Venous thromboembolism(VTE)is a prevalent complication in orthopedics.In recent years,machine learning has been widely applied in orthopedics.The essence of machine learning lies in utilizing algorithms to analyze vast amounts of data and construct risk prediction models that can accurately forecast unknown clinical outcomes.By integrating high-risk factors,machine learning aids medical professionals in precisely identifying and screening individuals with a high risk of VTE and offering them timely individualized interventions.This article reviews the concept and classification of machine learning,the advantages of machine learning in enhancing model prediction ability,and the current application status of machine learning in constructing risk prediction models for patients with VTE.

In recent years， adaptive design has gradually emerged in clinical trials， gaining widespread application and expansion. Compared with traditional randomized controlled trials， appropriate adaptive design can save time and resources to some extent， improve the validity of the trial results and strength of the evidence， align better with ethical requirements， and facilitate subsequent drug market research. However， due to the greater flexibility of adaptive design， controlling type Ⅰ error rates and truly enhancing research efficiency are challenging aspects of its application. Therefore， the normative application and optimization of adaptive design in research deserve early focus， including adhering to good adaptive design principles， conducting rigorous pre-design planning， scientifically applying various statistical methods such as Bayesian dynamic models and simulation methods， and utilizing monitoring tools like monitoring committees to ensure the scientific integrity of the research process. Moreover， the flexibility of adaptive design aligns well with the practical realities of traditional Chinese medicine （TCM） clinical practice. The main characteristics of TCM， such as the holistic concept and treatment based on syndrome differentiation， determine that TCM provides a dynamic and personalized treatment plan. In contrast， the traditional fixed development approach of western medicine limits the application of "human experience" in TCM. To address this bottleneck in TCM research and development， the innovative use of adaptive design， a modern scientific theory and methodology with multiplicity and variability， offers a feasible new approach for TCM clinical research and the innovation of Chinese herbal medicine. By learning from successful cases such as the first Food and Drug Administration （FDA）-approved oral botanical drug， Crofelemer， researchers can explore optimal solutions for adaptive design in clinical research protocols and its role in the "three-combination" evaluation system for Chinese herbal medicines， further refine trial design， clarify key trial points， and develop innovative research and development applications that adapt to the clinical positioning of TCM and highlight the characteristics and advantages of TCM， paving the way for the clinical efficacy， pharmacological mechanism exploration， and post-market efficacy evaluation of Chinese herbal compounds.

Objective:To investigate the association of urinary cadmium levels with peripheral leukocyte classification counts among middle-aged and older adults aged 40 to 89 years in selected areas of China.Methods:The research was based on the survey of the impact of soil quality of agricultural land on human health in typical areas conducted in 2019-2020. A total of 5 600 middle-aged and older adults aged 40 to 89 years were included by using a multi-stage stratified random sampling method. Baseline characteristics of the subjects were collected and physical examinations were performed. Random midstream urine was collected to measure urinary cadmium and urinary creatinine and fasting venous blood was collected to measure the leukocyte count, neutrophil count, lymphocyte count, monocyte count and eosinophil count. The linear mixed effect model was used to analyse the association of urinary cadmium levels with leukocyte classification counts, and the dose-response relationship between them was analyzed by using the restricted cubic spline (RCS) function.Results:The age of the subjects was (63.17±12.02) years; 2 851 (50.91%) were males; and the M ( Q 1, Q 3) of urinary creatinine-corrected urinary cadmium levels was 2.69 (1.52, 4.69) μg/g·creatinine. After adjusting for confounding factors, the results of linear mixed effects model analysis showed that for each 1-unit increase in urinary creatinine-corrected urinary cadmium level, the percentage change [% (95% CI)] of leukocyte count and lymphocyte count was -1.70% (-2.61%, -0.79%) and -1.57% (-2.86%, -0.26%), respectively. RCS function showed a negative linear relationship between urinary creatinine-corrected urinary cadmium levels and leukocyte counts and lymphocyte counts, respectively (all Pnon-linear>0.05). Conclusion:Urinary cadmium levels are negatively associated with leukocyte count and lymphocyte count among middle-aged and older adults aged 40 to 89 years in selected areas of China.

Objective:To investigate the association of urinary cadmium levels with peripheral leukocyte classification counts among middle-aged and older adults aged 40 to 89 years in selected areas of China.Methods:The research was based on the survey of the impact of soil quality of agricultural land on human health in typical areas conducted in 2019-2020. A total of 5 600 middle-aged and older adults aged 40 to 89 years were included by using a multi-stage stratified random sampling method. Baseline characteristics of the subjects were collected and physical examinations were performed. Random midstream urine was collected to measure urinary cadmium and urinary creatinine and fasting venous blood was collected to measure the leukocyte count, neutrophil count, lymphocyte count, monocyte count and eosinophil count. The linear mixed effect model was used to analyse the association of urinary cadmium levels with leukocyte classification counts, and the dose-response relationship between them was analyzed by using the restricted cubic spline (RCS) function.Results:The age of the subjects was (63.17±12.02) years; 2 851 (50.91%) were males; and the M ( Q 1, Q 3) of urinary creatinine-corrected urinary cadmium levels was 2.69 (1.52, 4.69) μg/g·creatinine. After adjusting for confounding factors, the results of linear mixed effects model analysis showed that for each 1-unit increase in urinary creatinine-corrected urinary cadmium level, the percentage change [% (95% CI)] of leukocyte count and lymphocyte count was -1.70% (-2.61%, -0.79%) and -1.57% (-2.86%, -0.26%), respectively. RCS function showed a negative linear relationship between urinary creatinine-corrected urinary cadmium levels and leukocyte counts and lymphocyte counts, respectively (all Pnon-linear>0.05). Conclusion:Urinary cadmium levels are negatively associated with leukocyte count and lymphocyte count among middle-aged and older adults aged 40 to 89 years in selected areas of China.