+ All Categories
Home > Health & Medicine > Good Laboratory practice and current good manufacturing practice

Good Laboratory practice and current good manufacturing practice

Date post: 13-Jul-2015
Category:
Upload: postgradomlcc
View: 297 times
Download: 15 times
Share this document with a friend
Popular Tags:
132
A primer Good laboratory practice and current good manufacturing practice For analytical laboratories
Transcript
Page 1: Good Laboratory practice and current good manufacturing practice

A primerGood laboratory practice and current good manufacturing practice

Foranalytical

laboratories

Page 2: Good Laboratory practice and current good manufacturing practice
Page 3: Good Laboratory practice and current good manufacturing practice

A primerGood laboratory practice and current good manufacturing practice

Ludwig Huber

Agilent Technologies Deutschland GmbHHewlett-Packard-Straße 8 76337 [email protected]

Foranalytical

laboratories

Page 4: Good Laboratory practice and current good manufacturing practice

Copyright © 2000-2002 Agilent Technologies Publication number 5988-6197ENPrinted in Germany 04/02

Page 5: Good Laboratory practice and current good manufacturing practice

III

FDA guidelines, industry validation groups

and reference books help toimplement compliance in

laboratories.

Good Laboratory Practice (GLP) deals with the organization,process and conditions under which laboratory studies areplanned, performed, monitored, recorded and reported.GLP data are intended to promote the quality and validityof test data.

(Current) Good Manufacturing Practice (cGMP) is thatpart of quality assurance which ensures that products areconsistently produced and controlled to the qualitystandards appropriate to their intended use.

Published GLP and cGMP regulations have a significantimpact on the daily operation of an analytical laboratory.Weller1 gave an excellent practical explanation on what isexpected from working in a regulated environment:

“If experimental work is conducted in compliance with

GLP, with or without the aid of computer, it should be

possible for an inspector, maybe four or five years hence,

to look at the records of the work and determine easily

why, how and by whom the work was done, who was in

control, what equipment was used, the results obtained,

any problems that were encountered and how they were

overcome”.

Unfortunately most laboratories have been in situations where they have had to interpret the regulationsthemselves. Procedures have been developed on an ad hoc basis, in isolation, in response to inspections by boththeir company’s Quality Assurance Unit (QAU) andregulatory bodies.

Preface

Good Laboratory Practice and current Good Manufacturing Practice

Page 6: Good Laboratory practice and current good manufacturing practice

IV

Preface

This situation has somewhat changed over the last coupleof years.

1.Regulatory agencies such as the United Stated Food or Drug Administration (FDA) and internationalorganizations such as the International Conference on Harmonization (ICH) have developed guidancedocuments for implementation.

2.Companies have formed validation groups anddeveloped procedures for qualification and validation ofequipment, computers and analytical methods

3.Validation reference books have been published withclear guidelines, checklists and Standard OperatingProcedures (SOPs) on how to validate and qualifycomputerized systems and other equipment andmethods used in analytical laboratories2,3.

In addition, some instrument vendors help users ofequipment and methods to comply with regulations. A good example is Agilent Technologies. Over the last ten years, at Agilent we have gained a goodunderstanding about the impact of regulations onanalytical laboratories. We also share this informationwith users of our equipment.

In 1993 and 1994 we published the first and secondeditions of this primer. It has been translated into tenlanguages and more than 50,000 copies have beendistributed. Simultaneously we started to develop anddeliver SOPs and services for the Installation andOperational Qualification (IQ/OQ) of our analyticalproducts.

The development and introduction of the 1100 SeriesHPLC was a breakthrough in validation & compliance. As a result of our experience and knowledge and of thetechnology available we were able to design automatedcalibration and validation features into the product.

The 1100 Series HPLC hardware and software was

designed for validation & compliance.

The GLP/GMP primer has been translated into 10 languages and more

than 100,000 copies have been distributed.

Page 7: Good Laboratory practice and current good manufacturing practice

V

The product was introduced with software for automatedverification of the installation and for operationalqualification. A validation binder or CD-ROM has sincebeen made available to proof documented evidence ofvalidation during development. The QualificationWorkbook was later added to document all validationactivities.

The results of several surveys made by LC/GC Magazineshow Agilent Technologies, formerly a part of Hewlett-Packard, as the number-one supplier for validation.

Over the last few years we received many requests toupdate the GLP/GMP primer with news on regulations andguidelines. Our original plan was to remove GLP and GMPbasics. However, surveys amongst the target audienceshowed that there is still a need for such information,especially for new employees. Therefore the first twochapters are dedicated to the GLP/GMP background andbasics. Additional requests were made for more specificson equipment qualification, computer validation andelectronic records & signatures. Chapters three to eightdeal with this. Chapter nine discusses possible vendorcontributions and gives examples.

Regulatory requirements, inspection and enforcementpractices are quite dynamic. What is appropriate todaymay not need to be appropriate tomorrow. Regulationschange but more often it is the inspection practices thatchange. In the early 90’s the focus of inspections was onbasic requirements of GLP and GMP, but then it changedto equipment hardware and later on to software andcomputer systems. Today, the clear focus is on datasecurity, traceability and integrity of electronic records,driven mainly but not only by FDA’s regulation 21 CFR Part 11.

Good Laboratory Practice and current Good Manufacturing Practice

Agilent is ranked as the preferred supplier

for validation of hardware,software methods

and data.

The focus of FDA inspections has changed

from equipment hardware tosoftware and now to data traceability, integrity and

security.

Page 8: Good Laboratory practice and current good manufacturing practice

VI

Preface

Paper documents are difficult to update. Therefore wehave decided to update this primer regularly on theInternet. In addition we would recommend the followingimportant sites where readers can get up-to-dateinformation.

www.agilent.com/chem/validation

www.fda.gov

US FDA website. You can find regulations, guidancedocuments and warning letters.

www.labcompliance.com

Website dedicated to regulatory and compliance inlaboratories. It includes many links to other websites andhas an open discussion forum. It also includes referenceliterature and other documents for download.

To comply with regulations can be quite expensive andsometimes it is just impossible to comply 100% even whenwilling, especially when new regulations are released. An example is 21 CFR Part 11 (electronic signatures &records) which was released in 1997 and nobody complied100% at that time.

On-line resourcesthrough the internetkeep you up-to-date,

day-by-day!

Page 9: Good Laboratory practice and current good manufacturing practice

VII

Good Laboratory Practice and current Good Manufacturing Practice

The challenge is to find a good compromise between notdoing enough and doing too much. Let’s take validation asan example. When complying right at the beginning of thevalidation process the additional value to each validationstep is tremendous. However, there is no added value intrying to validate each and every step and the incrementalcosts for validation goes up with each validation effort.The question is: ‘where is the optimum’ or ‘how muchvalidation is enough’. The challenge is to find the optimumand this requires a thorough risk analysis. With the help ofthis primer and listed references it is hoped that thereader will get enough guideline to find this optimum forhis or her specific process.

Ludwig HuberAgilent Technologies, April 2002

Page 10: Good Laboratory practice and current good manufacturing practice

VIII

Chapter 1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3Historical perspective (GLP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

Fraud and misinterpreted data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4FDA’s reaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5EPA’s reaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5GMP and cGMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

National and international GLP regulations . . . . . . . . . . . . . . . . . . . . . . . 7Memoranda of Understanding (MOU) and bilateral agreements . . . . . . . . . . . . . . . . . . 7

Who has to comply with GLP/cGMP regulations? . . . . . . . . . . . . . . . . 7Non-clinical laboratory studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8Analysis under GMP and cGMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8Availability of regulations and guidance documents . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

Chapter 2 GLP/GMP key provisions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11GLP organization and conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12Study director . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13Quality assurance unit (QAU) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14Standard operating procedures (SOPs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

Title page . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17Location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17Deviations and changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17Level of detail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17Language . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

Reagents and solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18Expiration date . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18Storage conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

Test and control articles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19Characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19Storage container . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

Raw data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20The laboratory notebook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20Transcriptions to computers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21Direct data capture by a computer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21Modification of raw data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21Integrity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

Storage and archiving . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22Facilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22Time period . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22Who has access? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23SOPs and operator manuals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23How should electronic records be stored? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23Subcontractors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

Personnel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24Documentation of qualification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24Part-time or full-time personnel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24Health precautions and safety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

Contents

Part 1Introduction to GLP/cGMP

basics

Page 11: Good Laboratory practice and current good manufacturing practice

IX

How to conduct a GLP study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25Identify speciments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25Meet data recording requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25Changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25Automated data collection systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

Enforcement of GLP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26Frequency of inspections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

Good Manufacturing Practice (GMP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27GMP inspections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28Guide to inspection of pharmaceutical quality control . . . . . . . . . . . . . . . . . . . . . . . . . 28Guide to investigating out of specifications (OOS) test results . . . . . . . . . . . . . . . . . . 29

Surviving an audit and inspection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29Conduct . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30Close . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

FDA 483 observations and warning letters . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30Inspection trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

FDA cGMP notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

Chapter 3 Validation overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35What is validation? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

Validation, versus qualification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

What has to be validated? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37Analysis method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38Analytical system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38Personnel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38Reference standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

When should validation be done? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39Steps Towards Equipment Validation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

Validation master plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39Validation committee . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40Inventory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40Design qualification (DQ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40Qualification of the vendor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40Installation qualification (IQ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41Operational qualification (OQ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41Performance qualification (PQ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41Data validation for consistency, security, integrity and traceability . . . . . . . . . . . . . . 41Validation report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41Individual validation plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

Useful resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

Good Laboratory Practice and current Good Manufacturing Practice

Part 2Impact of GLP/cGMP in the

analytical laboratory

Page 12: Good Laboratory practice and current good manufacturing practice

X

Chapter 4 Design qualification (DQ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45Setting the specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

Steps for design qualification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46Additional steps for computer systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47Steps for existing systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47DQ for instruments used for different applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47Help from instrument vendors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

The role of the vendor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49Tasks of the vendor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49Qualification of the vendor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

Chapter 5 Installation qualification (IQ)and operational qualification (OQ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53Steps for installation qualification (IQ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

Before installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54During installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54Additional steps for computer systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

Equipment inventory data base . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56Line between IQ and OQ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56Operational qualification (OQ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

Steps for OQ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57Additional steps for computer systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57Computer networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58Validation of home made programs and spreadsheets . . . . . . . . . . . . . . . . . . . . . . . . . 58OQ for existing systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

OQ discussions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59Selection of tests and acceptance criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59Module versus system test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60Frequency of tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60Test sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60Who can or should test the vendor? the user? a third party? . . . . . . . . . . . . . . . . . . . 61Preventive maintenance before the OQ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61OQ after repair . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61OQ after instrument move . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61Why should I do OQ at all, isn’t PQ enough? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

Chapter 6 Method validation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63Suitability for intended application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

Validation parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

Selectivity (specificity) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65Precision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67Accuracy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67Linearity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68Range . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68Limit of detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69Limit of quantitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69Ruggedness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

Contents

Page 13: Good Laboratory practice and current good manufacturing practice

XI

Robustness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69Strategy for validation of methods? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70When is method revalidation required? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73Validation report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74Bioanalytical methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

Chapter 7 On-going performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77Overview and importance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79Steps for PQ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

Additional steps for computer systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

Tests for PQ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79System suitability parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80Quality control samples and QC charts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81Additional tests? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

Chapter 8 Data security, integrity, traceability . . . . . . . . . . . . . . . . . . 8321 CFR Part 11 – electronic records and signatures . . . . . . . . . . . . . . . . . . 84

Who has to comply . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84Primary requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85System validation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86Secure retention of electronic records to instantly reconstruct the analysis . . . . . . 86Limited system access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87User independent computer generated, time-stamped audit trail . . . . . . . . . . . . . . . . 88Use of secure electronic signatures for closed and open systems . . . . . . . . . . . . . . . 89Training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89Hybrid systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89

Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90

Chapter 9 Vendor contribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95Importance of the vendor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96Help during design qualification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97Assistance for vendor qualification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98Contribution at installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99Help for operational qualification and requalification . . . . . . . . . . . . . . . 100Help for method validation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101Help to ensure on-going performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102Help to ensure system security, data integrity and traceability . . . . . 104

Appendix A Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108

Appendix B Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112

Good Laboratory Practice and current Good Manufacturing Practice

Part 3Vendor contributions

to GLP practices

Part 4Appendixes

Page 14: Good Laboratory practice and current good manufacturing practice
Page 15: Good Laboratory practice and current good manufacturing practice

Part 1Introduction to GLP/cGMP basics

These initial chapters review the historical background to the formation of GLP/cGMP guidelines and discuss therequirements involved in following them.

1

Page 16: Good Laboratory practice and current good manufacturing practice
Page 17: Good Laboratory practice and current good manufacturing practice

Chapter 1Background

Page 18: Good Laboratory practice and current good manufacturing practice

4

Public agencies are responsible for protecting their

citizens and their local environment from hazardous

materials. To make judgements on product safety

requires sound analytical data, traceable to source.

Good Laboratory Practices and Good Manufacturing

Practices are the result of this requirement.

Various national legislation for example, the Federal Food,Drug, and Cosmetic Act in the United States, places theresponsibility for establishing the safety and efficacy ofhuman and veterinary drugs (and devices) and the safetyof food and color additives on the sponsor (manufacturer)of the regulated product. Public agencies like the UnitedStates government’s Food and Drug Administration (FDA)or the Ministry of Health and Welfare in Japan (MOHW)are responsible for reviewing the sponsor’s test results anddetermining whether or not they can demonstrate theproduct’s safety and efficacy. The marketing of theproduct is permitted only when the agencies are satisfiedthat safety and efficacy have been established adequately.4

Until the mid-1970s, the underlying assumption at the FDA was that the reports submitted by the sponsors to theagency accurately described study conduct and preciselyreported the study data. Suspicion about this assumptionwas raised during the review of studies submitted by amajor pharmaceutical manufacturer in support of newdrug applications for two important therapeutic products.Data inconsistencies and evidence of unacceptablelaboratory practices came to light. The FDA requested a“for cause” inspection of the manufacturer’s laboratoriesto determine the cause and the extent of the discrepancies(a “for cause” inspection is one initiated at the request ofan agency when there are grounds for doubt surroundingan FDA regulated product), and revealed defects in design,conduct, and reporting of the studies. Further inspectionsat several other sites found similar problems.4

Chapter 1Background

Historicalperspective (GLP)

Fraud and

misinterpreted data

Page 19: Good Laboratory practice and current good manufacturing practice

5

The conclusion, that many of the studies on which proofof safety of regulated products had been based couldindeed be invalid, alarmed the FDA, the United StatesCongress, the public, and industry. Working groups weresoon formed to develop ways and means of ensuring thevalidity and reliability of all non-clinical safety studiessubmitted for FDA decision approval. They wouldeventually publish standards for measuring theperformance of research laboratories and define anenforcement policy.

Good Laboratory Practice (GLP) regulations were finallyproposed on November 19, 1976 for assuring a study’svalidity. The proposed regulations were designated as a new part, 3.e., of Chapter 21 of the Code of FederalRegulations. The final regulations were codified as 21CFR Part 58.

The United States Environmental Protection Agency(EPA) issued almost identical regulations in 1983 to coverthe required health and safety testing of agricultural andindustrial chemicals under the Federal Insecticide,Fungicide and Rodenticide Act (FIFRA)5 and the ToxicSubstances Control Act (TSCA)6 respectively. The GLPswere promulgated in response to problems encounteredwith the reliability of submitted studies. Some of thestudies were so poorly conducted that “the resulting data

could not be relied upon for the EPA’s regulatory decision

making process.” 7 The EPA regulations were extensivelyamended in 1989 and now essentially cover all testingrequired to be submitted to EPA under either Act.8,9

Both EPA GLP regulations are of a similar format andhave, with a few exceptions, the same wording.

Good Manufacturing Practice (GMP) regulatesmanufacturing and its associated quality control (in contrast to GLP which covers more drug developmentactivities). GMP predates GLP. Industries were alreadyfamiliar with GMP and thus GLP follows similar lines. Themost significant difference is in archiving requirements fortest samples and data.

Part 1Introduction to GLP/cGMP basics

FDA’s reaction

EPA’s reaction

GMP and cGMP

EPA

40 CFR 160

GoodLaboratoryPractice

Standards

21CFR58

Good Laboratory Practices

Page 20: Good Laboratory practice and current good manufacturing practice

6

Good Manufacturing Practice regulations have beendeveloped to ensure that medicinal (pharmaceutical)products are consistently produced and controlled to thequality standards appropriate to their intended use. Theyhave been developed and introduced in 1963 in responseto the US public’s concern about the safety, efficacy andoverall quality of drugs. In the United States theregulations are called current Good ManufacturingPractices (cGMP) to take into account that the regulationsare not static but rather dynamic. They are defined in Title21 of the U.S. Code of Federal Regulations: 21 CFR 210 –Current Good Manufacturing Practice for drugs, generaland 21 CFR 211 – Current Good Manufacturing Practicefor finished pharmaceuticals. In 1996 the FDA proposed asignificant revision of the regulation. Any drug marketedin the US must first receive FDA approval, and must bemanufactured in accordance with the US cGMPregulations. Because of this, FDA regulations have set aninternational regulation benchmark for pharmaceuticalmanufacturing.

In Europe local Good Manufacturing Practice regulationsexist in many countries. They are based on the EuropeanUnion (EU) directive: Good Manufacturing Practice forMedicinal Products in the European Community. This EU GMP is necessary to permit free trade in medicinalproducts between the member countries. Regulations inthe EU allow for the marketing of a new drug in thetwelve member countries with a single marketingapproval. The EU GMP is intended to establish a minimummanufacturing standard for all member states.

