OBJECTIVE: To identify the optimal machine learning algorithm for predicting post-stroke epilepsy (PSE) within one year following acute stroke, establish a nomogram model based on this algorithm, and perform external validation to achieve accurate prediction of secondary epilepsy. METHODS: A total of 870 acute stroke patients admitted to the emergency department of Xiang'an Hospital of Xiamen University from June 2019 to June 2023 were enrolled for model development (model group). An external validation cohort of 435 acute stroke patients admitted to the Fifth Hospital of Xiamen during the same period was used to validate the machine learning algorithms and nomogram model. Patients were classified into control and epilepsy groups based on the development of PSE within one year. Clinical and laboratory data, including baseline characteristics, stroke location, vascular status, complications, hematologic parameters, and National Institutes of Health Stroke Scale (NIHSS) score, were collected for analysis. Nine machine learning algorithms such as logistic regression, CN2 rule induction, K-nearest neighbors, adaptive boosting, random forest, gradient boosting, support vector machine, naive Bayes, and neural network were applied to evaluate predictive performance. The area under the curve (AUC) of receiver operator characteristic curve (ROC curve) was used to identify the optimal algorithm. Logistic regression was used to screen risk factors for PSE, and the top 10 predictors were selected to construct the nomogram model. The predictive performance of the model was evaluated using the ROC curve in both the model and validation groups. RESULTS: Among the 870 patients in the model group, 29 developed PSE within one year. Among the nine algorithms tested, logistic regression demonstrated the best performance and generalizability, with an AUC of 0.923. Univariate logistic regression identified several risk factors for PSE, including platelet count, white blood cell count, red blood cell count, glycated hemoglobin (HbA1c), C-reactive protein (CRP), triglycerides, high-density lipoprotein (HDL), aspartate aminotransferase (AST), alanine aminotransferase (ALT), activated partial thromboplastin time (APTT), thrombin time, D-dimer, fibrinogen, creatine kinase (CK), creatine kinase-MB (CK-MB), lactate dehydrogenase (LDH), serum sodium, lactic acid, anion gap, NIHSS score, brain herniation, periventricular stroke, and carotid artery plaque. Further multivariate logistic regression analysis showed that white blood cell count, HDL, fibrinogen, lactic acid and brain herniation were independent risk factors [odds ratio (OR) were 1.837, 198.039, 47.025, 11.559, 70.722, respectively, all P < 0.05]. In the external validation group, univariate logistic regression analysis showed that platelet count, white blood cell count, CRP, triacylglycerol, APTT, D-dimer, fibrinogen, CK, CK-MB, LDH, NIHSS score, and cerebral herniation were risk factors for PSE one year after acute stroke. Further multiple logistic regression analysis showed that APTT and cerebral herniation were independent predictors (OR were 0.587 and 116.193, respectively, both P < 0.05). The nomogram model, constructed using 10 key variables-brain herniation, periventricular stroke, carotid artery plaque, white blood cell count, triglycerides, thrombin time, D-dimer, serum sodium, lactic acid, and NIHSS score-achieved an AUC of 0.908 in the model group and 0.864 in the external validation group. CONCLUSIONS The logistic regression-based prediction model for epilepsy one year after acute stroke, developed using machine learning algorithms, showed optimal predictive performance. The nomogram model based on the logistic regression-derived predictors showed strong discriminative power and was successfully validated externally, suggesting favorable clinical applicability and generalizability.
Humans ; Machine Learning ; Stroke/complications* ; Nomograms ; Epilepsy/etiology* ; Algorithms ; Male ; Female ; Logistic Models ; Middle Aged ; Aged ; Risk Factors ; Bayes Theorem

