gms | German Medical Science

MAINZ//2011: 56. GMDS-Jahrestagung und 6. DGEpi-Jahrestagung

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie e. V.
Deutsche Gesellschaft für Epidemiologie e. V.

26. - 29.09.2011 in Mainz

Is there a danger of “biocreep” with non-inferiority trials?

Meeting Abstract

Search Medline for

  • Primrose Beryl - Uniklinikum Freiburg, Freiburg
  • Werner Vach - Uniklinikum Freiburg

Mainz//2011. 56. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie (gmds), 6. Jahrestagung der Deutschen Gesellschaft für Epidemiologie (DGEpi). Mainz, 26.-29.09.2011. Düsseldorf: German Medical Science GMS Publishing House; 2011. Doc11gmds080

DOI: 10.3205/11gmds080, URN: urn:nbn:de:0183-11gmds0804

Published: September 20, 2011

© 2011 Beryl et al.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by-nc-nd/3.0/deed.en). You are free: to Share – to copy, distribute and transmit the work, provided the original author and source are credited.


Outline

Text

Introduction: Non-inferiority(NI) trials test a hypothesis that a new treatment is not inferior to standard treatment. Bio-creep basically refers to the cyclical phenomenon where a slightly inferior treatment becomes the active control for the next generation of NI trials which over time leads to degradation of the efficacy of the investigational treatment. We studied the effect estimates from an unselected set of all the registered non-inferiority trials conducted within a seven-year period. The aim was to determine the pre- and post trial distribution of the true effect in NI trials from this data using meta analytic methods and simulations.

Methods: We did a search for all non-inferiority trials registered in the National Institute of Health’s Clinical trials register (www.ClinicalTrials.gov) which were carried out between January 2000 and December 2007. Trials studying non-inferiority of efficacy as the primary objective were only included. We did a search for information regarding the primary results from these trials in three steps: the NLM website, the International Federation of Pharmaceutical Manufacturers and Associations (IFPMA) website and Pubmed. Web-based search engines and personal communication were also used. Using the retrieved study results, a descriptive and exploratory analysis of the study characteristics and a meta-analysis of the effect estimates were performed using STATA 11.

Results: Of the 114 registered NI trials, 84 met the inclusion and exclusion criteria. Information about these trials could be traced for 77(92%) studies with the help of: NLM website-40, IFPMA-7, Pharmaceutical website-6, Pubmed-20 and others-4. The final results were available for 69 of the 72 completed studies. According to the published reports the new treatment was found to be superior in 7(14%), non-inferior in 55(79%) and inferior in 7(9%) of the trials. The effect estimate was positive among 39 (59%) of the 66 studies providing an effect estimate. We intend to derive the distribution of true effect estimates of NI trials and to present this at the conference.

Conclusion: We found a very high likelihood of retrieving results from registered clinical trials making it possible to calculate the prestudy distribution of the true effect in non-inferiority trials. The unanticipated finding of a positive average effect estimate suggests that a decline in standard treatment effect (biocreep) is not imminent, at least on average. However, the intimidating risk of approval of treatments with true negative effects reiterates the need for a careful choice of the margin in NI trials.


References

1.
Statistics in Medicineolume. 2010;29(27):2769-80.