Stomatology is a first-level discipline mainly focusing on maintaining and promoting oral health, as well as preventing and treating diseases of oral and maxillofacial system. Through the great efforts of generations of stomatologists, China's stomatologic causes has achieved remarkable results and rapid development. The number of stomatologists has reached 334 000, and the ratio of stomatologists to the population is 1:4 600, which have made China a major country in stomatology. However, compared with developed countries, there is still a considerable gap in the overall oral health level of our people. Strengthen the construction of stomatologist team, especially the training of stomatologists at the basis level; Optimize medical education in stomatology and improve the quality of graduates; Strengthen primary oral care and lay a solid foundation for oral care service; Regulate private dental institutions and strengthen their construction; Popularize knowledge about oral health, enhance public awareness of oral health, and improve public oral health behaviors; Give full play to the leading role of new ideas, knowledge and technologies in stomatology is a key link in developing stomatology and building a strong country in stomatology.
Oral Medicine/organization & administration* ; Humans ; China ; Oral Health

In recent years, China has continued to face a high prevalence of oral diseases, along with uneven access to high-quality dental care. Against this backdrop, artificial intelligence (AI), as a data-driven, algorithm-supported, and model-centered technology system, has rapidly expanded its role in transforming the landscape of stomatology. This review summarizes recent advances in the application of AI in stomatology across clinical care, biomedical and materials research, education, and hospital management. In clinical settings, AI has improved diagnostic accuracy, streamlined treatment planning, and enhanced surgical precision and efficiency. In research, machine learning has accelerated the identification of disease biomarkers, deepened insights into the oral microbiome, and supported the development of novel biomaterials. In education, AI has enabled the construction of knowledge graphs, facilitated personalized learning, and powered simulation-based training, driving innovation in teaching methodologies. Meanwhile, in hospital operations, intelligent agents based on large language models (LLMs) have been widely deployed for intelligent triage, structured pre-consultations, automated clinical documentation, and quality control, contributing to more standardized and efficient healthcare delivery. Building on these foundations, a multi-agent collaborative framework centered around an AI assistant for stomatology is gradually emerging, integrating task-specific agents for imaging, treatment planning, surgical navigation, follow-up prediction, patient communication, and administrative coordination. Through shared interfaces and unified knowledge systems, these agents support seamless human-AI collaboration across the full continuum of care. Despite these achievements, the broader deployment of AI still faces challenges including data privacy, model robustness, cross-institution generalization, and interpretability. Addressing these issues will require the development of federated learning frameworks, multi-center validation, causal reasoning approaches, and strong ethical governance. With these foundations in place, AI is poised to move from a supportive tool to a trusted partner in advancing accessible, efficient, and high-quality stomatology services in China.
Artificial Intelligence ; Humans ; Oral Medicine/trends* ; China ; Delivery of Health Care ; Machine Learning

The epidemiological research on mental health in China has undergone decades of development, transitioning from multi-regional surveys to nationally representative studies. In 1982, Academician Shen Yucun led a team to complete the first national survey in 12 regions, revealing a point prevalence rate of 10.54‰. In 1993, the point prevalence rate in the second national survey in 7 regions rose to 11.18‰. In 2002, the Composite International Diagnostic Interview (CIDI)-3.0 and Diagnostic and Statistical Manual of Mental Disorders, Fouth Edition (DSM-Ⅳ) standards were first applied in the surveys in urban Beijing and Shanghai to achieve international standards, but the representativeness of urban samples was limited. Subsequent regional studies contributed methodological insights toward a nationally representative survey. From 2013 to 2015, a research team led by Professor Huang Yueqin, in collaboration with 43 institutions, completed the China Mental Health Survey (CMHS), covering 32 552 community adults in 157 counties/districts in 31 provinces/autonomous. This study represents the first nationally representative epidemiological survey of mental disorders in China. The main results showed that the lifetime prevalence of mental disorders among adults in Chinese communities with depressive disorders was 16.6%, including 7.4% for mood disorders and 7.6% for anxiety disorders. Only 9.5% of patients with depressive disorders received treatment, and the full treatment rate was as low as 0.5%. Further surveys based on the CMHS framework in regions such as Ningxia, Urumqi, and Inner Mongolia confirmed the high risk of disease among rural women in western China and the widespread phenomenon of inadequate treatment. The results of CMHS methodology was transformed into the national Guidelines and Technical Standards for Epidemiological Investigation of Community Mental Disorders (2015 Edition) and software copyright, and the main data had been published in The Lancet Psychiatry. It has been used to guide the formulation of policies such as the Healthy China Action (2019-2030). CMHS establishing a complete process from scientific investigation to policy translation, filling the data gap at the national level, providing a replicable paradigm for the world, especially for developing countries, and marking a new stage of evidence-based decision-making in China's mental health epidemiological research.
