Objective:To develop and validate a prediction model for medication adherence among patients receiving allergen sublingual immunotherapy (SLIT).Methods:A prospective cross-sectional study was conducted, and a total of 288 patients who received SLIT treatment at an allergy center in the First Affiliated Hospital with Nanjing Medical University (Jiangsu Province Hospital) from December 2023 to July 2024 were assigned to the modeling group. Additionally, 122 patients from August to October 2024 were assigned to the validation group. Data of patients′ general information, medication beliefs, anxiety levels, social support, disease perception, and medication adherence were collected. Single-factor analysis and LASSO regression were utilized to identify potential predictors, and a prediction model for medication adherence was constructed using multifactorial logistic regression. A nomogram was then developed based on the model. The model′s discriminatory ability was evaluated using receiver operating characteristic curve (ROC), the area under curve (AUC), sensitivity, and specificity. The model was then validated in the validation cohort.Results:Single-factor analysis and LASSO regression identified a total of nine predictive factors. Logistic regression revealed that medical belief tendency [ OR (95% CI) =2.420 (1.116-5.248), P=0.025], the somatic control dimension in self-rating anxiety scales [ OR (95% CI)=1.404 (1.241-1.589), P<0.001], the subjective support dimension in social support assessment [ OR (95% CI)=0.784 (0.725-0.847), P<0.001], and the cognitive dimension in illness perception [ OR (95% CI)=0.725 (0.647-0.813), P<0.001] were independent predictors of medication adherence in patients undergoing SLIT. The AUC value of the model was 0.899 (95% CI=0.863-0.934) in the modeling group and 0.882 (95% CI=0.820-0.944) in the validation group, indicating good discriminatory ability. The optimal cutoff value of the model was 0.493, with a sensitivity of 81.1% and specificity of 85.7% in the modeling group, and a sensitivity of 87.3% and specificity of 82.5% in the validation group. Conclusion:The medication adherence prediction model developed in this study for patients undergoing SLIT exhibits good predictive performance and provides valuable guidance for early intervention by clinical healthcare professionals.

Objective:To develop CNN-Dinov2, a deep learning-based automatic classification model for Gram-stained images, enabling rapid diagnosis of three prevalent Gram-negative bacilli in bloodstream infections: Escherichia coli ( E.coli), Klebsiella pneumoniae ( K.pneumoniae), and Pseudomonas aeruginosa ( P.aeruginosa). Methods:This evaluation study analyzed 1 425 Gram-stained microscopic images from patients with bloodstream infections at Houjie Hospital, in Dongguan City, collected between January 2023 and January 2024. The images, all positive for blood culture and identified as target strains, were categorized into Escherichia coli (419 images), Klebsiella pneumoniae (411 images), Pseudomonas aeruginosa (413 images), and other Gram-negative bacilli (182 images). They were randomly split into a training set (1 141 images), a validation set (141 images), and a test set (143 images) in an 8∶1∶1 ratio. A hybrid CNN-Dinov2 model was developed by integrating ResNet′s local feature extraction with Dinov2′s global pre-trained features, followed by a fully connected layer. The model was optimized by inputting the preprocessed images and adjusting parameters through loss calculation and backpropagation. AlexNet, Dinov2, and ResNet18 served as control models. The models′ classification performance was assessed using accuracy, precision, weighted F1 score, and recall rate, derived from the confusion matrix. The PR curve and AP value further evaluated each model′s classification capability across the four image categories. Results:The CNN-Dinov2 model achieved a training accuracy of 99.74%, a validation accuracy of 98.12%, and a validation loss of 0.070 6, demonstrating robust generalization without overfitting. Validation metrics revealed superior performance with an accuracy of 98.60%, precision of 98.65%, a weighted F1 score of 98.60%, and a recall rate of 98.60%, outperforming other models. The confusion matrix confirmed its strong classification capability, with the highest sum of diagonal values for identifying four types of bacteria. The macro average precision (AP) values under the precision-recall (PR) curves were all 1, indicating excellent discrimination across all categories. Overall, the CNN-Dinov2 model exhibited the best performance among the four models evaluated.Conclusion:This study successfully developed CNN-Dinov2, an automated classification model for Gram staining images. It offers valuable support for the rapid diagnosis of bloodstream infections caused by Escherichia coli, Klebsiella pneumoniae, and Pseudomonas aeruginosa, demonstrating practical utility.

