Objective:To predict the risk of cryptogenic stroke (CS) patients with right-to-left shunt (RLS) by machine learning, and provide potential solutions for accurate and efficient prediction of CS.Methods:A retrospective analysis of clinical data on 289 subjects with positive RLS detected by contrast-enhanced transcranial Doppler tests (c-TCD) treated in the Department of Neurology at Laoshan Campus, the Affiliated Hospital of Qingdao University, from January 2018 to September 2023, including demographic information, medical history, laboratory test indicators, diagnosis, and treatment.The dataset was randomly divided into a training set and a testing set by the machine learning function train_test_split(), with a ratio of 8∶2.Risk prediction models for CS in RLS subjects were constructed by algorithms such as Logistic regression, decision trees, random forests, extreme gradient boosting, artificial neural networks, gradient boosting, extra trees, and adaptive Boosting.The model performance was evaluated by receiver operating characteristic curves (ROC), area under curve (AUC), confusion matrix, precision, recall, accuracy, F1 score, calibration curves, and decision curve analysis.The optimal model was subjected to interpretability analysis by feature importance and SHAP values.The t-test, Mann-Whitney U test and χ2 test were used for data analysis by SPSS 25.0 software.Delong test was used to compare the differences in AUC between the two models. Results:In 289 RLS subjects, there were 166 cases of CS (57.5%) and 123 cases of non-CS (42.5%).The statistical analysis results showed that blood biochemical indicators such as D-dimer, mean platelet volume, and fibrinogen in CS patients were higher than those in non-CS patients (all P<0.01).There were no statistically significant differences in variables between the training and testing sets(all P>0.05).Random forest model achieved the highest AUC (0.885), precision (0.806), recall (0.879), accuracy (0.810), and F1 score (0.841) for CS risk prediction in the testing set.The calibration curve showed that the random forest model was closest to the reference line, and the decision curve analysis indicated that it had a greater net benefit.The interpretability analysis revealed that high-risk factors included mean platelet volume, D-dimer, international normalized ratio, body mass index, and age. Conclusion:The random forest-based prediction tool exhibits excellent performance, demonstrating high accuracy in predicting CS risk in RLS population.

Objective:To investigate the surgical situations and perioperative outcome of pancreaticoduodenectomy in Jiangsu Province and the influencing factors for postoperative 90-day mortality.Methods:The retrospective case-control study was conducted. The clinicopathological data of 2 886 patients who underwent pancreaticoduodenectomy in 21 large tertiary hospitals of Jiangsu Quality Control Center for Pancreatic Diseases, including The First Affiliated Hospital of Nanjing Medical University, from March 2021 to December 2022 were collected. There were 1 732 males and 1 154 females, aged 65(57,71)years. Under the framework of the Jiangsu Provincial Pancreatic Disease Quality Control Project, the Jiangsu Quality Control Center for Pancreatic Diseases adopted a multi-center registration research method to establish a provincial electronic database for pancrea-ticoduodenectomy. Observation indicators: (1) clinical characteristics; (2) intraoperative and post-operative conditions; (3) influencing factors for 90-day mortality after pancreaticoduodenectomy. Measurement data with skewed distribution were represented as M( Q1, Q3) or M(IQR), and comparison between groups was conducted using the Mann-Whitney U test. Count data were expressed as absolute numbers or constituent ratio, and comparison between groups was conducted using the chi-square test, continuity correction chi-square test and Fisher exact probability. Maximal Youden index method was used to determine the cutoff value of continuous variables. Univariate analysis was performed using the corresponding statistical methods based on data types. Multivariate analysis was performed using the Logistic multiple regression model. Results:(1) Clinical characteristics. Of the 2 886 patients who underwent pancreaticoduodenectomy, there were 1 175 and 1 711 cases in 2021 and 2022, respectively. Of the 21 hospitals, 8 hospitals had an average annual surgical volume of <36 cases for pancreaticoduodenectomy, 10 hospitals had an average annual surgical volume of 36-119 cases, and 3 hospitals had an average annual surgical volume of ≥120 cases. There were 2 584 cases performed pancreaticoduodenectomy in thirteen hospitals with an average annual surgical volume of ≥36 cases, accounting for 89.536%(2 584/2 886)of the total cases. There were 1 357 cases performed pancrea-ticoduodenectomy in three hospitals with an average annual surgical volume of ≥120 cases, accounting for 47.020%(1 357/2 886) of the total cases. (2) Intraoperative and postoperative conditions. Of the 2 886 patients, the surgical approach was open surgery in 2 397 cases, minimally invasive surgery in 488 cases, and it is unknown in 1 case. The pylorus was preserved in 871 cases, not preserved in 1 952 cases, and it is unknown in 63 cases. Combined organ resection was performed in 305 cases (including vascular resection in 209 cases), not combined organ resection in 2 579 cases, and it is unknown in 2 cases. The operation time of 2 885 patients was 290(115)minutes, the volume of intra-operative blood loss of 2 882 patients was 240(250)mL, and the intraoperative blood transfusion rate of 2 880 patients was 27.153%(782/2 880). Of the 2 886 patients, the invasive treatment rate was 11.342%(327/2 883), the unplanned Intensive Care Unit (ICU) treatment rate was 3.087%(89/2 883), the reoperation rate was 1.590%(45/2 830), the duration of postoperative hospital stay was 17(11)days, the hospitalization mortality rate was 0.798%(23/2 882), and the failure rate of rescue data in 2 083 cases with severe complications was 6.529%(19/291). There were 2 477 patients receiving postoperative 90-day follow-up, with the 90-day mortality of 2.705%(67/2477). The total incidence rate of complication in 2 886 patients was 58.997%(1 423/2 412). The incidence rate of severe complication was 13.970%(291/2 083). The comprehensive complication index was 8.7(22.6) in 2 078 patients. (3) Influencing factors for 90-day mortality after pancreaticoduodenectomy. Results of multivariate analysis showed that age ≥ 70 years, postoperative invasive treatment, and unplanned ICU treatment were independent risk factors for 90-day mortality after pancreaticoduodenectomy ( odds ratio=2.403, 2.609, 16.141, 95% confidence interval as 1.281-4.510, 1.298-5.244, 7.119-36.596, P<0.05). Average annual surgical volume ≥36 cases in the hospital was an independent protective factor for 90-day mortality after pancreaticoduodenectomy ( odds ratio=0.368, 95% confidence interval as 0.168-0.808, P<0.05). Conclusions:Pancreaticoduodenectomy in Jiangsu Province is highly con-centrated in some hospitals, with a high incidence of postoperative complications, and the risk of postoperative 90-day mortality is significant higher than that of hospitallization mortality. Age ≥ 70 years, postoperative invasive treatment, and unplanned ICU treatment are independent risk factors for 90-day motality after pancreaticoduodenectomy, and average annual surgical volume ≥36 cases in the hospital is an independent protective factor.

