Intelligent prescription recommendation has become an important research direction in traditional Chinese medicine (TCM), offering new opportunities to support clinical decision-making and promote the modernization of TCM practice. With the rapid development of artificial intelligence (AI), a variety of computational approaches have been proposed to learn prescription patterns from clinical data and generate personalized treatment recommendations. However, despite increasing research activity, systematic and comprehensive reviews of AI-driven methods for TCM prescription recommendation remain limited. In this study, we present a comprehensive review of computational approaches for herbal prescription recommendation (HPR) in TCM. Existing methods are systematically categorized into several major paradigms, including traditional machine learning methods, topic model methods, sequential generative methods, deep learning and graph-based methods, and large language model–based frameworks. In addition to summarizing methodological developments, we also review commonly used public datasets and evaluation metrics in this field. Furthermore, representative models with publicly available implementations are experimentally evaluated on multiple benchmark datasets to provide a comparative analysis of their performance on the HPR task. Finally, we discuss the key challenges that hinder the practical deployment of intelligent prescription recommendation systems, including data heterogeneity, limited interpretability, and insufficient integration of TCM domain knowledge. Future research directions are outlined to facilitate the development of more reliable, interpretable, and clinically applicable AI-assisted HPR systems for TCM.

OBJECTIVE To confirm trigger items for adverse drug reaction （ADR） induced by bevacizumab， to identify and analyze the occurrence of related ADR， and to establish a predictive model for severe adverse reaction （SAR） caused by this drug. METHODS Based on the global trigger tool （GTT） theory， and referencing the GTT White Paper， drug package inserts and relevant literature， trigger items for bevacizumab-related ADR were confirmed using a single-round Delphi method. Utilizing these established items， electronic medical records of relevant patients at Guilin People’s Hospital from January 2020 to September 2024 were actively screened via the China Hospital Pharmacovigilance System. Pharmacists then identified and tallied the occurrence of bevacizumab-induced ADR. Data from patients with any positive trigger item served as the study subjects （divided into training and test sets at a ratio of 7∶3）， candidate feature variables were selected from 39 related variables using the Boruta algorithm， and the multivariable Logistic regression analysis was performed with the occurrence of SAR as the dependent variable. Based on these candidate features， Logistic Regression， Extreme Gradient Boosting， Light Gradient Boosting Machine， Random Forest， and Categorical Boosting models were constructed. Model performance was evaluated using metrics including the area under the curve （AUC） of receiver operating characteristic curve and recall rate. The Shapley Additive exPlanations （SHAP） method was applied to analyze and interpret the contribution of each variable. A nomogram was constructed based on the optimal model. RESULTS A total of 38 trigger items for active monitoring of bevacizumab-related ADR were determined， comprising 17 laboratory indicators， 13 clinical manifestations， and 8 intervention measures. In total， 483 patients with positive trigger items were included， and 318 patients with bevacizumab-induced ADR were identified， including 83 SARs. The positive predictive values for the trigger items and cases were 43.57% （708/1 625） and 63.84% （318/483）， respectively. Bevacizumab-induced ADR involved 7 systems/organs， with the hematological system being the most frequently involved （64.15%）. The Boruta algorithm selected 7 vari ables： serum potassium， hematocrit， albumin-to-globulin ratio， prealbumin， hypertension history， age and red blood cell count. Multivariable Logistic regression showed that elevated serum potassium levels were associated with a decreased risk of bevacizumab-induced SAR （OR＝0.234， P ＝0.002）， while a history of hypertension （OR＝2.642， P ＝0.006） and increased age （OR＝1.040， P ＝0.025） were associated with an increased risk. The Logistic Regression model demonstrated superior performance with higher AUC， F1 score and recall rate （0.761， 0.447， 0.607）， compared to other models. SHAP evaluation results indicated that variables such as serum potassium， hematocrit， and age ranked highest in importance. CONCLUSIONS Totally 38 trigger entries have been successfully identified for active screening of bevacizumab-related ADR. Elevated serum potassium levels are a protective factor against bevacizumab-induced SAR， whereas the hypertension history and increased age are risk factors. The Logistic Regression model is the optimal predictive model.