Objective:To develop and validate a prediction model for medication adherence among patients receiving allergen sublingual immunotherapy (SLIT).Methods:A prospective cross-sectional study was conducted, and a total of 288 patients who received SLIT treatment at an allergy center in the First Affiliated Hospital with Nanjing Medical University (Jiangsu Province Hospital) from December 2023 to July 2024 were assigned to the modeling group. Additionally, 122 patients from August to October 2024 were assigned to the validation group. Data of patients′ general information, medication beliefs, anxiety levels, social support, disease perception, and medication adherence were collected. Single-factor analysis and LASSO regression were utilized to identify potential predictors, and a prediction model for medication adherence was constructed using multifactorial logistic regression. A nomogram was then developed based on the model. The model′s discriminatory ability was evaluated using receiver operating characteristic curve (ROC), the area under curve (AUC), sensitivity, and specificity. The model was then validated in the validation cohort.Results:Single-factor analysis and LASSO regression identified a total of nine predictive factors. Logistic regression revealed that medical belief tendency [ OR (95% CI) =2.420 (1.116-5.248), P=0.025], the somatic control dimension in self-rating anxiety scales [ OR (95% CI)=1.404 (1.241-1.589), P<0.001], the subjective support dimension in social support assessment [ OR (95% CI)=0.784 (0.725-0.847), P<0.001], and the cognitive dimension in illness perception [ OR (95% CI)=0.725 (0.647-0.813), P<0.001] were independent predictors of medication adherence in patients undergoing SLIT. The AUC value of the model was 0.899 (95% CI=0.863-0.934) in the modeling group and 0.882 (95% CI=0.820-0.944) in the validation group, indicating good discriminatory ability. The optimal cutoff value of the model was 0.493, with a sensitivity of 81.1% and specificity of 85.7% in the modeling group, and a sensitivity of 87.3% and specificity of 82.5% in the validation group. Conclusion:The medication adherence prediction model developed in this study for patients undergoing SLIT exhibits good predictive performance and provides valuable guidance for early intervention by clinical healthcare professionals.

The drug resistance of the carbapenem-resistant Enterobacteriaceae(CRE)strains was mainly induced by multiple approaches such as production of carbapenemases,increase of bacterial outer membrane permeability,activation of active efflux pump system,formation of biofilm and drug modifying mechanisms.Those mecha-nisms involve deletion,mutation,insertion and posttranscriptional modification of relevant encoding genes,which may affect the susceptibility of the CRE strains to antibiotics.At present,the conventional clinical thera-pies include the use of traditional antibiotics,either the one-drug use or combined use of drugs.The development of novel antibacterial therapy is under way.The epidemiological characteristics of CRE infections,drug resist-ance mechanisms,current and prospective treatment strategies for CRE infections(covering new application of the drugs in available,the novel drugs such as ceftazidime/avibactam,meropenem/vaborbactam and imipenem/rele-bactam)were deeply reviewed in this article,so as to provide reliable reference for clinical prevention,control and treatment of CRE infections.

Objective To explore the influencing factors of cognitive impairment in non-dialysis patients with chronic kidney disease(CKD).Methods A total of 60 hospitalized non-dialysis patients with CKD in the Department of Nephrology of Northern Jiangsu People's Hospital Affiliated to Yangzhou University from September 2022 to September 2023 were enrolled as research objects.According to the estimated glomerular filtration rate(eGFR),they were divided into stage 1 to 2 of CKD group[eGFR ≥60 mL/(min·1.73 m2)]with 23 cases,the stage 3 of CKD group[eGFR 30～＜60 mL/(min·1.73 m2)]with 20 cases,and stage 4 to 5 of CKD group[eGFR＜30 mL/(min·1.73 m2)]with 17 cases.The Montreal Cognitive Assessment Scale(MoCA)was used to evaluate the cognitive function of the patients.Basic data and common clinical laboratory in-dicators on hospital admission were collected to analyze the differences in cognitive function levels under different renal function statuses and to explore the influencing factors of cognitive impairment.Results The incidence rates of cognitive impairment in the stage 1 to 2 of CKD group,stage 3 of CKD group,and stage 4 to 5 of CKD group were 47.8％,85.0％,and 94.1％respectively,the median MoCA scored 26,24 and 20 respectively,with statistically significant between-group differ-ences(P＜0.05).Cognitive function was significantly negatively correlated with age(r=-0.634,P＜0.001),blood urea nitrogen(BUN)(r=-0.574,P＜0.001),serum creatinine(Cr)(r=-0.417,P＜0.001),cystatin C(Cys-C)(r=-0.327,P=0.011),serum β2-microglobulin(β2-MG)(r=-0.259,P=0.046),and N-terminal pro-brain natriuretic peptide(NT-proBNP)(r=-0.474,P＜0.001),and was significantly positively correlated with hemoglobin(HB)(r=0.401,P=0.001)and eGFR(r=0.485,P＜0.001).Multivariate Logistic regression analysis showed that age(P=0.006)and NT-proBNP(P=0.041)were influencing factors of cognitive im-pairment in non-dialysis patients with CKD.Receiver operating characteristic(ROC)curve analysis showed that the area under the curve(AUC),sensitivity,and specificity of age for prediction were 0.860,0.864 and 0.812 respectively,the AUC,sensitivity,and specificity of NT-proBNP for pre-diction were 0.808,0.795 and 0.875 respectively,and the combined prediction of age and NT-proBNP had an AUC,sensitivity,and specificity of 0.893,0.955,and 0.750,respectively.Conclusion As renal function deteriorates,the incidence rate and severity of cognitive impairment in non-dialysis patients with CKD tend to increase.Advanced age,renal function deterioration,high NT-proBNP level,and anemia are associated with the occurrence of cognitive impairment in non-di-alysis patients with CKD,among which age and NT-proBNP are influencing factors for cognitive im-pairment.

