Background: Endoscopic surgery is widely used in clinical practice because of its smaller incisions and faster postoperative recovery. However, indirect visualization of the operative field can be challenging for less experienced surgeons, while technical performance has a substantial influence on clinical outcomes and prognosis. In addition, current assessment approaches often rely on subjective judgment and lack explicit depth-related information. Materials and methods: This study presents an approach for automated and objective assessment of endoscopic surgical skills by integrating surgical instrument segmentation, endoscopic depth rectification, and quantitative skill evaluation. A deep learning model is used to segment different types of surgical instruments from endoscopic videos, and a depth rectification algorithm is incorporated to improve the reliability of instrument tracking based on the segmentation results. Results: Three datasets comprising a total of 3110 images are included for instrument segmentation. Based on the tracking results, a quantitative evaluation framework covering four global surgical skill criteria is constructed. The proposed instrument segmentation framework achieves competitive accuracy across multiple endoscopic datasets (region of interest on RoboSeg2017: >0.85; Kvasir-Instrument: >0.94; Instrument-Crowdsourced: >0.89). In addition, the proposed quantitative criteria are consistent with manual assessment by experienced surgeons, supporting the feasibility of the framework for objective endoscopic skill assessment. Conclusions: Deep learning can support accurate analysis of surgical visual content and provide a basis for objective assessment of endoscopic performance. The proposed framework offers interpretable quantitative metrics that may assist in endoscopic skill evaluation and feedback generation. Further prospective and multi-center validation is needed to determine the robustness of the framework across diverse endoscopic procedures and clinical environments.
Objective: To conduct an exploratory analysis of BARSCORE, a novel classification of postoperative nutritional supplementation and support requirements, in a large multicenter derivation cohort, and to use the findings to define a revised version of the instrument, presented here, for prospective independent validation in a subsequent study. Background: Nutritional deficiencies remain a clinical challenge after bariatric surgery, yet the absence of standardized reporting limits comparison across procedures and centers. Materials and methods: BARSCORE was developed by multidisciplinary consensus among specialists from three high-volume university centers, defining six progressive grades (0-V, with subgrades IIA/IIB and IIIA/IIIB) of supplementation and nutritional support effort. Coherence with surgical procedure was assessed in consecutive adults undergoing primary laparoscopic bariatric surgery with at least 2 years’ follow-up, across 16 Spanish centers. The association between surgical procedure and BARSCORE grade was examined using proportional-odds ordinal logistic regression, with Firth-penalized logistic regression for the sparse grade IV-V comparison and a multinomial model as a sensitivity analysis; between-center variability was quantified with the intraclass correlation coefficient. Reliability and clarity were evaluated through two online surveys completed by clinicians from the participating centers who had no role in developing BARSCORE or grading the cohort, including classification of 14 clinical cases and assessment of interpretability. Results: Coherence analysis included 4,514 patients, showing a progressive shift toward higher grades with increasing intestinal nutrient diversion: grades 0-I predominated after sleeve gastrectomy (83.6%) and Roux-en-Y gastric bypass (65.9%), grade II-III predominated after single-anastomosis duodeno-ileal bypass with sleeve gastrectomy (65.4%) and duodenal switch (73.1%), and duodenal switch showed the highest proportion of grade IV-V (4.1% of DS patients). Among the 28 patients with grade IV-V overall, 17 (60.7%) had undergone Roux-en-Y gastric bypass, 9 (32.1%) duodenal switch, 2 sleeve gastrectomy, and none single-anastomosis duodeno-ileal bypass with sleeve gastrectomy; the highest-severity grades were therefore not exclusive to duodenal switch. Procedure was associated with BARSCORE grade in the ordinal model (Cramer’s V 0.21-0.25), robust to adjustment for sex, age, and preoperative BMI, though point estimates shifted materially (28-60%) between crude and adjusted models and the proportional-odds assumption did not hold fully for all procedures. Reliability and clarity surveys were completed by 133 clinicians (78 surgeons, 27 endocrinologists, 28 residents), none of whom had developed BARSCORE or graded the cohort. Inter-observer agreement (Fleiss’ kappa, correcting for chance) was 0.84 (95% CI 0.834-0.845) overall; 92.3% of respondents selected the pre-specified correct grade. On the clarity survey, 92.5% of respondents rated the system as simple. Conclusions: BARSCORE showed a coherent, though not fully proportional, association with the degree of intestinal nutrient diversion, high inter-observer reliability, and high rated clarity in this derivation cohort. On the basis of these findings the instrument was revised in five specific respects, and the resulting version, BARSCORE V2, is presented in this article. Prospective, independent multinational validation of BARSCORE V2 is the object of a planned second study, and is required before the score is used to benchmark nutritional outcomes across bariatric procedures and centers.
