In vitro dissolution similarity has been suggested as a surrogate for assessing equivalence between the pre-changed and post-changed formulations for postapproval changes of a drug. The difference factor f1, based on the absolute mean difference, has been proposed as a criterion for evaluating similarity between dissolution profiles. Statistical properties including density function, bias, and asymptotic distribution of a consistent estimator are investigated. Due to complexity of the distribution of the estimator, we suggest the use of the confidence intervals obtained from the bootstrap method for evaluation of dissolution similarity. A simulation was conducted to examine the size and power of the proposed Cl procedure. Comparisons with other criteria such as similarity factor are also provided. Numerical examples are used to illustrate the proposed Cl procedure.
|頁（從 - 到）||1089-1113|
|期刊||Communications in Statistics - Theory and Methods|
|出版狀態||Published - 2000 十二月 1|
All Science Journal Classification (ASJC) codes