@article {Buoro330, author = {Sabrina Buoro and Barbara Manenti and Michela Seghezzi and Paola Dominoni and Tiziano Barbui and Arianna Ghirardi and Alessandra Carobbio and Gianmariano Marchesi and Ivano Riva and Alessandra Nasi and Cosimo Ottomano and Giuseppe Lippi}, title = {Innovative haematological parameters for early diagnosis of sepsis in adult patients admitted in intensive care unit}, volume = {71}, number = {4}, pages = {330--335}, year = {2018}, doi = {10.1136/jclinpath-2017-204643}, publisher = {BMJ Publishing Group}, abstract = {Aims This study was aimed to investigate the role of erythrocyte, platelet and reticulocyte (RET) parameters, measured by new haematological analyser Sysmex XN and C reactive protein (CRP), for early diagnosis of sepsis during intensive care unit (ICU) stay.Methods The study population consisted of 62 ICU patients, 21 of whom developed sepsis during ICU stay and 41 who did not. The performance for early diagnosing of sepsis was calculated as area under the curve (AUC) of receiver operating characteristics curves analysis.Results Compared with CRP (AUC 0.81), immature platelet fraction (IPF) (AUC 0.82) showed comparable efficiency for identifying the onset of sepsis. The association with the risk of developing sepsis during ICU stay was also assessed. One day before the onset of sepsis, a decreased of RET\% was significantly associated with the risk of developing sepsis (OR=0.35, 95\% CI 0.14 to 0.87), whereas an increased of IPF absolute value (IPF$\#$) was significantly associated with the risk of developing sepsis (OR=1.13, 95\% CI 1.03 to 1.24) 2 days before the onset of sepsis. The value of CRP was not predictive of sepsis at either time points.Conclusions IPF$\#$ and RET\% may provide valuable clinical information for predicting the risk of developing sepsis, thus allowing early management of patients before the onset of clinically evident systemic infections.}, issn = {0021-9746}, URL = {https://jcp.bmj.com/content/71/4/330}, eprint = {https://jcp.bmj.com/content/71/4/330.full.pdf}, journal = {Journal of Clinical Pathology} }