Comparison of Artificial Intelligence to Resident Performance on Upper-Extremity Orthopaedic In-Training Examination Questions.

Purpose: Currently, there is a paucity of prior investigations and studies examining applications for artificial intelligence (AI) in upper-extremity (UE) surgical education. The purpose of this investigation was to assess the performance of a novel AI tool (ChatGPT) on UE questions on the Orthopaedic In-Training Examination (OITE). We aimed to compare the performance of ChatGPT to the examination performance of hand surgery residents. Methods: We selected questions from the 2020-2022 OITEs that focused on both the hand and UE as well as the shoulder and elbow content domains. These questions were divided into two categories: those with text-only prompts (text-only questions) and those that included supplementary images or videos (media questions). Two authors (B.K.F. and G.S.M.) converted the accompanying media into text-based descriptions. Included questions were inputted into ChatGPT (version 3.5) to generate responses. Each OITE question was entered into ChatGPT three times: (1) open-ended response, which requested a free-text response; (2) multiple-choice responses without asking for justification; and (3) multiple-choice response with justification. We referred to the OITE scoring guide for each year in order to compare the percentage of correct AI responses to correct resident responses. Results: A total of 102 UE OITE questions were included; 59 were text-only questions, and 43 were media-based. ChatGPT correctly answered 46 (45%) of 102 questions using the Multiple Choice No Justification prompt requirement (42% for text-based and 44% for media questions). Compared to ChatGPT, postgraduate year 1 orthopaedic residents achieved an average score of 51% correct. Postgraduate year 5 residents answered 76% of the same questions correctly. Conclusions: ChatGPT answered fewer UE OITE questions correctly compared to hand surgery residents of all training levels. Clinical relevance: Further development of novel AI tools may be necessary if this technology is going to have a role in UE education.

Unplanned Postoperative Phone Calls and Electronic Messages for Patients With and Without Opioid Prescriptions After Carpal Tunnel Release.

Purpose: The purpose of our study was to compare unplanned postoperative patient communication in the form of phone calls and/or electronic patient portal messages (EPPM) after carpal tunnel release (CTR) for patients with and without a postoperative opioid prescription. Methods: We identified all patients ≥18 years of age who underwent primary CTR between 2017 and 2022 without an opioid ordered within 90 days prior to surgery. The following two groups were created: cases with and without an opioid prescribed on the day of surgery. We recorded baseline demographics for all patients and recorded all unplanned communication (phone calls and EPPM) sent from a patient to the surgeon's office within 14 days after surgery. Unadjusted associations between unplanned communication and case characteristics were evaluated. Multiple logistic regression models were used to assess the relationship between opioid status and unplanned communication. Results: A total of 5,735 CTRs were included, and 54% of the patients were prescribed an opioid on the day of surgery. Forty-two percent of cases had unplanned postoperative communication, and 48.1% of cases, without an opioid prescription, had unplanned communication compared with 36.8% in the opioid group. Patients who were prescribed opioids were 0.62 times less likely to contact the surgeon's office via phone calls or EPPM (95% confidence interval [CI]: 0.56, 0.70). Increased age was associated with a reduction in the odds of unplanned contact (odds ratios [OR] = 0.95, 95% CI: 0.93, 0.97), whereas higher body mass index was significantly associated with increased communication (OR = 1.05, 95% CI: 1.01, 1.09). Conclusions: Patients prescribed opioids after CTR are 0.62 times less likely to contact the surgeon's office after surgery. Considering the 11% increase in unplanned postoperative communication after CTR, surgeons should consider alternative methods that have previously been demonstrated to reduce opioid consumption. Type of study/level of evidence: Prognostic II.

Sensitivity, Specificity, and Reliability of the CTS-6 for Carpal Tunnel Syndrome Administered by Medical Assistants.

