Publications

BACKGROUND: Over 100,000 patients undergo ACLR each year in the United States, providing rich electronic medical record data to improve outcomes, but manual chart review can be time-consuming and expensive. Natural language processing (NLP) methods have been successfully used to extract information for other orthopaedic operative procedures (e.g., hip and knee arthroplasties). Different subspecialities, however, use diverse terminology or "sublanguages" to describe clinical concepts and NLP models perform better concept extraction when trained on specialty-specific texts. Therefore, the purpose of this study was to develop and validate a reliable NLP pipeline to extract meaningful clinical data elements (e.g., graft type, meniscal involvement) from ACLR operative reports.

METHODS: Operative reports for a training and test set were randomly selected based on surgeon volume and year from individuals who underwent ACLR within a single healthcare system between 2013 and 2021. Clinical Language Annotation, Modeling, and Processing Toolkit (CLAMP) was utilized to train a domain-specific model and build a pipeline for data extraction of clinically meaningful data elements, including injury-related factors and surgical factors. Relevancy metrics were calculated using values identified by a single clinician as the gold standard.

RESULTS: Overall, the individuals selected for the training and test sets (n = 437 total) were 26.6 ± 10.9 years old, and 43.5% were female, similar to the age (26.7 ± 11.5 years) and proportion of females (43.0%) in the full data set (n = 5,818). Priority entities, including the side of surgery, ACL graft type, ACL procedure, and meniscal involvement, were identified with F1 scores between 0.87-1.

CONCLUSIONS: F1 scores of 0.87-1 were attained for several priority entities using a modest set of annotated ACL operative reports to train the model and an NLP tool designed for non-expert use. The NLP pipelines have the potential to extract relevant operative information for a large cohort of patients to support clinical research, such as identifying predictors of subsequent surgery after ACLR. This work can also inform the planning of future studies using larger training sets and more advanced NLP methods.

BACKGROUND: While medial meniscus posterior root repairs (MMPRRs) have been shown to reverse harmful biomechanical effects of medial meniscus posterior root tears, it is questionable whether meaningful improvement occurs in patients with advanced chondral loss.

PURPOSE/HYPOTHESIS: The objectives of this study were to compare midterm survival of MMPRR and to determine patient-reported outcomes (PROs) after MMPRR based on intraoperative Outerbridge grades (0-2 vs 3 and 4). It was hypothesized that patients undergoing MMPRR with Outerbridge grades 3 and 4 would have a decreased rate of MMPRR survival and inferior PROs compared to patients with Outerbridge grades 0 to 2.

STUDY DESIGN: Cohort study; Level of evidence, 3.

METHODS: Consecutive patients who underwent primary isolated MMPRR performed at a single institution between July 2003 and November 2022 with a minimum follow-up of 1 year were included and categorized according to medial compartment Outerbridge grade (0-2 or 3 and 4) based on diagnostic arthroscopy. Failure was defined as reoperation of the index knee. PROs collected included International Knee Documentation Committee (IKDC), visual analog scale (VAS) for pain, and Marx activity scale scores, which were compared to the previously established minimal clinically important difference (MCID) and/or clinically meaningful change. A minimum follow-up of 2 years was used for PRO analyses. Data were collected through electronic medical records and/or prospectively distributed surveys. A post hoc power analysis was performed, yielding a post hoc power of >0.99.

RESULTS: A total of 170 patients were included (74 in the Outerbridge grade 0-2 group and 96 in the Outerbridge grade 3 and 4 group). The mean follow-up time was 5 years. There were no significant descriptive differences between groups. The median survival was 11 years, and there were no significant differences between groups in the survival analysis. There was 80% survival at 8 years overall, with 80% survival at 8 years and 7 years for Outerbridge grades 0 to 2 and grades 3 and 4, respectively. The failure rate was 17.6%, with no significant differences between groups, and the mean time to reoperation was 5.3 years. The mean postoperative IKDC score was 64.1 ± 23.0. The mean IKDC change was 27.6 ± 29.7, reaching the MCID of 10.17, with no significant differences between groups. The VAS score decreased from 6.1 ± 2.2 preoperatively to 3.0 ± 3.0 postoperatively, with no significant differences between groups.