Objective: To systematically evaluate the influencing factors for medication compliance in patients with comorbidities of chronic diseases, so as to provide the evidence for improving medication compliance. Methods: Literature on influencing factors for medication compliance in patients with comorbidities of chronic diseases were retrived from CNKI, Wanfang Data, VIP, SinoMed, PubMed, Web of Science, Cochrane Library and Embase from inception to January 20, 2024. After independent literature screening, data extraction, and quality assessment by two researchers, a meta-analysis was performed using RevMan 5.4 and Stata 16.0 softwares. Literature were excluded one by one for sensitivity analysis. Publication bias was assessed using Egger's test. Results: Initially, 7 365 relevant articles were retrieved, and 35 of them were finally included, with a total sample size of about 150 000 individuals. There were 30 cross-sectional studies and 5 cohort studies; and 11 high-quality studies and 24 medium-quality studies. The meta-analysis showed that the demographic factors of lower level of education (OR=2.148, 95%CI: 1.711-2.696), lower economic income (OR=1.897, 95%CI: 1.589-2.264), male (OR=0.877, 95%CI: 0.782-0.985), living alone (OR=2.833, 95%CI: 1.756-4.569) and unmarried (OR=2.784, 95%CI: 1.251-6.196); the medication treatment factors of polypharmacy (OR=1.794, 95%CI: 1.190-2.706), potentially inappropriate medication (OR=2.988, 95%CI: 1.527-5.847), low frequency of daily medication (OR=0.533, 95%CI: 0.376-0.754) and adverse drug reactions (OR=3.319, 95%CI: 1.967-5.602); the disease factors of long course of disease (OR=2.118, 95%CI: 1.643-2.730), more comorbidities (OR=1.667, 95%CI: 1.143-2.431) and cognitive impairment (OR=2.007, 95%CI: 1.401-2.874); and the psychosocial factors of poor belief in taking medication (OR=1.251, 95%CI: 1.011-1.547), poor self-rated health (OR=1.990, 95%CI: 1.571-2.522) and being guided by healthcare professionals (OR=0.151, 95%CI: 0.062-0.368) were the influencing factors for medication compliance in patients with chronic comorbidities. Conclusion The medication compliance in patients with comorbidities of chronic diseases is associated with demographic factors, pharmacological factors, disease factors and psychosocial factors, mainly including living alone, adverse drug reactions, course of disease, number of comorbidities and medication beliefs.

Objective To investigate the dietary patterns of rural residents in the high-incidence areas of esophageal cancer (EC), and to explore the clustering and influencing factors of risk factors associated with high-incidence characteristics. Methods A special structured questionnaire was applied to conduct a face-to-face survey on the dietary patterns of rural residents in Yanting county of Sichuan Province from July to August 2021. Univariate and multivariate logistic regression models were used to analyze the influencing factors of risk factor clustering for EC. Results There were 838 valid questionnaires in this study. A total of 90.8% of rural residents used clean water such as tap water. In the past one year, the people who ate fruits and vegetables, soybean products, onions and garlic in high frequency accounted for 69.5%, 32.8% and 74.5%, respectively; the people who ate kimchi, pickled vegetables, sauerkraut, barbecue, hot food and mildew food in low frequency accounted for 59.2%, 79.6%, 68.2%, 90.3%, 80.9% and 90.3%, respectively. The clustering of risk factors for EC was found in 73.3% of residents, and the aggregation of two risk factors was the most common mode (28.2%), among which tumor history and preserved food was the main clustering pattern (4.6%). The logistic regression model revealed that the gender, age, marital status and occupation were independent influencing factors for the risk factors clustering of EC (P<0.05). Conclusion A majority of rural residents in high-incidence areas of EC in Yanting county have good eating habits, but the clustering of some risk factors is still at a high level. Gender, age, marital status, and occupation are influencing factors of the risk factors clustering of EC.

It is believed that retention due to deficient qi is an important pathogenesis of endometriosis （EMs）. Deficient qi is the root of the disease， mainly manifested as spleen deficiency， while retention is the branch pathogenesis of the disease， mainly with blood stasis， complicated with constraint， phlegm， heat， toxin and other pathological factors. Therefore， it is proposed to follow the treatment principle of supplementing deficiency and unblocking stagnation， and take the methods of replenishing qi and fortifying the spleen， removing stasis and eliminating concretions. Self-made Fuzheng Huayu Formula （扶正化瘀方） is taken as the basic formula， and can be modified with the symptoms in menstrual and non-menstrual periods. Additionally， the methods of moving qi， dispelling phlegm， clearing heat， relieving toxin and others can be combined， and it is recommended to treat the root and the branch simultaneously.

Objective:To summarize the best evidence of medication adherence interventions for patients with multiple chronic conditions.Methods:According to the "6S" evidence model, literature on medication adherence in patients with multiple chronic conditions was retrieved from BMJ Best Clinical Practice, UpToDate, Medlive, National Institute for Health and Clinical Excellence, Cochrane Library, Embase, PubMed, Web of Science, China Biology Medicine disc, China National Knowledge Infrastructure, WanFang data and so on. The search period was from establishing the database to August 30, 2023.Results:A total of 16 articles were included, including three guidelines, four expert consensus, seven systematic reviews, and two meta-analyses. Twenty-seven pieces of evidence were summarized from six aspects of compliance assessment, educational intervention, behavioral intervention, optimized treatment program, technical reminder intervention, and social-psychological-emotional intervention.Conclusions:The best evidence of medication adherence interventions for patients with multiple chronic conditions summarized provides a reference for medical and nursing staff to develop medication adherence interventions.

