Chuck Lam

Purpose: Most clinical evaluations of large language models assess factual recall rather than the multi-step reasoning behind operative plans. Reasoning-tuned models, post-trained to generate explicit intermediate reasoning before answering, may better approximate surgical decision-making. We compared two such models from different development ecosystems, OpenAI o1 and DeepSeek R1, on complex spine cases. Methods: Ten synthetic vignettes spanning deformity, degenerative, urgent, and cervical pathology were analysed; an eleventh was excluded owing to a survey error. Both models received identical prompts. Outputs were anonymised and randomised. Eight fellowship-trained spine surgeons (five consultants, three fellows) scored diagnostic accuracy, reasoning and thoroughness, surgical plan appropriateness, and clarity on 5-point Likert scales, giving 78 paired evaluations. Analysis used linear mixed-effects models with crossed random intercepts for rater and vignette (Bonferroni-adjusted alpha 0.0125); reliability was assessed by intraclass correlation. Results: o1 scored higher in all four domains, significantly for diagnostic accuracy (4.6 vs. 4.3; mean difference 0.24, 95% CI 0.11–0.38) and clarity (4.4 vs. 4.2; 0.26, 0.07–0.44); reasoning and thoroughness (0.21, 0.03–0.38), and surgical plan appropriateness (0.17, − 0.03 to 0.36) did not meet the adjusted threshold. Inter-rater reliability was poor (ICC 0.03–0.07). Clarity correlated strongly with the clinical domains (rho 0.74–0.79); adjusting for clarity attenuated the o1 advantage by 55% to over 100%, leaving no detectable clinical-domain difference. Conclusion: Both models produced plans surgeons usually rated good or excellent. o1 was rated higher, significantly only for diagnostic accuracy and clarity, most of it reflecting presentation rather than clinical content. Surgeon oversight and format-normalised evaluation designs remain essential.

Purpose: The posterior dynamic distraction device (PDDD) is a fusionless implant for adolescent idiopathic scoliosis (AIS). We aimed to quantify pooled major-curve Cobb correction, complication, revision, and infection rates after primary PDDD implantation for Lenke 1 or 5 curves, and to place these outcomes in context alongside posterior spinal fusion (PSF) and vertebral body tethering (VBT). Methods: Following PRISMA guidance (PROSPERO: CRD420251074442), PubMed, Embase, Scopus, and Web of Science were searched to May 2025. Random-effects meta-analyses were performed for Cobb correction, complications, and revisions, with leave-one-out sensitivity analyses. Results: Five studies (148 unique patients; mean age 13.1–16.0 years) met the inclusion criteria. Pooled mean Cobb correction was 22.8° (95% CI: 14.2–31.4°), approximately 51% of the preoperative magnitude; however, heterogeneity was extreme (I²=98%; prediction interval − 14.5° to 60.1°), and this pooled figure should not be interpreted as a reliable estimate of expected correction. The pooled complication rate was 24% (95% CI: 12–40%, I²=76%) and revision rate was 30% (95% CI: 19–45%, I²=64%), rising to 35% when elective exchanges for device outgrowth were included, with mechanical failures predominating. Primary surgical site infection occurred in 6% (95% CI: 3–12%, I²=0%). Risk of bias was moderate to serious across all studies. Conclusion: PDDD provides moderate short-term curve correction, but approximately one in four patients experiences a complication and nearly one in three requires reoperation at short-term follow-up. Reported perioperative advantages derive from only two to three uncontrolled cohorts. PDDD does not yet match PSF for reliability or percentage correction, nor VBT for reoperation rates. Pending larger comparative studies with long-term follow-up, PDDD should be considered investigational and patients counselled accordingly.

