Testing Computers Systems For FDA Mhra Compliance Computer Systems Val - ebooKOID

TESTING
COMPUTER
SYSTEMS FOR
FDA/MHRA
COMPLIANCE
David Stokes
Interpharm /CRC
Boca Raton London New York Washington, D.C.
Sue Horwood Publishing

Storrington, West Sussex, England
This edition published in the Taylor & Francis e-Library, 2005.

To purchase your own copy of this or any of Taylor & Francis or Routledges
collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.
/LEUDU\RI&RQJUHVV&DWDORJLQJLQ3XEOLFDWLRQ'DWD
&DWDORJUHFRUGLVDYDLODEOHIURPWKH/LEUDU\RI&RQJUHVV
7KLV ERRN FRQWDLQV LQIRUPDWLRQ REWDLQHG IURP DXWKHQWLF DQG KLJKO\ UHJDUGHG VRXUFHV 5HSULQWHG PDWHULDO LV TXRWHG ZLWK
SHUPLVVLRQDQGVRXUFHVDUHLQGLFDWHG$ZLGHYDULHW\RIUHIHUHQFHVDUHOLVWHG5HDVRQDEOHHIIRUWVKDYHEHHQPDGHWRSXEOLVK
UHOLDEOHGDWDDQGLQIRUPDWLRQEXWWKHDXWKRUDQGWKHSXEOLVKHUFDQQRWDVVXPHUHVSRQVLELOLW\IRUWKHYDOLGLW\RIDOOPDWHULDOV
RUIRUWKHFRQVHTXHQFHVRIWKHLUXVH
1HLWKHUWKLVERRNQRUDQ\SDUWPD\EHUHSURGXFHGRUWUDQVPLWWHGLQDQ\IRUPRUE\DQ\PHDQVHOHFWURQLFRUPHFKDQLFDO
LQFOXGLQJ SKRWRFRS\LQJ PLFUROPLQJ DQG UHFRUGLQJ RU E\ DQ\ LQIRUPDWLRQ VWRUDJH RU UHWULHYDO V\VWHP ZLWKRXW SULRU
SHUPLVVLRQLQZULWLQJIURPWKHSXEOLVKHU
7KHFRQVHQWRI&5&3UHVV//&GRHVQRWH[WHQGWRFRS\LQJIRUJHQHUDOGLVWULEXWLRQIRUSURPRWLRQIRUFUHDWLQJQHZZRUNV
RUIRUUHVDOH6SHFLFSHUPLVVLRQPXVWEHREWDLQHGLQZULWLQJIURP&5&3UHVV//&IRUVXFKFRS\LQJ
'LUHFWDOOLQTXLULHVWR&5&3UHVV//&1:&RUSRUDWH%OYG%RFD5DWRQ)ORULGD
7UDGHPDUN 1RWLFH 3URGXFW RU FRUSRUDWH QDPHV PD\ EH WUDGHPDUNV RU UHJLVWHUHG WUDGHPDUNV DQG DUH XVHG RQO\ IRU
LGHQWLFDWLRQDQGH[SODQDWLRQZLWKRXWLQWHQWWRLQIULQJH
9LVLWWKH&5&3UHVV:HEVLWHDWZZZFUFSUHVVFRP
E\&5&3UHVV//&
,QWHUSKDUPLVDQLPSULQWRI&5&3UHVV//&
1RFODLPWRRULJLQDO86*RYHUQPHQWZRUNV
,QWHUQDWLRQDO6WDQGDUG%RRN1XPEHU
/LEUDU\RI&RQJUHVV&DUG1XPEHU
ISBN 0-203-01133-3 Master e-book ISBN
Table of Contents
Authors Preface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
1
Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.1
What This Guideline Covers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.2
When Is This Guideline Applicable? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.3
Who Is This Guideline Intended For?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Why Do We Test? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3.1
Because the Regulators Require Us To... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3.2
Because the Quality Assurance Department Requires Us To... . . . . . . . . . . . . 5
3.3
Because Weve Always Done It This Way... . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.4
Because It Saves Money!. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
What to Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
4.1
GxP Priority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
4.2
Software/Hardware Category . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
4.3
Test Rationale and Test Policies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4.4
Testing or Verification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
The Test Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

5.1
Risk-Based Rationale. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
5.2
The Relationship between Test Specification(s) . . . . . . . . . . . . . . . . . . . . . . . 13
5.3
Integrating or Omitting the System Test Specification(s) . . . . . . . . . . . . . . . . 14
5.3.1 Hardware Acceptance Test Specification and Testing. . . . . . . . . . . . . . 15
5.3.2 Package Configuration Test Specification and Testing. . . . . . . . . . . . . 15
5.3.3 Software Module Test Specification and Testing . . . . . . . . . . . . . . . . . 15
5.3.4 Software Integration Test Specification and Testing. . . . . . . . . . . . . . . 15
5.3.5 System Acceptance Test Specification and Testing . . . . . . . . . . . . . . . 15
5.3.6 Integrating Test Specifications and Testing . . . . . . . . . . . . . . . . . . . . . 16
5.4
The Role of Factory and Site Acceptance Tests . . . . . . . . . . . . . . . . . . . . . . . 16
5.4.1 The Relationship between IQ, OQ and FATs and SATs . . . . . . . . . . . . 17
5.5
Roles and Responsibilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
5.5.1 Supplier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
5.5.2 User . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
5.5.3 Supplier Quality Assurance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
5.5.4 User Compliance and Validation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
5.5.5 Project Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
iii
iv
Testing Computer Systems for FDA/MHRA Compliance
5.6
5.5.6 Information Systems and Technology . . . . . . . . . . . . . . . . . . . . . . . . . 21

5.5.7 Supplier Software Test Team . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Relationships with Other Life Cycle Phases and Documents (Inputs and
Outputs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
5.6.1 Validation Plan and Project Quality Plan. . . . . . . . . . . . . . . . . . . . . . . 22
5.6.2 Design Specification(s). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
5.6.3 Tested Software and Hardware . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
5.6.4 System Test Specification(s) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
5.6.5 Factory/Site Acceptance Test Results and IQ, OQ and PQ . . . . . . . . . . 24
The Development Life Cycle of a Test Specification . . . . . . . . . . . . . . . . . . . . . . . 27

