Objective:To investigate the prognostic factors for esophageal cancer after radical resection and the application value of machine learning prediction model.Methods:The retrospective cohort study was conducted. The clinicopatholigical data of 406 esophageal cancer patients who were admitted to Qilu Hospital of Shandong University from January 2018 to March 2022 were collected. There were 357 males and 49 females, aged (64±8)years. All patients underwent radical resection of esophageal cancer. The 406 patients were randomly divided into a training set of 285 cases and a validation set of 121 cases at a 7∶3 ratio based on a random number table. The training set was used to construct prediction model, and the validation set was used to validate prediction model. Patients were divided into high-risk group and low-risk group based on risk scores. Observation indicators: (1) follow-up of patients and analysis of influencing factors for prognosis; (2) construction and validation of machine learning prediction models. Comparison of measurement data with normal distribution between groups was conducted using the independent sample t test. Comparison of measurement data with skewed distribution between groups was conducted using the Mann-Whitney U test. Comparison of count data between groups was conducted using the chi-square test. Comparison of ordinal data between groups was conducted using the rank sum test. The Kaplan-Meier method was used to calculate survival rate and plot survival curve, and the Log-rank test was used for survival analysis. The Cox proportional hazard regression model was used for univariate and multivariate analyses. Independent influencing factors were included, and data processing, machine learning model construction, and visualization were performed using R packages including random survival forest (RSF), gradient boosting machine (GBM), least absolute shrinkage and selection operator Cox regression (LASSO-Cox), Cox proportional hazards model boosting (CoxBoost), survival support vector machine (survivalsvm), extreme gradient boosting (XGBoost), supervised principal component analysis (SuperPC), and Cox partial least squares regression (plsRcox). Receiver operating characteristic (ROC) curves were drawn, and sensitivity, specificity, and area under the curve (AUC) were calculated. The Delong test was used to assess the differences in AUC among different models in the training set, and the time-dependent ROC was used to compare the predictive performance of different models. Calibration curves were used to evaluate model accuracy, and decision curve analysis (DCA) was used to evaluate overall net benefit. Results:(1) Follow-up of patients and analysis of influencing factors for prognosis. All 406 patients were followed up postoperatively for 28(range, 6-36)months, with 1- and 3-year overall survival rate of 86.5% and 40.9%, respectively. The 285 patients in the training set were followed up postoperatively for 30(range, 6-36)months, with 1- and 3-year overall survival rate of 85.1% and 35.5%, respectively. The 121 patients in the validation set were followed up postoperatively for 25(range, 6-36)months, with 1- and 3-year overall survival rate of 87.0% and 43.2%, respectively. There was no significant difference in postoperative overall survival rate between the training set and the validation set ( χ2=3.20, P>0.05). Results of multivariate analysis showed that left thoracic surgical approach, preopera-tive neutrophil count, vascular invasion, perineural invasion, pathological T2-4 stage, pathological N2-3 stage, and postoperative pneumonia were independent risk factors affecting postoperative survival of 285 patients in the training set ( hazard ratio=1.466, 1.037, 1.482, 1.549, 5.268, 7.727, 22.202, 2.539, 2.686, 1.425, 95% confidence interval as 1.026-2.096, 1.003-1.073, 1.008-2.179, 1.105-2.170, 1.201-23.099, 1.833-32.576, 4.734-104.128, 1.577-4.087, 1.631-4.422, 1.018-1.994, P<0.05). (2) Construction and validation of machine learning prediction models. Independent risk factors affecting postoperative survival were included to construct RSF, GBM, LASSO-Cox, CoxBoost, survivalsvm, XGBoost, SuperPC, and plsRcox machine learning prediction models. Results of Delong test showed that there were significant differences in the AUC of RSF and GBM from the other six models ( P<0.05). Results of time-dependent ROC curve showed that all 8 machine learning predic-tion models had good discriminative ability in the training cohort, among which the RSF machine learning prediction model had the best predictive performance. Results of calibration curve showed that the RSF machine learning prediction model fitted well for predicting postoperative 1-, 2-, and 3-year overall survival in the training cohort, with high consistency with actual results. Results of decision curve analysis showed that within a threshold range of 0-0.80, the RSF machine learning prediction model provided a better overall net benefit. Further analysis showed that in the validation set, the