| Summary: | 碩士 === 國立成功大學 === 統計學系碩博士班 === 90 === The Food and Drug Administration(FDA) draft guidance " Bioavailability and Bioequivalence Studies for Nasal Aerosols and Nasal Sprays for Local Action" (www.fda.gov/cder/) has recommended chi-square differences for in vitro profile comparison procedure. However, statistical properties of the method based on chi-square differences are not clear and not thoroughly investigated. Furthermore, the proposal in the draft guidance suggests the use of 95% upper confidence limits of the empirical sampling distribution for the conclusion of bioequivalence. The upper limit of this method can not determined since the function form of parameters in the hypothesis is unknown.
The objective of this paper is to provide a statistical assessment of in vitro profile bioequivalence and refers the drawback of FDA draft guidance.
Since the form of the criterion is complex, some statistical properties can be derived such as the asymptotic expectation and the asymptotic variance through the Taylor's expansion of chi-square difference at population proportions. But they are very complicated.
Then, some alternative procedures are proposed to assess the in vitro profile bioequivalence. Finally, a simulation study is conducted to investigate its empirical distribution. The FDA recommended criterion and several criteria proposed by Tsong(2000) are also compared.
|
|---|