The EU directive has been widely harmonized with the Guide to Good Manufacturing Practice for

Pharmaceutical Products as developed by thePharmaceutical Inspection Convention (PIC).10

Chapter 1Background

EU

Guide to GoodManufacturing

Practices

21 CFR 21021 CFR 211

Current GoodManufacturing

Practices

Page 21: Good Laboratory practice and current good manufacturing practice

7

Shortly after the US FDA introduced GLP regulations, theOrganization for Economic Cooperation and Development(OECD) published a compilation of Good LaboratoryPractices. OECD member countries have sinceincorporated GLP into their own legislations. In Europe,the Commission of the European Economic Community(EEC) has made efforts to harmonize the Europeanlaws.11 A list of national GLP authorities has beenpublished in reference 39. Guidelines on quality assurancefor measuring equipment and for calibration laboratorieshave also been published by technical committees of theInternational Organization for Standardization (ISO) andothers, for example in ISO/IEC 17025 (40).

To overcome trade differences and enable GLPs to berecognized abroad, bilateral memoranda of understanding(MOU) have been signed between many chemical tradingnations. For example, bilateral agreements have beensigned between all countries within the EuropeanEconomic Community. After signing such agreements,data generated and approved by national GLP authoritieswithin one country will be accepted by the national GLPauthority of the other country.

Originally, GLP regulations were intended for toxicitytesting only. It was reserved for labs undertaking animalstudies for pre-clinical work. Their general nature,applicable to any analytical instrument and method,enables implementation in all scientific disciplines and particularly in those which perform analyticalmeasurements. Some laboratories follow GLP’s wheneverthe studies are to be used to support applications forresearch or marketing studies to be submitted to the FDA,for example when doing biocompatibility testing of a newmaterial.

Part 1Introduction to GLP/cGMP basics

National andinternational GLP

regulations

Memoranda of

Understanding (MOU)

and bilateral agreements

Who has to complywith GLP/cGMP

regulations?

Page 22: Good Laboratory practice and current good manufacturing practice

8

Quality control of drugs typically is regulated underCurrent Good Manufacturing regulations. But GMP is notlimited to quality control laboratories in manufacturing. If,for example, a small volume ingredient is prepared in aresearch and development department, then the workshould also be performed under GMP. Similarly, anyproduction of material made for clinical trials also fallsunder GMP.

Some organizations publish their documents on theInternet. For example the US FDA cGMP regulations areavailable on FDA’s Website http://www.fda.gov

EU Directives can be downloaded fromhttp://dg3.eudra.org

For more links to national and international regulationsand guidelines see reference 39.

Chapter 1Background

Non-clinical laboratory

studies

GLP regulates all non-clinical safety studies that supportor are intended to support applications for research ormarketing permits for products regulated by the FDA orother similar national legislation (see table 1). Thisincludes medicinal and veterinary drugs, aroma and coloradditives in food, nutrition supplements for livestock, andbiological products.

GLP is needed for:Non clinical safety studies of development of drugs

Agricultural pesticide development

Development of toxic chemicals

Food control (food additives)

Test of substance with regard to explosive hazards

GLP is not needed for:Basic research

Studies to develop new analytical methods

Chemical tests used to derive thespecifications of a marketed food product

Analysis under GMP

and cGMP

Availability of

regulations and guidance

documents

Table 1

Page 23: Good Laboratory practice and current good manufacturing practice

9

ISO/IEC 17025

Ref. 40

Part 1Introduction to GLP/cGMP basics

For those who prefer regulations in paper format, you canorder them through various publishers: for example, theEnglish text of 27 national and international (current) GoodManufacturing Practice can be found in the bookInternational Drug GMP’s.12 International GMPs includethe most recent versions from the World HealthOrganization (WHO), Asia, Pharmaceutical InspectionConvention (PIC) and the European Union (EU).

Good guidance documents for laboratories are ISO/IECguides and standards especiallyISO/IEC 17025: “Generalrequirements for the competence and testing of calibrationand testing laboratories”. This standard is the inter-nationally recognized basic document for accreditation oflaboratories. The attainment of accreditation is mandatoryfor some regulatory work areas and frequently is the basisof contracts for analytical work.

Page 24: Good Laboratory practice and current good manufacturing practice

10

Page 25: Good Laboratory practice and current good manufacturing practice

11

Chapter 2GLP/GMP key provisions

Page 26: Good Laboratory practice and current good manufacturing practice

12

Chapter 2GLP/GMP key provisions

Analytical and other work carried out in a regulated

environment is different from that performed

independently of GLP and GMP rules. Either

additional responsibilities are imposed on analysts

or these responsibilities need to be carried out by

additional personnel. This new work must be

documented extensively and documents must be

archived for many years.

A laboratory which intends to conduct studies that areGLP compliant will have to be organized so that theconditions listed below in table 2 apply (the following listis not exhaustive).

GLP organizationand conditions

Study director (for toxicological studies)For each study to be performed, the facilitymanagement must appoint a study director– the individual responsible for the overallconduct of the study. He or she isresponsible for the technical conduct ofthe study, as well as for interpretation,analysis, documentation and reporting ofthe results.

Quality assurance unitA quality assurance unit (QAU) must bedesignated to audit the laboratory studiesand the accompanying data. It may be a separate department or an individualperson, either full- or part time, indeed anyperson other than the study director.

Usually, the QAU is also responsible for preparing a GLP inspection and forsupplying the data to the FDA or othercontrol agencies. The QAU is designatedby the testing facility management.

PersonnelMust be qualified through education, training and/or experience to follow directions andperform test procedures properly.

Standard operating proceduresAll laboratory activities must be performedin accordance with correctly written andproperly filed, management-approvedstandard operating procedures (SOPs). These must be readily available to the

personnel concerned. They should coverpolicies, administration, technicaloperation, equipment operation andanalytical methods

Control and test articlesMust be identified and characterized bystrength, purity, and stability. Reagents andsolutions must be labeled with informationon origin, identity, concentration, storageconditions, and expiration date.

EquipmentInstruments must be designed to meetanalytical requirements and regularlymaintained and calibrated and copies mustbe kept on these procedures.

Table 2

Page 27: Good Laboratory practice and current good manufacturing practice

13

Part 1Introduction to GLP/cGMP basics

The study director has overall responsibility for thetechnical conduct of the safety studies, as well as for theinterpretation, analysis, documentation and reporting ofthe results. He or she is designated by and receivessupport from management. The study director serves asthe single point of study control. It is important that thisis a single individual person and not a department or anyother grouping of people.

The study director may be the laboratory manager andmay be responsible for more than one study. However, heor she should not be over-burdened – an auditor couldotherwise get the impression that the study directorcannot monitor all studies carefully, see table 3.

Study director

Particular duties

The determination of the appropriateness of the test system is a scientific decision made bymanagement at the time of protocol approval. The study director need only assure thatprotocol specifications are followed.

The study director is not required to observe every data collection event, but should assurethat data is collected as specified by the protocol. The study director should also reviewdata periodically, or assure that such a review occurs.

Circumstances that may affect the quality and integrity of the study must be noted, then corrective action taken and documented.

Deviations from GLP requirements noted by QAU are reported periodically to themanagement and to the study director. If those reports indicate that corrective action is needed for any deviation from regulatory requirements, it is the study director’sresponsibility to assure that corrective action occurs.

A final statement is made in the study report that the study was conducted in compliance with GLP regulations.

All raw data, documentation, protocols, specimens, and final reports are transferred to thearchives during or at the close of the study.

ResponsibilitiesApproval of protocols and any subsequent changes.

Ensuring that the current revision of theprotocol is followed.

Ensuring correct recording of experimentaldata.

Collating records of, and verifying, allexperimental data, including observationsof unforeseen events.

Assure that all applicable GLP regulationsare followed.

Final statement on GLP compliance.

Assure timely archiving.

Table 3

Page 28: Good Laboratory practice and current good manufacturing practice

14

The quality assurance unit (QAU) serves as an internalcontrol function. It is responsible for monitoring eachstudy to assure management that facilities, personnel,methods, practices, records, controls, SOPs, final reports(for integrity), and archives are in conformance with theGLP regulations. For any given study, the QAU is entirelyseparate from and independent of the personnel engagedin the direction and conduct of that study.

As well as the immediate reporting of any problems, GLPregulations require the QAU to maintain and periodicallysubmit to laboratory management comprehensive writtenlisting findings and problems, actions recommended andtaken, and scheduled dates for inspection. A designatedrepresentative from the FDA or EPA may ask to see thewritten procedures established for the QAU’s inspectionand may request the laboratory’s management to certifythat inspections are being implemented, and followed-upin accordance with the regulations governing the QAU.

Part-time or full-time personnel may be used depending on whether the volume of work is sufficient to justifyemploying one or more full-time quality assuranceprofessionals. Full-time professionals are the preferredarrangement, because such an arrangement provides adegree of independence and removes the possibility thatthe demands of the person’s second job will interfere withhis or her performance of the QA function. For smallorganizations it might not be possible to designate a full-time person.

The regulation mandates that responsibilities andprocedures applicable to the QAU, the records maintainedby the QAU, and the method of indexing such records bemaintained. The regulation further requires that theseitems, including inspection dates, the description of thestudy inspected, the phase or segment of the study, andthe name of the individual performing the inspection, be made available for review by an authorized FDA agent.

Chapter 2GLP/GMP key provisions

Quality assuranceunit (QAU)

Page 29: Good Laboratory practice and current good manufacturing practice

15

Part 1Introduction to GLP/cGMP basics

Main activities of a QAU

✔ Maintain copy of master schedule sheetof all studies conducted. These are to be indexed by test article and must contain the test system, nature of study, datethe study was initiated, current statusof each study, identity of the sponsor,and name of the study director.

✔ Maintain copies of all protocolspertaining to the studies for which QAU is responsible.

✔ Inspect studies at adequate intervals toassure the integrity of the study andmaintain written and correctly signedrecords of each periodic inspection.These records must show the date ofthe inspection, the study inspected, the

phase or segment of the studyinspected, the person performing theinspection, findings and problems,action recommended and taken toresolve existing problems, and anyscheduled date for reinspection. Any problems discovered which arelikely to affect study integrity are to bebrought to the attention of the studydirector and management immediately.

✔ Periodically submit to management and the study director written statusreports on each study, noting problemsand corrective actions taken.

✔ Determine whether deviations fromprotocols and SOPs were made withproper authorization anddocumentation.

✔ Review the final study report to ensurethat it accurately describes themethods and SOPs and that thereported results accurately reflect the raw data of the study.

✔ Prepare and sign a statement to beincluded with the final study report thatspecifies the dates of audits and datesof reports to management and to thestudy director.

✔ Audit the correctness of the statement,made by the study director, on the GLPcompliance of the study.

✔ Audit laboratory equipment.

The FDA agent cannot request findings of a QAU audit,see table 4.

The QAU activies also include oversight of the laboratoryequipment and procedures. This does not require that theQAU staff become experts in computer operations, butrather that they are familiar with the test procedure andhave sufficient competence to inspect and audit thesystem procedures and practices to evaluate theircompliance to GLPs.

Table 4

Page 30: Good Laboratory practice and current good manufacturing practice

16

Standard operatingprocedures (SOPs)

Routine inspection, cleaning, maintenance, testing, calibration and standardization ofinstruments.

Actions to be taken in response to equipment failure.

Analytical methods.

Definition of raw data.

Data handling, storage, and retrieval.

Qualification of personnel.

Health and safety precautions.

Authorized access to equipment.

Receipt, identification, storage, mixing, and method sampling of test and control articles

Record keeping, reporting, storage, and retrieval of data.

Coding of studies, handling of data, including the use of computerized data systems.

Operation of quality assurance personnel in performing and reporting study audits,inspections, and final study report reviews.

SOPs shall be established for, but not limited to:

Standard operating procedures (SOPs) are writtenprocedures for a laboratory’s program. They define how to carry out protocol-specified activities, and are often written in a chronological listing of action steps, see table 5.

Chapter 2GLP/GMP key provisions

SOPs should preferably be written in the laboratory closeto the instrument, and not in an office. It should be eitherwritten or thoroughly reviewed by the instruments’operators. SOPs should not be written to explain howprocedures are supposed to work, but how they work.

This ensures that the information is adequate and that thedocument invites rather than discourages routine use.Content should cover:

• SOP unique number and revision number,

• page number and total number of pages,

• for equipment testing: performance acceptance criteria, recommended corrective actions, and a template for continuous entries of test results and corrective actions,

• printing history.

Table 5

Page 31: Good Laboratory practice and current good manufacturing practice

17

Part 1Introduction to GLP/cGMP basics

Copies of SOPs for equipment should be located close to the instruments and must be easily accessible byoperators.

Deviations from SOPs in a study must be authorized bythe study director and significant changes in establishedSOPs must be authorized in writing by management.

How specific should a SOP be or how general can it be? If written too restrictively, SOPs will frequently needrevising. On the other hand, if the details are insufficient,instructions will fail to provide adequate direction to studypersonnel. SOPs should be detailed enough to providemeaningful direction to study personnel. The level of

Title page should include

Location

Deviations and changes

Level of detail

Page 32: Good Laboratory practice and current good manufacturing practice

18

Chapter 2GLP/GMP key provisions

Reagents andsolutions

Expiration date

Storage conditions

Language

detail depends mainly on the education, training, andexperience of the study personnel. Things that maychange frequently, for example the suppliers of materialsshould not be specified in a SOP.

Standard operating procedures should be drafted in alanguage understood in the workplace.

All reagents and solutions in the laboratory areas shall be labeled to indicate identity, titer or concentration,storage requirements, and expiration date. Deteriorated or outdated reagents and solutions should not be used. If reagents and solutions used for non-regulated work are stored in the same room as reagents for regulatedstudies, all reagents must be labeled. Reagents that are not adequately labeled, even if not intended for use inregulated studies, may have an adverse effect on regulatedlaboratory work. It is also good practice to include theDate opened. This can be critical for some chemicals such as ether.

The expiration date depends on the nature of thechemical. Sodium chloride has practically no expirationdate. In these cases it might be acceptable to indicateNONE or Not applicable (N/A) on the label forexpiration date. The laboratory must be prepared to justifythis designation. Formal studies are not required to justifyassigned expiration dates. It is sufficient to assignexpiration dates based on literature references and/orlaboratory experience.

The label should indicate special environmental conditions, for example Refrigerate or Protect from light.

Page 33: Good Laboratory practice and current good manufacturing practice

19

Control articles (referred to as reference substances in the OECD principles) are of utmost importance becausethey are commonly used to calibrate the instrument. The accuracy of the reference substances, typicallysimultaneously, determines the accuracy of the analyticalmethod and therefore the interest in the certification andhandling of control articles.

The identity, strength, purity, composition and/or othercharacteristics, which will appropriately define the test or control article, should be determined for each batchand documented. Methods of synthesis, fabrication, orderivation of test and control articles should also bedocumented. Copies of this documentation must beincluded with the study and must be available for FDAinspection.

The stability of each test or control article should bedetermined. This can be done either before studyinitiation, or concomitantly according to written SOPswhich provide for periodic reanalysis of each batch.

Each storage container for a test or control article shouldbe labeled by name, chemical abstract number or codenumber, batch number, expiration date, if any, and, whereappropriate, storage conditions necessary to maintain theidentity, strength, purity and composition. Storagecontainers should be assigned to a particular test articlefor the length of the study.

Certified reference standards can be purchased fromappropriate suppliers. If standards are not available, therecommendation is to take a lot of your own material, andanalyze, certify and use it as the standard. Sometimescertified standards are too expensive for day-to-dayroutine use. In this case homemade laboratory standardscan be used as working standards. However, they shouldbe made from high purity material and be comparedagainst the primary standard to ensure the traceabilitychain. For the comparison, validated test methods should

Part 1Introduction to GLP/cGMP basics

Test and controlarticles

Characterization

Stability

Storage container

Standards

Page 34: Good Laboratory practice and current good manufacturing practice

20

be used. All reference material, either purchased or homemade, should be subject to a quality control procedure.This includes regular checks of purity, identity andconcentrations. Section “Certified Reference Standard” in Chapter two describes how to prepare and qualifyreference standards.

Raw data refers to any laboratory worksheets, records,memoranda, notes, or exact copies thereof, that are theresults of original observations and activities of a study.The term covers all data necessary for the reconstructionof the report of the study. Raw data may include hand-written notes, photographs, microfiche copies, computerprint-outs, magnetic media, dictated observations, andelectronically recorded data from automated instruments. Examples include records of animal receipt, results ofenvironmental monitoring, instrument calibration records,and integrator output from analytical equipment. Raw datamay also be entries in a worksheet used to read and noteinformation from the LED display of an analyticalinstrument.

For raw data entries, it is recommended to use controlledforms or a laboratory notebook for each study. Thisshould be robust, bound and have the pages numbered. Allentries should be made in indelible ink. Scientists andtechnicians sometimes record raw data on scraps of paperor even on paper towels. Their intention is to neatlytranscribe the information to official data forms at a latertime and to discard the originally recorded data. Thispractice should be discouraged, because the scraps ofpaper are the real raw data, and must be retained.

More recently electronic notebooks are used instead of paper notebooks. For US-FDA GLP/GMP regulatedlaboratories the regulation on electronic records andsignatures, 21 CFR Part11 applies (for details, see chapter eight of this primer).

Chapter 2GLP/GMP key provisions

Raw data

The laboratory notebook

When electronic notebooks are used, 21 CFR Part 11 applies

Page 35: Good Laboratory practice and current good manufacturing practice

21

If raw data is transferred to a computer data base, neitherthe electronically stored data nor its paper print out cansubstitute for the original.

