Objectives: . Idiopathic sudden sensorineural hearing loss (ISSNHL) is a significant cause of hearing loss. Intratympanic steroid injection (ITSI) is commonly used as an initial or salvage treatment; however, the lack of a standardized treatment protocol has resulted in variability in clinical practice. In addition, no efficient prediction model currently exists to support personalized management. Therefore, this study aimed to develop tailored management strategies for ISSNHL using a machine-learning model. Methods: . This retrospective multicenter cohort study was conducted between January 2015 and December 2020, with data analysis performed between January 2021 and March 2024. Patients were selected based on the International Classification of Diseases, 10th Revision criteria for ISSNHL, along with relevant medication and procedure codes. Patients with pure-tone audiogram results not meeting ISSNHL criteria, better initial hearing in the affected ear, an identifiable etiology, no post-treatment audiogram, or delayed treatment (>6 weeks) were excluded. We included 770 patients diagnosed with ISSNHL who received ITSI. The primary outcome was the area under the receiver operating characteristic curve for prediction performance. Recovery status was determined using the last pure-tone audiogram. Modeling was conducted on the Quanta for Medical Care AI platform using five machine-learning algorithms and a nested cross-validation framework, in which feature selection and hyperparameter tuning were performed in the inner folds and model performance was evaluated in the outer folds. Results: . A random forest classifier outperformed the other models in predicting hearing outcomes, achieving an area under the receiver operating characteristic curve of 0.788. Time to ITSI was the most influential treatment-related factor, with ITSI administered within 10 days of hearing loss being associated with better outcomes. This model can be used to provide personalized prognostic estimates under different treatment protocols. Conclusion . The machine-learning-based prediction model facilitates personalized treatment strategies and timely treatment adjustments for ISSNHL, thereby optimizing the likelihood of complete recovery.

BACKGROUND: Severe stroke has high rates of mortality and morbidity. This study aimed to investigate the clinical course, causes of worsening, and outcomes of severe ischemic stroke. METHODS: This prospective, multicenter cohort study enrolled adult patients admitted ≤30 days after ischemic stroke from nine hospitals in China between September 2017 and December 2019. Severe stroke was defined as a score of ≥15 on the National Institutes of Health Stroke Scale (NIHSS). Clinical worsening was defined as an increase of 4 in the NIHSS score from baseline. Unfavorable functional outcome was defined as a modified Rankin scale score ≥3 at 3 months and 1 year after stroke onset, respectively. We performed Logistic regression to explore baseline features and reperfusion therapies associated with clinical worsening and functional outcomes. RESULTS: Among 4201 patients enrolled, 854 patients (20.33%) had severe stroke on admission. Of 3347 patients without severe stroke on admission, 142 (4.24%) patients developed severe stroke in hospital. Of 854 patients with severe stroke on admission, 33.95% (290/854) experienced clinical worsening (median time from stroke onset: 43 h, Q1-Q3: 20-88 h), with brain edema (54.83% [159/290]) as the leading cause; 24.59% (210/854) of these patients died by 30 days, and 81.47% (677/831) and 78.44% (633/807) had unfavorable functional outcomes at 3 months and 1 year respectively. Reperfusion reduced the risk of worsening (adjusted odds ratio [OR]: 0.24, 95% confidence interval [CI]: 0.12-0.49, P  <0.01), 30-day death (adjusted OR: 0.22, 95% CI: 0.11-0.41, P  <0.01), and unfavorable functional outcomes at 3 months (adjusted OR: 0.24, 95% CI: 0.08-0.68, P <0.01) and 1 year (adjusted OR: 0.17, 95% CI: 0.06-0.50, P  <0.01). CONCLUSIONS: Approximately one-fifth of patients with ischemic stroke had severe neurological deficits on admission. Clinical worsening mainly occurred in the first 3 to 4 days after stroke onset, with brain edema as the leading cause of worsening. Reperfusion reduced the risk of clinical worsening and improved functional outcomes. REGISTRATION ClinicalTrials.gov , NCT03222024.
Humans ; Male ; Female ; Prospective Studies ; Ischemic Stroke/mortality* ; Aged ; Middle Aged ; Aged, 80 and over ; Stroke ; Brain Ischemia

