Douglas W. Lundy
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Douglas W. Lundy has a total of 179 co-authors. Here are the top co-authors by number of shared publications:
- Dustin Adam Greenhill
St. Luke’s University Health Network
3 shared publications - Ashley E. Levack
Department of Orthopaedic Surgery & Rehabilitation, Loyola University Medical Center, Maywood, IL
2 shared publications - Guy Putzeys
Orthopedic and Trauma Department, AZ Groeninge hospital, Kortrijk, Belgium
2 shared publications - Hans-Christoph Pape
Department of Trauma Surgery, University Hospital Zurich, University of Zurich, Zurich, Switzerland
2 shared publications - Inger Birgitta Schipper
Department of Trauma Surgery, Leiden University Medical Center, Leiden, Netherlands
2 shared publications - Jonathan McKeeman
Department of Orthopaedic Surgery St. Luke’s University Health Network Bethlehem Pennsylvania U.S.A.
2 shared publications
Latest Publications
- 1 December 2026
- Published by Wolters Kluwer Health in Journal of Orthopaedic Trauma
- Vol. 40 (12) , 270-271
Response Letter to the Editor Regarding Article “Orthopaedic Trauma Association Annual Meetings Consistently Provide Academic Value for Attendees”
- 1 December 2026
- Published by Wolters Kluwer Health in Journal of Orthopaedic Trauma
- Vol. 40 (12) , 270-271
Response Letter to the Editor Regarding Article “Orthopaedic Trauma Association Annual Meetings Consistently Provide Academic Value for Attendees”
- Open Access
- 22 October 2026
- Published by Elsevier in SSRN Electronic Journal
Automated AO/OTA Fracture Classification from Radiology Reports and Verbal Clinical Communication Using Large Language Models
Introduction: The AO/OTA Fracture and Dislocation Compendium is important in clinical communication and is critical for research, yet it is rarely documented systematically in electronic medical records. This study evaluates the accuracy of a Retrieval-Augmented Generation (RAG) Large Language Model (LLM) pipeline in automating AO/OTA fracture classification from plain-text radiology reports and clinical communication via verbal presentations during morning fracture conferences. Methods: We collected 71 de-identified radiology reports from a Level 1 trauma center. An expert OTA Classification Committee of eight attending surgeons established three "ground truths": one based on report text, one based on radiographs, and one based on resident verbal presentations. A HIPAA-compliant LLM (Versa, based on GPT-o3) was integrated with a RAG pipeline using the AO/OTA Compendium as its reference library. The LLM’s performance was compared against the committee's labels and a baseline cosine similarity NLP method using Accuracy and Cohen’s Kappa. Results: The RAG-augmented LLM demonstrated Near Perfect agreement for foundational categories, including Bone (97.6%, kappa=0.972) and Location (95.1%, kappa=0.938). For core morphology, the model reached Moderate agreement for Type (68.3%, kappa=0.579) and Fair for Group (62.2%, kappa=0.396). Cumulatively, the LLM maintained Substantial agreement through the Bone/Location/Type level (65.9%, kappa=0.645), significantly outperforming the baseline NLP method (22.5%, kappa=0.214). At the Bone/Location level, LLM agreement with imaging ground truth (81.7%) approached the committee's own inter-rater consistency (87.8%). When applied to resident verbal presentations during fracture conference, the LLM achieved Substantial agreement for Bone (78.6%, kappa = 0.758) and Location (72.9%, kappa = 0.648), and Fair agreement at the Bone/Location/Type level (41.4%, kappa = 0.403). Conclusion: A RAG-enhanced LLM can classify fractures from plain text and verbal clinical communication with reliability comparable to expert surgeons, both from structured radiology reports and unstructured resident verbal presentations. While manual refinement is still required for granular universal modifiers, the substantial agreement at the Bone/Location/Type level supports the implementation of AI-assisted documentation tools. Such tools offer a lightweight, scalable solution to enhance EMR data quality and facilitate large-scale orthopaedic research. Further integration of LLMs into clinical workflows is needed to provide real-time feedback that improves radiology reporting and clinical communication.
