1.Application of time series and machine learning models in predicting the trend of sickness absenteeism among primary and secondary school students in Shanghai
WANG Zhengzhong, ZHANG Zhe, ZHOU Xinyi, YUAN Linlin, ZHAI Yani, SUN Lijing, LUO Chunyan
Chinese Journal of School Health 2025;46(3):426-430
Objective:
To analyze the temporal variation patterns of sickness absenteeism among primary and secondary school students in Shanghai, so as to explore models suitable for predicting peaks and intensity of absenteeism rates.
Methods:
The seasonal and trend decomposition using loess (STL) method was used to analyze the seasonal and long term trend changes in sickness absenteeism among primary and secondary school students from September 1 in 2010 to June 30 in 2018, in Shanghai. A hierarchical clustering method based on Dynamic Time Warping (DTW) was employed to classify absenteeism symptoms with similar temporal patterns. Based on historical data, the study constructed and evaluated different time series algorithms and machine learning models to optimize the accuracy of predicting the trend of sickness absenteeism.
Results:
During the research period, the average new absenteeism rate due to illness was 16.86 per 10 000 person day for every academic year, and the trend of sickness absenteeism exhibited both seasonality and a long term upward trend, reaching its highest point in the 2017 academic year (22.47 per 10 000 person day). The symptoms of absenteeism were divided into three categories: high incidence in winter and spring (respiratory symptoms, fever and general discomfort, etc.), high incidence in summer (eye symptoms, nosebleeds, etc.) and those without obvious seasonality (skin symptoms, accidental injuries, etc.).The constructed time series models effectively predicted the trend of absenteeism due to illness, although the accuracy of predicting peak intensity was relatively low. Among them, the multi layer perceptron (MLP) model performed the best, with an root mean squared error (RMSE) of 8.96 and an mean absolute error (MAE) of 4.37, reducing 36.51% and 39.02% compared to the baseline model.
Conclusion
Time series models and machine learning algorithms could effectively predict the trend of sickness absenteeism, and corresponding prevention and control measures can be taken for absenteeism caused by different symptoms during peak periods.
2.Comparison of small-sample multi-class machine learning models for plasma concentration prediction of valproic acid
Xi CHEN ; Shen’ao YUAN ; Hailing YUAN ; Jie ZHAO ; Peng CHEN ; Chunyan TIAN ; Yi SU ; Yunsong ZHANG ; Yu ZHANG
China Pharmacy 2025;36(11):1399-1404
OBJECTIVE To construct three-class (insufficient, normal, excessive) and two-class (insufficient, normal) models for predicting plasma concentration of valproic acid (VPA), and compare the performance of these two models, with the aim of providing a reference for formulating clinical medication strategies. METHODS The clinical data of 480 patients who received VPA treatment and underwent blood concentration test at the Xi’an International Medical Center Hospital were collected from November 2022 to September 2024 (a total of 695 sets of data). In this study, predictive models were constructed for target variables of three-class and two-class models. Feature ranking and selection were carried out using XGBoost scores. Twelve different machine learning algorithms were used for training and validation, and the performance of the models was evaluated using three indexes: accuracy, F1 score, and the area under the working characteristic curve of the subject (AUC). RESULTS XGBoost feature importance scores revealed that in the three-class model, the importance ranking of kidney disease and electrolyte disorders was higher. However, in the two-class model, the importance ranking of these features significantly decreased, suggesting a close association with the excessive blood concentration of VPA. In the three-class model, Random Forest method performed best, with F1 score of 0.704 0 and AUC of 0.519 3 on the test set; while in the two-class model, CatBoost method performed optimally, with F1 score of 0.785 7 and AUC of 0.819 5 on the test set. CONCLUSIONS The constructed three-class model has the ability to predict excessive VPA blood concentration, but its prediction and model generalization abilities are poor; the constructed two-class model can only perform classification prediction for insufficient and normal blood concentration cases, but its model performance is stronger.
