GROUP-BASED TRAJECTORY MODELING OF PLATELET IN PATIENTS WITH APLASTIC ANEMIA: A STUDY BASED ON THE MIMIC DATABASE

Abstract

Background: The main components that support coagulation and hemostasis are platelets. Nevertheless, there was no sufficient research done on how variations in platelet counts during hospital stays affect aplastic anemia (AA) patients' prognoses.

Objective: Using Group-based Trajectory Modeling (GBTM), this study proposed to evaluate the association between alterations in platelet levels and illness risk in patients with AA.

Methods: GBTM was used to group AA patients based on changes in platelet levels. Cox regression models were used to evaluate the relationship between platelet levels and the 30-day survival status of patients. Kaplan-Meier (K-M) survival curve analysis was used to assess the impact of platelet transfusion on survival among different trajectory groups of patients.

Results: Three trajectory patterns were recognized by GBTM: Class 1, Class 2, and Class 3. Even after controlling for confounding variables, the Cox risk estimates showed that AA patients had a higher chance of surviving in Class 1 (OR>1, P<0.05). Class 2 patients had the greatest survival, according to K-M (Log-rank P<0.001). According to landmark research, Class 1 patients' survival was not improved by platelet transfusion.

Conclusion: Patients with AA who had increasing platelet trajectories during their hospital stay had a higher 30-day survival rate; hence, patients with low platelet counts might not be good candidates for platelet transfusion treatment.