A novel scoring system for predicting disease severity without CT imaging in acute diverticulitis.

Abstract

Purpose: Clinical scoring could help physicians identify patients with suspected acute diverticulitis who would benefit from further evaluation using computed tomography imaging. The aim of the study was to identify risk factors for complicated acute diverticulitis and create a risk score to predict disease severity in acute diverticulitis.

Methods: Patients diagnosed with CT-verified acute diverticulitis between 2015 and 2017 were included. Data on patients' clinical and laboratory findings and medical histories were collected retrospectively. Risk factors for complicated acute diverticulitis were identified using univariate and multivariate analyses. Continuous laboratory values were categorised by cut-off points determined using receiver operating characteristic (ROC) analysis. The Acute Diverticulitis Severity Score was formulated using logistic regression analysis.

Results: Of the total 513 patients included in the study, 449 (88%) had UAD, and 64 (12%) had CAD. Older age, significant comorbidities, C-reactive protein level, leucocyte count, vomiting, and body temperature were found to be independently associated with a higher risk for CAD. The novel Acute Diverticulitis Severity Score could reliably detect patients with CAD. The area under the ROC curve was 0.856 (p < 0.001) in discriminating disease severity. While higher scores indicate radiological studies, patients with low scores face an almost non-existent risk for complicated disease, making such studies possibly redundant.

Conclusions: The Acute Diverticulitis Severity Score accurately separated patients with uncomplicated disease from those at risk for complicated disease. This score can be applied in daily clinical practice to select patients requiring further investigation, consequently reducing healthcare costs and burdens.