6.1
Recommended Phasing; Interfaces between and the Dependencies
of Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
6.2
Milestones in the Process. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
6.3
Inputs to the Development of a Test Specification . . . . . . . . . . . . . . . . . . . . . 29
6.4
Document Evolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
6.4.1 The Review Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
6.5
Constraints on the Development of a Test Specification . . . . . . . . . . . . . . . . . 31
6.6
Constraints on the Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
6.7
Conducting the Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
6.7.1 Test Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
6.7.2 Manual Data Input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
6.7.3 Formal Acceptance of Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
6.8
Outputs from the Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Recommended Content for System Test Specification(s) . . . . . . . . . . . . . . . . . . . 39

7.1
Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
7.1.1 Front Page/Title Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
7.1.2 QA Review Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
7.1.3 Scope of Document. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
7.2
General Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
7.2.1 Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
7.2.2 General Principles and Test Methodology . . . . . . . . . . . . . . . . . . . . . . 41
7.2.3 General Test Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
7.2.4 Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
7.3
Individual Test Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
7.3.1 Unique Test Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
7.3.2 Name of Hardware Item, Software Module or Function Under Test. . . 47
7.3.3 Cross Reference to Functional Description or Design Detail . . . . . . . . 48
7.3.4 Specific Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
7.3.5 Particular Test Methods and Test Harnesses. . . . . . . . . . . . . . . . . . . . . 48
7.3.6 Acceptance Criteria. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
7.3.7 Data Recording . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
7.3.8 Further Actions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
7.3.9 The Use of Separate Test Record Sheets . . . . . . . . . . . . . . . . . . . . . . . 52
Good Testing Practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

8.1
Prepare for Success . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
8.2
Common Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
8.3
8.4
8.5
8.6
8.7
8.8
8.9
8.10
8.11
8.12
8.13
8.14
8.15
8.16
8.17
8.18
8.2.1 Untestable Requirements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

8.2.2 Start Early. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
8.2.3 Plan for Complete Test Coverage . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
8.2.4 Insufficient Detail in the Test Scripts. . . . . . . . . . . . . . . . . . . . . . . . . . 56
8.2.5 Design Qualification Start When You Are Ready . . . . . . . . . . . . . . . 57
8.2.6 Taking a Configuration Baseline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
Testing in the Life Science Industries is Different . . . . . . . . . . . . . . . . . . . . . 58
Prerequisite Training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
An Overview of the Test Programme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Roles and Responsibilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
8.6.1 Test Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
8.6.2 Lead Tester . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
8.6.3 Tester . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
8.6.4 Test Witness (or Reviewer) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
8.6.5 Quality/Compliance and Validation Representative . . . . . . . . . . . . . . . 60
8.6.6 Test Incident Manager. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
Managing a Test Programme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
Checking Test Scripts In and Out . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
Recording Test Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
To Sign or Not to Sign. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
The Use of Test Witnesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Capturing Test Evidence (Raw Data) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
Proceed or Abort? (Test Incident Management) . . . . . . . . . . . . . . . . . . . . . . . 65
Categorising Test Incidents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Impact Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Test Execution Status. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
Test Data Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
Test Log-On Accounts (User IDs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
Supplier System Test Reports/Qualification Reports . . . . . . . . . . . . . . . . . . . . . . 69
10
The Use of Electronic Test Management and Automated Test Tools. . . . . . . . . . . 71

10.1 The Need for Test Tools in the Pharmaceutical Industry . . . . . . . . . . . . . . . . . 71
10.2 Test Tool Functionality. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
10.3 Electronic Records and Electronic Signature Compliance . . . . . . . . . . . . . . . 72
10.4 The Availability of Suitable Test Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
10.5 Test Script Life Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
10.6 Incident Life Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
10.7 Flexibility for Non-GxP Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
10.8 Project and Compliance Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
10.9 Testing Test Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
10.10 Test Record Integrity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
10.11 Features to Look Out For . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
11
Appendix A Hardware Test Specification and Testing . . . . . . . . . . . . . . . . . . . . 79

11.1 Defining the Hardware Test Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
11.2 Standard Test Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
11.3 Manual Testing of Component Hardware . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
11.3.1 Automated Test Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
vi
11.4
11.3.2 Burn-In/Heat Soak Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

11.3.3 Standard Integrated Hardware Tests . . . . . . . . . . . . . . . . . . . . . . . . . . 81
11.3.4 Automated Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
11.3.5 Hardware Acceptance Test Methods . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Performance Baseline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
12
Appendix B Package Configuration Test Specifications and Testing. . . . . . . . . 85

12.1 Defining the Package Configuration Test Strategy . . . . . . . . . . . . . . . . . . . . . 85
12.2 Configurable Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
12.3 Verifying the Package Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
12.4 Functional Testing of the Package Configuration . . . . . . . . . . . . . . . . . . . . . . 87
12.5 Stress Testing of the Package Configuration. . . . . . . . . . . . . . . . . . . . . . . . . . 87
12.6 Configuration Settings in Non-Configurable Systems . . . . . . . . . . . . . . . . . 88
13
Appendix C Software Module Test Specifications and Testing . . . . . . . . . . . . . 89

13.1 Defining the Software Module Test Strategy . . . . . . . . . . . . . . . . . . . . . . . . . 89
13.2 Examples of Software Modules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
13.3 Stress (Challenge) Testing of Software Modules . . . . . . . . . . . . . . . . . . . . . . 89
14
Appendix D Software Integration Test Specifications and Testing . . . . . . . . . . 91

14.1 The Purpose and Scope of Software Integration Testing . . . . . . . . . . . . . . . . . 91
14.2 System Integration Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
15
Appendix E System Acceptance Test Specifications and Testing . . . . . . . . . . . . 93

15.1 The Purpose of System Acceptance Testing . . . . . . . . . . . . . . . . . . . . . . . . . . 93
15.2 The Nature of System Acceptance Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
15.3 Establishing a Performance Monitoring Baseline . . . . . . . . . . . . . . . . . . . . . . 93
16
Appendix F Risk-Based Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
17
Appendix G Traceability Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

17.1 The Development of the Test Specifications. . . . . . . . . . . . . . . . . . . . . . . . . 100
17.2 The Development of the Test Scripts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
17.3 Test Execution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
17.4 Test Reporting and Qualification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
18
Appendix H Test Script Templates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

18.1 Basic Template for a Test Script . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
18.2 Example of a Specific Test Script . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
18.3 Example of a Test Script with Detailed Instructions . . . . . . . . . . . . . . . . . . . 109
19
Appendix I Checklists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113

19.1 Checklist 1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
19.2 Checklist 2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
19.3 Checklist 3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
19.4 Checklist 4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
19.5 Checklist 5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
19.6 Checklist 6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117

20
vii
Appendix J References and Acknowledgments. . . . . . . . . . . . . . . . . . . . . . . . . 119

