ChatGPT-4 generates orthopedic discharge documents faster than humans maintaining comparable quality: a pilot study of 6 cases.

BACKGROUND AND PURPOSE: Large language models like ChatGPT-4 have emerged. They hold the potential to reduce the administrative burden by generating everyday clinical documents, thus allowing the physician to spend more time with the patient. We aimed to assess both the quality and efficiency of discharge documents generated by ChatGPT-4 in comparison with those produced by physicians. PATIENTS AND METHODS: To emulate real-world situations, the health records of 6 fictional orthopedic cases were created. Discharge documents for each case were generated by a junior attending orthopedic surgeon and an advanced orthopedic resident. ChatGPT-4 was then prompted to generate the discharge documents using the same health record information. The quality assessment was performed by an expert panel (n = 15) blinded to the source of the documents. As secondary outcome, the time required to generate the documents was compared, logging the duration of the creation of the discharge documents by the physician and by ChatGPT-4. RESULTS: Overall, both ChatGPT-4 and physician-generated notes were comparable in quality. Notably, ChatGPT-4 generated discharge documents 10 times faster than the traditional method. 4 events of hallucinations were found in the ChatGPT-4-generated content, compared with 6 events in the human/physician produced notes. CONCLUSION: ChatGPT-4 creates orthopedic discharge notes faster than physicians, with comparable quality. This shows it has great potential for making these documents more efficient in orthopedic care. ChatGPT-4 has the potential to significantly reduce the administrative burden on healthcare professionals.

Application of deep learning for automated diagnosis and classification of hip dysplasia on plain radiographs.

BACKGROUND: Hip dysplasia is a condition where the acetabulum is too shallow to support the femoral head and is commonly considered a risk factor for hip osteoarthritis. The objective of this study was to develop a deep learning model to diagnose hip dysplasia from plain radiographs and classify dysplastic hips based on their severity. METHODS: We collected pelvic radiographs of 571 patients from two single-center cohorts and one multicenter cohort. The radiographs were split in half to create hip radiographs (n = 1022). One orthopaedic surgeon and one resident assessed the radiographs for hip dysplasia on either side. We used the center edge (CE) angle as the primary diagnostic criteria. Hips with a CE angle < 20°, 20° to 25°, and > 25° were labeled as dysplastic, borderline, and normal, respectively. The dysplastic hips were also classified with both Crowe and Hartofilakidis classification of dysplasia. The dataset was divided into train, validation, and test subsets using 80:10:10 split-ratio that were used to train two deep learning models to classify images into normal, borderline and (1) Crowe grade 1-4 or (2) Hartofilakidis grade 1-3. A pre-trained on Imagenet VGG16 convolutional neural network (CNN) was utilized by performing layer-wise fine-turning. RESULTS: Both models struggled with distinguishing between normal and borderline hips. However, achieved high accuracy (Model 1: 92.2% and Model 2: 83.3%) in distinguishing between normal/borderline vs. dysplastic hips. The overall accuracy of Model 1 was 68% and for Model 2 73.5%. Most misclassifications for the Crowe and Hartofilakidis classifications were +/- 1 class from the correct class. CONCLUSIONS: This pilot study shows promising results that a deep learning model distinguish between normal and dysplastic hips with high accuracy. Future research and external validation are warranted regarding the ability of deep learning models to perform complex tasks such as identifying and classifying disorders using plain radiographs. LEVEL OF EVIDENCE: Diagnostic level IV.

Trends and projection of forearm fractures including elbow fractures of the Olecranon in Sweden: an analysis of 363 968 fractures using public aggregated data.

BACKGROUND: Forearm and olecranon fractures are a common orthopaedic injury. This study aimed to analyse whether the incidence of forearm injury is changing and identifying trends in the number of forearm and olecranon fractures using public aggregated data in Sweden. METHODS: The number of forearm and olecranon fractures as defined by the number of registered diagnoses with the ICD-10 code of S52 were collected and normalized per 100,000 inhabitants and stratified per sex, age, and month. Age-adjusted incidence for forearm and olecranon fractures were calculated using the direct method. Poisson regression was used to analyse monthly, seasonal and yearly change in forearm and olecranon fracture incidence. Logistical regression was used to predict future trends of forearm and olecranon fractures. RESULTS: The findings revealed a slight decreasing trend in forearm and olecranon fractures. The average incidence rate during the study period was 333 with women having a higher incidence rate than men. More fractures occurred in the winter months. Fluctuations in the number of forearm and olecranon fractures were observed during 2020 which may be influenced by the COVID-19 pandemic. Based on current data, forearm and olecranon fractures are expected to decrease in Sweden by 2035. CONCLUSION: This study describes the trend of forearm and olecranon fractures among individuals according to sex and age in Sweden using easily obtainable data. Trends in forearm and olecranon fractures are dependent on sex and age but generally show a decreasing trend. More precise studies are needed in order to properly quantify the specific incidence of various subtypes of forearm and olecranon fractures and associated risk factors.