Background/Aims: Acetyl coenzyme A (acetyl-CoA) is one of the most essential metabolites in cell metabolism but its function and concentration in hepatocellular carcinoma (HCC) remain elusive and controversial. Methods: A comprehensive analysis of acetyl-CoA levels and acetyl-CoA synthetase 2 (ACSS2) expression across a range of samples, including patient specimens from both hepatitis B virus (HBV) positive and HBV negative HCC individuals, HBV-transgenic mouse HCC models, and multiple cell lines. Furthermore, to evaluate the functional significance of ACSS2 in HBV-related HCC, we implemented both genetic and pharmacological inhibition strategies targeting ACSS2. Molecular mechanism and mitophagy assessment were revealed by cleavage under target and tagmentation sequencing, RNA sequencing, bioinformatic analyses, transmission electron microscopy and JC-1 staining. Results: Our study revealed a distinct metabolic signature of HBV-related HCC, marked by elevated acetyl-CoA, which was driven by ACSS2. ACSS2 was upregulated by the carbohydrate response element-binding protein in HBV-related HCC. Furthermore, ACSS2 improved tumor cell proliferation, an effect that was dependent on its enzymatic activity. Mechanistically, ACSS2-induced acetyl-CoA accumulation activated voltage-dependent anion channels 1 transcription through increased H3K27ac occupancy, which subsequently promoted mitophagy and HBV-related HCC tumorigenesis. Notably, targeting ACSS2 by depletion or inhibition with a catalytic inhibitor significantly suppressed tumor growth. Conclusions These findings not only illustrate the interplay between metabolic reprogramming, epigenetic modification, and tumorigenesis in the context of HBV infection, but also highlight ACSS2 as a novel metabolic vulnerability in HBV-related HCC. Therefore, targeting ACSS2 could be a novel strategy against HBV-related HCC.

Objective:To investigate the drug resistance of newly diagnosed HIV-1 infected patients in Shanghai and to provide reference value for clinical antiretroviral therapy (ART).Methods:The peripheral venous blood plasma of 196 newly diagnosed HIV-1 infected patients screened according to the inclusion and exclusion criteria at the Shanghai Public Health Clinical Center from April to June 2023 was collected, HIV-1 RNA was extracted, the pol region was amplified by reverse transcription-polymerase chain reaction (RT-PCR) for sequencing, the mutation sites and ART drug resistance were analyzed.Results:The plasma of 196 newly diagnosed HIV-1 infected patients was amplified successfully in 162 cases (amplification success rate was 82.65%). The subtypes consisted of CRF07_BC(51.23%), CRF01_AE (27.78%), and others (6.79%), CRF55_01B (5.56%), B (3.70%), CRF01_AE/B (3.70%) and CRF08_BC (1.23%). The overall transmitted drug resistance rate was 7.41%, the protease inhibitors (PIs), non-nucleoside/nucleotide reverse transcriptase inhibitors (NNRTIs), nucleoside/nucleotide reverse transcriptase inhibitors (NRTIs), integrase inhibitors (INSTIs) resistance rates were 3.09%, 3.70%, 0.00% and 0.62%, respectively. The proportion of NNRTIs-related mutation sites in B (66.67%) and CRF55_01B (88.89%) was higher than that in CRF07_BC (13.25%); the proportion of NNRTIs-related mutation sites in CRF55_01B (88.89%) was higher than that in CRF01_AE (22.22%) and other subtypes (18.18%), the difference was statistically significant (all P<0.05). Multivariate logistic regression analysis showed that the probability of PIs-related mutation sites in CRF01_AE/B was 21.71 times that of CRF07_BC[odds ratio ( OR)=21.71, 95% confidence interval ( CI): 3.36-140.27, P=0.001]. Conclusions:The transmitted drug resistance among newly diagnosed HIV-1 infected patients in Shanghai is at the moderate epidemic level, mainly NNRTIs and PIs-related drug resistance, and the INSTIs resistance rate is low, the use of INSTIs in ART regimens should be considered.

Objective To compare and find an optimal model for predicting the risk of DKD occurrence in patients with type 2 diabetes mellitus(T2DM).Methods A total of 2005 patients with T2DM were enrolled in this study from The Second Hospital of Shijiazhuang City during December 2017 to December 2022.All the subjects were divided into a training set(n=1403)and a validation set(n=602)according to the ratio of 3∶1 by simple random sampling.With the occurrence of DKD as the outcome variablein the training set,important feature variables were screened by LASSO regression.Six different machine learning models were established according to the feature variables,thenthe optimal model was determined by comparison,and anonlinerisk predictor for DKD occurrence was constructed in patients with T2DM.Results Taking the occurrence of DKD as the outcome variable in the training set,the results of LASSO regression analysis showed that the optimal value of the model was 10-fold cross validation lambda.1se=0.01662473,and 15 characteristic variables with nonzero coefficient were screened out to be related to the occurrence of DKD.The data included sex,age,family history of DM,DM duration,LDL-C,HbA1c,WBC,PDW,Scr,urine α1-microglobulin,urine β2-microglobulin,urine microalbumin,hypertension,hypokalemia,and DR.In the training set and validation set,the prediction performance of XGBoost model was better than that of other models(AUC=0.872,0.893,95%CI 0.853～0.891,0.865～0.921),the sensitivity was 0.779,0.863,and the specificity was 0.721,0.758,respectively.The F1 scores were 0.774 and 0.787.DCA analysis showed that the XGBoost model had a greater net benefit and threshold probability.According to the XGBoost model,the online predictor of DKD risk in T2DM patients was laid out,and two patients were selected for application,the results showed that the predictive value of the model was 0.185 in non-DKD patients,and the predictive value was 0.510 in DKD patients.Conclusions The XGBoost model is the best model for predicting the occurrence of DKD in T2DM patients,and an online predictor was successfully built.