- 22 October 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessAutomated AO/OTA Fracture Classification from Radiology Reports and Verbal Clinical Communication Using Large Language Models
Introduction: The AO/OTA Fracture and Dislocation Compendium is important in clinical communication and is critical for research, yet it is rarely documented systematically in electronic medical records. This study evaluates the accuracy of a Retrieval-Augmented Generation (RAG) Large Language Model (LLM) pipeline in automating AO/OTA fracture classification from plain-text radiology reports and clinical communication via verbal presentations during morning fracture conferences. Methods: We collected 71 de-identified radiology reports from a Level 1 trauma center. An expert OTA Classification Committee of eight attending surgeons established three "ground truths": one based on report text, one based on radiographs, and one based on resident verbal presentations. A HIPAA-compliant LLM (Versa, based on GPT-o3) was integrated with a RAG pipeline using the AO/OTA Compendium as its reference library. The LLM’s performance was compared against the committee's labels and a baseline cosine similarity NLP method using Accuracy and Cohen’s Kappa. Results: The RAG-augmented LLM demonstrated Near Perfect agreement for foundational categories, including Bone (97.6%, kappa=0.972) and Location (95.1%, kappa=0.938). For core morphology, the model reached Moderate agreement for Type (68.3%, kappa=0.579) and Fair for Group (62.2%, kappa=0.396). Cumulatively, the LLM maintained Substantial agreement through the Bone/Location/Type level (65.9%, kappa=0.645), significantly outperforming the baseline NLP method (22.5%, kappa=0.214). At the Bone/Location level, LLM agreement with imaging ground truth (81.7%) approached the committee's own inter-rater consistency (87.8%). When applied to resident verbal presentations during fracture conference, the LLM achieved Substantial agreement for Bone (78.6%, kappa = 0.758) and Location (72.9%, kappa = 0.648), and Fair agreement at the Bone/Location/Type level (41.4%, kappa = 0.403). Conclusion: A RAG-enhanced LLM can classify fractures from plain text and verbal clinical communication with reliability comparable to expert surgeons, both from structured radiology reports and unstructured resident verbal presentations. While manual refinement is still required for granular universal modifiers, the substantial agreement at the Bone/Location/Type level supports the implementation of AI-assisted documentation tools. Such tools offer a lightweight, scalable solution to enhance EMR data quality and facilitate large-scale orthopaedic research. Further integration of LLMs into clinical workflows is needed to provide real-time feedback that improves radiology reporting and clinical communication.
- 9 September 2026
- Published by Wolters Kluwer Health in Journal of Bone and Joint Surgery
Business Aspects of Orthopaedics: The Private Equity Case
Was not provided for this article.
- 9 September 2026
- Published by Wolters Kluwer Health in Journal of Bone and Joint Surgery
Business Aspects of Orthopaedics: The Private Equity Case
Was not provided for this article.
- Open Access
- 1 May 2026
- Published by Wolters Kluwer Health in OTA International
- Vol. 9 (2)
Deriving the OTA/AO fracture classification from routinely collected radiology reports using a large language model
Objectives: Fracture classification plays a pivotal role in research and quality assurance; despite its wide acceptance, the OTA/AO classification is seldom documented in patients' electronic medical records, which impedes fracture registry creation and effective interdisciplinary communication. In this study, we investigate “off-the-shelf” large language models (LLMs) in translating free text in radiology reports into OTA/AO classification labels. Methods: We employed a Health Insurance Portability and Accountability Act-compliant LLM to classify 109 fracture descriptions from randomly selected radiology reports in a deidentified electronic medical record database. Ground-truth classifications were assigned by expert orthopaedic traumatologists based on corresponding radiographs. Multiple prompting strategies were tested, including zero-shot prompting, zero-shot chain-of-thought prompting, and retrieval-augmented generation. We additionally asked the LLM to assign classification labels to “ideal” fracture descriptions written according to the 2018 OTA/AO Fracture and Dislocation Classification Compendium. Model performance was assessed using Cohen kappa and accuracy against ground-truth labels. Results: The 3 prompting strategies tested yielded similar classification performance on radiology report fracture descriptions, with almost perfect agreement at the bone and bone and location levels. Performance declined to slight agreement at the subgroup level. The best performance was observed using ideal fracture descriptions with retrieval-augmented generation, in which the agreement between the full LLM-generated and ground-truth labels remained moderate. Classification errors were largely due to imprecise descriptions, hallucinated information, or incorrect application of factually correct information. Conclusions: Our study demonstrates some potential for LLMs to translate free-text fracture descriptions into OTA/AO classifications, allowing for efficient labeling of large datasets of radiology reports. Future work should focus on refining model classification capabilities using more sophisticated prompting methods. Level of Evidence: Level III.
