Objective To construct machine learning models based on preoperative inflammatory and radiological features for the prediction of glioma grading and isocitrate dehydrogenase (IDH) mutation status, and to analyze application values of these models and identify the optimal predictive models. Methods A retrospective analysis was conducted on the data of pathologically confirmed glioma patients admitted to Tongji Hospital Affiliated to Tongji University from March 2019 to March 2023. LASSO regression was used to screen feature variables, and predictive models were constructed based on logistic regression (LR), random forest (RF), support vector machine (SVM), gradient boosting decision tree (XGBoost) and K-nearest neighbor (KNN) algorithms. The model performance was comprehensively evaluated using metrics including discrimination ability, area under the precision-recall curve (AUC), accuracy, F1 score and Brier score. The DeLong test was adopted to compare the AUC values among different models; Friedman rank-sum test was used to determine the overall performance differences of the models, with the Nemenyi test applied for multiple comparison correction. Results In the task of glioma grading prediction, the LR model achieved the highest comprehensive score (0.726), and no significant difference was observed between the LR model and the other four models; age was positively correlated with glioma grading (P=0.003). In the task of IDH mutation status prediction, the XGBoost model obtained the highest comprehensive score (0.832), which was superior to the LR (0.762, P=0.035) and KNN models (0.754, P=0.025), while no statistical differences were found between the XGBoost model and the RF or SVM models. Conclusions The LR model for glioma grading prediction and XGBoost model for IDH mutation prediction constructed based on a task-oriented strategy achieve a favorable interpretability while ensuring optimized performance, thereby providing reliable decision support for the individualized diagnosis and treatment of glioma.

Chronic heart failure （CHF） is a group of complex clinical syndromes caused by abnormal changes in the structure and/or function of the heart due to various reasons， resulting in disorders of ventricular contraction and/or diastole. CHF is a condition where primary diseases such as coronary heart disease， hypertension and pulmonary heart disease recur frequently and persist for a long time， presenting blood stasis in meridians and collaterals， stagnation of water and dampness， and accumulation of Qi in collaterals. Its pathogenesis is complex and may involve myocardial energy metabolism disorders， oxidative stress responses， myocardial cell apoptosis， autophagy， inflammatory responses， etc. According to the theory of restraining hyperactivity to acquire harmony， we believe that under normal circumstances， the adenosine monophosphate-activated protein kinase （AMPK） signaling pathway functions normally， maintaining human physiological activities and energy metabolism. Under pathological conditions， the AMPK signaling pathway is abnormal， causing energy metabolism disorders， inflammatory responses， and myocardial fibrosis. Traditional Chinese medicine （TCM） can regulate the AMPK signaling pathway through multiple mechanisms， targets， and effects， effectively curbing the occurrence and development of CHF. It has gradually become a research hotspot in the prevention and treatment of this disease. Guided by the theory of TCM， our research group， through literature review， summarized the relationship between the AMPK pathway and CHF and reviewed the research progress in the prevention and control of CHF with TCM active ingredients， TCM compound prescriptions， and Chinese patent medicines via regulating the AMPK pathway. The review aims to clarify the mechanism and targets of TCM in the treatment of CHF by regulating the AMPK pathway and guide the clinical treatment and drug development for CHF.

