Background: Given the significant impact of delayed graft function (DGF) on transplant outcomes, the aim of this study was to develop and validate machine learning (ML) models capable of predicting the risk of DGF in deceased-donor kidney transplantation (DDKT). Methods: This retrospective cohort study was conducted using clinical and histopathological data collected between 2018 and 2022 at Ramathibodi Hospital from DDKT donors, recipients, and post-implantation time-zero kidney biopsy samples to develop predictive models. The performance of three ML models (neural network, random forest, and extreme gradient boosting [XGBoost]) and traditional logistic regression on an independent test data set was evaluated using the area under the receiver operating characteristic curve (AUROC) and Brier score calibration. Results: Among 354 DDKT recipients, 64 (18.1%) experienced DGF. The key contributing factors included a donor body mass index > 23 kg/m², donor diabetes mellitus, a prolonged cold ischemia time, a male recipient, and an interstitial fibrosis/tubular atrophy score of 2–3 in the time-zero kidney biopsy sample. The random forest model had a specificity of 99.96% and an AUROC of 0.9323, the neural network model had a specificity of 97.43% and an AUROC of 0.844, and the XGBoost model had a specificity of 99.81% and an AUROC of 0.989. A traditional statistical model had a specificity of 84.4% and an AUROC of 0.769. Conclusions: Predictive models, especially XGBoost models, have potential as tools for assessing DGF risk post-DDKT, guiding acceptance decisions, and avoiding risky biopsy, and they may be crucial in resource-limited settings.

Objectives To assess the real-world effectiveness and safety of sodium-glucose co-transporter-2 inhibitors (SGLT2i) on cardiovascular outcomes in patients with type 2 diabetes mellitus (T2D) and heart failure (HF) and to evaluate the associated risks of adverse events. Design A retrospective cohort study using propensity score analysis to control confounding variables. Setting Data were collected from the electronic health records of two large tertiary care hospitals in Thailand over a 12-year period (2010–2022). Participants Adults aged 18 years and older with a diagnosis of T2D and HF were included in the study. Patients who received SGLT2i for a minimum of 3 months were compared with those in a non-SGLT2i group. Participants with a diagnosis of HF that preceded their diagnosis of T2D were excluded from the analysis. Primary and secondary outcome measures The primary outcome was heart failure hospitalisation (HFH). Secondary outcomes included non-fatal stroke, non-fatal myocardial infarction (MI), all-cause mortality and adverse events (urinary tract infections, hypoglycaemia and acute kidney injury). Results A total of 11 758 patients were included in the study, with a median follow-up of 2.44 (IQR: 0.72–5.02) years. After applying inverse probability of treatment weighting, covariates were balanced, ensuring the validity of the treatment effect model’s assumptions. SGLT2i use was associated with a 59% reduction in HFH (HR 0.41, 95% CI 0.28 to 0.61), a 54% reduction in stroke (HR 0.46, 95% CI 0.33 to 0.63), a 51% reduction in MI (HR 0.49, 95% CI 0.36 to 0.67) and a 76% reduction in in-hospital all-cause mortality (HR 0.24, 95% CI 0.14 to 0.42). Additionally, SGLT2i use was associated with fewer adverse events, including lower rates of urinary tract infections and hypoglycaemia, compared with the non-SGLT2i group. Conclusions SGLT2i significantly improved cardiovascular outcomes in patients with T2D and HF in a real-world clinical setting. These findings support the incorporation of SGLT2i in the management of high-risk patients with T2D and HF. Further research is warranted to explore long-term outcomes and barriers to SGLT2i prescription in routine practice.