Background: Ewing sarcoma is a rare malignancy of bone and soft tissue that most commonly affects children, adolescents, and young adults. Although multimodal treatment has improved survival in localised disease, outcomes remain poorer in adults, particularly in those with metastatic disease, larger tumours, or anatomically complex primary sites. Current staging systems are useful for treatment allocation but do not fully capture individual prognostic variation. This study aimed to develop and internally validate a predictive nomogram for disease‐specific survival in adult patients with Ewing sarcoma. Methods: A retrospective population‐based cohort study was performed using the SEER‐17 database from 2000 to 2015. Adult patients aged 18 years or older with histologically confirmed primary Ewing sarcoma of bone or soft tissue were included. The primary outcome was disease‐specific survival. Predictor variables included age, sex, tumour site, tumour size, metastatic status, surgery, radiotherapy, and chemotherapy. Multivariable Cox proportional hazards regression was used to identify independent prognostic factors and construct a nomogram estimating 1‐, 3‐, and 5‐year disease‐specific survival. Internal validation was performed using 150 bootstrap resamples, and clinical utility was assessed using decision curve analysis. Results: A total of 937 adult patients were included. The mean age was 33.9 years, and 556 patients were male. Extraosseous disease was the most common primary site, followed by axial bone and appendicular bone disease. Metastatic disease was present in 294 patients. Multivariable analysis identified metastatic disease, increasing age, larger tumour size, and absence of surgery as independent adverse prognostic factors. The nomogram incorporated age, clinical stage, primary site group, tumour size, surgery, and chemotherapy to estimate 1‐, 3‐, and 5‐year disease‐specific survival. Internal validation demonstrated an optimism‐corrected C‐index of 0.688, with acceptable calibration across the three time points. Decision curve analysis suggested that the full nomogram provided greater clinical utility than stage alone across intermediate threshold probabilities. Conclusion: This study developed and internally validated a population‐based prognostic nomogram for adult patients with Ewing sarcoma. The model demonstrated acceptable discrimination and calibration, and may support individualised risk assessment, prognostic communication, and clinical decision‐making. External validation in independent cohorts, incorporating molecular and treatment‐intensity data, is required to further improve clinical applicability.

Adolescent Idiopathic Scoliosis: A Comprehensive Review of Epidemiology, Etiology, and Interventions

Supracondylar humeral fractures (SHF) are among the most common paediatric elbow injuries. While the majority are managed successfully with reduction and fixation, a subset presents with vascular compromise requiring surgical intervention. Autologous vein grafting is a recognised technique for arterial reconstruction in adults; however, its long-term efficacy and safety in children remain unclear. This systematic review investigates the patency, complication rates, and functional outcomes of autologous venous grafts used in the management of paediatric SHF with vascular injury. This study adhered to PRISMA and Cochrane guidelines and was prospectively registered on PROSPERO (CRD42024553679). A systematic search was conducted across PubMed, MEDLINE, Embase, Cochrane, and Web of Science. Studies were included if they reported clinical, ultrasonographic, or functional outcomes of paediatric patients (< 18 years) undergoing autologous venous grafting following SHF. Risk of bias was assessed using a modified ROBINS-I, and data synthesis was performed using random effects meta-analysis. Five studies comprising 21 paediatric patients were included. The mean age was 6.1 years with an average follow-up of 7.0 years. All patients underwent brachial artery reconstruction using either a basilic or saphenous vein graft. All grafts remained patent at final follow-up, with 90.5% confirmed via ultrasonography. Complications occurred in 95% of patients, however minor vessel ectasia predominated. Aneurysm (9.9%), thrombus (3.8%), and calcification (8.3%) were rare and did not require further surgery. No cases of ischemic contracture were reported. Subgroup analysis showed no significant difference in outcomes between graft types. Autologous venous grafting is an effective strategy for vascular reconstruction in paediatric supracondylar humeral fractures, demonstrating high patency and low long-term morbidity. However, the high incidence of vessel ectasia and wide variation in outcome reporting underscore the need for standardised follow-up and further research. Systematic review of Level IV studies Level of Evidence III.