ObjectiveTo investigate the therapeutic effect of Astragali Radix-mediated changes in gut microbiota on treating type 2 diabetes （T2DM）. MethodsA 12-week randomized， placebo-controlled clinical trial enrolled eighty patients with newly diagnosed type 2 diabetes and poor glycemic control in the Qi deficiency type. All patients received insulin therapy. The observation group （40 cases） was administered with Astragali Radix Granules， while the control group （40 cases） received a placebo. Both treamtents were taken orally twice daily. Changes in gut microbiota were assessed by 16s rDNA sequencing. Serum glucagon-like peptide-1 （GLP-1） levels were measured using enzyme-linked immunosorbent assay （ELISA）. Glucose metabolism indicators including fasting blood glucose （FPG）， 2-hour postprandial blood glucose （2 h PG），glycated albumin（GA）， and glycated hemoglobin （HbA1c） were evaluated. Pancreatic function was evaluated using fasting C-peptide （FCP）， 2-hour postprandial C-peptide （2 h CP）， and C-peptide area under the curve （AUC_cp）. Traditional Chinese medicine （TCM） syndrome scores， clinical efficacy， and safety indicators were also observed. ResultsIn terms of glucose metabolism indicators， compared with the baseline， both groups exhibited significantly lower FPG， 2 h PG， GA and HbA1C （P<0.01），while FCP， 2 h CP and AUC_cp were significantly higher （P<0.01）. Compared with the control group after the treatment， the observation group showed significantly lower FPG， 2 h PG， GA and HbA1C（P<0.05， P<0.01），and significantly higher FCP， 2 h CP and AUC_cp （P<0.05， P<0.01）， indicating that Astragali Radix can improve glucose metabolism. In terms of the diversity of gut microbiota， no significant differences were detected in the Chao1， Shannon and Simpson indexes of the two groups compared with their respective baselines. However， compared with the post-treatment control group， the observation group demonstrated significant increases in the Chao1， Shannon and Simpson indexes （P<0.05， P<0.01）. The β-diversity analysis showed significant separation in gut microbiota composition before and after treatment in both groups， indicating that Astragali Radix can significantly alter the structure and improve the diversity of gut microbiota. At the phylum level， compared with the baseline， both groups showed a significant increase in the relative abundance of Bacteroidota（P<0.01）. The relative abundance of the potentially harmful phylum Proteobacteria was significantly lower in the observation Group after treatment （P<0.01）. Compared with the post-treatment control group， the observation group had a significantly higher relative abundance of Bacteroidota（P<0.01）. No significant difference was found in Firmicutes/Bacteroidota （F/B） ratio between the two groups after treatment， and other phyla showed no significant differences. At the genus level， compared with the baseline， the observation group exhibited a significant increase in Bacteroides （P<0.01） and a significant decrease in Escherichia-Shigella （P<0.01）， whereas no significant difference was seen in the control group . Compared with the control group after treatment， the observation group after treatment had a significantly higher relative abundance of Bacteroides （P<0.01）. No significant differences were seen in other genera. Linear discriminant analysis （LDA） identified potential characteristics taxa： in the observation group， Bacteroidota at the phylum level and Bacteroides and Dubosiella at the genus level， in the control group， Proteobacteria at the phylum level as well as Barnesiella and Staphylococcus at the genus level. Correlation analysis based on a heatmap revealed that GLP-1 levels were positively correlated with Firmicutes， F/B ratio and Fusobacterium， and negatively correlated with Bacteroidota， Proteobacteria， Bacteroides and Escherichia-Shigella. In terms of clinical efficacy， compared with the control group， the total effective rate of the observation group was significantly higher （P<0.05）. Compared with the baseline， the scores for shortness of breath， fatigue， weakness， spontaneous sweating and reluctance to speak significantly decreased in both groups （P<0.01）. Compared with the control group after treatment， the score for weakness was significantly lower in the observation group （P<0.01），indicating that Astragali Radix could improve clinical symptoms and alleviate weakness symptoms. In terms of safety， compared with the baseline， alanine aminotransferase （ALT） levels significantly decreased in both groups （P<0.05，P<0.01），indicating that Astragali Radix did not induce any significant abnormalities in liver and kidney functions. ConclusionAstragali Radix demonstrates the potential to significantly improve the gut microbiota environment in patients of newly diagnosed type 2 diabetes with Qi deficiency. The therapeutic effect may contribute to glycemic control， possibly mediated by an elevation in GLP-1 level. These findings may support its further clinical investigations and potential applications.

