Editor—We do not agree that our Cochrane review clearly shows that lactulose is ineffective. The fact that we found no evidence of effect does not imply that there is evidence of no effect.1 It is difficult to prove that a treatment has no effect.1
A survey of 989 abstracts of Cochrane reviews showed that inappropriate claims of no effect were made in 240 (22.5%) abstracts.2 In our review, we found that high quality trials found no significant effect of lactulose on the risk of no improvement of hepatic encephalopathy (relative risk 0.92, 95% confidence interval 0.42 to 2.04). The confidence interval indicates that we cannot exclude that lactulose may benefit (reduce the risk of no improvement by up to 58%). On the other hand, lactulose may also harm (increase the risk of no improvement by up to 104%). Our meta-analysis is based on only two trials with a total of 46 patients. Accordingly, our analysis has low power to detect clinically beneficial or harmful effect.
A consequence of claiming no effect would be that further research assessing the effect of lactulose is unnecessary. We find that a large, randomised, parallel, double blind trial is warranted. It would be interesting to compare lactulose, another laxative (magnesium or sorbitol) prepared to appear and taste like non-absorbable disaccharides, and a placebo of similar taste and appearance but without a cathartic effect (such as glucose).
We agree that lactulose should not be part of standard treatment of hepatic encephalopathy, and that placebo controlled trials are mandatory in this field, where none of the current treatments has proved clinically effective.
Competing interests: None declared.
References
- 1.Altman DG, Bland JM. Absence of evidence is not evidence of absence. BMJ 1995;311: 485. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Alderson P. Absence of evidence is not evidence of absence. BMJ 2004;328: 476-7. (28 February.) [DOI] [PMC free article] [PubMed] [Google Scholar]