Orthopedics最新文献

Background: Artificial intelligence (AI) generative large language models are powerful and increasingly accessible tools with potential applications in health care education and training. The annual Orthopaedic In-Training Examination (OITE) is widely used to assess resident academic progress and preparation for the American Board of Orthopaedic Surgery Part 1 Examination.

Materials and methods: Open AI's ChatGPT and Google's Bard generative language models were administered the 2022 OITE. Question stems that contained images were input without and then with a text-based description of the imaging findings.

Results: ChatGPT answered 69.1% of questions correctly. When provided with text describing accompanying media, this increased to 77.8% correct. In contrast, Bard answered 49.8% of questions correctly. This increased to 58% correct when text describing imaging in question stems was provided (P<.0001). ChatGPT was most accurate in questions within the shoulder category, with 90.9% correct. Bard performed best in the sports category, with 65.4% correct. ChatGPT performed above the published mean of Accreditation Council for Graduate Medical Education orthopedic resident test-takers (66%).

Conclusion: There is significant variability in the accuracy of publicly available AI models on the OITE. AI generative language software may play numerous potential roles in the future in orthopedic education, including simulating patient presentations and clinical scenarios, customizing individual learning plans, and driving evidence-based case discussion. Further research and collaboration within the orthopedic community is required to safely adopt these tools and minimize risks associated with their use. [Orthopedics. 2024;47(3):e146-e150.].

Background: With pressures to decrease the financial burden of total hip arthroplasty (THA), it is imperative to understand the cost drivers of this procedure. This study evaluated operative and total encounter costs for two surgical approaches to THA-posterior (P) and direct anterior (DA).

Materials and methods: A total of 233 THAs (134 P and 99 DA) performed by two fellowship-trained arthroplasty surgeons from 2017 to 2022 were reviewed. Demographics, comorbidities, mobility status, operative time, length of stay, implants used, discharge location, and complications until final follow-up were recorded. Total encounter cost was collected and itemized. Multivariable regression analyses evaluated predictors of cost.

Results: There were differences in age (67 years for DA and 63 years for P; P=.03), body mass index (28.0 kg/m² for DA and 33.8 kg/m² for P; P<.01), Elixhauser Comorbidity Index score (4.6 for DA and 5.6 for P; P=.04), and operative time (2.1 hours for DA and 1.9 hours for P; P<.01) between the two cohorts. The DA cohort trended toward shorter length of stay, with the highest percentage of patients discharged home (86.9%; P=.02). The P cohort had the lowest encounter ($9601 for DA and $9100 for P; P=.20) and intraoperative (including implant used) ($7268 for DA and $6792 for P; P<.01) costs. The DA cohort had a significantly higher cost of radiology during the encounter ($244; P<.01). Regression analysis demonstrated that length of stay and DA approach were both predictors of increased encounter cost.

Conclusion: The DA cohort had improved measures of health; however, this approach was associated with a higher operative cost and was predictive of increased encounter cost despite a shorter length of stay. [Orthopedics. 2024;47(3):e151-e156.].

Background: Despite increasing attention, disparities in outcomes for Black and Hispanic patients undergoing orthopedic surgery are widening. In other racial-ethnic minority groups, outcomes often go unreported. We sought to quantify disparities in surgical outcomes among Asian, American Indian or Alaskan Native, and Native Hawaiian or Pacific Islander patients across multiple orthopedic subspecialties.

Materials and methods: The National Surgical Quality Improvement Program was queried to identify all surgical procedures performed by an orthopedic surgeon from 2014 to 2020. Multivariable logistic regression models were used to investigate the impact of race and ethnicity on 30-day medical complications, readmission, reoperation, and mortality, while adjusting for orthopedic subspecialty and patient characteristics.

Results: Across 1,512,480 orthopedic procedures, all patients who were not White were less likely to have arthroplasty-related procedures (P<.001), and Hispanic, Asian, and American Indian or Alaskan Native patients were more likely to have trauma-related procedures (P<.001). American Indian or Alaskan Native (adjusted odds ratio [AOR], 1.005; 95% CI, 1.001-1.009; P=.011) and Native Hawaiian or Pacific Islander (AOR, 1.009; 95% CI, 1.005-1.014; P<.001) patients had higher odds of major medical complications compared with White patients. American Indian or Alaskan Native patients had higher risk of reoperation (AOR, 1.005; 95% CI, 1.002-1.008; P=.002) and Native Hawaiian or Pacific Islander patients had higher odds of mortality (AOR, 1.003; 95% CI, 1.000-1.005; P=.019) compared with White patients.