If data are captured directly by a computer (for example if a balance is connected to a computer), until 1997 thelaboratory could elect to treat either the electronicallyrecorded information or a hard copy print out as raw data.If a hard copy was retained, the magnetic copy may have been deleted. With the release of 21CFR Part 11 (electronic records, electronic signatures) in 1997 this has changed.As soon as any data hit a durable storage device such as acomputer hard disk, electronic records become the rawdata and must be kept for the duration as required by thepredicate rule, for example, GLP or cGMP’s.

When raw data are recorded on paper all changes shouldbe made by drawing a single line through the data beingchanged, recording the corrected information and the dateof change, and indicating a reason for the change. Theperson making the change should be identified by asignature or initial.

Special rules apply in the case of automated datacollection systems: the instrument and person responsiblefor data collection must be identified at the time of datainput. Changes in automated data entries must be made insuch a way that the original entry is saved, and the personresponsible for making the changes must be identified.Any changes to data must be automatically recorded bythe computer together with a time stamp as part of theautomated audit trail. When working in a GLPenvironment, the reason for the change must also berecorded.

Part 1Introduction to GLP/cGMP basics

Transcriptions to

computers

Direct data capture

by a computer

Modification of raw data

Corrections on paper must be legible,legitimized and authorized. When usingelectronic notebooks, changes toelectronic records must not obscureoriginal data.

Don’t do this ...

... but this!

Page 36: Good Laboratory practice and current good manufacturing practice

22

When working with computer systems, special care shouldbe taken regarding data integrity and security. There isfrequently a higher risk that unauthorized users haveaccess to these data and changes to electronic files maybe more difficult to recognize than changes on paperrecords Controlled limited access to the computerhardware and/or log on control through biometric devices or a combination of password and user I.D. aremechanisms used to ensure security.

All raw data, documentation, SOPs, protocols, finalreports, and specimens (except those specimens obtainedfrom mutagenicity tests and wet specimens of blood,urine, feces, and biological fluids) should be retained.

There should be archives for orderly storage andexpedient retrieval of all raw data, documentation,protocols, specimens, and interim and final reports.Appropriate storage (temperature, humidity) shouldminimize deterioration of documents and specimens inaccordance with the requirements for the time period oftheir retention and the nature of the documents orspecimens. For example, paper documents should not besubjected to long periods of high humidity. Separatestorage rooms should be available if certain subjects candeteriorate others. For instance, samples containingformaldehyde should not be stored in the same rooms aspaper. Storage conditions should be monitored so thatdeviations from proper storage conditions can bepromptly rectified.

The length of time in which documentation and specimensmust be archived varies from country to country and maybe up to 15 years. As a general rule it is desirable thatmaterial should be retained as long as the test substance isin use. However, specimens should be retained only for aslong as it could reasonably be expected that the quality ofthe preparation would permit evaluation.

Chapter 2GLP/GMP key provisions

Integrity

Storage andarchiving

Facilities

Time period

Page 37: Good Laboratory practice and current good manufacturing practice

23

An individual should be identified as responsible for the archives. Only authorized personnel may enter thearchives. The personnel who may enter the archivesshould be defined by SOPs. If materials are removed fromthe archives, a record should be kept of what is removedand by whom. Material retained or referred to in thearchives should be indexed to permit rapid retrieval.

FDA requests that all revisions of SOPs be archived. Thisis a good reason to write SOPs so that they don’t need tobe frequently revised. Operator manuals must be archivedif they are cited in SOPs. If different software revisions areused during a study, all revisions of all manuals must bearchived.

There is a lot of discussion about long-term storage ofcomputer captured raw data. Alternatives are electronic media such as tapes or discs. In theory all media can beused as long as it is ensured that the data can be madeavailable for the entire time which is required for datastorage. The biggest problem is the availability of softwareand computer hardware to ‘instantly’ replay the raw datasuch that the same final results will be obtained as during the original reprocessing. One solution may be validated fileconversion to new computer systems. Conversion routinesshould also include ‘meta data’ such as chromatographicintegration parameters and calibration tables. For moredetails see chapter eight.

It is important to understand that the sponsor has theultimate responsibility for archiving parts of the study orquality control work subcontracted to other companies.This is particularly important should the subcontractor goout of business.

Part 1Introduction to GLP/cGMP basics

Who has access?

SOPs and operator

manuals

How should electronic

records be stored?

Subcontractors

21 CFR Part 11

Electronic Records andSignatures

Page 38: Good Laboratory practice and current good manufacturing practice

24

Each individual engaged in the conduct of, or responsiblefor the supervision of a study or analytical analysis shouldhave education, training, and experience or a combinationthereof, to enable that individual to perform the assignedfunction. Personnel must be qualified to do the work.Operators of instruments should have sufficient trainingand/or experience to correctly operate the instrument andalso to identify an instrument.

Each facility should maintain a current summary oftraining and experience and job description for eachindividual engaged in or supervising the conduct of a non-clinical laboratory study: job description, participation intraining courses, other technical instruction on GLP/GMPand instrumentation. This documentation should be keptseparate from personnel records. The documentationneeds to be regularly updated, retained, and archived.

Personnel may be employed as part or full-time for GLPstudies, as long as they have sufficient training to do thejob properly.

Laboratory management should take the health and safetyof employees into account. Minimum precautions shouldensure that employees working in an analytical laboratorywear laboratory coats and safety glasses when workingwith hazardous material. A laboratory should also have ageneric policy for safe handling of chemicals. People whohave an illness which may adversely affect the quality andintegrity of the study should be excluded from directcontact with test systems and from test and controlarticles.

Chapter 2GLP/GMP key provisions

Personnel

Documentation of

qualification

Part-time or full-time

personnel

Health precautions

and safety

Page 39: Good Laboratory practice and current good manufacturing practice

25

Equipment used in generation, measurement, orassessment of data and equipment used for facilityenvironmental control should be of appropriate design andadequate capacity to function according to the protocoland be suitably located for operation, inspection, cleaning,and maintenance. The equipment should undergo a validation process to ensure that it will consistently function as intended. Examples are analytical equipment such aschromatographs, spectrophotometers, computerizedequipment for instrument control, direct data capture,data transmission, data evaluation, printing, archiving andretrieval. Chapters three to seven will describe equipmentvalidation and qualification in more detail.

A protocol must be in place for each study. The protocolshould be written before the start of the study. In generalthe protocol must be followed. However, scientificallyjustified changes can be made, if the changes aredocumented and authorized by the study director.

The proper identification of specimens is important andincludes test systems, nature and/or date of collection.

Data entries must be recorded directly, promptly andlegibly in indelible ink, to prevent improper erasures andcorrections. All records must be signed and dated. Thisdoes not mean that every individual piece of data must besigned off. It is sufficient, for example, to provide onesignature and date for all data collected during a singledata collection session.

Changes must not obscure the original and must beexplained and signed with full signature or initials. Withthe exception of automated data collection systems, allchanges in data should be made by drawing a single linethrough the data being changed.

Part 1Introduction to GLP/cGMP basics

Equipment

How to conduct a GLP study

Identify specimens

Meet data recording

requirements

Changes

Page 40: Good Laboratory practice and current good manufacturing practice

26

Similar principles apply to automated data collectionsystems. The individual responsible for direct data inputshould be identified at the time of data input. Any changein automated data entries should be made as not toobscure the original entry, indicating the reason forchange, the date and the identity of the responsibleindividual. User I.D. and password entries can be used foridentification and treated as signatures. For more detailssee chapter 8.

Enforcement of GLP – certification and audits ofanalytical laboratories – is the responsibility of the FDAand the EPA in the United States and the Departments ofHealth and Social Affairs in the OECD and EU membercountries. Each of these government agencies mayperform detailed examinations of any laboratory facilitywithin its jurisdiction at reasonable times and in areasonable manner. Such audits involve the inspection ofthe facility, equipment records, and specimens and mayalso include the investigation of an experiment in depthfrom raw data to final reports.

This varies from country to country. In the US, the FDAhas two different types of inspections:

The routine inspection constitutes a periodicdetermination of the facility’s compliance with theregulations. The toxicological facilities and animalhandling areas may be inspected annually, but frequentlythe laboratory itself is not inspected. A data audit may bedone.

Cause inspections are conducted less frequently. Theassignment of this inspection is sometimes initiated by routine inspections when serious non-compliance withGLP regulations is found. Laboratories are not notifiedbeforehand.

Chapter 2GLP/GMP key provisions

Automated data

collection systems

Enforcement of GLP

Frequency of inspections

Page 41: Good Laboratory practice and current good manufacturing practice

27

Good Manufacturing Practice (GMP) is concerned withboth production and quality control. The basicrequirements of quality control are that:10

• adequate facilities, trained personnel and approvedprocedures are available for sampling, inspecting and testing starting materials, packaging materials,intermediate bulk, and finished products, and whereappropriate for monitoring environmental conditions for GMP purposes;

• samples of starting materials, packaging materials,intermediate products, bulk products and finishedproducts are taken by personnel and by methodsapproved by quality control;

• test methods are validated;

• records are made manually and/or by recordinginstruments, which demonstrate that all requiredsampling, inspecting and testing procedures wereactually carried out. Any deviations are fully recordedand investigated;

• the finished products contain active ingredientscomplying with the qualitative and quantitativecomposition of the marketing authorization, are of thepurity required, and are enclosed within their propercontainer and correctly labeled;

• records are made of the results of inspection and thattesting of materials, intermediate, bulk and finishedproducts are formally assessed against specification.Product assessment includes a review and evaluation ofrelevant production documentation and an assessmentof deviations from specified procedures;

• no batch of product is released for sale or supply priorto certification, by an authorized person, that it is inaccordance with the requirements of the marketingauthorization;

Part 1Introduction to GLP/cGMP basics

Good ManufacturingPractice (GMP)

Page 42: Good Laboratory practice and current good manufacturing practice

28

• sufficient reference samples of starting materials andproducts are retained to permit future examination ofthe product if necessary and that the product is retainedin its final pack unless exceptionally large packs areproduced.

Typically GMP regulations and guidelines within a specificcountry apply to all medical products manufactured inthat country or imported from other countries. Eachproduct must have a marketing authorization before it canbe sold.

In Europe, inspections are made either by an individualmember state authority in case of national marketingauthorization, or on behalf of the European Agency for theEvaluation of Medicinal Products (EMEA) for Europeanmarketing authorization.

In the US the FDA makes general system inspections andproduct specific preapproval inspections. Inspections arecarried out about every two years, and are also requiredbefore foreign firms can ship commercial products intothe USA.

The US FDA published a Guide to Inspection of

Pharmaceutical Quality Control Laboratories.13

Even though it was written as a guideline for FieldInvestigators, it is a useful document for Quality Controllaboratories. It has a large chapter on the handling of‘Failure (out of specification) Laboratory Results’ and on‘Retesting’. Other chapters give guidelines on laboratoryrecords and documentation, laboratory standardssolutions, methods validation, equipment, raw materialtesting, in process control, computerized laboratory dataacquisition systems and on laboratory management.

Chapter 2GLP/GMP key provisions

GMP inspections

Guide to inspection of

pharmaceutical quality

control

Guide to inspections ofpharmaceutical quality

control laboratories

Ref. 13

Page 43: Good Laboratory practice and current good manufacturing practice

29

Retesting out of specification samples without aninvestigation is one of most frequent laboratory controldeficiencies. In response to this the US FDA developed aspecial guidance14 on this topic. It provides the agency’s thinking on how to evaluate suspect, or out of specification (OOS), test results. The term OOS results includes all

suspect results that fall outside the specifications oracceptance criteria established in new drug applications,official compendia, or by the manufacturer.

This applies to laboratory testing during the manufactureof active pharmaceutical ingredients, excipients, and othercomponents and the testing of finished products to theextent that current good manufacturing practices (cGMP)regulations apply (21 CFR parts 210 and 211). Specifically,the guidance discusses how to investigate suspect, or OOStest results, including the responsibilities of laboratorypersonnel, the laboratory phase of the investigation,additional testing that may be necessary, when to expandthe investigation outside the laboratory, and the finalevaluation of all test results.

Internal audits are a key element of any quality system.Inspections are conducted to evaluate a company’scompliancewith regulatory expectations, applicationapprovals and company SOP/standards. A goodpreparation together with some recommendations on thispage should help to successfully survive any audit andinspection.

• Obtain all data and documentation for studies to beaudited (don’t let auditors search in file cabinets. Askand bring the requested material).

• Assign a technical contact to review the files and answerquestions. The assigned technical contact should bepresent all the time.

Part 1Introduction to GLP/cGMP basics

FDA guidance to inspections of

quality systems

Ref. 15

Surviving an auditand inspection

Preparation

Guide to investigating

out of specifications

(OOS) test results

Guide to investigating out of specifications

(OOS) test result

Ref. 14

Page 44: Good Laboratory practice and current good manufacturing practice

30

Chapter 2GLP/GMP key provisions

• Review the QA files. Prepare an agenda for theinspection. Set up a work area for the inspectors.Review the master schedule. Present a floor plan of test facility. Prepare staff. An audit may be a toughexperience for all people involved. Therefore they needto be informed on what will happen and the questionswhich may be asked.

• Maintain a continuous log of the inspection.

• Provide copies (do not give originals away!).

• Keep duplicates of all information supplied to auditors.

• Take immediate corrective action, when appropriate.

• Hold a daily debriefing meeting to assess the progress.

• Keep all documents in the work area.

• Accompany the inspector all the time.

• Be courteous and co-operative.

• Answer only questions that are asked.

• If you are unable to answer, tell the inspector openly.

• Protect proprietary information.

Conduct an exit review and ask if there are any questionsor cause for dissatisfaction. Finally, create a file of theinspection material and prepare an audit report.

One problem for analysts in laboratories is the change inenforcement and inspection practices. The industry canuse a variety of information from the FDA to stay abreastof the areas of most concern. The sources includepresentations from FDA inspectors; warning letters, pre-approval withhold recommendations and 483observations. Some of this information is available in the Internet through FDA and other websites, e.g.http://www.fda.gov/cder/warn/index.htm

http://www.fda.gov/foi/warning.htm

http://www.fdawarningletter.com

Conduct

Close

FDA 483observations and

warning letters

Page 45: Good Laboratory practice and current good manufacturing practice

31

Part 1Introduction to GLP/cGMP basics

During the1980’s FDA investigators focused their activities on process control, and in the 90’s they paidmore attention to laboratories. Zareth16 reported thatlaboratory controls were most frequently cited byinvestigators.

• Laboratory controls 67%

• Records 67%

• Process validation 50%

• Process controls 45%

• Stability 43%

Laboratory control deficiencies include the following:

• Retests without appropriate investigations.

• Use of unvalidated computer systems and software.

• Use of uncalibrated equipment.

• Use of unvalidated test methods.

• No investigation of abnormal or missing data.

• Incorrect use of secondary reference standards.

cGMP notes are another useful source of information from the FDA. There is a periodic memo on Current Good Manufacturing Practice Issues on Human UsePharmaceuticals available, issued by the Division ofManufacturing and Product Quality, HFD-320, Office ofCompliance, Center for Drug Evaluation and Research,U.S. Food and Drug Administration, 7520 Standish Place,Rockville, MD 20855. The memo is an internal FDAissuance intended to enhance field/headquarterscommunications on cGMP issues in a timely manner. It isa forum to hear and address cGMP questions, provideupdates on cGMP projects, and clarify and help applyexisting policy to day to day activities of FDA staff.

Inspection trends

FDA cGMP notes

Page 46: Good Laboratory practice and current good manufacturing practice
Page 47: Good Laboratory practice and current good manufacturing practice

33

Part 2Impact of GLP/cGMP in the analytical laboratory

The following chapters discuss in more detail how adoptingGood Laboratory Practice and current Good ManufacturingPractice will affect analytical instrumentation and methods.

2

Page 48: Good Laboratory practice and current good manufacturing practice

34

Page 49: Good Laboratory practice and current good manufacturing practice

35

Chapter 3Validation overview

Page 50: Good Laboratory practice and current good manufacturing practice

36

Chapter 3Validation overview

One of the key requirements of GLP and GMP

regulations for analytical laboratories is validation:

equipment hardware, software, systems, methods

and data. Validation is not a one-time event but on-

going covering all phases of a product or process.

Validation is the evaluating of processes, products oranalytical methods to ensure compliance with product ormethod requirements. Prerequisites to fulfill theserequirements for analytical laboratories are properlyfunctioning and well documented instruments (hardwareand firmware), computer hardware and software andvalidated analytical methods. When the equipment and aparticular method have been selected and found to bevalidated, the equipment for that method goes through asystem suitability test before and during sample analysis.Validation includes also checking functions related to dataintegrity, security and traceability. One of the mostpopular definitions for validation came from the US FDA’General Principles of Validation from 1987 17:

The terms validation and qualification are frequentlymixed up and there is also some overlap. Equipmentqualification means checking an instrument forcompliance with previously defined functional andperformance specifications. For Operational Qualification,generic standards and analytical conditions are usedrather than real sample conditions. Validation relates moreto the entire but sample specific process including samplepreparation, analysis, and data evaluation. For software,validation includes the whole process from design toretirement of the product. It should include processes thatadress on-going support and (change) control of thesystem. Qualification here is more concerned with testingthe compliance of individual phases with specifications.

What is validation?

