Treatment of irritable bowel syndrome in primary careBMJ 2008; 337 doi: https://doi.org/10.1136/bmj.a2213 (Published 13 November 2008) Cite this as: BMJ 2008;337:a2213
- Roger Jones, professor
Irritable bowel syndrome is a common condition with a community prevalence of 10-15% of the general population.1 2 The annual incidence in primary care is around 0.8%, and the prevalence of patients diagnosed in primary care is about 3-4%.3 The disorder is difficult to treat, hence the wide range of treatments used—dietary exclusion, fibre supplements, and probiotics; antispasmodic drugs, antidiarrhoeal agents, and laxatives; antidepressants, hypnotherapy, and cognitive behavioural therapy. This unusual spectrum of drug and non-drug treatments also highlights our ignorance about the cause of the condition. In the linked systematic review (doi:10.1136/bmj.a2313), Ford and colleagues summarise the effects of three different agents—fibre, antispasmodic drugs, and peppermint oil—in people with the syndrome.4
In the 1990s a range of new agents acting on 5-hydroxytryptamine type 3 and type 4 receptors in the enteric nervous system held considerable therapeutic promise. Most of them, however, failed to find a place in the routine drug treatment of irritable bowel syndrome because of lack of efficacy, serious adverse effects, or both. These disappointments added to the general scepticism about treating the syndrome with drugs, which is compounded by the high placebo response seen in therapeutic trials.
Such reservations are reflected in the neutral treatment advice given to patients by the American Gastroenterology Association,5 and the qualified suggestions for drug treatment given by CORE—the UK digestive diseases charity6—and the British Society for Gastroenterology.1 Complementary and alternative approaches feature strongly in the recently published National Institute for Health and Clinical Excellence (NICE) guidelines,7 with cognitive behavioural therapy and hypnotherapy suggested if symptoms persist beyond 12 months.
Ford and colleagues’ systematic review and meta-analysis includes data on more than 2500 patients with irritable bowel syndrome.4 At first glance their conclusions look like good news for patients (and prescribers), with numbers needed to treat (NNTs) for fibre, antispasmodics, and peppermint oil of 11.5, 5, and 2.5, respectively. However, as always, the devil is in the detail.
Although the trials of fibre show an overall benefit, analysis of the effect of different kinds of fibre shows that bran is not effective and that only ispaghula significantly reduces symptoms (NNT 6). However, the effect is no longer significant when only the highest quality studies are analysed. Nevertheless, these findings add support to NICE guidance,7 which advises against the use of insoluble fibre (such as bran) and recommends the use of soluble fibre (such as ispaghula).
The analysis of antispasmodic agents includes studies on a dozen different agents, with the most impressive therapeutic effects being shown for otilonium, cimetropium, hyoscine, and pinaverium. There was some evidence of publication bias, and the authors expressed reservations about the strength of their conclusions for otilonium and cimetropium because of heterogeneity between trials. The best evidence of efficacy was for hyoscine. Hyoscine butylbromide is an antimuscarinic agent extracted from the cork wood tree. It is not widely used in primary care in the United Kingdom at present—10 times more prescriptions are written for mebeverine than for hyoscine.8 Hyoscine is available without prescription from pharmacists in many countries, including the UK and United States.
Peppermint oil, which is also available without prescription, seems to be the most promising agent—NNT 2.5—although this figure was based on only four trials of fewer than 400 patients in total. However, secondary analysis of the three highest quality of these trials showed a similar treatment effect, with little heterogeneity between trials. None of these agents had significant adverse effects.
Limitations of the meta-analysis include the lack of information on the subtype of irritable bowel syndrome (constipation predominant, diarrhoea predominant, or alternating pattern), drug dosage, and patterns of administration. The analysis does not provide guidance on patient selection for particular agents on the basis—for example, of demographic factors, disease subtype, or clinical history, which limits the implementation of the findings.
It may be a little premature to follow the authors’ recommendation that national guidelines should be updated to include therapeutic guidance on these agents, but the results should reawaken an interest in the pharmacotherapy of irritable bowel syndrome and stimulate further research. Trials should have sufficient power and patients be better characterised so that predictors of response to treatment can be identified. There may also be a place for “N of 1 trials” in individual patients to determine individual therapeutic responses.9 None of these data, of course, invalidate the importance of making a “holistic” diagnosis in irritable bowel syndrome—that takes into account physical, psychological, and social factors—and of planning an integrated approach to treatment, which deals with all of these factors.10
Cite this as: BMJ 2008;337:a2213
Competing interests: None declared.
Provenance and peer review: Commissioned; not externally peer reviewed.