Objective To explore the relationship between triglyceride-glucose(TyG)index and hypertension in patients with cerebral hemorrhage.Methods A total of 1718 patients with cere-bral hemorrhage admitted to our hospital from January,2013 to May,2023 were enrolled in this study.According to the TyG index quartile,437 cases were assigned into Q1 group(≤8.375),424 cases into Q2 group(TyG index 8.376～8.737),429 cases into Q3 group(TyG index 8.738～9.087),and 428 cases into Q4 group(≥9.088).The general clinical data were compared in the four groups.Logistic regression analysis was used to study the correlation between TyG index and hy-pertension.Results There were significant differences among the four groups in terms of age,hy-pertension,diabetes,SBP,DBP,TC,TG,HDL-C,LDL-C,FBG,glycated hemoglobin and TyG in-dex(P＜0.05,P＜0.01).Logistic regression analysis showed that when the TyG index was a con-tinuous variable,it was significantly correlated with the risk of hypertension(OR=1.999,95％CI:1.393-2.869,P=0.001).When the index was used as a categorical variable,with Q1as a ref-erence,TyG index in Q3 and Q4 was associated with an increase in OR of hypertension(OR=1.869,95％CI:1.220-2.865,P=0.004;OR=1.844,95％CI:1.125-3.020,P=0.015).After ad-justing cofounders,the association of TyG index and risk of hypertension was stronger in the fe-males(OR=2.618,95％CI:1.312-5.221,P=0.006)than the males(OR=1.783,95％CI:1.151-2.761,P=0.010),and in the patients ≥65 years old(OR=3.277,95％CI:1.600-6.741,P=0.001)than those＜65 years old(OR=1.782,95％CI:1.076-2.949,P=0.025).Conclusion TyG index is closely associated with hypertension in patients with cerebral hemorrhage,especially in women and elderly.

Objective To compare and find an optimal model for predicting the risk of DKD occurrence in patients with type 2 diabetes mellitus(T2DM).Methods A total of 2005 patients with T2DM were enrolled in this study from The Second Hospital of Shijiazhuang City during December 2017 to December 2022.All the subjects were divided into a training set(n=1403)and a validation set(n=602)according to the ratio of 3∶1 by simple random sampling.With the occurrence of DKD as the outcome variablein the training set,important feature variables were screened by LASSO regression.Six different machine learning models were established according to the feature variables,thenthe optimal model was determined by comparison,and anonlinerisk predictor for DKD occurrence was constructed in patients with T2DM.Results Taking the occurrence of DKD as the outcome variable in the training set,the results of LASSO regression analysis showed that the optimal value of the model was 10-fold cross validation lambda.1se=0.01662473,and 15 characteristic variables with nonzero coefficient were screened out to be related to the occurrence of DKD.The data included sex,age,family history of DM,DM duration,LDL-C,HbA1c,WBC,PDW,Scr,urine α1-microglobulin,urine β2-microglobulin,urine microalbumin,hypertension,hypokalemia,and DR.In the training set and validation set,the prediction performance of XGBoost model was better than that of other models(AUC=0.872,0.893,95%CI 0.853～0.891,0.865～0.921),the sensitivity was 0.779,0.863,and the specificity was 0.721,0.758,respectively.The F1 scores were 0.774 and 0.787.DCA analysis showed that the XGBoost model had a greater net benefit and threshold probability.According to the XGBoost model,the online predictor of DKD risk in T2DM patients was laid out,and two patients were selected for application,the results showed that the predictive value of the model was 0.185 in non-DKD patients,and the predictive value was 0.510 in DKD patients.Conclusions The XGBoost model is the best model for predicting the occurrence of DKD in T2DM patients,and an online predictor was successfully built.