China/epidemiology* ; Humans ; Mental Disorders/epidemiology* ; Health Surveys/history* ; Mental Health/statistics & numerical data* ; Prevalence ; Adult ; Female ; Male

OBJECTIVE: To evaluate the effect of palliative care on drug use, medical service utilization and medical expenditure of patients with advanced cancer. METHODS: A cohort of patients including pal-liative care and standard care was constructed using the medical records of the patients in Peking University Cancer Hospital from 2018 to 2020, and coarsened exact matching was used to match the two groups of patients. The average monthly opioid consumption, hospitalization rate, intensive care unit (ICU) rate and operation rate, and the average monthly total cost were selected to evaluate drug use, medical service utilization and medical expenditure. Chi-square test and Wilcoxon signed rank test were used to compare the differences between the two groups before and after exposure and the change in the palliative care group. The net impact of palliative care on the patients was calculated using the difference-in-differences analysis. RESULTS: In this study, 180 patients in the palliative care group and 3 101 patients in the stan-dard care group were finally included in the matching, and the matching effect of the two groups was good (L₁ < 0.1). Before and after exposure, the average monthly opioid consumption in the palliative care group was significantly higher than that in the standard care group (Before exposure: 0.3 DDD/person-month vs. 0.1 DDD/person-month, P < 0.01; After exposure: 0.7 DDD/person-month vs. 0.1 DDD/person-month, P < 0.01; DDD refers to defined daily dose), palliative care significantly increased the average monthly opioid consumption in the patients (0.3 DDD/person-month, P < 0.01). The hospitalization rate (48.9% vs. 74.3%, P < 0.01) and operation rate (3.9% vs. 8.8%, P < 0.01) of the patients in palliative care group were significantly lower than those in standard care group, and the ICU rate became similar between the two groups (1.1% vs. 1.6%, P=0.634). Palliative care significantly reduced the patients ' hospitalization rate (-25.6%, P < 0.01), ICU rate (-4.9%, P < 0.01) and operation rate (-14.5%, P < 0.01). Before and after exposure, the average monthly total costs of pal-liative care group were slightly higher than those of standard care group (Before exposure: 20 092.3 yuan vs. 19 132.8 yuan, P=0.725; After exposure: 9 719.8 yuan vs. 8 818.8 yuan, P=0.165). Palliative care increased the average monthly total cost by 2 208.8 yuan, but it was not statistically significant (P=0.316). CONCLUSION Palliative care can increase the opioid consumption in advanced cancer patients, reduce the rates of hospitalization, ICU and surgery, but has no significant effect on medical expenditure.