AUC of RSF machine learning prediction model for postoperative 1-, 2-, and 3-year survival prediction were 0.786 (95% confidence interval as 0.609-0.962), 0.774 (95% confidence interval as 0.676-0.873), and 0.750 (95% confidence interval as 0.652-0.848), respectively. Results of calibration curve showed that the RSF machine learning prediction model fitted well for predicting postopera-tive 1-, 2-, and 3-year overall survival in the validation set, with high consistency with actual results. In the training set, the optimal cutoff value of the RSF machine learning prediction model risk score was 11.7. Patients with risk score ≥11.7 were classified as the high-risk group, and those with risk score <11.7 as the low-risk group. The median survival times of the two groups were 18.0 months and >36.0 months, respectively, showing a significant difference between them ( χ2=73.30, P<0.05). In the validation set, the optimal cutoff value of the RSF machine learning prediction model risk score was 11.7. Patients with risk score ≥11.7 were classified as the high-risk group, and those with risk score<11.7 as the low-risk group. The median survival times of the two groups were 17.0 months and>36.0 months for the high-risk and low-risk groups, respectively, showing a significant difference between them ( χ2=35.20, P<0.05). Conclusions:Left thoracic surgical approach, preoperative neutrophil count, vascular invasion, perineural invasion, pathological T2-4 stage, pathological N2-3 stage, and postoperative pneumonia are independent risk factors affecting survival of esophageal cancer patients after radical resection. The RSF machine learning prediction model constructed based on these factors can effectively distinguish the survival prognosis of high-risk and low-risk patients.

Background: Nasopharyngeal carcinoma (NPC) is characterized by high programmed death-ligand 1 (PD-L1) expression and abundant infiltration of non-malignant lymphocytes, which renders patients potentially suitable candidates for immune checkpoint blockade therapies. Palate, lung, and nasal epithelium clone (PLUNC) inhibit the growth of NPC cells and enhance cellular apoptosis and differentiation. Currently, the relationship between PLUNC (as a tumor-suppressor) and PD-L1 in NPC is unclear. Methods: We collected clinical samples of NPC to verify the relationship between PLUNC and PD-L1. PLUNC plasmid was transfected into NPC cells, and the variation of PD-L1 was verified by western blot and immunofluorescence. In NPC cells, we verified the relationship of PD-L1, activating transcription factor 3 (ATF3), and β-catenin by western blot and immunofluorescence. Later, we further verified that PLUNC regulates PD-L1 through β-catenin. Finally, the effect of PLUNC on β-catenin was verified by co-immunoprecipitation (Co-IP). Results: We found that PLUNC expression was lower in NPC tissues than in paracancer tissues. PD-L1 expression was opposite to that of PLUNC. Western blot and immunofluorescence showed that β-catenin could upregulate ATF3 and PD-L1, while PLUNC could downregulate ATF3/PD-L1 by inhibiting the expression of β-catenin. PLUNC inhibits the entry of β-catenin into the nucleus. Co-IP experiments demonstrated that PLUNC inhibited the interaction of DEAD-box helicase 17 (DDX17) and β-catenin. Conclusions PLUNC downregulates the expression of PD-L1 by inhibiting the interaction of DDX17/β-catenin in NPC.

Background: Nasopharyngeal carcinoma (NPC) is characterized by high programmed death-ligand 1 (PD-L1) expression and abundant infiltration of non-malignant lymphocytes, which renders patients potentially suitable candidates for immune checkpoint blockade therapies. Palate, lung, and nasal epithelium clone (PLUNC) inhibit the growth of NPC cells and enhance cellular apoptosis and differentiation. Currently, the relationship between PLUNC (as a tumor-suppressor) and PD-L1 in NPC is unclear. Methods: We collected clinical samples of NPC to verify the relationship between PLUNC and PD-L1. PLUNC plasmid was transfected into NPC cells, and the variation of PD-L1 was verified by western blot and immunofluorescence. In NPC cells, we verified the relationship of PD-L1, activating transcription factor 3 (ATF3), and β-catenin by western blot and immunofluorescence. Later, we further verified that PLUNC regulates PD-L1 through β-catenin. Finally, the effect of PLUNC on β-catenin was verified by co-immunoprecipitation (Co-IP). Results: We found that PLUNC expression was lower in NPC tissues than in paracancer tissues. PD-L1 expression was opposite to that of PLUNC. Western blot and immunofluorescence showed that β-catenin could upregulate ATF3 and PD-L1, while PLUNC could downregulate ATF3/PD-L1 by inhibiting the expression of β-catenin. PLUNC inhibits the entry of β-catenin into the nucleus. Co-IP experiments demonstrated that PLUNC inhibited the interaction of DEAD-box helicase 17 (DDX17) and β-catenin. Conclusions PLUNC downregulates the expression of PD-L1 by inhibiting the interaction of DDX17/β-catenin in NPC.

