It is crucial to identify high-risk patients with infectious conditions for appropriate management. We previously found that inflammatory markers added value to vital signs for predicting mortality in patients with suspected infection. In this study, the aim was to externally validate the added value of the inflammatory markers and to develop a new prediction model. For the external validation, consecutive adult patients with suspected infection admitted to the department of general medicine at two acute care hospitals were evaluated. A prognostic model for 30-day in-hospital mortality based on vital signs (systolic blood pressure, respiratory rate, and mental status) was compared with an extended model that also included four inflammatory markers (C-reactive protein, neutrophil-lymphocyte ratio, mean platelet volume, and red cell distribution width). Similar to the previous finding, all inflammatory markers except C-reactive protein showed significant contributions. Subsequently, a prediction model was developed using vital signs and markers with significant added value using a dataset that combined the external validation data with the data of the previous study. The new model was compared with a model based on the quick Sequential (sepsis-related) Organ Failure Assessment (qSOFA) score. The newly developed model showed a higher c-index than the qSOFA model [0.756 (95% CI 0.726-0.786) vs. 0.663 (0.630-0.696), p < 0.001]. Using the new model, 9.0% of patients who died were correctly reclassified compared with the qSOFA model at the threshold of 10% mortality risk. The new model including these markers showed potential to outperform the qSOFA model.

Download full-text PDF

Source
http://dx.doi.org/10.1007/s11739-024-03815-0DOI Listing

Publication Analysis

Top Keywords

inflammatory markers
20
vital signs
16
patients suspected
12
suspected infection
12
external validation
12
model
9
markers vital
8
signs predicting
8
predicting mortality
8
mortality patients
8

Similar Publications

Want AI Summaries of new PubMed Abstracts delivered to your In-box?

Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!