Skip navigation
FACT
Focus on Alternative and Complementary Therapies

Does publication bias work against complementary medicine papers?

Abbot NC, Ernst E
Department of Complementary Medicine, University of Exeter, 25 Victoria Park Road, Exeter, UK

Objective

To determine whether journal reviewers in orthodox medicine are more likely to reject complementary medicine papers reporting positive rather than negative results, and to assess the effect of scientific quality on the reviewer decision.

Methods and results

200 corresponding authors of research papers on vascular disease were selected from MEDLINE 1994–6, and assigned by block randomisation to four groups. Each group was sent a different version of the same fictitious short communication: a good quality (G) and a poor quality (P) version, each reporting either positive (+) or negative result (-). Reviewers marked the paper on a VAS score sheet. 125 reviewer sheets were received back. The P+ version had a significantly (P<0.05) higher rejection score (mean 5.2, 95% CI 3.7 to 6.8) than the G+ version (3.1, 1.8 to 4.4), but there were no significant differences between any other versions. In terms of quantity, significantly more P+ were rejected than G+ (55% vs 16%; χ2, P<0.05). Combining data from the positive and negative studies to examine the effect of quality alone, poor versions were significantly more likely (χ2, P<0.01) to be rejected than good versions (rejection rate 50% vs. 23% respectively). By contrast, pooling the results by outcome rather than quality gave no significant difference between the negative and positive versions.

Conclusion

The scientific quality of manuscript was found to be more important than the direction of the results in this study, an encouraging result for the refereeing process.

Top | Next: Uncovering adverse effects of complementary medicine (CM)»
© Pharmaceutical Press 2009
Accessibility | Terms and Conditions