CONCLUSION: Midterm survival and clinical outcomes after MMPRR were comparable between patients with intraoperative Outerbridge grades 0 to 2 and grades 3 and 4, with no significant differences in failure rate or survival. The overall median survival was 11 years, with an 80% survival rate at 8 years. The change in IKDC score exceeded the MCID in both groups. These findings indicate that MMPRR is durable and effective and intraoperative Outerbridge classification grades should not dissuade orthopaedic surgeons from performing MMPRR.

Inertial measurement units (IMUs) offer a promising pathway to extend functional-recovery assessment after anterior cruciate ligament reconstruction (ACLR) beyond laboratory or in-clinic evaluations. Yet, it remains unclear how joint kinematics change in natural environments over time. In this study, 26 participants wore five sticker-like IMUs on the pelvis, thighs, and shanks for up to seven days, three and nine months after their surgery. They also participated in a laboratory assessment at each timepoint using standard marker- and inertial-based motion capture systems. Knee extension excursion (KEE) was estimated, and inter-limb KEE asymmetry was used as the primary outcome. Median KEE asymmetry decreased significantly from 7.7° at three months to 2.6° at nine months post-surgery, through a modest, 1.0° increase in the reconstructed KEE and a more salient reduction of 4.3° in the contralateral KEE. In the laboratory, however, only a 2.2° decrease in KEE asymmetry was observed, attributed to a 3.3° increase in reconstructed KEE and no significant change in the contralateral KEE. Sensitivity analyses revealed that findings were robust to walking-bout duration and number of days included in the analysis, and all reported group changes in KEE asymmetry exceeded disagreement between IMU- vs. marker-based tracking. Participant compliance to remote monitoring was high, with nearly 95% of the sessions completed. Together, these findings demonstrate that joint-level biomechanics can be captured longitudinally in natural environments after ACLR and that ecologically valid monitoring outside the laboratory can reveal aspects of functional recovery not observed in controlled settings.

BACKGROUND: Bariatric surgery (BS) is an increasingly utilized intervention for the treatment of obesity. However, BS is also associated with postoperative nutritional deficiencies that may affect healing rates after orthopaedic procedures.

PURPOSE: To compare failure rates and patient-reported outcomes after arthroscopic rotator cuff repair (RCR) between patients with and without a history of BS.

STUDY DESIGN: Cohort study; Level of evidence: 3.

METHODS: Patients in a single institution with a history of BS who underwent arthroscopic RCR for full-thickness supraspinatus tears were identified. These patients were matched in a 1:3 ratio by age, sex, and body mass index to patients without a history of BS who underwent arthroscopic RCR. The minimum follow-up was 24 months. The primary outcome was surgical failure. Secondary outcomes assessed included the numeric rating scale (NRS) score for pain, Single Assessment Numeric Evaluation (SANE) score, American Shoulder and Elbow Surgeons (ASES) Shoulder Score, need for manipulation under anesthesia (MUA) or arthroscopic lysis of adhesions (LOA), infection requiring reoperation, and conversion to reverse total shoulder arthroplasty (rTSA).

RESULTS: A total of 34 arthroscopic patients with a history of BS who underwent RCR were matched to 102 patients without BS. The BS group had significantly higher overall failure rates (20.6% vs 6.9%; P = .044) than patients without BS. The BS group had significantly higher postoperative NRS pain scores (3.9 vs 1.3; P < .001), lower SANE scores (77.7 vs 87.7; P = .041), and lower ASES scores (72.6 vs 90.4; P < .001) at the final follow-up. Rates of revision RCR, reoperation for MUA or LOA, and conversion to rTSA were not statistically significantly different (P > .050 for all). No postoperative infections were reported.

CONCLUSION: A history of BS is associated with increased failure rates, worse postoperative pain, and worse patient-reported outcomes after arthroscopic RCR. Patients with a history of BS and those considering BS before arthroscopic RCR should be counseled regarding a possible risk for inferior outcomes after surgery.

PURPOSE: To investigate location- and sex-specific glenoid labrum morphology of healthy young adults in vivo, and the labrum's effects on depth and radius of curvature (ROC) of the glenohumeral joint.

METHODS: Healthy young adults with no history of shoulder surgery, injury, or instability underwent computed tomography and magnetic resonance imaging scans. Bone, cartilage, and labrum tissues were segmented from the scans and used to create 3-dimensional models for each participant. Measurements were made on the coregistered 3-dimensional models. The labral thickness, height, and depth were expressed according to a clockface on the glenoid (3, 6, 9, and 12 o'clock locations). Glenoid depth and ROC were also measured. Comparisons were made between sexes and among clockface locations.