Objective The stress distribution of the lumbar spine L1-5,fibrous rings and nucleus pulposus of the patient in mid-phase of single-leg support under slow-walking gait was studied,to determine the optimal correction angle of the osteotomy block in curved periacetabular osteotomy(CPO),and provide an individualized plan for clinical surgery.Methods The femur-pelvis-lumbar spine DICOM data of a patient and a healthy volunteer were obtained using CT scanning to construct a three-dimensional finite element model.The cortical bone,cancellous bone,and cartilages were delineated using the modeling software,and the model was analyzed by finite element method using simulation software.The patient's lateral center edge angle(LCEA)and anterior center edge angle(ACEA)were 15°,and 16 different postoperative models(LCEA=15°,25°,35°,45° and ACEA=15°,25°,35°,45°)were obtained by computer simulation of the surgical osteotomy process.The stress differences in the regions of interest of the model were compared and analyzed,which were also compared with those of patient before surgery and the healthy volunteer,so as to obtain the optimal surgical plan.Results The stresses applied to the lumbar spine decreased with increasing LCEA and ACEA angles,with the lowest stresses applied to the lumbar cones,nucleus pulposus,and annulus fibrosus in the LCEA=35°,ACEA=35° models;then,the stresses applied increased with increasing angles.Conclusions The optimal correction angle for LCEA and ACEA can be obtained using the finite element method,and this method is of great significance to improve the accuracy and efficiency of CPO for different patients.

Objective The stress distribution of the lumbar spine L1-5,fibrous rings and nucleus pulposus of the patient in mid-phase of single-leg support under slow-walking gait was studied,to determine the optimal correction angle of the osteotomy block in curved periacetabular osteotomy(CPO),and provide an individualized plan for clinical surgery.Methods The femur-pelvis-lumbar spine DICOM data of a patient and a healthy volunteer were obtained using CT scanning to construct a three-dimensional finite element model.The cortical bone,cancellous bone,and cartilages were delineated using the modeling software,and the model was analyzed by finite element method using simulation software.The patient's lateral center edge angle(LCEA)and anterior center edge angle(ACEA)were 15°,and 16 different postoperative models(LCEA=15°,25°,35°,45° and ACEA=15°,25°,35°,45°)were obtained by computer simulation of the surgical osteotomy process.The stress differences in the regions of interest of the model were compared and analyzed,which were also compared with those of patient before surgery and the healthy volunteer,so as to obtain the optimal surgical plan.Results The stresses applied to the lumbar spine decreased with increasing LCEA and ACEA angles,with the lowest stresses applied to the lumbar cones,nucleus pulposus,and annulus fibrosus in the LCEA=35°,ACEA=35° models;then,the stresses applied increased with increasing angles.Conclusions The optimal correction angle for LCEA and ACEA can be obtained using the finite element method,and this method is of great significance to improve the accuracy and efficiency of CPO for different patients.