Background: As the global prevalence of dementia rises, an increasing number of patients with cognitive impairment require Total Joint Arthroplasty (TJA). However, the specific impact of dementia on postoperative outcomes remains underreported. This systematic review evaluates the complications, mortality, and healthcare utilization associated with dementia in patients undergoing Total Hip (THA) and Total Knee Arthroplasty (TKA). Methods: A systematic search of PubMed, Scopus, Web of Science, Embase, and the Cochrane Library was conducted following PRISMA guidelines. Eligible studies included comparative cohorts of patients with and without a diagnosis of dementia undergoing primary TJA. Methodological quality was appraised using the Newcastle-Ottawa Scale (NOS). Data were synthesized regarding mortality, readmissions, implant-related complications, and discharge disposition. Results: Seven retrospective cohort studies comprising 13,816 patients with dementia and 869,061 controls were included. The mean NOS score was 8.4/9, indicating high methodological quality. Patients with dementia exhibited significantly worse outcomes across both procedures, including higher rates of postoperative delirium (OR: 4.25–6.40), mortality (HR: 1.43–3.05), and discharge to skilled nursing facilities (OR: 1.87). Stratification by procedure revealed distinct risk profiles: while both cohorts faced high readmission rates, THA patients demonstrated specific vulnerability to mechanical complications, including increased risks of dislocation, periprosthetic fracture (OR: 2.07), and revision surgery. These mechanical failures were frequently driven by falls and poor compliance with postoperative precautions rather than infection alone. Conclusion: Dementia is a robust independent predictor of adverse outcomes after TJA. While both THA and TKA carry elevated systemic risks, THA poses unique mechanical challenges that may warrant the use of high-stability implants (e.g., dual-mobility cups) to mitigate dislocation risks. Perioperative strategies should prioritize caregiver-led surveillance and strict fall prevention protocols to improve safety in this vulnerable population.

Purpose: Early onset scoliosis comprises spinal deformities in children younger than 10, creating challenges in diagnosis, risk assessment, and management. Timely intervention is vital, because untreated deformity can lead to cardiopulmonary compromise. Artificial intelligence and machine learning are reshaping orthopaedic care by improving detection, forecasting progression, and guiding treatment. This scoping review maps current use in this patient population. Methods: Following PRISMA ScR standards, we systematically searched PubMed, Embase, Web of Science, Cochrane, and Scopus for studies that developed, applied, or validated AI models to diagnose, manage, or predict outcomes in EOS. Results: After removing duplicates, 352 records were screened, 22 full texts were reviewed, and 11 studies met inclusion criteria. Most investigations (63.6%) employed convolutional neural networks (CNNs) such as Mask R CNN, EfficientNet, and U Net. Ensemble learning with gradient boosting, random forest, and logistic regression (9.1%), Gaussian Naïve Bayes (9.1%), sparse additive machines (9.1%), and unsupervised clustering (9.1%) were also used. Image analysis dominated (72.7%), automating radiographic measurements (Cobb angle, skeletal maturity) and monitoring growing-rod distraction. Predictive models (27.3%) estimated prolonged hospital stay, unplanned reoperation, or postoperative complications. Mean accuracy was 91.2% (range 86.1% to 94.0%). Common limitations were small sample sizes, single-centre data, and limited external validation. Conclusion: AI shows promise for EOS imaging and risk prediction, yet translation is hindered by methodological heterogeneity and scarce external validation. Future work should adopt standardised reporting, aggregate multicentre datasets, and test models prospectively in large cohorts.

Background/Objectives: Osteoporotic vertebral fractures (OVFs) are common in older adults. While surgery is generally reserved for unstable or painful fractures, some OVFs conceal underlying malignancies, including metastatic and hematologic cancers. This study aimed to determine the pooled prevalence of unsuspected malignancy in patients initially diagnosed with OVFs. Methods: A systematic search of PubMed and Scopus was conducted from inception to September 2025 in accordance with PRISMA guidelines. Eligible studies included adults with presumed OVFs who underwent vertebral biopsy and histopathological evaluation. Prevalence estimates were pooled using a random-effects model, and study quality was assessed with the Newcastle–Ottawa Scale. Results: Thirteen studies involving 3513 patients were included. The pooled prevalence of malignancy was 8.0% (95% CI: 5.4–10.6), comprising metastatic solid tumors (4.9%; 95% CI: 2.3–7.4) and multiple myeloma (2.6%; 95% CI: 1.3–3.9). Malignancy was detected in 2.7% (95% CI: 1.8–4.1) of routine biopsy cohorts versus 36.8% (95% CI: 22.1–54.4) of clinically suspected cases. Diagnostic yield exceeded 45% in patients selected by combined history, imaging, or known malignancy. No biopsy-related complications or procedure-related mortality were reported. Moderate heterogeneity was observed, mainly in suspected cohorts. Conclusions: Vertebral biopsy is a safe and diagnostically valuable procedure in vertebral compression fractures. Its yield ranges from about one in 30 patients in routine settings to nearly one in two in high-risk groups, underscoring the importance of structured patient selection to facilitate timely cancer detection and referral.