Article type
Year
Abstract
Background: Non-inferiority (NI) randomized clinical trials (RCTs) commonly evaluate efficacy of new antiretroviral (ARV) drugs in ARV-naïve HIV patients. Their reporting and interpretation have not been systematically evaluated.
Objectives: To evaluate the reporting of NI RCTs of new ARV drugs in ARV-naïve HIV patients according to the CONSORT statement; to evaluate the degree of misinterpretation of RCTs when NI was inconclusive or not established.
Methods: Systematic review of NI RCTs evaluating drugs in ARV-naïve HIV patients. Pubmed, The Web of Science, and Scopus were reviewed until March 2011. Selection and extraction was independently done by two reviewers. Key reporting information included: similarity to prior RCTs of the active comparator, description of method to determine the NI margin, use of confidence interval (CI) method to interpret the primary outcome, use of blinding, and use of intention-to-treat (ITT) vs. per protocol (PP) statistical analysis. When NI was inconclusive or not established, we evaluated whether authors highlighted NI and distracted readers with secondary results.
Results: Fourteen RCTs were selected (range 71-3316 patients). None of the RCTs gave information about prior RCTs of the active comparator, and all used 2-sided CIs. All studies described the NI margin between 10% and 15%, but only 5 explained the method to determine it. Blinding was used in 4 studies, and the appropriate PP was the primary analysis in 4 studies. Five studies with NI inconclusive or not established highlighted NI or equivalence, and they distracted readers with positive secondary results in the abstract, results and conclusions.
Conclusions: There is poor reporting and interpretation of NI RCTs done in ARV-naïve HIV patients. Maximizing the reporting of the method of NI margin determination, the use of blinding and per-protocol analyses, and interpreting negative NI according to actual primary findings will improve the understanding and translation of results into clinical practice.
Objectives: To evaluate the reporting of NI RCTs of new ARV drugs in ARV-naïve HIV patients according to the CONSORT statement; to evaluate the degree of misinterpretation of RCTs when NI was inconclusive or not established.
Methods: Systematic review of NI RCTs evaluating drugs in ARV-naïve HIV patients. Pubmed, The Web of Science, and Scopus were reviewed until March 2011. Selection and extraction was independently done by two reviewers. Key reporting information included: similarity to prior RCTs of the active comparator, description of method to determine the NI margin, use of confidence interval (CI) method to interpret the primary outcome, use of blinding, and use of intention-to-treat (ITT) vs. per protocol (PP) statistical analysis. When NI was inconclusive or not established, we evaluated whether authors highlighted NI and distracted readers with secondary results.
Results: Fourteen RCTs were selected (range 71-3316 patients). None of the RCTs gave information about prior RCTs of the active comparator, and all used 2-sided CIs. All studies described the NI margin between 10% and 15%, but only 5 explained the method to determine it. Blinding was used in 4 studies, and the appropriate PP was the primary analysis in 4 studies. Five studies with NI inconclusive or not established highlighted NI or equivalence, and they distracted readers with positive secondary results in the abstract, results and conclusions.
Conclusions: There is poor reporting and interpretation of NI RCTs done in ARV-naïve HIV patients. Maximizing the reporting of the method of NI margin determination, the use of blinding and per-protocol analyses, and interpreting negative NI according to actual primary findings will improve the understanding and translation of results into clinical practice.