OBJECTIVES: To explore the measures to improve the follow-up rate of preterm infants after discharge, and to evaluate the effectiveness of these measures using quality improvement methodology. METHODS: The follow-up status of preterm infants discharged from March to May 2017 was used as the baseline before quality improvement, and a specific quality improvement goal for the follow-up rate was proposed. The Pareto chart was used to analyze the causes of follow-up failure, and a key driver diagram was constructed based on the links involved in improving follow-up rate. The causes of failure were analyzed to determine the key links and intervention measures for quality improvement, and the follow-up rate was monitored weekly using a control chart until the quality improvement goal was achieved. RESULTS: The follow-up rate of preterm infants after discharge was 57.92% (117/202) at baseline before quality improvement, and the quality improvement goal was set to increase the follow-up rate of preterm infants from baseline to more than 80% within 12 months. The Pareto chart analysis showed that the main causes of follow-up failure were deficiencies in follow-up file management and irregular follow-up times (33.70%, 31/92), insufficient follow-up education and poor communication (25.00%, 23/92), and the inability to meet the diverse needs of parents (18.48%, 17/92). Based on the key links for quality improvement and the main causes of follow-up failure, the following intervention measures were adopted: (1) strengthen follow-up publicity and education; (2) build a follow-up team; and (3) establish a follow-up platform and system. The control chart indicated that with the implementation of the above intervention measures, the weekly follow-up rate increased to 74.09% (306/413) in July 2017 and 83.09% (511/615) in December 2017, finally achieving the quality improvement goal. During the COVID-19 pandemic, the follow-up rate of preterm infants fluctuated between 23.54% (460/1 954) and 70.97% (1 931/2 721), and subsequently, it returned to pre-pandemic levels starting in February 2023. CONCLUSIONS The application of quality improvement methodology can help to formulate intervention measures based on the main causes of follow-up failure, thereby improving the follow-up rate of preterm infants after discharge. This quality improvement method is feasible and practical and thus holds promise for clinical application.
Humans ; Quality Improvement ; Infant, Premature ; Infant, Newborn ; Patient Discharge ; Follow-Up Studies ; Female ; Male

Objective To explore the relevant risk factors of H.pylori infection,and provide reference for prevention and treatment of H.pylori in this area,and further provide theoretical basis for the prevention and treatment of gastric cancer.Methods A total of 1200 residents in four districts of Haikou city were investigated by questionnaire and urea 14 C breath test by holistic stratified random sampling to calculate the population infection rate and analyze the risk factors of infection.Results The total infection rate was 32.5％,which was lower than the national H.pylori infection rate.No consumption of fruits and vegetables,no habit of washing hands before meals,and people with gastrointestinal symptoms,are independent risk factors of H.pylori infection.No consumption of pickled products is of great significance to prevent H.pylori infection.Conclusion The prevalence of H.pylori infection in the population of Haikou is lower than the national average,and H.pylori infection is closely related to the poor living habits of residents.

Objective: To evaluate the prognostic value of left ventricular ejection fraction (LVEF) reserve assessed by gated SPECT myocardial perfusion imaging (SPECT G-MPI) for major adverse cardiovascular event (MACE) in patients with coronary artery disease. Methods: This is a retrospective cohort study. From January 2017 to December 2019, patients with coronary artery disease and confirmed myocardial ischemia by stress and rest SPECT G-MPI, and underwent coronary angiography within 3 months were enrolled. The sum stress score (SSS) and sum resting score (SRS) were analyzed by the standard 17-segment model, and the sum difference score (SDS, SDS=SSS-SRS) was calculated. The LVEF at stress and rest were analyzed by 4DM software. The LVEF reserve (ΔLVEF) was calculated (ΔLVEF=stress LVEF-rest LVEF). The primary endpoint was MACE, which was obtained by reviewing the medical record system or by telephone follow-up once every twelve months. Patients were divided into MACE-free and MACE groups. Spearman correlation analysis was used to analyze the correlation between ΔLVEF and all MPI parameters. Cox regression analysis was used to analyze the independent factors of MACE, and the optimal SDS cutoff value for predicting MACE was determined by receiver operating characteristic curve (ROC). Kaplan-Meier survival curves were plotted to compare the difference in the incidence of MACE between different SDS groups and different ΔLVEF groups. Results: A total of 164 patients with coronary artery disease [120 male; age (58.6±10.7) years] were included. The average follow-up time was (26.5±10.4) months, and a total of 30 MACE were recorded during follow-up. Multivariate Cox regression analysis showed that SDS (HR=1.069, 95%CI: 1.005-1.137, P=0.035) and ΔLVEF (HR=0.935, 95%CI: 0.878-0.995, P=0.034) were independent predictors of MACE. According to ROC curve analysis, the optimal cut-off to predict MACE was a SDS of 5.5 with an area under the curve of 0.63 (P=0.022). Survival analysis showed that the incidence of MACE was significantly higher in the SDS≥5.5 group than in the SDS<5.5 group (27.6% vs. 13.2%, P=0.019), but the incidence of MACE was significantly lower in the ΔLVEF≥0 group than in theΔLVEF<0 group (11.0% vs. 25.6%, P=0.022). Conclusions: LVEF reserve (ΔLVEF) assessed by SPECT G-MPI serves as an independent protective factor for MACE, while SDS is an independent risk predictor in patients with coronary artery disease. SPECT G-MPI is valuable for risk stratification by assessing myocardial ischemia and LVEF.
Humans ; Male ; Middle Aged ; Aged ; Coronary Artery Disease/diagnostic imaging* ; Stroke Volume ; Myocardial Perfusion Imaging ; Retrospective Studies ; Ventricular Function, Left ; Myocardial Ischemia

