关键词: 护理人员, 核心能力, logistic回归, 决策树, 影响因素

Abstract:

Objective　To analyze the factors influencing core competence in oncology nursing among nursing staff in Shaanxi Province using logistic regression and decision tree interaction detection, and to provide some evidences for optimizing training and intervention strategies. Methods　A total of 1 529 nurses from 62 hospitals in Shaanxi Province from August to September 2022 were selected as the survey subjects, including 26 males and 1 503 females. There were 397 nurses, 580 senior nurses, 485 nurses-in-charge, and 67 chief nurses. A self-designed general information and core competency questionnaire was used for the cross-sectional survey. χ² test, logistic regression analysis, and decision tree interaction detection were used to analyze the differences between the results of the two models. Results　A total of 1 650 questionnaires were distributed, and 1 529 valid ones were ultimately collected, with an effective response rate of 92.67%. The total core competency score was 186.24±59.76, indicating a moderate proficiency. The logistic regression model showed that obtaining municipal-level oncology certification (OR=0.43, 95%CI 0.26-0.73), participating in oncology training (OR=2.19, 95%CI 1.31-3.60), and working in oncology departments (OR=0.51, 95%CI 0.37-0.71) were the predictive factors of core competence in oncology nursing. The decision tree analysis prioritized work satisfaction as the primary factor, followed by certification, training, and oncology department affiliation. The area under the ROC (AUC) of the logistic regression model was 0.745, and the AUC of the decision tree model was 0.721. There is no statistical difference in the prediction efficiency between the two prediction models (Z=1.943; P>0.05). Conclusions　Core competence in oncology nursing of nursing staff in Shaanxi Province is at a moderate level. Work satisfaction, certification, training, and oncology department affiliation are critical influencing factors. Integrating logistic regression's quantitative risk assessment with decision tree's dynamic pathway analysis enhances prediction accuracy. Managers should adopt dual-model strategies to establish baseline standards and design tiered interventions for high-risk subgroups.

Key words: Nursing staff, Core competence, Logistic regression, Decision tree, Influencing factors