Nonlinear relationship between platelet count and 30-day in-hospital mortality in ICU acute respiratory failure patients: a multicenter retrospective cohort study

Abstract Background Limited evidence exists regarding the link between platelet count and 30-day in-hospital mortality in acute respiratory failure (ARF) patients. Thus, this study aims to investigate this association among ICU patients experiencing acute respiratory failure. Methods We conducted a retrospective cohort study across multiple centers, utilizing data from the US eICU-CRD v2.0 database covering 22,262 patients with ARF in the ICU from 2014 to 2015. Our aim was to investigate the correlation between platelet count and 30-day in-hospital mortality using binary logistic regression, subgroup analyses, and smooth curve fitting. Results The 30-day in-hospital mortality rate was 19.73% (4393 out of 22,262), with a median platelet count of 213 × 10 9 /L. After adjusting for covariates, our analysis revealed an inverse association between platelet count and 30-day in-hospital mortality (OR = 0.99, 95% CI 0.99, 0.99). Subgroup analyses supported the robustness of these findings. Furthermore, a nonlinear relationship was identified between platelet count and 30-day in-hospital mortality, with the inflection point at 120 × 10 9 /L. Below the inflection point, the effect size (OR) was 0.89 (0.87, 0.91), indicating a significant association. However, beyond this point, the relationship was not statistically significant. Conclusion This study establishes a clear negative association between platelet count and 30-day in-hospital mortality among ICU patients with ARF. Furthermore, we have identified a nonlinear relationship with saturation effects, indicating that among ICU patients with acute respiratory failure, the lowest 30-day in-hospital mortality rate occurs when the baseline platelet count is approximately 120 × 10 9 /L.