Systemic inflammation response index predicts short-term mortality in intensive care unit patients with bacterial bloodstream infections.

BACKGROUND

Bloodstream infections (BSI) in intensive care units (ICU) are associated with high mortality, requiring reliable prognostic tools. The systemic inflammatory response index (SIRI) may be a novel biomarker that may help prediction of outcomes in these patients. This study aims to evaluate the prognostic value of SIRI in predicting mortality and discharge outcomes in ICU patients with bloodstream infection (BSI).

METHODS

This retrospective study included 199 adult patients diagnosed with bloodstream infections at Sakarya University Training and Research Hospital between 2018 and 2022. SIRI and other clinical parameters were calculated at the onset of BSI. The receiver operating characteristic (ROC) was used to determine the optimal cut-off SIRI value to predict the mortality. The multivariate regression analysis analyzed the correlation between mortality and SIRI.

RESULTS

The median SIRI (mSIRI) value at the onset of BSI was 5003. Notably, higher SIRI levels were significantly associated with an increased risk of both 14-day and 30-day mortality (p = 0.001 and p = 0.011, respectively). In the multivariate analysis, SIRI ≥ 5003 demonstrated an independent predictive value for 14-day mortality (p = 0.02). Furthermore, logistic regression analysis revealed that higher initial CRP and SIRI levels, elevated neutrophil counts, and the presence of COVID-19 or malignancy were independently associated with increased mortality risk.

CONCLUSION

Our results suggest higher SIRI levels are independently associated with the BSI and 14-day mortality. SIRI is a valid and cost-effective prognostic marker of mortality in ICU patients with bloodstream infections. Its use in clinical practice can improve risk stratification and guide therapeutic decisions. Further studies are needed to validate these results in different populations. RANDOMIZED CONTROLLED TRIAL NUMBER: Clinical trial number: not applicable. Since this study has a retrospective design, it is not possible to assign a Randomized Controlled Trial number to it.