Objective: To establish a comprehensive diagnostic classification model of lateral cephalograms based on artificial intelligence (AI) to provide reference for orthodontic diagnosis. Methods: A total of 2 894 lateral cephalograms were collected in Department of Orthodontics, Capital Medical University School of Stomatology from January 2015 to December 2021 to construct a data set, including 1 351 males and 1 543 females with a mean age of (26.4± 7.4) years. Firstly, 2 orthodontists (with 5 and 8 years of orthodontic experience, respectively) performed manual annotation and calculated measurement for primary classification, and then 2 senior orthodontists (with more than 20 years of orthodontic experience) verified the 8 diagnostic classifications including skeletal and dental indices. The data were randomly divided into training, validation, and test sets in the ratio of 7∶2∶1. The open source DenseNet121 was used to construct the model. The performance of the model was evaluated by classification accuracy, precision rate, sensitivity, specificity and area under the curve (AUC). Visualization of model regions of interest through class activation heatmaps. Results: The automatic classification model of lateral cephalograms was successfully established. It took 0.012 s on average to make 8 diagnoses on a lateral cephalogram. The accuracy of 5 classifications was 80%-90%, including sagittal and vertical skeletal facial pattern, mandibular growth, inclination of upper incisors, and protrusion of lower incisors. The acuracy rate of 3 classifications was 70%-80%, including maxillary growth, inclination of lower incisors and protrusion of upper incisors. The average AUC of each classification was ≥0.90. The class activation heat map of successfully classified lateral cephalograms showed that the AI model activation regions were distributed in the relevant structural regions. Conclusions: In this study, an automatic classification model for lateral cephalograms was established based on the DenseNet121 to achieve rapid classification of eight commonly used clinical diagnostic items.
Male ; Female ; Humans ; Young Adult ; Adult ; Artificial Intelligence ; Deep Learning ; Cephalometry ; Maxilla ; Mandible/diagnostic imaging*

Objective: To develop a multi-classification orthodontic image recognition system using the SqueezeNet deep learning model for automatic classification of orthodontic image data. Methods: A total of 35 000 clinical orthodontic images were collected in the Department of Orthodontics, Capital Medical University School of Stomatology, from October to November 2020 and June to July 2021. The images were from 490 orthodontic patients with a male-to-female ratio of 49∶51 and the age range of 4 to 45 years. After data cleaning based on inclusion and exclusion criteria, the final image dataset included 17 453 face images (frontal, smiling, 90° right, 90° left, 45° right, and 45° left), 8 026 intraoral images [frontal occlusion, right occlusion, left occlusion, upper occlusal view (original and flipped), lower occlusal view (original and flipped) and coverage of occlusal relationship], 4 115 X-ray images [lateral skull X-ray from the left side, lateral skull X-ray from the right side, frontal skull X-ray, cone-beam CT (CBCT), and wrist bone X-ray] and 684 other non-orthodontic images. A labeling team composed of orthodontic doctoral students, associate professors, and professors used image labeling tools to classify the orthodontic images into 20 categories, including 6 face image categories, 8 intraoral image categories, 5 X-ray image categories, and other images. The data for each label were randomly divided into training, validation, and testing sets in an 8∶1∶1 ratio using the random function in the Python programming language. The improved SqueezeNet deep learning model was used for training, and 13 000 natural images from the ImageNet open-source dataset were used as additional non-orthodontic images for algorithm optimization of anomaly data processing. A multi-classification orthodontic image recognition system based on deep learning models was constructed. The accuracy of the orthodontic image classification was evaluated using precision, recall, F1 score, and confusion matrix based on the prediction results of the test set. The reliability of the model's image classification judgment logic was verified using the gradient-weighted class activation mapping (Grad-CAM) method to generate heat maps. Results: After data cleaning and labeling, a total of 30 278 orthodontic images were included in the dataset. The test set classification results showed that the precision, recall, and F1 scores of most classification labels were 100%, with only 5 misclassified images out of 3 047, resulting in a system accuracy of 99.84%(3 042/3 047). The precision of anomaly data processing was 100% (10 500/10 500). The heat map showed that the judgment basis of the SqueezeNet deep learning model in the image classification process was basically consistent with that of humans. Conclusions: This study developed a multi-classification orthodontic image recognition system for automatic classification of 20 types of orthodontic images based on the improved SqueezeNet deep learning model. The system exhibitted good accuracy in orthodontic image classification.
Humans ; Male ; Female ; Child, Preschool ; Child ; Adolescent ; Young Adult ; Adult ; Middle Aged ; Deep Learning ; Reproducibility of Results ; Radiography ; Algorithms ; Cone-Beam Computed Tomography

