PURPOSE: Adjusted indirect comparisons can be used to investigate bioequivalence between generic products that are bioequivalent with a common reference product. In previous work with generic tuberculosis medicines prequalified by the WHO, it was observed that although indirect comparisons are an effective approach for confirming the interchangeability of generics, the approach is subject to less precision than direct comparisons. The objective of this investigation was to explore this by examining the influence of point estimates and power of bioequivalence studies versus the reference on the ability to show equivalence in indirect comparisons. METHODS: Power was considered as a determining factor instead of variability and sample size, because sample size is calculated based on variability and desired power. Scenarios were computed combining a range of point estimate differences (0-14 %) and statistical power of the studies (50-99.99 %). RESULTS: The indirect comparisons could conclude equivalence between generics only when (a) point estimate differences between generics were low (≤ 5.5 %) for any sufficiently powered study (> 80 %), or (b) the differences were large (but less than 14 %) and both bioequivalence studies were overpowered (e.g., 10 % difference and power ≥ 95 %). CONCLUSIONS: In summary, the ability to demonstrate interchangeability between generics is dependent not only on the real differences between the products but also on the design of the original generic vs. reference bioequivalence studies being combined, as earmarked by their respective power.
PURPOSE: Adjusted indirect comparisons can be used to investigate bioequivalence between generic products that are bioequivalent with a common reference product. In previous work with generic tuberculosis medicines prequalified by the WHO, it was observed that although indirect comparisons are an effective approach for confirming the interchangeability of generics, the approach is subject to less precision than direct comparisons. The objective of this investigation was to explore this by examining the influence of point estimates and power of bioequivalence studies versus the reference on the ability to show equivalence in indirect comparisons. METHODS: Power was considered as a determining factor instead of variability and sample size, because sample size is calculated based on variability and desired power. Scenarios were computed combining a range of point estimate differences (0-14 %) and statistical power of the studies (50-99.99 %). RESULTS: The indirect comparisons could conclude equivalence between generics only when (a) point estimate differences between generics were low (≤ 5.5 %) for any sufficiently powered study (> 80 %), or (b) the differences were large (but less than 14 %) and both bioequivalence studies were overpowered (e.g., 10 % difference and power ≥ 95 %). CONCLUSIONS: In summary, the ability to demonstrate interchangeability between generics is dependent not only on the real differences between the products but also on the design of the original generic vs. reference bioequivalence studies being combined, as earmarked by their respective power.
Authors: Marta Herranz; Susana Morales-Alcelay; Ma Teresa Corredera-Hernández; José María de la Torre-Alvarado; Antonio Blázquez-Pérez; Ma Luisa Suárez-Gea; Covadonga Alvarez; Alfredo García-Arieta Journal: Eur J Clin Pharmacol Date: 2012-11-30 Impact factor: 2.953
Authors: L Gwaza; J Gordon; J Welink; H Potthast; H Leufkens; M Stahl; A García-Arieta Journal: Clin Pharmacol Ther Date: 2014-07-02 Impact factor: 6.875