Inter-observer reliability of computed tomographic classifications of diverticulitis.

Area of Science:

• Radiology
• Gastroenterology
• Surgical Pathology

Background:

• Computed tomography (CT) is widely used for diagnosing diverticulitis.
• New CT-based classifications for diverticulitis have emerged, but their reliability requires evaluation.

Purpose of the Study:

• To assess the inter-observer agreement among radiologists for different diverticulitis CT classifications.
• To compare the reproducibility of the modified Hinchey, Ambrosetti, and Dharmarajan classifications.

Main Methods:

• 100 CT examinations of diverticulitis patients were independently reviewed by three blinded radiologists.
• Inter-observer agreement was calculated using kappa values for the modified Hinchey, Ambrosetti, and Dharmarajan classifications.
• Agreement with a reference standard was assessed for the modified Hinchey and Ambrosetti classifications.

Main Results:

• The Ambrosetti classification showed almost perfect inter-observer agreement (median kappa 0.83).
• The Dharmarajan (median kappa 0.76) and modified Hinchey (median kappa 0.72) classifications demonstrated substantial inter-observer agreement.
• Both Ambrosetti and modified Hinchey classifications had substantial agreement with the reference standard.

Conclusions:

• The Ambrosetti classification is the most reproducible among the evaluated systems.
• The Ambrosetti and modified Hinchey classifications are reliable for diagnosing diverticulitis.
• The Dharmarajan classification serves as a valuable complementary tool for complicated diverticulitis cases.