Vries, Willemijn de (2025) Impact of Cardiovascular Factors on PPG: A Non-Inferior Alternative to NEWS for Predicting Sepsis-Related Deterioration. thesis, Medicine.
Full text available on request.Abstract
Background: Sepsis is a dysregulated host response to infection and a leading cause of global mortality. Hemodynamic instability can lead to septic shock, making early risk identification crucial for better outcomes. Photoplethysmography (PPG) detects vascular changes, offering potential for identifying vascular alterations associated with clinical deterioration. This study aims to assess the impact of cardiovascular medication on the PPG signal and its implications for deterioration prediction. Methods: This post-hoc analysis used prospectively collected Acutelines biobank data. Patients with suspected infection and high-quality PPG recordings were included. Linear regression assessed associations between cardiovascular medication dosages and PPG features, while logistic regression and receiver operating curves evaluated predictive model performance. Results: A total of 576 patients were included. Several PPG features, including pulse width (PW), perfusion index (PI), and crest time (CT), were significantly associated with diuretic and beta-blocker dosages. Adjusting the PPG-based sepsis deterioration model for age, cardiovascular disease (CVD), gender, CCI-socre, and beta-blocker/diuretic dose resulted in a model that performed non-inferior to the National Early Warning Score (NEWS) (area under the curve (AUC) (0.749 vs. 0.759; p=0.810)). Conclusion: Continuous waveform analysis using PPG demonstrated comparable performance to the NEWS score for predicting early sepsis deterioration during triage at the emergency department (ED). Presenting a new alternative option for detecting early signs of sepsis-related deterioration, offering continuous and non-invasive monitoring.
| Item Type: | Thesis (UNSPECIFIED) |
|---|---|
| Supervisor name: | Bouma, prof. dr. H.R. and Horst, drs. S. Ter |
| Faculty: | Medical Sciences |
| Date Deposited: | 22 May 2026 12:28 |
| Last Modified: | 22 May 2026 12:28 |
| URI: | https://umcg.studenttheses.ub.rug.nl/id/eprint/3958 |
Actions (login required)
 |
View Item |