Objective To investigate the value of pulse oxygen perfusion index(PI)and blood lactic acid(BLA)concentration in early prediction of retinopathy of prematurity(ROP).Methods A retrospective case-control study was conducted on 128 preterm infants who met the inclusion criteria and were admitted to the neonatal intensive care unit of our hospital from September 2018 to December 2022.Among them,46 patients with ROP were in the ROP group,and 82 pa-tients without ROP were in the non-ROP group.Basic data of these preterm infants were recorded after admission.PI val-ues were continuously monitored with the Masimo Radical-7(USA)SpO2 blood oxygen saturation detector,and BLA con-centrations were detected with the ABL90FLEX blood gas analyzer.The receiver operating characteristic(ROC)curve and area under curve(AUC)were used to evaluate the value of PI and BLA concentration in early prediction of ROP.Results There were no significant differences in gestational age,birth weight,sex,and delivery mode between the two groups(all P＞0.05).The PI values after birth were significantly different between the two groups(Fgroup=15.393,Pgroup＜0.001).The PI values of preterm infants in the ROP group decreased significantly at 1 h,12 h and 24 h after birth and slightly at 48 h to 96 h after birth compared with the non-ROP group.The PI values of preterm infants in the two groups sta-bilized at 96 h after birth.The PI values of preterm infants in the ROP group were lower than those in the non-ROP group at all time points within 96 h after birth(all P＜0.05).The PI values showed interaction effects between the two groups at different time points(Finteraction=5.061,Pinteraction＜0.001).There was a significant difference in BLA concentration between the two groups after birth(Fgroup=91.158,Pgroup＜0.001).In the ROP group,the BLA concentration increased significantly at 1 h after birth and slightly at 12 h and 24 h after birth compared with the non-ROP group.The BLA concentration in the ROP group was higher than that in the non-ROP group at all time points after birth(all P＜0.05).The BLA concentration showed no interaction effects between the two groups at different time points(Finteraction=0.567,Pinteraction＞0.05).The AUC of PI values at 1 h,12 h and 24 h after birth and BLA concentration at 1 h after birth for predicting ROP was 0.77,0.82,0.83,and 0.82,respectively.The AUC of combined PI values at 1 h,12 h and 24 h after birth and BLA concentration at 1 h after birth for predicting ROP was 0.94,higher than the predictive value of a single indicator.Conclusion PI and BLA concentration have good clinical value for early prediction of ROP.

The 2019 guideline update for the Diagnosis and Management of atrial fibrillation was jointly published by the American College of Cardiology(ACC), American Heart Association(AHA), American College of Chest Physicians(ACCP), and Heart Rhythm Society(HRS)on November 30, 2023.This article aims to interpret key updates in the new guideline, including recommendations for different stages of atrial fibrillation, modification and prevention of atrial fibrillation risk factors, flexibility in using clinical risk scores beyond CHA2DS2-VASc for predicting stroke and systemic embolism, consideration of stroke risk modifiers, early rhythm control, and the use of catheter ablation as a first-line therapy in selected patients.The guideline also provides recommendations for catheter ablation of atrial fibrillation in patients with heart failure with reduced ejection fraction, updates for device-detected atrial fibrillation, higher level Class of Recommendation for left atrial appendage occlusion devices, and guidance for managing atrial fibrillation in patients identified during medical illness or surgery.

