ObjectiveTo investigate the value of predictive models established based on machine learning methods in predicting the risk of metabolic associated fatty liver disease （MAFLD）， and to analyze its key risk factors. MethodsA retrospective analysis was performed for the 50 variables of 2 168 healthy individuals who underwent physical examination in Department of Health Assessment， Xiyuan Hospital， China Academy of Chinese Medical Sciences， from January 2021 to December 2024， including body composition， past history， and laboratory tests， and according to whether they were diagnosed with MAFLD or not， they were divided into MAFLD group with 265 individuals and non-MAFLD group with 1 903 individuals. The Mann-Whitney U test was used for comparison of continuous data between two groups， and the chi-square test was used for comparison of categorical data between two groups. Randomly split the research data into a training set and a validation set in a 70% to 30% ratio. Predictive factors were screened from the training set data using univariate analysis， LASSO regression， and multivariate Logistic regression analysis. Predictive models were then constructed using seven machine learning methods： Logistic regression， decision tree， random forest （RF）， eXtreme gradient boosting， light gradient boosting machine， support vector machine， and artificial neural network. Model performance was evaluated by plotting receiver operating characteristic curve for the validation set and calculating the area under the curve （AUC）， sensitivity， specificity， and Youden index for each model. Furthermore， the SHapley Additive exPlanation （SHAP） method was used to analyze the contribution of variables in the optimal model. ResultsThe prevalence rate of MAFLD among the 2 168 subjects was 12.22% （265/2 168）. Smoking， diastolic blood pressure， phase angle， visceral fat area， muscle fat ratio， waist-to-hip ratio， aspartate aminotransferase， non-HDL-C/HDL-C ratio， triglyceride-glucose index， and gallstones were independent risk factors for MAFLD （all P<0.05）. The seven predictive models of support vector machine， eXtreme gradient boosting， decision tree， light gradient boosting machine， artificial neural network， RF， and Logistic regression had an AUC of 0.738， 0.754， 0.757， 0.786， 0.795， 0.796， and 0.815， respectively， in the validation set， among which the RF model had the best discriminatory ability （AUC=0.796， 95% confidence interval： 0.754 — 0.839）， with a sensitivity of 81.01%， a specificity of 63.16%， and a Youden index of 44.17%. The SHAP analysis showed that visceral fat area， waist-to-hip ratio， and diastolic blood pressure were the top three predictive factors in terms of importance. ConclusionThe RF model， constructed based on body composition and clinical indicators， has a good performance in predicting the risk of MAFLD， and its interpretability can help to identify high-risk individuals in the early stage in clinical practice.

OBJECTIVE To establish a scientific， systematic， multi-dimensional performance management system for pharmacy practice， so as to improve the efficiency and quality of pharmacy practice performance management in public hospitals. METHODS Based on the four dimensions of the balanced scorecard theory， finance， customer， internal process， learning and growth， reference indicators for pharmacy practice performance management were summarized. The Delphi method was used to screen indicators， and the analytic hierarchy process was applied to determine the weights of indicators. A pharmacy practice performance management system was then constructed. Based on this system， action plans were formulated and implemented. The effectiveness was evaluated from two aspects： customer reviews and changes in pharmacy practice outcomes. RESULTS A total of 28 reference indicators were summarized， and a performance management system for pharmacy practice was constructed， consisting of 4 primary indicators， 9 secondary indicators， and 20 tertiary indicators. Compared with action plans implementation before， the satisfaction of clinical departments was significantly improved， and 11 pharmacy practice performance management indicators were optimized after implementation. CONCLUSIONS A scientific and systematic performance management system for pharmacy practice has been successfully established， which can provide a reference for the innovation of hospital pharmacy practice management and the high quality development of pharmacy practice.