Validation, versus

qualification

FDA’s Glossary of Computer Systems

Software DevelopmentTerminology (August 1995)

Ref. 18

”Establishing documented evidence which

provides a high degree of assurance that a specific process will consistently

produce a product meeting it’spredetermined specifications

and quality attributes.“

FDA 1987Ref. 17

Page 51: Good Laboratory practice and current good manufacturing practice

37

Part 2Impact of GLP/cGMP in the analytical laboratory

Many laboratory managers associate validation withincreased workload in the laboratory or increased paperwork. However, validation is essentially nothing new. Eversince the development of analytical instrumentation andmethods, statistics have been used to prove the correctfunctioning, reliability and precision of the equipment andmethods. New to most existing validation procedures isthe disciplined planning of validation and documentationof all testing experiments.

Validation efforts in the analytical laboratory should bebroken down into separate components addressing theequipment (both the instrument and the computercontrolling it) and the analytical methods run on thatequipment. After these have been verified separately they should be checked together to confirm expectedperformance limits (so-called system suitability testing),and finally the sample analysis data collected on such asystem should be authenticated with suitable validationcheckouts. Other activities include checking referencestandards and qualification of people.

All (computerized) equipment that is used to create,modify, maintain, archive, retrieve, or distribute criticaldata for cGMP/GCP/GLP purposes should be validated.Validation of hardware includes testing the instrumentaccording to the documented specifications. Even thoughthis may include word processing systems to create andmaintain SOPs, in this primer we only will cover analyticalsystems. If instruments consist of several modules, amodular HPLC system for example, the entire systemshould be validated. Validation of computer systems mustinclude the qualification of hardware and software.

What has to bevalidated?

Equipment

“Validation means nothing else than well-organized, well-documented

common sense”

Ken Chapman

Page 52: Good Laboratory practice and current good manufacturing practice

38

Chapter 3Validation overview

Validation covers testing of significant methodcharacteristics, for example sensitivity and reproducibility.

The system combines instrument, computer and analyticalmethod. This validation usually referred to as system

suitability testing, tests the system for documented performance specifications for the specific analysis method.

When analyzing samples the data must be validated. Thevalidation process includes documentation and checks fordata plausibility, consistency, integrity and traceability. Acomplete audit trail must be in place, which allows tracingback the final result to the raw data for integrity.

People should be qualified for their jobs. This includeseducation, training and/or experience.

Reference standards should be checked for purity,identity, concentrations and stability.

Analysis method

Analytical system

Data

Personnel

Reference standards

ObjectivesProve suitability for intended use

System suitability testing

Hard- and softwarevalidation

Analytical method validation

Page 53: Good Laboratory practice and current good manufacturing practice

39

Part 2Impact of GLP/cGMP in the analytical laboratory

Instrument hardware should be validated prior to routineuse, after repair and at regular time intervals. Computersystems should be validated during and at the end of thedevelopment process, during installation, prior and duringroutine use and after software updates.

Computer systems with complex software are frequentlydeveloped over many years. It is critical to note thatquality cannot be tested into a product or system at the final testing stages. To ensure quality during thedevelopment process the life cycle concept was developed.Adding this concept to the definition of validation, thecomplete concept of validation incorporates verifying thedevelopmental activities as they are accomplished, and theformal testing of the end product system.

Analytical methods should be validated prior to routineuse and after changing method parameters. Analyticalsystems should be tested for system suitability prior toand during routine use, practically on a day by day basis.

Validation of equipment starts when somebody has an ideaabout a product and it ends when the product has beenretired from the laboratory and all methods and data havebeen successfully converted to a new system.

It is a good practice to document all validation activities ina validation master plan or in an equivalent document. TheFDA does not specifically demand a validation masterplan. However, inspectors want to know what thecompany’s approach towards validation is. The validationmaster plan is an ideal tool to communicate this approachinternally and to inspectors.

When shouldvalidation be done?

Steps towards equipmentvalidation

Validation of software starts when

you have an idea and it ends when you have removed

the product from operation and all data are transferred

to and validated on the new system

Validation master plan

Best Practice

Validation Master Plan

Ludwig Huber

Ref. 41

Page 54: Good Laboratory practice and current good manufacturing practice

40

Chapter 3Validation overview

For complex computerized equipment a validation team should be formed. Members should include alldepartments that have anything to do with the equipment.Such members typically come from the analytical lab, theQA department, validation groups, and also from the ITdepartment. Responsibilities are defined in the validationmaster plan and generally include identifying equipmentrequiring validation, prioritization of the validation to beperformed, developing revisions of the validation masterplan and establishing procedures for computer systemvalidation.

An equipment inventory that includes all equipment in thelaboratory should be available. It should list all hardwareand software in use within the laboratory and is the firststep in identifying systems that require validation. Theinventory should include information on the validationstatus and on the criticality of data generated by thesystem. This inventory may also be the starting point atinspections.

Any validation should start with setting and documentingthe specifications for user requirements, instrumentfunctions and performance. The specifications of theinstrument’s design should be compared with the userrequirement specifications. It is a simple rule of thumb:without specifications there is no validation. Asking forspecifications is also frequently the initial step for specificequipment inspection. DQ is the most important step inthe validation process. Errors made in this phase can havea tremendous impact on the workload during later phases.

The user has the ultimate responsibility for validation.Some validation activities, especially during developmentof software, can only be carried out by the vendor. Usersshould qualify vendors for compliance with theirvalidation needs.

Validation committee

Inventory

Design qualification

(DQ)

Qualification of

the vendor

Page 55: Good Laboratory practice and current good manufacturing practice

41

Part 2Impact of GLP/cGMP in the analytical laboratory

Installation qualification verifies and documents that theinstrument has arrived as purchased and that it has beenproperly installed.

Operational qualification verifies and documents that theinstrument functions and performs in the users laboratoryas defined in the DQ phase.

On-going performance qualification includes preventivemaintenance and regular tests such as system suitabilityand quality control analyses with creation of QC-charts.For computer systems it also includes regular data backup, virus checks and change control procedures.

This is the most critical step for computerized systemsand attracts the most attention at FDA inspections. Itincludes authorized and traceable access to systems,applications, methods and data. It also includes electronicaudit trail and mechanisms to delete or change records.

On completion of the installation and operationalqualification, documentation should be available thatconsists of:

• Validation plan and protocols.

• User requirement and functional specifications.

• Evidence of vendor qualification.

• The installation qualification document (includes description of hardware and software).

• Operating and maintenance manuals and SOPs fortesting.

• Qualification test reports with signatures and dates.

• Summary of test results and a formal statement that the system has been accepted.

• Approval of user, validation department and qualityassurance.

Installation qualification

(IQ)

Operational

qualification (OQ)

Performance

qualification (PQ)

Data validation for

consistency, security,

integrity and traceability

Validation report

Page 56: Good Laboratory practice and current good manufacturing practice

42

Chapter 3Validation overview

Individual validation

plan

System scopeShould explain the purpose of the system in sufficient detail to understand the major functions.

System definitionDefines the user requirement and functional specifications, the test environment withdescription of hardware, software, communications and other applications that comprisethe whole system. Also included are: security considerations, special hardwareconsiderations, and related documentation.

ResponsibilityThe individuals who are responsible for preparing and approving the validation plan mustbe specifically designated. The plan should also include the names of the people whoexecute the test. Modifications after review should be documented and authorized.

Test dataThe data to be used in the validation plan together with limitations should be specified, for instance if data sets do not cover all possible events. Data sets can come from previousexperiments or studies and should be kept for revalidation.

Expected resultsThe expected results of each test should be listed in the plan. This output will be used to determine if the acceptance (validation) testing is successful.

Acceptance criteriaThe plan must include acceptance criteria for formally accepting the system. The test plan with expected test results and acceptance criteria must be signed before the tests start.

Revalidation criteriaThe plan should include criteria for revalidation of the system after a change anywhere in the system. Depending on the extent of the change, revalidation may or may not benecessary.

Sign-offThe plan should be signed off by the person executing the tests and by management. It should include a statement that the system is validated.

Individual validation plans should be developed for largeprojects, especially for complex computer systems.Examples of the contents are shown in table 6 below:

➜➜

➜➜

➜➜

➜➜

Table 6

Page 57: Good Laboratory practice and current good manufacturing practice

43

Part 2Impact of GLP/cGMP in the analytical laboratory

In this primer we will mainly focus on strategies forequipment validation and qualification and on methodvalidation. It will not be possible to go into too muchdetail and to give a lot of practical examples. Other topicsrelated to validation such as details on computers andsoftware, macro-programs, reference compounds andother details, examples, checklists, etc can be found inreference books and official technical papers such as thePDA and Euarachem.

A few examples are given below. A more complete list canbe found on the website www.labcompliance.com. On thissite you can also find regularly updated information onvalidation and compliance issues in laboratories.

Literature:

PDA Technical Paper Number 31: Validation andqualification of Computerized Laboratory Data AcquisitionSystems (LDAS), November 1999, www.pda.org

L. Huber, “Validation and Qualification in AnalyticalLaboratories”, published by Interpharm, Buffalo Grove, IL,USA, November 1998, Agilent P/N: 5956-0036, more information: www.labcompliance.com/books

P. Bedson and M. Sargent, The development andapplication of guidance on equipment qualification ofanalytical instruments, Accreditation and QualityAssurance, 1 (6), 265-274 (1996).

L. Huber, “Validation of Computerized Analytical andNetworked Systems”, published by IHS Interpharm,Englewood, Co. USA, April 2002.

More information: www.labcompliance.com/books

Useful resources

PDA

Validation and Qualification of

Computerized LaboratoryData Acquisition

Systems

Ref. 19

Page 58: Good Laboratory practice and current good manufacturing practice

44

Page 59: Good Laboratory practice and current good manufacturing practice

45

Chapter 4Design qualification (DQ)

Page 60: Good Laboratory practice and current good manufacturing practice

46

Chapter 4Design qualification (DQ)

“Design qualification (DQ) defines the functional

and operational specifications of the instrument and

details the conscious decisions in the selection of

the supplier”20

DQ should ensure that instruments have all the necessaryfunctions and performance criteria that will enable themto be successfully implemented for the intendedapplication and to meet business requirements. Errors insetting the functional and operational specifications canhave a tremendous technical and business impact, andtherefore a sufficient amount of time and resources shouldbe invested in the DQ phase. For example, setting wrongoperational specifications can substantially increase theworkload for OQ testing, and selecting a vendor withinsufficient support capability can decrease instrument uptime with a negative business impact.

While IQ, OQ and PQ are being performed in mostregulated laboratories, DQ is a relatively new concept tomany laboratories. It is rarely officially performed anddocumented in those cases where the equipment isplanned to be used not for a specific but for multipleapplications.

The recommended steps that should be considered forinclusion in a design qualification are listed below:

• Description of the analysis problem.

• Selection of the analysis technique.

• Description of the intended use of the equipment.

• Preliminary selection of functional and performance or operational specifications (technical, environmental, safety).

• Preliminary selection of the supplier.

Setting thespecifications

Steps for design

qualification

Page 61: Good Laboratory practice and current good manufacturing practice

47

Part 2Impact of GLP/cGMP in the analytical laboratory

• Instrument tests (if the technique is new).

• Final selection of the equipment.

• Final selection of the supplier.

• Development and documentation of the final functional and operational specifications.

DQ for computer systems should include a description ofthe intended IT environment, including the current andfuture anticipated operating system, the networkenvironment and computer system policies.

Steps are the same as for new systems. Describe what thesystem should do, which functions it should have, and theenvironment in which it is located.

It is frequently the case that instruments are used fordifferent applications with different functional andperformance or operational requirements. In this case, the recommendation is to describe the most importantintended applications and to specify the functional andperformance specifications so that they meet the criteriafor all applications. It is also possible to develop a genericDQ for instrument categories that will be used for similarapplications.

To set the functional and performance specifications, thevendor’s specification sheets can be used as guidelines.However, we would not recommend simply writing up thevendor’s specifications because compliance to thefunctional and performance specifications as described in the DQ document should be verified later on in theprocess during operational qualification and performancequalification. Specifying too many functions and settingthe values too stringently will significantly increase theworkload for OQ.

Additional steps for

computer systems

Steps for existing

systems

DQ for instruments used

for different applications

Help from instrument

vendors

Page 62: Good Laboratory practice and current good manufacturing practice

48

Chapter 4Design qualification (DQ)

Table 7 includes a few selected examples of items that canbe included in a design qualification.

Design qualificationIntended use

Analysis technique

User requirement specification for the HPLC analysis

Functional Pump

Detector

Autosampler

Column compartment

Computer

Operational

User instructions

Validation/qualification

Maintenance

Training

Selected examplesAnalysis of impurities in drugs with quantitation limit 0.1%

High performance liquid chromatography for analysis.

• 20 samples / day

• Automated over-night analysis

• Limit of quantitation: 0.1%

• Automated confirmation of peak identity and purity with diode-array detection

• Automated compound quantitation and printing of report

Binary or higher gradient

UV-vis diode-array, 190 to 400 nm

100 samples, 0.5 to 100 µl sample volume

25 to 40 °C, Peltier-controlled

System control, data acquisition for signals and spectra, peak integration and quantitation,spectral evaluation for peak purity and compound confirmation.

Electronically save all chromatograms together with meta data like integration parameters

• Detector: Baseline noise: < 5 x 10–5 AU

• Sampler: Precision inj. volume: <0.5 % RSD

• Pump: precision of retent.time: <0.5 % RSD

• Operational manual on paper

• Computer based tutorial

Vendor must provide IQ and OQ procedures and services

• Vendor must deliver maintenance procedure and recommend schedule

• Instrument must include early maintenance feedback for timely exchange of most important maintenance parts

• Maintenance procedures must be supplied on multimedia CD-ROM

Vendor must provide familiarization and training

Table 7

Page 63: Good Laboratory practice and current good manufacturing practice

49

Part 2Impact of GLP/cGMP in the analytical laboratory

Even though the user of a system has ultimateresponsibility for validation, the vendor also plays a major role. As explained earlier, the validation covers thecomplete life of a product, starting with the design anddevelopment. For commercial off the shelf systems theuser has hardly any influence on how the software is being developed and validated, but he can check throughdocumentation to see if the vendor followed anacknowledged quality process.

The vendor should:

• Develop and validate software following documented procedures.

• Test the system and document test cases, acceptance criteria and test results.

• Retain the tests protocols and source code for review at the vendor’s site.

• Provide procedures for IQ and OQ/PV.

• Implement a customer feedback, change control and response system

• Provide fast telephone, e-mail and/or on-site support

As part of the design qualification process, the vendorshould be qualified. The question is, how should this bedone? Is an established and documented quality systemenough, for example ISO 9001? Should there be a directaudit? Is there another alternative between these twoextremes?

There may be situations where a vendor audit isrecommended: for example, when complex computersystems are being developed for a specific user. However,this is rarely the case for analytical equipment. Typically,off-the-shelf systems are purchased from a vendor withlittle or no customization for specific users.

The role of thevendor

Tasks of the vendor

Qualification of the

vendor

Page 64: Good Laboratory practice and current good manufacturing practice

50

Chapter 4Design qualification (DQ)

The exact procedure to qualify a vendor depends verymuch on the individual situation, for example, is thesystem in mind employing mature or new technology? Isthe specific system in widespread use either within yourown laboratory or your company, or are there referencesin the same industry? Does the system include complexcomputer hardware and software? For example, if theequipment does not include a complex (networked)computer system, then a good reputation, their ownexperiences or good references from other users togetherwith ISO 9001 certification can be sufficient.

When the equipment to be purchased is an commercialoff-the-shelf system that includes a computer forinstrument control and data handling, we recommend the steps described in table 8 below. in so far as there is no previous experience with this vendor in yourcompany.

Page 65: Good Laboratory practice and current good manufacturing practice

51

Part 2Impact of GLP/cGMP in the analytical laboratory

1. Develop a vendor qualification checklist. This list should include questions on how the equipment is developed, validated, installed and supported (a more complete example of such a list is shown in reference 3). The most important questions are:

• Does the vendor have a documented and certified quality system, for example ISO 9001? (please note: ISO 9002 or 9003 is insufficient because they don’t cover development!)

• Is equipment hardware and computer software developed and validated according to a documented procedure, for example, according to a product life cycle?

• Is the vendor prepared to make product development, validation records and source codes accessible to regulatory agencies?

• For equipment hardware: does the vendor provide a certificate or declaration of conformity to documented manufacturing specifications?

• Does the vendor provide assistance in design qualification, equipment installation, qualification, maintenance and timely repair through qualified people?

• Is there a customer feedback and response system in case the user reports a problem or there is an enhancement request?

• Is there a change control system with suitable notification to users after the changes?

2. Send the checklist to the vendor.If the vendor answers all the questions satisfactorily within a given time frame, then the vendor is qualified.

3. If the vendor does not answer the questions satisfactorily,another vendor should be considered. If there is no other vendor who could provide an instrument that meets the operational and functional specifications, a direct audit should be considered.

Table 8

Page 66: Good Laboratory practice and current good manufacturing practice

52

Page 67: Good Laboratory practice and current good manufacturing practice

53

Chapter 5Installation qualification (IQ) andoperational qualification (OQ)

Page 68: Good Laboratory practice and current good manufacturing practice

54

Chapter 5Installation qualification (IQ) and operational qualification (OQ)

“Installation qualification” (IQ) establishes that the

instrument is received as designed and specified,

that it is properly installed in the selected

environment, and that this environment is suitable

for the operation and use of the instrument.

“Operational qualification” (OQ) is the process

of demonstrating that an instrument will function

according to its operational specification in the

selected environment.20

Steps for IQ include activities prior and during installationof the equipment. The following steps are recommendedbefore and during installation:

• Obtain manufacturer’s recommendations for installationsite requirements.

• Check the site for the fulfillment of the manufacturer’srecommendations (utilities such as electricity, water andgases and environmental conditions such as humidity,temperature and dust).

