Jubilee Mission Medical College and Research Institute is a private, non-profit Christian minority medical college, hospital and research institute located at Thrissur in Kerala, India. The establishment is administered by the Jubilee Mission Hospital Trust, a charitable organisation under the Catholic Archdiocese of Thrissur.
Rotational thromboelastometry (ROTEM) offers rapid, point-of-care assessment of hemostasis and is increasingly used in massive bleeding [1,2]. However, its high capital and reagent costs limit widespread adoption in low- and middle-income countries like India. This study evaluated the diagnostic performance and turnaround time (TAT) of ROTEM compared with conventional coagulation tests (CCT) and explored the feasibility and cost implications of replacing or supplementing CCT with ROTEM in massive bleeding. In this exploratory cross-sectional study, ROTEM assays (EXTEM, INTEM, FIBTEM) were performed in adult patients in parallel with PT, aPTT, fibrinogen, and platelet count. Correlation and diagnostic performance were assessed using Pearson correlation and ROC analysis. TAT for ROTEM and CCT was compared. Contextual cost and feasibility analysis used locally available device and reagent price ranges and published economic evaluations of viscoelastic hemostatic assays. A total of 109 patients were included. Trauma was the most common indication. Mean TAT for ROTEM was significantly lower than CCT (11.2 ± 2.4 v/s 33.5 ± 5 min). FIBTEM parameters A5, A10 and MCF correlated excellently with fibrinogen (r values of 0.858, 0.856,0.855) while EXTEM A5, A10 and MCF correlated strongly with platelet counts (r values of 0.558,0.554, 0.528). In contrast, PT showed only moderate correlations with CT and CFT of EXTEM (r = 0.472, 0.403), and so is APTT with CT and CFT of INTEM (r = 0.452, 0.377). EXTEM versus fibrinogen ROC curve AUC for A5, A10, and MCF EXTEM were 0.812, 0.813 and 0.810 with cut-offs around 35–56.5 mm, yielding sensitivities and specificities in the 70–74
We introduce MediX-R1, an open-ended reinforcement learning (RL) framework for medical multimodal large language models (MLLMs) that enables clinically grounded, free-form answers beyond multiple-choice formats. MediX-R1 fine-tunes a baseline vision–language backbone with Group Relative Policy Optimization (GRPO) and a composite reward tailored for medical reasoning: an LLM-based accuracy reward that judges semantic correctness with a strict YES/NO decision, a medical embedding–based semantic reward to capture paraphrases and terminology variants, and lightweight format and modality rewards that enforce interpretable reasoning and modality recognition. This multi-signal design provides stable, informative feedback for open-ended outputs where traditional verifiable or MCQ-only rewards fall short. To measure progress, we propose a unified evaluation framework for both text-only and image+text tasks that uses an LLM-as-judge in place of brittle string-overlap metrics, capturing semantic correctness, reasoning, and contextual alignment. Despite using only $\sim 50$K instruction examples, MediX-R1 achieves excellent results across standard medical LLM and VLM benchmarks, outperforming strong open-source baselines and delivering particularly large gains on open-ended clinical tasks (e.g., radiology summarization and report generation). Our results demonstrate that open-ended RL with comprehensive reward signals and LLM-based evaluation is a practical path toward reliable medical reasoning in multimodal models. Our trained models, curated datasets and source code will be publicly released.
Anti-tuberculosis drug–induced hepatotoxicity is a major challenge in tuberculosis (TB) management, particularly in high-burden countries like India. Early liver function test (LFT) derangements during daily fixed-dose combination (FDC) therapy may reflect subclinical hepatocellular stress and guide timely intervention; however, evidence from Indian programmatic settings remains limited. To determine the prevalence, severity, demographic associations, biochemical trends, and clinical outcomes of early (two-week) LFT abnormalities in adult TB patients initiating daily FDC therapy under the National Tuberculosis Elimination Programme (NTEP). A prospective observational study was conducted among adults with newly diagnosed TB at a tertiary center in Kerala (January-June 2025). Baseline and two-week LFTs were compared using paired t-tests. Associations with demographic and clinical variables were assessed using Chi-square tests. LFT derangement was defined as per the WHO/NTEP criteria. Among 146 participants, 41 (28.1%) developed LFT derangements at two weeks (95% CI: 20.9–36.4%). No significant associations were observed with age (χ²=0.05, p=0.977), sex (χ²=0.11, p=0.898), TB type (χ²=0.00, p=1.000), or weight band (p>0.05). Mean alanine aminotransferase (ALT) increased from 26.6 to 58.1 IU/L (mean difference +31.5 IU/L; 95% CI: 25.8-37.2; p<0.001), and aspartate aminotransferase (AST) from 29.5 to 55.1 IU/L (+25.6 IU/L; 95% CI: 20.1-31.0; p<0.001). Total bilirubin rose from 0.63 to 0.83 mg/dL (+0.20 mg/dL; 95% CI: 0.09-0.31; p<0.01). Severity grading showed 65.9% Grade 1, 22.0% Grade 2, and 12.1% Grade 3 abnormalities; no Grade 4 hepatotoxicity occurred. Clinically, 29/41 (70.7%) patients continued therapy with monitoring, 8 (19.5%) required temporary interruption, 3 (7.3%) were successfully rechallenged, and 1 (2.4%) required permanent regimen modification. No patient developed jaundice, hepatic failure, or required hospitalization. Early LFT derangements are common but predominantly mild and clinically manageable among adults initiating daily FDC ATT. Significant early rises in ALT and AST highlight the value of routine two-week monitoring. Structured early biochemical surveillance under NTEP may prevent severe outcomes and minimize treatment disruption.