OBJECTIVES: To investigate the incidence of extrauterine growth retardation (EUGR) and its risk factors in very preterm infants (VPIs) during hospitalization in China. METHODS: A prospective multicenter study was performed on the medical data of 2 514 VPIs who were hospitalized in the department of neonatology in 28 hospitals from 7 areas of China between September 2019 and December 2020. According to the presence or absence of EUGR based on the evaluation of body weight at the corrected gestational age of 36 weeks or at discharge, the VPIs were classified to two groups: EUGR group (n=1 189) and non-EUGR (n=1 325). The clinical features were compared between the two groups, and the incidence of EUGR and risk factors for EUGR were examined. RESULTS: The incidence of EUGR was 47.30% (1 189/2 514) evaluated by weight. The multivariate logistic regression analysis showed that higher weight growth velocity after regaining birth weight and higher cumulative calorie intake during the first week of hospitalization were protective factors against EUGR (P<0.05), while small-for-gestational-age birth, prolonged time to the initiation of total enteral feeding, prolonged cumulative fasting time, lower breast milk intake before starting human milk fortifiers, prolonged time to the initiation of full fortified feeding, and moderate-to-severe bronchopulmonary dysplasia were risk factors for EUGR (P<0.05). CONCLUSIONS It is crucial to reduce the incidence of EUGR by achieving total enteral feeding as early as possible, strengthening breastfeeding, increasing calorie intake in the first week after birth, improving the velocity of weight gain, and preventing moderate-severe bronchopulmonary dysplasia in VPIs.
Female ; Fetal Growth Retardation ; Gestational Age ; Hospitalization ; Humans ; Incidence ; Infant ; Infant, Newborn ; Infant, Premature ; Infant, Very Low Birth Weight ; Prospective Studies ; Risk Factors