OBJECTIVE To introduce the development of an intelligent prescription review decision-support system for anti-tumor drugs and assess its clinical application outcomes. METHODS Relevant data sources， including national and local pharmaceutical administration policies， clinical practice guidelines/consensus， hospital information systems data， and genetic testing results， were integrated. Adhering to the principles of structure， standardization and dynamic updating， a knowledge base covering chemotherapeutic， targeted and immunotherapeutic agents was constructed using a dual-dimensional modeling approach that combined “drug attributes” and “clinical contexts”. This knowledge base was then embedded into the hospital’s electronic medical order system to establish the prescription review decision-support system. The application and performance of the system were evaluated at Shanghai Sixth People’s Hospital Affiliated to Shanghai Jiao Tong University School of Medicine. RESULTS A knowledge base containing 18 318 prescription review rules for anti-tumor drugs was constructed， and a closed-loop prescription review system was successfully established， encompassing pre-prescription real-time intervention， in-process interactive review， and post-prescription evaluation and analysis. From 2021 to 2024， the system generated a total of 57 879 alerts for prescriptions of five typical categories of anti-tumor drugs. For platinum-containing prescriptions， 22 577 alerts were generated， with Cisplatin for injection （lyophilized） being the most frequently alerted drug （13 445 alerts）， and “ototoxicity risk due to combined use” alerts remained high （7 682 alerts）. For methotrexate-containing prescriptions， 3 721 alerts were recorded， primarily related to “precaution-related issues” （76.4%， 2 843/3 721）. For doxorubicin-containing prescriptions， 17 301 alerts were triggered， primarily related to “dosage and administration” （14 315 alerts）. For human epidermal growth factor receptor 2-targeted agents-containing prescriptions， 1 007 alerts were issued， mostly related to “reimbursement restrictions” （956 alerts）. For programmed death-1/programmed death-ligand 1 inhibitors-containing prescriptions， the alerts increased year by year， totaling 13 273 alerts， primarily related to “inappropriate indication” （9 118 alerts）. Over the 4 years， the physician response rates to system alerts were 21.4%， 27.1%， 33.5% and 51.6%， respectively. CONCLUSIONS An intelligent decision-support system for anti-tumor drug prescription review， encompassing a closed-loop process of “real-time pre-event intervention， interactive in-event prescription review， post-event evaluation and analysis”， has been successfully constructed and implemented throughout the entire workflow. There is a discernible trend in this hospital， where the focus on monitoring anti-tumor drugs is shifting towards immunotherapy drugs. Additionally， the acceptance rate of physicians regarding prescription review opinions has been steadily increasing year by year.

ObjectiveTo investigate the patterns of chronic disease comorbidity among permanent elderly residents in Xuhui District of Shanghai, and to analyze the impact of different comorbidity patterns on the elderly’s ability to perform activities of daily living. MethodsBased on data from the 2015 and 2021 Surveys on Health Status and Health Service Utilization among Permanent Residents in Xuhui District of Shanghai, the study included permanent residents who fully participated in both surveys, were aged 60 years or older at the 2015 survey, self-reported having a chronic disease, and reported no significant changes in their chronic disease status between the two surveys. A prospective cohort study design was adopted, using the 2015 data as the baseline and the 2021 survey data as follow-up data. Physical examinations and questionnaire surveys were conducted by the Shanghai Xuhui District Center for Disease Control and Prevention in the subdistricts and towns within the jurisdiction from April to July in both 2015 and 2021. Latent class analysis (LCA) was used to classify comorbidity patterns across 11 types of chronic diseases. The smaller the values of the Akaike information criterion (AIC), the Bayesian information criterion (BIC), and the sample-corrected Bayesian information criterion (aBIC), the