Deep learning classification of shoulder fractures on plain radiographs of the humerus, scapula and clavicle.

In this study, we present a deep learning model for fracture classification on shoulder radiographs using a convolutional neural network (CNN). The primary aim was to evaluate the classification performance of the CNN for proximal humeral fractures (PHF) based on the AO/OTA classification system. Secondary objectives included evaluating the model's performance for diaphyseal humerus, clavicle, and scapula fractures. The training dataset consisted of 6,172 examinations, including 2-7 radiographs per examination. The overall area under the curve (AUC) for fracture classification was 0.89, indicating good performance. For PHF classification, 12 out of 16 classes achieved an AUC of 0.90 or greater. Additionally, the CNN model had excellent overall AUC for diaphyseal humerus fractures (0.97), clavicle fractures (0.96), and good AUC for scapula fractures (0.87). Despite the limitations of the study, such as the reliance on ground truth labels provided by students with limited radiographic assessment experience, our findings are in concordance with previous studies, further consolidating CNN as potent fracture classifiers in plain radiographs. The inclusion of multiple radiographs with different views from each examination, as well as the generally unselected nature of the sample, contributed to the overall generalizability of the study. This is the fifth study published by our group on AI in orthopaedic radiographs, which has consistently shown promising results. The next challenge for the orthopaedic research community will be to transfer these results from the research setting into clinical practice. External validation of the CNN model should be conducted in the future before it is considered for use in a clinical setting.

Timing of adverse events in patients undergoing acute and elective hip arthroplasty surgery: a multicentre cohort study using the Global Trigger Tool.

OBJECTIVE: To explore timing in relation to all types of adverse events (AEs), severity and preventability for patients undergoing acute and elective hip arthroplasty. DESIGN: A multicentre cohort study using retrospective record review with Global Trigger Tool methodology in combination with data from several registers. SETTING: 24 hospitals in 4 major regions of Sweden. PARTICIPANTS: Patients ≥18 years, undergoing acute or elective total or hemiarthroplasty of the hip, were eligible for inclusion. Reviews of weighted samples of 1998 randomly selected patient records were carried out using Global Trigger Tool methodology. The patients were followed for readmissions up to 90 days postoperatively throughout the whole country. RESULTS: The cohort consisted of 667 acute and 1331 elective patients. Most AEs occurred perioperatively and postoperatively (n=2093, 99.1%) and after discharge (n=1142, 54.1%). The median time from the day of surgery to the occurrence of AE was 8 days. The median days for different AE types ranged from 0 to 24.5 for acute and 0 to 71 for elective patients and peaked during different time periods. 40.2% of the AEs, both major and minor, occurred within postoperative days 0-5 and 86.9% of the AEs occurred within 30 days. Most of the AEs were deemed to be of major severity (n=1370, 65.5%) or preventable (n=1591, 76%). CONCLUSIONS: A wide variability was found regarding the timing of different AEs with the majority occurring within 30 days. The timing and preventability varied regarding the severity. Most of the AEs were deemed to be preventable and/or of major severity. To increase patient safety for patients undergoing hip arthroplasty surgery, a better understanding of the multifaceted nature of the timing of AEs in relation to the occurrence of differing AEs is needed.

Evaluation of Three-Dimensional Bioprinted Human Cartilage Powder Combined with Micronized Subcutaneous Adipose Tissues for the Repair of Osteochondral Defects in Beagle Dogs.

