Prognostic implications of system inflammation response index in atrial fibrillation patients with type 2 diabetes mellitus

Figures

• 

  Fig. 1. Selection Flowchart for Patient Analysis: The MIMIC-IV database (version 2.2) holds detailed medical records of 299,712 patients covering 431,243 hospital admissions and 73,181 ICU stays. Within this comprehensive dataset, 4441 patients were diagnosed with both AF and T2DM. From these, individuals who met the exclusion criteria were carefully filtered out, resulting in a total of 2387 patients being removed from the study. The study ultimately included 2054 patients, of which 192 succumbed during the observation period.

• 

  Fig. 2. ROC Curves for Predictors of In-Hospital and One-Year Mortality in Critical Patients: (A) In-hospital mortality; (B) 1-year mortality. The red solid line traces the ROC curve for the SIRI with its AUC indicating a comparatively good predictive performance in this study. The DeLong test outcomes indicate a superior AUC for the SIRI compared to the MLR (in-hospital mortality: 0.731 vs. 0.669, p < 0.001; 1-year mortality: 0.715vs. 0.694, p = 0.007) and SII (in-hospital mortality: 0.731 vs. 0.651, p < 0.001; 1-year mortality: 0.715 vs. 0.653, p < 0.001). The AUC value of SIRI was superior to SAPSII (0.715 vs. 0.649, p < 0.001), SIRS (0.715 vs. 0.568, p < 0.001) and SOFA (0.715 vs. 0.625, p < 0.001) for predicting one-year mortality.

• 

  Fig. 3. Kaplan-Meier survival analysis curves for all-cause mortality. Footnote SIRI tertiles: Tertile1 < 2.591, Tertile2 2.591–5.701, Tertile3 > 5.701. Kaplan–Meier curves showing cumulative probability of all-cause mortality according to groups in hospital (A) and 1 year (B).

• 

  Fig. 4. Non-Linear Association Between SIRI and Mortality Shown via RCS Regression: (A) Hospitalization mortality; (B) 1-year follow-up mortality. The adjusted hazard ratios (HR) are denoted by solid blue lines, with 95% confidence intervals (CI) marked by blue dashed lines. The analysis demonstrates a non-linear correlation between SIRI and both hospitalization and one-year mortality rates. The hospital mortality risk map’s cutoff value stood at 7.336, while it was 4.53 for the one-year mortality.

• 

  Fig. 5. Forest Plot of In-Hospital and One-Year Mortality Hazard Ratios Across Subgroups: This plot displays hazard ratios for in-hospital (A) and one-year (B) mortality, segmented by various subgroups, and includes an interaction analysis probing the influence of SIRI on in-hospital mortality and one-year mortality within these diverse cohorts.