better the model fits. The Barthel index (BI) and the Lawton and Brody Instrumental Activities of Daily Living Scale (LB-IADL) were used to assess the participants’ activities of daily living (ADL) and instrumental activities of daily living (IADL). Multiple linear regression and multivariate logistic regression models were used to analyze the association between different comorbidity patterns and ADL and IADL. ResultsA total of 1 608 study participants were enrolled. Among them, 671 (41.73%) self-reported having no chronic diseases. The prevalence of 11 types of chronic diseases was as follows: hypertension in 743 participants (46.21%), heart disease in 148 participants (9.20%), cerebrovascular disease in 37 participants (2.30%), diabetes or abnormal blood glucose in 310 participants (19.28%), other endocrine system diseases (non-diabetic) in 39 participants (2.43%), malignant tumors in 37 participants (2.30%), chronic lung disease in 16 participants (1.00%), gout in 8 participants (0.50%), musculoskeletal disorders in 34 participants (2.11%), gastric or digestive system diseases in 27 participants (1.68%), and dyslipidemia in 15 participants (0.93%). LCA was used to classify comorbidity patterns of chronic diseases among older adults into 1 to 9 latent category combinations. Model 5 (i.e., 5 latent categories) had the lowest BIC (7 112.69) and aBIC (6 909.38), with a relatively low AIC (6 768.20), indicating the best model fit; thus, five latent categories represented the optimal combination in this study. The five latent categories (and their respective proportions of study participants) were the healthy control group (41.73%), the hypertension group (31.28%), the diabetes-hypertension group (19.28%), the severe illness group (5.29%), and the other metabolic diseases group (2.43%). According to multiple linear regression analyses, compared with the healthy control group, the diabetes-hypertension group had significantly lower BI-20 scores (b=-0.952, 95%CI: -1.317‒ -0.587, P<0.001) and lower IADL scores (b=-0.744, 95%CI: -0.999‒ -0.489, P<0.001); the severe illness group (b=-0.644, 95%CI: -1.072‒ -0.217, P=0.003) and the hypertension group (b=-0.344, 95%CI: -0.562‒ -0.125, P=0.002) had lower IADL scores. According to multivariate logistic regression analyses, compared with the healthy control group, older adults in the diabetes-hypertension group had a higher risk of disability (OR=2.835, 95%CI: 1.392‒5.774), impaired ADL function (OR=2.470, 95%CI: 1.637‒3.726), and impaired IADL function (OR=1.739, 95%CI: 1.264‒2.392); older adults in the severe illness group had a higher risk of impaired ADL function (OR=2.206, 95%CI: 1.171‒4.155). ConclusionAmong the elderly participating in the Survey on Health Status and Health Service Utilization of Permanent Residents in Xuhui District of Shanghai, diabetes-hypertension is one of the key chronic disease comorbidity combinations and has adverse effects on both ADL and IADL. Community health management should prioritize the elderly with diabetes-hypertension comorbidity, and strengthen early screening and comprehensive interventions to slow the decline in ADL and IADL.

ObjectiveTo observe the effect of Yifei Jianpi prescription on the of signal transducer and activator of transcription protein 1 （STAT1）/interferon regulatory factor 3 （IRF3） signaling pathway in a pneumonia model induced by lipopolysaccharide （LPS） and to explore the mechanism of Yifei Jianpi prescription in improving lung immune and inflammatory responses. MethodsSixty male SPF SD rats were used in this study. Ten rats were randomly assigned to the normal control group， and the remaining 50 were instilled with LPS in the trachea to establish a pneumonia model. After successful modeling， the rats were randomly divided into the model group， dexamethasone group （0.5 mg·kg^-1）， and Yifei Jianpi prescription high-dose （12 mg·kg^-1）， medium-dose （6 mg·kg^-1）， and low-dose （3 mg·kg^-1） groups， with 10 rats in each group. Treatment was administered once daily， and the normal control and model groups received the same volume of normal saline. After 14 days， flow cytometry was used to detect the classification of whole blood lymphocytes. Enzyme-linked immunosorbent assay （ELISA） was used to measure serum levels of immunoglobulin G （IgG）， immunoglobulin A （IgA）， immunoglobulin