Objective To analyze the effectiveness of reasonable control measures for prophylactic use of antibiotics in urinary system lithotripsy. Methods By antimicrobial stewardship, strengthening special comments on antibiotics and information notification on rational use of antibiotics, adding and improving the pre-review rules for antibiotics prescriptions, conducting in-depth clinical training and consultation by clinical pharmacists, strengthening innovation in rational use of drugs, and taking various measures to actively improve rational use of prophylactic antibiotics of lithotripsy in urology department, the changes of indexes related to antibiotics in urology department from 2019 to 2022 were analyzed. Results After active and reasonable control, Antibiotics Use Density in urology department decreased year by year. The utilization rate of antibiotics in inpatients decreased from 94.27% in 2019 to 77.47% in 2022. Various rate of microbial inspection reached the standard in 2022. The imipenem and cilastatin sodium for injection ranking of prophylactic use of antibiotics consumption DDDs for urinary system lithotripsy decreased from the 4th place in 2019 to the 8th place in 2022. The ranking of the urology department on carbapenem consumption DDDs in the whole hospital decreased from the 8th place in 2019 to the 12th place in 2022. At the same time, the incidence of urinary tract lithotripsy postoperative infection showed a decreasing trend year by year, from 0.84% in 2019 to 0.49% in 2022. Conclusion Positive control measures can promote the rational use of prophylactic antibiotics for urinary system lithotripsy.

Background: Nasopharyngeal carcinoma (NPC) is characterized by high programmed death-ligand 1 (PD-L1) expression and abundant infiltration of non-malignant lymphocytes, which renders patients potentially suitable candidates for immune checkpoint blockade therapies. Palate, lung, and nasal epithelium clone (PLUNC) inhibit the growth of NPC cells and enhance cellular apoptosis and differentiation. Currently, the relationship between PLUNC (as a tumor-suppressor) and PD-L1 in NPC is unclear. Methods: We collected clinical samples of NPC to verify the relationship between PLUNC and PD-L1. PLUNC plasmid was transfected into NPC cells, and the variation of PD-L1 was verified by western blot and immunofluorescence. In NPC cells, we verified the relationship of PD-L1, activating transcription factor 3 (ATF3), and β-catenin by western blot and immunofluorescence. Later, we further verified that PLUNC regulates PD-L1 through β-catenin. Finally, the effect of PLUNC on β-catenin was verified by co-immunoprecipitation (Co-IP). Results: We found that PLUNC expression was lower in NPC tissues than in paracancer tissues. PD-L1 expression was opposite to that of PLUNC. Western blot and immunofluorescence showed that β-catenin could upregulate ATF3 and PD-L1, while PLUNC could downregulate ATF3/PD-L1 by inhibiting the expression of β-catenin. PLUNC inhibits the entry of β-catenin into the nucleus. Co-IP experiments demonstrated that PLUNC inhibited the interaction of DEAD-box helicase 17 (DDX17) and β-catenin. Conclusions PLUNC downregulates the expression of PD-L1 by inhibiting the interaction of DDX17/β-catenin in NPC.