3.Drug resistance in 38 cases of carbapenem-resistant Serratia marcescens infection in a grade Ⅲ level A general hospital in Shanghai
Yuan LI ; Liang TIAN ; Chunyan LI ; Yun LIU ; Wei JI
Shanghai Journal of Preventive Medicine 2025;37(6):484-489
ObjectiveTo understand the infection characteristics and drug resistance of carbapenem-resistant Serratia marcescens (CRSM) in a general hospital in Shanghai, and to provide a theoretical basis for clinical anti-infective treatment and prevention of drug-resistant bacteria. MethodsClinical data on cases with CRSM infections detected in clinical specimens at a gradeⅢ level A general hospital in Shanghai from June 2022 to June 2024 were retrospectively collected, and their clinical distributions, factors of hospital-acquired infections, prognosis, and drug-resistant situation were analyzed simultaneously. ResultsA total of 38 cases with CRSM were detected from June 2022 to June 2024, and the number of CRSM strains accounted for 25.00% (38/152) of the number of SM strains. The 38 CRSM infection samples were all derived from sputum. CRSM were distributed in 9 clinical departments, and the top 3 departments having the highest percentages of CRSM among SM strains, were intensive care unit (ICU) (78.79%, 26/33), gastrointestinal surgery department (57.14%, 4/7), and thyroid hernia surgery department (50.00%, 1/2). Among the 38 patients with CRSM infections, 8 cases were identified as hospital-acquired infection, resulting in a hospital-acquired infection rate of 21.05. The mortality rate of the 38 cases of CRSM infected patients within 30 days after detection of CRSM was 23.68% (9/38). The results of multivariate logistic regression analysis showed that sequential organ failure assessment (SOFA) score
4.Effect of enriched environment combined with acupuncture at head points on behavior in rats with autism spectrum disorder
Zichen MU ; Qiang TANG ; Yunqiu SHI ; Yan WANG ; Shuwei ZHU ; Ya'nan ZHUANG ; Danshuang XU ; Hongyu LI ; Baolong LI ; Chunyan ZHANG ; Mengke YUAN
Chinese Journal of Rehabilitation Theory and Practice 2024;30(2):176-182
ObjectiveTo observe the effect of enriched environment (EE) combined with acupuncture at head point (HA) on behavior in rats with autism spectrum disorder. MethodsHealthy female Wistar rats were given peritoneal injection of sodium valproate at 12.5 days of gestation. Twenty-four male offspring rats were randomly selected and then randomly divided into model group (n = 6), EE group (n = 6), HA group (n = 6) and EE combined with HA group (the combined group, n = 6). Six male offspring rats born from female mice injected with the same amount of saline intraperitoneally were as control group. After four weeks of treatment, all the five groups were tested with three-chamber test and marble burying test, and the sociability index, the social novelty index and the number of buried marbles were recorded. The levels of interleukin (IL)-1β and IL-6 in peripheral blood were determined by enzyme-linked immunosorbent assay (ELISA). ResultsAfter treatment, compared with the model group, the sociability index and the social novelty index improved (P < 0.05), the number of buried marbles reduced (P < 0.05), and the levels of IL-6 and IL-1β in peripheral blood decreased in EE group, HA group and the combined group (P < 0.05); while the combined group was the best (P < 0.01). ConclusionBoth EE or acupuncture at HA could improve behavioral symptoms, and reduce the expression of inflammatory factors in rats with autism spectrum disorder. The combination of the two methods showed the best result.