Objective: This study proposes a clustering framework for Chinese materia medica (CMM) based on a large language model (LLM), aiming to explore potential compatibility patterns among CMMs from the semantic perspective of CMM property theory. Methods: First, a CMM property knowledge base was constructed based on Chinese Materia Medica, including 567 commonly used CMMs characterized by four properties, five flavors, and meridian tropism. Then, 49 CMMs derived from 10 prescriptions for Zangdu (脏毒, pathogenic toxins) recorded in Waike Zhengzong (《外科正宗》, Orthodox Manual of External Medicine) and Yangke Xinde Ji (《疡科心得集》, Collected Insights on Ulcer Medicine) were selected as the experimental dataset. Five semantic representation methods—One-Hot, Word2Vec, Bidirectional Encoder Representations from Transformers (BERT), Beijing Academy of Artificial Intelligence General Embedding (BGE), and Qwen—were applied to encode CMM property information into vector representations. Subsequently, t-distributed Stochastic Neighbor Embedding (t-SNE) was used for nonlinear dimensionality reduction on high-dimensional semantic vectors, followed by k-means clustering (k = 7). Clustering performance was evaluated using the Silhouette Score (SS), Davies-Bouldin Index (DBI), and Calinski-Harabasz Index (CHI). Results: The Qwen-based clustering method, CMM-EmbedCluster, achieved the highest SS (0.607 4) and CHI (158.057 2), as well as the lowest DBI (0.499 5), indicating improved cluster separation and compactness compared with other methods. Visualization of CMM clustering results showed that the clusters were well separated in the low-dimensional space, with strong inter-cluster discrimination and high intra-cluster functional consistency. Further interpretability analysis of CMM clustering results revealed stable structural differences among clusters in terms of four properties, five flavors, and meridian tropism, forming functional partitions consistent with CMM property theory. Conclusion CMM-EmbedCluster utilizes an LLM to achieve semantic-level representation and clustering of CMMs within the framework of CMM property theory, providing support for exploring potential compatibility patterns among CMMs from the perspective of CMM property semantics.

Diabetic foot ulcer （DFU） is one of the most severe and costly complications of diabetes， with its refractory nature largely attributed to the persistent vicious cycle of inflammation and oxidative stress. Conventional single-target therapeutic strategies often fail to effectively break this cycle. Traditional Chinese medicine （TCM）， leveraging its unique philosophy of ''holistic regulation and multi-target intervention''， has demonstrated significant advantages in promoting DFU healing. This review introduced a ''systemic intervention'' perspective to systematically elucidate how TCM， through multi-component synergistic networks， precisely deconstructs and intervenes in this pathological loop. Firstly， this study provided an in-depth analysis of how， under hyperglycemic conditions， the crosstalk between the nuclear factor-kappa B （NF-κB） and NOD-like receptor family pyrin domain-containing protein 3 （NLRP3） inflammasome pathways， the imbalance in macrophage polarization and the disruption of redox homeostasis interact to form a self-sustaining vicious cycle that impedes wound repair. Subsequently， the core content systematically discussed the strategies by which TCM breaks this deadlock： （1） Active ingredients from single herbs （e.g.， luteolin and astragaloside Ⅳ） can precisely modulate key nodes such as the phosphatidylinositol 3‑kinase/protein kinase B （PI3K/Akt） and TLR4/NF-κB pathways， thereby suppressing inflammation and oxidative stress. （2） Classical compound formulas （e.g.， Simiao Yong'an decoction and Taohong Siwu decoction） synergistically improve microcirculation and the immune microenvironment through multi-component cooperation. （3） External preparations （e.g.， Shengji Yuhong Ointment and Jinhuang powder） potently exert local anti-inflammatory and pro-repair effects by activating pathways such as nuclear factor erythroid 2-related factor 2/heme oxygenase 1 （Nrf2/HO-1）. Collectively， these three modalities embody a synergistic mechanism characterized by ''internal treatment modulating the internal milieu to address the root cause， and external treatment targeting the lesion to alleviate symptoms''.” Existing clinical evidence has confirmed that the aforementioned multi-target interventions can effectively promote healing and improve symptoms. However， this field still faces persistent challenges， including an unclear material basis for the efficacy of compound formulas， and insufficient standardization in quality control and clinical protocols. These challenges stem from the inherent tension between the complex systems-based characteristics of TCM and the requirements of modern standardization. Future efforts urgently require to leverage cutting-edge technologies， such as network pharmacology， spatial multi-omics， and artificial intelligence to propel the paradigm shift in the prevention and treatment of DFU with TCM from ''empirical multi-target approaches'' toward ''precise systems regulation''. This will provide a theoretical foundation for developing innovative strategies for managing difficult-to-heal wounds based on an integrated traditional Chinese and Western medicine approach.

