engleski

Neutrophil-lymphocyte ratio, hematocrit, blood urea nitrogen and red-cell distribution width in the early prediction of clinically relevant outcomes in patients with acute pancreatitis

Introduction: Various clinical and biochemical factors have been evaluated as potential predictors of severity in acute pancreatitis (AP). Early detection of severe cases may prompt more intensive treatment measures resulting in better clinical outcomes. Aims & Methods: To investigate the usefulness of hematocrit (HTC), neutrophil-limphocyte ratio (NLR), blood urea nitrogen (BUN), and red cell distribution width (RDW) measured on admission in prediction of peri/pancreatic necrosis, organ failure (OF), local complications (LC), and mortality in AP. A retrospective analysis of prospectively collected data was performed on 382 patients with AP admitted consecutively at a tertiary healthcare facility from January 2014 to December 2016. Patients were diagnosed with AP based on the 2012 Revised Atlanta criteria. Patients with chronic pancreatitis or pain lasting for >72 hours before admission were excluded. Pearson's rank correlation test was performed. Receiver operating curves (ROC) were constructed to calculate sensitivity and specifity for outcomes with significant correlations. Results: We showed significant correlations between NLR and all outcomes, mortality (r=0.147 ; P=0.019), OF (r=0.185 ; P=0.004), necrosis (r=0.179 ; P=0.007), al LC (r=0.129 ; P=0.044). HTC showed significant correlations with necrosis (r=0.137 ; P=0.011) and LC (r=0.246 ; P<0.0001). Significant correlations of RDW were established with mortality and OF (r=0.121 ; P=0.019) and (r=0.126 ; P=0.018), respectively. Strong correlations were shown for BUN and mortality (r=0.294 ; P<0.0001) and OF (r=0.329 ; P<0.0001). With a cut-off value for HTC>0.41 we showed for necrosis and LC an AUC 0.597 (P=0.016 ; Se 69% ; Sp 50%) and AUC 0.650 (P<0.0001 ; Se 68% ; Sp 58%), respectively. BUN>9 showed for mortality AUC 0.819 (P<0.0001 ; Se 59% ; Sp 85%) and OF AUC 0.765 (P<0.0001 ; Se 51% ; Sp 87%). Conclusion: Most of the variables measured at admission showed a significant correlation with clinically important outcomes. When specifically analysed as single diagnostic tools they have high degree of false positive results and are not reliable in prediction of severe cases by themselves. Use of several different might achieve better accuracy in early detection of pts at risk of unfavorable clinical outcomes.

acute pancreatitis ; clinical outcomes

nije evidentirano