Humans ; Palliative Care/economics* ; Neoplasms/drug therapy* ; Analgesics, Opioid/economics* ; Male ; Female ; Middle Aged ; Aged ; Hospitalization/economics* ; Intensive Care Units/statistics & numerical data* ; Health Expenditures/statistics & numerical data* ; Adult ; Drug Utilization/statistics & numerical data* ; Patient Acceptance of Health Care/statistics & numerical data*

OBJECTIVES: Obtaining informed consent from research participants is an ethical and legal obligation for medical researchers in clinical studies. Electronic informed consent (eIC) is increasingly being adopted in clinical research worldwide. However, there is limited data on Chinese medical researchers' knowledge and attitudes toward eIC. This study aims to investigate their knowledge, attitudes, and influencing factors regarding eIC use in clinical research. METHODS: This cross-sectional study was conducted using stratified random sampling. From June to August 2022, medical researchers from 8 tertiary hospitals were surveyed via an online platform (Wenjuanxing). A self-developed eIC knowledge questionnaire and attitude scale were used to assess participants' understanding and perceptions of eIC. Univariate analysis was employed to explore factors influencing attitude scores and the correlation between knowledge and attitudes. A generalized linear model was used to analyze associations between demographic characteristics and attitude scores, including the frequency of difficulties in using smartphones or computers, preferred device for using eIC, and their interaction effects. Stratified analysis was further performed for significant interactions. RESULTS: A total of 399 valid questionnaires were collected. The mean accuracy rate on the eIC knowledge questionnaire was (94.88±15.50)%. Of the respondents, 74.9% had heard of eIC, and 84.5% preferred using mobile devices over computers to access eIC. The median attitude score was 3.41 (3.18, 3.76), indicating generally positive attitudes. Specifically, 81.7% found eIC more convenient than paper-based consent, 79.7% considered it more efficient, and 51.1% believed it could fully replace paper forms. However, 60.7% expressed concerns about data security and privacy, and 89.7% believed that relevant laws and regulations need improvement. Spearman correlation analysis showed a weak positive correlation between knowledge and attitude scores (r=0.171, P=0.001). Univariate analysis indicated that the frequency of difficulty using devices and preferred device for eIC were significantly associated with attitude scores (P<0.05). After adjusting for confounding factors, the generalized linear model demonstrated that participants who occasionally experienced had difficulty using devices had significantly lower attitude scores compared to those who never had difficulty (β=-0.040, 95% CI -0.071 to -0.009, P=0.012). Those who preferred using PCs had significantly lower attitude scores than those who preferred mobile devices (β=-0.066, 95% CI -0.108 to -0.023, P=0.002). Interaction analysis showed a significant interaction analysis showed a significant interaction between age and preferred device (P=0.011), particularly among participants aged ≥45-year (P<0.001). No other interactions were found to be significant (all P>0.05). CONCLUSIONS Medical researchers in China generally have a high level of knowledge and positive attitudes toward eIC, though concerns remain regarding data security and privacy. Future promotion of eIC in Chinese clinical research should be grounded in ethical considerations and address the specific needs of older users and mobile device users, while also enhancing researchers' competencies in using digital tools and eIC systems.
Humans ; Cross-Sectional Studies ; Informed Consent ; Surveys and Questionnaires ; Female ; Male ; Health Knowledge, Attitudes, Practice ; Adult ; Biomedical Research ; Research Personnel/psychology* ; Middle Aged ; China

OBJECTIVES: The 24-hour movement behaviors, comprising physical activity, sedentary behavior, and sleep, are crucial factors affecting children's mental health. This study aims to explore the longitudinal association between 24-hour movement behaviors and psychological well-being in overweight and obese children, providing empirical evidence for mental health promotion in this population. METHODS: A total of 445 overweight and obese children were recruited via stratified cluster random sampling from a provincial capital city in China and followed up for one year. Measures included objectively assessed physical activity and sleep duration using triaxial accelerometers (ActiGraph GT3X+), parent-reported sedentary screen-based time (SST), and self-reported psychological well-being. RESULTS: After one year, the proportion of children meeting all 3 movement guidelines increased from 10.11% to 11.68%, while those meeting none increased from 11.24% to 15.06%. After adjusting for relevant covariates, children who met individual guidelines for moderate-to-vigorous physical activity (MVPA) (β=0.377, 95% CI 0.209 to 0.545), sleep (β=0.187, 95% CI 0.042 to 0.332), or guideline combinations of MVPA+SST (β=0.545, 95% CI 0.377 to 0.713) and MVPA+sleep (β=0.602, 95% CI 0.449 to 0.755) showed significant improvements in psychological well-being after one year. Additionally, an increase in the number of guidelines met was significantly associated with improved well-being (β=0.113, 95% CI 0.011 to 0.214). CONCLUSIONS Adherence to any single movement guideline, especially MVPA or sleep, and combinations such as MVPA+SST or MVPA+sleep is significantly associated with enhanced psychological well-being in overweight and obese children. Integrated behaviors may be an effective strategy to improve mental well-being in this population.
