the procedure for degenerative conditions. This study was designed to fill the void in this knowledge gap. We hypothesized that patients undergoing RTSA for fracture treatment would experience worse clinical outcomes than those undergoing elective RTSA. METHODS:A prospectively collected database was queried for patients undergoing RTSA between 2007 and 2016. Patients were sorted based on the indication for RTSA: treatment of acute proximal humeral fracture vs. "elective" treatment of degenerative conditions of the shoulder. Baseline demographic characteristics, intraoperative and perioperative complications, and clinical, functional, and radiographic outcomes were collected. Only patients with ≥2 years' follow-up were included. Final outcomes were compared between the fracture and elective groups. RESULTS:In total, 1984 patients met the inclusion criteria, with 1876 in the elective group and 108 in the fracture group. Compared with the elective RTSA group, the group undergoing RTSA for fracture treatment was older, was female dominant, and was less likely to have undergone a previous operation on the ipsilateral shoulder. RTSA for fracture was associated with a longer hospital length of stay and greater intraoperative blood loss. The incidence of postoperative adverse events was 7.1% in the elective group vs. 4.6% in the fracture group. Functional outcomes did not differ beyond 1 year or at mean final follow-up > 40 months. CONCLUSION/CONCLUSIONS:Despite differences in patient demographic characteristics, the outcome and complication profiles are similar between patients undergoing RTSA for acute fracture and those indicated for the treatment of degenerative conditions of the shoulder.