- 1 May 2026
- Published by Wolters Kluwer Health in OTA International
- Vol. 9 (2)
Open AccessDeriving the OTA/AO fracture classification from routinely collected radiology reports using a large language model
Objectives: Fracture classification plays a pivotal role in research and quality assurance; despite its wide acceptance, the OTA/AO classification is seldom documented in patients' electronic medical records, which impedes fracture registry creation and effective interdisciplinary communication. In this study, we investigate “off-the-shelf” large language models (LLMs) in translating free text in radiology reports into OTA/AO classification labels. Methods: We employed a Health Insurance Portability and Accountability Act-compliant LLM to classify 109 fracture descriptions from randomly selected radiology reports in a deidentified electronic medical record database. Ground-truth classifications were assigned by expert orthopaedic traumatologists based on corresponding radiographs. Multiple prompting strategies were tested, including zero-shot prompting, zero-shot chain-of-thought prompting, and retrieval-augmented generation. We additionally asked the LLM to assign classification labels to “ideal” fracture descriptions written according to the 2018 OTA/AO Fracture and Dislocation Classification Compendium. Model performance was assessed using Cohen kappa and accuracy against ground-truth labels. Results: The 3 prompting strategies tested yielded similar classification performance on radiology report fracture descriptions, with almost perfect agreement at the bone and bone and location levels. Performance declined to slight agreement at the subgroup level. The best performance was observed using ideal fracture descriptions with retrieval-augmented generation, in which the agreement between the full LLM-generated and ground-truth labels remained moderate. Classification errors were largely due to imprecise descriptions, hallucinated information, or incorrect application of factually correct information. Conclusions: Our study demonstrates some potential for LLMs to translate free-text fracture descriptions into OTA/AO classifications, allowing for efficient labeling of large datasets of radiology reports. Future work should focus on refining model classification capabilities using more sophisticated prompting methods. Level of Evidence: Level III.
- 4 March 2026
- Published by Wolters Kluwer Health in Journal of Bone and Joint Surgery
- Vol. 108 (10) , 701-706
Early Subspecialization in Orthopaedic Residency: Pathway to Expertise or Compromise in Training
Orthopaedic surgery residency is progressively shifting toward earlier and more extensive subspecialization, with residents seeking increased exposure to their intended subspecialty in order to gain experience and to increase their readiness for a fellowship. This exposure may involve formal 6 to 8-week elective rotations or flexible, customizable schedules. Although early subspecialization is well established in other surgical fields, its role in producing broadly competent orthopaedic surgeons is still largely debated. At the 2025 American Orthopaedic Association (AOA) Annual Leadership Meeting, a symposium entitled “Early Subspecialization in Orthopaedic Residency: Pathway to Expertise or Compromise in Training” brought together 5 orthopaedic leaders to discuss their perspectives on this evolving trend. In conjunction, a survey was distributed to symposium attendees and AOA members to evaluate current residency structures, attitudes toward early subspecialization, perceived impacts on general competence, and the preparedness of fellows. Quantitative survey data were analyzed using descriptive statistics, and the results were supplemented with qualitative insights from the panel discussion. Of 209 total respondents, 43% reported that their program provides subspecialty-focused opportunities, most often through elective rotations. Only 17.5% of respondents felt that early subspecialization produces safer and more skilled surgeons, whereas 83% believed that it compromises general orthopaedic competence. Similarly, 74% did not feel that subspecialization better prepares residents for practice. However, 75% of fellowship educators also reported working with underprepared fellows; of those, 57% noted that this influenced their perceptions of residency programs. Symposium discussions highlighted recurring themes, such as the educational advantages of targeted rotations, the risks of narrowing exposure too early, competency-based training as a potential framework, and workforce implications. Early subspecialization in orthopaedic surgery residency reflects broader trends in surgical education; however, within orthopaedics, there is continued discourse on how to balance breadth and depth of training. Proponents argue that early subspecialization may result in confidence and readiness for a fellowship, whereas critics highlight concerns over general training and call preparedness. Future efforts should focus on structured, competency-based models that balance subspecialty development with the essential goal of producing broadly capable orthopaedic surgeons.
