Physiological Phenotypes in Comatose ICU Patients: A Retrospective Multidimensional Analysis.

Background/Objectives: Comatose patients admitted to intensive care represent a heterogeneous critical population, in which prognosis can be influenced by the interaction between inflammation, coagulation, tissue damage and respiratory dysfunction. This study aimed to identify physiological phenotypes in comatose ICU patients and evaluate their association with ICU outcome. Methods: A retrospective observational study was performed on 227 adult comatose patients admitted to the ICU. Clinical variables, inflammatory and coagulation markers, LDH, blood gas parameters, APACHE and SOFA scores and ICU outcome were analyzed. Spearman correlations, hierarchical cluster analysis, Kruskal-Wallis tests, Dunn post hoc, effect sizes and adjusted logistic regression were used. Results: ICU mortality was 65.6%. Cluster analysis suggested three tentative phenotypes: inflammatory-coagulopathic, relatively stable, and a severe respiratory-inflammatory profile. Differences between clusters were mainly determined by D-dimers, CRP, procalcitonin, INR, and LDH, while APACHE and SOFA scores did not differ significantly between phenotypes. ICU outcome differed significantly between clusters (χ2 = 18.370; p < 0.001; Cramer's V = 0.284), with the maximum mortality in Cluster 3. After adjustment, cluster membership did not remain an independent predictor of mortality. Conclusions: Comatose patients admitted to the ICU can be classified into exploratory physiological phenotypes, associated with different clinical outcomes in the unadjusted analysis. These results support the utility of multidimensional biological assessment but require external validation before clinical application.