Active‐control trials with binary data: A comparison of methods for testing superiority or non‐inferiority using the odds ratio

Abstract

This paper considers methods for testing for superiority or non-inferiority in active-control trials with binary data, when the relative treatment effect is expressed as an odds ratio. Three asymptotic tests for the log-odds ratio based on the unconditional binary likelihood are presented, namely the likelihood ratio, Wald and score tests. All three tests can be implemented straightforwardly in standard statistical software packages, as can the corresponding confidence intervals. Simulations indicate that the three alternatives are similar in terms of the Type I error, with values close to the nominal level. However, when the non-inferiority margin becomes large, the score test slightly exceeds the nominal level. In general, the highest power is obtained from the score test, although all three tests are similar and the observed differences in power are not of practical importance.

A Non-Inferiority, Individually Randomized Trial of Intermittent Screening and Treatment versus Intermittent Preventive Treatment in the Control of Malaria in Pregnancy

“…This resulted in a target sample size of 5000 women (1250 per center). Defining a non-inferiority margin a priori for LBW and placental malaria (PM) was complicated by the fact the exact risk of these outcomes by center was not known prior to the start of the trial; the odds ratio (OR) was used as a practical solution to this problem [ 21 ]. For LBW, we specified an OR based on a consensus among the investigators that a 4% risk difference in the prevalence of LBW would be clinically acceptable if the prevalence was 20% in the IPTp-SP group i.e.…”

Modern Issues and Methods in Biostatistics

“…Readers can find the relevant materials in the papers by Siqueira et al (2008) (with Comments by Brittain and Hu 2009). Wang and her colleagues (2001) discuss group sequential test strategies for superiority and noninferiority hypotheses.…”

Adjusting for covariates in non‐inferiority studies with margins defined as risk differences

“…The treatment effect in trials with binary outcomes is typically measured as a risk difference, odds ratio (OR) or relative risk. Various statistical methods for testing non‐inferiority based on the different measures of the treatment effect have been discussed in the statistical literature, with preference given to the use of the odds‐ratio parameterisation . However, in the review conducted earlier, the margin was predominantly defined as a risk difference, with 17 of 21 studies using a risk difference.…”

“…Various statistical methods for testing non-inferiority based on the different measures of the treatment effect have been discussed in the statistical literature, with preference given to the use of the † Supporting information may be found in the online version of this article. odds-ratio parameterisation [14,15]. However, in the review conducted earlier, the margin was predominantly defined as a risk difference, with 17 of 21 studies using a risk difference.…”