• Allow sufficient shelf space for the equipment, SOPs,operating manuals and software.

• Compare equipment, as received, with purchase order(including software, accessories, spare parts)

• Check documentation for completeness (operatingmanuals, maintenance instructions, and standardoperating procedures for testing, safety and validationcertificates).

• Check equipment for any damage.

• Install hardware (computer, equipment, fittings andtubings for fluid and gas connections, columns in HPLC and GC, power cables, data flow and instrument control cables).

• Switch on the instruments and ensure that all modulespower up and perform an electronic self-test.

• List equipment manuals and SOPs.

• Prepare an installation report.

Steps for installation qualification (IQ)

Before installation

During installation

Page 69: Good Laboratory practice and current good manufacturing practice

55

Part 2Impact of GLP/cGMP in the analytical laboratory

Computer systems should be well documented with modelnumber, serial and revision numbers and the softwareshould be documented with model and revision numbers.Documentation should include items like size of the harddisk, internal memory (RAM), installed type and version ofoperating software, standard application software anduser contributed software, for example MACRO programs.This information is important because all items caninfluence the overall performance of a computer system.The information should be readily available when aproblem occurs with the computer system.

Recommended steps for computer systems are as follows:

• Install software on computer following themanufacturer’s recommendations.

• Verify correct software installation, for example, are allfiles loaded. Utilities to do this should be included in thesoftware itself.

• Make back-up copy of software.

• Configure peripherals like printers and equipmentmodules.

• Identify and make a list with a description of allhardware, operating system software, and applicationsoftware and include drawings where appropriate.

• Make a list with a description of all software installed onthe computer.

Additional steps for

computer systems

Page 70: Good Laboratory practice and current good manufacturing practice

56

Chapter 5Installation qualification (IQ) and operational qualification (OQ)

For a larger laboratory we recommend entering theequipment data into a spreadsheet or database. Itemsshould include:

• Unique in-house identification number (asset number)

• Name of the item of equipment

• The manufacturer’s name, address and phone number forservice calls and service contract number, if there is one

• Serial number and firmware revision number ofequipment

• Computer hardware with information on the processor,hard disk space, memory and the monitor

• Software with product and revision number

• Date received

• Date placed in service

• Current location

• Size, weight

• Condition when received, for example, new, used,reconditioned

• List with authorized users and person responsible.

One question that frequently arises is whether any type oftesting should be done as part of IQ. Functional andoperational testing belong to OQ. IQ should only includetests to verify that the software and hardware are installedproperly and that all electrical and fluid connections arecorrect. Therefore IQ should include switching on theinstrument and checking for any error messages. Correctloading of computer software should be checked bysuitable verification software. For a system that consistsof several modules, such as a modular HPLC system, IQcan include injection and qualitative evaluation of astandard. In this way the correct installation of all fluidand electrical tubings and cables can be checked.

Line between IQ and OQ

Equipment inventory

data base

Page 71: Good Laboratory practice and current good manufacturing practice

57

Part 2Impact of GLP/cGMP in the analytical laboratory

OQ should prove that the instrument is suitable for itsintended use. OQ is not required to prove that theinstrument meets the manufacturer’s performancespecifications. This is a frequent misunderstanding and wehave experienced that many operators prefer to use themanufacturer’s specifications because usually these arereadily available.

• Define intended functions to be tested.

• Define test cases and acceptance criteria. For an HPLCsystem such tests include precision of retention timesand peak areas, wavelength accuracy of UV detectors,gradient accuracy and precision, system carry over,baseline noise and detector linearity.

• Perform tests and compare the results with theacceptance criteria.

Validating a computer system can be a complex andexpensive task. It mainly depends on the complexity ofthe system. If a stand-alone computer controls andevaluates data from a single instrument, the computer canbe treated and tested as a module of the complete system.For example in chromatography, ‘critical functions’ such as instrument control, method sequencing, dataacquisition, peak integration, quantitation, data storageand retrieval and printing are all executed during thechromatographic equipment tests. It is advisable to listthese functions in the OQ protocol and classify them asbeing tested.

Additional tests should include limited system access toauthorized people through user I.D. and password. Testcases would be to use a wrong password/user I.D.combination. Other tests should include properfunctioning of electronic time, stamped audit trail andcorrect storage and retrieval of ‘meta data’.

Operationalqualification (OQ)

Steps for OQ

Additional steps for

computer systems

Page 72: Good Laboratory practice and current good manufacturing practice

58

Chapter 5Installation qualification (IQ) and operational qualification (OQ)

If a single computer can control and evaluate data frommultiple instruments, tests should include the ‘worst case’.This can be done by acquisition of data from themaximum number of specified instruments and at thehighest data acquisition rate. In this ‘worst case’ thesystem should not crash or lose data.

For networked systems the network functions should bespecified and tested. In complex networks, it is difficult totest all combinations. Develop test cases which arerepresentative for the network. It is a good practice to test each individual subsystem of the network beforetesting the network.

For a networked computer system, operationalqualification can mean, for example, verifying correctcommunication between the computers and peripherals.Data sets should be developed and input at one part of the network. The output at some other part should becompared with the input. For example, if a server is usedto secure and archive data from a chromatographic datastation, results should be printed on:

1. The chromatographic data system.

2. The server after storage and retrieval of the files.

The results should be compared, either manually orautomatically.

If the network links to other in-house systems, correctfunction of the linkage should be verified using well-characterized data sets.

Any program written in the user’s laboratory should bevalidated and documented by the user. The same goes forspreadsheet formulas. Correct functioning should betested using typical data and then outsider data.

It is out of the scope of this primer to give more advice.Relevant information can be found in reference 3.

Computer networks

Validation of home made

programs and

spreadsheets

Page 73: Good Laboratory practice and current good manufacturing practice

59

Part 2Impact of GLP/cGMP in the analytical laboratory

In principle, existing systems should be treated the sameway as new systems. Functional and performance oroperational requirements should be specified and criticalfunctions should be tested and compared withspecifications.

For computer systems, information on development andvalidation during development might not be available.Instead there should be a lot of experience on past use,which can be used to judge the quality and reliability of asystem.

There is still some uncertainty about specifics on OQ.Here we discuss the most frequently asked questions.

Before OQ testing is done, one should always considerwhat the instrument will be used for. Testing may be quiteextensive if the instrument is to be used for all types ofapplications and where some of these put high demandson the performance of the system. For example, if achromatograph is intended for use with certainapplications that work at low limits of quantitation (LOQ),and others require quantitation of large amounts, theinstrument’s capability to quantitate trace levels and largeamounts should be verified. In this case we recommendusing generic standards that test the instrument for itsgeneral purpose.

On the other hand, if the instrument is to be used for oneapplication only, the tests and acceptance criteria shouldbe limited to that application. In this case, the testcompound can be the same as the compounds analyzed inunknown samples.

OQ discussions

Selection of tests and

acceptance criteria

Selecting Parameters and Limits for

Equipment OperationalQualification

Ref. 21

OQ for existing systems

Page 74: Good Laboratory practice and current good manufacturing practice

60

Chapter 5Installation qualification (IQ) and operational qualification (OQ)

If a system comprises several modules, it is recommendedto perform system tests rather than performing testsmodule by module. Individual module tests should beperformed as part of the diagnosis if the system fails. Thisholistic approach for the validation of computerized HPLCsystems was first promoted by Furman and Layloff, twoUS FDA employees.22 For very complex computer systemslike complex networks, it is recommended to checkindividual subsystems before the network functions aretested.

The frequency of OQ tests depends not only on the type of instrument and the stability of the performanceparameters, but also on the specified acceptance criteria.In general, the time intervals should be selected such thatthe probability is high that all parameters are still withinthe operational specifications. Otherwise, analyticalresults obtained with that particular instrument arequestionable. Here the importance of proper selection of the procedures and acceptance limits becomes veryapparent. For example, if the baseline noise of a UV-visibledetector is set at the lowest possible limit, the lamp willhave to be changed more frequently than if it is set afactor of 5 higher.

Let’s assume the instrument is used for differentapplications, which means different samples, differentcolumns and different calibration standards. In this case itis recommended to use a generic standard for the sameinstrument category. We would also recommend using thesame approach if multiple instruments in a lab performdifferent applications. If there are just one or twoinstruments that run one type of application with onecalibration standard, it makes sense to also use thatstandard for OQ.

Module versus system

test

Frequency of tests

Test sample

Page 75: Good Laboratory practice and current good manufacturing practice

61

Part 2Impact of GLP/cGMP in the analytical laboratory

This is both a resource and business question, in additionto the technical aspect. In principle, this can be done bythe user and by the vendor. The technical question relatesto the procedure the vendor offers: does it check thecritical characteristics of the instrument? As long as testprocedures relate to the intended use of the instrument, it may be more economical if a vendor carries them out.The advantage for the user is that he or she does not have to be careful about the traceability of tools such as thermometers, because the vendor’s representativesupplies everything. Also, for whatever reason, someauditors prefer to see an OQ stamp on the equipment that comes from outside the user’s lab.

Preventive maintenance prior to an OQ reduces the risk of failing the test. However, when doing so there is noevidence that the instrument was performing properly allthe time. Before a decision is made, one should thinkabout the purpose of an OQ: is it proof that the equipmentdid and does perform according to specification all thetime, or should it make sure that the equipment is fit justfor future work? The answer to this question will alsoanswer the question if maintenance should be done before OQ.

Whenever any instrument is repaired, an OQ should bedone. The number of tests depends on the repair itself.Only those tests should be repeated which could beaffected by the repair itself. Instrument vendors shouldprepare a list with recommendations on what type of testsshould be repeated after which repair. Required testingshould be defined in an SOP.

If the building, the environment and scope of theinstrument remain the same, a full OQ is not necessary.Correct functioning and performance of parts that couldbe affected by the move should be verified. An example isto wavelength accuracy of variable wavelength detectors.If the instrument is moved to another building a full OQ ofequipment hardware is recommended.

Who can or should test

the vendor? the user?

a third party?

Preventive maintenance

before the OQ

OQ after repair

OQ after instrument

move

Page 76: Good Laboratory practice and current good manufacturing practice

62

Chapter 5Installation qualification (IQ) and operational qualification (OQ)

Why should I do OQ at

all, isn’t PQ enough?

The final question that arises is: why should I do OQ at allon a regular basis and why is PQ not enough? This is avalid question for many users. PQ has several advantages:it is done on a more frequent basis, and it is more specificto the user’s application. So, if the instrument is used justfor one or maybe only a few specific applications, and ifthe PQ tests include all relevant performance criteria, theregular OQ test may be omitted.

Page 77: Good Laboratory practice and current good manufacturing practice

63

Chapter 6Method validation

Page 78: Good Laboratory practice and current good manufacturing practice

64

Chapter 6Method validation

The ultimate objective of the method validation

process is to provide evidence that the method does

what it is intended to do, accurately, reliably and

reproducibly.

Method validation is the process for establishing thatperformance characteristics of the analytical method aresuitable for the intended application. Chromatographicmethods need to be validated or revalidated

• before their introduction into routine use

• whenever the conditions for which the method has beenvalidated change, for example in the case of instrumentwith different characteristics or samples with a differentmatrix

• whenever the method is changed, and the change isoutside the original scope of the method.

To obtain the most accurate results, all the variables of the method should be considered, including samplingprocedure, sample preparation, chromatographicseparation, detection and data evaluation, using the samematrix as that of the intended sample. The proposedprocedure must go through a rigorous process ofvalidation. All validation experiments used for makingclaims or conclusions about the validity of the methodshould be documented in a report.

The criteria for what constitutes a validatedchromatographic method has received considerableattention in the literature and from regulatory agencies.The International Conference on Harmonization (ICH) of Technical Requirements for the Registration ofPharmaceuticals for Human Use 23,24 has tried toharmonize criteria and methodology and developed aconsensus text on the validation of analytical procedures.The final text has been adopted by the Committee forProprietary Medicinal (CPMP) Products of the European

Suitability forintended

application

ICH:

Validation of analytical procedures:

Methodology

Ref. 23

Page 79: Good Laboratory practice and current good manufacturing practice

Part 2Impact of GLP/cGMP in the analytical laboratory

Union as CPMP/ICH/381/95 European Union in 1995. The Ministry of Health and Welfare (MOHW) in Japanimplemented it in 1997 and it was also adopted as a guideline by the US FDA.

In August 2000 the FDA released a draft guidance onAnalytical Procedures and Methods Validation.44 Theguidance provides recommendations to applicants onsubmitting analytical procedures, validation data, andsamples to support the documentation of the identity,strength, quality, purity, and potency of drug substancesand drug products. It is intended to assist applicants inassembling information, submitting samples, andpresenting data to support analytical methodologies

The parameters for method validation have been definedby different working groups of national and internationalcommittees and are described in the literature.Unfortunately some of the definitions are different in thedifferent organizations. An attempt at harmonization wasmade for pharmaceutical applications through theInternational Conference on Harmonization.23,24 Thererepresentatives from the industry and regulatory agenciesfrom USA, Europe and Japan defined parameters,requirements and, to some extent, also methodology foranalytical methods validation. Wiechert26 described theefforts and results of the harmonized method validation.In this chapter of the primer we will give a short summaryof parameters for method validation. More details havebeen discussed in reference 25.

The terms selectivity and specificity are often usedinterchangeably. A detailed discussion of this term asdefined by different organizations has been made byVessmann27. He particularly pointed out the differencebetween the specificity as defined by IUPAC/WELAC andICH (IUPAC: International Union of Pure and AppliedChemistry, WELAC: Western European LaboratoryAccreditation Conference).

Validation parameters

Validation of Analytical Methods:Review and Strategy

Ref. 25

Selectivity(specificity)

65

Guidance for Industry-Analytical Procedures and

Methods Validation

Draft, August 2000, 37 pages

Page 80: Good Laboratory practice and current good manufacturing practice

66

Chapter 6Method validation

Although inconsistent with ICH, the term specific

generally refers to a method that produces a response fora single analyte, while the term selective refers to amethod which provides responses for a number ofchemical entities that may or may not be distinguishedfrom each other. If the response is distinguished from allother responses, the method is said to be selective. Sincethere are very few methods that respond to only oneanalyte, the term selectivity is usually more appropriate.The USP monograph27 defines selectivity of an analyticalmethod as its ability to measure accurately an analyte inthe presence of interference, such as synthetic precursors,excipients, enantiomers and known (or likely) degradationproducts that may be expected to be present in the samplematrix. Selectivity in liquid chromatography is obtained bychoosing optimal columns and setting chromatographicconditions, such as mobile phase composition, columntemperature and detector wavelength.

It is a difficult task in chromatography to ascertainwhether the peaks within a sample chromatogram arepure or consist of more than one compound. While in thepast, chromatographic parameters such as mobile phasecomposition or the columns were modified, more recently,the application of spectroscopic detectors coupled on-lineto the chromatograph had also been suggested. UV-visiblediode-array detectors and mass-spectrometers acquirespectra on-line throughout the entire chromatogram. The spectra acquired during the elution of a peak arenormalized and overlaid for graphical presentation. If thenormalized spectra are different, the peak consists of atleast two compounds.

The principles of diode-array detection in highperformance liquid chromatography (HPLC) and theirapplication and limitations to peak purity are described in the literature29. Examples of pure and impure HPLCpeaks are shown. While the chromatographic signalindicates no impurities in either peak, the spectralevaluation identifies the peak on the left as impure. Thelevel of impurities that can be detected with this method

Page 81: Good Laboratory practice and current good manufacturing practice

67

Part 2Impact of GLP/cGMP in the analytical laboratory

depends on the spectral difference, on the detector’sperformance and on the software algorithm. Under ideal conditions, peak impurities of 0.05 to 0.1% can be detected.

The precision of a method is the degree of similarityamong individual test results when the procedure isapplied repeatedly to multiple samplings. Precision ismeasured by injecting a series of standards. According tothe ICH guidelines the measured standard deviation issubdivided into three categories, repeatability,

intermediate precision and reproducibility. Repeatablity

is obtained if the analysis is carried out in one laboratoryby one operator, using one piece of equipment over arelatively short time span. Intermediate precision is alsomeasured in one laboratory but over several days and/orusing different analysts.

Reproducibility is defined as the variability of themeasurement process in different laboratories withdifferent operators and different instruments. Thereproducibility standard deviation is typically two- tothreefold larger than that for repeatability.

The accuracy of an analytical method is the extent towhich test results generated by the method and the truevalue agree. The true value for accuracy assessment canbe obtained in several ways.

One alternative is to compare the results of the methodwith results from an established reference method. Thisapproach assumes that the uncertainty of the referencemethod is known. Secondly, accuracy can be assessed byanalyzing a sample with known concentrations, forexample, a certified reference material, and comparing themeasured value with the true value as supplied with thematerial. If such certified reference material is notavailable, a blank sample matrix of interest can be spikedwith a known concentration by weight or volume. Afterextraction of the analyte from the matrix and injection

Precision

Accuracy

Page 82: Good Laboratory practice and current good manufacturing practice

68

Chapter 6Method validation

into the analytical instrument, its recovery can bedetermined by comparing the response of the extract withthe response of the reference material dissolved in a puresolvent. Because this accuracy assessment measures theeffectiveness of sample preparation, care should be takento mimic the actual sample preparation as closely aspossible.

The concentration should cover the range of concern andshould particularly include one concentration close to thequantitation limit. The expected recovery depends on thesample matrix, the sample processing procedure and onthe analyte concentration.