20.1 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
20.2 Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
List of Tables
Table 4.1
Table 4.2
Table 4.3
Table 5.1
Table 6.1
Table 16.1
Table 16.2
Table 17.1
Table 17.2
Table 17.3
Table 17.4
Example of Software Testing Criticality. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

Example of Hardware Testing Criticality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Example of Test Approaches Based Upon Software or Hardware Criticality. . . 9
Summary of Testing Roles and Responsibilities . . . . . . . . . . . . . . . . . . . . . . . 19
Constraints on Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Example of System Risk Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Example of Test Approaches Based Upon Risk Factors . . . . . . . . . . . . . . . . . 97
Test Specifications Traceability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Test Script Traceability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Test Execution Traceability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
Test Reporting Traceability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
List of Figures
Figure 5.1
Figure 5.2
Figure 5.3
Figure 6.1
Figure 6.2
Figure 10.1
Figure 10.2
The Relationship between Test Specifications and Test Activities . . . . . . . . . . 14

Relationship between Design Specifications, Test Specifications, FATs,
SATs and IQ, OQ and PQ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Output Tested Hardware and Software as Inputs to Subsequent Tests . . . . . . . 24
The Dependencies: Various Life Cycle Documents and Activities . . . . . . . . . 28
The Evolutionary Development of Test Specification and Associated
Test Scripts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
The Test Script Life Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
The Test Incident Life Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
Authors Preface
This version of Testing Computer Systems For FDA/MHRA Compliance replaces and updates
four previous guides that specifically covered the topics of software module, software
integration, hardware and system acceptance testing. It consolidates much of the original
material on how to test, and includes valuable additional material on why we test, what to test,
and how to test. The MHRA (Medicines and Healthcare Products Regulatory Agency) was
formerly known as the MCA (Medicines Control Agency) and is based in London.
This version brings together current best practice in computer systems testing in the
regulatory environment specifically the pharmaceutical and related healthcare manufacturing
industries. We reference content from the latest GAMP 4 Guide [1] (Package Configuration, the
revised software and hardware categories and risk analysis) and show how the principles
detailed in GAMP 4 can be used to define a pragmatic approach to testing.
Much of this best testing practice has been established for a number of years, and many of
the basic ideas date back to the 1980s (and even earlier). Although the specific regulations vary
from industry to industry, the approach and ideas contained in this guideline can certainly be
used in other regulated sectors, such as the nuclear and financial industries.
In the two years since publication of the original guidelines the world of information
technology (IT) has continued to move forward apace. Despite the bursting of the dot.com
bubble some useful tools have emerged from the Internet frenzy and are now available for
testing of computer systems.
Most recent developments have been driven by the need to test large Internet based systems,
and some manufacturers have invested the time and money to provide automated test tools that
can be used in a manner which comply with the stringent requirements of regulations such as
21CFR Part 11 (Electronic Records and Electronic Signatures).
New content is included in this guideline, covering the compliant use of such tools, which
will be of specific interest and value to those companies and individuals thinking of investing
in such technology.
Additional thought has been given to trying to clarify the relationship and responsibilities of
the system user and supplier. This includes where testing starts in the project lifecycle, who
does what testing, where the lines of responsibility start and end and the differences in the
terminology used in the healthcare and general IT sectors.
We have tried to produce guidance that reflects the renewed approach of the FDA and other
regulatory agencies towards systematic inspections and risk-based validation with an underlying
scientific rationale. While the acceptability of some of the ideas put forward will no doubt be
subject to discussion in many Life Science companies, we hope the guide will prove to be a
valuable starting point
David Stokes, Spring 2003

ix
CHAPTER 1
Purpose
The purpose of this guideline is to:
Demonstrate the value of a systematic approach to computer systems testing (why we

test).
Provide a pragmatic method of determining the degree of testing necessary for any given
system (what to test).
Provide a detailed guide to the recommended contents of computer systems test

specifications and how to produce these in the most cost effective manner possible.
Show where computer system testing sits in the full validation life cycle and where the tests
sit in relation to the overall project.
Provide practical advice on how to conduct computer system tests (how to test).
Provide guidance on the use of automated test tools in a compliant environment.
CHAPTER 2
Scope
2.1 What This Guideline Covers

This guideline covers the following areas:
i.
ii.
iii.
iv.
v.
vi.
vii.
viii.
ix.
x.
The cost/benefits of conducting an appropriate degree of system testing.

A practical approach to determining exactly what is an appropriate degree of system
testing and how this can be justified (and documented) from a regulatory perspective.
The life cycle management relating to the development of Test Specifications and the
conducting of these system tests.
The roles and responsibilities of those involved with the development of Test
Specifications and the execution of these system tests.
The relationship between System Test Specification(s) and other project documentation.
The relationship between the system tests and other aspects of the project implementation.
Recommended content for inclusion in System Test Specification(s).
A traceability matrix defining how the System Test Specification(s) relate to the System
(design) Specification(s).
The selection, implementation and use of compliant automated test tools.
References and Appendices, including:
A checklist of questions to be used when developing System Test Specification(s)

Templates for documenting typical system test results
In this guideline the term System Test Specification(s) refers to any of the following separate
Test Specifications defined in GAMP 4:
Hardware Test Specification

Software Module Test Specification(s)
Software Integration Test Specification
Package Configuration Test Specification(s)
System Acceptance Test Specification
Further details on the specific purpose and content of such Test Specification(s) is given later
in this guideline, as well as other commonly defined testing such as Factory Acceptance Test
Specifications, Site Acceptance Test Specifications and so on.
2.2 When Is This Guideline Applicable?
This guideline can be used for any project where there is a requirement for system testing and
may be used to help test planning, Test Specification development, test execution, test reporting
and test management.
3

