Lampang Hospital

Abstract Background Accurate intra-operative assessment of knee extension angle is essential in total knee arthroplasty (TKA), as residual flexion contracture or hyperextension is associated with inferior post-operative function and increased implant wear. Visual estimation remains widely used, but prior research has demonstrated substantial inaccuracy. This study evaluated the accuracy and reliability of a digital inclinometer technique for measuring knee extension angle during TKA, using computer navigation as the reference standard. Methods A prospective comparative study was performed on 130 knees undergoing primary posterior-stabilized TKA. After trial implant insertion, knee extension was assessed using two methods: visual estimation and a digital inclinometer method utilizing the femoral chisel holder and anterior border of the tibial crest. All measurements were compared with OrthoPilot navigation. Accuracy was defined as measurements within ± 2° of navigation. Correlation, agreement, and intra- and inter-observer reliability were evaluated. Results Visual estimation underestimated extension and showed a mean difference of − 3.5° ± 4.1° versus navigation (p < 0.001). The proportion of knees within ± 2° of navigation was only 26.2%. In comparison, the digital inclinometer demonstrated a significantly smaller mean difference of 0.1° ± 1.6° (p = 0.753), with 86.9% of knees meeting the ± 2° accuracy threshold. Superiority of the inclinometer persisted across all subgroups of pre-operative deformity. Correlation with navigation was weak for visual estimation (r = 0.320; r 2 = 0.102), but very strong for the inclinometer technique (r = 0.894; r 2 = 0.799; both p < 0.001). Bland–Altman analysis demonstrated minimal bias and no proportional error for the inclinometer, whereas visual estimation showed wide limits of agreement. Intra- and inter-observer reliability for inclinometer measurements was excellent (ICCs of 0.91 and 0.93), superior to visual estimation. Conclusion The digital inclinometer technique provides accurate and reproducible intra-operative knee extension measurements, outperforming visual estimation and approximating navigation-based assessment. This low-cost approach is feasible where navigation is unavailable. Trial registration Thai Clinical Trials Registry (TCTR20180828002). Registered on 28 August 2018. Prospectively registered.

Bootstrap-based methods have been recommended for assessing prediction instability in clinical prediction models, but their performance relative to cross-validation (CV) remains unclear. We propose a CV-based approach for assessing prediction instability and compare it with a bootstrap-based approach in logistic regression and random forest models. We conducted a resampling-based empirical experiment using a clinical cohort of 19,418 emergency department patients. Development samples were generated under events-per-variable (EPV) scenarios of 10, 30, and 50, and results were compared with those from the full dataset. Models were evaluated using bootstrap validation and repeated 5-fold CV; nested CV was used for random forest tuning. Predictive performance was assessed using AUC, calibration slope, and calibration-in-the-large. Prediction instability was quantified using mean absolute prediction error (MAPE). For logistic regression, bootstrap validation and repeated 5-fold CV produced broadly similar discrimination and calibration, especially at higher EPV values. For random forest, apparent performance consistently overestimated empirical discrimination. Bootstrap validation and repeated 5-fold CV gave comparable discrimination, but repeated 5-fold CV produced calibration slope estimates closer to the empirical value. Prediction stability improved as EPV increased for both modelling approaches. At EPV 30, bootstrap-derived MAPE was higher than CV-derived MAPE for both logistic regression (median, 0.042 versus 0.020) and random forest (median, 0.077 versus 0.027). A CV-based approach can assess prediction instability while also providing internally validated performance. These findings support CV-based instability assessment as a practical alternative to bootstrap-based assessment, particularly when comparing instability across multiple modelling algorithms.

