Immature platelet fraction in predicting sepsis in critically ill patients

Intensive Care Med. 2013 Apr;39(4):636-43. doi: 10.1007/s00134-012-2725-7. Epub 2012 Oct 24.


Purpose: To establish whether in critically ill patients without sepsis at intensive care unit (ICU) admission the percentage immature platelet fraction (IPF%) is a cellular marker predicting sepsis to verify a possible correlation between IPF% changes and manifest sepsis and describe the IPF% time course after ICU admission.

Methods: Prospective, observational 7-day study of 64 adult patients admitted to a general ICU at a University Hospital with no sepsis criteria. We measured daily IPF%, procalcitonin (PCT), C-reactive protein, platelets, white blood cell count and coagulation variables. Thirty-one patients with sepsis at ICU admission were studied as controls.

Results: The only variable we tested at ICU admission that predicted sepsis was plasma IPF% (p < 0.001; >4.7 %: sensitivity 56.2 % IC 37.7-73.6; specificity 90.0 % IC 73.4-97.8). IPF% and PCT values were higher for the patients who had sepsis at admission and during the study than in patients in whom sepsis never developed (IPF%: p = 0.017; PCT: p = 0.030). Among the outcome variables, logistic regression was identified as the only variable related to the development of sepsis, IPF% (r = 0.51; p = 0.004). In patients who developed sepsis IPF% was inversely correlated with platelet count (r = -0.60; p < 0.001) and had high values before sepsis became manifest, decreasing significantly on the 2nd day thereafter.

Conclusions: In patients without sepsis at ICU admission IPF% increases before sepsis becomes manifest. Measuring IPF% through an easily available technology can therefore provide an early cellular marker predicting the development of sepsis.

MeSH terms

  • Biomarkers / blood
  • C-Reactive Protein / analysis*
  • Calcitonin / blood*
  • Calcitonin Gene-Related Peptide
  • Critical Illness
  • Early Diagnosis
  • Female
  • Humans
  • Intensive Care Units / statistics & numerical data
  • Leukocyte Count*
  • Logistic Models
  • Male
  • Middle Aged
  • Platelet Count*
  • Predictive Value of Tests
  • Prognosis
  • Prospective Studies
  • Protein Precursors / blood*
  • Sepsis / blood
  • Sepsis / diagnosis*
  • Thrombocytopenia / diagnosis
  • Thrombocytopenia / etiology*


  • Biomarkers
  • CALCA protein, human
  • Protein Precursors
  • Calcitonin
  • C-Reactive Protein
  • Calcitonin Gene-Related Peptide