Systemic inflammation response index as a prognostic predictor in patients with acute ischemic stroke: A propensity score matching analysis

Abstract

Background: Acute ischemic stroke (AIS), the most common type of stroke, is a major cause of morbidity and mortality worldwide. A growing number of studies have demonstrated that inflammation is a critical mechanism in AIS. Being an easily available and effective inflammatory marker, the systemic inflammation response index (SIRI) shows a high association with mortality in patients with cancer and intracerebral hemorrhage. In this study, we evaluated the potential prognostic role of SIRI in critically ill patients with AIS.

Methods: Clinic data were extracted from the Medical Information Mart data for the Intensive Care IV (MIMIC-IV) database. The optimal cutoff value of SIRI was determined by X-tile software. The primary outcome was the 90-day all-cause mortality, and the secondary outcomes were 30-day and 1-year all-cause mortality of patients with AIS. Cox proportional hazards regression analyses were used to assess the association between SIRI levels and all-cause mortality, and survival curves were estimated using the Kaplan-Meier method. Furthermore, a 1:1 propensity score matching (PSM) method was performed to balance the influence of potential confounding factors.

Results: A total of 2,043 patients were included in our study. X-tile software indicated that the optimal cutoff value of the SIRI for 90-day mortality was 4.57. After PSM, 444 pairs of score-matched patients were generated. Cox proportional hazard model showed that after adjusting for possible confounders, high SIRI level (≥4.57) was independently associated with the 90-day all-cause mortality in the cohort before PSM (HR = 1.56, 95% CI: 1.30-1.89, p < 0.001) and the PSM subset (HR = 1.47, 95% CI: 1.16-1.86, p = 0.001). The survival curves showed that patients with SIRI ≥4.57 had a significantly lower 90-day survival rate in the cohort before PSM (56.7 vs. 77.3%, p < 0.001) and the PSM subset (61.0 vs. 71.8%, p = 0.001). Consistently, AIS patients with high SIRI levels (≥4.57) presented a significantly high risk of 30-day and 1-year all-cause mortality before and after PSM.

Conclusion: A higher SIRI (≥4.57) was associated with a higher risk of 90-day, 30-day, and 1-year mortality and was an independent risk factor of mortality in patients with acute ischemic stroke.

Keywords: ischemic stroke; mortality; predictor; propensity score matching (PSM); systemic inflammation response index.

Figure 1

Workflow chart for the patient selection process.

Figure 2

Correlation coefficient between the variables. A color scale represents positive correlation (in blue) to negative correlation (in red). The size of the circle represents the absolute values of correlation coefficients.

Figure 3

The Kaplan–Meier survival plots of overall survival. A significantly lower 90-day (A), 30-day (B), and 1-year (C) survival rate was observed among patients in the higher SIRI (≥4.57) before PSM. It also showed significantly lower 90-day (D), 30-day (E), and 1-year (F) survival rates in the third and fourth SIRI quartiles compared with the first quartile in patients with AIS. After PSM, a similar trend in these outcomes was also found in 90-day (G), 30-day (H), and 1-year (I) overall survival. p-value was calculated by log-rank test and indicated in the plot. SIRI, systemic inflammation response index; PSM, propensity score matching.

Figure 4

The number of acute ischemic strokes in a different sequence of diagnosis.

Figure 5

The absolute standardized differences for variables used to match the two groups.

Figure 6

Subgroup analysis for the effect of SIRI on 90-day all-cause mortality in critically ill patients with AIS.

Figure 7

The time-dependent ROC analysis for prognostic value in patients with acute ischemic stroke. (A) ROC curves corresponding to 90-day all-cause mortality. (B) ROC curves corresponding to 30-day all-cause mortality. (C) ROC curves corresponding to 1-year all-cause mortality.