Objective: To analyze the early discriminative predictive information regarding the immunophenotype components of patients with sepsis, and its potential use as a prognosis tool.
Design: Observational prospective clinical study.
Setting: Intensive care unit (ICU) in a University Hospital.
Patients: Thirty-five patients admitted with severe sepsis.
Measurements: Analysis of peripheral blood on admission and 48 h later of the absolute white cell count and the immunophenotype of lymphocyte (CD3, CD3-HLADR, CD4, CD8, CD4/CD8 ratio, CD19, and CD25) and monocyte (CD13, CD13-HLADR, CD14, CD14-HLADR, CD13-CD14, and CD4) subpopulations.
Results: Due to its high correlation, the immunophenotypic profile studied at admission and 48 h later showed the same prognosis power regardless of the time of performance. The univariate analysis between groups (survival versus death) confirmed the prognostic significance of the total monocyte count and its subpopulations; significant differences were observed from the beginning only in the CD19 lymphocyte subpopulation. Multivariate analysis was performed using logistic regression with survival as the dependent variable. The final model comprised monocytes beta = 0.002 (P = 0.025) and CD13-HLADR beta = 0.016 (P = 0.029). The monocytes receiver operating characteristic (ROC) area obtained was 0.819 (confidence interval 0.663-0.976 at 95 %), the CD13-HLADR ROC area was 0.810 (confidence interval 0.658-0.963), and the monocytes + CD13-HLADR ROC area was 0.918 (confidence interval 0.807-1.000).
Conclusions: A single blood sample test obtaining the absolute monocyte and CD13-HLADR subpopulation count in the first days of admission could contribute to simplifying the classification of patients with severe sepsis into high- and low-mortality risk.