Development of an algorithm to improve on the National Early Warning Score 2 (NEWS2) system's accuracy in predicting critical outcomes using additional patient data and amendments to the scoring process
Introduction: The second National Early Warning Score (NEWS2) is a widely used tool for the systematic identification and documentation of clinical deterioration. It is effective at reducing in-hospital mortality,1,2 facilitates effective communication between clinicians and enables timely intervent...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-07-01
|
| Series: | Clinical Medicine |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S1470211825000983 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Introduction: The second National Early Warning Score (NEWS2) is a widely used tool for the systematic identification and documentation of clinical deterioration. It is effective at reducing in-hospital mortality,1,2 facilitates effective communication between clinicians and enables timely interventions to improve patient outcomes, but NEWS2 has limited positive and negative predictive accuracy,3 particularly in predicting adverse events beyond 24 h.4The use of digital technologies in healthcare presents an opportunity to evaluate whether the inclusion of additional routinely collected variables, and/or changes in the use of existing data, would improve the predictive accuracy of an early warning score and, thus, reduce patient risk.1,5Older adults are particularly vulnerable to sudden changes in physiological status,6 but this group was not included in the development of NEWS2.7 Improved predictive accuracy in this growing population group could significantly improve patient outcomes.8 Methods and Methods: The project is a retrospective cohort study using ∼8 years (2017–2024) of anonymised patient data from the Newcastle upon Tyne Hospitals NHS Foundation Trust (NUTH). The study includes any patients who were admitted to NUTH between 2017 and 2024 and had physiological observations recorded in the form of NEWS2. People who had opted out of the use of their de-identified data for research either locally or nationally, children (age <16 years old) and maternity admissions were excluded (Fig 1).A scoping review of six databases (CINAHL, PubMed, Embase, ScienceDirect, Cochrane Library and Web of Science) was conducted9 to determine which variables may improve the risk prediction of NEWS2. Data analysis: De-identified data from the Trust’s clinical data warehouse will be divided into training and testing subsets. Using these datasets, we will train and test an algorithm that optimises the variables and their weightings to predict the risk of key clinical outcomes, including mortality, intensive care unit admission, sepsis and cardiac arrest, to demonstrate a proof of concept for a modified scoring system. Intended outputs: By June 2025, an algorithm based on 8 years of historical patient data will have been trained and tested on NUTH datasets to create a new tool that is of national and international importance. The aim of this project is to improve the predictive accuracy of the NEWS2 scoring system, particularly over more than 24-h, and in older patients. If the new system performs better than NEWS2, we will conduct implementation studies in Newcastle and then in other NHS Trusts to assess adoption of the system and gather initial data on patient outcomes to construct a learning healthcare system to evaluate real-word performance of the algorithm and facilitate continual improvement.10 Ethics: An application for the ethical approval of this study has been submitted to the UK The Health Research Authority Integrated Research Approval System and has undergone proportionate review. |
|---|---|
| ISSN: | 1470-2118 |