Plasma, the cell-free component of blood, offers a minimally invasive and rich source of biomolecular information, making it ideal for investigating systemic diseases. The advent of high-throughput proteomic technologies has revolutionized plasma analysis by enabling large-scale identification and quantification of circulating proteins. These advances have facilitated biomarker discovery for early diagnosis, prognosis, and therapeutic targeting of various disorders, including cancers and inflammatory diseases. However, the inherent complexity of plasma, along with inconsistent sample preparation and analysis methods, has historically limited proteome coverage and reproducibility. Early approaches relying on crude plasma analysis often suffered from poor consistency and significant batch effects, impeding reliable comparative studies across clinical cohorts. To overcome these limitations, the current study introduces a robust and scalable plasma proteomics workflow. The protocol integrates key preparatory steps such as immunodepletion, molecular weight cutoff filtration, lyophilization, protein quantification and normalization, enzymatic digestion, and LC-MS/MS profiling. This optimized strategy significantly enhances proteome depth and reduces variability, thereby enabling more accurate and reproducible differential protein expression analysis between disease cases and healthy controls. The workflow provides a valuable platform for researchers aiming to generate high-quality proteomic data from plasma, ultimately contributing to the advancement of biomarker-driven diagnostics and personalized medicine.
Background and objective Cesarean section rates are rising globally, including in India. Following the ARRIVE trial, elective induction at 39 weeks has become more common. Cervical length is included in the Bishop score for assessing cervical favorability; however, the Bishop score is subjective and has limited predictive value. This study proposes that an objective third-trimester transvaginal ultrasound measurement of cervical length at 36 weeks may better predict spontaneous labor onset and guide delivery management. The objective of this study is to determine the correlation between cervical length at 36-36+6 weeks and spontaneous labor onset and to assess its effectiveness in predicting mode of delivery. Methods This prospective observational study included 95 eligible pregnant women at a tertiary hospital over 18 months. All underwent transvaginal cervical length measurement between 36 and 36+6 weeks. Demographic and obstetric data (including Bishop score) were collected, and participants were followed until delivery. Outcomes were spontaneous labor by 39 weeks and the mode of delivery. Data were analyzed using chi-square tests for categorical variables and Pearson's correlation for continuous variables, with significance at p < 0.05. Results Cervical length at 36-36+6 weeks showed no significant correlation with spontaneous labor by 39 weeks; all 95 women required induction. Longer cervical lengths were significantly associated with prolonged induction-to-delivery intervals (r = 0.491). Women with longer cervices more often required Foley catheter ripening (p = 0.001), while none with very short cervices (≤2.5 cm) required emergency cesarean delivery. Cervical length was not associated with gestational age at delivery (p = 0.065). These findings align with evidence that shorter cervical lengths predict faster inductions and higher vaginal delivery rates, whereas longer cervices are linked to the need for multiple induction methods and increased cesarean likelihood. Conclusions Cervical length measurement at 36-36+6 weeks can inform induction strategies at term. Although not predictive of spontaneous labor onset, longer cervical length was associated with prolonged induction, greater need for mechanical ripening, and higher cesarean risk. These results support incorporating cervical length assessment into preinduction evaluation. Identifying a longer cervix at 36 weeks can alert clinicians to the likely need for intensive induction methods (e.g., Foley catheter) and the potential for cesarean delivery if the cervix remains unfavorable. The findings align with guidelines advocating individualized induction based on objective cervical assessment. Combining the Bishop score with sonographic cervical length at 36-37 weeks may improve the prediction of induction success and guide delivery planning. Further research should refine how cervical length measurement can be integrated into standard induction protocols to improve maternal and neonatal outcomes.
