ESTABLISHING EQUIVALENCE OF 2 TREATMENTS AND SAMPLE-SIZE REQUIREMENTSIN MATCHED-PAIRS DESIGN

Statistical methods for testing the null hypothesis of a nonzero diffe rence between two treatments and the sample size determination for mat ched-pairs studies are investigated. A Wald-type test proposed by Lu a nd Bean (1995, Statistics in Medicine 14, 1831-1839) is anticonservati ve, i.e., its false positive error rate is greater than specified. Sco re method and normal deviate based on a restricted maximum likelihood estimation are presented. These two test statistics are shown to be al gebraically equal. Their actual type I error probabilities are satisfa ctorily close to a nominal level. Numerical examinations demonstrate t hat the sample size formulas using these alternative methods are reaso nable while that by Lu and Bean is not. The efficiency of matching in equivalence studies is positively related to intercorrelation or the k appa coefficient of agreements. We recommend the score or hit methods to establish equivalence of two treatments for individually matched sa mples.