2.3 Who Is This Guideline Intended For?
This guideline is of value to:
Those involved with developing Validation Master Plans (VMP) and Validation Plans (VP)
Those involved with developing Project Quality Plans (PQP)
Those involved in reviewing and approving Test Specifications
Those responsible for developing System (Design) Specification(s) (to ensure the
testability of the overall software design)
Those involved with the development and execution of the System Test Specification(s)
Project Managers whose project scope includes system testing
CHAPTER 3
Why Do We Test?
There are a number of reasons given in answer to the question why do we test? Some of the
answers are more useful than others; it is important that anyone involved in testing understands
the basic reason why computer systems are tested.
3.1 Because the Regulators Require Us To
Testing is a fundamental requirement of current best practice with regard to achieving and
maintaining regulatory compliance. Although the need to test computer systems is defined by
certain regulations and in supporting guidance documents, the way in which computer systems
should be tested is not defined in detail.
Although the nature and extent of computer systems testing must be defined and justified on
a system by system basis, it is a basic premise that most computer systems will require some
degree of testing.
Failure to test will undermine any validation case and the compliant status of the system.
Where exposed, during regulatory inspection, this may lead to citations and warning letters
being issued and possibly a failure to grant new drug/device licenses, license suspension,
products being placed on import restrictions, etc.
Regulatory expectation is based on the premise that computer systems be tested in order to
confirm that user and functional requirements have been met and in order to assure data
integrity. These, in turn, are driven by a regulatory need to assure patient safety and health.
3.2 Because the Quality Assurance Department Requires Us To
The role of the Quality Assurance (QA) department (Department of Regulatory Affairs,
Compliance and Validation department, etc.) in many organisations is a proactive and supportive
one. In such organisations the QA department will provide independent assurance that regulations
are met and will help to define policies outlining the need for, and approach to, testing.
However, in some companies this may lead to a situation where the QA department becomes
responsible for policing the validation of computer systems and often defines the need to test
computer systems within an organisation. The danger here is that testing is conducted purely
because the QA department requires it other reasons for testing are not understood.
This QA role is defined at a corporate level and those organisations where the IT and
Information Systems (IS) departments and QA work hand-in-hand usually conduct the most
appropriate and pragmatic level of testing.
This is not always the case. In some organisations, one standard of testing may be inappropriately applied to all systems, simply because this has always been the approach in the past.
It is important that computer systems validation policies state and explain the need for testing,
rather than mandate an approach that must be followed, regardless of the system under test.
5

3.3 Because Weve Always Done It This Way
In many organisations there is a single standard or level of testing mandated for all.
However, one standard cannot be appropriately applied to systems that may range in scope
from a global Enterprise Resource Planning (ERP) system to a small spreadsheet. In this
guideline the term system covers all such systems including embedded systems. A scaleable,
cost effective and risk-based approach must therefore be taken, as defined in Section 4.1.
3.4 Because It Saves Money!
So far, the only justifiable requirement for testing is based upon meeting regulatory expectation;
if this were the only reason, industries not required to meet regulatory requirements would
possibly not test systems at all. There is, however, an overriding reason for testing computer
systems.
This primary reason for testing systems is that it is more cost effective to go live with
systems that are known to function correctly. Regulatory expectations are therefore fully in-line
with business benefits.
Most people involved with projects where there has been insufficient testing know that those
problems only exposed after go live will be the most time consuming and most expensive to
correct.
In many Life Science organisations there is political pressure to implement systems in
unrealistic timescales and at the lowest possible capital cost. This often leads to a culture where
testing is minimised in order to reduce project timescales and implementation costs.
Although this may often succeed in delivering a system, the real effect is to:
Reduce the effectiveness and efficiency of the system at go live.

Increase the maintenance and support costs.
Require a costly programme of corrective actions to be implemented, to correct faults and
meet the original requirements.
At worst, roll out a system which does not meet the basic user requirements.
The net effect is to increase the overall cost of implementing the system (although this may be
hidden on an operational or support budget) and to delay, or prevent the effective and efficient
use of the system.
When a system is appropriately tested it is more likely to operate correctly from go-live.
This improves user confidence and improves overall acceptance of the system (it is no
coincidence that system or user acceptance testing is an important part of the test process). The
system will operate more reliably and will cost less to maintain and support.
Senior management and project sponsors need to understand that testing is not an
unnecessary burden imposed by the regulators or internal QA departments. Proper testing of the
system will ensure that any potential risk to patient safety is minimised; one of the main
business justifications is that it will save time and money.
CHAPTER 4
What to Test
Having stated that a one-size-fits-all approach to system testing is no longer appropriate, the
challenge is to define a justifiable approach to testing; to minimise the time and cost of testing,
while still meeting regulatory expectations.
This comes down to the basic (and age-old) questions of:
How much testing to conduct?

What should be tested for?
Some systems are extremely complex and the concern of the regulatory agencies is that there
are almost infinite numbers of paths through the software. This stems from a concern that,
unless all paths through the software are tested, how can patient safety be assured under all
circumstances?
In large or complex systems it is practically impossible to test each path, but the reasoning
for not testing certain paths, options or functions is often made on an arbitrary basis. What is
needed is an approach that will allow testing to focus on areas of highest potential risk, but to
do so in a justifiable and documented manner
4.1 GxP Priority
Appendix M3 in GAMP 4 defines a methodology for determining the GxP Priority of a system.
More usefully, this approach can be used to determine the GxP Priority of specific functions in
a large or complex system.
In order to determine a sensible approach to testing a system it is useful to determine the GxP
Priority of the system or the GxP Priority of different parts (functions) of the system. This can
then be used in justifying the approach to testing. Different component parts of the system may
be more GxP critical than others, for example, Quality versus Financial functions. Assessing the
GxP criticality of each function allows testing to be focused on the areas of greatest risk. There
are other risks, which may need to be considered and these are discussed in Appendix F.
4.2 Software/Hardware Category
Appendix M4 in GAMP 4 defines categories of software and hardware.
With the exception of some embedded systems, most systems will be made up of software of
different categories. For instance, a system may consist of an Operating System (software
category 1) and a configurable application (software category 4).
Most systems will be based upon standard hardware (hardware category 1), although some
systems may be based upon custom hardware (hardware category 2).
7
Once the component parts of a system have been categorised they can be used to help
determine a justifiable approach to cost effective testing.
4.3 Test Rationale and Test Policies
Based upon a combination of the GxP criticality of a system or function, and the software and/or
hardware category of a system or function, it is possible to define a consistent and justifiable
approach to system testing.
Based upon the GAMP 4 GxP Priority and the software/hardware category of the system, a
consistent approach to testing can be documented. This may be in the form of a Test Strategy
Matrix and defined Test Approaches, examples of which are given below. There are also other
risk factors that should be borne in mind, and these are discussed in Appendix F.
Note that the examples shown are provided as a case-in-point only. An organisation may wish
to define their own corporate Testing Policy and a standard Test Strategy Matrix and Test
Approaches, based upon the principles given below. Other approaches and notation may be
defined (in the examples below S refers to software testing, H to hardware testing and F to
Functional Testing).
Once an organisation has agreed upon standard approaches to risk-based testing, they can be
used as the basis for defining system specific Test Strategy Matrices and Test Approaches for
the testing of individual systems.
Table 4.1 and Table 4.2 show how GxP Criticality and software/hardware category can be
cross-referenced to a test approach.
Table 4.1 Example of Software Testing Criticality
GAMP 4
Software
Category
1*
2
3
4
5
GAMP 4 GxP Criticality