Conclusion: Disparities regarding surgical outcome and utilization rates persist across orthopedic surgery. American Indian or Alaskan Native and Native Hawaiian or Pacific Islander patients, who are under-represented in research, have lower rates of arthroplasty but higher odds of medical complication, reoperation, and mortality. This study highlights the importance of including these patients in orthopedic research to affect policy-related discussions. [Orthopedics. 2024;47(3):e131-e138.].

Finger amputations are commonly encountered. These may be revised in the emergency department (ED) or the operating room (OR). Previous studies have demonstrated the cost-effectiveness associated with procedures performed in the ED. Patient outcomes have not been described. We retrospectively reviewed patients who presented to our level 1 trauma center with a traumatic partial or complete finger amputation through flexor tendon zone I. All were treated with revision amputation performed in either the ED or the OR between January 2012 and December 2017. A total of 172 patient charts were included. Ninety-three of the revision amputations were performed in the ED, while 79 were performed in the OR. There was no difference in age, race, sex, having a manual labor job, medical comorbidities, or mechanism of injury between the groups. Compared with procedures performed in the ED, procedures performed in the OR had a higher rate of delayed healing, a longer stay in the hospital, and a higher referral to therapy postoperatively. Length of follow-up and number of follow-up visits were not statistically different based on location of procedure. There was no difference in post-procedural infection rate or need for revision procedure between the groups. Our data support the efficacy of performing revision amputation procedures in the ED. Recorded patient complications and subsequent treatment after revision amputations performed in the ED vs the OR were comparable. Those performed in the ED potentially decrease the burden placed on the patient and the health care system. [Orthopedics. 2024;47(3):152-156.].

This study compared the results of endoscopic cubital tunnel release (eCuTR) with those of open cubital tunnel release (oCuTR) for the management of cubital tunnel syndrome (CuTS). In this retrospective study, 35 patients underwent eCuTR or oCuTR. Group I and group II consisted of 16 patients undergoing eCuTR and 19 patients undergoing oCuTR, respectively. Patients were asked to report paresthesia and pain, and electromyography was performed. The Dellon and Bishop classifications were used. The Disabilities of the Arm, Shoulder and Hand (DASH) and visual analog scale (VAS) pain scores were recorded, as well as the key pinch strength and two-point discrimination. The incision length and operation duration were noted. The mean follow-up was 39 months. The mean operating time was longer in the endoscopy group (43 vs 22 minutes). Overall, 34.3% (n=12) of the cases were classified as Dellon grade II and 65.7% (n=23) were classified as Dellon grade III. According to the Bishop score, excellent or good results were obtained for 75% of the patients in the eCuTR group and 78.9% of the patients in the oCuTR group. In the eCuTR and oCuTR groups, all outcome measures improved after surgery: DASH score (preoperative, 37.7 vs 30.7; postoperative, 15.4 vs 20), VAS score (preoperative, 7.8 vs 7.3; postoperative, 4.3 vs 4.1), pinch strength (preoperative, 74 vs 66; postoperative, 93 vs 84), and two-point discrimination (preoperative, 5.6 vs 6.6; postoperative, 4.9 vs 4.5). No significant difference was apparent between the two techniques in outcomes. However, the endoscopic release had a higher reoperation rate and took twice as long to perform despite having a shorter incision. [Orthopedics. 2024;47(3):e119-e124.].

Background: Identification of the hip center of rotation (HCOR) before total hip arthroplasty (THA) is crucial for achieving optimal implant position and size, and for restoring native biomechanics around the hip joint. Current techniques for determining the HCOR in cases of abnormal hip anatomy are limited and unreliable. This study presents a novel technique using open-access software for preoperative THA templating for patients with significantly abnormal hip anatomy due to unilateral hip arthrosis. The aim is to reliably predict the HCOR and acetabular implant size compared with a traditional intraoperative method.

Materials and methods: This retrospective study involved 20 patients with unilateral hip arthrosis who underwent THA. Preoperative templating was performed using the experimental technique, and the position of the HCOR was measured on preoperative and postoperative radiographs. The positions of the predicted and actual HCOR were compared, along with the inclination and size of the acetabular component.

Results: The difference between the predicted and actual HCOR positions was insignificant (0.43±0.22 mm vertically and 0.18±0.20 mm horizontally), and there was a positive correlation between them (r=0.78, P<.005; r=0.72, P<.005). The agreement between the predicted and actual acetabular implant sizes was 85%, with near-perfect interobserver agreement (Cohen's kappa=0.827).

