The non-inferiority and superiority (NI/S) formulations to evaluate a new treatment are frequently used in active-controlled clinical trials. A key assumption for the NI/S statistical tests that compare 2 independent proportions is that corresponding critical regions are Barnard convex sets. This assumption allows significant reduction in the computation time required to calculate test sizes for these tests. This study presents arguments that both types of testing procedures (NI/S) require the corresponding critical regions to be Barnard convex sets. Otherwise, these tests may become meaningless. Notably, the critical regions of the well-known Blackwelder and Hauck-Anderson tests are not Barnard convex sets for many sample sizes.
ChenJTsongYKangS. Tests for equivalence or non-inferiority between two proportions. Drug Inf J. 2000;34:569–578.
2.
BarnardGA. Significance tests for 2x2 tables. Biometrika. 1947;34:123–138.
3.
Martín AndrésASilva MatoA. Choosing the optimal unconditioned test for comparing two independent proportions. Comput Stat Data Anal. 1994;17:555–574.
4.
Martin AndrésAHerranz TejedorI. Exact unconditional non-classics tests on the difference of two proportions. Comput Stat Data Anal. 2004;45:373–388.
5.
Silva MatoAMartín AndrésA. Optimal unconditional tables for comparing two independent proportions. Biom J. 1995;37:821–836.
6.
Silva MatoAMartín AndrésA. Simplifying the calculation of the P-value for Barnard’s test and its derivatives. Stat Comput. 1997,7:137–143.
7.
Herranz TejedorIMartín AndrésA. A numerical comparison of several unconditional exact tests in problems of equivalence based on the difference of proportions. J Stat Comput Simul. 2008;78:969–981.
8.
RöhmelJMansmannU. Unconditional non-asymptotic one sided tests for independent binomial proportions when the interest lies in showing non-inferiority and or superiority. Biom J. 1999;41:149–170.
9.
FrickH. Undominated p-values and property C for unconditional one-sided two sample binomial tests. Biom J. 2000;42:715–728.
10.
Almendra-AraoFSotres-RamosD. Some properties of non-inferiority tests for two independent probabilities. Commun Stat Theory Methods. 2012;41:1636–1646.
11.
RöhmelJMansmannU. Letter to the editor on Exact test of equivalence and efficacy with a non-zero lower bound for comparative studies. Stat Med. 1999;18:1734–1735.
12.
RöhmelJ. Statistical considerations of FDA and CPMP rules for the investigation of new anti-bacterial products. Stat Med. 2001,20:2561–2571.
13.
RöhmelJ. Problems with existing procedures to calculate exact unconditional p-values for non-inferiority/superiority and confidence intervals for two binomial and how to solve them. Biom J. 2005;47:37–47.
14.
Almendra-AraoF. A study of the classical asymptotic non-inferiority test for two binomial proportions. Drug Inf J. 2009;43:567–571.
15.
Almendra-AraoF. Behavior of the asymptotic pooled z-statistic. JP J Biostat. 2009;3:247–256.
16.
Almendra-AraoFSotres-RamosD. Comparison of some non-inferiority asymptotic tests for two independent proportions. Agrociencia. 2009;43:163–172.
17.
Almendra-AraoF. Barnard convex sets. Commun Stat Theory Methods. 2011;40:2574–2582.
18.
Almendra-AraoF. Efficient calculation of test sizes for non-inferiority. Comput Stat Data Anal. 2012;56:4138–4145.
19.
SkipkaGMunkAFreitagG. Unconditional exact tests for the difference of binomial probabilities—contrasted and compared. Comput Stat Data Anal. 2004;47:757–773.
20.
MunkASkipkaGStratmannB. Testing general hypotheses under binomial sampling: the two sample case—asymptotic theory and exact procedures. Comput Stat Data Anal. 2005;49:723–739.
21.
WellekS. Statistical methods for the analysis of two-arm non-inferiority trials with binary outcomes. Biom J. 2005;47:48–61.