RESULTS: Sixty shoulders of 30 individuals (15 males/15 females, average age: 25 ± 7 years, body mass Index: 25.4 kg/m2) were included. Repeated segmentation of a subset of the data showed the average absolute differences in labrum depth ranged from 0.3 to 0.4 mm, the labrum thickness differences ranged from 0.3 to 0.8 mm, and the labrum height differences ranged from 0.4 to 0.8 mm between the original and resegmented scans. For the complete dataset, no location-specific differences in labrum morphology were found between males and females. The 12 o'clock location was at least 5.0 mm deeper, 4.3 mm thicker, and 8.4 mm taller than any other clockface location (all P < .001). The labrum increased the average depth of the glenoid between 2.4 and 5.6 mm in the anterior/posterior and superior/inferior locations, respectively (P < .001), contributing an average of 60.1% of the overall glenoid depth, and decreasing the average ROC by 22.6 mm in the anterior/posterior locations and by 8.8 mm in the superior/inferior locations.

CONCLUSIONS: The labrum in young healthy adults was found to be larger than previously reported in cadaver studies. The labrum at 12 o'clock is deeper, thicker, and taller than at other locations; however, no location-specific differences were found between men and women. The labrum contributed significantly to the depth and created a more congruent joint by minimizing differences between humerus and glenoid ROC.

CLINICAL RELEVANCE: Improved understanding of the morphology of the glenoid labrum in young adults may be used to guide surgical repair and design of anatomical shoulder replacements.

BACKGROUND: The impact of arthroscopic Bankart repair (ABR) alone vs. ABR with remplissage (ABR + R) on athlete return-to-play rates and number of games played after surgery is poorly understood. The objective of this study was to utilize online sports databases to compare number of games played and return-to-play rates between athletes who underwent ABR vs. ABR + R for "on-track" Hill-Sachs lesions (HSLs). We hypothesized that there would be no difference in relative games played nor return-to-play rates post-operatively between patients undergoing ABR vs. ABR + R.

METHODS: Patients aged 14-40 years with "on-track" HSLs who underwent either ABR or ABR + R between 2007 and 2022 for anterior shoulder instability were retrospectively reviewed. Exclusion criteria included revision surgery, <1-year follow-up, "off-track" HSLs, >20% glenoid bone loss, nonathletes, and missing data in online sports databases. Athletes were queried in online sports databases and games played were recorded the season before and 3 seasons after surgery, if available. The primary outcome was relative change in games played in the seasons after surgery compared to the season before surgery ( R e l a t i v e G a m e s P l a y e d = # G a m e s i n S e a s o n A f t e r S u r g e r y # G a m e s i n S e a s o n B e f o r e S u r g e r y ). Secondary outcomes included recurrent anterior shoulder instability, defined as recurrent dislocation and/or subluxation, and return-to-play.

RESULTS: Eighty-one patients (ABR: 60 | ABR + R: 21) were included in the analysis, with average age of 18 ± 2 years and average follow-up of 6.5 ± 3.7 years (range, 1.0-14.4 years). "Near-track" HSLs (distance-to-dislocation <10 mm) were present in 20% of the ABR group vs. 76% of the ABR + R group (P < .01). Return-to-play rates were similar between groups (ABR: 75% vs. ABR + R: 81%, P = .58). In the first season after surgery, the ABR group had significantly higher relative games played compared to the ABR + R group (ABR: 1.5 ± 1.5; n = 38 | ABR + R: 0.8 ± 0.5; n = 13, P = .02). This difference did not persist in the second and third seasons after surgery (P > .05). In an adjusted Generalized Estimating Equation model, surgical technique was not significantly associated with relative games played after surgery. Rates of recurrent anterior shoulder instability were not statistically different between groups (ABR: 28% (17/60) vs. ABR + R: 10% (2/21), P = .13).

CONCLUSION: Among athletes with "on-track" HSLs, ABR + R resulted in similar long-term game participation and return-to-play rates and compared to isolated ABR alone. Although patients undergoing ABR + R demonstrated fewer relative games played in the first post-operative season, this difference did not persist in subsequent seasons. Remplissage augmentation may therefore have a temporary impact on early post-operative performance without compromising overall return-to-play outcomes.

