Background: Because acute cholecystitis has a different prognosis according to the degree of inflammation, early detection and prompt operation of severe cholecystitis are critical to the success of treatment. However, computed tomography (CT) has a low discriminative value for differentiating between simple and severe cholecystitis. Therefore, to enhance the diagnostic accuracy of CT scan, the imaging studies should be supplemented by preoperative clinical variables.
Methods: Patients undergoing laparoscopic cholecystectomy for simple and severe cholecystitis between 2007 and 2014 were compared. Severe cholecystitis included hemorrhagic, gangrenous, emphysematous, xanthogranulomatous, and perforated cholecystitis. Prediction models for severe cholecystitis were developed based on multivariate analyses of preoperative clinical and radiologic variables.
Results: Independent factors related with severe cholecystitis were age ≥65 years, male gender, body mass index (BMI) ≥25, serum leukocyte count ≥10,000/mm3, serum neutrophil fraction ≥80%, serum platelet count ≥20,000/mm3, serum alanine transaminase (ALT) level ≥40 IU/L, admission via the emergency department, and radiologic features of gallbladder wall thickening ≥4 mm, and presence of pericholecystic fluid collection (p < 0.05). A standard risk assessment scale (range: 0-77) for severe cholecystitis was developed based on the individual hazard rate of these variables. Patients scoring ≥28 on the risk assessment scale showed an 8.6 higher odds of severe cholecystitis than those scoring <28 (p < 0.01).
Conclusion: Standard and quick-and-easy predictive models for severe cholecystitis have been developed based on preoperative radiological and clinical variables, which is expected to help improve surgical outcome of patients with cholecystitis.
Keywords: CT scan; cholecystitis; complication; predictive model; prognosis.
Copyright © 2016. Published by Elsevier Taiwan.