Background: Post-stroke cognitive impairment (PSCI) severely affects patients' quality of life and survival, and no curative treatment is currently available. Therefore, early detection and prevention are crucial.
Objective: To develop and internally validate a nomogram for predicting the risk of PSCI within 6 months after stroke.
Study design: Bidirectional observational cohort study.
Methods: Demographic, clinical, and cognitive data were collected from patients with stroke treated at the First Affiliated Hospital of Air Force Medical University between January 2019 and January 2024, within 40 days of stroke onset. Six-month cognitive outcomes were obtained from medical records for the retrospective component (January 2019 to August 2023), while patients in the prospective component (August 2023 to January 2024) were followed up for 6 months. Patients were categorized into PSCI and non-PSCI groups based on the 6-month cognitive assessment. The least absolute shrinkage and selection operator regression and multivariable logistic regression were used to identify independent predictors and construct the prediction model. Model accuracy was evaluated and validated internally using the bootstrap method.
Results: Of 286 patients with stroke assessed, 107 presented with PSCI at the 6-month follow-up. The key predictors identified were baseline Mini-Mental State Examination score, time from lesion to rehabilitation facility, lesion location, and hypoalbuminemia. The risk prediction model using these factors showed strong performance, with an apparent area under the curve (AUC) of 0.940 (95% confidence interval [CI], 0.913-0.967), indicating good discrimination. The mean apparent AUC across 1000 bootstrap training samples was 0.947 (95% CI: 0.918-0.970). After optimism correction, the internally validated AUC was 0.937 (mean optimism = 0.003), indicating minimal overfitting.
Conclusion: The developed nomogram-based model predicted the risk of PSCI. It may aid in counseling patients with stroke and their families, and help clinicians identify high-risk patients for early intervention to improve outcomes.

Background: In the context of diagnosis-related group (DRG) payment, evidence on structural changes in nursing charge composition in general wards remains limited.
Objective: This study aimed to examine the nursing charge structures and their changes in priority departments within a DRG environment, and to identify whether nursing care is financially undervalued.
Study design: Retrospective observational study. Methods: This retrospective observational study analyzed 215,961 DRG-paid inpatient records (2022-2024) from a large general hospital in Shandong Province. Patients were stratified by relative weight (RW). The analysis was conducted using logistic regression, the entropy weight method, and structural change analysis.
Results: Surgery (OR=46.78, 95%CI 43.76-50.01) and age≥ 60 (OR=4.27, 95%CI 3.99-4.57) were the strongest factors associated with high RW. From 2022-2024, nursing charges in the cardiology department cumulatively decreased by 2.16%, with a contribution rate of 4.63% to the structural changes. Graded nursing care consistently accounted for over 93% each year, while specialized nursing care was consistently maintained at a low level. In the cardiovascular surgery department, nursing charges decreased cumulatively by 0.05% during this period, with a contribution rate of 0.22% to structural changes. Graded nursing care constituted the predominant share, accounting for 79.49%-82.54%. Specialized nursing care represented a key component, consistently exceeding 17%.
Conclusion: This study analyzed nursing charge structure changes in internal medicine and surgery departments. The observed shifts in nursing charge patterns may indicate that the value of nursing care is not fully captured under DRG-based payment. However, this interpretation is based solely on billing data and should therefore be made cautiously. Further research incorporating more granular, nursing-sensitive measures is needed to validate these findings. Integrating nursing workload into DRG may be important for stabilizing valuation and workforce.

Background: While adverse childhood experiences (ACEs) increase suicide attempt risk, the psychosocial pathways underlying this association remain uncl
Objective: To explore the potential mediating roles of negative life events (NLEs) and multiple psychological factors in the association between ACEs and suicide attempts.
Study design: A 1:1 matched case–control study.
Methods: A total of 442 participants (221 suicide attempters and 221 age- and sex-matched controls) were included from rural areas of Shandong Province, China. Information on ACEs, NLEs, and five psychological variables (resilience, impulsivity, depressive symptoms, anxiety, and loneliness) was collected via face-to-face interviews. Psychological status was identified using latent profile analysis. Conditional logistic regression and exploratory mediation analyses (simple, parallel, and serial mediation models) were used to explore the association of ACEs with suicide attempts and the potential mediating roles of NLEs and psychological factors.
Results: Individuals with ≥ 4 ACEs showed higher suicide attempt risk than those with 0–3 ACEs (odds ratio = 5.24, 95% confidence interval: 1.21–22.75). Simple mediation analyses showed that NLEs and psychological factors significantly contributed to the association between ACEs and suicide attempts, with indirect effects accounting for 19.9%–42.8% of the total effect. In parallel mediation models, resilience, impulsivity, depressive symptoms, and psychological status remained significant mediators after accounting for NLEs. Serial mediation analyses identified potential sequential pathways involving NLEs followed by multiple psychological factors.
Conclusions: ACEs were associated with suicide attempt risk, potentially through NLEs and psychological pathways. These results highlight the importance of considering stress exposure and psychological functioning, while the identified pathways require validation in longitudinal studies.

