Background: Irritable bowel syndrome (IBS) is a complex, heterogeneous disease which can be challenging to diagnose. No study has identified and assessed the accuracy of all available methods of diagnosing IBS.

Aim: To conduct a systematic review of the literature to identify and assess accuracy of symptom-based diagnostic criteria, biomarkers, psychological markers or combinations thereof.

Methods: MEDLINE, EMBASE and EMBASE Classic were searched (until April 2015) to identify studies reporting accuracy of available methods to diagnose IBS in adult populations. Eligible studies assessed accuracy of these diagnostic tests against an accepted reference standard. Data were extracted to calculate positive and negative likelihood ratios, with 95% confidence intervals (CIs), of the diagnostic test utilised. Where more than one study used the same test, data were pooled in a meta-analysis.

Results: Twenty-two studies (7106 patients) were eligible. Positive and negative likelihood ratios of the current gold standard, the Rome III criteria, were 3.35 (95% CI: 2.97-3.79) and 0.39 (95% CI: 0.34-0.46), similar to other symptom-based criteria. Eleven biomarkers performed no better than symptom-based criteria. Psychological markers performed well in one study. Five different combinations were assessed. The best in terms of positive likelihood ratio was faecal calprotectin, intestinal permeability and Rome I criteria (26.4; 95% CI: 11.4-61.9), and in terms of negative likelihood ratio serum-based biomarkers and psychological markers (0.18; 95% CI: 0.12-0.25).

Conclusions: Symptom-based diagnostic criteria, biomarkers and psychological markers performed modestly in predicting IBS. Combining symptoms with markers appears more effective, and may represent the way forward in the diagnosis of IBS.