Identifying contributing-factor configurations in reported bed falls: a Bayesian network-based exploratory study

基于贝叶斯网络的探索性研究:识别报告的跌倒事件中的促成因素

阅读:1

Abstract

OBJECTIVE: This study aimed to explore the conditional dependency structure of contributing factors related to bed falls and to identify stable dependency configurations within reported bed falls incident reports using Bayesian network analysis. METHODS: A retrospective analysis was conducted using 102 inpatient bed-fall incident reports collected from a tertiary hospital between 2014 and 2024. Twenty-two previously identified contributing factors were encoded as binary variables. Bayesian network structure learning was performed using a hill-climbing algorithm with the Bayesian Information Criterion, without imposing prior structural constraints. Structural stability was assessed through non-parametric bootstrap resampling. Conditional inference and posterior probability analyses were used to examine probability changes and co-occurrence patterns under specific clinical conditions. RESULTS: Bootstrap analysis identified several structurally stable directed edges in the Bayesian network. Sudden changes in patient consciousness were repeatedly connected to medication-related hypotension and safety measure-related nodes, forming a central structural component of the network. A stable directed relationship was also observed between nighttime sedative use with toileting and unassisted bed exit. In addition, delayed patient assistance was more frequently linked to delayed responses by on-site caregivers than to complete caregiver absence. Together, these results indicate that reported bed falls are associated with a limited number of recurrent and structurally consistent contributing-factor configurations. CONCLUSION: Reported bed falls appear to arise from interactions among a small number of contributing factors under specific clinical contexts rather than from isolated causes. Bayesian network analysis offers a useful approach for identifying contributing-factor configurations and informing scenario-based fall prevention strategies in clinical nursing practice.

特别声明

1、本页面内容包含部分的内容是基于公开信息的合理引用;引用内容仅为补充信息,不代表本站立场。

2、若认为本页面引用内容涉及侵权,请及时与本站联系,我们将第一时间处理。

3、其他媒体/个人如需使用本页面原创内容,需注明“来源:[生知库]”并获得授权;使用引用内容的,需自行联系原作者获得许可。

4、投稿及合作请联系:info@biocloudy.com。