Positive Health Online

Abstract

ZIMPEL and WINDELER, Abteilung Medizinische Biometrie, Universitatsklinikum Heidelberg, Heidelberg, Germany investigated whether a publication bias could be shown for the results of medical theses on all kinds of complementary therapies and what factors influencing publication could be identified.

Background

The problem of 'publication bias' has been highlighted in recent years as a threat to the validity of meta-analyses and systematic literature reviews . Successful strategies for dealing with this problem have yet to be established.

Methodology

140 medical theses for the years 1982-1992 on all kinds of complementary therapies were identified and all were included in the study. Data from the theses were analysed. Publications arising from the theses were searched for on MEDLINE and by personal communication with the authors and authors' supervisors. Factors that might influence the likelihood of publication were identified by bivariate analysis and logistic regression analysis.

Results

Publications were found for 53 (37.9% ) of the 140 medical theses examined. The main factors influencing the likelihood of publication were: 'positive result' (odds ratio 2.337); 'high-level statistical analysis' (odds ratio 1.483); 'supervisors with a high publication output' (odds ratio 1.477); and 'candidates of younger age' (odds ratio 0.691).

Conclusion

As has been found for other medical subjects, there was an apparent publication bias for complementary medicine, with positive results having a greater chance of publication than negative ones.

References

Zimpel T and Windeler J. (Publications of dissertations on unconventional medical therapy and diagnosis procedures – a contribution to 'publication bias'.) Forschende Komplementaermedizin und Klassische Naturheilkunde 7 (2): 71-4. Apr 2000.