BACKGROUND: Calcific tendinitis of the shoulder is a common, painful rotator cuff disorder with both nonoperative and operative treatment options. The optimal nonoperative modality remains unclear, and it is not well understood how previous nonoperative treatments influence eventual surgical outcomes.

PURPOSE/HYPOTHESIS: The purpose of this study was to compare success rates, defined as avoidance of surgery, among 3 nonoperative treatments for calcific tendinitis: physical therapy (PT), corticosteroid injection (CSI), and ultrasound-guided barbotage (USB). For patients who underwent surgery, outcomes were compared according to previous nonoperative management. It was hypothesized that success rates and postoperative outcomes would not differ significantly between modalities.

STUDY DESIGN: Cohort study; Level of evidence, 3.

METHODS: A retrospective review of patients diagnosed with calcific tendinitis from 2009 to 2023 was performed. Exclusion criteria were lack of radiographic confirmation, <6 months follow-up, or incomplete electronic medical record data. Patients were categorized by attempted nonoperative treatment (none, PT, CSI, USB, multiple) and by final management (nonoperative vs operative). Patient-reported outcomes included the visual analog scale (VAS), Subjective Shoulder Value (SSV), and range of motion (ROM), collected at initial and final presentation. Radiographic findings were extracted from radiology reports. Statistical testing used parametric or nonparametric methods as well as a multivariable Cox proportional hazards model to predict nonoperative failure. Significance was set at P < .05.

RESULTS: A total of 257 patients (mean age 55 ± 11 years) were analyzed with a mean follow-up of 18 ± 16 months with an overall nonoperative success rate of 63%. Success rates did not differ significantly among PT (59%), CSI (75%), and USB (72%), but patients with multiple tendon involvement or calcifications >3 cm were more likely to fail nonoperative management. Patients completing successful nonoperative management improved in VAS, SSV, and ROM, with no between-group differences. All patients who attempted multiple modalities (18/18; 100%) required surgery (P < .01). Among 121 patients undergoing surgery, final VAS, SSV, and ROM outcomes did not differ based on previous nonoperative management.

CONCLUSION: PT, CSI, and USB demonstrate similar rates of avoiding surgery for calcific tendinitis. For patients ultimately requiring surgery, outcomes are not influenced by previous nonoperative management. Surgical intervention may be indicated after failure of a single nonoperative modality.

The purpose of this study was to evaluate how surgical technique and prosthesis design affect muscle activation after reverse shoulder arthroplasty (RSA) during hand-to-head (H2H) and hand-to-back (H2B) movements. Surface electromyography data from 8 muscles were collected while 28 RSA patients performed H2H and H2B. Muscle activation onset (AO) and total muscle activation (TMA) were calculated. Multiple regression was used to identify surgical technique and prosthesis design factors that predicted muscle AO and TMA. Correlation was used to identify associations between muscle activation and patient reported outcomes (CMS and DASH). During H2H, neck shaft angle predicted anterior deltoid, pectoralis, and latissismus muscle AO and lateral humeral offset (LHO) predicted latissimus AO. Glenosphere eccentricity, glenosphere tilt, glenosphere size, and LHO predicted TMA for anterior deltoid, middle deltoid, teres minor, and pectoralis muscles, respectively. During H2B, glenosphere lateralization predicted trapezius, anterior deltoid, and teres minor muscle AO, and glenosphere eccentricity and humeral retroversion change predicted trapezius TMA. During H2H, later middle deltoid AO was associated with better CMS, and later pectoralis major AO was associated with worse DASH scores. Additionally, higher teres minor TMA was associated with worse internal rotation score. During H2B, later latissimus AO was associated with better DASH scores. Improved understanding of how implant design and surgical technique affect muscle activation and clinical outcomes after RSA may help surgeons and therapists to optimize shoulder function after RSA.

BACKGROUND: To evaluate how patient-reported outcomes (PROs) trend over time following arthroscopic Bankart repair (ABR) for anterior shoulder instability with on-track Hill-Sachs lesions (HSLs).