ObjectiveTo systematically assess the public health governance capacity in Zhejiang Province, to conduct an in-depth analysis of its strengths and weaknesses, so as to provide scientific basis and strategic recommendations for further enhancement. MethodsA systematic collection of policy documents, public information reports, and research literature related to public health governance capacity in Zhejiang Province from 2002 to 2023 was conducted (encompassing a total of 1 263 policy documents, 138 pieces of information reports and 631 research articles). Based on the evaluation criteria suitable for public health systems previously developed by the research team, the basic status and magnitude of change in public health governance capacity in Zhejiang Province was evaluated. Additionally, normative gap analyses were employed to identify the strengths and weaknesses. ResultsZhejiang Province ranked 4th nationwide in terms of public health governance capacity with a score of 733.4 points (1 000.0-point maximum). The province has effectively implemented the principle of health first (scoring 698.5 points in the assessment of health-first strategy implementation) and attached sufficient importance to health-related goals (scoring 658.2 points in the scientific rationality of goal setting). However, the implementation of inter-departmental coordination and incentive mechanisms only scored 178.7 points, the feasibility of management and monitoring mechanisms scored even lower at only 144.0 points, and the coverage of incentive mechanisms scored 286.0 points. ConclusionZhejiang Province has effectively implemented its health first strategy and attached great importance to health targets, but still needs to strengthen cross-departmental coordination mechanisms and health-oriented incentives.

ObjectiveThis paper aims to explore the risk factors for chronic atrophic gastritis （CAG） with turbidity toxin accumulation syndrome and establish a prediction model. MethodsClinical data of 180 patients with CAG who participated in the "clinical study of Xianglian Huazhuo Particles blocking CAG cancer transformation" of Hebei Sheng Zhong Yi Yuan from July 2021 to March 2022 were collected. After confounding factors were controlled by propensity score matching， patients were divided into a training set （namely dev） and a validation set （namely vad） in a seven to three ratio. The risk factors for CAG with turbidity toxin accumulation syndrome in the training set were investigated by using univariate Logistic regression analysis and least absolute shrinkage and selection operator （namely Lasso） regression algorithms. Subsequently， a model， named model 1se， was developed by using the training set data to predict the risk factors for CAG with turbidity toxin accumulation syndrome. The accuracy of the prediction model was assessed by using various methods， including the receiver operating characteristic （ROC） curve， Hosmer-Lemeshow test （H-L）， calibration plot， and decision curve analysis （DCA）. ResultsAge， body mass index （BMI）， family history of cancer， job and life satisfaction， yellow and greasy fur with slippery pulse， and heavy body sensation were independent risk factors of the model. The prediction model showed excellent predictive value for both the training and validation sets. ConclusionThe established prediction model for CAG with turbidity toxin accumulation syndrome has high discrimination and excellent calibration， which could provide an excellent clinical basis for disease diagnosis and individualized treatment of patients.

ObjectiveThis paper aims to explore the risk factors for chronic atrophic gastritis （CAG） with turbidity toxin accumulation syndrome and establish a prediction model. MethodsClinical data of 180 patients with CAG who participated in the "clinical study of Xianglian Huazhuo Particles blocking CAG cancer transformation" of Hebei Sheng Zhong Yi Yuan from July 2021 to March 2022 were collected. After confounding factors were controlled by propensity score matching， patients were divided into a training set （namely dev） and a validation set （namely vad） in a seven to three ratio. The risk factors for CAG with turbidity toxin accumulation syndrome in the training set were investigated by using univariate Logistic regression analysis and least absolute shrinkage and selection operator （namely Lasso） regression algorithms. Subsequently， a model， named model 1se， was developed by using the training set data to predict the risk factors for CAG with turbidity toxin accumulation syndrome. The accuracy of the prediction model was assessed by using various methods， including the receiver operating characteristic （ROC） curve， Hosmer-Lemeshow test （H-L）， calibration plot， and decision curve analysis （DCA）. ResultsAge， body mass index （BMI）， family history of cancer， job and life satisfaction， yellow and greasy fur with slippery pulse， and heavy body sensation were independent risk factors of the model. The prediction model showed excellent predictive value for both the training and validation sets. ConclusionThe established prediction model for CAG with turbidity toxin accumulation syndrome has high discrimination and excellent calibration， which could provide an excellent clinical basis for disease diagnosis and individualized treatment of patients.