PURPOSE: The CTS-6 can be used clinically to diagnose carpal tunnel syndrome (CTS) and has demonstrated high levels of interrater reliability when administered by nonexpert clinicians. Our purpose was to assess sensitivity (Sn), specificity (Sp), and interrater reliability of the CTS-6 when administered by medical assistants (MAs). METHODS: A series of patients presenting to an academic, upper-extremity surgery clinic were screened using CTS-6 between May and June of 2023. The CTS-6 was first administered by one of seven MAs and then by one of four fellowship-trained upper-extremity surgeons. In addition to recording baseline demographics, the results of each of the six history and examination components of the CTS-6 were recorded, as was the cumulative CTS-6 score (0-26). Surgeons were blinded to the scores obtained by the MAs. Interrater reliability (Cohen's kappa) was determined between the groups with respect to the diagnosis of CTS and the individual CTS-6 items. Sensitivity/specificity was calculated for the MA-administered CTS-6, using the score obtained by the surgeon as the reference standard. A CTS-6 score >12 was considered diagnostic of CTS. RESULTS: Two hundred eighteen patients were included, and 26% had a diagnosis of CTS. The MA group demonstrated a Sn/Sp of 84%/91% for the diagnosis of CTS. Interrater reliability was substantial (Cohen's kappa: 0.72, 95% confidence interval: 0.62-0.83) for MAs compared with hand surgeons for the diagnosis of CTS. For individual CTS-6 components, agreement was lowest for the assessment of two-point discrimination (fair) and highest for the assessment of subjective numbness (near perfect). CONCLUSIONS: The CTS-6 demonstrates substantial reliability and high Sn/Sp when administrated by MAs in an upper-extremity clinic. These data may be used to inform the development of CTS screening programs and future investigations in the primary care setting. TYPE OF STUDY/LEVEL OF EVIDENCE: Diagnostic II.

Surgeon Factors and Trends Associated With the Use of Subacromial Decompression at the Time of Rotator Cuff Repair.

PURPOSE: Subacromial decompression (SAD) has historically been described as an essential part of the surgical treatment of rotator cuff disorders. However, investigations throughout the 21st century have increasingly questioned the need for routine SAD during rotator cuff repair (RCR). Our purpose was to assess for changes in the incidence of SAD performed during RCR over a 12-year period. In addition, we aimed to characterize surgeon and practice factors associated with SAD use. METHODS: Records from two large tertiary referral systems in the United States from 2010 to 2021 were reviewed. All cases of RCR with and without SAD were identified. The outcome of interest was the proportion of SAD performed during RCR across years and by surgeon. Surgeon-specific characteristics included institution, fellowship training, surgical volume, academic practice, and years in practice. Yearly trends were assessed using binomial logistic regression modeling, with a random effect accounting for surgeon-specific variability. RESULTS: During the study period, 37,165 RCR surgeries were performed by 104 surgeons. Of these cases, 71% underwent SAD during RCR. SAD use decreased by 11%. The multivariable model found that surgeons in academic practice, those with lower surgical volume, and those with increasing years in practice were significantly associated with increased odds of performing SAD. Surgeons with fellowship training were significantly more likely to use SAD over time, with the greatest odds of SAD noted for sports medicine surgeons (odds ratio = 3.04). CONCLUSIONS: Although SAD use during RCR appears to be decreasing, multiple surgeon and practice factors (years in practice, fellowship training, volume, and academic practice) are associated with a change in SAD use. CLINICAL RELEVANCE: These data suggest that early-career surgeons entering practice are likely driving the trend of declining SAD. Despite evidence suggesting limited clinical benefits, SAD remains commonly performed; future studies should endeavor to determine factors associated with practice changes among surgeons.

A Comparison of Early Complication Rates of Endoscopic Carpal Tunnel Release With and Without Resident Involvement.

