Systematic evaluation of non-inferiority and equivalence randomized trials of anti-infective drugs

Abstract

The non-inferiority (NI) and equivalence (EQ) design is used widely in the clinical trials of anti-infective drugs, but still many arguments for, and against, conducting active control NI/EQ trials rather than simple placebo controlled trials. We searched Pubmed database and conducted a systematic literature review (1992–2011) to assess the methodological aspects of NI and EQ randomized trials of anti-infective drugs. A total of 335 publications with 337 trials were included. Of them, 235 trials reported a pre-specified margin of 10–15%. A proportion (e.g., cure, successful, failure) was used as the primary outcome in 316 trials (93.8%). Test treatments were non-inferior or equivalent to the control treatment in 325 trials (93.4%). The historical evidence for the effect of the control drug was specified in 38 trials (11.3%). For the literature of NI/EQ trials in anti-infective treatment, aspects that need improvement include the description of study participation, trial implementation, historical evidence and endpoint for the efficacy of control, inclusion of flow diagrams and figures that present margins and confidence intervals according to CONSORT criteria.

Keywords::

• equivalence
• infectious disease
• margin
• non-inferiority
• systematic review

Acknowledgement

The authors are grateful to JH Powers, G Li and the peer reviewers who has read the manuscript with great care and offered invaluable and informative suggestions. Authors’ contributions: Y Li read all of the articles, analyzed the data, wrote the manuscript, and made the figures and tables. Y He developed the idea, built the structure of this paper, read all of the articles, and co-wrote the manuscript. Y Sheng, K Wang, J Yang, L Li, J Huang, J Yang and J Wang helped analyzing data. Q Zheng inspired the idea and critically reviewed the manuscript. All authors read and approved the final manuscript.

Financial & competing interests disclosure

This study was supported by the National Science and Technology Supporting Projects of China (2008BAI51B03, 2012ZX09303009-001, 2012ZX09303-003), and the Project of the Shanghai Municipal Education Commission (085ZY1202, J50303, ZYSNXD-CC-LCPJ). The authors are grateful to Prof John H Powers, Dr Guofu Li and the peer reviewers who have read the manuscript with great care and offered invaluable and informative suggestions. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. The authors declare that they have no competing interests.

No writing assistance was utilized in the production of this manuscript.

Key issues

• • The non-inferiority (NI) and equivalence (EQ) design is used widely in the clinical trials of anti-infective drugs.

• • We systematically reviewed the NI and EQ randomized clinical trials of anti-infective drugs in order to evaluate the methodological aspects and reporting characteristics of these trials.

• • The available evidence suggests that there are still some aspects need improvement, like description of study participation, trial implementation, historical evidence and endpoint for the efficacy of control, inclusion of flow diagrams and figures that present margins and confidence intervals.

• • The evidence of the margin based should refer the previous placebo controlled trials and this process should be specified in reports. NI and EQ design is not suitable for the trials treating acute bacterial sinusitis, acute bacterial otitis media and acute bacterial exacerbation of chronic bronchitis, because a consistent and reliable estimate of the efficacy of active treatments relative to placebos has not been established.

• • Both intention-to-treat and per protocol analyses are required and are equally important when reaching a conclusion. Explicit statements about post-randomization exclusions should replace the ambiguous terminology of mITT.

Notes