Humans ; Child ; Exercise/psychology* ; Sleep ; Sedentary Behavior ; Female ; Male ; Pediatric Obesity/psychology* ; Overweight/psychology* ; Mental Health ; China ; Accelerometry ; Psychological Well-Being

OBJECTIVES: Urinary calculi are characterized by a high recurrence rate, and patients' adherence to self-management after discharge directly affects health outcomes. Traditional offline follow-up models often face problems such as poor compliance and uneven allocation of medical resources, making it difficult to meet individualized health management needs. Remote follow-up provides a novel solution to optimize long-term management, improve health literacy, and enhance clinical outcomes. This study aims to evaluate the effect of remote follow-up under an intelligent medical collaborative model on quality of life and health-promoting lifestyle in patients with urinary calculi, and to assess its short-term impact on clinical outcomes. METHODS: A total of 118 patients with urinary calculi admitted to a tertiary hospital in Hunan Province between August and November 2024 were recruited and randomly assigned to a control group (n=59) or an intervention group (n=59). The control group received routine departmental follow-up, while the intervention group underwent remote follow-up based on an intelligent medical collaborative model for one month. Assessments were conducted before discharge (T0), 15 days after discharge (T1), and one month after discharge (T2), using the Wisconsin Stone Quality of Life Questionnaire and the Health-Promoting Lifestyle Profile. At T2, the incidence of forgotten ureteral stents (FUS), ureteral stent-related complications, unplanned readmissions, and patient satisfaction were evaluated. RESULTS: No significant differences were observed between groups at T0 in baseline characteristics or outcome measures (all P>0.05). At T1 and T2, the intervention group had significantly higher health-related quality of life scores than the control group (P<0.05). Generalized estimating equation (GEE) analysis showed significant between-group effects (Wald's χ²=22.961, P<0.001), time effects (Wald's χ²=23.065, P<0.001), and interaction effects (Wald's χ²=6.930, P<0.05). Similarly, at T1 and T2, the intervention group scored significantly higher on health-promoting lifestyle than the control group (P<0.05), with significant between-group effects (Wald's χ²=22.936, P<0.001), time effects (Wald's χ²=10.694, P<0.001), and interaction effects (Wald's χ²=18.921, P<0.05). No significant differences were found between groups in the incidence of FUS, ureteral stent-related complications, or unplanned readmissions (all P>0.05). Patient satisfaction was significantly higher in the intervention group (t=4.089, P<0.001). CONCLUSIONS Remote follow-up under an intelligent medical collaborative model helps improve quality of life, promote health-oriented lifestyles, and enhance patient satisfaction among individuals with urinary calculi.
Humans ; Quality of Life ; Male ; Female ; Urinary Calculi/therapy* ; Health Promotion/methods* ; Middle Aged ; Adult ; Follow-Up Studies ; Treatment Outcome

OBJECTIVES: Accurate identification of risk factors for precocious puberty is essential for clinical diagnosis and management, yet the performance of natural language processing methods applied to unstructured electronic medical record (EMR) data remains to be fully evaluated. This study aims to assess the performance of a prompt engineering method for extracting individual risk factors of precocious puberty from EMRs. METHODS: Based on the capacity and role-insight-statement-personality-experiment (CRISPE) prompt framework, both simple and optimized prompts were designed to guide the large language model GLM-4-9B in extracting 10 types of risk factors for precocious puberty from 653 EMRs. Accuracy, precision, recall, and F1-score were used as evaluation metrics for the information extraction task. RESULTS: Under simple and optimized prompt conditions, the overall accuracy, precision, recall, and F1-score of the model were 84.18%, 98.09%, 81.99%, and 89.32% versus 97.15%, 98.31%, 98.16%, and 98.23%, respectively. The optimized prompts achieved more stable performance across age (<9 years vs ≥9 years) and visit-time (<2023 vs ≥2023) subgroups compared with simple prompts. The accuracy range for extracting each risk factor was 60.03%-97.24%, while with optimized prompts, the range improved to 92.19%-99.85%. The largest performance improvement occurred for "beverage intake" (60.03% vs 92.19%), and the smallest for "maternal age of menarche" (97.24% vs 99.23%). In comparing distributions among simple prompts, optimized prompts, and ground truth, statistically significant differences were observed for snack intake, beverage intake, soy milk intake, honey intake, supplement use, tonic use, sleep quality, and sleeping with the light on (all P<0.001), while exercise (P=0.966) and maternal menarche age (P=0.952) showed no significant differences. CONCLUSIONS Compared with simple prompts, optimized prompts substantially improved the extraction performance of individual risk factors for precocious puberty from EMRs, underscoring the critical role of prompt engineering in enhancing large language model performance.
