Assumptionsofnormality, additivityoftheusedstatistical model and the homogeneity of the variance are of pivotal importance to most statistical procedures used in the assessment of bioequivalence (BE).With regard to the normality assumption, most international guidelines require that such data be transformed to its natural logarithm prior to any statistical treatment. It is claimed that transformation would either improves or restores the distribution of the random subject and random error terms to normal. This study clearly demonstrate that LnT has invariably failed to achieve the desirable impact on the normality distribution of BE data, thus it should be undertaken with extreme caution and adequate justification.
биоэквивалентность, логарифмическое преобразование, bioequivalence, logarithmtransform
1. D. J. Schuirmann, J. Pharmacokinet / Biopharm. - 1987. - Vol. 15, № 6. - C. 657-680
2. M. I. Evgen'ev, S. Yu. Garmonov, N. S. Shitova V. I. Pogorel'cev / Vestnik Kazanskogo tehnologicheskogo universiteta. - 2004. - №1-2. - S. 74-81.
3. Ministerstvo zdravoohraneniya i social'nogo razvitiya RF, Federal'naya sluzhba po nadzoru v sfere zdravoohraneniya i social'nogo razvitiya. Ocenka bioekvivalentnosti lekarstvennyh sredstv: Metodicheskie ukazaniya. Moskva: FGU «Nauchnyy centr ekspertizy sredstv medicinskogo primeneniya». - 2008. - 33 C.
4. G. G. Islamova, T. Z. Lygina, A. M. Gubaydullina / Vestnik Kazanskogo tehnologicheskogo universiteta. - 2010. - №8. - C. 257-262
5. V. M. Zaycev, V. G. Liflyandskiy, V. I. Marinkin, Prikladnaya medicinskaya statistika: Uchebnoe posobie. Spb.: FOLIANT, 2003, 432 S.
6. J. P. Lui and C. S. Weng / Commun Stat-Theor M. - 1994. - Vol. 23. - S. 421-434