Background: Many patients with type 2 diabetes (T2D) develop diabetic retinopathy (DR) despite achieving target glycated hemoglobin (HbA1c) levels. Continuous glucose monitoring (CGM)-derived metrics capture glucose variability beyond HbA1c, yet systematic evaluations of their associations with DR across multiple CGM parameters remain limited.
Objective: This study aimed to evaluate correlations between five CGM metrics and DR in T2D.
Study design: Systematic review and meta-analysis.
Methods: This systematic review was registered on the International Prospective Register of Systematic Reviews (PROSPERO) (CRD42024613889). Six electronic databases were systematically searched from inception to April 1, 2026, for observational studies in T2D. The exposures of interest were CGM-derived metrics, including time in range (TIR), mean amplitude of glycemic excursions (MAGE), standard deviation of blood glucose (SDBG), mean blood glucose (MBG), and coefficient of variation (CV). The primary outcome was DR. Standardized mean differences (SMDs) with 95% confidence intervals (CIs) were synthesized.
Results: Ten studies (N = 6434) were included. Compared to non-DR controls, patients with DR had significantly lower TIR (SMD = −0.83, 95% confidence interval [CI]: −1.18 to −0.48; P < 0.001) and higher glycemic parameters, including MAGE (SMD = 1.53, 95% CI: 0.91–2.15; P < 0.001), SDBG (SMD = 1.10, 95% CI: 0.61–1.60; P < 0.001), MBG (SMD = 0.67, 95% CI: 0.25–1.09; P = 0.002), and CV (SMD = 0.38, 95% CI: 0.08–0.69; P = 0.014).
Conclusions: CGM metrics, particularly MAGE, SDBG, and TIR, are associated with prevalent DR in T2D, offering exploratory clinical insights alongside HbA1c. However, longitudinal prospective evidence is currently limited to TIR. Given the substantial heterogeneity across primary studies and the Very Low certainty of evidence per the Grading of Recommendations Assessment, Development and Evaluation (GRADE) framework, these metrics should be interpreted with caution and require prospective validation before clinical translation.

Background: Identifying clinical research articles can facilitate evidence synthesis and research evaluation; however, journal categories and keyword-based approaches may not accurately capture studies at the article level.
Objective: To develop and evaluate a text-based classifier for distinguishing clinical from non-clinical research articles.
Study design: We conducted a bibliographic classification study using Web of Science records published in 2023 by authors affiliated with Chinese hospitals, with temporal evaluation based on records published in 2024 from the same source.
Methods: We assembled 135,932 records published in 2023 and developed a model using a fixed subset of 20,000 records. Bidirectional Encoder Representations from Transformers (BERT), Bidirectional Encoder Representations from Transformers for Biomedical Text Mining (BioBERT), PubMedBERT, and Biomedical Language Model (BioMedLM) were evaluated using titles and keywords alone, with original abstracts, or with DeepSeek-V3-reformulated abstracts for training and validation as well as original abstracts for testing.
Results: Adding original abstracts improved the mean performance of BERT, BioBERT, and PubMedBERT. BioBERT-D3 demonstrated the highest observed accuracy (0.926), precision (0.926), F1 score (0.919), and AUC (0.972). PubMedBERT-D3 achieved the highest recall (0.922). On the 2024 temporal set, BioBERT-D3 achieved an accuracy of 0.980, a precision of 0.981, a recall of 0.970, an F1 score of 0.976, and an AUC of 0.992.
Conclusions: The framework may support first-pass literature screening and article-level bibliometric evaluation of clinical research output. Clinically ambiguous or mixed-content articles may still require human review. Although the model performed well on records published in 2024 from the same database and institutional setting, additional validation is required using other databases, datasets from different countries, and independently annotated reference labels.