METHODS: A retrospective review was performed of patients undergoing ABR with or without remplissage between 2007 and 2023. Exclusion criteria included age <14 years, glenoid bone loss >20%, off-track HSLs, and prior shoulder surgery. PROs and clinical outcomes, including Western Ontario Shoulder Index (WOSI), pain visual analog scale (pVAS), Subjective Shoulder Value (SSV), and recurrent instability, were collected at final follow-up. Spearman rank correlation assessed relationships between PROs and time from surgery. Rates of achieving Patient Acceptable Symptom State for WOSI and pVAS were compared across follow-up intervals (<5 years, 5-9 years, ≥10 years) and between patients with and without recurrent instability. Significance was set at P < .05.

RESULTS: A total of 121 patients (age: 23 ± 8 years, 27% females) were included at mean follow-up of 7.1 years (range: 1.6-17.2). Longer follow-up duration was associated with worsening WOSI (r = 0.239; P = .008), pVAS (r = 0.180; P = .049), and SSV (r = -0.186; P = .041) scores. Patients undergoing isolated ABR demonstrated significant deterioration in PROs over time (WOSI: r = 0.331, P = .002; pVAS: r = 0.261, P = .015; SSV: r = -0.216, P = .045), whereas those undergoing remplissage showed no change in PROs over time (all P > .05). Patients with recurrent instability had lower rates of achieving WOSI Patient Acceptable Symptom State at <5 years (56% vs. 88%; P = .029), 5-9 years (22% vs. 87%; P < .001), and ≥10 years (20% vs. 86%; P < .001) compared to patients without recurrent instability. On linear regression analysis, longer follow-up length remained an independent predictor of worse WOSI scores after controlling for recurrent instability.

CONCLUSION: PROs may decline over time following ABR for anterior shoulder instability for patients who experience recurrent anterior shoulder instability and undergo isolated ABR compared to ABR with remplissage. These findings may be relevant for pre-operative counseling regarding long-term subjective outcomes.

BACKGROUND: Previous literature has demonstrated that an increased number of preoperative anterior shoulder instability episodes is associated with recurrent anterior shoulder instability after arthroscopic Bankart repair (ABR). However, a threshold for the number of preoperative instability episodes that increases the risk of recurrent anterior shoulder instability is not well established.

PURPOSE: To establish a threshold value for the number of preoperative instability episodes that predicts recurrent anterior shoulder instability after ABR and to compare glenoid bone loss, the Hill-Sachs interval, and the distance to dislocation between patients who have surpassed the threshold and those who have not.

STUDY DESIGN: Retrospective cohort study; Level of evidence, 3.

METHODS: This retrospective review included consecutive patients with "on-track" Hill-Sachs lesions who underwent primary ABR for anterior shoulder instability at a single institution between 2007 and 2019. Patients with an unknown number of preoperative instability episodes, >20% glenoid bone loss, <2 years' follow-up, or age >40 or <14 years were excluded. Logistic regression assessed associations between preoperative anterior shoulder instability episodes and recurrent anterior shoulder instability after ABR, defined as a recurrent subluxation or dislocation. Receiver operating characteristic analysis determined the optimal threshold of preoperative anterior shoulder instability episodes to predict recurrent anterior shoulder instability. Significance was set as P < .050.

RESULTS: A total of 151 patients (mean age, 20 ± 5 years; mean follow-up, 6.0 ± 3.1 years) were included, of whom 28 (19%) experienced recurrent anterior shoulder instability. Multiple thresholds showed increased odds of recurrent anterior shoulder instability: ≥2 preoperative anterior shoulder instability events (odds ratio [OR], 9.70 [95% CI, 2.63-35.70]; P = .001), ≥3 events (OR, 3.47 [95% CI, 1.37-8.80]; P = .009), and ≥4 events (OR, 3.08 [95% CI, 1.17-8.08]; P = .023). Receiver operating characteristic analysis revealed that ≥2 preoperative anterior shoulder instability events was the strongest predictor of recurrent anterior shoulder instability (area under the curve = 0.72).

CONCLUSION: A threshold of ≥2 preoperative anterior shoulder instability episodes best predicted recurrent anterior shoulder instability after ABR. Stratification beyond 1 versus ≥2 preoperative anterior shoulder instability episodes did not increase predictive ability. This finding may help surgeons to counsel patients and consider earlier surgical stabilization in those who have sustained anterior shoulder instability episodes.