OBJECTIVE: To explore clinical efficacy of percutaneous endoscopic discectomy with a lateral approach and dual-channel method in treating highly free lumbar disc herniation(LDH). METHODS: A retrospective analysis was conducted on 54 patients with highly free LDH who were treated with spinal endoscopic techniques from January 2021 to December 2022. Twenty-seven patients were treated with lateral approach dual-channel(lateral approach dual-channel group), including 16 males and 11 females, with an average age of (54.6±10.5) years old. Twenty-seven patients were treated with unilateral biportal endoscopic (UBE group), including 17 males and 10 females, with an average age of (52.9±12.3) years old. The number of intraoperative fluoroscopy, operation time and hospital stay, as well as visual analogue scale (VAS) and Oswestry diability index (ODI) of low back and leg pain between two patients before operation, 1 day, 1, 3, and 12 months after operation, and the efficacy was evaluated by the modified MacNab criteria at 12 mohths after operation. RESULTS: All patients were successfully completed surgical and were followed up, the time raged from 12 to 22 months with an average of (13.57±4.12) months. There was no statistically significant difference in operation time between two groups (P>0.05). The hospital stay of lateral approach dual-channel group was (3.9±1.1) days, which was shorter than that of UBE group (6.5±1.4) days, the number of intraoperative fluoroscopy in lateral approach dual-channel group was (12.7±2.1) times, which was more than that in UBE group (6.6±1.3) times, the differences were statistically significant (t=5.197, -7.532;P<0.05). VAS and ODI for low back pain at 1 day and 1 month after operation, and VAS for leg pain at 1 day after operation of lateral approach dual-channel group were superior to those of UBE group, and the differences were statistically significant (P<0.05). However, there were no statistically significant differences in VAS and ODI for low back and leg pain between two groups before operation and 3 and 12 months after operation (P>0.05). VAS and ODI of low back and leg pain were significantly improved at each time point before and after operation in both groups, and the difference were statistically significant (P<0.05). At 12 months after operation, according to the modified MacNab criteria, the excellent and good rates of therapeutic effects between lateral approach dual-channel group and UBE group were 92.6% (25/27) and 88.9% (24/27), respectively, and the difference was not statistically significant (χ²=0.22, P>0.05). CONCLUSION For patients with highly free lumbar intervertebral disc protrusion, both of lateral approach dual-channel method and UBE endoscopic surgery are safe and effective. Endoscopic surgery with lateral approach and dual-channel method could be performed under local anesthesia, allowing for the removal of the nucleus pulposus under direct vision. It is simpler, more efficient.
Humans ; Male ; Female ; Intervertebral Disc Displacement/surgery* ; Middle Aged ; Diskectomy, Percutaneous/methods* ; Lumbar Vertebrae/surgery* ; Endoscopy/methods* ; Adult ; Retrospective Studies ; Aged

This study explores whether the current external quality assessment (EQA) level and acceptable bias for basic semen analysis in China are clinically useful. We collected data of semen EQA from Andrology laboratories in the Hunan Province (China) in 2022 and searched for data in the published literature from January 2000 to December 2023 in China. On the basis of these data, we analyzed the coefficients of variation and acceptable biases of different quality control materials for basic semen analysis through robust statistics. We compared these findings with quality specifications based on biological variation from optimal, desirable, and minimum levels of bias to seek a unified and more suitable semen EQA bias evaluation standard for China's national conditions. Different sources of semen quality control material exhibited considerable variation in acceptable biases among laboratories, ranging from 8.2% to 56.9%. A total of 50.0% of the laboratories met the minimum quality specifications for progressive motility (PR), whereas 100.0% and 75.0% of laboratories met only the minimum quality specifications for sperm concentration and total motility (nonprogressive [NP] + PR), respectively. The Z value for sperm concentration and PR+NP was equivalent to the desirable performance specification, whereas the Z value for PR was equivalent only to the minimum performance specification. This study highlights the feasibility of operating external quality assessment schemes for basic semen analysis using quality specifications based on biological variation. These specifications should be unified among external quality control (EQC) centers based on biological variation.
Semen Analysis/standards* ; Humans ; China ; Male ; Quality Control ; Sperm Motility ; Sperm Count/standards*

Objective: Machine learning (ML) has been reported to have better predictive capability than traditional statistical techniques. The aim of this study was to assess the efficacy of ML algorithms and logistic regression (LR) for predicting depressive symptoms during the COVID-19 pandemic. Methods: Analyses were carried out in a national cross-sectional study involving 21,916 participants. The ML algorithms in this study included random forest (RF), support vector machine (SVM), neural network (NN), and gradient boosting machine (GBM) methods. The performance indices were sensitivity, specificity, accuracy, precision, F1-score, and area under the receiver operating characteristic curve (AUC). Results: LR and NN had the best performance in terms of AUCs. The risk of overfitting was found to be negligible for most ML models except for RF, and GBM obtained the highest sensitivity, specificity, accuracy, precision, and F1-score. Therefore, LR, NN, and GBM models ranked among the best models. Conclusion Compared with ML models, LR model performed comparably to ML models in predicting depressive symptoms and identifying potential risk factors while also exhibiting a lower risk of overfitting.

