Multi-Biomarker Prediction Models for Multiple Infection Episodes Following Blunt Trauma.

Severe trauma predisposes patients to multiple independent infection episodes (MIIEs), leading to augmented morbidity and mortality. We developed a method to identify increased MIIE risk before clinical signs appear, which is fundamentally different from existing approaches entailing infections' detection after their establishment. Applying machine learning algorithms to genome-wide transcriptome data from 128 adult blunt trauma patients' (42 MIIE cases and 85 non-cases) leukocytes collected ≤48 hr of injury and ≥3 days before any infection, we constructed a 15-transcript and a 26-transcript multi-biomarker panel model with the least absolute shrinkage and selection operator (LASSO) and Elastic Net, respectively, which accurately predicted MIIE (Area Under Receiver Operating Characteristics Curve [AUROC] [95% confidence intervals, CI]: 0.90 [0.84-0.96] and 0.92 [0.86-0.96]) and significantly outperformed clinical models. Gene Ontology and network analyses found various pathways to be relevant. External validation found our model to be generalizable. Our unique precision medicine approach can be applied to a wide range of patient populations and outcomes.

Denver and Marshall scores successfully predict susceptibility to multiple independent infections in trauma patients.

Trauma patients are at risk of repeated hospital-acquired infections, however predictive scores aiming to identify susceptibility to such infections are lacking. The objective of this study was to investigate whether commonly employed disease-severity scores can successfully predict susceptibility to multiple independent infectious episodes (MIIEs) among trauma patients. A secondary analysis of data derived from the prospective, longitudinal study "Inflammation and the Host Response to Injury" ("Glue Grant") was performed. 1,665 trauma patients, older than 16, were included. Patients who died within seven days from the time of injury were excluded. Five commonly used disease-severity scores [Denver, Marshall, Acute Physiology and Chronic Health Evaluation II (APACHE II), Injury Severity Score (ISS), and New Injury Severity Score (NISS)] were examined as independent predictors of susceptibility to MIIEs. The latter was defined as two or more independent infectious episodes during the index hospital stay. Multivariable logistic regression was used for the statistical analysis. 22.58% of the population was found to be susceptible to MIIEs. Denver and Marshall scores were highly predictive of the MIIE status. For every 1-unit increase in the Denver or the Marshall score, there was a respective 15% (Odds Ratio:1.15; 95% CI: 1.07-1.24; p < 0.001) or 16% (Odds Ratio:1.16; 95% CI: 1.09-1.24; p < 0.001) increase in the odds of MIIE occurrence. APACHE II, ISS, and NISS were not independent predictors of susceptibility to MIIEs. In conclusion, the Denver and Marshall scores can reliably predict which trauma patients are prone to MIIEs, prior to any clinical sign of infection. Early identification of these individuals would potentially allow the implementation of rapid, personalized, preventative measures, thus improving patient outcomes and reducing healthcare costs.