Presepsin and Procalcitonin Levels as Markers of Adverse Postoperative Complications and Mortality in Cardiac Surgery Patients

2019 
: Backgound: This study was aimed at evaluating the presepsin and procalcitonin levels to predict adverse postoperative complications and mortality in cardiac surgery patients. METHODS: A total of 122 cardiac surgery patients were enrolled for the study. Presepsin and procalcitonin levels were measured 48 h after the procedure. The primary endpoints were adverse renal, respiratory, and cardiovascular outcomes and mortality. RESULTS: Presepsin and procalcitonin levels were significantly higher in patients with adverse renal and respiratory outcome (p < 0.001 and 0.0081). The presepsin levels were significantly higher in patients with adverse cardiovascular outcome (p = 0.023) and the procalcitonin values in patients with sepsis (p = 0.0013). Presepsin levels were significantly higher in patients who died during hospitalization (382 pg/mL, interquartile range [IQR] 243-717.5 vs. 1,848 pg/mL, IQR 998-5,451.5, p = 0.049). In addition, the predictive value for in-hospital, 30-days, and 6-months mortality was higher for presepsin, with a significant difference between the 2 biomarkers (p = 0.025, p = 0.035, p = 0.003; respectively). Presepsin and procalcitonin seem to have comparable predictive value for adverse renal, cardiovascular, and respiratory outcome in cardiac surgery patients. Although a positive trend was notable for presepsin and adverse renal outcome (area under the ROC [receiver operating characteristic] curves [AUC] of 0.760, 95% CI 0.673-0.833 versus procalcitonin: AUC 0.692; 95% CI 0.601-0.773): no statistically significant difference was evident between the AUC of the 2 biomarkers (p = 0.25). CONCLUSIONS: Presepsin and -procalcitonin seem to have comparable predictive value for -adverse renal, cardiovascular, and respiratory outcome in cardiac surgery patients. Also, presepsin possesses a better predictive value for in-hospital, 30-days, and 6-months mortality.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    0
    References
    10
    Citations
    NaN
    KQI
    []