- 4 March 2026
- Published by Wolters Kluwer Health in Journal of Bone and Joint Surgery
- Vol. 108 (10) , 701-706
Early Subspecialization in Orthopaedic Residency: Pathway to Expertise or Compromise in Training
Orthopaedic surgery residency is progressively shifting toward earlier and more extensive subspecialization, with residents seeking increased exposure to their intended subspecialty in order to gain experience and to increase their readiness for a fellowship. This exposure may involve formal 6 to 8-week elective rotations or flexible, customizable schedules. Although early subspecialization is well established in other surgical fields, its role in producing broadly competent orthopaedic surgeons is still largely debated. At the 2025 American Orthopaedic Association (AOA) Annual Leadership Meeting, a symposium entitled “Early Subspecialization in Orthopaedic Residency: Pathway to Expertise or Compromise in Training” brought together 5 orthopaedic leaders to discuss their perspectives on this evolving trend. In conjunction, a survey was distributed to symposium attendees and AOA members to evaluate current residency structures, attitudes toward early subspecialization, perceived impacts on general competence, and the preparedness of fellows. Quantitative survey data were analyzed using descriptive statistics, and the results were supplemented with qualitative insights from the panel discussion. Of 209 total respondents, 43% reported that their program provides subspecialty-focused opportunities, most often through elective rotations. Only 17.5% of respondents felt that early subspecialization produces safer and more skilled surgeons, whereas 83% believed that it compromises general orthopaedic competence. Similarly, 74% did not feel that subspecialization better prepares residents for practice. However, 75% of fellowship educators also reported working with underprepared fellows; of those, 57% noted that this influenced their perceptions of residency programs. Symposium discussions highlighted recurring themes, such as the educational advantages of targeted rotations, the risks of narrowing exposure too early, competency-based training as a potential framework, and workforce implications. Early subspecialization in orthopaedic surgery residency reflects broader trends in surgical education; however, within orthopaedics, there is continued discourse on how to balance breadth and depth of training. Proponents argue that early subspecialization may result in confidence and readiness for a fellowship, whereas critics highlight concerns over general training and call preparedness. Future efforts should focus on structured, competency-based models that balance subspecialty development with the essential goal of producing broadly capable orthopaedic surgeons.
- Open Access
- 1 October 2025
- Published by Wolters Kluwer Health in OTA International
- Vol. 8 (6S)
Musculoskeletal trauma rehabilitation: a global comparison
Rehabilitation after musculoskeletal trauma is globally acknowledged as essential for restoring function and improving quality of life; however, significant variation exists in rehabilitation frameworks and services worldwide. Although certain principles—such as early intervention, structured care pathways, and multidisciplinary approaches—are universally recognized as important determinants of outcomes in trauma patients, their implementation and effectiveness vary considerably due to differences in health care infrastructure, economic resources, policy frameworks, and cultural contexts. Drawing on comparative analyses of rehabilitation frameworks across regions, including North America, Latin America, Europe, Asia-Pacific, and Africa, this paper identifies structural variations, systemic challenges, and innovations that can inform global rehabilitation efforts. Strategic international collaboration focusing on shared learning, standardized protocols, technology integration, and targeted investments is key to creating equitable and sustainable rehabilitation models, especially in resource-limited settings and aging populations.