Objective:To investigate the prognostic factors for esophageal cancer after radical resection and the application value of machine learning prediction model.Methods:The retrospective cohort study was conducted. The clinicopatholigical data of 406 esophageal cancer patients who were admitted to Qilu Hospital of Shandong University from January 2018 to March 2022 were collected. There were 357 males and 49 females, aged (64±8)years. All patients underwent radical resection of esophageal cancer. The 406 patients were randomly divided into a training set of 285 cases and a validation set of 121 cases at a 7∶3 ratio based on a random number table. The training set was used to construct prediction model, and the validation set was used to validate prediction model. Patients were divided into high-risk group and low-risk group based on risk scores. Observation indicators: (1) follow-up of patients and analysis of influencing factors for prognosis; (2) construction and validation of machine learning prediction models. Comparison of measurement data with normal distribution between groups was conducted using the independent sample t test. Comparison of measurement data with skewed distribution between groups was conducted using the Mann-Whitney U test. Comparison of count data between groups was conducted using the chi-square test. Comparison of ordinal data between groups was conducted using the rank sum test. The Kaplan-Meier method was used to calculate survival rate and plot survival curve, and the Log-rank test was used for survival analysis. The Cox proportional hazard regression model was used for univariate and multivariate analyses. Independent influencing factors were included, and data processing, machine learning model construction, and visualization were performed using R packages including random survival forest (RSF), gradient boosting machine (GBM), least absolute shrinkage and selection operator Cox regression (LASSO-Cox), Cox proportional hazards model boosting (CoxBoost), survival support vector machine (survivalsvm), extreme gradient boosting (XGBoost), supervised principal component analysis (SuperPC), and Cox partial least squares regression (plsRcox). Receiver operating characteristic (ROC) curves were drawn, and sensitivity, specificity, and area under the curve (AUC) were calculated. The Delong test was used to assess the differences in AUC among different models in the training set, and the time-dependent ROC was used to compare the predictive performance of different models. Calibration curves were used to evaluate model accuracy, and decision curve analysis (DCA) was used to evaluate overall net benefit. Results:(1) Follow-up of patients and analysis of influencing factors for prognosis. All 406 patients were followed up postoperatively for 28(range, 6-36)months, with 1- and 3-year overall survival rate of 86.5% and 40.9%, respectively. The 285 patients in the training set were followed up postoperatively for 30(range, 6-36)months, with 1- and 3-year overall survival rate of 85.1% and 35.5%, respectively. The 121 patients in the validation set were followed up postoperatively for 25(range, 6-36)months, with 1- and 3-year overall survival rate of 87.0% and 43.2%, respectively. There was no significant difference in postoperative overall survival rate between the training set and the validation set ( χ2=3.20, P>0.05). Results of multivariate analysis showed that left thoracic surgical approach, preopera-tive neutrophil count, vascular invasion, perineural invasion, pathological T2-4 stage, pathological N2-3 stage, and postoperative pneumonia were independent risk factors affecting postoperative survival of 285 patients in the training set ( hazard ratio=1.466, 1.037, 1.482, 1.549, 5.268, 7.727, 22.202, 2.539, 2.686, 1.425, 95% confidence interval as 1.026-2.096, 1.003-1.073, 1.008-2.179, 1.105-2.170, 1.201-23.099, 1.833-32.576, 4.734-104.128, 1.577-4.087, 1.631-4.422, 1.018-1.994, P<0.05). (2) Construction and validation of machine learning prediction models. Independent risk factors affecting postoperative survival were included to construct RSF, GBM, LASSO-Cox, CoxBoost, survivalsvm, XGBoost, SuperPC, and plsRcox machine learning prediction models. Results of Delong test showed that there were significant differences in the AUC of RSF and GBM from the other six models ( P<0.05). Results of time-dependent ROC curve showed that all 8 machine learning predic-tion models had good discriminative ability in the training cohort, among which the RSF machine learning prediction model had the best predictive performance. Results of calibration curve showed that the RSF machine learning prediction model fitted well for predicting postoperative 1-, 2-, and 3-year overall survival in the training cohort, with high consistency with actual results. Results of decision curve analysis showed that within a threshold range of 0-0.80, the RSF machine learning prediction model provided a better overall net benefit. Further analysis showed that in the validation set, the AUC of RSF machine learning prediction model for postoperative 1-, 2-, and 3-year survival prediction were 0.786 (95% confidence interval as 0.609-0.962), 0.774 (95% confidence interval as 0.676-0.873), and 0.750 (95% confidence interval as 0.652-0.848), respectively. Results of calibration curve showed that the RSF machine learning prediction model fitted well for predicting postopera-tive 1-, 2-, and 3-year overall survival in the validation set, with high consistency with actual results. In the training set, the optimal cutoff value of the RSF machine learning prediction model risk score was 11.7. Patients with risk score ≥11.7 were classified as the high-risk group, and those with risk score <11.7 as the low-risk group. The median survival times of the two groups were 18.0 months and >36.0 months, respectively, showing a significant difference between them ( χ2=73.30, P<0.05). In the validation set, the optimal cutoff value of the RSF machine learning prediction model risk score was 11.7. Patients with risk score ≥11.7 were classified as the high-risk group, and those with risk score<11.7 as the low-risk group. The median survival times of the two groups were 17.0 months and>36.0 months for the high-risk and low-risk groups, respectively, showing a significant difference between them ( χ2=35.20, P<0.05). Conclusions:Left thoracic surgical approach, preoperative neutrophil count, vascular invasion, perineural invasion, pathological T2-4 stage, pathological N2-3 stage, and postoperative pneumonia are independent risk factors affecting survival of esophageal cancer patients after radical resection. The RSF machine learning prediction model constructed based on these factors can effectively distinguish the survival prognosis of high-risk and low-risk patients.