ObjectiveTo investigate the molecular mechanism by which the Bushen Kaixuan Tongluo prescription exerts a renal protective effect in mice with diabetic kidney disease （DKD） by regulating the cyclic adenosine monophosphate （cAMP） signaling pathway. MethodsThirty specific pathogen-free （SPF） male db/db mice were adaptively fed for three weeks. Mice with a random tail vein blood glucose level ≥ 11.1 mmol·L^-1 and urinary albumin-creatinine ratio （ACR） ≥ 30 mg·g^-1 were considered successfully modeled. The successfully modeled mice were randomly divided into five groups with six mice in each group： the model group， the low-， medium-， and high-dose Bushen Kaixuan Tongluo prescription groups （administered at doses of 7， 14， 28 g·kg^-1·d^-1 respectively）， and the positive drug irbesartan group （administered at a dose of 20 mg·kg^-1·d^-1）. Additionally， six db/m mice were selected as the blank group. Mice in each group were given intragastric administration of the Bushen Kaixuan Tongluo prescription at the corresponding concentrations， irbesartan， or an equal volume of pure water， and the intervention lasted for 12 weeks. During the experiment， the general conditions， body weight changes， and renal function indicators of the mice were dynamically monitored. After the intervention， a blood glucose meter was used to measure the fasting blood glucose （FBG） of the mice. An automatic biochemical analyzer was employed to detect the levels of serum creatinine （SCr）， blood urea nitrogen （BUN）， urinary microalbumin （uALB）， ACR， aspartate aminotransferase （AST）， alanine aminotransferase （ALT）， total cholesterol （TC）， triglycerides （TG）， leptin （LEP）， glycosylated serum protein （GSP）， and insulin （INS） in the mice. Renal tissues were collected for hematoxylin-eosin （HE） staining， periodic acid-Schiff （PAS） staining， and Masson's trichrome staining to observe the histopathological changes. Immunohistochemistry （IHC） was used to detect the expressions of protein kinase A （PKA） and cAMP response element-binding protein （CREB） in the mice. Western blot analysis was performed to determine the expression levels of PKA， phosphorylated protein kinase A （p-PKA）， CREB， phosphorylated cAMP response element-binding protein （p-CREB）， and B-cell lymphoma-2 （Bcl-2） proteins in the renal tissues of the mice. Real-time quantitative polymerase chain reaction （Real-time PCR） was used to detect the mRNA expression levels of PKA， CREB， and Bcl-2 in the renal tissues of the mice. ResultsCompared with the blank group， the mice in the model group showed listlessness， decreased activity， and a significant increase in body weight （P<0.01）. Biochemical indicators revealed that the levels of BUN， uALB， ACR， AST， ALT， TC， TG， FBG， LEP， GSP， and INS were significantly increased （P<0.01）， while SCr showed an increasing trend with no statistically significant difference. Compared with the model group， the mice in the Bushen Kaixuan Tongluo prescription intervention groups had improved general conditions and a decreasing trend in body weight. Biochemical indicators showed that the levels of BUN， uALB， ACR， TC， GSP， and INS were significantly decreased （P<0.05）， while SCr， AST， ALT， TG， and LEP showed a decreasing trend with no statistically significant difference. Renal histopathological analysis showed that the model group exhibited typical DKD pathological features such as thickening of the glomerular basement membrane， expansion of the mesangial matrix， and deposition of collagen fibers in the renal tubulointerstitium， and all treatment groups could alleviate the above pathological damages. The IHC results showed that compared with the blank group， the expression levels of p-PKA and p-CREB in the renal tissues of the model group were significantly decreased （P<0.01）. Compared with the model group， the expression level of p-PKA in the medium-dose Bushen Kaixuan Tongluo prescription group was significantly increased （P<0.01）， while the expression level of p-CREB showed an increasing trend with no statistically significant difference. Western blot results showed that compared with the blank group， the expression levels of p-PKA/PKA， p-CREB/CREB， and Bcl-2 in the model group were significantly decreased （P<0.05）. Compared with the model group， the expression levels of these proteins in the medium-dose Bushen Kaixuan Tongluo prescription group were significantly increased （P<0.01）. Real-time PCR results showed that compared with the blank group， the mRNA expressions of PKA， CREB， and Bcl-2 in the model group were significantly down-regulated （P<0.05）. Compared with the model group， the mRNA expressions of these genes in the medium-dose Bushen Kaixuan Tongluo prescription group were significantly up-regulated （P<0.05）. ConclusionThe Bushen Kaixuan Tongluo prescription can improve the liver and kidney functions of db/db mice， correct lipid metabolism disorders and glucose metabolism imbalance. Its renal protective effect is associated with up-regulating the cAMP signaling pathway to improve renal fibrosis and reduce the level of oxidative stress， thereby protecting renal function.