Background/Objectives: Several chemotherapeutic regimens and targeted therapies are currently established as standard second-line treatments for patients with advanced biliary tract cancer (BTC). However, evidence regarding the benefits of treatment after first-line therapy failure remains limited, particularly among Thai populations. This study aimed to explore the efficacy of second-line chemotherapy in patients with advanced BTC. Methods: We conducted a single-institution, retrospective study including patients with locally advanced or metastatic BTC who experienced disease progression following first-line treatment between January 2017 and December 2019. Overall survival (OS) was defined as the primary endpoint. The secondary endpoint was the restricted mean survival time (RMST). To minimize confounding, propensity scores were estimated and applied using inverse probability of treatment weighting (IPTW). Results: A total of 110 patients were included, of whom 69 (62%) received second-line chemotherapy in combination with best supportive care (2LCMT + BSC), while 41 (38%) received best supportive care (BSC) alone. The majority of cases were intrahepatic cholangiocarcinoma (73.9% and 70.7% in each group, respectively). The median OS was 5.3 months (95% CI 3.5-7.0) in the 2LCMT + BSC group and 1.0 months (95% CI 0.5-1.9) in the BSC-only group (unadjusted HR 0.40, 95% CI 0.26-0.59; p < 0.001). In IPTW-adjusted flexible parametric regression analysis, second-line chemotherapy was associated with a 53% reduction in the risk of death compared with BSC alone (p = 0.009). The restricted mean survival time (RMST) differences between groups at 3, 6, and 12 months were 1.3 months (95% CI 0.9-1.6; p < 0.001), 2.6 months (95% CI 1.9-3.3; p < 0.001), and 3.9 months (95% CI 2.7-5.1; p < 0.001), sequentially. Conclusions: These findings demonstrate that second-line chemotherapy provides a significant overall survival benefit compared with best supportive care alone in patients with advanced BTC.

AIM:To compare the effects of individual glucagon-like peptide-1 receptor agonists (GLP-1RAs) on direct body composition among adults with overweight or obesity, with or without type 2 diabetes (T2D). METHODS:We systematically searched six electronic databases and grey literature from inception to 3 November 2025, to identify all randomised controlled trials (RCTs) that investigated individual GLP-1RAs on change in body composition from baseline among adults with overweight or obesity with or without T2D, without language restrictions. The main outcomes included total body fat (%), fat mass (kg), visceral adipose tissue (VAT) area (cm2), subcutaneous adipose tissue (SAT) area (cm2), liver fat content (%), total lean tissue (%) and lean mass (kg). A frequentist network meta-analysis was employed using a random-effects model and reported as standardised mean differences (SMDs). Certainty of evidence was appraised using the Confidence in Network Meta-Analysis. RESULTS:We identified 43 unique RCTs with 3379 participants that compared 17 different interventions or dosages with a control group. Regarding variations in treatment benefits on body composition, GLP-1RAs subcutaneously were more efficacious than the control group in decreasing total body fat, fat mass, VAT, SAT and liver fat from baseline. No significant differences for change in total lean tissue were observed. However, subcutaneously, liraglutide 1.8 mg/day, semaglutide 1.0 mg-weekly and tirzepatide 15 mg-weekly (SMDs ranged from -1.09 to -0.50) significantly decreased lean mass from baseline. CONCLUSIONS:GLP-1RAs substantially improve body composition-related excess adiposity, but unfavourable effects on lean mass were observed, particularly at high doses.

Background: The triglyceride–glucose (TyG) index is a surrogate marker of insulin resistance and has been associated with stroke in general populations. However, evidence regarding its prognostic value for incident stroke in elderly patients with type 2 diabetes mellitus (T2DM) remains limited. Objectives: To evaluate the association between the TyG index and incident stroke in elderly patients with T2DM. A retrospective cohort study conducted at a provincial tertiary care center in Northern Thailand. Materials and Methods: This study included 1024 patients aged ≥60 years with T2DM. The TyG index was calculated from fasting triglyceride and fasting plasma glucose levels at baseline and categorized into quartiles. The primary outcome was computed tomography-confirmed incident stroke during a 5-year follow-up. Cox proportional hazards regression models were used to estimate hazard ratios and 95% confidence intervals, adjusting for age, sex, hypertension, and smoking status. Kaplan–Meier survival analysis and Cox proportional hazards regression. Results: Over a mean follow-up of 4.7 years, 51 patients (5.0%) developed incident stroke. Although higher TyG index quartiles were associated with less favorable metabolic profiles, no significant differences in stroke-free survival were observed across TyG quartiles. In multivariable analyses, the TyG index was not independently associated with incident stroke, whether analyzed as quartiles or as a continuous variable. Conclusions: The TyG index was not associated with incident stroke in elderly patients with T2DM, suggesting limited prognostic utility of the TyG index for stroke risk prediction in this population.