PURPOSE: The purpose of this study was to compare the complication rates of endoscopic carpal tunnel release (ECTR) relative to orthopedic resident trainee involvement in the procedure. METHODS: All patients undergoing isolated, elective ECTR by two attending surgeons within a 59-month period were analyzed. Cases were categorized as the following according to the degree of resident involvement: ECTR performed by attending with either no resident or a resident as an assistant (group 1), resident performing a portion of the procedure (group 2), and resident performing the entire procedure (group 3). Early postoperative complications and/or intraoperative conversion to an open procedure were the outcomes of interest. We used a noninferiority design, hypothesizing that resident involvement would not be associated with inferior outcomes compared with cases without resident involvement. Multiple logistic regression models, adjusted for patient demographic and surgical characteristics, were fit to assess the relationship between resident involvement groups and complication/conversion outcomes. RESULTS: A total of 1,167 ECTR cases (895 patients) were performed and returned for postoperative follow-up for at least 2 weeks after surgery. Operative time was significantly shorter for group 1 cases versus group 2 and 3 cases. The early postoperative complication and conversion rates were 1.7% and 1.0%, respectively. Superficial infection (1.2%), deep infection (0.3%), and transient neuropraxia (0.3%) occurred infrequently and did not differ relative to resident involvement. No differences in the odds of complication and/or conversion relative to resident involvement were observed. CONCLUSIONS: The results of ECTR performed entirely or in part by attending-supervised resident trainees were not inferior to ECTR performed by an attending surgeon regarding the odds of experiencing complications or conversion to an open procedure. With appropriate supervision, ECTR can be performed safely by orthopedic and plastic surgery residents. TYPE OF STUDY/LEVEL OF EVIDENCE: Therapeutic II.

Statistical Literacy in Hand and Upper-Extremity Patients.

Purpose: Statistical literacy is the ability of a patient to apply basic statistical concepts to their health care. Understanding statistics is a critical component of shared decision making. The purpose of this investigation was to define levels of statistical literacy in an upper-extremity (UE) patient population. We aimed to determine if patient demographics would be associated with statistical literacy. Methods: An electronic survey was administered to a consecutive series of UE patients at a single institution. We recorded baseline demographics, Single Assessment Numeric Evaluation scores, the Berlin Numeracy Test (BNT), and General Health Numeracy Test. We also included a surgical risk question, which asked: "Approximately 3% of patients who get carpal tunnel surgery develop an infection. If 100 patients get this surgery, how many would you expect to develop an infection?" A covariate-controlled adjusted odds ratio reflecting the association between each statistical literacy outcome measure and patient characteristics was reported. Results: A total 254 surveys were administered, 148 of which were completed and included. Fifty percent of respondents had a high-school education or less. For the BNT, 78% scored in the bottom quartile, and 52% incorrectly answered all questions. For the General Health Numeracy Test, 34% answered 0 or 1/6 questions correctly. For the surgical risk question, 24% of respondents answered incorrectly. Respondents who had a college or graduate degree had 2.62 times greater odds (95% confidence interval, 1.09-6.32) of achieving a BNT score in a higher quartile than patients who did not have a college or graduate degree. Conclusions: Overall levels of statistical literacy are low for UE patients. Clinical relevance: When engaging in management discussions and shared decision making, UE surgeons should assume low levels of statistical literacy. Consideration of alternative formats, such as frequencies, video-based materials, and pictographs, may be warranted when discussing outcomes and risks of surgical procedures.

Defining Nonunion for Metacarpal Fractures: A Systematic Review.

Purpose: Our purpose was to assess how nonunion of the metacarpals has been defined in prior investigations with respect to both clinical and radiographic criteria. We hypothesized that the definitions of nonunion would be highly variable. Methods: A systematic review was conducted using MEDLINE and Embase databases for clinical articles related to the treatment of metacarpal fractures (surgical and nonsurgical) from 2010 to 2021. Included articles were searched to assess how nonunion was defined based on clinical and radiographic criteria. We assessed the treatment type, method of union assessment, time to union, and incidence of union as well as article factors such as the following: date of publication, level of evidence, and publishing journal. Results: A total of 641 articles were identified, of which 102 were included for a definition of nonunion and 97 were included for the assessment of clinical management and outcomes. Of the included articles, 62% contained level IV evidence. A definition of nonunion was provided in 47% of the articles. Radiographic criteria alone, clinical criteria alone, or a combination of the 2 was used in 22%, 6%, and 19% of the cases, respectively, to define nonunion. The most common definition of nonunion was presence of fracture-site tenderness (with no time defined) in 20 articles (20%), followed by lack of radiographic healing at 6 months (15%). In the 97 included articles, the total number of fracture cases was 4,435 and nonunion was reported in 0.45%. Cases with nonunion were reported in a total of six articles that used a variety of treatment modalities. Conclusions: The definition of metacarpal nonunion remains highly variable and lacks standardization with respect to clinical and radiographic criteria. Clinical relevance: Standardizing the definition of nonunion for metacarpal fractures would allow for more accurate assessments of the incidence of this complication and may aid in improving diagnostic and management strategies.