5.The role of CB2 in accelerating orthodontic tooth movement
Dengying FAN ; Haoyan ZHAI ; Huijuan LIU ; Yuan ZHAO ; Dongna LI ; Xing QIAO ; Wenjing KANG ; Dechao ZHU ; Chunyan LIU
Acta Universitatis Medicinalis Anhui 2024;59(2):212-218
Objective To explore the effect of cannabinoid receptor 2(CB2)on orthodontic tooth movement(OTM)rate and periodontal tissue reconstruction of pressure area in mice.Methods Thirty CB2-/-male mice and thirty littermate control WT male mice were individually accepted the orthodontic appliance at their age of 6 weeks.The mice were respectively scarified at 3 days,7 days,14 days and 21 days after the operation.Then the tooth movement distance was examined through the stereomicroscope.Hematoxylin-eosin staining was performed to explore the biological responses of periodontium at the distal mesial root pressure area.Anti-tartrate acid phospha-tase staining was performed to calculate the number and distribution of osteoclasts at the distal mesial root pressure area,and MMP-9 was evaluated by immunohistochemistry to examine the number of MMP-9(+)monocytes and multinucleated cells in the same district as the TRAP staining.Results Compared with those WT mice at 3,7,14 and 21 days,OTM distance showed a gradual increased tendency according with experimental time over 21 days.The widths of periodontal ligament on the pressure side were markedly greater in CB2-/-mice than WT mice at 7,14 and 21 days(P<0.000 1).The numbers of TRAP positive osteoclasts were significantly greater in CB2-/-mice than those in WT mice at 14 days of OTM(P<0.001).MMP-9 immunohistochemical staining showed that the number of MMP-9(+)monocytes and multinucleated cells was more in CB2-/-mice than that in WT mice at 14 days of OTM(P<0.05).Conclusion The absence of CB2 accelerates orthodontic tooth movement under or-thodontic force.The absence of CB2 reinforces bone resorption in orthodontic tooth movement compressive area dur-ing orthodontic tooth movement.
6.The role of reactive oxygen species in periodontitis and periodontal tissue regeneration
Haoyan ZHAI ; Yuan ZHAO ; Dengying FAN ; Chunyan LIU
Chinese Journal of Tissue Engineering Research 2024;28(14):2254-2260
BACKGROUND:Reactive oxygen species is a double-edged sword in the development of periodontitis and the regeneration of periodontal tissue.Low concentration of reactive oxygen species induces the differentiation of periodontal fibroblasts,and excessive reactive oxygen species will cause damage to periodontal tissue.In the process of inflammation,the accumulation of reactive oxygen species in periodontal tissue induces damage to cells and tissues through a variety of signaling pathways or through redox reactions. OBJECTIVE:To review the double-edged sword effect of reactive oxygen species in periodontitis and periodontal tissue regeneration,thereby providing potential targets and treatment ideas for the clinical treatment of periodontitis and periodontal tissue regeneration. METHODS:Databases of CNKI and PubMed were searched for relevant articles published from April 1990 to April 2023 with the key words of"periodontal tissue engineering,periodontal defect,regeneration of periodontal tissue,chronic periodontitis,reactive oxygen species,oxidative stress,antioxidative stress,oxidative injuries,free radicals,reactive nitrogen species"in Chinese and English,respectively.By reading the titles and abstracts,repetitive studies or irrelevant literatures were excluded.Finally,77 articles were included for review. RESULTS AND CONCLUSION:Reactive oxygen species are a kind of free radicals with high reactivity.When bacteria invade,reactive oxygen species are released in large quantities by the respiratory explosion of neutrophils and play a double-edged sword role in the body through their redox reactions or as pleiotropic physiological signal transmitters.In periodontitis,low concentrations of reactive oxygen species can kill invading pathogenic bacteria,but high concentrations of reactive oxygen species promote the secretion of inflammatory factors through JNK,RANK,Wnt/β-Catenin and other pathways,promote immune damage or directly damage tissues through oxidative reactions or through other ways to aggravate periodontitis.In the process of periodontal tissue regeneration,low concentrations of reactive oxygen species can promote the proliferation and differentiation of periodontal ligament stem cells through Nrf2 and other pathways and can promote the secretion of vascular endothelial growth factor to promote vascular regeneration.This provides seeds and a nutrient environment for periodontal tissue regeneration,which is extremely important for promoting periodontal tissue regeneration.However,high concentrations of reactive oxygen species will reduce the activity of periodontal ligament stem cells and damage endothelial cells,which are not conducive to vascular regeneration.This will inhibit wound healing and periodontal tissue regeneration.Therefore,it is important to explore the role of reactive oxygen species in the development of periodontitis and periodontal tissue regeneration and to discover the potential mechanism of its action and to explore the appropriate concentration for its role in reducing periodontal inflammation and promoting periodontal tissue regeneration for the future treatment of periodontitis and periodontal tissue regeneration in clinical practice.Using reactive oxygen species as a target to explore ways to reduce periodontal inflammation and promote periodontal ligament stem cell activity and vascular regeneration may become a clinically effective method for treating periodontitis and promoting periodontal tissue regeneration.