Low
Medium
High
Test Approach F
Test Approach F
Test Approach F
Test Approach F
Test Approach S4
Test Approach F
Test Approach F
Test Approach F
Test Approach S2
Test Approach S5
Test Approach F
Test Approach S1
Test Approach S1
Test Approach S3
Test Approach S6
* No testing of Category 1 (Operating System) is required this is tested in situ with the application it is supporting.
Table 4.2 Example of Hardware Testing Criticality
GAMP 4
Hardware Category
1*
2
GAMP 4 GxP Criticality

Low
Medium
High
Test Approach F
Test Approach H1
Test Approach F
Test Approach H2
Test Approach F
Test Approach H3
* No testing of Category 1 (standard hardware components) is required this is implicitly tested by the integration
testing of the system.
Table 4.3 describes the testing and review required for each of the Test Approaches defined.
What to Test
Table 4.3 Example of Test Approaches based upon Software or Hardware Criticality
Test
Approach
Description
No specific hardware or software testing is required.

Hardware and software will be tested as part of overall System Acceptance Testing (Functional Testing).
S1
Software will be tested as part of overall System Acceptance Testing (Functional Testing).
Testing outside standard operating ranges is required in order to predict failure modes.
100% of System Acceptance Test Specifications and Results are subject to Quality function review and
approval.
S2
In additional to System Acceptance (Functional) Testing, software must be subject to stress testing during
normal operating conditions to challenge:
Basic system (log-in) access

User (role) specific functional access
System administration access
Network security
All Test Specifications and Results are subject to peer review.

50% of Package Configuration Test Specifications and 50% of related Results are subject to independent
Quality function review and approval.
100% of all System Acceptance Test Specifications and Results are subject to independent Quality
function review and approval.
S3
In addition to System Acceptance (Functional) Testing, software must be subject to comprehensive stress
testing across normal and abnormal operating conditions in order to challenge:

Network security

100% of Package Configuration Test Specifications and 100% of related Results are subject to
independent Quality function review.
function review.
S4
Software Module Testing is mandated prior to System Integration Tests and System Acceptance Testing.
Testing is required only within standard operating range.
25% of Software Module Test Specifications and 10% of all Software Module Test Results are subject
to independent Quality function review.
25% of all Software Integration Specification and related test Results are subject to independent Quality
function review.
function review.
S5
Software Module Testing mandated prior to System Integration Tests and System Acceptance Testing.
Testing only within standard operating range required for Software Module Tests.
Testing outside standard operating range required for Software Integration Tests in order to predict failure
modes.
function review.
function review.
10
Test
Approach
Description
S6
Software Module Testing mandated prior to System Integration Tests and System Acceptance Testing.
modes.
function review.
function review.
H1
No hardware specific testing required. Will be tested as part of overall System Acceptance Testing
(Functional Testing)
function review.
H2
Hardware assembled from custom components procured from single supplier requires hardware
integration tests to be performed to test adequate performance across all normal operating ranges. May
be conducted by supplier so long as acceptable documentary proof is provided.
Hardware assembled from custom components procured from multiple suppliers requires hardware
integration tests to be performed to test adequate performance across all normal operating ranges.
H2
50% of all Hardware Test Specification(s) and related test Results subject to independent Quality
function review.
function review.
H3
Hardware assembled from custom components procured from single supplier requires hardware
integration tests to be performed to test adequate performance across all normal operating ranges. Should
be witnessed by user representative if conducted by supplier.
Hardware assembled from custom components procured from multiple suppliers requires hardware
integration tests to be performed to test adequate performance across all normal operating ranges. Also
requires hardware integration tests to be performed outside normal operating ranges in order to predict
failure modes.
100% of all Hardware Test Specification(s) and related test Results subject to independent Quality
function review.
function review.
Such an approach may be used to justify the nature and level of both testing and review to be
applied to any individual system, or to the specific parts (functions) of a complex system.
However, the move away from full (100%) review of test specifications and test results by an
independent QA function needs to be justified within any organisation. For this to be accepted
as a risk-based approach to testing (validation), based upon a justifiable and rigorous scientific
approach, it is important to have proof that the integrity or quality of the testing process is not
compromised.
This can best be obtained by monitoring long-term trends in the testing process, and will
almost certainly require the QA department to monitor the efficacy and integrity of the peer
What to Test
11
review process, with subsequent traceable changes to the testing policy. This can be achieved by
comparing test statistics taken during the testing process and by sampling and reviewing a
random selection of test specifications and results subjected to a peer review process. As with
any sampling, there must be a scientific rationale for the sample size taken.
If the peer review process adversely impacts upon the quality or integrity of the testing
process, corrective actions must be taken. This may include further training for those involved
in testing, or the creation of a dedicated test team. If this does not improve the situation, the
policy may need to increase the level of QA review and approval, until such a time as acceptable
standards of peer review are achieved. When defining such an approach in a test policy, the
following key points must be borne in mind:
The method and rationale for revising the test policy and test approaches must be
explained, including the basis for any sampling.
It is better to start with a closely monitored peer review process and relax the QA review
over time, rather than initially remove all QA reviews and tighten up again at a later date.
Such a test policy can never mandate the scope, nature and level of testing for any specific
system. The policy should seek to provide consistent guidance and also identify
circumstances where a more thorough testing approach may be appropriate.
Note: more complex risk-based criteria can also be used to allocate different software modules
to appropriate testing approaches (see Appendix F Risk-Based Testing).
4.4 Testing or Verification
The terms testing and verification are often used when validating a system and the two are often
(incorrectly) used interchangeably. Testing is different from verification and there should be
clarity as to which parts of a system are to be subject to testing and which parts will be verified.
In simple terms, components that can be subjected to a repeatable set of input criteria, which
will produce a predictable and repeatable set of output criteria (results), can be tested. This
means that a test script can be written which defines both input criteria and expected output
criteria, and upon which the actual output criteria (results) may be recorded.
It may not be possible or practical to subject other components of the system to input criteria,
or it may be difficult to observe the resultant output criteria. In these cases it may be possible
to verify the correct operation of the system or component by other means.
As an example, consider a set of data being imported from a legacy system (about to be
decommissioned) into a replacement system.
Data migration from the old system to the new system will possibly involve data (format)
conversion, data export, data cleansing, data import and final data conversion. Software routines
can be written to perform all of these functions, but must be tested to ensure that they work in
a predictable and repeatable manner across a wide range of datasets. These tests should include
the use of out-of-range data, corrupted data and illegal data formats. This ensures that the
results of routines can be predicted and assured for all datasets that are to be migrated and that
any errors will be trapped and flagged.
Where a large number of complex datasets are to be converted, it is obviously cost effective
to develop and test such routines. This may not be cost-justified when there is only a single
dataset, which contains just 16 floating-point numbers.
In certain cases it will be more cost effective to verify the data in the new system. In the
simple case quoted above, this may involve a simple data import (entering the data directly into
the new system), and manually checking data in the new system against the data in the old
12
system (either on-screen or as hard copy). For data classified as medium or high GxP criticality,
this may be subject to independent checking by a second person. This manual process would not
test the data transport mechanism but would verify the results of the process.
In a similar way, other parts of a system build cannot be tested, but must be verified by
manual inspection. Examples of this may include:
Checking the version of an installed operating system.