Objective: To investigate the values of serum 8-hydroxydeoxyguanosine (8-OHdG) in predicting disease progression and prognosis of patients with sepsis. Methods: The prospective observational research methods were used. A total of 124 patients with sepsis who met the inclusion criteria were admitted to the Department of Emergency of the First Affiliated Hospital of Wenzhou Medical University from April 2015 to July 2016, including 79 males and 45 females, aged (62±15) years. The sepsis-related organ failure assessment (SOFA) scores of all patients on admission and on the second day of admission and their difference (ΔSOFA) were calculated. The patients were divided into non-progression group with ΔSOFA score <2 (n=101) and progression group with ΔSOFA score ≥2 (n=23), and according to the survival during hospitalization, the patients were divided into survival group (n=85) and death group (n=39). Data of patients between non-progression group and progression group, survival group and death group were compared, including the gender, age, days in emergency intensive care unit (ICU), smoking, hypertension, diabetes mellitus, serum white blood cell count, serum C-reactive protein, and serum procalcitonin on admission, and serum 8-OHdG within 24 h of admission. The multivariate logistic regression analysis was used to screen the independent risk factors of disease progression and death during hospitalization in 124 patients with sepsis, the receiver's operating characteristic (ROC) curves were drawn according to the independent risk factors, and the area under the curve (AUC), the best threshold, and the sensitivity and specificity under the best threshold were calculated. The patients were divided into high 8-OHdG group (n=35) and low 8-OHdG group (n=89) according to the best threshold in ROC curve of death during hospitalization. The data including the gender, age, SOFA score on admission, SOFA score on the second day of admission, and ΔSOFA score of patients in the two groups were compared. The survival rates of patients within 90 d of admission in the two groups were compared by the Kaplan-Meier method. Data were statistically analyzed with independent sample t test, Mann-Whitney U test, chi-square test, and Log-rank test. Results: The gender, age, days in emergency ICU, smoking, complicated with hypertension, complicated with diabetes mellitus, serum white blood cell count, serum C-reactive protein, and serum procalcitonin on admission of patients in non-progression group and progression group were similar (P>0.05). The serum 8-OHdG within 24 h of admission of patients in progression group was significantly higher than that in non-progression group (Z=-2.31, P<0.05). Multivariate logistic regression analysis showed that the serum 8-OHdG within 24 h of admission was the independent risk factor for disease progression of 124 patients with sepsis (odds ratio=1.06, with 95% confidence interval of 1.01-1.11, P<0.05). The AUC under the ROC curve of serum 8-OHdG within 24 h of admission to predict disease progression of 124 patients with sepsis was 0.65 (with 95% confidence interval of 0.52-0.79, P<0.05), the optimal threshold was 32.88 ng/mL, and the sensitivity and specificity under the optimal threshold was 52.2% and 79.2%, respectively. The gender, age, days in emergency ICU, smoking, complicated with hypertension, complicated with diabetes mellitus, and serum white blood cell count, serum C-reactive protein, and serum procalcitonin on admission of patients in survival group and death group were similar (P>0.05). The serum 8-OHdG within 24 h of admission of patients in death group was significantly higher than that in survival group (Z=-2.37, P<0.05). Multivariate logistic regression analysis showed that the serum 8-OHdG within 24 h of admission was the independent risk factor for death of 124 patients with sepsis (odd ratio=1.04, with 95% confidence interval of 1.00-1.09, P<0.05). The AUC under the ROC curve of serum 8-OHdG within 24 h of admission to predict death of patients during hospitalization was 0.63 (with 95% confidence interval of 0.52-0.75, P<0.05), the optimal threshold was 32.43 ng/mL, the sensitivity and specificity under the optimal threshold was 51.3% and 84.7%, respectively. The gender and age of patients in high 8-OHdG group and low 8-OHdG group were similar (P>0.05). The SOFA score on admission, SOFA score on the second day of admission, and ΔSOFA score of patients in high 8-OHdG group were significantly higher than those in low 8-OHdG group (with Z values of -2.49, -3.01, and -2.64, respectively, P<0.05 or P<0.01). The survival rate within 90 d of admission of patients in low 8-OHdG group was significantly higher than that in high 8-OHdG group (χ²=14.57, P<0.01). Conclusions: Serum 8-OHdG level is an independent risk factor for disease progression and death in sepsis patients with limited ability for predicting disease progression and prognosis of sepsis of patients. The patients with higher serum 8-OHdG level have higher death risk within 90 d of admission.
8-Hydroxy-2'-Deoxyguanosine ; Aged ; Disease Progression ; Female ; Humans ; Male ; Middle Aged ; Prognosis ; ROC Curve ; Retrospective Studies ; Sepsis

Objective::To study the protective effect of astragalus polysaccharide (APS) on micronucleus and sister chromatid exchange (SCE) in human bone marrow mesenchyml stem cell (BMSCs) exposed to formaldehyde, in order to initially explore the potential mechanism. Method::BMSCs were cultured in vitro, cells were randomly divided into five groups: control group, formaldehyde group, and APS 40, 100, 400 mg·L^-1 groups. BMSCs were infected with 120 μmol·L^-1 formaldehyde, meanwhile, APS 40, 100, 400 mg·L^-1 groups were co-cultured with 40, 100, 400 mg·L^-1 APS. Cell morphology was observed by inverted phase contrast microscope, micronucleus were detected by micronucleus test, SCE was detected by SCE test, and mRNA and protein expressions of proliferating cell nuclear antigen (PCNA), xeroderma pigmentosum B, D, F, G (XPB, XPD, XPF, XPG) were detected by quantitative real-time fluorescence polymerase chain reaction(Real-time PCR)and Western blot. Result::Compared with control group, cell counts decreased, and cell morphology of BMSCs in formaldehyde group significantly changed, they were all recovered gradually in 40, 100, 400 mg·L^-1 APS groups. Compared with control group, the micronucleus and SCE increased significantly (P<0.01), PCNA mRNA and protein expressions down-regulated significantly (P<0.05), while XPB, XPD, XPF, XPG mRNA and protein expressions up-regulated significantly (P<0.05, P<0.01). Compared with formaldehyde group, BMSCs were treated with APS at 40, 100, 400 mg·L^-1, micronucleus and SCE decreased significantly (P<0.01), and mRNA and protein expressions of PCNA, XPB, XPD, XPF and XPG up-regulated significantly (P<0.05, P<0.01). Among them, the 100 mg·L^-1 APS group had the most obvious effect. Conclusion::APS can protect formaldehyde-induced BMSCs micronucleus and SCE, especially 100 mg·L^-1 APS has the most obvious effect. The mechanism may be associated with the up-regulation of expressions of PCNA, XPB, XPD, XPF and XPG in the nucleotide exicision repair pathway (NER), which promoted the damage repair.