The linearity of an analytical method is its ability to elicittest results that are directly, or by means of well definedmathematical transformations, proportional to theconcentrations of analytes in samples within a givenrange. Linearity is determined by a series of injections ofstandards at about six different concentrations that span50–150 % of the expected working range assay. Theresponse should be linearly related to the concentrationsof standards. A linear regression equation applied to theresults should have an intercept not significantly differentfrom zero. If a significant non-zero intercept is obtained, itshould be demonstrated that there is no effect on theaccuracy of the method.

The range of an analytical method is the interval betweenthe upper and lower levels (including these levels) thathave been demonstrated to be determined with precision,accuracy, and linearity using the method as written. Therange is normally expressed in the same units as testresults (for example percent, parts per million) obtainedby the analytical method.

Linearity

Range

Page 83: Good Laboratory practice and current good manufacturing practice

69

Part 2Impact of GLP/cGMP in the analytical laboratory

The limit of detection is the point at which a measuredvalue is larger than the uncertainty associated with it. It is the lowest concentration of analyte in a sample that can be detected, but not necessarily quantified. Inchromatography the detection limit is the injected amountwhich results in a peak with a height at least twice as highas the baseline noise.

The limit of quantification is the injected amount, whichresults in a reproducible measurement of peak areas(equivalent to amounts). Peak heights are typicallyrequired to be about 10- to 20-times higher than thebaseline noise.

Ruggedness is not addressed in the ICH documents.18,19

Its definition has been replaced by reproducibility whichhas the same meaning as ruggedness, as defined by theUSP to be: the degree of reproducibility of resultsobtained under a variety of conditions, such as differentlaboratories, different analysts, different instruments,environmental conditions, operators and materials.Ruggedness is a measure of reproducibility of test resultsunder normal, expected operational conditions fromlaboratory to laboratory and from analyst to analyst.Ruggedness is determined by the analysis of aliquots from homogeneous lots in different laboratories.

Robustness tests examine the effect that operationalparameters have on the analysis results. For thedetermination of a method’s robustness, a number ofmethod parameters, for example pH, flow rate, columntemperature, injection volume, detection wavelength ormobile phase composition are varied within a realisticrange, and the quantitative influence of the variables isdetermined. If the influence of the parameter is within apreviously specified tolerance, the parameter is said to bewithin the method’s robustness range.

Ruggedness

Robustness

Limit of detection

Limit of quantitation

Page 84: Good Laboratory practice and current good manufacturing practice

70

Chapter 6Method validation

Obtaining data on these effects helps to assess whether a method needs to be revalidated when one or moreparameters are changed, for example to compensate forcolumn performance over time. In the ICH document23

it is recommended to consider the evaluation of amethod’s robustness during the development phase, andany results that are critical for the method should bedocumented. This is not, however, required to be includedas part of a registration application.

The validity of a specific method should be demonstratedin laboratory experiments using samples or standards thatare similar to the unknown samples analyzed in theroutine. The preparation and execution should follow avalidation protocol, preferably written in a step by stepinstruction format. Possible steps for a complete methodvalidation are listed in the table below. This proposedprocedure assumes that the instrument has been selected,the method has been developed and meets criteria such asease of use, ability to be automated and to be controlledby computer systems, costs per analysis, samplethroughput, turnaround time and environmental, healthand safety requirements.

Strategy forvalidation of

methods?

Page 85: Good Laboratory practice and current good manufacturing practice

71

Part 2Impact of GLP/cGMP in the analytical laboratory

1. Develop a validation protocol, an operating procedure or avalidation master plan for the validation.

2. Define the application, purpose and scope of the method.3. Define the performance parameters and acceptance criteria.4. Define validation experiments.5. Verify relevant performance characteristics of equipment6. Qualify materials, e.g. standards and reagents for purity,

accurate amounts and sufficient stability.7. Perform pre-validation experiments.8. Adjust method parameters and/or acceptance criteria if

necessary.9. Perform full internal (and external) validation experiments.10. Develop SOPs for executing the method in the routine.11. Define criteria for revalidation.12. Define type and frequency of system suitability tests and/or

analytical quality control (AQC) checks for the routine.13. Document validation experiments and results in the

validation report.

Successful acceptance of the validation parameters andperformance criteria, by all parties involved, requires thecooperative efforts of several departments includinganalytical development, quality control, regulatory affairsand the individuals requiring the analytical data. Theoperating procedure or the validation master plan shouldclearly define the roles and responsibilities of eachdepartment involved in the validation of analyticalmethods.

Table 9

Page 86: Good Laboratory practice and current good manufacturing practice

72

Chapter 6Method validation

The scope of the method and its validation criteria shouldbe defined early in the process. These include:

• what analytes should be detected?

• what are the expected concentration levels?

• what are the sample matrices?

• are there interfering substances expected and, if so,should they be detected and quantified?

• are there any specific legislative or regulatoryrequirements?

• should information be qualitative or quantitative?

• what are the required detection and quantitation limits?

• what is the expected concentration range?

• what precision and accuracy is expected?

• how robust should the method be?

• which type of equipment should be used, is the methodfor one specific instrument or should it be used by allinstruments of the same type?

• will the method be used in one specific laboratory orshould it be applicable in all laboratories?

• what skills do the anticipated users of the method have?

The method’s performance characteristics should be basedon the intended use of the method. It is not alwaysnecessary to validate all analytical parameters that areavailable for a specific technique. For example, if themethod is to be used for qualitative trace level analysis,there is no need to test and validate the method’s limit ofquantitation, or the linearity over the full dynamic range ofthe equipment. Initial parameters should be chosenaccording to the analyst’s experience and best judgment.Final parameters should be agreed between the lab or theanalytical chemist performing the validation and the lab orthe individual applying the method.

Page 87: Good Laboratory practice and current good manufacturing practice

73

When is methodrevalidation

required?

Part 2Impact of GLP/cGMP in the analytical laboratory

Operating ranges should be defined for each methodbased on experience with similar methods, or they shouldbe investigated during method developments. Theseranges should be verified during method validation inrobustness studies and should be part of the methodcharacteristics. Availability of such operating rangesmakes it easier to decide when a method should berevalidated. A revalidation is necessary whenever amethod is changed and the new parameter is outside theoperating range. If, for example, the operating range of the column temperature has been specified to be between30 and 40 °C, the method should be revalidated if, forwhatever reason, the new operating parameter has beenselected as 41 °C. Revalidation is also required if thesample matrix changes and if the instrument type changes,for example if a brand with significantly differentinstrument characteristics is used. For example, arevalidation is necessary, if a high performance liquidchromatographic method has been developed andvalidated on a pump with a delay volume of 5 ml and thenew pump only has 0.5 ml.

Part or full revalidation may also be considered if systemsuitability tests or the results of quality control sampleanalysis are out of preset acceptance criteria and thesource of the error cannot be tracked back to instrumentsor anything else.

Page 88: Good Laboratory practice and current good manufacturing practice

74

Chapter 6Method validation

Once the method has been developed and validated, a validation report should be prepared that includes:

• objective and scope of the method (applicability, type)

• summary of methodology

• type of compounds and matrix

• all chemicals, reagents, reference standards, qualitycontrol samples with purity, grade, their source ordetailed instructions on their preparation

• procedures for quality checks of standards andchemicals used

• safety precautions

• a plan and procedure for method implementation frommethod development lab to routine

• method parameters

• critical parameters taken from robustness testing

• listing of equipment and its functional and performancerequirements like cell dimensions, baseline noise,column temperature range. For complex equipment apicture or schematic diagrams may be useful.

• detailed conditions on how the experiments wereconducted, including sample preparation. The reportmust be detailed enough to ensure that it can bereproduced by a competent technician with comparableequipment.

• statistical procedures and representative calculations

• procedures for quality control in the routine, like systemsuitability tests

• representative plots like chromatograms, spectra andcalibration curves

• performance data for method acceptance limit

• the expected uncertainty of measurement results

• criteria for revalidation

Validation report

Page 89: Good Laboratory practice and current good manufacturing practice

Guidance for Industry-Bioanalytical Method

Validation

Draft, May 2001, 25 pages

75

Bioanalyticalmethods

Part 2Impact of GLP/cGMP in the analytical laboratory

• the person who developed and initially validated themethod

• references, if there are any

• summary and conclusions

• approval with names, titles, date and signature of thoseresponsible for the review and approval of the analyticaltest procedure

The FDA has published a validation guidance forbioanalytical methods36. The information in the guidanceis generally applicable to gas chromatography or high-pressure liquid chromatography analytical methodsperformed on drugs and metabolites obtained frombiological matrices such as blood, serum, plasma, or urine. The guidance should also apply to other analyticaltechniques such as immunological and microbiologicalmethods or other biological matrices, such as tissuesamples including skin samples, although in these cases a higher degree of variability may be observed. Inaddition to parameters discussed earlier in this chapter,storage conditions should be determined to control thestability of the analyte in the matrix under study.

Page 90: Good Laboratory practice and current good manufacturing practice

76

Page 91: Good Laboratory practice and current good manufacturing practice

77

Chapter 7On-going performance

Page 92: Good Laboratory practice and current good manufacturing practice

78

Chapter 7On-going performance

“Performance Qualification” (PQ) is the process of

demonstrating that an instrument consistently

performs according to a specification appropriate

for its routine use.20

Most important in the definition of PQ is the word‘consistently’. PQ should ensure that the instrumentproduces reliable, consistent and accurate data on a day-by-day basis. Each laboratory should have acomprehensive preventive maintenance, that is well-understood, accepted and followed by individuals as well as by laboratory organizations, to prevent, detect and correct problems. The purpose is to ensure that theequipment is running without problems and that analyticalresults have the highest probability of being of acceptablequality.

On-going generation of accurate data is important tomaximize the efficiency of a laboratory. In developmentlaboratories data that are imprecise or inaccurate result inadditional work for re-analyzing and in wrong conclusionsbased on these results.

In pharmaceutical manufacturing, results that are out ofspecifications (OOS) initiate a failure investigation, whichcan be quite time consuming. Complete productionbatches must be held from release until the investigationis complete and concludes that the batch can still bereleased, again a big financial impact.

Overview andimportance

Page 93: Good Laboratory practice and current good manufacturing practice

79

Part 2Impact of GLP/cGMP in the analytical laboratory

Steps to ensure PQ can include:

• Preventive maintenance.

• Tests of critical functions, e.g., through systemsuitability tests or analysis of quality control samples.

• Instrument calibration.

• Analysis of blanks.

• Changes of hardware, firmware and software in acontrolled manner.

• Proper error recording and handling system.

• Participation in proficiency testing schemes.

• Training programs for new employees.

• Regular virus checks.

• Regular data back-up.

• Regular removal of unnecessary files, e.g., temporaryfiles to avoid data overflow.

The test frequency is much higher than for OQ. Anotherdifference is that PQ should always be performed underconditions that are similar to routine sample analysis. Fora chromatograph this means using the same column, thesame analysis conditions, e.g., mobile phase and detectorwavelength, and the same or similar test compounds.

PQ should be performed on a daily basis or whenever theinstrument is used. The test frequency not only dependson the stability of the equipment but on everything in thesystem that may contribute to the analysis results. For aliquid chromatograph, this may be the chromatographiccolumn or a detector’s lamp. The test criteria andfrequency should be determined during the developmentand validation of the analytical method.

Steps for PQ

Additional steps for

computer systems

Tests for PQ

Page 94: Good Laboratory practice and current good manufacturing practice

80

Chapter 7On-going performance

In practice, PQ can mean system suitability testing, wherecritical key system performance characteristics aremeasured and compared with documented, preset limits.For example, a well-characterized standard can beinjected five or six times and the standard deviation ofamounts are then compared with a predefined value. If thelimit of detection and/or quantitation is critical, the lamp’sintensity profile or the baseline noise should be tested.

For testing we would recommend the following steps.

1. Define the performance criteria and test procedures. Because of the high test frequency, the selection and automated execution of the test is of key importance.

2. Select critical parameters. For a liquid chromatography system this can be

• precision of the amounts

• precision of retention times

• resolution between two peaks

• peak width at half height or peak tailing

• limit of detection and limit of quantitation

• wavelength accuracy of a UV-visible wavelength detector.

Some of the parameters are related to the instrument and others to the column.

3. Define the test intervals, e.g.,

• every day

• every time the system is used

• before, between and after a series of runs

4. Define corrective actions on what to do if the system does not meet the criteria, in other words if the system is out of specification.

System suitability

parameters

Page 95: Good Laboratory practice and current good manufacturing practice

81

Quality control samples

and QC charts

Additional tests?

Part 2Impact of GLP/cGMP in the analytical laboratory

The analysis of quality control (QC) samples withconstruction of quality control charts has been suggestedas another way of performing PQ. Control samples withknown amounts are interspersed among actual samples atintervals determined by the total number of samples, thestability of the system and the specified precision. Theadvantage of this procedure is that the systemperformance is measured more or less continuously underconditions that are very close to the actual application.With suitable software the samples are automaticallyanalyzed and the results can be presented in a graphicalway as quality control charts.

As with system suitability testing, test procedures andacceptance limits should be specified during methodvalidation. Documented procedures should exist toinstruct the operators on what to do if the system does not meet the criteria.

A frequently discussed question is if, either systemsuitability testing or the analysis of QC samples aresufficient to prove on-going system performance, orwhether additional checks should be performed. Theanswer to this question depends very much on theconditions under which the control samples are analyzed.For example, if the system is used for trace analysis andthe amounts of the control sample do not include tracelevel amounts, the capability of the system to measure lowamounts should be verified. In HPLC, this could be aroutine check of the wavelength accuracy, the baselinenoise or the intensity of the UV lamp.

Page 96: Good Laboratory practice and current good manufacturing practice

82

Page 97: Good Laboratory practice and current good manufacturing practice

83

Chapter 8Data security, integrity, traceability

Page 98: Good Laboratory practice and current good manufacturing practice

84

Chapter 8Data security, integrity, traceability

Protecting the integrity, security, and traceability of

electronic records is most critical for any business

and regulatory environment. Success in complying

with new regulations such as the FDA’s 21 CFR

Part 11 (electronic signatures and records) hinges

on securing the authenticity and integrity of data

you generate.

Since the mid 90’s the FDA has paid a lot of attention todata integrity and authenticity. Several warning lettershave even been issued regarding this topic. Data integritybecame even more important in 1997 when 21 CFR Part 11was issued.30 With this regulation, electronic records andsignatures can be equivalent to paper records andhandwritten signatures. The regulation applies to allindustry segments regulated by the FDA that includesGood Laboratory Practice (GLP), Good Clinical Practice(GCP) and current Good Manufacturing Practice (cGMP).

Laboratories have to comply with Part 11 when threecriteria are present:

1.When computers are used to create, modify, maintain,archive, retrieve, or transmit data.

2.When at any time electronic records hit a durablestorage device.

3.When the laboratory intends to create records that areintended to be submitted to or required by the FDA.

For most analytical work numbers 1 and 2 apply, so theopen question is only with reference to number 3.Laboratories can decide to do signatures on paper, butthey have no choice on records. They must be keptelectronically. (Status as of January 2000).

21 CFR Part 11 –Electronic records

and signatures

Who has to comply

Page 99: Good Laboratory practice and current good manufacturing practice

85

Part 2Impact of GLP/cGMP in the analytical laboratory

The primary requirements of the regulation for analyticallaboratories are:

• Limited system access to authorized individuals.

• Use of validated existing and new computer systems.

• Secure retention of electronic records to instantlyreconstruct the analysis.

• User independent computer generated time-stampedaudit trails.

• Ensure system and data security, data integrity andconfidentiality through limited authorized systemaccess.

• Use of secure electronic signatures for closed and opensystems.

• Use of digital signatures for open systems.

Implementing the new rule will have a significant impacton the instrumentation, the work processes and on thepeople in analytical pharmaceutical laboratories:

• The current process of generating signatures should beevaluated (who has to sign what and when?).

• New procedures have to be developed in the companyand in the laboratory for limited authorized access tosystems and data (who can do what?).

• Computerized systems used for implementation must beupdated or replaced to ensure correct functionality.

• The manner of using and handling I.D. codes andpasswords as a basis for ‘legally’ binding signatures mayhave to be changed.

• New specialists, for example ‘electronic archivist, maybe required.

Primary requirements

Implementing21CFR Part 11 –

ElectronicSignatures and Records in

Analytical Laboratories

Ref. 31 and 32

21 CFR Part 11

Electronic Records andSignatures

Page 100: Good Laboratory practice and current good manufacturing practice

86

Chapter 8Data security, integrity, traceability

All computer systems used to generate maintain

and archive electronic records must be validated to

ensure accuracy, reliability, consistent independent

performance and the ability to discern invalid or

altered records.

This holds true for new as well as existing systems. It is basically nothing new for laboratories using computers ina regulated environment. Validating computer systems hasbeen very well described3,19 and most companies havedeveloped strategies for implementation. The problem liesnot as much with new or fairly new systems but more withthe older systems. They require a formal evaluation and astatement on their validation status. If they cannot bevalidated they cannot be used under 21CFR Part 11.

Procedures should be in place to generate accurate and

complete copies of records in both human readable and

electronic form suitable for inspection, review, and

copying by the agency. Records must be protected to

enable their accurate and ready retrieval throughout

the records retention period.

The FDA expects final results to be kept with the originaldata and the procedures for processing the data (‘metadata’). The FDA wants to be able to trace the final resultsback to the raw data using the same tools as the userwhen the data were generated. This is probably one of themost difficult requirement to implement. Knowing thatthey are subject to the predicate rule, the records must bekept for ten years or more, and computer hardware andsoftware have a much shorter lifetime, one can anticipateproblems with this paragraph.