Checking the serial numbers of hardware components installed within a system.
When considering what to test, it should be appreciated that, when it is impossible to test some
items, they must still be verified. Where testing is possible, but verification is the chosen route
(for reasons of cost effectiveness or efficiency), this should be justified as part of the test
strategy.
CHAPTER 5
The Test Strategy
Any approach to testing must be documented in order to demonstrate that an appropriate riskbased approach has been taken.
For small or simple systems the test strategy may be obvious at the start of the project. Where
it is possible to reference a corporate (division or site) policy, this may be included in the
Validation Master Plan (VMP), or possibly in the Project Quality Plan (PQP).
For larger or more complex systems it is useful to define a test strategy in a separate
document. This may be either a specific test strategy document, or a high-level qualification
protocol document. For very large or complex systems, multiple test strategy documents may
be produced, one for each level of testing in addition to an overall summary describing the
relationship between the various types of testing and associated test strategies.
For the purposes of this guide, the term test strategy refers to the testing rationale and
justification, whether this is included in the VMP, PQP, a separate document, or the installation
qualification (IQ), operational qualification (OQ), and performance qualification (PQ)
protocols.
Test strategies should include the following sections (where applicable).
5.1 Risk-Based Rationale
The approach taken to testing should be based on risk and should be included as part of the test
strategy. This includes GxP Priority (see Section 4.1) as well as other risks (see Appendix F
Risk-Based Testing).
5.2 The Relationship between Test Specification(s)
Depending upon the complexity and size of the system, different test specifications will be
needed. As we have seen above, GAMP 4 defines five types of test specification, namely:

Software Module Test Specification(s)
Package Configuration Test Specification(s)
Further specific information on each of these types of testing is given in Appendices 1 to 5.

Which of these types of test are needed depends upon the nature of the system (GxP Criticality,
software category and hardware category).
Guidance on which of these tests is required can be defined in a test strategy. GAMP 4
includes these types of testing in both the Documentation in the Life Cycle Model (Figure 8.1
13
14
in GAMP 4) and the Standalone Systems Lifecycle Activities and Documentation Model
(Figure 9.3 in GAMP 4).
Extracting the test specific documentation and activities from these models produces the
following diagram, which clearly shows the relationship between the various test
specification(s) and test activities.
The order in which these are shown should never be varied. The sequencing of the various
test activities should be defined as prerequisites in the test strategy. In summary these are:
All test specification(s) must be approved before the corresponding test activities
commence.
Any software module testing should be completed prior to the software integration tests
commencing. It should be noted that in large systems some parallel testing will take place.
As an example, integration testing may commence before all module testing is complete. It
is recommended that this is limited to informal testing; formal integration testing is not
performed until all software module testing is complete.
All hardware acceptance tests, package configuration verification, software module and
software integration testing must be complete, and signed off, prior to system acceptance
testing commencing.
5.3 Integrating or Omitting the System Test Specification(s)
Not all of these system test specifications will be needed for every system; some are optional,
depending upon the nature of the system being tested. For instance, if a new application is being
installed on an existing server, no hardware testing will be required. If there is no bespoke
(customised) software, no software module testing will be required. In the case of small or simple
systems, all the testing may be defined in a single test specification, which may include elements
of all hardware and software testing.
Where test specifications are omitted or integrated the reasons for this should be clearly
documented and the rationale justified in the test strategy.
Figure 5.1 The Relationship between Test Specifications and Test Activities.
The Test Strategy
15
5.3.1 Hardware Acceptance Test Specification and Testing