M （IgM）， and the content of tumor necrosis factor-α （TNF-α）， interleukin-8 （IL-8）， interleukin-6 （IL-6）， and interleukin-10 （IL-10） in alveolar lavage fluid （BALF）. Hematoxylin-eosin （HE） staining was used to observe lung tissue pathology and score the damage. Thymus weight， spleen weight， and wet-to-dry weight ratio （W/D） were recorded. Real-time fluorescence quantitative polymerase chain reaction （Real-time PCR） was used to detect the mRNA expression of STAT1， IRF3， IL-6， and interferon-alpha （IFN-α） in lung tissues， while Western blot was performed to assess the protein expression of STAT1， IRF3， IL-6， and IFN-α. ResultsCompared with the normal control group， the model group showed significantly increased proportion of B lymphocytes in peripheral blood， decreased proportions of NK cells and CD4⁺/CD8⁺ （P<0.05， P<0.01）， decreased serum levels of IgG and IgA， significantly increased IgM levels （P<0.01）， significantly elevated content of TNF-α， IL-6， and IL-8 in BALF， and significantly decreased IL-10 levels （P<0.01）. Lung tissue damage was evident， with significant increases in thymus and spleen weights and a higher W/D ratio （P<0.01）. The mRNA and protein expression of STAT1， IRF3， IFN-α， and IL-6 in lung tissues was significantly upregulated （P<0.05，P<0.01）. Compared with the model group， the Yifei Jianpi prescription groups showed significantly reduced proportions of B lymphocytes in peripheral blood， increased proportions of NK cells and CD4⁺/CD8⁺ ratios （P<0.05， P<0.01）， significantly increased serum levels of IgG and IgA， significantly decreased IgM levels （P<0.05， P<0.01）， significantly reduced levels of TNF-α， IL-6， and IL-8 in BALF， and significantly increased IL-10 levels （P<0.01）. Lung tissue damage was alleviated， thymus and spleen weights were significantly reduced， and the W/D ratio was markedly decreased （P<0.01）. The mRNA and protein expression of STAT1， IRF3， IFN-α， and IL-6 in lung tissues was significantly downregulated （P<0.05， P<0.01）. ConclusionYifei Jianpi prescription can alleviate lung tissue damage and improve immune and inflammatory responses in LPS-induced pneumonia rats. The mechanism may be related to the inhibition of STAT1/IRF3 signaling pathway activation.

ObjectiveTo introduce the three main techniques for tuberculosis screening currently used in China, to systematically evaluate their accuracy in diagnosing latent tuberculosis infection (LTBI), so as to provide scientific basis and recommendations for the formulation of China’s tuberculosis screening strategy. MethodsLiterature on the diagnosis of tuberculosis by tuberculin skin test (TST), interferon-γ release assay (IGRA), and recombinant Mycobacterium tuberculosis fusion protein (EC) skin test from January 1, 2010 to August 22, 2024 was comprehensively retrieved from PubMed, China National Knowledge Infrastructure (CNKI), and Wanfang Database through computerized search. Besides, all the literature was screened in accordance to the inclusion criteria for diagnostic tests, and characteristic information of the literature selected was extracted simultaneously. Meta-analysis was performed using Stata 17.0 software, with a random-effects model used for weighted quantitative synthesis of included literature, calculating pooled sensitivity, specificity, positive likelihood ratio, negative likelihood ratio, and their 95% confidence intervals (CI). ResultsA total of 543 relevant articles were retrieved, with 105 ultimately included. Among them, 33 articles reported diagnostic data for TST, with a pooled sensitivity of 0.68 (95%CI: 0.62‒0.73), specificity of 0.67 (95%CI: 0.60‒0.73), positive likelihood ratio of 2.0 (95%CI: 1.7‒2.5), and negative likelihood ratio of 0.48 (95%CI: 0.40‒0.58). Ninety-four articles reported the diagnostic value of IGRAs test, with a pooled sensitivity of 0.88 (95%CI: 0.87‒0.89), specificity of 0.82 (95%CI: 0.79‒0.84), positive likelihood ratio of 4.8 (95%CI: 4.2‒5.6), and negative likelihood ratio of 0.15 (95%CI: 0.13‒0.17). Data on EC skin test was limited, but preliminary analysis showed that it had high sensitivity and specificity. ConclusionIGRA has a significant advantage in diagnosing LTBI, and EC skin test also shows good diagnostic performance, although relevant data is limited. TST remains suitable for large-scale screening due to its cost-effectiveness.