Objective:To investigate the efficacy of arthroscopic superior capsular reconstruction using composite autologous patch graft combined with tenodesis of the long head of the biceps tendon in the treatment of irreparable massive rotator cuff tears (IMRCT).Methods:A retrospective case series study was performed on 11 IMRCT patients who were admitted to Affiliated Fuyang Hospital of Bengbu Medical University (Fuyang People′s Hospital) from May 2020 to June 2022, including 7 males and 4 females, aged 54-74 years [(62.6±7.3)years]. All the patients were treated with arthroscopic superior capsular reconstruction using composite patch graft combined with tenodesis of the long head of the biceps tendon. The Visual Analogue Scale (VAS), Acromiohumeral Distance (AHD), Constant-Murley score and University of California Los Angeles (UCLA) score and active range of motion of the shoulder joint before, at 6 months after surgery and at the last follow-up were compared. At the last follow-up, the integrity of reconstructed superior capsule and the long head of the biceps tendon was evaluated using MRI of the shoulder joint. Postoperative complications were observed.Results:All the patients were followed up for 13-39 months [16(13, 36)months]. The VAS score, AHD, Constant-Murley score, and UCLA score were 2(2, 3)points, (9.1±1.1)mm, (56.1±5.4)points, and (19.7±2.8)points respectively at 6 months after surgery, which were all significantly improved from those before surgery [6(5, 7)points, (5.1±1.2)mm, (37.9±2.2)points, and (11.8±1.2)points] ( P<0.05). The VAS score, AHD, Constant-Murley score, and UCLA score were 0(0, 1)points, (8.4±0.9)mm, (83.6±3.8)points, and (28.2±2.3)points respectively at the last follow-up, which were all significantly improved from those before surgery ( P<0.05). At the last follow-up, the VAS score or AHD were not significantly improved from those at 6 months after surgery ( P>0.05); Constant-Murley score and UCLA score were both significantly improved from those at 6 months after surgery ( P<0.05). At 6 months after surgery, shoulder active ranges of motion in forward flexion, abduction and external rotation were (134.6±13.5)°, (124.6±18.6)° and 45(40, 50)° respectively, which were all significantly improved compared with those before surgery [(63.2±36.1)°, (65.0±23.1)°, and [30(20, 40)°] ( P<0.05). At the last follow-up, shoulder active ranges of motion in forward flexion, abduction and external rotation were (144.1±12.6)°, (139.6±15.4)° and 60(45, 65)° respectively, which were all significantly improved compared with those before surgery ( P<0.05). There were no significant differences in active range of motion of the shoulder in forward flexion, abduction and external rotation between 6 months after surgery and the last follow-up ( P>0.05). At the last follow-up, MRI revealed integrity of the reconstructed superior joint capsule and the long head of the biceps tendon in 10 patients. One patient developed resorption of the greater tuberosity and 1 showed a partial tear of the supraspinatus tendon at 1 year after surgery. Conclusion:Arthroscopic superior capsular reconstruction using composite autologous patch graft combined with tenodesis of the long head of the biceps tendon can relieve shoulder pain, decrease upward displacement of the humerus head, improve the function and range of motion of the shoulder joint, and reduce complications in the treatment of IMRCT.

Traumatic supraorbital fissure syndrome (TSOFS) is a symptom complex caused by nerve entrapment in the supraorbital fissure after skull base trauma. If the compressed cranial nerve in the supraorbital fissure is not decompressed surgically, ptosis, diplopia and eye movement disorder may exist for a long time and seriously affect the patients′ quality of life. Since its overall incidence is not high, it is not familiarized with the majority of neurosurgeons and some TSOFS may be complicated with skull base vascular injury. If the supraorbital fissure surgery is performed without treatment of vascular injury, it may cause massive hemorrhage, and disability and even life-threatening in severe cases. At present, there is no consensus or guideline on the diagnosis and treatment of TSOFS that can be referred to both domestically and internationally. To improve the understanding of TSOFS among clinical physicians and establish standardized diagnosis and treatment plans, the Skull Base Trauma Group of the Neurorepair Professional Committee of the Chinese Medical Doctor Association, Neurotrauma Group of the Neurosurgery Branch of the Chinese Medical Association, Neurotrauma Group of the Traumatology Branch of the Chinese Medical Association, and Editorial Committee of Chinese Journal of Trauma organized relevant experts to formulate Chinese expert consensus on the diagnosis and treatment of traumatic supraorbital fissure syndrome ( version 2024) based on evidence of evidence-based medicine and clinical experience of diagnosis and treatment. This consensus puts forward 12 recommendations on the diagnosis, classification, treatment, efficacy evaluation and follow-up of TSOFS, aiming to provide references for neurosurgeons from hospitals of all levels to standardize the diagnosis and treatment of TSOFS.

