Statistically Based Release Testing Alternatives for Uniformity of Dosage Units

Abstract This article provides a systematic framework to assess alternative testing approaches for USP <905 > uniformity of dosage units (UDU) for products formulated with a 100% label claim target. For illustration purposes several alternative release tests for UDU were evaluated, and these fall into three categories: (i) Tolerance Interval-based plans, (ii) CUDAL-ASTM E2709/E2810 plans, and (iii) Counting test-based plans. For consistency with the USP <905 > test, the same sample size per stage was employed: 10 dosage units at stage 1, and 30 dosage units at stage 2 (including the 10 units from stage 1). Also, in order to develop alternative tests that provide assurance of passing the USP <905 >, alternative UDU tests were designed to have an 80% probability of detecting a 20% failure rate of the USP <905 >. The 80/20 probabilistic matching point could be adjusted as part of a proposed risk-based decision framework.