Humans ; Puberty, Precocious/epidemiology* ; Risk Factors ; Electronic Health Records ; Female ; Child ; Natural Language Processing

OBJECTIVES: With the accelerating aging of the population and the rising prevalence of chronic diseases, the number of patients with pressure injuries (PIs) has increased markedly, prolonging the period of disease-related care. Informal caregivers play a critical role in the daily care of patients with pressure injuries, and their care burden has become increasingly prominent. This study aims to investigate the current status and influencing factors of care burden among informal caregivers of patients with PIs, providing evidence for targeted intervention strategies. METHODS: A total of 170 informal caregivers of patients with PIs were selected by convenience sampling from the Third Xiangya Hospital of Central South University. General demographic and clinical data of both patients and caregivers were collected. The Zarit Caregiver Burden Inventory (ZBI), Knowledge-Attitude-Practice Scale for Informal Caregivers of Patients with PIs, General Self-Efficacy Scale (GSES), and Family Caregiver Task Inventory (FCTI) were used to assess caregiving burden, knowledge-attitude-practice level, self-efficacy, and caregiving ability, respectively. Pearson correlation analysis was conducted to evaluate relationships among ZBI, Knowledge-Attitude-Practice Scale for Informal Caregivers of Patients with PIs, GSES, and FCTI scores. Stepwise multiple linear regression analysis was used to identify factors influencing caregiving. RESULTS: Among the 170 patients with pressure injuries, the age was (65.52±15.88) years; 118 (69.41%) were male and 52 (30.59%) were female. The duration of PIs was less than 1 month in 108 (63.53%) cases and 1 to 6 months in 40 cases (23.53%). Stage II injuries were predominant (135 cases, 79.41%). A total of 193 pressure injury sites were recorded, most commonly located at the sacrococcygeal region (127 sites, 65.80%), followed by the head (3 sites, 1.55%), shoulder and back (9 sites, 4.66%), feet (24 sites, 12.44%), and other regions (30 sites, 15.55%). Informal caregivers were 48.82% aged 46 to 59 years, 54.71% female, 41.77% primarily spouses and 47.06% children of the patients, and 77.06% lived with the patients. Caregivers who received assistance from others or had higher family per-capita monthly income reported significantly lower caregiver burden scores than those without assistance or with lower income (all P<0.001). The total ZBI score was 50.89±14.95, indicating a moderate burden. The total scores of the Knowledge-Attitude-Practice Scale for Informal Caregivers, GSES, and FCTI were 50.61±7.22, 26.03±7.11, and 14.76±8.70, respectively. Pearson correlation analysis revealed that ZBI scores were correlated with scores on the Knowledge-Attitude-Practice Scale for Informal Caregivers of Patients with PIs (r=-0.543, P<0.001), GSES scores (r=-0.545, P<0.001), and FCTI scores (r=0.800, P<0.001). The scores on Knowledge-Attitude-Practice Scale for Informal Caregivers of patients with PIs were correlated with GSES scores (r=0.500, P<0.001) and FCTI scores (r=-0.461, P<0.001); GSES scores was negatively correlated with FCTI scores (r=-0.415, P<0.001). Stepwise multiple linear regression analysis showed that assistance availability, family per-capita monthly income, total scores on the Knowledge-Attitude-Practice Scale for Informal Caregivers of Patients with PIs, total GSES score, and total FCTI score were the main influencing factors of caregiver burden, jointly explaining 79.38% of its variance. CONCLUSIONS The main factors influencing the caregiving burden of informal caregivers of patients with PIs include the availability of assistance, family per-capita monthly income, total score on the Knowledge-Attitude-Practice Scale for Informal Caregivers of PI patients, total score on the GSES, and total score on the FCTI. Developing targeted intervention strategies addressing these factors may help alleviate the caregiving burden among informal caregivers of patients with PIs.