Conclusion: This novel technique provides a reliable method for predicting HCOR and acetabular implant size for THA in cases of unilateral hip arthrosis. This technique may help optimize biomechanics and improve outcomes in challenging cases. Further research and validation are warranted to establish its broader applicability. [Orthopedics. 2024;47(3):e139-e145.].

Objective: Literature reviews are essential to the scientific process and allow clinician researchers to advance general knowledge. The purpose of this study was to evaluate if the artificial intelligence (AI) programs ChatGPT and Perplexity.AI can perform an orthopedic surgery literature review.

Materials and methods: Five different search topics of varying specificity within orthopedic surgery were chosen for each search arm to investigate. A consolidated list of unique articles for each search topic was recorded for the experimental AI search arms and compared with the results of the control arm of two independent reviewers. Articles in the experimental arms were examined by the two independent reviewers for relevancy and validity.

Results: ChatGPT was able to identify a total of 61 unique articles. Four articles were not relevant to the search topic and 51 articles were deemed to be fraudulent, resulting in 6 valid articles. Perplexity.AI was able to identify a total of 43 unique articles. Nineteen were not relevant to the search topic but all articles were able to be verified, resulting in 24 valid articles. The control arm was able to identify 132 articles. Success rates for ChatGPT and Perplexity. AI were 4.6% (6 of 132) and 18.2% (24 of 132), respectively.

Conclusion: The current iteration of ChatGPT cannot perform a reliable literature review, and Perplexity.AI is only able to perform a limited review of the medical literature. Any utilization of these open AI programs should be done with caution and human quality assurance to promote responsible use and avoid the risk of using fabricated search results. [Orthopedics. 2024;47(3):e125-e130.].

Objective: Treating high-level athletes involves a balance between early and safe return to play. Various types of protective immobilization have been recommended after operatively treated Bennett's fracture. The purpose of this study was to investigate if hand-based immobilization offers protection equivalent to forearm-based immobilization.

Materials and methods: A cadaveric model of Bennett's fracture was created in 8 fresh-frozen, cadaveric forearms. Osteosynthesis was performed using a single headless compression screw. Three matched pairs were casted in either hand-based or forearm length, thumb spica casts, while 2 specimens remained un-casted as controls. Specimens were mounted on a custom testing apparatus. Weights were added in 6.8-kg increments until fixation failed and the fracture displaced. Fluoroscopy was performed after each trial. We used the Kruskal-Wallis non-parametric test to compare the groups. We considered P<.05 statistically significant.

Results: Failure of fixation occurred at 6.8 kg in the control specimens. Fixation failed in hand-based and forearm length casts at a mean of 18.1±5.1 kg. We did not find a statistically significant difference between median values of load at failure in kilograms across control specimens and 2 immobilization categories (P=.114). All specimens in the hand-based group sustained additional wrist injuries, while no additional injuries were noted in the forearm length group.

Conclusion: Our study results showed that hand-based immobilization provides equivalent protection against fixation failure for operatively treated Bennett's fractures but may predispose athletes to increased risk of wrist injury compared with traditional, forearm-based casting. [Orthopedics. 2024;47(3):157-160.].

Objective: This study sought to understand trends in industry payments for research awarded to orthopedic surgeons.

Materials and methods: The Centers for Medicare & Medicaid Services Open Payments database was queried for the years 2016 to 2021 for industry payments for research. Financial analyses were performed to understand temporal trends and differences by orthopedic subspecialty and principal investigator characteristics such as sex. The threshold for statistical significance was set at .05.

Results: A total of 2014 orthopedic surgeons were identified, among whom 542 adult reconstruction (27%) and 460 sports medicine (23%) surgeons were major beneficiaries. Seventy-one female orthopedic surgeons comprised the minority (4%). Total research payments awarded during the study period aggregated to $266,633,592, with adult reconstruction ($88,819,047; 33%) and sports medicine ($57,949,822; 22%) receiving the highest amounts. Total research payments awarded trended upward yearly except for a decline in 2020 that subsequently rebounded (P<.001). Median annual research payment per orthopedic surgeon was $13,375. Median total industry payments per orthopedic surgeon differed between specialties (P <.001), with the highest amounts for adult reconstruction ($44,063) and sports medicine ($34,567) and the lowest amounts for hand ($12,052) and foot and ankle ($19,233). Median total payments did not differ significantly when stratified by sex (P=.276) and region (P=.906). Specialties in which the respective top three companies offered the majority of the research funding were musculoskeletal oncology (90%), pediatric orthopedics (66%), and shoulder and elbow (64%).

Conclusion: These results can be used as a primer for orthopedic surgeons seeking to leverage industry relationships to fund translational research. [Orthopedics. 2024;47(3):172-178.].