Use of a predictive equation for diagnosis of acute gangrenous cholecystitis

Abstract

Background: Factors previously identified by multivariate logistic regression that were predictive for gangrenous cholecystitis (GC) were used to develop a predictive equation. Our objective was to evaluate the sensitivity, specificity, and positive (PPV) and negative (NPV) predictive values of this equation for detecting GC in patients with acute cholecystitis (AC).

Methods: Medical records of patients who presented to a tertiary care hospital with AC were reviewed. Twenty-one patient and clinical variables were recorded. We prospectively tested the results of the following equation against pathologic diagnosis: P=e((0.7116+0.9944.DM+1.7157.WBC-1.0319.ALT.2.0518.ALP+2.7078.PCF))/(1+e([-0.7116+0.9944.DM+1.7157.WBC-1.0319.ALT-2.0518.ALP+2.7078.PCF])), where P = predicted value; DM = diabetes mellitus; WBC = white blood cell count; ALT = alanine aminotransferase; AST = aspartate aminotransferase; and PCF = pericholecystic fluid.

Results: Ninety-eight patients presented with AC and 18% had GC (18 of 98). Using a cutoff of P = 0.724, our equation had a specificity of 93%, sensitivity of 83%, PPV of 71%, and NPV of 96%, P <0.001 for the detection of GC.

Conclusions: Our study demonstrates the equation may be useful in detecting the subset of AC patients who have GC.