Analysis of Terminated Hand and Wrist-Related Clinical Trials.

Purpose: While clinical trials provide high-quality evidence guiding medical decision-making, early trial termination can result in both lost time and resources. Our purpose was to investigate the rate of and reasons for clinical trial termination for hand and wrist-related conditions and identify study characteristics associated with early trial termination. Methods: The ClinicalTrials.gov database was queried for all hand and wrist-related clinical trials. All terminated and completed trials were reviewed, with characteristics and reasons for termination recorded. Study characteristics included type, purpose, intervention assessed, enrollment, group allocation, blinding, trial phase, sponsor type, and geographic region. Chi-square test was used to identify associations between trial characteristics and terminated versus completed status. Results: A total of 793 hand and wrist-related clinical trials were identified, with 77 trials (10%) terminated prior to completion. The most common reason for termination was "recruitment/retention difficulty," reported in 37 (48%) terminated trials. In comparing competed versus terminated trials, primary purpose (nonobservational studies), enrollment (<50 patients), and geographic region (North America) were all significantly more likely to be terminated. Terminated trials were more likely to have an intervention type investigating a specific device or drug. Conclusions: Early trial termination for hand and wrist-related conditions is common (10%), with patient recruitment and retention identified as the leading cause of termination. Trials involving potential commercial incentives (those investigating a device or drug) were associated with an increased rate of trial termination. Clinical relevance: An emphasis on patient enrollment during study design may aid in mitigating the most common cause of early clinical trial termination.

The Effect of an Intraoperative Real-Time Counter on Radiation Exposure Events During Operative Treatment of Distal Radius Fractures.

PURPOSE: Occupational radiation exposure can have adverse health consequences for surgeons. The purpose of this study was to determine if utilization of an intraoperative, real-time radiograph counter results in decreased radiation exposure events (REEs) during open reduction and internal fixation (ORIF) of distal radius fractures (DRFs). METHODS: We reviewed all cases of isolated ORIF DRFs performed at a single center from January 2021 to February 2023. All cases performed on or after January 1, 2022 used an intraoperative radiograph counter, referred to as a "shot-clock" (SC) group. Cases prior to this date were performed without a SC and served as a control group (NoSC group). Baseline demographics, fracture, and surgical characteristics were recorded. Final intraoperative radiographs were reviewed to record reduction parameters (radial inclination, volar tilt, and ulnar variance). REEs, fluoroscopy exposure times, and total radiation doses milligray (mGy) were compared between groups. RESULTS: A total of 160 ORIF DRF cases were included in the NoSC group, and 135 were included in the SC group. The NoSC group had significantly more extra-articular fractures compared with the SC group. Reduction parameters after ORIF were similar between groups. The mean number of REEs decreased by 48% in the SC group. Cases performed with the SC group had significantly lower total radiation doses (0.8 vs 0.5 mGy) and radiation exposure times (41.9 vs 24.2 seconds). Mean operative times also decreased for the SC group (70 minutes) compared with that for the NoSC group (81 minutes). CONCLUSIONS: A real-time intraoperative radiograph counter was associated with decreased REEs, exposure times, and total radiation doses during ORIF DRFs. Cases performed with a SC had significantly shorter operative times without compromising reduction quality. Using an intraoperative SC counter during cases requiring fluoroscopy may aid in decreasing radiation exposure, which serves as an occupational hazard for hand and upper-extremity surgeons. TYPE OF STUDY/LEVEL OF EVIDENCE: Therapeutic II.

