PHARMACEUTICAL483 SAMPLER pptx

*************************** PHARMACEUTICAL 483 SAMPLER *************************** FIVE 483 OBSERVATIONS ON PHARMACEUTICAL SYSTEMS THAT YOU SHOULD KNOW EDITED BY DAVE CORTES, PE 483 Summaries.com PHARMACEUTICAL SAMPLER All Rights Reserved. Copyright 2012 by Dave Cortes Cover Art 2012 Dave Cortes Cartoons 2012 Dave Cortes No original part of this publication may be reproduced or distributed in any form or by any means, electronic or mechanical, including photocopying, recording, or by any information or retrieval system, without the prior written permission of the editor. Full 483 Forms are publicly available through the Electronic Freedom of Information Act, fees may apply. For Information address: 17101 Costa Esmeralda Ceiba P.R. USA 00735 customer@483summaries.com Contents INTRODUCTION 5 1. WHY THIS BOOK IS DIFFERENT? 5 2. HOW THIS BOOK IS ORGANIZED 12 OBSERVATION #1: TRAINING 15 STATED OBSERVATION: 15 CODE COMMENTS 16 EDITOR COMMENTS: 17 OBSERVATION #2: FACILITES AND EQUIPMENT 19 STATED OBSERVATION: 19 CODE COMMENTS 20 EDITOR COMMENTS: 21 TIPS FOR JUSTIFICATIONS: 22 OBSERVATION #3: QUALITY SYSTEMS 23 STATED OBSERVATION: 23 CODE COMMENTS 24 EDITOR COMMENTS: 25 TIPS FOR JUSTIFICATIONS: 26 OBSERVATION #4: PRODUCTION SYSTEMS 27 STATED OBSERVATION: 27 CODE COMMENTS 28 EDITOR COMMENTS: 28 OBSERVATION #5: CLEANING VALIDATIONS 30 STATED OBSERVATION: 30 CODE COMMENTS 33 EDITOR COMMENTS: 34 INTRODUCTION 1. WHY THIS BOOK IS DIFFERENT? If today an FDA auditor comes to your operation site, there is a 54.1% that you will receive a dreaded 483-form full of non compliance observations. On 2010 the FDA issued 646 483’s to the pharmaceutical plants they visited. The question then becomes how to prepare or mitigate the probability of receiving a 483 observation? We believe that the best way to prepare for this occurrence is to gain experience from past examples of 483’s observations and use them as Business Cases Studies. Case studies illustrate psychological principles and Technical knowledge in a way that abstract generalizations and statistics cannot. In addition they construct a more detailed depiction of a problem and its boundaries better than any other method. Until now the Drug Manufacturing industry has relied on a descriptive process to assess its Regulatory Vulnerability. Specifically, the preferred way to evaluate the vulnerability of the drug manufacturing systems is by performing Audits, Assessment and Evaluations. These processes are done by “experts” that evaluate the different manufacturing systems and express an opinion on the individual system in question. After a typical one or two week assessment, the “expert” auditors produce an assessment report with a summary of good practices and gap finding. Given that there is a 54.1% chance of a pharmaceutical operation to receive a 483 Form, it seems this approach has a pretty bad performance record. But why would the Auditors predict so badly? Can it be improved? Behavioral Meta studies have shown that “expert judgments” are no more accurate than those of lightly trained novices and are less accurate that actuarial mathematical models. 1 This is a troublesome fact that seems to be a dirty little secret of quality departments. But let’s reflect on this methodology for a moment, using a physical metaphor. Imagine a Pharmaceutical Manufacturing System as a Three dimensional figure, and a regulatory audit as a cross sectional sample of this three- dimensional figure. 1 Camerer and Johnson, (1993), “The process-performance paradox in expert judgment”. The prior figure illustrates the concept, and shows two dimensional cross section samples (A,B, and C), taken from the three dimensional figure. As it can be seen in the figure, these cross sectional samples can give a good understanding of the complete figures, but only on the chosen path of the cross section. Note that they cannot explain the true nature of the three dimensional contour of the figures, as this will require the sum of an infinite number of cross sections to have a full understanding of the three dimensional figure. In the same way, due to practical constraints, a regulatory audit seeks to provide only reasonable assurance that the systems are free from material error, not the true status of the vulnerability of the system. This is a huge realization, as the TRILLION DOLLARS Pharmaceutical industry relies completely on a process that could only glimpse the true nature of the vulnerability of its systems. Empirically, this is the reason why many concurrent regulatory assessments diverge significantly from each other; they depend on limited cross sectional samples of the dynamical system. Compounding this fact is that, by design, the audits are path dependant exercises with completely random initial variables like auditor preference and expertise, current customer complaints or documentation samples. For example, a corporate auditor that knows or likes laboratory systems will focus on them, but will not choose to focus, or know how to focus, on the interdependency of the facility or the equipment. Another empirical example is that, it is typical to have corporate assessors produce similar audit results year after year, and then receiving an FDA assessment that diverges completely from the corporate assessment. This occurrence could give the perception of a lack of rigor from corporate assessments, but in reality is the systemic futility of attempting to define a three dimensional system by randomly selecting cross sections of the system. The insight of this introspection is that this is not the result of audit weakness, is more profound, it is a systemic flaw of the only tool available to diagnose Regulatory Vulnerability. We can go even farther and state that a descriptive vulnerability Audit can do more harm than good, as they provide a false sense of security with the illusion of understanding the non-linear relationships of a dynamical system. . *************************** PHARMACEUTICAL 483 SAMPLER *************************** FIVE 483 OBSERVATIONS ON PHARMACEUTICAL SYSTEMS THAT YOU SHOULD KNOW EDITED BY DAVE CORTES, PE 483 Summaries.com PHARMACEUTICAL SAMPLER All Rights