A machine learning-based approach for precision risk stratification and multifactorial analysis of needlestick
Xiulan Deng1, Jiayang Han2, Linlin Yin3
1Shandong Provincial Hospital Affiliated to Shandong First Medical University, Jinan, Shandong, 250014, China.
Background:
Needlestick injuries are a significant occupational hazard for oral and maxillofacial surgery nurses. This hazard results from complex procedures, limited workspace, and frequent handling of sharp instruments. This study uses advanced clustering and dimensionality reduction techniques to identify high-risk groups and key contributing factors.
Methods:
A structured questionnaire was administered to 224 nurses in five hospitals in Shandong Province. Spearman correlation analysis was used to identify critical risk variables, while K-means clustering and t-SNE visualization were used for risk stratification.
Results:
The study results showed that 33.93% of the nurses were classified as high risk, 35.27% as medium risk, and 30.8% as low risk. Analysis revealed that nurses in the high-risk category experienced significantly poor working conditions and suboptimal instrument management (P < 0.001), as well as lower levels of patient cooperation and more challenging surgical environments (P < 0.001).
Conclusions:
These findings underscore the urgent need for data-driven, targeted interventions to improve the surgical environment, optimize instrument management, and enhance patient cooperation, providing a robust framework for reducing needlestick injuries in oral and maxillofacial surgical care.
Trial Registration:
Not applicable.