Hardware acceptance tests are only generally required for customised hardware (Hardware
Category 2). Where the system is based upon standard hardware in widespread use, hardware
specific acceptance testing is usually not required (although testing of connected hardware
components may be required). See Appendix A for further details
5.3.2 Package Configuration Test Specification and Testing
Certain types of system require configuration (the setting of various software parameters that
determine how the package functions) rather than programming. Typical examples of such
systems are Enterprise Resource Planning (ERP) systems and Laboratory Information Systems
(LIMS). Some systems combine a degree of configuration with traditional programming
(coding).
Any system that includes any degree of configuration setting should have a package
configuration specification and should be subjected to site acceptance testing (SAT) as a
minimum. For highly configurable systems, it is useful to verify the correct configuration of the
system prior to the SAT. It may also be possible to perform a lower level of package configuration
testing prior to the full SAT (see Appendix B for details).
5.3.3 Software Module Test Specification and Testing
Software module testing is normally required where customised or bespoke software modules
have been developed as part of the system or application (GAMP software category 5). This
may include customised components of systems that generally consist of category 3 or 4
software.
Where the system or application does not include customised software (lines of code, ladder
logic, etc.) then the software module test specification or testing may be omitted (see Appendix
C for further details).
5.3.4 Software Integration Test Specification and Testing
Systems constructed from multiple bespoke modules, or multiple standard software
components or packages, require a software integration test specification; thus providing
adequate proof that the various modules/components integrate in an acceptable manner, and that
integration is robust and functional.
Systems that consist solely of software categories 1, 2, 3 or 4 may require little or no software
module integration testing, so long as the component packages have a proven track record in the
Life Sciences market place. In this case the software integration test specification or testing may
be omitted. Package integration testing may still be required if individual software packages
have not been used in combination before.
If any of these packages have been modified (and is therefore treated as software category 5),
or if an unproven package is integrated as part of the overall solution, software integration
testing is required (see Appendix D for further details).
5.3.5 System Acceptance Test Specification and Testing
For simple or small systems, a separate phase of system acceptance testing may not be required.
This will normally be the case for systems comprised solely of software category 1, 2 and
16
possibly category 3 software. This is usually justified when the equipment or system is in
widespread use in the Life Science industries and is known to meet the defined business (user)
and functional requirements. Although some form of acceptance testing against user
requirements will still be required, a separate system acceptance test specification or phase of
testing may be omitted (see Appendix E for further details).
5.3.6 Integrating Test Specifications and Testing
In the case of small or simple systems, it is not cost effective or justifiable to produce separate
test specifications for each type of test. The various types of testing may be combined in a lesser
number of test specification documents, possibly combining the recommended content for each
type. In very small or very simple systems all levels of testing may be combined into a single
test specification.
Consolidation of test specifications is especially useful in the case of embedded systems,
where it may be difficult to separate software testing from hardware testing (and possibly
electrical and/or pneumatic testing etc.).
5.4 The Role of Factory and Site Acceptance Tests
Depending upon the scope of the project, Hardware Acceptance Testing, Software Module
Testing, Package Configuration Testing, Software Integration Testing and some System
Acceptance Testing may be performed at the suppliers premises known as factory acceptance
testing (FAT), or on site, known as site acceptance testing (SAT).
Many suppliers use the terms FAT and SAT to describe the standard testing they perform on
their systems or equipment and this is more common with suppliers to a broader range of
industries than just the Life Science industries. These are often contractual milestones, on which
a stage payment may be based (assuming successful completion). The following paragraphs are
provided to help explain how these tests may be leveraged to reduce the scope of any additional
(or duplicate) testing.
Usually the system will not be deemed as having been subject to system acceptance testing
until at least some of these tests have been performed in situ/on site. This is because the
Functional Testing of some system features can only be performed when the system is properly
installed in its final environment, with all interfaces and support infrastructure in place.
FATs are usually preceded by formal development testing this is part of the suppliers
software development life cycle/quality management system. Formal client testing may
commence with the factory acceptance test, but additional site acceptance testing is useful to
ensure that:
The system actually delivered to site is the system that was tested in the factory (by
checking hardware serial numbers, software version numbers, ID tags, etc.).
The system has suffered no damage during shipment that would adversely affect the
functional performance of the system.
System functions that can only be properly tested in situ can be performed.
Although unusual, it may be possible or necessary to omit any separate factory acceptance
testing and perform the full system acceptance test on site, as part of a site acceptance test. From
a project perspective this is not desirable, since testing on site is usually more time consuming
and more costly than factory testing. From a financial perspective, as much of the testing as is
practical should be performed as part of the standard factory acceptance test.
The Test Strategy
17
If appropriate, the System Acceptance Test Specification may be split into two parts, one
covering the Factory Acceptance Tests and one covering the Site Acceptance Tests.
Where both Factory Acceptance Testing and Site Acceptance Testing are performed, these
will have a relationship with the IQ and OQ as follows:
Factory Acceptance Testing will be performed first, executing as much of testing as is

practical.
The system will be shipped to the site and the Installation Qualification will be performed.
This usually relates to the system hardware (and possibly the firmware, operating system
and database software see Appendix A).
Site Acceptance Test will then be performed, executing the remaining content of the
System Acceptance Testing.
Operational Qualification will then be performed.
5.4.1 The Relationship between IQ, OQ and FATs and SATs

It should be noted that the Site Acceptance Testing and the Operation Qualification Testing
largely fulfil the same objectives (testing against the Functional Specification) and that these
may usefully be performed at the same time or combined.
Note that decisions on the need for Factory and Site Acceptance Testing, the timing of these
with the respect to the IQ and OQ and the possible combination of these may be taken in the
early stages of the project. This decision should be documented as part of the Validation Master
Plan or in the Project Quality Plan. If this is not the case, the relationship between Factory
Acceptance Testing and Site Acceptance Testing should be documented in the Test Strategy.
Wherever possible it is desirable to reference the Suppliers standard testing. Formal IQ and
OQ Reports may reference the Suppliers standard testing, which may be conducted as part of
standard Factory or Site Acceptance Testing. This may significantly reduce the scope of
additional or duplicate user testing and assumes that the level of documented evidence is
sufficient to support the validation case, which is in turn dependant upon the GxP criticality of
the system (risk).
The relationship between the various Design Specifications, Test Specifications, FATs and
SATs is shown in the following diagram (Figure 5.2).
Figure 5.2 shows that:
The development of Test Specifications takes place at the same time as the correspondent
Design Specification (this is of course done by a separate team). This reduces the project
implementation time scales and helps ensure that the Functional and Design Specifications
are testable.
Hardware Acceptance Testing is more likely to take place as part of the FAT, but some
elements of hardware testing may only be completed in situ, on site, during the SAT.
Software Module Testing and Package Configuration Testing are more likely to take place
as part of the FAT but some may only be completed on site during the SAT.
Software Integration Testing starts during the FAT, but some of this can only be conducted
on site during the SAT.
The results of the Hardware and Software Module Testing can all be referenced or
summarised as part of the Installation Qualification.
The results of the Package Configuration and Software Integration Testing can all be
referenced or summarised as part of the Operational Qualification.
Some System Acceptance Testing may be conducted as part of the FAT, but many
Acceptance Tests can only be conducted as part of the SAT.
18
Figure 5.2 Relationship between Design Specifications, Test Specifications, FATs, SATs and IQ, OQ and PQ.
The results of the System Acceptance Testing can be referenced or summarised as part of
the Operation Qualification or Performance Qualification, depending upon the exact
nature of the tests concerned.
Note that these interrelationships are summarised in a traceability matrix in Appendix G

Traceability Matrices.
5.5 Roles and Responsibilities
The usual roles and responsibilities associated with the preparation of the System Test
Specifications and the conducting of the associated tests should be defined in the Test Strategy,
as listed below. Note however, that these roles and responsibilities may be changed or shared,
according to the specific requirements of a project.
Specifically, the role of supplier may be fulfilled by a user internal function such as IT,
Information Systems, Internal Support or an Engineering group.
The Test Strategy
19
In addition, the contractual relationship and/or a good long term working relationship may
allow the supplier to assume more of the responsibilities usually associated with the role of the
user.
There is also an opposite to this situation, where there is a new working relationship, or where
the Validation Plan requires the user to put additional validation activities in place in make up
for deficiencies in the suppliers quality system. In this case the user may perform more of the
suppliers traditional role, or it may require that the user conduct more tests than would usually
be the case.
The key roles and responsibilities are usually assigned as summarised in Table 5.1 and in the
corresponding explanatory text.
Table 5.1 Summary of Testing Roles and Responsibilities