7.Investigation on pharmacognosic identification of two Yi Medicine in Yunnan province
LIN Chunyan ; FU Xingqing ; REN Jie ; DONG Yuan ; ZHANG Wenjie
Drug Standards of China 2024;25(1):010-017
Objective: To understand their appearance and microscopic characteristics, as well as their differences by studying the pharmacognosy of Yi medicine Elsholtzia rugulosa and Elsholtzia bodinieri, in order to provide a basis for identification and improvement of quality standards.
Methods: Stereo microscopy and optical microscopy and the macroscopic and microscopic identification methods were adopted to compare identification and digital representation for Elsholtzia rugulosa and Elsholtzia bodinieri from overall character, local characteristics, the microscopic identification characteristics, the transverse section and the powder.
Results:There were significant differences in the the macroscopic and the microscopic identification characteristics of Elsholtzia rugulosa and Elsholtzia bodinieri.
Conclusion: This study summarized the exclusive and practical features in pharmacognosic identification of Elsholtzia rugulosa and Elsholtzia bodinieri, it provides a useful reference for supervision the clinical medication,inspection,and standard drafting.
8.Hemolytic diseases of the fetus and newborn caused by anti-Dib: a case report and related research
Zhijian LIAO ; Shuangshuang JIA ; Jizhi WEN ; Chunyan MO ; Yuan SHAO ; Runqing ZHANG ; Guangping LUO ; Yanli JI
Chinese Journal of Blood Transfusion 2024;37(2):158-164
【Objective】 To identify the specificity of alloantibody against high-frequency antigens in one case suffering with severe hemolytic diseases of the fetus and newborn (HDFN) and to screen for matching blood for transfusion. 【Methods】 The HDFN test and the antibody serological identification tests in the mother were performed. Several common high frequency antigens of maternal red blood cells (RBCs) were determined. IgG subtype coated on the RBCs of the newborn was determined. The phagocytic efficiency of the antibody was tested using the monocyte phagocytosis of sensitized erythrocyte by flow cytometry in vitro. Sanger sequencing of DI gene was performed in the mother, father and mother’s brother. The diluted maternal plasma was used for large scale screening of matching blood using IAT in Coomb’s gel card. 【Results】 Di(b-) phenotype was identified in the mother of the newborn and anti-Dib (titer: 512) related HDN was detected in the newborn. IgG1 and IgG2 subtypes of anti-Dib were detected and the rate of monocyte phagocytosis was 88.83%(74.7/84.09). The compatible blood was not detected in the maternal relatives. Subsequently, the newborn received the matching RBCs of two Di(b-) donors identified from 5 520 blood donors and discharged from the hospital. We screened out 17 Di(b-) donors out of 51 334 blood donors, indicating that the distribution frequency of Di(b-) among blood donors in Guangzhou was about 0.033% (17/51 334). 【Conclusion】 By serology and molecular biology methods, the newborn was identified with HDFN caused by anti-Dib, and an effective large-scale screening method for Di (b -) rare blood types was established to find matching blood, which supported the establishment of rare Di(b-) blood database.