One problem is to decide exactly which records should belogged and retained. The situation is most complex forquantitative chromatographic analyses. Usually inchromatography data acquisition, evaluation and printoutis done automatically using preprogrammed methods.However, occasionally the pre-programmed integrationmethod can be inappropriate which becomes obvious on

‘Meta data’ in chromatography

Integration parameters (threshold, area reject,

peak width)Calibration tables

Report layoutsPost-run macros

System validation

Secure retention of

electronic records to

instantly reconstruct the

analysis

Guidance for Industry-21 CFR Part 11, Electronic

Records; Electronic SignatureValidation, Draft, August 2001

Ref. 43

Page 101: Good Laboratory practice and current good manufacturing practice

87

Part 2Impact of GLP/cGMP in the analytical laboratory

the chromatogram and peak baseline printout. In this caseanalysts have to work with the raw data and adjustparameters to generate more appropriate measurementsof peak integrations. This is a manual iterative process,which frequently is subjective to the user. A few years agoit was sufficient to keep the original data and the finalresults together with the final method used to develop the final results. Now, the expectation is to keep allintegration methods in between as well.

A second problem is the availability of the recordsthroughout the retention period. The problem is not somuch the durability of storage devices such as CD-ROM’sbut more the computer hardware, operating systems andapplication software that is required to reconstruct theanalysis. If all this was available, it would be difficult tofind the people who could operate this old equipment.However, as Paul Motise stated at a conference in Berlin33:“The agency does not expect companies to save computerhardware and software for the sole purpose of recreatingevents. We anticipated that it would be possible to makean accurate and complete copy of those electronicrecords”. The expectation is that data and ‘meta data’ could be accurately converted to future systems.

Procedures should be in place to limit the access

to authorized users.

This can be ensured through physical and/or logicalsecurity mechanisms. Most companies already have suchprocedures in place. Typically users have to log on to asystem with user I.D. and password. Problems have beenreported with practical implementation in analyticallaboratories when computer controlled systems arecollecting data over time, especially when more than oneperson operates a computer at similar times usingdifferent applications and during a shift change in aroutine lab. Group users I.D.s. and passwords can be usedto log on the system, but unique identification throughindividual application specific passwords must beavailable for binding signatures with records.

Compliance Policy Guide:21 CFR Part 11;

Electronic records,Electronic Signatures

(CPG 7153.17)

Ref. 34

Limited system access

Deviations from part 11 are handled on a case by case basis

Page 102: Good Laboratory practice and current good manufacturing practice

88

Chapter 8Data security, integrity, traceability

An alternative to the autorization through the combinationpassword/user I.D. are biometric devices such as face orfinger print scanners. While these devices may be used inthe future, currently there are some problems withrobustness and accuracy.

Procedures should be available to use secure, computer-

generated, time-stamped audit trails to independently

record the date and time of operator entries and actions

that create, modify, or delete electronic records. Record

changes shall not obscure previously recorded

information.

This paragraph generates a lot of questions anddiscussions. The problem lies mainly in the manner inwhich it is implemented, especially which details arerecorded. For example, what should be recorded whengenerating a calibration table? Should each typing error berecorded when entering a compound name, should eachline be recorded when the return key is pressed or shouldthose entries be recorded only when the session is closedat the end of the calibration table entries? Too manyconfirmation steps will have an impact on the analyst’sproductivity. To implement a computer generated timestamped audit trail requires new software. Users ofcomputerized systems are advised to talk to the vendorabout possible upgrades with this function.

User independent

computer generated,

time-stamped audit trail

Page 103: Good Laboratory practice and current good manufacturing practice

89

Part 2Impact of GLP/cGMP in the analytical laboratory

The establishment of, and adherence to, written policies

that hold individuals accountable and responsible for

actions initiated under their electronic signatures, in

order to determine record and signature falsification.

This definitely requires not only the development ofprocedures but also behavioral changes on using I.D.codes and passwords. The barrier to share a passwordwith a colleague is usually much lower than to teachsomebody how to abuse a handwritten signature.However, in the sense of Part 11 both have the sameconsequence. A second problem is how to ensure theintegrity of records once they have been signed. This isonly possible with a well functioning audit trail.

Personnel in the lab should be trained on Part 11 and onthe meaning of electronic signatures. The training shouldbe documented and after the training attendees shouldsign a paragraph stating, for example: “I understand thatelectronic signatures are legally binding and have thesame meaning as handwritten signatures”.

Part 11 does not mandate electronic signatures. Signaturescan still be made on paper. Such systems are called hybridsystems. Companies should inform the FDA when theyintend using electronic signatures with a letter like:

“This is to certify that “My Company” intends that allelectronic signatures executed by our employees, agents,or representatives, located anywhere in the world, are thelegally binding equivalent of traditional handwrittensignatures.”

Use of secure electronic

signatures for closed and

open systems

Training

Hybrid systems

Page 104: Good Laboratory practice and current good manufacturing practice

90

Chapter 8Data security, integrity, traceability

Implementing the regulation on electronic signatures andrecords will have major consequences. This situation iscomparable with implementing Good Laboratory Practicesat the beginning of the eighties and validation in the firsthalf of the nineties.

They are as follows:

1. Define all work in your organization or laboratory thatwill fall under 21 CFR Part 11.

2. Form a working group with members from the ITdepartment, if existing, QA personnel and laboratorystaff.

3. Develop an implementation plan for your organizationand laboratory.

4. Decide whether you will use full electronic records andsignatures or hybrid systems, e.g., records in electronicand paper format. Report your decision to the FDA.

5. Create awareness for the rule among all employees,especially for the accountability of electronicsignatures.

6. Train the people in the organization and in thelaboratory on other contents and consequences.

7. Decide if all signatures in use today are really neededfrom a regulatory point of view.

8. Make an inventory of all computer systems.

9. Categorize computer systems into those• that must comply with Part 11• that are used to create critical data

Recommendations

Page 105: Good Laboratory practice and current good manufacturing practice

91

Part 2Impact of GLP/cGMP in the analytical laboratory

10. Look at the requirements of Part 11 and thefunctionality available in the system. Make a gapanalysis for each system for each required function

11. Make a risk assessment for those systems that arecurrently not compliant.

12. Based on the risk assessment develop a plan on how totake the lowest risk with minimal effort. The goal ofthe plan is to bring equipment into compliance withrequirements. The plan should include information onhuman and financial resources.

13. Document the plan and make sure that all departmentsand management buy into it.

Page 106: Good Laboratory practice and current good manufacturing practice
Page 107: Good Laboratory practice and current good manufacturing practice

93

Part 3Vendor contributions to GLP and GMP practices

The ultimate responsibility for validation lies with the user of a system. However, the vendor can help to be more efficientthrough providing validation products and services.

3

Page 108: Good Laboratory practice and current good manufacturing practice

94

Page 109: Good Laboratory practice and current good manufacturing practice

95

Chapter 9Vendor contribution

Page 110: Good Laboratory practice and current good manufacturing practice

96

Chapter 9Vendor contribution

As portrayed in previous sections, the burden of

responsibility for regulated facilities may appear

daunting. However, with a little foresight, you

can lighten the load with the assistance of the

laboratory equipment vendor.

The economic costs of running a facility to GLP/GMPstandards have been estimated at between 20 to 30 %higher than normal running costs, a lot of which stemsfrom the additional resources required to validateinstruments and methods, to audit data integrity and to archive data and ‘meta data’.

Some of the validation tasks can be shared with thevendor, development of SOPs for equipment qualificationfor example, and other tasks can only be performed withthe help of the vendor, validation during development for example. Some of the more recent regulatoryrequirements like computer generated time-stamped audit trail or the instant retrieval and replay of data years after the analysis was done, requires new softwarefunctionality. The feature set and availability of suchsoftware must be negotiated with the vendor.

Importance of thevendor

Page 111: Good Laboratory practice and current good manufacturing practice

97

Part 3Vendor contributions to GLP practices

During design qualification users must make sure that thedesign of the equipment meets user requirements.Individual steps include setting user requirement andfunctional qualification for the equipment.

A useful source are the specification sheets for softwareand hardware products which should be available fromthe vendor. Especially for complex chromatographysoftware. Not all functions that are typically included inthe equipment are required for the user’s application,therefore it does not make much sense just to completelycopy these vendor specifications into the user functionalspecifications. Otherwise, once they are in they alsoshould be validated. For example, if a software offers thecapability for a peak confirmation and peak purity analysisbased on spectral comparison, but the user only needs achromatographic signal for quantitation, peak purity andconfirmation should not be included in the userrequirement specifications. he recommendation is toselectively copy those functions that are also used lateron. Vendors can help by providing not only thespecifications but also the electronic files, which makes it easy to cut and paste.

Similarly for equipment hardware, vendors typically alsoprovide performance specification, for example, baselinenoise for an HPLC UV-visible detector or precision of peakarea for a complete HPLC system. Users should use thesetypes of specifications as guidelines but should use valuesthat are required by the application as the performancelimits for operational qualification.

Help during designqualification

Agilent provides clear functionalspecifications for its ChemStation andCerity data systems

Page 112: Good Laboratory practice and current good manufacturing practice

98

Chapter 9Vendor contribution

Users of equipment should have documented evidencethat the products have been developed and validated andthat they can be supported to ensure initial and consistenton-going quality.

Vendors can help by:

• Having a documented quality system in place that is inline with generally accepted quality standards, e.g.,ISO9001 or equivalent.

• For software vendors, an ISO 9001 certification thatshould be extended to comply with ISO9000-3.38

• Software development and validation should be in linewith practices required by the pharmaceutical industryand published by the Computer System ValidationCommittee of the US Pharmaceutical ManufacturerAssociation. Such practices include development andvalidation according to a software developmentlifecycle.

• Sending out a certificate or declaration on successfuldevelopment and validation with each product.

• Providing documentation on product development andvalidation procedures. This is usually done under a non-disclosure agreement.

• Retaining the source code and validation documents forpossible inspection at the developer’s site.

• Provide a certificate on safety and environmental issues.

• Supply, if requested, the credentials of staff involved inthe development, installation or maintenance ofequipment.

• Answering checklist questions from user firms preciselyand quickly. User firms can assist by avoidingunnecessary questions that in some cases have extendedthe number of questions to several hundred.

• Allowing the user company to perform an audit, if thismakes sense. User firms can make this more efficient bymaking vendor audits a corporate wide effort.

Assistance forvendor qualification

Agilent offers a CD-ROM providinginformation on software development and validation. A non-disclosureagreement is required.

Page 113: Good Laboratory practice and current good manufacturing practice

99

Part 3Vendor contributions to GLP practices

Installation of equipment can be performed by the vendor or by the user. Larger equipment such as mass-spectrometers are usually installed by the vendor, smallerones like simple pH-meters by the user. In any case thevendor should:

• Provide a clear requirement list for the site where theequipment will be installed. This includes requirementsfor gas and power supply, space requirements andenvironmental conditions such as humidity andtemperature. This list should be sent to the user ofequipment a few days or weeks prior to instrumentarrival.

• Provide a form to document and authorize theinstallation process.

• Offer installation and installation qualification as astandard or as an optional service.

For software the vendor should also provide:

• Validated software to verify an accurate and completecopy of software from CD-ROMS etc to the computer’shard disk.

• Services to perform installation qualification forcomputer systems.

Contribution atinstallation

Page 114: Good Laboratory practice and current good manufacturing practice

100

Chapter 9Vendor contribution

Operational qualification and requalification after changesare typically the most time consuming tasks. Despite allthe testing at the vendor’s site, OQ tests should also beperformed at the user’s site. Vendors can help through:

• Recommendations on what to test. Vendors shouldknow the best methods of testing their instrument sothat it will generate consistent, precise and accuratedata.

• Standard operating procedures for qualification tests.Preferably these tests and test protocols should be

offered in electronic format such that theuser, according to the intended use of theequipment, can easily customize them.

• Built in equipment hardware and firmware for convenient calibration, for example, holmium oxide filters for wavelength calibration of HPLC UV-visible detectors.

• Validated software to perform the tests automatically, to the extend which is possible. This speeds up the tests, reduces operator’s time, increases instrument uptime for sample measurement and ensures consistency of tests and documentation.

• Services to perform and document the operational qualification and requalification.

• The people who deliver the OQ service should have documented evidence about their qualification, e.g., training certificates.

Help for operationalqualification and

requalification

Verification file: C:\DATA\OQ_PV\VERIFY\1100.REG

Test Results Summary

Test Limit Measured Status Result

1. Flow Accuracy/Precision PassedAccuracy (%) 5 1.23 PassedPrecision (% RSD) 0.5 0.208008 Passed

2. Flow Accuracy/Precision PassedAccuracy (%) 5 0.75 PassedPrecision (% RSD) 0.5 0.340233 Passed

3. Temperature Accuracy PassedAccuracy at left [°C] 2 0.4 PassedAccuracy at right [°C] 2 0.16 Passed

4. Noise/Temperature Stability PassedASTM Noise [mAU] 0.04 0.010204 PassedWander [mAU] 0.2 0.022274 PassedDrift [mAU/h] 0.5 0.043427 Passed

5. Wavelength Accuracy PassedAccuracy at 1st max [nm] 2 1 PassedAccuracy at 2nd max [nm] 2 1 PassedAccuracy at minimum [nm] 2 1 Passed

6. Holmium PassedWavelength Accuracy [nm] 2 0.1 Passed

7. Injector Precision/Carry Over PassedPrecision Area [% RSD] 1 0.653937 PassedPrecision Height [%RSD] 2 0.192927 PassedCarry Over Area [% of prev] 0.2 0.014533 PassedCarry Over Height [% of prev] 0.4 0.021462 Passed

8. Response Linearirty PassedCorrelation [low limi[ 0.999 0.999893 Passed

9. Gradient Composition PassedRipple [%B] 0.5 0.101262 PassedAccuracy [%B] 0.7 2.279632 Passed

Print-out of an automated operational qualificationof the Agilent 1100 Series HPLC

Page 115: Good Laboratory practice and current good manufacturing practice

101

Part 3Vendor contributions to GLP practices

• Recommendations on what to test after maintenancerepair and upgrades of equipment.

• Recommendations on what to test after maintenance,repair and upgrades of computer hardware.

• Recommendations on what to test after upgrades ofoperating systems and application software.

Analytical methods used in GxP environment should bevalidated after development and after updates. Forstandard and compendial methods a laboratory mustdemonstrate the competence to run the method byrepeating most critical validation experiments. Similarly,proper functioning of a method should be demonstrated ina QA/QC laboratory after it has been transferred from thedevelopment laboratory.

Considering that one validation study can require up to 70experiments with up to 10 or more components runningsuch a study and evaluating the results can be quite timeconsuming.Vendors can contribute through offeringsoftware for automated method validation. This helps toimprove efficiency and consistent implementation.Software attributes should include• The software should be developed in a well documented

Software Quality Assurance environment• The vendor should provide documented evidence of

successful validation during and at the end of development

• The vendor should provide SOPs and services for IQ and OQ

• The software should be compatible with FDA regulations like 21 CFR Part 11. This means it should include functions like electronic audit trail, revision management, electronic signatures and electronic records with either handwritten or electronic signatures.

Help for methodvalidation

Agilent offers a CD-ROM providingsolutions for validation of analyticalmethods45.

Page 116: Good Laboratory practice and current good manufacturing practice

102

Chapter 9Vendor contribution

• The software should handle all major national andinternational regulations and guidelines for methodvalidation. Examples are ICH, USP, EP, BP and JP.

• The software should be seamlessly integrated into majorHPLC systems like the Agilent ChemStation Plus withthe 1100 Series HPLC

• The software should allow to enter variables, likenumber of injections for precision studies and numberof calibration points for linearity studies, and to includeadditional information, such as SOP's for samplepreparations.

• For robustness studies the software should allow thevalidation of methods across instruments, depending onthe scope even from different vendors, by providingmeans to import data from other data systems.

• The report should be flexible such that the user canchoose between table and graphical reports and reportsshould be easily transferred to word processingprograms like MS Word.

Performance qualification should be performed with thefully integrated systems, preferably using conditions asclose as possible to those used for sample analysis.

Vendors typically lack the experience to run instrumentsunder all these different conditions and they cannot offer

all the different compounds that arerequired for the tests. Therefore, typically,the users themselves performPerformance qualification by testing thecomplete system under sample analysisconditions. Vendors can contributethrough providing software for automatedtesting. PQ tests should be done on a dayto day basis. Performing and documentingsuch tests manually can take quite sometime. Using software, e.g., to run system

Help to ensure on-going

performance

Page 117: Good Laboratory practice and current good manufacturing practice

103

Part 3Vendor contributions to GLP practices

suitability tests or analysis of quality control samplesinterspaced with unknown samples using automatedgeneration of quality control charts can significantlyreduce the workload for these tests.

In addition to these PQ tests the vendor can significantlycontribute to get accurate, consistent and reliable resultson a daily basis through:

• Designing reliable instruments to increase uptime.

• Software and/or firmware to alert the user to carry outinstrument maintenance after a specified usage time ofmaintenance parts. For example, an HPLC detector canalert the user to exchange the detector lamp if the usagetime approaches a prespecified time. This is called earlymaintenance feedback (EMF).

• Fast response in case an instrument fails. This could bephone support to remotely diagnose and fix the problemand onsite service.

• Procedures and services for preventive equipmentmaintenance.

• A users feedback and response system that the user canuse to report instrument problems and to get feedbackeither about a fix or a work around solution.

• An electronic logbook to record unusual events.

For software and computers the vendor can help through:

• Procedures for file back up/recovery.

• Procedures and software to check limited andauthorized access to applications and the system.

