To determine whether preoperative platelet (PLT) count independently predicts in-hospital mortality in patients receiving venoarterial extracorporeal membrane oxygenation (VA-ECMO).

Secondary analysis of a prospectively maintained institutional registry.

A single academic medical center.

A total of 582 adult patients were supported with VA-ECMO between January 2015 and December 2019.

None.

Pre-cannulation PLT counts were used to categorize patients into four groups: G1, 100 × 10⁹/L or less; G2, greater than 100 × 10⁹/L to 200 × 10⁹/L; G3, greater than 200 × 10⁹/L to 300 × 10⁹/L; and G4, greater than 300 × 10⁹/L. The primary outcome was in-hospital mortality. Multivariable logistic regression, curve fitting analysis, and interaction analyses were performed. After adjustment, each 10 × 10⁹/L increase in PLT count was associated with a 4% lower odds of in-hospital mortality (adjusted odds ratio, 0.96; 95% confidence interval, 0.94-0.98; p = 0.0002). Compared with G1, the adjusted odds ratios for mortality were 0.51 for G2, 0.32 for G3, and 0.28 for G4 (p for trend < 0.0001). Subgroup analyses revealed significant interactions with diabetes and preoperative anticoagulation (p < 0.05), with a more pronounced predictive value in anticoagulated patients.

A lower preoperative PLT count is independently associated with increased in-hospital mortality in VA-ECMO patients. This parameter may aid in early risk stratification and help identify patients who could benefit from closer monitoring of hemostatic parameters.