Objective To know the difficulties in the communication between the nurse and the patient during the ocean-go-ing medical support, so that reference could be provided to the training of those who were engaged in such kind of missions in the fu-ture.Methods Qualitative research was used in the study.In-depth interviews were made with the 12 nurses who were engaged in the nursing of patients during the ocean-going medical support mission by Colaizzi method.Results As a first step, 4 subjects were summa-rized:i.e.psychological stress derived from exchanges with foreign patients in English, poor mastery of English medical terms, heavy dependence on non-verbal communication, and low awareness of multi-culture nursing care.Conclusion The administrator in charge of nursing care should intensify efforts for English language training of the nursing staff, pay more attention to the training of the ability in the communication between the nurse and the patient and heighten awareness of multi-culture nursing, so as to help the nursing staff to fulfill nursing tasks successfully during the ocean-going medical support.

Objective To identify the medical qualities and requirements for standardized training of medical officers in basic naval units , so as to provide evidence for the implementation of targeted standardized medical training in basic naval units .Methods Medical officers in the basic units of a certain fleet were chosen as research subjects and a questionnaire survey was conducted accord -ingly .The results obtained were analyzed statistically by using the statistical software .Results The scores of insufficiency for "medical manipulating skills"were the highest , comprehensive scores of work ability , work performance and work environment were moderate (7.97 ±1.55).The daily workload of medical officers in naval basic units was relatively lower and the diseases they commonly en -countered were upper respiratory tract infection , trauma, training injury and dermatosis , etc.The skills for the diagnosis and treatment of cardiovascular-cerebrovascular diseases , infectious diseases , gastrointestinal diseases , as well as manipulating skills of medical e-quipment demanded further improvement .71.13%of medical officers were in the opinion that it was absolutely necessary for them to receive standardized training , with a training duration of one year as the most desirable .Training focuses should be directed at the fol-lowing:the training of clinical analysis and solution of medical problems (80.06%), the training of basic skills and manipulating skills (67.56%) and extension of comprehensive knowledge (56.55%).Conclusion The professional qualities of navy medical officers were relatively weak, for this reason, there was an urgent requirement for them to have standardized medical training .It was recommen-ded that the mechanism of training , employment and management be developed in accordance with the practical conditions of basic u -nits, the time for training be properly arranged , standardized training subjects be optimized , and most importantly , the system for the training of medical officers in basic naval units be perfected step by step , so as to promote medical support capability of front-line units.

Objective To investigate the effect of "five-tube"treatment mode on emergency care of very severe burn pa-tients.Methods Eighteen very severe burn patients admitted into the hospital from January 2011 to January 2013 were chosen as the experimental group .The"five-tube"( i.e.the intravenous catheter , the oxygen-breathing tube , the tracheotomy tube , the indwelling urinary catheter and the indwelling gastric tube ) treatment mode was applied in the emergency care of the patients .The patients in the control group were 21 very severe burn patients admitted into the hospital from January 2008 to December 2010.They all received con-ventional emergency treatment and nursing care during the whole treatment course .Then, statistical analysis was made on the time of ef-fective emergency care.Results The time for the establishment of intravenous passage was (7.7 ±4.0)min, the time for the improve-ment of hypoxia was(11.5 ±6.2)min, and the time for early effective treatment was (35.9 ±9.5)min, which were all significantly shortened , as compared with those of the control group .Rate of mortality for the experimental group was also significantly reduced , and statistical differences could be seen, when comparisons were made between the 2 groups (P<0.01).Conclusion The "five-tube"treatment mode was worth further clinical extension in the emergency care of severe burn patients .

Objective To investigate the training requirements of military civilian nursing backbones in scientific nursing re-search.Methods A questionnaire survey was conducted among the 93 military civilian personnel taking part in the 8t h working-post training class of military civilian nursing backbones by using the questionnaire specifically developed for the study.Results Ninety per cent of the military civilian nursing backbones held that it was absolutely necessary to conduct training on scientific nursing research. The top 3 training requirements for scientific nursing research were:scientific nursing research design (accounting for 89%),medical statistics (accounting for 83%)and thesis writing skill (accounting for 80%).The top 3 requirements of training forms were:scientific research training courses (accounting for 85%),seminars (accounting for 82%)and expert instruction (accounting for 71%).A 2 to 4-week training period was most appropriate for scientific research training.Conclusion Attention should be paid on the part of the nursing administration to the selection of appropriate scientific research training programs and training methods,as well as flexible ar-rangement of training time,so as to improve scientific research capacity of the military civilian nursing backbones.