[Construction of integrated platform for emergency clinical scientific research based on big data].

OBJECTIVE: To explore clinical rules based on the big data of the emergency department of the Second Affiliated Hospital of Guangzhou Medical University, and to establish an integrated platform for clinical research in emergency, which was finally applied to clinical practice. METHODS: Based on the hospital information system (HIS), laboratory information system (LIS), emergency specialty system, picture archiving and communication systems (PACS) and electronic medical record system of the Second Affiliated Hospital of Guangzhou Medical University, the structural and unstructured information of patients in the emergency department from March 2019 to April 2022 was extracted. By means of extraction and fusion, normalization and desensitization quality control, the database was established. In addition, data were extracted from the database for adult patients with pre screening triage level III and below who underwent emergency visits from March 2019 to April 2022, such as demographic characteristics, vital signs during pre screening triage, diagnosis and treatment characteristics, diagnosis and grading, time indicators, and outcome indicators, independent risk factors for poor prognosis in patients were analyzed. RESULTS: (1) The data of 338 681 patients in the emergency department of the Second Affiliated Hospital of Guangzhou Medical University from March 2019 to April 2022 were extracted, including 15 modules, such as demographic information, triage information, visit information, green pass and rescue information, diagnosis information, medical record information, laboratory examination overview, laboratory information, examination information, microbiological information, medication information, treatment information, hospitalization information, chest pain management and stroke management. The database ensured data visualization and operability. (2) Total 140 868 patients with pre-examination and triage level III and below were recruited from the emergency department database. The gender, age, type of admission to the hospital, pulse, blood pressure, Glasgow coma scale (GCS) and other indicators of the patients were included. Taking emergency admission to operating room, emergency admission to intervention room, emergency admission to intensive care unit (ICU) or emergency death as poor prognosis, the poor prognosis prediction model for patients with pre-examination and triage level III and below was constructed. The receiver operator characteristic curve and forest map results showed that the model had good predictive efficiency and could be used in clinical practice to reduce the risk of insufficient emergency pre-examination and triage. CONCLUSIONS: The establishment of high-quality clinical database based on big data in emergency department is conducive to mining the clinical value of big data, assisting clinical decision-making, and improving the quality of clinical diagnosis and treatment.

Machine learning-based models to support decision-making in emergency department triage for patients with suspected cardiovascular disease.

BACKGROUND: Accurate differentiation and prioritization in emergency department (ED) triage is important to identify high-risk patients and to efficiently allocate of finite resources. Using data available from patients with suspected cardiovascular disease presenting at ED triage, this study aimed to train and compare the performance of four common machine learning models to assist in decision making of triage levels. METHODS: This cross-sectional study in the second Affiliated Hospital of Guangzhou Medical University was conducted from August 2015 to December 2018 inclusive. Demographic information, vital signs, blood glucose, and other available triage scores were collected. Four machine learning models - multinomial logistic regression (multinomial LR), eXtreme gradient boosting (XGBoost), random forest (RF) and gradient-boosted decision tree (GBDT) - were compared. For each model, 80 % of the data set was used for training and 20 % was used to test the models. The area under the receiver operating characteristic curve (AUC), accuracy and macro- F1 were calculated for each model. RESULTS: In 17,661 patients presenting with suspected cardiovascular disease, the distribution of triage of level 1, level 2, level 3 and level 4 were 1.3 %, 18.6 %, 76.5 %, and 3.6 % respectively. The AUCs were: XGBoost (0.937), GBDT (0.921), RF (0.919) and multinomial LR (0.908). Based on feature importance generated by XGBoost, blood pressure, pulse rate, oxygen saturation, and age were the most significant variables for making decisions at triage. CONCLUSION: Four machine learning models had good discriminative ability of triage. XGBoost demonstrated a slight advantage over other models. These models could be used for differential triage of low-risk patients and high-risk patients as a strategy to improve efficiency and allocation of finite resources.

Behavioral activation for dementia caregivers: scheduling pleasant events and enhancing communications.

BACKGROUND: Dementia caregiving is often associated with increase in depressive symptoms and strained relationships. This study tested whether telephone-delivered psychoeducation combined with an enhanced behavioral activation (BA) module had a better effect on the well-being of Alzheimer's caregivers than psychoeducation alone. The focus is on enhancing the competent use of coping skills via BA. The program is delivered by telephone to increase accessibility and sustainability for caregivers. Senior citizens are trained as paraprofessionals to deliver the BA module to increase the potential for sustainability of the program. METHODS AND SUBJECTS: The study compared two telephone interventions using a 4-month longitudinal randomized controlled trial. For the first 4 weeks, all participants received the same psychoeducation program via telephone. Then for the following 4 months, eight biweekly telephone follow-up calls were carried out. For these eight follow-up calls, participants were randomized into either one of the two following groups with different conditions. For the psychoeducation with BA (PsyED-BA) group, participants received eight biweekly sessions of BA practice focused on pleasant event scheduling and improving communications. For the psychoeducation only (PsyED only) group, participants received eight biweekly sessions of general discussion of psychoeducation and related information. A total of 62 family caregivers of persons living with dementia were recruited and 59 (29 in the PsyED-BA group and 30 in the PsyED only group) completed the whole study. RESULTS: As compared to the group with psychoeducation and discussion, the group with enhanced BA had decreased levels of depressive symptoms. The study had a low attrition rate. CONCLUSION: Results suggested that competence-based training could be effectively administered through the telephone with the help of senior citizens trained and engaged as paraprofessionals. Results contribute to the present literature by offering some framework for developing effective, accessible, sustainable, and less costly interventions.

Talking about others facilitates theory of mind in Chinese preschoolers.

A longitudinal study and a training study were conducted to show that simply referring to others facilitated theory of mind (ToM) development in Chinese children. In Study 1, 3- to 4-year-old Chinese children (N = 52) were tested on ToM and autobiographical memory (AM). One year later, in the group of children who initially failed the false belief tasks, only those who increased their references to others in AM recall passed the tasks. In Study 2, Chinese preschoolers who were trained to talk about others through storytelling showed improvement in their ToM performance. These findings suggest alternative pathways for ToM development in non-Euro-American context.