Jalalabad Ragib-Rabeya Medical College

Background: The accuracy of Glasgow Coma Scale (GCS) for assessing mortality rate within ICU settings associated with traumatic brain injury (TBI) is not clearly explored within the South Asian context. This study evaluates the accuracy of GCS for predicting mortality rate among TBI patients in Bangladesh. Objective: To evaluate the discrimination and calibration performance of admission Glasgow Coma Scale (GCS) in predicting ICU mortality among adult traumatic brain injury (TBI) patients in Bangladesh. Methodology: This study was conducted aiming to assess the accuracy level of admission Glasgow Coma Scale (GCS) for the purpose of predicting the mortality rate of 300 adults within the ICU settings with a focus on five different medical colleges. The study took place for a period of one and a half years. The predictor variable was admission GCS and the primary outcome was mortality within ICU settings. The diagnostic performance was evaluated by using AUROC, specificity, sensitivity and calibration analysis. Results: Those who survived (190 out of 300 patients) had significantly higher GCS than the non-survivors with ROC showing good discrimination. The GCS admission also confirmed adequate calibration (Hosmer-Lemeshow p = 0.312). Conclusion: Altogether admission GCS was concluded as a reliable, resource-independent prognostic tool for mortality prediction, supporting clinical decisions associated with ICU admission within the South Asia context. Bangladesh Crit Care J September 2026; 14 (2): 110-115

BACKGROUND:Cardiovascular disease remains the leading cause of mortality worldwide, with residual cardiovascular (CVS) risk remaining high despite pharmacological interventions. Cholesteryl ester transfer protein (CETP) inhibitors have shown promise in improving lipid profiles, but their comparative effectiveness in reducing CVS outcomes remains unclear. This network meta-analysis and network meta-regression aimed to evaluate the comparative effectiveness of CETP inhibitors (Evacetrapib, Anacetrapib, Dalcetrapib, and Torcetrapib) on CVS outcomes, including CVS mortality, all-cause mortality, major adverse cardiovascular events, myocardial infarction, need for revascularization, and stroke. METHODS:A Bayesian network meta-analysis was conducted, incorporating data from 10 randomized controlled trials with 84,134 patients diagnosed with cardiovascular disease or at high risk. We compared various CETP inhibitors for their effectiveness in reducing CVS events. Network meta-regression was used to assess the impact of covariates such as age, sex, and baseline lipid levels. RESULTS:Evacetrapib ranked highest for CVS mortality (relative risk [RR]: 0.92, 95% credible interval [CrI]: 0.39-1.87; Surface Under the Cumulative Ranking Curve [SUCRA]: 71.91%) and Anacetrapib for major adverse cardiovascular events (RR: 0.40, 95% CrI: 0.07-1.06; SUCRA: 84.25%). Dalcetrapib showed favorable effects on all-cause mortality (RR: 0.72, 95% CrI: 0.16-1.42; SUCRA: 67.51%) and stroke (RR: 0.75, 95% CrI: 0.41-1.26; SUCRA: 83.03%). Anacetrapib had the highest rank in reducing myocardial infarction (RR: 0.83, 95% CrI: 0.46-1.15; SUCRA: 86.06%). CONCLUSIONS:This analysis suggests that Evacetrapib and Anacetrapib are the most effective CETP inhibitors for improving CVS outcomes, with Dalcetrapib showing beneficial effects in certain outcomes. Further studies are needed to confirm these findings.

Background: Diabetic retinopathy is among the leading preventable causes of blindness worldwide, but the shortage of trained ophthalmologists and the growing burden of screening still remain a considerable challenge, particularly in resource-constrained settings. Purpose: The aim of this study was to assess the diagnostic accuracy of a deep learning algorithm for detecting multiple retinal lesions associated with diabetic retinopathy in fundus photographs and to compare its performance with that of ophthalmologists. Methods: This retrospective analytical study used a combined dataset of approximately 38,788 annotated colour fundus photographs drawn from the publicly available EyePACS and APTOS 2019 datasets. A deep learning model based on EfficientNet-B4 was developed using supervised learning. The dataset was divided into training, validation, and testing sets. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC), sensitivity, specificity, and F1-score. A subset of test images was independently assessed by ophthalmologists for benchmarking. Findings: The deep learning model achieved a mean AUC of 0.92 across five categories of retinal lesions. The highest performance was observed for hard exudate detection (AUC = 0.96) and retinal haemorrhage detection (AUC = 0.94), with relatively lower performance for neovascularization (AUC = 0.87). The model performed similarly to ophthalmologists and demonstrated higher sensitivity in detecting hard exudates. Conclusion: Deep learning algorithms are capable of high diagnostic accuracy in multi-lesion detection of diabetic retinopathy in fundus photographs and represent a promising adjunct to ophthalmologist-led screening; however, prospective real-world validation and further work on model interpretability are needed before routine clinical deployment. J.Natl.Inst.Ophthalmol.2025;8(1): 56-62