Background: There is limited evidence on fracture risk in patients with thymic epithelial tumors (TETs). We investigated this risk after surgical resection, stratified by treatment modality and time since surgery. Methods: Using the Korean National Health Insurance Service database, 1655 TET patients and 4965 age- and sex-matched controls (1:3) were included. Participants were followed from 2009 to 2022 for incident fractures (any, vertebral, and hip), ascertained within 1 year and at 1 year or more after surgery. Competing risk analyses, with all-cause death as a competing event, were performed. Results: Overall, TET patients had a higher risk of any fracture [cause-specific hazard ratio (csHR) 1.31, 95% confidence interval (CI) 1.04–1.64] and vertebral fracture (csHR 2.24, 95% CI 1.42–3.53) compared with controls, with a pronounced risk among those receiving surgery with radiation therapy (RT) (csHR 1.58, 95% CI 1.16–2.14 for any fracture; csHR 2.73, 95% CI 1.52–4.88 for vertebral fracture). Within the first year after surgery, fracture risk was markedly elevated among patients receiving surgery with chemoradiotherapy (CRT) (csHR 4.40 for any fracture; csHR 11.95 for vertebral fracture; csHR 13.90 for hip fracture). Beyond 1 year, the elevated risk persisted (csHR 1.30 for any fracture; csHR 2.18 for vertebral fracture) and was most prominent in patients receiving RT (csHR 1.60 for any fracture; csHR 2.69 for vertebral fracture). Conclusions: TET patients had a higher risk of fractures than non-cancer controls, particularly among those who underwent surgery with RT or CRT, and the elevated risk persisted beyond 1 year. Active surveillance and multidisciplinary management may help reduce the fracture burden in this population.
Objective: This updated systematic review and meta-analysis evaluates the safety and effectiveness of drain omission in pancreaticoduodenectomy (PD) and distal pancreatectomy (DP). Background: Routine prophylactic abdominal drainage after pancreatic resection has long been standard practice, despite growing evidence questioning its clinical utility. Prior meta-analyses were limited by methodological heterogeneity, outdated definitions, and failure to stratify outcomes by procedure type. Methods: A comprehensive search of PubMed, Scopus, the Cochrane Library, and Google Scholar from inception to December 2025 identified studies of adult patients undergoing PD or DP with or without abdominal drainage. The analysis was conducted using RevMan. Risk ratios (RRs) and mean differences (MDs) were computed, along with subgroup analyses by procedure and study design. Results: A total of 40 422 patients underwent drain placement, while 8758 patients were in the no-drain group. Drain omission significantly reduced major morbidity (RR: 0.87), clinically relevant postoperative pancreatic fistula (RR: 0.57), interventional drainage (RR: 0.79), and readmissions (RR: 0.79), and shortened hospital stay (MD: −1.20 days). No significant differences were observed in mortality, reoperation, delayed gastric emptying, post-pancreatectomy hemorrhage, or abscess formation. Conclusion: This study establishes a clear surgical dichotomy. For DP, drain omission is safe and superior. Conversely, for PD, randomized evidence demonstrates no clinical advantage of omission, while the potential consequences of an undrained anastomotic leak remain serious. Therefore, routine omission in PD cannot currently be recommended and should be restricted to clinical trials or carefully selected low-risk patients.
Background: The prevalence of patients with extreme obesity [body mass index (BMI) ≥ 50] has significantly increased over the past three decades, surpassing the rise in the prevalence of overweight individuals in the general population. Ring-augmented one-anastomosis gastric bypass (RA-OAGB) is a modification of standard one-anastomosis gastric bypass (OAGB) designed to enhance the durability of long-term weight-loss and potentially reduce pouch dilation, particularly in patients with extreme obesity. While short-term outcomes appear comparable to other procedures, long-term comparative data remain limited. Methods: In this retrospective single-center study, we compared patients who underwent RA-OAGB with those who underwent OAGB or sleeve gastrectomy (SG) procedures, all with a minimum follow-up of ≥5 years. Data collected included demographics, comorbidities, perioperative outcomes, long-term complications, and patient-reported quality of life. Results: A total of 111 patients were included in our study during the study period – 24 underwent RA-OAGB, 44 OAGB, and 43 SG. Loss to direct follow-up at 5 years was 15% (94/111 patients had direct follow-up), although clinical data were available for all patients through medical records. The RA-OAGB group had a higher baseline BMI than the OAGB and SG groups (55.4 vs. 51.2 and 51.3 kg/m 2 ; P = 0.006). In the unadjusted analysis, median percentage of excess weight loss (%EWL) was 74%, 63%, and 52%, respectively ( P = 0.027), while median percentage of total weight loss (%TWL) was 42.4%, 37.5%, and 28.9%, respectively ( P = 0.017). After adjustment for baseline BMI and follow-up duration, BMI was significantly lower after RA-OAGB compared with SG ( P = 0.025); RA-OAGB remained associated with greater %EWL ( P = 0.001) and %TWL ( P = 0.002) compared with SG. Adjusted associations with symptomatic gastroesophageal reflux disease (GERD) were not statistically significant. Iron-deficiency anemia was more frequent after bypass procedures than after SG. Quality of life and patient satisfaction were high across all groups, with no significant differences in Bariatric Analysis and Reporting Outcome System (BAROS) scores. Conclusion: In this retrospective single-center study, RA-OAGB was associated with greater long-term weight loss than OAGB and SG. However, given the observational design, baseline differences between groups, and limited sample size, these findings should be considered hypothesis-generating and require confirmation in larger prospective multicenter studies.
