About the Risk Prediction Tool
Background
In a prospective direct observational study conducted in Neonatal Intensive Care Units (NICUs) of five public hospitals in Malaysia, we investigated the prevalence of medication administration errors (MAEs), factors associated with these errors, and potential clinical and economic outcomes. The error rate recorded in this study was 68.0%, affecting 92.4% of the neonates in the NICU. The majority of these errors were categorised as potentially moderate. The estimated potential economic outcome of the observed MAEs was MYR 43,664.16 (USD 27,452.10).
Purpose
These findings highlighted the burden of MAEs and the critical need to prioritise and develop strategies to identify and potentially reduce their occurrence. Consequently, we developed a machine learning-based risk prediction tool to predict the presence of MAEs in NICUs.
Methodology
Our rigorous methodology included:
Direct observational study
Dataset analysis
Application of advanced machine learning techniques
We employed these methods to address the complexity of factors contributing to MAEs and to overcome the limitations of traditional statistical methods in capturing nonlinear relationships within the data.
Key Findings
Our study identified AdaBoost as the most effective machine learning algorithm for predicting MAEs, outperforming other algorithms. This tool aids in predicting the presence of MAEs in NICUs.
Impact
While the implementation of this tool alone may not directly reduce MAEs, it can guide healthcare providers in making timely and appropriate interventions. By facilitating proactive measures, this model has the potential to enhance patient safety and improve neonatal health outcomes in NICUs.