Utility of Ranson score, computed tomography severity index, and CRP criteria in risk stratification on the day of hospital admission in patients with acute pancreatitis: A cross-sectional analysis.

Abstract

The early identification of severe acute pancreatitis (AP) remains a great challenge in clinical practice and novel predictors are needed to complement available scoring systems. This study aimed to investigate utility of Ranson score, and computed tomography severity index (CTSI) and C-reactive protein (CRP) criteria in determination of risk prognostic status in AP. A total of 104 patients with AP (median age: 71.5 (range, 21-102) years, (59.6% were males) were included in this cross-sectional study. Patients were divided into two groups according to risk prognostic status including good prognosis (n=67) and poor prognosis (n=37) groups, based on presence of at least one of the poor prognostic criteria including Ranson score ≥3, presence of pseudocyst and necrotizing fluid collection on ultrasonography or computed tomography imaging and CRP levels >15 mg/L. Data on patient demographics, etiology of AP, smoking, blood biochemistry and hemogram findings and inflammatory markers including CRP (mg/L), mean platelet volume (fL), neutrophil-lymphocyte ratio, and platelet-lymphocyte ratio were recorded. Overall, 37 (35.6) patients with at least one these criteria comprised the poor prognosis group. Most of patients were considered to be in the poor prognosis group based on CTSI only (35.1%), CTSI + CRP (18.9%), and CTSI + Ranson (16.2%). Overall, 6 (5.8%) patients died, and all of them were in the poor prognosis group (p=0.002). Patients with poor versus good prognosis had significantly higher median (min-max) values for creatinine (1 [0.57-10.0] vs. 0.76 [0.5-8.4] mg/dL, p=0.004) and urea (48.0 [9.0-247.0] vs. 27.0 [10.0-111.0] mg/dL, p<0.001), and lower albumin values (3.5 [2.4-4.3] vs. 3.6 [2.7-4.6] g/L, p=0.021). Kappa values indicated presence of a moderate agreement between CTSI and CRP (kappa: 0.408), a fair agreement between CTSI and Ranson (kappa: 0.312), and a none to slight agreement between Ranson and CRP (kappa: 0.175). CTSI was able to discriminate all 6 patients (100.0%) with mortality, whereas Ranson and CRP each discriminated only 2 (33.3%) of 6 patients with mortality. Our findings suggest a stronger individual prognostic value of CTSI alone, rather than CRP or Ranson score alone, in risk stratification of AP patients for severity of disease and related mortality risk on the day of admission, whereas emphasize the likelihood of using CRP or Ranson score complementary to CTSI to enable further identification of poor prognostic status.