Objective: Machine learning (ML) has been reported to have better predictive capability than traditional statistical techniques. The aim of this study was to assess the efficacy of ML algorithms and logistic regression (LR) for predicting depressive symptoms during the COVID-19 pandemic. Methods: Analyses were carried out in a national cross-sectional study involving 21,916 participants. The ML algorithms in this study included random forest (RF), support vector machine (SVM), neural network (NN), and gradient boosting machine (GBM) methods. The performance indices were sensitivity, specificity, accuracy, precision, F1-score, and area under the receiver operating characteristic curve (AUC). Results: LR and NN had the best performance in terms of AUCs. The risk of overfitting was found to be negligible for most ML models except for RF, and GBM obtained the highest sensitivity, specificity, accuracy, precision, and F1-score. Therefore, LR, NN, and GBM models ranked among the best models. Conclusion Compared with ML models, LR model performed comparably to ML models in predicting depressive symptoms and identifying potential risk factors while also exhibiting a lower risk of overfitting.

Objective: Machine learning (ML) has been reported to have better predictive capability than traditional statistical techniques. The aim of this study was to assess the efficacy of ML algorithms and logistic regression (LR) for predicting depressive symptoms during the COVID-19 pandemic. Methods: Analyses were carried out in a national cross-sectional study involving 21,916 participants. The ML algorithms in this study included random forest (RF), support vector machine (SVM), neural network (NN), and gradient boosting machine (GBM) methods. The performance indices were sensitivity, specificity, accuracy, precision, F1-score, and area under the receiver operating characteristic curve (AUC). Results: LR and NN had the best performance in terms of AUCs. The risk of overfitting was found to be negligible for most ML models except for RF, and GBM obtained the highest sensitivity, specificity, accuracy, precision, and F1-score. Therefore, LR, NN, and GBM models ranked among the best models. Conclusion Compared with ML models, LR model performed comparably to ML models in predicting depressive symptoms and identifying potential risk factors while also exhibiting a lower risk of overfitting.

Objective: Machine learning (ML) has been reported to have better predictive capability than traditional statistical techniques. The aim of this study was to assess the efficacy of ML algorithms and logistic regression (LR) for predicting depressive symptoms during the COVID-19 pandemic. Methods: Analyses were carried out in a national cross-sectional study involving 21,916 participants. The ML algorithms in this study included random forest (RF), support vector machine (SVM), neural network (NN), and gradient boosting machine (GBM) methods. The performance indices were sensitivity, specificity, accuracy, precision, F1-score, and area under the receiver operating characteristic curve (AUC). Results: LR and NN had the best performance in terms of AUCs. The risk of overfitting was found to be negligible for most ML models except for RF, and GBM obtained the highest sensitivity, specificity, accuracy, precision, and F1-score. Therefore, LR, NN, and GBM models ranked among the best models. Conclusion Compared with ML models, LR model performed comparably to ML models in predicting depressive symptoms and identifying potential risk factors while also exhibiting a lower risk of overfitting.

As a new type of production factor that can empower the development of new quality productivity, the data element is an important engine to promote the high quality development of the industry. Traditional Chinese medicine(TCM) dispensing is the most basic work of TCM clinical pharmacy, and its quality directly affects the clinical efficacy of TCM. The integration of data elements and TCM dispensing can stimulate the innovation and vitality of the TCM dispensing industry and promote the high-quality and sustainable development of the industry. A large-scale, detailed, and systematic study on TCM dispensing was conducted. The innovative practice path of data fusion construction in the whole process of TCM dispensing was investigated by integrating the digital resources "nine full activities" of TCM dispensing, creating the digital dictionary of "TCM clinical information data elements", and exploring innovative applications of TCM dispensing driven by data and technology, so as to promote the standardized, digital, and intelligent development of TCM dispensing in medical health services. The research content of this project was successfully selected as the second batch of "Data element×" typical cases of National Data Administration in 2024, which is the only selected case in the field of TCM.
Medicine, Chinese Traditional/methods* ; Drugs, Chinese Herbal ; Humans