Page 118: Good Laboratory practice and current good manufacturing practice

104

Chapter 9Vendor contribution

The FDA regulation 21 CFR Part 11 is probably theregulation that requires most help from the vendors. A good understanding of the regulation and itsinterpretation is a prerequisite for Part 11 support. Thesituation is similar to ten years ago with GLP and fiveyears ago with computer validation. A lot of uncertaintyexists.

While procedures for equipment qualification and even forvalidation of computer systems are now understood andcan be developed by the user, a computer generated time-stamped audit trail must be included in the software. Thesame holds for other requirements such as data integrityand long term archiving with the capability to instantlyreplay the data.

The vendor should provide software with functions for:

• Limited and authorized systems access through logicalsecurity controls. This is typically done throughselecting an appropriate operating system, Microsoft NT,for example.

• Limited and authorized access to selected tasks andapplications. This is typically done through theapplication itself. The I.D. of the person who performedthe tasks should be recorded. Based on this informationit must be possible to identify the person responsible forrecords generated at any time.

• Secure, operator independent electronic time-stampedaudit trail with information on who changed what. Toenter a reason for a change should be optional.

• The original data should not be overwritten during anyreprocessing procedure.

Help to ensuresystem security,

data integrity andtraceability

Using Agilent ChemStation Plus to ensure analytical data complies with FDA

21 CFR Part 11

Ref. 35

Page 119: Good Laboratory practice and current good manufacturing practice

105

Part 3Vendor contributions to GLP practices

• For systems with UV-visible diode-array detectors:preselecting signal and spectral acquisition modes toensure that only data that are relevant to the user’s workare transferred to the computer. This may be just one ortwo signals, or signals plus spectra when the peak elutesor all spectra. Be careful to store all spectra during theentire run.

• ‘Meta data’ like audit trail, integration parameters andcalibration tables should be stored together with the rawdata file in such a way that the originally obtained finalresults can be ‘instantly’ reconstructed from the rawdata.

• When vendors change to new systems, files including‘meta data’ should be converted to the new system suchthat the originally obtained final results can be ‘instantly’reconstructed from the raw data on the new system. If,for example, a chromatographic integrator algorithm hasbeen changed for the new system, the final numbersmay not be the same for all digits, but the newlycalculated result should be within the originalspecification. File conversion should be validated.

• The system should be able to generate electronicsignatures. The signature should include the name, timeand date, and the meaning of the signature.

• Electronic records either signed on paper orelectronically, should not be altered without electronicaudit trail.

• The software should allow to mirror a company’sprocedure on password handling, e.g., expiration date,character length and type of password.

Vendors should offer procedures and help for validatingfunctions as described above and as laid out in previouschapters of this primer. Validation procedures and servicesshould also be available for those software functions thatare required for Part 11, such as computer generated time-stamped audit trails.

Page 120: Good Laboratory practice and current good manufacturing practice
Page 121: Good Laboratory practice and current good manufacturing practice

107

Part 4Appendixes

You will find here a glossary and literature references

4

Page 122: Good Laboratory practice and current good manufacturing practice

108

Appendix AGlossary

Appendix AGlossary

The procedure by which an authoritative body givesformal recognition that a body is competent to carry outspecific tasks.

The degree of agreement of a measured value with theactual expected value.

Association of Official Analytical Chemists

American society for testing and materials.

The set of operations that establish, under specifiedconditions, the relationship between values indicated by a measuring instrument or measuring system, or valuesrepresented by material measure and the correspondingvalues of the measurand.

Co-Operation on International Traceability in AnalyticalChemistry. A forum for worldwide cooperationcollaboration on the mechanisms needed to ensure the validity and comparability o analytical data on a global basis.

Current Good Manufacturing Practice.

Collection of all regulations issued by U.S. governmentagencies. The individual titles making up the regulationsare numbered the same way as the federal laws on thesame topic. For example, the Federal Food, Drug, andCosmetic Act is found in Title 21 of United States Codeand the companion regulations implementing the law arefound in 21 CFR.

A system composed of computer(s), peripheral equipmentsuch as disks, printers and terminals, and the softwarenecessary to make them operate together (ANSI/IEEEStandard 729-1983).

Accreditation

Accuracy

AOAC

ASTM

Calibration

CITAC

cGMP

Code of Federal

Regulations (CFR)

Computer system

Page 123: Good Laboratory practice and current good manufacturing practice

109

Part 4Appendixes

A system that has a computer as a major, integral part.The system is dependent on the computer software tofunction

Environmental Protection Agency of the United StatesGovernment. A regulatory body who develops andenforces all aspects of environmental monitoring. Thisincludes development of analytical methods.

European Pharmacopeia, Official compendium of themember states of the Council of Europe, which includesall EC and EFTA countries.

Food and Drug Administration, U.S. agency, part of theDepartment of Health and Human Services, responsiblefor regulating clinical research and approval of marketingpermits for food, drugs, medical devices and cosmetics inthe U.S.

Good Automated Laboratory Practice.

Good Automated Manufacturing Practice.

Good Clinical Practice.

Good Laboratory Practice

Good Manufacturing Practice.

International Conference on Harmonization of TechnicalRequirements for Registration of Pharmaceuticals forHuman Use.

International Laboratory Accreditation Cooperation.Working for international acceptance of data generated byaccredited organizations. Developed the ISO Guide 25.

Documented verification that all key aspects of hardwareinstallation adhere to appropriate codes and approveddesign intentions and that the recommendations of themanufacturer have been suitably considered.

International Organization for Standardization. Agencyresponsible for developing international standards; over160 technical committees, 650 sub-committees and 1500working groups; more than 6000 ISO standards published;represents more than 90 countries. Founded in 1947.

Computerized system

EPA

EP

FDA

GALP

GAMP

GCP

GLP

GMP

ICH

ILAC

Installation

Qualification (IQ)

ISO

Page 124: Good Laboratory practice and current good manufacturing practice

110

Appendix AGlossary

Japanese Pharmacopeia, Official pharmacopoeia of Japan.

Limit of detection.

Limit of quantification.

National Institute for Standards and Technology in theUnited States.

Organization for Economic Cooperation and Development.

Documented verification that the equipment relatedsystem or subsystem performs as intended throughoutrepresentative or anticipated operating ranges

A service offered by Agilent’s Analytical Products Groupsupport organization. It verifies that the system at theuser’s site performs according to the specifications asagreed between the vendor and the purchaser.

Documented verification that the process and/or the totalprocess related system performs as intended throughoutall anticipated operating ranges.

Documented verification that the process and/or the totalprocess related system performs as intended throughoutall anticipated operating ranges.

Pharmaceutical Inspection Convention, a multinationalorganization (primarily of European countries) whosemembers have agreed to mutual recognition of facilityinspections for good manufacturing practice.

Pharmaceutical Manufacturers Association in the UnitedStates. A trade association that represents more than 100 firms, collectively producing more than 90 percent of American prescription drugs.

Action of proving that any equipment works correctly andactually leads to the expected results. The word validationis sometimes widened to incorporate the concept ofqualification.

JP

LOD

LOQ

NIST

OECD

Operational

Qualification (OQ)

Performance Verification

(PV)

Performance

Qualification (PQ)

Performance

Qualification (PQ)

PIC

PMA

Qualification

Page 125: Good Laboratory practice and current good manufacturing practice

111

Part 4Appendixes

A material or substance, one or more properties of whichare sufficiently well established to be used for calibratingan apparatus, assessing a measurement method or forassigning values to materials

Establishing documented evidence that a system doeswhat it purports to do based on review and analysis ofhistoric information.31

An indication of how resistant the process is to typicalvariations in operation, such as those to be expected whenusing different analysts, different instruments anddifferent reagent lots.

An original computer program in a legible form(programming language), translated into machine-readableform for execution by the computer.

Person in the laboratory responsible for the outcome ofthe GLP validation.

United States Pharmacopeia

Letter issued by U.S. Food and Drug Administration tomanufacturer containing adverse findings and giving themanufacturer 15 days in which to reply. It replaced theRegulatory Letter and the Notice of Adverse Findings.

Establishing documented evidence that provides a highdegree of assurance that a specific process willconsistently produce a product meeting its predeterminedspecifications and quality attributes

Confirmation by examination and provision of evidencethat specified requirements have been met.

Reference material

Retrospective validation

Ruggedness

Source code

Study director

USP

Warning Letter

Validation

Verification

Page 126: Good Laboratory practice and current good manufacturing practice

112

Appendix BLiterature

1.D.L.M. Weller, “GLP andQuality Assurance”, Anal.

Proc. 198, Vol 25, 199/200.

2.L.Huber, “Validation andQualification in AnalyticalLaboratories”, Interpharm,Buffalo Grove, IL, USA, ISBN 1-57491-080-9, Agilent part number 5965-0036,November 1998. Free samplechapters, ordering informationand book updates available from www.labcompliance.com.

3.L.Huber, “Validation ofComputerized Analytical andNetworked Systems” TextBook, IHS-Interpharm, ISBN1-57491-133-3, April 2002.

4.A.F. Hirsch, “Good LaboratoryPractice Regulations”, Marcel

Dekker, New York, 1989, ISBN 0-8247-8101-5.

5.Environmental ProtectionAgency (US) “FederalInsecticide, Fungicide andRodenticide Act (FIFRA): FoodLaboratory PracticeStandards”, Fed.Reg 48, Nr. 230, 53946-53969, Nov 29,1983, effective: May 2, 1984,United States Government Printing Office, Washington DC.

6.“Toxic Substance Control Act(TSCA): Good LaboratoryPractice Standards”, Fed.Reg 48, Nr. 230, 53922-53944, Nov 29, 1983,effective: December 1983,United States Government Printing Office, Washington DC.

7. “Automated LaboratoryStandards: Good laboratorypractice for EPA Programs”,June 1990, p5. EPA OIRMGALP publications, US EPA,Research Triangle Park, NC,USA.

8. “40 CFR Part 160: FederalInsecticide, Fungicide andRodenticide Act (FIFRA): FoodLaboratory PracticeStandards”, Fed.Reg 54, Nr. 158, 34053-34074, Aug 17,1989, effective: Oct 16, 1989,United States Government Printing Office, Washington DC.

9.“40 CFR Part 792: ToxicSubstance Control Act (TSCA):Good Laboratory PracticeStandards”, Fed.Reg 54, Nr. 158, 34034-34052, Aug 17, 1989,effective: Sept 18, 1989,United States Government Printing Office, Washington DC.

Appendix BLiterature

Page 127: Good Laboratory practice and current good manufacturing practice

113

Part 4Appendixes

10.Pharmaceutical Inspection ofConvention, “Guide to GoodManufacturing practice forPharmaceutical Products” Part 1.3 in International DrugGMP’s, June 1993,Interpharm Press, Inc., ISBN 0-935184-17-1.

11.Commision of the Council ofthe European Community,directives 87/18/EEC, 1987,88/320/EEC, 1988 and90/18/EEC 1990.

12.“International Drug G.M.P.’s”:editied by M.H. Anisfeld,Interpharm Press Inc. 1993,ISBN 0-935184-17-1.

13.United States Food & DrugAdministration, Guide toInspection of PharmaceuticalQuality control Laboratories,The Division of FieldInvestigations, Office ofRegional Operations, Office ofRegulatory Affairs, July 1993.

14. United States Food & DrugAdministration, Guide toInvestigating out ofSpecifications (OOS) TestResults, 1999.

15. United States Food & DrugAdministration / ORAInspectional References –Guide to Inspections of Quality Systems, August 1999.

16. E.H. Zaret, GMP Complianceand Trends, Part I,Pharmaceutical Formulation &Quality Control, May/June 1999, 23/24.

17. United States Food & DrugAdministration, Generalprinciples of validation,Rockville, MD, Center forDrug Evaluation and Research(CDER), May 1987.

18. United States Food & DrugAdministration’s Glossary ofComputer Systems SoftwareDevelopment Terminology(August 1995),www.fda.gov/ora/inspect-

ref/igs/gloss.html.

19. PDA Technical Report Number 31: Validation andQualification of ComputerizedLaboratory Data AcquisitionSystems – was prepared by thePhRMA CSVWG (now the PDAComputer Validation IssuesTask Group) and the PDAComputer Related Systems-Laboratory Systems TaskGroup. Order fromwww.pda.org.

Page 128: Good Laboratory practice and current good manufacturing practice

114

Appendix BLiterature

20. P. Bedson and M. Sargent, Thedevelopment and applicationof guidance on equipmentqualification of analyticalinstruments, Accreditation

and Quality Assurance, 1 (6), 265/274 (1996)

21. L. Huber and L. Welebob,Selecting Parameters andLimits for EquipmentOperational Qualification,Accreditation and QualityAssurance, 2 (7), 316/322(1997)

22. W.B. Furman, T.P. Layloff andR.F. Tetzlaff, Validation ofcomputerized liquidchromatographic systems,J.AOAC Intern, 77 (5),1314/1318 (1994).

23. International Conference onHarmonization (ICH) ofTechnical Requirements forthe Registration ofPharmaceuticals for HumanUse, Q2A: Validation of

analytical procedures, ICH-Q2A, Geneva, 1995.

24. International Conference onHarmonization (ICH) ofTechnical Requirements forthe Registration of Pharma-ceuticals for Human Use, Q2B:Validation of analytical

procedures: Methodology, ICH-Q2B, Geneva, 1996.25.

L. Huber, Validation ofAnalytical Methods: Reviewand Strategy, LC/GC International,February 1998, 96-105

26. E. Wiechert, Harmonizing theValidation of AnalyticalProcedures, BioPharm,August 1999, 18/24, and 51

27. J. Vessman, Selectivity orspecificity? Validation ofanalytical methods from theperspective of an analyticalchemist in the pharmaceuticalindustry, J.Pharm&Biomed

Analysis, 14 (1996) 867/869

28. General Chapter <1225>,Validation of compendialmethods, United States

Pharmacopeia XXIII, NationalFormulary, XVIII, Rockville,MD, The United States

Pharmacopeial Convention,

Inc, 1995, 1710-1612

Page 129: Good Laboratory practice and current good manufacturing practice

115

Part 4Appendixes

29. L. Huber and S. George, Diode-array detection in high-performance liquidchromatography, Marcel

Dekker, New York, ISBN 0-8247-4, 1993

30. United States Food & DrugAdministration, Code ofFederal Regulations, Title 21,Food and Drugs, Part 11"Electronic Records;Electronic Signatures; FinalRule; Federal Register 62 (54),13429/13466, 1997.

31. L. Huber, Implementing 21CFR Part 11 – ElectronicSignatures and Records inAnalytical Laboratories, Part 1: Overview andRequirements, BioPharm,November 1999, 28/34.

32. W. Winter and L. Huber,Implementing 21CFR Part 11 –Electronic Signatures andRecords in AnalyticalLaboratories, Part 2: Securityaspects for systems andapplications, BioPharm,January 2000

33. Paul Motise at the ECAConference: FDA andComputer in AnalyticalLaboratories, Berlin,September 1999.Question/Answers can bedownloaded fromwww.labcompliance.com/

conferences/august99.

34. United States Food & DrugAdministration, CompliancePolicy Guide: 21 CFR Part 11;Electronic records, ElectronicSignatures (CPG 7153.17)www.fda.gov/ora/compliance

_ref/cpg/cpggenl/cpg160-

850.htm

35. C. Nickel, Using AgilentChemstation Plus to ensureanalytical data complies withFDA 21 CFR Part 11, Agilent

Technologies technical note, publication number 5968-7930E,2000.

36. United States Food & DrugAdministration, BioanalyticalMethod Validation May 2001, 25 pages

Page 130: Good Laboratory practice and current good manufacturing practice

116

Appendix BLiterature

37. Shah, V. P., K. K. Midha, S. V. Dighe, et al., Analytical Methods Validation:Bioavailability, Bioequivalenceand Pharmacokinetic Studies(Conference report).Pharmaceutical Research,1992; 9:588/592.

38. ISO 9000-3:1991(E), Qualitymanagement and qualityassurance standards, Part 3:Guidelines for the applicationof ISO 9001 to the development,supply and maintenance ofsoftware, InternationalOrganization forStandardization, case postale56, CH-1211 Geneve 20,Switzerland.

39. L. Huber, National regulatoryagencies, published onLabcompliance in 1999,www.labcompliance.com/

regulations/national-

authorities.htm.

40. International Standard ISO/IEC17025: “General Requirementsfor the Competence of Testingand Calibration Laboratories”,2000

41. L.Huber “Validation MasterPlan”,www.labcompliance.com/books/masterplan.htm, April

2002

42. W.Winter and L.Huber“Implementing 21CFR Part 11– Electronic Signatures andRecords in AnalyticalLaboratories, Part 3: EnsuringData Integrity of ElectronicRecords”, BioPharm, March

2000

43. United States Food and DrugAdministration, “Guidance forIndustry, Electronic Records;Electronic SignaturesValidation”, Draft, August

2001, 24 pages

44.United States Food and DrugAdministration, “Guidance forIndustry, AnalyticalProcedures and MethodsValidation”, Draft, August

2000, 37 pages

45.“Validation of AnalyticalMethods: Regulatoryrequirements, strategy andnew automted software”,Agilent Technologies CD-

ROM, publication number5988-5819EN, April 2002.

Page 131: Good Laboratory practice and current good manufacturing practice

117

Page 132: Good Laboratory practice and current good manufacturing practice

www.agilent.com/chem

Copyright © 2000 Agilent Technologies

The information in this publication issubject to change without notice.

All rights reserved. Reproduction,adaptation or translation without priorwritten permission is prohibited, except as allowed under the copyright laws.

Printed in Germany 03/00Publication Number 5968-6193E

U4 sieh

e Date

i:

5968

-6193

_cov

er.qx

d


Recommended