- 1 October 2025
- Published by Wolters Kluwer Health in OTA International
- Vol. 8 (6S)
Open AccessMusculoskeletal trauma rehabilitation: a global comparison
Rehabilitation after musculoskeletal trauma is globally acknowledged as essential for restoring function and improving quality of life; however, significant variation exists in rehabilitation frameworks and services worldwide. Although certain principles—such as early intervention, structured care pathways, and multidisciplinary approaches—are universally recognized as important determinants of outcomes in trauma patients, their implementation and effectiveness vary considerably due to differences in health care infrastructure, economic resources, policy frameworks, and cultural contexts. Drawing on comparative analyses of rehabilitation frameworks across regions, including North America, Latin America, Europe, Asia-Pacific, and Africa, this paper identifies structural variations, systemic challenges, and innovations that can inform global rehabilitation efforts. Strategic international collaboration focusing on shared learning, standardized protocols, technology integration, and targeted investments is key to creating equitable and sustainable rehabilitation models, especially in resource-limited settings and aging populations.
- Open Access
- 24 September 2025
- Published by Wolters Kluwer Health in OTA International
- Vol. 8 (4)
Biomechanical comparison of cephalomedullary and reconstruction nails used in the treatment of subtrochanteric femur fractures
Objectives: There are several studies comparing various implants for treating subtrochanteric femur fractures. However, there is no known comparison of the DePuy Synthes TFN-ADVANCED and the DePuy Synthes FRN-ADVANCED. This study compares the biomechanical characteristics of 2 nail-composite femur models designed to simulate reconstructed subtrochanteric femur fractures. Methods: Eight composite femora were osteotomized to produce models of subtrochanteric femur fractures (Arbeitsgemeinschaft für Osteosynthesefragen/Orthopaedic Trauma Association 32-C3.i). The simulated fractures were fixed using either a DePuy Synthes TFN-ADVANCED (1 Screw) proximal femoral nailing system or the DePuy Synthes FRN-ADVANCED (2 screw) femoral recon nailing system. The reconstructed fractures were subjected to compressive mechanical testing protocol consisting of cyclic loading and destructive ramp-to-failure. Interfragmentary motion of the proximal and distal bone segments was tracked at multiple locations using 3-dimensional motion-capture. Results: The TFN-ADVANCED withstood a significantly higher maximum force (3512.5 N ± 261.1 vs. 3055.22 N ± 174.1, P = 0.027) than the FRN-ADVANCED and difference in energy to failure approached significance (35.6 J ± 12.01 vs. 18.30 J ± 7.50, P = 0.050). The TFN-ADVANCED also exhibited higher magnitudes of proximal segment displacement at the greater trochanter (7.65 mm vs. 4.72 mm, P = 0.045) but not at the distal segment (3.52 mm vs. 2.62 mm, P = 0.205). There were no differences in overall construct stiffness (506.81 N/mm ± 81.7 vs. 409.74 N/mm ± 112.3, P = 0.212). In all models, the implant underwent plastic deformation as the mode of failure. Conclusions: In simulated subtrochanteric femur fractures, the TFN-ADVANCEDTM implant is stronger and allows more proximal segment motion before failure compared with the FRN-ADVANCEDTM. In addition to strength and deformation, the nails exhibited different failure mechanisms with the TFNA failing proximally and the FRNA failing distally. Osteotomy site motion was the same, which is clinically relevant when considering strain at a healing fracture. Level of Evidence: III Biomechanical Study.
