Diagnostic accuracy of tests suitable in primary care for acute diverticulitis: a systematic review and meta-analysis

Natasja Vijfschagt, Michiel De Boer, Marjolein Berger, Huib Burger, Gea Holtman

Keywords: acute diverticulitis, diagnostic accuracy, c-reactive protein, ultrasound

Background:

Timely recognition of acute diverticulitis is important, because of (a small) risk for severe complications. The differentiation between acute diverticulitis and other gastro-intestinal disorders is challenging as symptoms overlap. Clinical tests could assist the clinician with this diagnostic challenge. Previous reviews have focused on prognostic questions and imaging examinations in a secondary care setting.

Research questions:

What is the diagnostic accuracy of clinical tests suitable in primary care for acute diverticulitis in suspected patients?

Method:

We have systematically searched MEDLINE, EMBASE and Web of sciences for studies about the diagnostic accuracy of tests suitable in the primary care (signs, symptoms, laboratory tests and ultrasound) compared to a reference standard (computed tomography, colonoscopy, surgery findings (histopathology) and/or follow-up) in patients (>16 years old) suspected of acute diverticulitis. Three independent reviewers selected studies, extracted data, and assessed study quality with the QUADAS-2 tool. We have constructed dumbbell plots with corresponding likelihood ratio´s (LR) for signs and symptoms and meta-analyzed laboratory tests and ultrasound results.

Results:

Sixteen studies were included (N = 4.283), all studies were performed in secondary care with a median prevalence of 49% (IQR 40 - 59). Signs and symptoms showed a LR+ range of 0.46 – 6.47 and a LR- range of 0.22 – 1.81. C-reactive protein >10 mg/L (4 studies) had the best diagnostic accuracy of the laboratory tests with a pooled sensitivity and specificity of 0.95 (95%CI 0.89 - 0.98) and 0.39 (0.28 - 0.52) respectively. Ultrasound (11 studies) had a high pooled sensitivity and specificity of 0.92 (95%CI 0.85 - 0.96) and 0.95 (95%CI 0.89 - 0.98) respectively.

Conclusions:

None of individual signs and symptoms were informative. CRP showed potential for ruling out acute diverticulitis and ultrasound had good diagnostic accuracies. More research is needed about the diagnostic accuracy of these tests in patients suspected of acute diverticulitis in primary care.

Points for discussion:

How do you establish the diagnosis of acute diverticulitis at the GP’s office?

What do you think about the applicability of C-reactive protein?

Ultrasound showed a high diagnostic accuracy in our study, do you think ultrasound could be applicable in a primary care setting for acute diverticulitis or abdominal pain in general?

#127