Abstract
BACKGROUND: Although suboptimal platelet transfusion (PT) response in critically ill patients with thrombocytopenia remains a challenge in clinical practice. This study aimed to investigate PT response during intensive care unit (ICU) stay among thrombocytopenic patients without underlying hematologic disease. METHODS: This retrospective single-center analysis included thrombocytopenic patients without primary hematologic disorders who received PT in ICU between June 2021 and December 2023. Clinical and laboratory variables were analyzed using a generalized linear mixed-effects model (GLMM), with the results visualized through a nomogram. The 28-day survival curves, stratified by receiving single or multiple PT episodes, were established using the Kaplan-Meier method. RESULTS: Suboptimal PT response was observed in 522 episodes (77.9%, 522/670) and in 291 patients (79.9%, 291/364). The GLMM identified sepsis, splenomegaly, mechanical ventilation, higher APACHE II score, and longer time interval of post-PT platelet count as independent predictors of suboptimal response, while higher white blood cell count at ICU admission and the PT episode number in ICU were independently protective. A nomogram based on these seven variables demonstrated good predictive performance. Suboptimal PT episodes were associated with higher red blood cell and fresh frozen plasma requirements. The 28-day survival probability was significantly higher in the single transfusion group with optimal response versus the suboptimal response. CONCLUSIONS: Repeat PT may enhance the PT response and survival. Suboptimal PT response was associated with increased RBC and FFP transfusion requirements. The established nomogram demonstrated strong predictive accuracy and may provide a practical tool for optimizing PT practices in the ICU.