- 24 September 2025
- Published by Wolters Kluwer Health in OTA International
- Vol. 8 (4)
Open AccessBiomechanical comparison of cephalomedullary and reconstruction nails used in the treatment of subtrochanteric femur fractures
Objectives: There are several studies comparing various implants for treating subtrochanteric femur fractures. However, there is no known comparison of the DePuy Synthes TFN-ADVANCED and the DePuy Synthes FRN-ADVANCED. This study compares the biomechanical characteristics of 2 nail-composite femur models designed to simulate reconstructed subtrochanteric femur fractures. Methods: Eight composite femora were osteotomized to produce models of subtrochanteric femur fractures (Arbeitsgemeinschaft für Osteosynthesefragen/Orthopaedic Trauma Association 32-C3.i). The simulated fractures were fixed using either a DePuy Synthes TFN-ADVANCED (1 Screw) proximal femoral nailing system or the DePuy Synthes FRN-ADVANCED (2 screw) femoral recon nailing system. The reconstructed fractures were subjected to compressive mechanical testing protocol consisting of cyclic loading and destructive ramp-to-failure. Interfragmentary motion of the proximal and distal bone segments was tracked at multiple locations using 3-dimensional motion-capture. Results: The TFN-ADVANCED withstood a significantly higher maximum force (3512.5 N ± 261.1 vs. 3055.22 N ± 174.1, P = 0.027) than the FRN-ADVANCED and difference in energy to failure approached significance (35.6 J ± 12.01 vs. 18.30 J ± 7.50, P = 0.050). The TFN-ADVANCED also exhibited higher magnitudes of proximal segment displacement at the greater trochanter (7.65 mm vs. 4.72 mm, P = 0.045) but not at the distal segment (3.52 mm vs. 2.62 mm, P = 0.205). There were no differences in overall construct stiffness (506.81 N/mm ± 81.7 vs. 409.74 N/mm ± 112.3, P = 0.212). In all models, the implant underwent plastic deformation as the mode of failure. Conclusions: In simulated subtrochanteric femur fractures, the TFN-ADVANCEDTM implant is stronger and allows more proximal segment motion before failure compared with the FRN-ADVANCEDTM. In addition to strength and deformation, the nails exhibited different failure mechanisms with the TFNA failing proximally and the FRNA failing distally. Osteotomy site motion was the same, which is clinically relevant when considering strain at a healing fracture. Level of Evidence: III Biomechanical Study.
- 17 September 2025
- Published by Wolters Kluwer Health in Journal of the American Academy of Orthopaedic Surgeons
- Vol. 34 (6) , e873-e884
Periprosthetic Femur Fractures in Hemiarthroplasty Correlated With Stem Type: An Analysis From the American Joint Replacement Registry
Background: Periprosthetic femur fractures are a known complication after hip hemiarthroplasty (HA) in geriatric patients. The relationship between femoral stem design and fracture risk remains unclear. This study aimed to assess (1) the association between stem geometry and fixation on periprosthetic femur fracture risk and (2) the effect of a femoral stem collar. Methods: The American Joint Replacement Registry data were analyzed for HA cases in patients aged 70+ years from 2012 to 2021. We identified 56,828 primary HAs for the diagnosis of femoral neck fracture. Patient demographics and revision surgery for periprosthetic femur fracture in the form of open reduction with internal fixation or revision arthroplasty were documented. Stems were categorized into tapered wedge or fit and fill, and cemented stems were categorized into composite beam or taper slip. Cox models and Benjamini-Hochberg adjustments were used for statistical analysis. Results: Cemented composite beam stems accounted for 38% of implants, followed by fit-and-fill stems at 32%, tapered wedge stems at 29%, and cemented taper slip stems at 1%. A collar was present in all cemented composite beam stems, 23% of tapered wedge stems, and 20% of fit-and-fill stems. After applying inverse probability weighting adjustment, cemented composite beam stems showed a markedly lower periprosthetic femur fracture risk over other stems. Compared with cemented composite beam stems, collarless tapered wedge demonstrated a markedly higher hazard ratio (HR) (3.63; P < 0.001), as did both collarless (HR, 3.4; P < 0.001) and collared fit-and-fill stems (HR, 3.45; P < 0.001). The presence of a collar reduced periprosthetic femur fracture risk across all designs, with the most pronounced reduction seen in tapered wedge stems (0.90% versus 0.41%). Conclusion: For HA in patients aged 70 years and older, cemented composite beam stems with a collar were associated with the lowest risk of periprosthetic femur fractures in the American Joint Replacement Registry database. However, these findings reflect implant utilization patterns up to 2021 and may not fully account for the recent rise in triple taper collared stem use. Surgeons should consider these implants when selecting fixation strategies for HA. Level of Evidence: Level III