Objective:This multicenter retrospective study aimed to analyze the clinicopathological features of duodenal adenocarcinoma (DA) and identify prognostic factors for postoperative survival.Methods:Demographic characteristics, clinicopathological features, treatment outcomes and survival of DA patients undergoing surgical treatment at 18 Chinese medical centers from January 2012 to December 2023 were retrospectively analyzed.Results:Among the 2 056 DA patients included, 46.8% (963) had extra-ampullary DA (EA-DA), and 53.2% (1 093) had peri-ampullary DA (PA-DA). The 1-, 3-, and 5-year overall survival (OS) rates for patients who underwent radical surgery were 93.2%, 71.0%, and 57.2%, respectively. The median overall survival was 76 months, and the median progression-free survival (PFS) was 65 months. No differences in survival were observed between the laparotomy group and minimally invasive surgery (MIS) group either before or after propensity score matching (OS: 76 vs. 75 months before PSM, P=0.986; OS: 75 vs. 75 months after PSM, P=0.602). Furthermore, there were no significant differences between-group in operation time and postoperative complications ( P＞0.05). The MIS group experienced less intraoperative blood loss and shorter hospital stays. The multivariate Cox regression analysis revealed that advanced age ( HR=1.43,95% CI:1.18-1.73), elevated carbohydrate antigen 19-9 levels ( HR=1.24,95% CI:1.02-1.51), perineural invasion ( HR=1.44,95% CI:1.14-1.81), vascular invasion ( HR=1.35,95% CI:1.07-1.71), advanced T stage (T3-4 vs. T1-2: HR=1.86,95% CI:1.49-2.31), regional lymph node metastasis ( HR=1.93,95% CI:1.58-2.36), preoperative biliary drainage ( HR=1.26,95% CI:1.04-1.53), intraoperative blood loss ( HR=1.34,95% CI:1.11-1.62), clinically significant postoperative pancreatic fistulas ( HR=1.53,95% CI:1.12-2.09), and postoperative hemorrhage ( HR=1.62,95% CI:1.14-2.29) were independent risk factors for poor prognosis after surgery (all P＜0.05). Conclusions:Radical surgery is associated with favorable overall survival among DA patients, and no difference in survival is observed between EA-DA and PA-DA patients. MIS is a reliable alternative for DA treatment.

Objective To explore the relationship between triglyceride-glucose(TyG)index and hypertension in patients with cerebral hemorrhage.Methods A total of 1718 patients with cere-bral hemorrhage admitted to our hospital from January,2013 to May,2023 were enrolled in this study.According to the TyG index quartile,437 cases were assigned into Q1 group(≤8.375),424 cases into Q2 group(TyG index 8.376～8.737),429 cases into Q3 group(TyG index 8.738～9.087),and 428 cases into Q4 group(≥9.088).The general clinical data were compared in the four groups.Logistic regression analysis was used to study the correlation between TyG index and hy-pertension.Results There were significant differences among the four groups in terms of age,hy-pertension,diabetes,SBP,DBP,TC,TG,HDL-C,LDL-C,FBG,glycated hemoglobin and TyG in-dex(P＜0.05,P＜0.01).Logistic regression analysis showed that when the TyG index was a con-tinuous variable,it was significantly correlated with the risk of hypertension(OR=1.999,95％CI:1.393-2.869,P=0.001).When the index was used as a categorical variable,with Q1as a ref-erence,TyG index in Q3 and Q4 was associated with an increase in OR of hypertension(OR=1.869,95％CI:1.220-2.865,P=0.004;OR=1.844,95％CI:1.125-3.020,P=0.015).After ad-justing cofounders,the association of TyG index and risk of hypertension was stronger in the fe-males(OR=2.618,95％CI:1.312-5.221,P=0.006)than the males(OR=1.783,95％CI:1.151-2.761,P=0.010),and in the patients ≥65 years old(OR=3.277,95％CI:1.600-6.741,P=0.001)than those＜65 years old(OR=1.782,95％CI:1.076-2.949,P=0.025).Conclusion TyG index is closely associated with hypertension in patients with cerebral hemorrhage,especially in women and elderly.

Objective:To examine the manifestations and causes of administrative burden among primary healthcare professionals,and to explore its impact on job burnout through the mediating role of role conflict,providing theoretical and empirical support for governance-level burden-reduction strategies.Methods:An exploratory sequential mixed-methods design was employed,focusing on primary healthcare professionals in Shandong Province.In the first phase,in-depth interviews were conducted with 175 participants;in the second phase,a questionnaire survey of 1,096 participants and follow-up interviews with 107 participants were carried out.Results:The proportions of respondents who reported"heavy"or"very heavy"burdens were 62.7%for inspection,54.8%for documentation,51.8%for reporting,and 24.4%for meetings.Structural equation modeling showed that administrative burden had a direct effect on job burnout(0.150)and an indirect effect through role conflict(0.093).Qualitative findings further indicated that administrative burden largely stemmed from public health traceability requirements and medical insurance policies,and operated through both resource-based and value-based conflicts.Conclusions:Primary healthcare professionals face considerable administrative burdens,which may heighten job burnout through role conflict.Governance reforms should optimize inspection and assessment,streamline data reporting,refine record-keeping,and promote collaborative governance to break the chain of institutional pressure leading to burnout.