Letters to the Editor
To the Editor: Our cross-over study on the effects of probiotics (1) is not the first to choose a 2-week washout period (2–3). Others have also reported reversion to “normal” microbiota 2 weeks after discontinuing VSL#3 (4). Only half of our patients began the study on VSL#3; the other half began on placebo. Assuming that the probiotic's effect persisted in half of our patients over the entire 8 weeks on placebo, this would have obviously resulted in a favorable effect erroneously attributed to placebo. Thus, the efficacy of VSL#3 would have been greater than reported!
While the suitability of crossover design for studies on functional GI disorders is debatable, investigators continue to use this method for these studies (2–3,5–8). The crossover design allows (1) each subject to be its own control, minimizing enrollment bias and issues associated with randomization, and (2) a more manageable sample size with low variability compared with a simple double-blind design (1).
The nonparametric Wilcoxon signed rank test is preferred for small sample size, in contrast to the 2-way ANOVA that requires a homogeneous population with normal distribution. We used the Wilcoxon signed rank test to compare the magnitude of the change in score between study start and end of treatment for each subject. Each subject was evaluated and tested for all 5 endpoints to determine if the improvement reported during the use of VSL#3 was indeed significantly different from improvement reported during the use of placebo for the same subject. As for the “intention to treat” vs “per-protocol,” the 8 subjects excluded from the analysis had withdrawn from the trial during phase I (4 from placebo and 4 from treatment groups): 6 study subjects/parents were unwilling to complete the questionnaires and withdrew from the study before visit 2 (day 15); the other 2 subjects (1 from each study group) were withdrawn because they were not compliant. Obviously, these subjects did not even enter into the crossover phase, so including these patients as part of the intent to treat population would have been inappropriate for the purpose of this study. Hence, our choice of a “per-protocol” analysis is valid.
Our answer to the question “Does VSL#3 Really Improve Symptoms in Children with IBS?”: yes.
1. Guandalini S, Magazzu G, Chiaro A, et al. VSL#3 improves symptoms in children with irritable bowel syndrome: a multicenter, randomized, placebocontrolled, double-blind, crossover study. J Pediatr Gastroenterol Nutr
2. O'Sullivan MA, O’Morain CA. Bacterial supplementation in the irritable bowel syndrome. A randomised double-blind placebo-controlled crossover study. Dig Liver Dis
3. Silk DB, Davis A, Vulevic J, Tzortzis G, Gibson GR. Clinical trial: the effects of a trans-galactooligosaccharide prebiotic on faecal microbiota and symptoms in irritable bowel syndrome. Aliment Pharmacol Ther
4. Venturi A, Gionchetti P, Rizzello F, et al. Impact on the composition of the faecal flora by a new probiotic preparation: preliminary data on maintenance treatment of patients with ulcerative colitis. Aliment Pharmacol Ther
5. Klooker TK, Leliefeld KE, Van Den Wijngaard RM, Boeckxstaens GE. The cannabinoid receptor agonist delta-9-tetrahydrocannabinol does not affect visceral sensitivity to rectal distension in healthy volunteers and IBS patients. Neurogastroenterol Motil
6. Ligaarden SC, Axelsson L, Naterstad K, Lydersen S, Farup PG. A candidate probiotic with unfavourable effects in subjects with irritable bowel syndrome: a randomised controlled trial. BMC Gastroenterol. 2010; 10:16.
7. Ong DK, Mitchell SB, Barrett JS, et al. Manipulation of dietary short chain carbohydrates alters the pattern of gas production and genesis of symptoms in irritable bowel syndrome. J Gastroenterol Hepatol
8. Gonlachanvit S, Mahayosnond A, Kullavanijaya P. Effects of chili on postprandial gastrointestinal symptoms in diarrhoea predominant irritable bowel syndrome: evidence for capsaicin-sensitive visceral nociception hypersensitivity. Neurogastroenterol Motil