- 17 September 2025
- Published by Wolters Kluwer Health in Journal of the American Academy of Orthopaedic Surgeons
- Vol. 34 (6) , e873-e884
Periprosthetic Femur Fractures in Hemiarthroplasty Correlated With Stem Type: An Analysis From the American Joint Replacement Registry
Background: Periprosthetic femur fractures are a known complication after hip hemiarthroplasty (HA) in geriatric patients. The relationship between femoral stem design and fracture risk remains unclear. This study aimed to assess (1) the association between stem geometry and fixation on periprosthetic femur fracture risk and (2) the effect of a femoral stem collar. Methods: The American Joint Replacement Registry data were analyzed for HA cases in patients aged 70+ years from 2012 to 2021. We identified 56,828 primary HAs for the diagnosis of femoral neck fracture. Patient demographics and revision surgery for periprosthetic femur fracture in the form of open reduction with internal fixation or revision arthroplasty were documented. Stems were categorized into tapered wedge or fit and fill, and cemented stems were categorized into composite beam or taper slip. Cox models and Benjamini-Hochberg adjustments were used for statistical analysis. Results: Cemented composite beam stems accounted for 38% of implants, followed by fit-and-fill stems at 32%, tapered wedge stems at 29%, and cemented taper slip stems at 1%. A collar was present in all cemented composite beam stems, 23% of tapered wedge stems, and 20% of fit-and-fill stems. After applying inverse probability weighting adjustment, cemented composite beam stems showed a markedly lower periprosthetic femur fracture risk over other stems. Compared with cemented composite beam stems, collarless tapered wedge demonstrated a markedly higher hazard ratio (HR) (3.63; P < 0.001), as did both collarless (HR, 3.4; P < 0.001) and collared fit-and-fill stems (HR, 3.45; P < 0.001). The presence of a collar reduced periprosthetic femur fracture risk across all designs, with the most pronounced reduction seen in tapered wedge stems (0.90% versus 0.41%). Conclusion: For HA in patients aged 70 years and older, cemented composite beam stems with a collar were associated with the lowest risk of periprosthetic femur fractures in the American Joint Replacement Registry database. However, these findings reflect implant utilization patterns up to 2021 and may not fully account for the recent rise in triple taper collared stem use. Surgeons should consider these implants when selecting fixation strategies for HA. Level of Evidence: Level III
- 5 September 2025
- Published by Wolters Kluwer Health in Journal of Bone and Joint Surgery
- Vol. 107 (20) , 2334-2338
Resident Unions in Orthopaedic Surgery
Across the United States, there is an increasing trend among medical residents to join unions. Even though these entities have existed for almost 100 years, escalating concerns regarding compensation, working conditions, and other factors have influenced thousands of residents to join unions. Unionization can both positively and negatively affect residents, hospital networks, and the communities that the system serves. Most resident unions are nationally organized, but the laws and rules governing these bodies vary among the 50 states. The individual states regulate the collective bargaining of the union with the health-care system, and physician leaders should be aware of the nuances regarding unions at their workplace setting.
- 5 September 2025
- Published by Wolters Kluwer Health in Journal of Bone and Joint Surgery
- Vol. 107 (20) , 2334-2338
Resident Unions in Orthopaedic Surgery
Across the United States, there is an increasing trend among medical residents to join unions. Even though these entities have existed for almost 100 years, escalating concerns regarding compensation, working conditions, and other factors have influenced thousands of residents to join unions. Unionization can both positively and negatively affect residents, hospital networks, and the communities that the system serves. Most resident unions are nationally organized, but the laws and rules governing these bodies vary among the 50 states. The individual states regulate the collective bargaining of the union with the health-care system, and physician leaders should be aware of the nuances regarding unions at their workplace setting.