Objective To observe the effects of Yifei Jianpi Prescription on airway mucus hypersecretion and protein expressions of EGFR/PKC/NF-κB pathway in lipopolysaccharide(LPS)-induced acute lung injury(ALI)model rats;To explore its mechanism in the treatment of ALI.Methods Ten of 60 SPF SD rats were randomly selected as blank group,and the other rats were intratracheal instilled with LPS to establish ALI model.The model rats were randomly divided into model group,dexamethasone group and Yifei Jianpi Prescription high-,medium-and low-dosage groups,with 8 rats in each group.Each treatment group was given corresponding drug solution by gavage,and the blank group and model group were given equal volume of normal saline by gavage,once a day for 14 days.The pulmonary functions of rats were measured[peak expiratory flow(PEF),tidal volume(TV),expiratory volume(EV),50%expiratory flow rate(EF50)],HE staining was used to observe the morphology of lung tissue,AB-PAS staining was used to evaluate the proliferation and mucus secretion of goblet cells,the expressions of epidermal growth factor receptor(EGFR),protein kinase C(PKC),nuclear factor-κB(NF-κB)p65 and MUC5AC in lung tissue were detected by immunofluorescence staining,the mRNA expressions of EGFR and MUC5AC in lung tissue were detected by fluorescent quantitative PCR,and the content of MUC5AC in lung tissue was detected by ELISA.Results Compared with the blank group,PEF,TV,EV and EF50 of the model group rats significantly decreased(P<0.01);the bronchial wall was significantly thickened,the lumen narrowed,pulmonary interstitial edema and hyperemia,the thickness of alveolar wall increased,accompanied by a large number of inflammatory cells infiltration,and the lung tissue injury score increased significantly(P<0.01);goblet cells proliferated significantly,mucus secretion increased significantly(P<0.01);the protein expressions of EGFR,PKC,NF-κB p65,MUC5AC and mRNA expressions of EGFR and MUC5AC in lung tissue increased significantly(P<0.01),and the content of MUC5AC in lung tissue increased significantly(P<0.01).Compared with the model group,PEF,TV,EV and EF50 in dexamethasone group and Yifei Jianpi Prescription each dosage groups increased in varying degrees;the pathological injury of lung tissue was alleviated to varying degrees,the score of lung tissue injury was reduced;the proliferation of goblet cells was reduced,and the secretion of mucus was reduced,the expressions of EGFR,PKC,NF-κB p65,MUC5AC protein and EGFR,MUC5AC mRNA in lung tissue decreased,and the content of MUC5AC in lung tissue decreased.There was statistical significance in dexamethasone group and Yifei Jianpi Prescription high-and medium-dosage groups(P<0.01).Conclusion Yifei Jianpi Prescription can inhibit the hypersecretion of airway mucus and the high expression of EGFR/PKC/NF-κB pathway protein in rats with ALI induced by LPS.