Cartilage lesions are difficult to repair due to low vascular distribution and may progress into osteoarthritis. Despite numerous attempts in the past, there is no proven method to regenerate hyaline cartilage. The purpose of this study was to investigate the ability to use a 3D printed biomatrix to repair a critical size femoral chondral defect using a canine weight-bearing model. The biomatrix was comprised of human costal-derived cartilage powder, micronized adipose tissue, and fibrin glue. Bilateral femoral condyle defects were treated on 12 mature beagles staged 12 weeks apart. Four groups, one control and three experimental, were used. Animals were euthanized at 32 weeks to collect samples. Significant differences between control and experimental groups were found in both regeneration pattern and tissue composition. In results, we observed that the experimental group with the treatment with cartilage powder and adipose tissue alleviated the inflammatory response. Moreover, it was found that the MOCART score was higher, and cartilage repair was more organized than in the other groups, suggesting that a combination of cartilage powder and adipose tissue has the potential to repair cartilage with a similarity to normal cartilage. Microscopically, there was a well-defined cartilage-like structure in which the mid junction below the surface layer was surrounded by a matrix composed of collagen type I, II, and proteoglycans. MRI examination revealed significant reduction of the inflammation level and progression of a cartilage-like growth in the experimental group. This canine study suggests a promising new surgical treatment for cartilage lesions.

Considerations on the Demography of BIA-ALCL in European Countries Based on an E(A)SAPS Survey.

BACKGROUND: A growing body of evidence indicates that breast implant-associated anaplastic large cell lymphoma (BIA-ALCL) is associated with the use of certain breast implants. Regional variations have been reported, and a genetic susceptibility has also been suggested. However, large variations in the ability to correctly diagnose BIA-ALCL and to further report and register cases exist between countries and may in part explain variations in the demography. MATERIAL AND METHODS: A survey was conducted by The European Association of Societies of Aesthetic Plastic Surgery E(A)SAPS and sent to 48 European countries. The primary aim was to identify the total number of confirmed cases of and deaths from BIA-ALCL in each country during four consecutive measurements over a two-year period. RESULTS: An increase in BIA-ALCL cases during four repeated measurements from a total of 305 in April 2019 to 434 in November 2020 was reported by 23 of the 33 responding countries. A nearly 100-fold variation in the number of cases per million inhabitants was noted, where Netherlands had the highest rate (4.12) followed by Finland (1.99). Countries with the lowest reported rates were Austria (0.078), Romania (0.052) and Turkey (0.048). CONCLUSION: The current study displays a notable variation ßin the number of confirmed BIA-ALCL cases across Europe, even for countries with established breast implant registers. Variations in diagnosis and reporting systems may explain the differences, but the influence of genetic variations and the prevalence of high-risk implants cannot be excluded. Incomplete sales data along with medical tourism preclude an absolute risk assessment. LEVEL OF EVIDENCE IV: This journal requires that authors assign a level of evidence to each article. For a full description of these Evidence-Based Medicine ratings, please refer to the Table of Contents or the online Instructions to Authors www.springer.com/00266 .

Natural language processing with deep learning for medical adverse event detection from free-text medical narratives: A case study of detecting total hip replacement dislocation.

BACKGROUND: Accurate and timely detection of medical adverse events (AEs) from free-text medical narratives can be challenging. Natural language processing (NLP) with deep learning has already shown great potential for analyzing free-text data, but its application for medical AE detection has been limited. METHOD: In this study, we developed deep learning based NLP (DL-NLP) models for efficient and accurate hip dislocation AE detection following primary total hip replacement from standard (radiology notes) and non-standard (follow-up telephone notes) free-text medical narratives. We benchmarked these proposed models with traditional machine learning based NLP (ML-NLP) models, and also assessed the accuracy of International Classification of Diseases (ICD) and Current Procedural Terminology (CPT) codes in capturing these hip dislocation AEs in a multi-center orthopaedic registry. RESULTS: All DL-NLP models outperformed all of the ML-NLP models, with a convolutional neural network (CNN) model achieving the best overall performance (Kappa = 0.97 for radiology notes, and Kappa = 1.00 for follow-up telephone notes). On the other hand, the ICD/CPT codes of the patients who sustained a hip dislocation AE were only 75.24% accurate. CONCLUSIONS: We demonstrated that a DL-NLP model can be used in largescale orthopaedic registries for accurate and efficient detection of hip dislocation AEs. The NLP model in this study was developed with data from the most frequently used electronic medical record (EMR) system in the U.S., Epic. This NLP model could potentially be implemented in other Epic-based EMR systems to improve AE detection, and consequently, quality of care and patient outcomes.

