Background Patients with cardiovascular-kidney-metabolic (CKM) syndrome have high mortality risk. The C-reactive protein–triglyceride glucose index (CTI) integrates inflammation and insulin resistance, yet its stage-specific prognostic value is undefined. Methods and results This retrospective cohort study included 8,632 CKM patients (2020–2025), stratified by CTI quartiles. All-cause mortality was assessed using adjusted Cox models. Among 860 deaths (10.0%), CTI was an independent mortality predictor in both stage 0–3 (HR 1.68, 95% CI 1.30–2.18) and stage 4 CKM (HR 1.54, 1.33–1.77). Compared with Q1, Q4 patients had significantly higher risk (stage 0–3: HR 4.57, 2.32–8.99; stage 4: HR 2.26, 1.64–3.13). A linear dose-response relationship was confirmed (P for nonlinear >0.05). CTI showed superior 1-year predictive performance over TyG index (AUC: stage 0–3, 0.73 vs. 0.62; stage 4, 0.69 vs. 0.51). Conclusion CTI is a potent, independent predictor of all-cause mortality in CKM, with a linear dose-response relationship. By integrating inflammatory and metabolic pathways, it provides superior risk stratification over TyG index, highlighting its clinical utility.
Jiacai Lin, Shaobin Qiu, Shuling Su et al.· Frontiers in Cardiovascular...· 0 citations
Managing diabetic kidney disease (DKD) is inherently complex, requiring clinicians to synthesize patient history, fluctuating biomarkers, and evolving treatment guidelines. While large language models (LLMs) show promise in medical decision support, their clinical adoption is hindered by factual inaccuracies and a lack of specific reasoning required for individualized patient management. To address this, we developed a hierarchical multi-agent system that integrates a locally deployed retrieval-augmented generation (RAG) framework with a cloud-based advanced reasoning engine, grounding responses in a curated corpus of clinical guidelines. We conducted a multi-center retrospective validation using 267 patient cases. The system's performance was evaluated against baseline models through a blinded review by twelve independent physicians across clinical dimensions including accuracy, safety, and factuality. Our evaluation reveals that the RAG-enhanced system significantly outperforms unaugmented models in providing accurate, guideline-compliant recommendations. Notably, it substantially reduced safety-critical errors, particularly in identifying medication contraindications related to renal function stages, while achieving high inter-rater reliability. This study demonstrates that anchoring LLMs with authoritative knowledge effectively mitigates hallucination risks and enhances clinical reliability. The proposed framework functions as a reliable on-demand assistant for DKD management, providing guideline-grounded decision support for primary care providers.
Xuan Tao, Lan Tian, Chenhao Fang et al.· npj Digital Medicine· 0 citations