Objective:To study the clinical characteristics，diagnosis，treatment and prognosis of neonatal appendicitis.Methods:From January 2019 to December 2022，19 neonates with appendicitis（appendicitis group）and 38 neonates with sepsis（sepsis group）admitted to the Neonatal Department of Xi'an Children's Hospital Affiliated to Xi'an Jiaotong University were studied.The characteristics of clinical manifestation，imaging，treatment and prognosis of neonates in two groups were analyzed,retrospectively.Results:Among 19 neonates with appendicitis，31.6% were premature，the mean birth weight was（2 927.9±796.2）g，male∶female=2.17∶1.Abdominal distention（8/19，42.1%）and fever（8/19，42.1%）were the first symptoms of appendicitis，and the first symptoms of sepsis were mainly fever（20/38，52.6%）and poor reaction（7/38，18.4%）.In the appendicitis group，the proportions of abdominal distension（89.5% vs. 5.3%），vomiting（36.8% vs. 2.6%），breast resistance（84.2% vs. 39.5%），mental reaction changing（94.7% vs. 71.1%）and abdominal positive signs（84.2% vs. 5.3%）were significantly higher than those in sepsis group（ P<0.05）.C-reactive protein（CRP）was elevated in 16 neonates with appendicitis and 13 neonates with sepsis，and elevated gradually in 14 neonates with appendicitis. Compared with sepsis group，CRP was higher in appendicitis group（ P<0.05）.Fifteen（78.9%）neonates with appendicitis were diagnosed only by ultrasound，mainly manifested as low echo area or liquid dark area in the right abdomen，thickening of the appendix wall or effusion in the cavity，and liquid exudation.Three（15.8%）neonates with appendicitis were diagnosed by ultrasound and CT.Eight（42.1%）neonates with appendicitis were complicated appendiceal perforation.Fifteen neonates with appendicitis were treated by conservative treatment，four cases were treated by operation，and all of them were cured and discharged. Conclusion:Abdominal ultrasonography should be improved as soon as possible in neonates with fever and septicemia，especially those with abdominal symptoms or signs，or CRP increased during treatment，and CT or surgical exploration if necessary,to confirm the diagnosis of neonatal appenditis and early treatment.

Neonatal onset multisystem inflammatory disease(NOMID), also known as chronic infantile neurological cutaneous and articular syndrome(CINCA), originates from perinatal period and mainly manifests urticaria, joint lesions, and central nervous system lesions.It is an autoinflammatory disease associated with mutations of NLRP3 located on chromosome 1q44.The early atypical clinical symptoms are prone to misdiagnosis.NOMID/CINCA should be differentiated from infectious diseases, familial cold autoinflammatory syndrome, Muckle-Wells syndrome, systemic juvenile idiopathic arthritis, mevalonate-kinase deficiency, tumor necrosis factor receptor-associated periodic syndrome, and other diseases.NOMID/CINCA is mainly diagnosed based on clinical symptoms, while genetic testing provides an essential supplementary for patients with atypical clinical manifestations.IL-1 targeted therapies including anakinra, rilonacept, and canakinumab, have been proven with sustained efficacy in treating NOMID/CINCA.This article reviews the progress on diagnosis and treatment of NOMID/CINCA.

Objective:To evaluate the efficacy of Janus kinase (JAK) inhibitors in the treatment of 5 children with severe alopecia areata, especially those with complicated nail damage.Methods:A total of 5 children with severe alopecia areata were enrolled and treated with oral JAK inhibitors (tofacitinib or baricitinib). The improvement of hair loss was assessed by using the severity of alopecia tool (SALT) at 12, 24, 36, and 48 weeks after the start of treatment. For 3 children with complicated nail damage, the improvement of diseased nails and toenails was evaluated by using the modified nail psoriasis severity index after treatment. During the treatment, adverse reactions were monitored.Results:The 5 children with severe alopecia areata were aged 2 - 11 years, with the disease duration ranging from 5 to 120 months, and the treatment with JAK inhibitors lasted 24 - 48 weeks. After 12-week treatment, 2 children achieved a 50% improvement in SALT (SALT50) ; after 24-week treatment, 3 achieved SALT95, and 1 achieved SALT75 and then withdrew baricitinib for personal reasons; after 36-week treatment, 3 achieved SALT99, and then received half-dose treatment; after 48-week treatment, 1, 1, 1 and 1 patient achieved SALT99, SALT83, SALT31, and SALT0, respectively, and 2 of them experienced gradually aggravated hair loss 1 - 2 months after the start of half-dose treatment. Among the 3 children with complicated nail damage, the improvement rates of nail severity index scores were 67.5%, 45.4%, and 25% respectively, and the improvement rates of toenail severity index scores were 42.5%, 71.4%, and 5% respectively after 12-week treatment; after 48-week treatment, the improvement rate of nail severity index scores were 100%, 100%, and 50% respectively, and the improvement rate of toenail severity index scores were 96.2%, 100%, 50% respectively. During the treatment, the uric acid level increased in 2 children, and one of them was accompanied by increased serum levels of low-density lipoprotein cholesterol and high-density lipoprotein cholesterol; 1 suffered from respiratory tract infections twice during the treatment, and was recovered after symptomatic treatment; there were no adverse reactions leading to drug withdrawal.Conclusion:JAK inhibitors can be used as a treatment option for severe alopecia areata in children.