Measuring adverse events following hip arthroplasty surgery using administrative data without relying on ICD-codes.

INTRODUCTION: Measure and monitor adverse events (AEs) following hip arthroplasty is challenging. The aim of this study was to create a model for measuring AEs after hip arthroplasty using administrative data, such as length of stay and readmissions, with equal or better precision than an ICD-code based model. MATERIALS AND METHODS: This study included 1 998 patients operated with an acute or elective hip arthroplasty in a national multi-centre study. We collected AEs within 90 days following surgery with retrospective record review. Additional data came from the Swedish Hip Arthroplasty Register, the Swedish National Patient Register and the Swedish National Board of Health and Welfare. We made a 2:1 split of the data into a training and a holdout set. We used the training set to train different machine learning models to predict if a patient had sustained an AE or not. After training and cross-validation we tested the best performing model on the holdout-set. We compared the results with an established ICD-code based measure for AEs. RESULTS: The best performing model was a logistic regression model with four natural age splines. The variables included in the model were as follows: length of stay at the orthopaedic department, discharge to acute care, age, number of readmissions and ED visits. The sensitivity and specificity for the new model was 23 and 90% for AE within 30 days, compared with 5 and 94% for the ICD-code based model. For AEs within 90 days the sensitivity and specificity were 31% and 89% compared with 16% and 92% for the ICD-code based model. CONCLUSION: We conclude that a prediction model for AEs following hip arthroplasty surgery, relying on administrative data without ICD-codes is more accurate than a model based on ICD-codes.

The influence of depression on patient-reported outcomes for hip-fracture patients 1 year after surgery: a prospective cohort study.

BACKGROUND: Depression is common in elderly hip-fracture patients and together with cognitive impairment is associated with increased risk of mortality. AIM: We aimed to examine the influence depression has on patient-reported outcome up to 1 year after acute hip fracture. METHODS: 162 hip-fracture patients participated in the prospective observational cohort study and were followed up at baseline, and 3 and 12 months using patient-reported outcome scores. Patients with cognitive impairment were excluded. Depression was defined as a score ≥ 8 on the depression subscale of the Hospital Anxiety Depression Scale (HADS D), having a diagnosis of depression or being treated with anti-depressant medication. Hip function was assessed using Harris Hip Score (HHS), EQ-5D was used to assess health status and Quality of life, and the Pain Numerical Rating Scale (PRNS) was used to assess pain levels. A linear regression model adjusted for group, age, sex, and ASA class was used to identify risk factors for functional outcome 12 months after fracture. RESULTS: 35 patients were included in the depression group versus 127 in the control group. No statistical differences were found in the demographic data (age, sex, ASA class, fracture type, operation method, living situation, activities of daily living ADL and clinical pathway) between the groups. In the regression model, we found no correlation between depression and the patient-reported outcome. CONCLUSION: In young elderly hip fracture patients without cognitive dysfunction, depression may not be of major importance for the rehabilitation of hip function in the short term.

Exploring the incidence and nature of nursing-sensitive orthopaedic adverse events: A multicenter cohort study using Global Trigger Tool.