Objective:To develop CNN-Dinov2, a deep learning-based automatic classification model for Gram-stained images, enabling rapid diagnosis of three prevalent Gram-negative bacilli in bloodstream infections: Escherichia coli ( E.coli), Klebsiella pneumoniae ( K.pneumoniae), and Pseudomonas aeruginosa ( P.aeruginosa). Methods:This evaluation study analyzed 1 425 Gram-stained microscopic images from patients with bloodstream infections at Houjie Hospital, in Dongguan City, collected between January 2023 and January 2024. The images, all positive for blood culture and identified as target strains, were categorized into Escherichia coli (419 images), Klebsiella pneumoniae (411 images), Pseudomonas aeruginosa (413 images), and other Gram-negative bacilli (182 images). They were randomly split into a training set (1 141 images), a validation set (141 images), and a test set (143 images) in an 8∶1∶1 ratio. A hybrid CNN-Dinov2 model was developed by integrating ResNet′s local feature extraction with Dinov2′s global pre-trained features, followed by a fully connected layer. The model was optimized by inputting the preprocessed images and adjusting parameters through loss calculation and backpropagation. AlexNet, Dinov2, and ResNet18 served as control models. The models′ classification performance was assessed using accuracy, precision, weighted F1 score, and recall rate, derived from the confusion matrix. The PR curve and AP value further evaluated each model′s classification capability across the four image categories. Results:The CNN-Dinov2 model achieved a training accuracy of 99.74%, a validation accuracy of 98.12%, and a validation loss of 0.070 6, demonstrating robust generalization without overfitting. Validation metrics revealed superior performance with an accuracy of 98.60%, precision of 98.65%, a weighted F1 score of 98.60%, and a recall rate of 98.60%, outperforming other models. The confusion matrix confirmed its strong classification capability, with the highest sum of diagonal values for identifying four types of bacteria. The macro average precision (AP) values under the precision-recall (PR) curves were all 1, indicating excellent discrimination across all categories. Overall, the CNN-Dinov2 model exhibited the best performance among the four models evaluated.Conclusion:This study successfully developed CNN-Dinov2, an automated classification model for Gram staining images. It offers valuable support for the rapid diagnosis of bloodstream infections caused by Escherichia coli, Klebsiella pneumoniae, and Pseudomonas aeruginosa, demonstrating practical utility.

The drug resistance of the carbapenem-resistant Enterobacteriaceae(CRE)strains was mainly induced by multiple approaches such as production of carbapenemases,increase of bacterial outer membrane permeability,activation of active efflux pump system,formation of biofilm and drug modifying mechanisms.Those mecha-nisms involve deletion,mutation,insertion and posttranscriptional modification of relevant encoding genes,which may affect the susceptibility of the CRE strains to antibiotics.At present,the conventional clinical thera-pies include the use of traditional antibiotics,either the one-drug use or combined use of drugs.The development of novel antibacterial therapy is under way.The epidemiological characteristics of CRE infections,drug resist-ance mechanisms,current and prospective treatment strategies for CRE infections(covering new application of the drugs in available,the novel drugs such as ceftazidime/avibactam,meropenem/vaborbactam and imipenem/rele-bactam)were deeply reviewed in this article,so as to provide reliable reference for clinical prevention,control and treatment of CRE infections.

To investigate the genomic features and perform cluster analysis of Carbapenem-resistant Klebsiella pneumoniae (CRKP) to provide an experimental basis for guiding the prevention and treatment of CRKP infections.A retrospective case-cohort study was conducted on 19 non-redundant CRKP strains isolated from the Tenth Affiliated Hospital of Southern Medical University between January and June 2023. Whole genome sequencing (WGS) and multilocus sequence typing (MLST) were performed to compare genomic features and analyze the resistance genes and homology of the strains.The results showed that the 19 CRKP strains were isolated from 8 different clinical departments, mainly from respiratory specimens. The whole genome sequencing revealed that the genomic lengths of CRKP ranged from 4.90 to 5.85 Mbp, with contigs N50 values>20 kb for each genome. The median overall GC content was 57.0% (50.4%-57.1%). Comparative genomic analysis identified three regions with high genomic variability. WGS detected 32 resistance genes across 11 categories. All 19 strains carried carbapenem resistance genes ( blaKPC-2 and blaOXA-48), blaTEM-1B extended-spectrum β-lactamase resistance genes, qnrS1 quinolone resistance gene, and fosA fosfomycin resistance gene, with each strain carrying only one carbapenemase gene. The detection rate of blaKPC-2 was 94.7% (18/19). MLST identified three sequence types: ST11, ST437 and ST147, with ST11 being predominant (89.5%, 17/19). Clustering analysis based on acquired resistance genes revealed three clonal transmission patterns among strains 72 and 90, and strains 88, 84, 66 and 79.In conclusion, CRKP strains carry multiple resistance genes, and clustering analysis indicating that nosocomial clonal transmission is closely related to acquired resistance genes. The ST11- blaKPC-2 type strain is the predominant clone. Strengthened surveillance and effective control strategies are necessary to reduce nosocomial transmission of CRKP.