Objective To study the genotoxicities of raceanisodamine hydrochloride injection. Methods Bacterial reverse mutation test, in vitro Chromosomal aberration test and in vivo Micronucleus test were performed to investigate the genotoxicities of raceanisodamine hydrochloride injection. Results The Ames test showed that raceanisodamine hydrochloride injection did not increase mutagenicity for TA1535, TA102, TA100, TA98 and TA97 strains at the dosage of 0.5, 5, 50, 500, 5000 μg per plate under two parallel system conditions (±S9). Results of CA test indicated that there was no statistical difference between raceanisodamine hydrochloride injection groups (doses of 58.75,117.5 and 235.0 μg/ml) and the solvent control group under two parallel system conditions (±S9). In MNT test, with doses of 7.5, 15.0 and 30.0 mg/kg respectively, the micronucleus induction rate of bone marrow of ICR mice was not statistically significant (P>0.05) when compared with that of vehicle control group in all dose groups. Conclusion Under the conditions of these study, the results indicated that raceanisodamine hydrochloride injection had no mutagenicity to Salmonella typhimurium, had no aberration effect on the chromosome of mammalian cultured cells, and had no effect on inducing micronucleus of bone marrow polychromatic erythrocytes in ICR mouse. All test results showed that raceanisodamine hydrochloride injection had no potential carcinogenicities and genetic toxicities under the test conditions.

Objective:To investigate the diagnostic accuracy of serological indicators and evaluate the diagnostic value of a new established combined serological model on identifying the minimal hepatic encephalopathy (MHE) in patients with compensated cirrhosis.Methods:This prospective multicenter study enrolled 263 compensated cirrhotic patients from 23 hospitals in 15 provinces, autonomous regions and municipalities of China between October 2021 and August 2022. Clinical data and laboratory test results were collected, and the model for end-stage liver disease (MELD) score was calculated. Ammonia level was corrected to the upper limit of normal (AMM-ULN) by the baseline blood ammonia measurements/upper limit of the normal reference value. MHE was diagnosed by combined abnormal number connection test-A and abnormal digit symbol test as suggested by Guidelines on the management of hepatic encephalopathy in cirrhosis. The patients were randomly divided (7∶3) into training set ( n=185) and validation set ( n=78) based on caret package of R language. Logistic regression was used to establish a combined model of MHE diagnosis. The diagnostic performance was evaluated by the area under the curve (AUC) of receiver operating characteristic curve, Hosmer-Lemeshow test and calibration curve. The internal verification was carried out by the Bootstrap method ( n=200). AUC comparisons were achieved using the Delong test. Results:In the training set, prevalence of MHE was 37.8% (70/185). There were statistically significant differences in AMM-ULN, albumin, platelet, alkaline phosphatase, international normalized ratio, MELD score and education between non-MHE group and MHE group (all P<0.05). Multivariate Logistic regression analysis showed that AMM-ULN [odds ratio ( OR)=1.78, 95% confidence interval ( CI) 1.05-3.14, P=0.038] and MELD score ( OR=1.11, 95% CI 1.04-1.20, P=0.002) were independent risk factors for MHE, and the AUC for predicting MHE were 0.663, 0.625, respectively. Compared with the use of blood AMM-ULN and MELD score alone, the AUC of the combined model of AMM-ULN, MELD score and education exhibited better predictive performance in determining the presence of MHE was 0.755, the specificity and sensitivity was 85.2% and 55.7%, respectively. Hosmer-Lemeshow test and calibration curve showed that the model had good calibration ( P=0.733). The AUC for internal validation of the combined model for diagnosing MHE was 0.752. In the validation set, the AUC of the combined model for diagnosing MHE was 0.794, and Hosmer-Lemeshow test showed good calibration ( P=0.841). Conclusion:Use of the combined model including AMM-ULN, MELD score and education could improve the predictive efficiency of MHE among patients with compensated cirrhosis.