Sensitivity and Specificity of Examination Maneuvers for Carpal Tunnel Syndrome: A Meta-Analysis.

Our purpose was to assess the diagnostic validity (sensitivity (Sn) and specificity (Sp)) of physical examination maneuvers for carpal tunnel syndrome (CTS). This meta-analysis utilized the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) checklist. Studies assessing exam maneuvers (including components of the CTS-6) for CTS were identified in MEDLINE (Medical Literature Analysis and Retrieval System Online) and Embase (Excerpta Medica Database) databases. Assessed maneuvers assessed included: Phalen's test, Tinel's sign, Durkan test, scratch-collapse test, Semmes-Weinstein monofilament (SWM), and static 2-point discrimination (2PD) test. Data extracted included: article name, total number of subjects/hands, type of exam, and exam Sn/Sp. Forest plots were presented to display the estimated Sn/Sp and boxplots were used to demonstrate the locality, spread, and skewness of the Sn/Sp through the quartiles. After screening 570 articles, 67 articles involving 8924 hands were included. Forty-eight articles assessed Phalen's test, 45 assessed Tinel's sign, 21 assessed the Durkan test, seven assessed the scratch-collapse test, 11 assessed SWM, and six assessed the static 2PD test. Phalen's test demonstrated the greatest median Sn (0.70, (Q1, Q3): (0.51, 0.85)), followed by the Durkan test (0.67, (Q1, Q3): (0.46, 0.82)). 2PD demonstrated the highest median Sp (0.90, (Q1, Q3): (0.88, 0.90)), followed by SWM (0.85, (Q1, Q3): (0.51, 0.89)). There is considerable variability with respect to the validity of physical exam tests used in the diagnosis of CTS. Upper-extremity surgeons should be aware of inherent limitations for individual exam maneuvers. In the absence of a uniformly accepted diagnostic gold standard, a combination of exams, along with pertinent patient history, should guide the diagnosis of CTS.

Reliability and Validity of the Hook Test for Diagnosis of Distal Biceps Tendon Ruptures.

PURPOSE: Although the initial description of the distal biceps tendon (DBT) hook test (HT) reported 100% sensitivity (Sn) and specificity (Sp), subsequent retrospective series have demonstrated imperfect validity. The purpose of this investigation was to prospectively assess the validity and reliability of the HT for complete DBT ruptures. We aimed to determine the Sn/Sp and interrater reliability for the HT. METHODS: A consecutive series of adult patients presenting to our outpatient clinics with an elbow complaint was prospectively examined. Patients were included if they had undergone advanced imaging (magnetic resonance imaging or ultrasound) that imaged the DBT and underwent DBT repair. There were four participating surgeons, all of whom were blinded to magnetic resonance imaging/ultrasound prior to performing the HT. To determine the Sn/Sp of the HT and advanced imaging, intraoperative findings served as the primary reference standard. The interrater reliability of the HT was calculated for cases in which a primary examiner (surgeon) and secondary examiner (physician assistant or resident) performed the HT. RESULTS: Of 64 patients who had undergone advanced imaging, 28 (44%) underwent DBT surgery and were included in the assessment of Sn/Sp. The mean age was 49 years, and all patients were men. The Sn and Sp of the HT were 96% and 67%, respectively. Advanced imaging demonstrated 100% Sn and Sp. Twenty-five patients were evaluated by a primary and secondary examiner. The interrater reliability was substantial (Cohen kappa, 0.71). CONCLUSIONS: The Sn and Sp of the HT were 96% and 67%, respectively, when assessed prospectively. Advanced imaging findings (magnetic resonance imaging/ultrasound) demonstrated 100% Sn and Sp. The HT can be performed reliably by examiners with varying experience levels. Considering the imperfect validity of the HT, we caution against the use of this examination alone to diagnose DBT ruptures. TYPE OF STUDY/LEVEL OF EVIDENCE: Diagnostic II.