9.Application of monocyte monolayer assay on hemolytic disease of fetus and newborn caused by IgG anti-M
Chunyan MO ; Shuangshuang JIA ; Siying ZHU ; Yanli JI ; Yuan SHAO ; Zhijian LIAO ; Guangping LUO ; Ling WEI
Chinese Journal of Blood Transfusion 2024;37(6):643-647
Objective To elucidate the prediction ability of monocyte monolayer assay(MMA)used in hemolytic dis-ease of fetus and newborn(HDFN)caused by IgG anti-M.Methods Plasma from eight pregnant women containing IgG an-ti-M were collected,and were divided into two groups(4 cases with HDFN,with severe clinical symptoms such as fetal hy-drops,and 4 cases without HDFN)according to the clinical outcomes.M antigen positive cells were sensitized with dithioth-reitol(DTT)treated plasma from eight pregnant women respectively.MMA was performed by coincubation with monocytes and sensitized M cells,along with negative and positive control set up.T-test was conducted to compare the difference in phagocytic efficiency between two groups.Results The phagocytic efficiency in group with HDFN were 15.37%,13.05%,9.17%and 24.50%respectively,with the mean value of 15.52%,while the group without HDFN were 8.74%,11.07%,5.12%and 6.23%respectively,with the mean value of 7.79%.There was no significant difference in phagocytic efficiency between two groups(P>0.05).The mean values of both groups were not significantly different from the negative control(P>0.05),but both were significantly lower than positive control(P<0.05).Conclusion The low phagocytic efficiency couldn't convince that the MMA is an effective predictor for the HDFN caused by IgG anti-M,indicating that another mech-anism might be responsible for it rather than monocyte phagocytosis.The assessment of the peak systolic velocity in middle cerebral artery of the fetal should be considered in the management for pregnant women who produce IgG anti-M to estimate the situation of fetal anemia.
10.Carrier screening for 223 monogenic diseases in Chinese population:a multi-center study in 33 104 individuals
Wei HOU ; Xiaolin FU ; Xiaoxiao XIE ; Chunyan ZHANG ; Jiaxin BIAN ; Xiao MAO ; Juan WEN ; Chunyu LUO ; Hua JIN ; Qian ZHU ; Qingwei QI ; Yeqing QIAN ; Jing YUAN ; Yanyan ZHAO ; Ailan YIN ; Shutie LI ; Yulin JIANG ; Manli ZHANG ; Rui XIAO ; Yanping LU
Journal of Southern Medical University 2024;44(6):1015-1023
Objective To investigate the epidemiological characteristics and mutation spectrum of monogenic diseases in Chinese population through a large-scale,multicenter carrier screening.Methods This study was conducted among a total of 33 104 participants(16 610 females)from 12 clinical centers across China.Carrier status for 223 genes was analyzed using high-throughput sequencing and different PCR methods.Results The overall combined carrier frequency was 55.58%for 197 autosomal genes and 1.84%for 26 X-linked genes in these participants.Among the 16 669 families,874 at-risk couples(5.24%)were identified.Specifically,584 couples(3.50%)were at risk for autosomal genes,306(1.84%)for X-linked genes,and 16 for both autosomal and X-linked genes.The most frequently detected autosomal at-risk genes included GJB2(autosomal recessive deafness type 1A,393 couples),HBA1/HBA2(α-thalassemia,36 couples),PAH(phenylketonuria,14 couples),and SMN1(spinal muscular atrophy,14 couples).The most frequently detected X-linked at-risk genes were G6PD(G6PD deficiency,236 couples),DMD(Duchenne muscular dystrophy,23 couples),and FMR1(fragile X syndrome,17 couples).After excluding GJB2 c.109G>A,the detection rate of at-risk couples was 3.91%(651/16 669),which was lowered to 1.72%(287/16 669)after further excluding G6PD.The theoretical incidence rate of severe monogenic birth defects was approximately 4.35‰(72.5/16 669).Screening for a battery of the top 22 most frequent genes in the at-risk couples could detect over 95%of at-risk couples,while screening for the top 54 genes further increased the detection rate to over 99%.Conclusion This study reveals the carrier frequencies of 223 monogenic genetic disorders in the Chinese population and provides evidence for carrier screening strategy development and panel design tailored to the Chinese population.In carrier testing,genetic counseling for specific genes or gene variants can be challenging,and the couples need to be informed of these difficulties before testing and provided with options for not screening these genes or gene variants.


Result Analysis
Print
Save
E-mail