BACKGROUND: For decades, patient safety has been recognized as a critical global healthcare issue. However, there is a gap of knowledge of all types of adverse events sensitive to nursing care within hospitals in general and within orthopaedic care specifically. OBJECTIVES: The aim of this study is to explore the incidence and nature of nursing-sensitive adverse events following elective or acute hip arthroplasty at a national level. DESIGN: A retrospective multicenter cohort study. OUTCOME VARIABLES: Nursing-sensitive adverse events, preventability, severity and length of stay. METHODS: All patients, 18 years or older, who had undergone an elective (degenerative joint disease) or acute (fractures) hemi or total hip arthroplasty surgery at 24 hospitals were eligible for inclusion. Retrospective reviews of weighted samples of 1998 randomly selected patient records were carried out using the Swedish version of the Global Trigger Tool. The patients were followed for readmissions up to 90 days postoperatively throughout the whole country regardless of index hospital. RESULTS: A total of 1150 nursing-sensitive adverse events were identified in 728 (36.4%) of patient records, and 943 (82.0%) of the adverse events were judged preventable in the study cohort. The adjusted cumulative incidence regarding nursing-sensitive adverse events for the study population was 18.8%. The most common nursing-sensitive adverse event types were different kinds of healthcare-associated infections (40.9%) and pressure ulcers (16.5%). Significantly higher proportions of nursing-sensitive adverse events were found among female patients compared to male, p < 0.001, and patients with acute admissions compared to elective patients, p < 0.001. Almost half (48.5%) of the adverse events were temporary and of a less severe nature. On the other hand, 592 adverse events were estimated to have contributed to 3351 extra hospital days. CONCLUSIONS: This study shows the magnitude of nursing-sensitive adverse events. We found that nursing-sensitive adverse events were common, in most cases deemed preventable and were associated with different kinds of adverse events and levels of severity in orthopaedic care. Registered nurses play a vital role within the interdisciplinary team as they are the largest group of healthcare professionals, work 24/7 and spend much time at the bedside with patients. Therefore, nursing leadership at all hospital levels must assume responsibility for patient safety and authorize bedside registered nurses to deliver high-quality and sustainable care to patients.

Only 8% of major preventable adverse events after hip arthroplasty are filed as claims: a Swedish multi-center cohort study on 1,998 patients.

Background and purpose - Hip arthroplasty is one of the most performed surgeries in Sweden, and the rate of adverse events (AEs) is fairly high. All patients in publicly financed healthcare in Sweden are insured by the Mutual Insurance Company of Swedish County Councils (Löf). We assessed the proportion of patients that sustained a major preventable AE and filed an AE claim to Löf.Patients and methods - We performed retrospective record review using the Global Trigger Tool to identify AEs in a Swedish multi-center cohort consisting of 1,998 patients with a total or hemi hip arthroplasty. We compared the major preventable AEs with all patient-reported claims to Löf from the same cohort and calculated the proportion of filed claims.Results - We found 1,066 major preventable AEs in 744 patients. Löf received 62 claims for these AEs, resulting in a claim proportion of 8%. 58 of the 62 claims were accepted by Löf and received compensation. The claim proportion was 13% for the elective patients and 0.3% for the acute patients. The most common AE for filing a claim was periprosthetic joint infection; of the 150 infections found 37 were claimed.Interpretation - The proportion of filed claims for major preventable AEs is very low, even for obvious and serious AEs such as periprosthetic joint infection.

Validation of adverse events after hip arthroplasty: a Swedish multi-centre cohort study.

OBJECTIVES: Preventing adverse events (AEs) after orthopaedic surgery is a field with great room for improvement. A Swedish instrument for measuring AEs after hip arthroplasty based on administrative data from the national patient register is used by both the Swedish Hip Arthroplasty Register and the Swedish Association of Local Authorities and Regions. It has never been validated and its accuracy is unknown. The aim of this study was to validate the instrument's ability to detect AEs, and to calculate the incidence of AEs following primary hip arthroplasties. DESIGN: Retrospective cohort study using retrospective record review with Global Trigger Tool methodology in combination with register data. SETTING: 24 different hospitals in four major regions of Sweden. PARTICIPANTS: 2000 patients with either total or hemi-hip arthroplasty were recruited from the SHAR. We included both acute and elective patients. PRIMARY AND SECONDARY OUTCOME MEASURES: The sensitivity and specificity of the instrument. Adjusted cumulative incidence and incidence rate. RESULTS: The sensitivity for all identified AEs was 5.7% (95% CI: 4.9% to 6.7%) for 30 days and 14.8% (95% CI: 8.2 to 24.3) for 90 days, and the specificity was 95.2% (95% CI: 93.5% to 96.6%) for 30 days and 92.1% (95% CI: 89.9% to 93.8%) for 90 days. The adjusted cumulative incidence for all AEs was 28.4% (95% CI: 25.0% to 32.3%) for 30 days and 29.5% (95% CI: 26.0% to 33.8%) for 90 days. The incidence rate was 0.43 AEs per person-month (95% CI: 0.39 to 0.47). CONCLUSIONS: The AE incidence was high, and most AEs occurred within the first 30 days. The instrument sensitivity for AEs was very low for both 30 and 90 days, but the specificity was high for both 30 and 90 days. The studied instrument is insufficient for valid measurements of AEs after hip arthroplasty.