ObjectiveTo develop a full-process data platform of renal rehabilitation, diagnosis and quality control information. MethodsA hierarchical architectural design was proposed, adhering to clinical pathway models and standardized data protocols. The platform comprehensively covered assessment, intervention, follow-up and quality control for maintenance hemodialysis (MHD) patients. By integrating multidisciplinary resources and standardizing rehabilitation workflows, it delivered standardized and intelligent rehabilitation services. ResultsThe platform achieved standardized and intelligent management of rehabilitation services, effectively improved the physiological function, psychological state and quality of life convenience for MHD patients, while significantly reduced the economic and care burden on patients' families and society. ConclusionThe rehabilitation service model based on a full-process data platform may provide scientific and systematic support for MHD patients.

ObjectiveTo investigate the effects of Huangqi Jianzhongtang （HQJZ） on macrophage polarization and fat consumption in cancer cachexia （CC） mice. MethodsUltra-performance liquid chromatography-quadrupole/electrostatic field Orbitrap high-resolution mass spectrometry （UPLC-Q-Orbitrap HRMS） was used to control the quality of HQJZ. （1） In vitro experiment： HQJZ-containing serum was prepared， and the optimal concentration was determined by cytotoxicity assay. Mouse monocyte-derived macrophages （RAW264.7） were cultured and randomly divided into six groups， including a blank group， a classically activated macrophages （M1） group， an alternatively activated macrophages （M2） group， a HQJZ + blank group， a HQJZ+M1 group， and a HQJZ + M2 group. The relative expression of macrophage marker genes CD86， inducible nitric oxide synthase （iNOS）， CD206， and arginase-1 （Arg1） was detected by real-time quantitative polymerase chain reaction （Real-time PCR ）. （2） In vivo experiment： Thirty-two BALB/c mice were randomly divided into a control group， a model group， a medroxyprogesterone acetate （MPA） group， and a HQJZ group. Except for the control group， the other mice were injected with CT-26 colon cancer cells to establish a CC model. Mice in the MPA and HQJZ groups were given MPA （0.13 g·kg^-1·d^-1） or HQJZ （13.13 g·kg^-1·d^-1） by gavage， respectively， while mice in the control and model groups were given an equal volume of saline by gavage， with interventions continued for 10 d. Real-time PCR was used to detect the expression of macrophage markers （iNOS， Arg1， CD86， CD206） and fat browning-related genes uncoupling protein 1 （UCP1） and peroxisome proliferator-activated receptor γ （PPARγ） in epididymal adipose tissue. Western blot （WB） was used to detect protein expression levels of UCP1 and PPARγ. Micro-computed tomography （micro-CT） was used to measure residual fat volume， and hematoxylin-eosin （HE） staining was used to assess fat browning and calculate pathological scores. ResultsIn vitro， the dominant effective concentration of HQJZ-containing serum was 12.5%. Real-time PCR results showed that， compared with the blank group， Arg1 expression decreased in the HQJZ+blank group （P<0.05）， CD206 showed a downward trend without statistical significance， while iNOS and CD86 expression were significantly increased （P<0.05）. Compared with the M1 group， Arg1 and CD206 expression decreased in the HQJZ+M1 group （P<0.05）. Compared with the M2 group， CD206 expression decreased in the HQJZ+M2 group （P<0.05）， CD86 expression increased significantly （P<0.01）. In vivo， Real-time PCR results showed that， compared with the control group， CD86 and CD206 expression levels were significantly increased in the model group （P<0.01）. Compared with the model group， CD206 expression in the MPA group was significantly decreased （P<0.01）. In the HQJZ group， CD206 was significantly decreased （P<0.01）. WB results showed that， compared with the model group， protein expression of UCP1 and PPARγ was significantly reduced in the HQJZ group （P<0.05， P<0.01）. micro-CT results showed that the total white fat volume in the HQJZ group was greater than that in the model group （P<0.05）. HE staining results showed that pathological scores in the HQJZ group were lower than those in the model group （P<0.05）. ConclusionHQJZ may inhibit white adipose tissue browning by promoting macrophage M1 polarization and suppressing M2 polarization， thereby delaying fat consumption in CC mice.