evaluate its efficacy. Informations regarding the diagnostic and prognostic role of PCT in the critically ill subject are lacking. Aim of this study was to evaluate the value of PCT in the diagnosis of bacterial infections and its prognostic weight in the unstable patient. Methods: we enrolled 1063 consecutive, critically ill subjects admitted to our Internal Subintensive Medicine Department in the period 2008-2010, evaluating age, sex, haemodynamic parameters, blood exams, PCT and blood cultures. In particular we collected the absolute number of peripheral white blood cells, troponin I (TnI) and serum creatinine. Among hemodynamic parameters, we evaluated the presence of shock (defined as systolic blood pressure < 90 mmHg, low urine output, < 0.5 ml/kg/h and reduced cutaneous perfusion). The degree of severity of the pathology was assessed with the SAPS-II score. We’ve set as outcomes mortality or intensive therapy transfer, overall survival and length of hospital admission. Results: We observed positive cultures in 375 subjects, whose mean PCT levels were significantly higher than in patients without positive cultures (0,84 ng/ml versus 0,20 ng/ml; p<0,0001). ROC curve analysis, however, underlined a sub-optimal role of PCT in predicting bacterial isolation (AUC:0,58;95% CI: 0,54-0,62). 172 patients died, and their mean PCT values were significantly higher than survivors (2,62 ng/ml versus 0,17 ng/ ml; p<0,05). Calculating an optimal cutoff of 0,50 ng/ml, mean time without events among subjects with lower PCT was 44 days compared with 26 days observed in patient with high PCT. The prognostic weight of PCT was superior than TnI, and PCT was correlated with prognosis independently from bacterial infection. A model including SAPS2, troponin I and PCT had a good AUC (0,734; 95%CI: 0,667-0,775) in predicting in-hospital events, such as death or transfer to intensive-care unit. Discussion: Procalcitonin is now a widely used marker of bacterial infection, despite several papers confirmed its low diagnostic yeld. In our population, we confirmed this observation by showing a low correlation between increased PCT and positive bacterial isolates (blood, sputum or urine cultures). However, an increased PCT value was associated, independently from the presence of sepsis or infection, to an increased rate of adverse events and a lower event-free survival. When combined in a model including clinical and serologic markers of severity, such as SAPS2 and troponin I, PCT showed a good prognostic value for in-hospital death or complications by increasing the AUC of the model. This derived model was able to predict adverse event in all the critically ill patients enrolled, independently from the cause of admission. Conclusion: Among critically ill patients PCT has a low diagnostic yeld, with a poor predictive value for positive bacterial cultures. PCT, however, maintains a good prognostic significance in predicting both event-free survival and adverse events during the hospitalization. In particular, PCT increases the prognostic value of commonly used scores, such as SAPS2, and well-recognised prognostic markers, as troponin I. A model including SAPS2, troponin I and PCT is highly predictive for in-hospital adverse events among critically ill patients, and could be used to predict the risk of in-hospital mortality and the probability of transfer to intensive therapy unit.
The role of procalcitonin in emergency medicine
Falsetti L
Writing – Original Draft Preparation
;
2017
Abstract
evaluate its efficacy. Informations regarding the diagnostic and prognostic role of PCT in the critically ill subject are lacking. Aim of this study was to evaluate the value of PCT in the diagnosis of bacterial infections and its prognostic weight in the unstable patient. Methods: we enrolled 1063 consecutive, critically ill subjects admitted to our Internal Subintensive Medicine Department in the period 2008-2010, evaluating age, sex, haemodynamic parameters, blood exams, PCT and blood cultures. In particular we collected the absolute number of peripheral white blood cells, troponin I (TnI) and serum creatinine. Among hemodynamic parameters, we evaluated the presence of shock (defined as systolic blood pressure < 90 mmHg, low urine output, < 0.5 ml/kg/h and reduced cutaneous perfusion). The degree of severity of the pathology was assessed with the SAPS-II score. We’ve set as outcomes mortality or intensive therapy transfer, overall survival and length of hospital admission. Results: We observed positive cultures in 375 subjects, whose mean PCT levels were significantly higher than in patients without positive cultures (0,84 ng/ml versus 0,20 ng/ml; p<0,0001). ROC curve analysis, however, underlined a sub-optimal role of PCT in predicting bacterial isolation (AUC:0,58;95% CI: 0,54-0,62). 172 patients died, and their mean PCT values were significantly higher than survivors (2,62 ng/ml versus 0,17 ng/ ml; p<0,05). Calculating an optimal cutoff of 0,50 ng/ml, mean time without events among subjects with lower PCT was 44 days compared with 26 days observed in patient with high PCT. The prognostic weight of PCT was superior than TnI, and PCT was correlated with prognosis independently from bacterial infection. A model including SAPS2, troponin I and PCT had a good AUC (0,734; 95%CI: 0,667-0,775) in predicting in-hospital events, such as death or transfer to intensive-care unit. Discussion: Procalcitonin is now a widely used marker of bacterial infection, despite several papers confirmed its low diagnostic yeld. In our population, we confirmed this observation by showing a low correlation between increased PCT and positive bacterial isolates (blood, sputum or urine cultures). However, an increased PCT value was associated, independently from the presence of sepsis or infection, to an increased rate of adverse events and a lower event-free survival. When combined in a model including clinical and serologic markers of severity, such as SAPS2 and troponin I, PCT showed a good prognostic value for in-hospital death or complications by increasing the AUC of the model. This derived model was able to predict adverse event in all the critically ill patients enrolled, independently from the cause of admission. Conclusion: Among critically ill patients PCT has a low diagnostic yeld, with a poor predictive value for positive bacterial cultures. PCT, however, maintains a good prognostic significance in predicting both event-free survival and adverse events during the hospitalization. In particular, PCT increases the prognostic value of commonly used scores, such as SAPS2, and well-recognised prognostic markers, as troponin I. A model including SAPS2, troponin I and PCT is highly predictive for in-hospital adverse events among critically ill patients, and could be used to predict the risk of in-hospital mortality and the probability of transfer to intensive therapy unit.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.