Humans ; Caregivers/psychology* ; Pressure Ulcer/nursing* ; Female ; Male ; Middle Aged ; Cost of Illness ; Adult ; Aged ; Surveys and Questionnaires ; Health Knowledge, Attitudes, Practice ; Self Efficacy ; Caregiver Burden ; China

OBJECTIVES: Older adults in rural areas with multiple chronic conditions (MCC) generally exhibit poorer medication adherence than the general elderly population. Considering individual heterogeneity helps to design precise subgroup-based interventions. This study aims to identify latent profile types of medication adherence mechanisms among rural older adults with MCC based on the capability-opportunity-motivation-behavior (COM-B) model, and to explore factors influencing medication adherence. METHODS: A multistage sampling method was used to recruit 349 rural older adults with MCC from 10 administrative villages in Jianghua County, Yongzhou City, Hunan Province, between July and September, 2024. Participants were surveyed using a general information questionnaire, the Health Literacy Scale for Chronic Patients, the Beliefs about Medicines Questionnaire-Specific, the Multidimensional Scale of Perceived Social Support, and the Morisky Medication Adherence Scale. Latent profile analysis based on the COM-B model was conducted to identify subgroups of medication adherence mechanisms. Univariate and Logistic regression analyses were used to identify influencing factors associated with different latent profiles and adherence levels. RESULTS: Among the participants, 33.5% demonstrated good medication adherence. The 5 most prevalent chronic diseases were hypertension (86.5%), diabetes (36.7%), arthritis or rheumatism (34.4%), stroke (21.8%), and heart disease (17.5%). Overall, rural older adults with MCC exhibited relatively good medication capability, opportunity, and motivation. Their medication adherence mechanisms were classified into 3 latent profiles: "family-support restrained type" (5.2%), "family-support driven type" (52.1%), and "comprehensive advantage type" (42.7%). Significant differences were observed among the three profiles in terms of education level, marital status, living arrangement, and per capita monthly household income (all P<0.05). Multivariate Logistic regression revealed that higher education level was a protective factor for belonging to the "comprehensive advantage type" rather than the "family-support driven type" [OR=0.277, 95% CI (PL) 0.126 to 0.614, P=0.002]. Furthermore, significant differences in education level, self-rated health status, and latent profile type were found between participants with good and poor adherence (P<0.05). Binary Logistic regression indicated that with each one-level increase in self-rated health status, the risk of poor adherence increased by 293.9% [OR=3.939, 95% CI (PL) 1.610 to 9.636, P=0.003]. Compared with the "family-support restrained type", individuals classified as the "comprehensive advantage type" had a 96.8% [OR=0.032, 95% CI (PL) 0.008 to 0.123, P<0.001] lower risk of poor medication adherence. CONCLUSIONS The mechanisms underlying medication adherence among rural older adults with MCC show clear heterogeneity. Primary healthcare providers should focus on the "family-support restrained type" subgroup, strengthen social support networks, and implement targeted interventions to improve medication adherence.
Humans ; Aged ; Rural Population ; Male ; Female ; China ; Medication Adherence/psychology* ; Surveys and Questionnaires ; Chronic Disease/drug therapy* ; Multiple Chronic Conditions/drug therapy* ; Social Support ; Motivation ; Middle Aged ; Health Literacy ; Aged, 80 and over