Herewith are presented the results of an investigation the statistical power of USP compendial release tests and recommended alternatives.
<br>The U.S. drug supply chain, formerly protected by a closed distribution network, is now threatened by the legal and illegal importation of drug products. Whereas quality can never be inspected into final products, compendial release standards may represent the only valid assessment that products of dubious origin would receive. Reliable tests for content uniformity and dissolution are required to protect the safety of the supply chain. A study was designed to test the hypothesis that existing compendial tests for content uniformity and dissolution would protect the supply chain against substandard and counterfeit drugs if basic field tests failed.
<br>Compendial tests for content uniformity and dissolution were evaluated for statistical power using simulation studies. The results revealed that the revised content uniformity test, based on tolerance analysis, was subject to an unacceptable level of consumers' risk. The Bergum method proved to be an excellent secondary standard for product assessment and is recommended as an alternative to the USP method.
Simulations with the USP dissolution test revealed significant weaknesses and inconsistencies in the test structure. Theoretical models and power assessments confirmed that the coverage specification of the dissolution test was an unacceptably high 50% coverage with 50% confidence.
<br>A Bayesian D-optimal design program was used to investigate alternative methods to improve the coverage capability of the USP dissolution test. The result of this program was the identification of two alternatives to the existing USP procedure. The first alternative is based on the addition of attribute coverage tests to stages 2 and 3 of the USP test, whereas the second alternative is based on the concept of tolerance analysis.
<br>Validation studies confirmed that both alternatives significantly improved the statistical power of the USP dissolution test without increasing the sample size or modifying the current three-stage procedure. The attribute test is non-parametric and behaves similarly to the existing USP with improved coverage, whereas the continuous alternative is more sensitive and is consistent with the recent revisions to the content uniformity test. / Mylan School of Pharmacy and the Graduate School of Pharmaceutical Sciences / Pharmaceutics / PhD / Dissertation
| Identifer | oai:union.ndltd.org:DUQUESNE/oai:digital.library.duq.edu:etd/154147 |
| Date | 29 June 2012 |
| Creators | Lunney, Phillip |
| Contributors | James K. Drennen, Carl A. Anderson, Ira S. Buckner, Khalid M. Kamal, Peter L. D. Wildfong |
| Source Sets | Duquesne University |
| Detected Language | English |
| Rights | Worldwide Access |
Page generated in 0.0019 seconds