Objective To observe the synergistic effect of Shengxian Huaxian Prescription and its disassembled prescription on pulmonary fibrosis;To explore whether its mechanism is related to regulating the STAT6/PPAR-y pathway to promote polarization of M2 type macrophages towards M1 type.Methods Ten SD rats were randomly selected from 70 rats as blank group,and the remaining rats were re-established pulmonary fibrosis model by intratracheal infusion of bleomycin.After modeling,the rats were divided into model group,positive group,Shengxian Huaxian Prescription group,Shengxian group,Tongluo group and Bushen group,with 10 rats in each group.Shengxian Huaxian Prescription group,Shengxian group,Tongluo group and Bushen group were given 12.60,7.65,3.60 and 2.25 g/kg of corresponding TCM solution,respectively;the positive group was given 0.12 g/kg of pirfenidone suspension;the blank group and the model group were given equal volume of normal saline,once a day,for consecutive 28 days.The lung function of rats was detected,the contents of IL-6 and TGF-β1 in serum were detected by ELISA,the pathological changes in lung tissue were observed by Masson staining,the expression of CD68,iNOS and CD206,Arg-1 in lung tissue were detected by immunofluorescence,the expression of SOCS1,SOCS3,STAT6,p-STAT6 and PPAR-γ in lung tissue were detected by Western blot.Results Compared with the blank group,the PEF,PIF and EF50 in model group rats significantly decreased,and the contents of serum IL-6 and TGF-β1 significantly increased,Masson staining showed a large amount of collagen fiber deposition,Ashcroft score significantly increased,CD206,Arg-1,STAT6,p-STAT6,PPAR-y protein expression significantly increased(P＜0.01),the expressions of SOCS1 and SOCS3 protein significantly decreased(P＜0.01),while the expression of CD68 and iNOS were not significantly changed(P＞0.05).Compared with the model group,the PEF,PIF and EF50 in all administration groups significantly increased,and the contents of serum IL-6 and TGF-β1 significantly decreased,collagen fiber deposition in lung tissue were decreased to varying degree,Ashcroft score significantly decreased,the expression of CD206,Arg-1,STAT6,p-STAT6 and PPAR-γ protein significantly decreased(P＜0.01),the expressions of CD68,iNOS,SOCS1 and SOCS3 protein significantly increased(P＜0.05).The above indicators showed the most significant changes in Shengxian Huaxian Prescription group,followed by Shengxian group(P＜0.01,P＜0.05).Conclusion Both Shengxian Huaxian Prescription and its disassembled prescription have anti pulmonary fibrosis effects,and their mechanism may related to regulating the STAT6/PPAR-y pathway and promoting polarization of M2 type macrophages towards M1 type.Shengxian Huaxian Prescription group has the best effect,while Shengxian group play an important role in the prescription,and the compatibility between each group has a synergistic effect.

Objective To develop a risk prediction model for moderate-to-severe catheter-related bladder discomfort(CRBD)after non-transurethral surgery based on various machine-learning algorithms and to compare the performance of the models,so as to provide a reference for accurately identification and prevention of the postoperative moderate-to-severe CRBD.Methods A convenience sampling method was employed to recruit 719 patients as study subjects.The patients received non-transurethral surgery and intraoperative urinary catheterisation in a Tier-ⅢA hospital in Shanxi Province between January and May 2024.The clinical data were collected,with 70％of the randomly selected data was assigned to a training dataset(n=503)for the model building and the rest of 30％of data was used as the testing dataset(n=216)for internal model validation.Predictors were determined using least absolute shrinkage and selection operators(LASSO).Seven machine learning methods of logistic regression,K-nearest neighbours,random forest,artificial neural network,decision tree,light gradient boosting machine(LightGBM)and elastic net were employed to establish the risk prediction models.Performance of the models was evaluated based on the area under receiver operating characteristic curve(AUR-ROC),accuracy,precision,recall and F1 score.Results A total of 719 patients who underwent non-transurethral surgery were included in the study.It was found that 154(21.4％)patients presented with moderate to severe CRBD and 565(78.6％)patients were without or only with a mild CRBD.The predictors were deduced to six variables:gender,abdominal surgery,type of surgery,administration of dexmedetomidine before surgery,intraoperative administration of flurbiprofenate,and use of tramadol by the completion of surgery.It was found that the LightGBM model demonstrated a high stability,with 0.793 in AUC-ROC,0.763 in accuracy,0.879 in precision,0.747 in recall and 0.808 in F1.Conclusion The risk prediction model established through LightGBM for moderate-to-severe CRBD after a non-transurethral surgery exhibits a high stability.It offers a reference for medical practitioners to identify the patients with high-risk of moderate-to-severe CRBD and prepares for relevant interventions.