Introduction Small intestine bacterial overgrowth (SIBO) is characterised by diarrhoea and malabsorption. Identifying those at risk is key to diagnosis and treatment. We reviewed all glucose-hydrogen breath tests (GHBTs) performed for suspected SIBO, over a 6-year period in a single teaching hospital to identify associated risk factors and assessed the effectiveness of antibiotic treatment among those with a positive test.

Methods We collected data retrospectively for all GHBTs performed to investigate possible SIBO from 2006 to 2011. Demographic data and information concerning potential risk factors for SIBO were collected by review of clinic letters. A positive GHBT was defined as a rise of post-glucose end-tidal hydrogen reading > 20 parts per million from pre-dose baseline during the 2-hour 20 minute test period. Frequency of potential risk factors for SIBO among those with a positive GHBT compared to those with a negative GHBT were assessed using an odds ratio (OR) along with a 95% confidence interval (CI). Success of treatment with antibiotics for confirmed SIBO was judged according to patient report.

Results 316 patients underwent GHBT during the 6-year period. Of these, 17 were tertiary referrals and were excluded. Among the remaining 299 patients median age was 52 years (range:17–91) and 201 (66.9%) were female. 59 (19.7%) patients had a positive GHBT, 232 were negative, and 8 had equivocal results. Among these 59 patients median age was 61 years (range:20–91) and 39 (66.1%) were female. Of those with a positive test the principal indications for GHBT were diarrhoea in 35 (63.6%), diarrhoea and bloating in 9 (16.4%), high stoma output in 4, bloating alone in 3, abdominal pain in 3, and weight loss in 1. Presence of type II diabetes (OR 2.71; 95% CI 1.08–6.54) and previous intestinal surgery (OR 2.44; 95% CI 1.31–4.56) were significantly associated with a positive GHBT. Proton pump inhibitor (PPI) use (OR 0.98; 95% CI 0.49–1.88), previous radiotherapy (OR 0.84; 95% CI 0.30–2.09), presence of scleroderma (OR 1.27; 95% CI 0.39–3.51), opiate-use (OR 1.21; 95% CI 0.55–2.52), and presence of Crohn’s disease (OR 1.35; 95% CI 0.65–2.70) were not significantly associated with a positive GHBT.

In total, 43 patients with a positive GHBT received antibiotic treatment. Of these 21 (48.8%) improved, 18 (41.8%) had no improvement and in 4 the response to therapy could not be ascertained.

Conclusion Almost 20% of patients undergoing GHBT tested positive. Risk factors among our series of patients included type II diabetes and previous surgery. Interestingly, despite assertions from others, PPI use was not significantly associated with a positive GHBT. Almost 50% of patients with a positive GHBT responded to antibiotic treatment on clinical grounds.

Disclosure of Interest None Declared.