Predictive value of procalcitonin for diagnosis of infections in patients with chronic kidney disease: a comparison with traditional inflammatory markers C-reactive protein, white blood cell count, and neutrophil percentage.

Purpose: To investigate the predictive value of procalcitonin (PCT) in patients with stage 1-4 and stage 5 chronic kidney disease (CKD).
Methods: Five hundred and forty-one CKD inpatients were retrospectively analyzed and divided into CKD stage 1-4 (CKD1-4) and CKD stage 5 (CKD5) groups. Each group was further divided into non-infection, local infection, and sepsis subgroups. The clinical characteristics and inflammatory indexes of each subgroup including PCT, C-reactive protein (CRP), white blood cell count (WBC), and neutrophil percentage (N%) were compared, and the receiver operating characteristic curves to predict local infection and sepsis were plotted.
Results: Our research showed that the incidence and severity of infection in CKD5 group were significantly higher than those of CKD1-4 group; the baseline PCT level in CKD patients increased as renal function decreased and strongly correlated with CKD staging (r = 0.749); for CKD1-4 group, PCT, WBC, and N% could predict sepsis with the area under the curve (AUC) of 0.956, 0.854, and 0.917, respectively, but only CRP could predict local infection with AUC of 0.729, and for CKD5 group, only PCT and CRP could predict local infection with AUC of 0.715 and 0.780, respectively, and only PCT and N% could predict sepsis with AUC of 0.823 and 0.683, respectively.
Conclusions: The baseline PCT level of CKD patients is negatively correlated with renal function. In both CKD1-4 and CKD5 patients, the predictive value of PCT for local infection is not as good as that of CRP, while it has a significant advantage in predicting sepsis.