Supplier
QA
Develop Test Policy

Develop VMP (VP)
Review and Approve
VMP (VP)
Develop PQP
Review and Approve
PQP
Develop and Review
Test Strategy
Approve Test Strategy
Develop Test Specs
Review Test Specs
Approve Test Specs
Prepare for Tests
Conduct Tests
Support Tests
Monitor Tests
Review and Approve
Test Results
IS/IT
Test
Team
User
PM
Validation
IS/IT
Project
Team
PM
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
5.5.1 Supplier
It is the responsibility of the supplier to:
Develop the Project Quality Plan that identifies the need for supplier specific testing.
Develop the Hardware Test Specification (if appropriate).
Develop the Software Module Test Specification (if appropriate).
Develop the Software Integration Test Specification (if appropriate).
Develop the Package Configuration Test Specification (if appropriate).
Develop the System Acceptance Test Specification (if appropriate).
Physically prepare for the actual tests.
Conduct the appropriate tests (including recording the results and any retesting as
required).
20
5.5.2 User
It is the responsibility of the user to:
Define the need for the various System Test Specification(s) (usually in the Validation
Plan).
Physically prepare for those tests that will be performed on site.
Assist with those System Acceptance Tests that will be performed on site.
Witness the System Acceptance Tests (and any others that may need to be witnessed).
This may be no more that identifying that the System Test Specification(s) are a deliverable of
the supplier and the user may choose to delegate all further responsibility to the supplier. This
may be acceptable in the case of a reputable supplier with whom the user has worked before.
The System Acceptance Tests are the first major test of overall functionality of the system
and it is usual for the user to witness the System Acceptance Tests in order to verify that the
system to be supplied meets the agreed Functional Specification.
Where a user chooses not to witness some or all of the System Acceptance Tests the following
may suffice as an acceptable alternative:
Review and/or approve the final System Acceptance Test Specification prior to the System
Acceptance Tests commencing.
Review the results of the System Acceptance Tests and associated documentation as part of
the Operational Qualification.
Where the supplier audit has revealed deficiencies in the suppliers testing regime the user may
choose to review/approve other Test Specifications and/or witness additional tests (either at the
premises of the supplier or on site). These may include the Software Module Tests, the Software
Integration Tests, the Package Configuration Tests or the Hardware Tests.
5.5.3 Supplier Quality Assurance
It is the role of the suppliers Quality Assurance function to ensure that:
The System Test Specification(s) are written in accordance with the Project Quality Plan.
The System Test Specification(s) are approved (by the supplier and, if required, the user)
prior to the corresponding level of System Testing commencing.
The System Tests are conducted in accordance with the requirements of the corresponding
Test Specification, including the recording of results.
System Tests are only performed once the prerequisite Tests have been completed and
signed off.
The completed System Tests are fully signed off by appropriate supplier (and possibly user)
personnel.
5.5.4 User Compliance and Validation

It is the role of the users Compliance and Validation (C &V) function to ensure that:
The Test Strategy is appropriate to the GxP criticality of the system and the size and
complexity of the system
The Test Strategy
21
The need for the various System Test Specification(s) (and associated tests) are clearly
defined in the Validation Plan or Test Strategy, along with the scope and outline content of
the Specification (or the reasons for omitting or combining them).
The need for and nature of reviewing and approving System Test Specification(s) and
witnessing the System Tests by the user are clearly defined in the Validation Plan (the
rationale for the review? Who reviews it? When do they review it? How do they review it?
How they formally reject/accept? Who witnesses the tests? Who formally accepts the
results of the tests?).
The System Test Specification(s) are traceable to the Validation Master Plan, the Project
Quality Plan, the Test Strategy and the corresponding System (Design) Specification.
The level of user involvement in conducting/signing off the System Tests is clearly defined
and justified in the Validation Plan.
The acceptable level of System Test documentation is clearly defined in the Validation Plan
or Test Strategy (details required, authorised signatories allowed to sign off the tests etc).
The need to review the System Test documentation as part of the Qualifications (IQ, OQ
and PQ) and the degree of the review is clearly defined in the Validation Plan or Test
Strategy (including who reviews it, when and how they review it and how they formally
accept or reject it).
5.5.5 Project Manager

It is the role of the users and suppliers Project Managers to ensure that:
All of the documentation required by the users Validation Plan, the suppliers Project
Quality Plan and the Test Strategy is developed in a timely and properly sequenced manner
and to the required standard:
The System Test Specification(s)

The System Test Sheets(s)
The System Test Result(s)
Incident Reports (if required)
All hold points are properly observed, and that user reviews are conducted before moving
on to subsequent (dependant) phases of the project life cycle.
The review of System Test Specification(s) is conducted prior to conducting the
corresponding System Tests.
The System Tests are conducted in a timely manner, all results are properly recorded and
any necessary retests performed and signed off prior to moving onto subsequent tasks.
Testing integrity is not compromised due to budgetary or time constraints.
5.5.6 Information Systems and Technology

It is the role of the suppliers Information Systems and Technology function to ensure that:
The necessary test facilities and infrastructure are available to allow the System Tests to be
conducted (i.e. network infrastructure, printers, test equipment, simulation software).
The System Tests are properly supported as required (with regards to resources, facilities,
witnesses etc.).
It is the role of the users Information Systems and Technology function to ensure that:
22

The necessary test facilities and infrastructure are available to allow the Site Acceptance
Tests to be conducted (i.e. network infrastructure, printers, test equipment, simulation
software).
5.5.7 Supplier Software Test Team

It is the role of the suppliers software testing function (the Test Team) to ensure that:
The System Test Specification(s) are developed in a timely manner, and in accordance with
the requirements of the users Master Validation Plan and the suppliers Project Quality Plan.
The System Test Specification(s) are submitted to internal review and approval as per the
suppliers Project Quality Plan (and if required, by the user as per the users Validation
Plan).
The System Test Specification(s) are traceable to the corresponding System (Design)
Specification, the users Validation Plan and the suppliers Project Quality Plan.
Formal System Tests are conducted in a timely manner, in accordance with the
corresponding System Test Specification(s).
The results of all formal System Tests are recorded in accordance with the requirements of
the corresponding System Test Specification(s).
Any necessary retesting is conducted in a timely manner, in accordance with the
requirements of the System Test Specification(s).
All System Tests are signed off in accordance with the requirements of the System Test
Specification(s).
Incident reports are generated for any exceptional results or circumstances that are likely
to have a wider knock-on effect and will need further consideration.
Note that it is good testing practice on large projects for one set of developers or engineers to
develop the System (design) Specification(s), a different team to develop the System Test
Specification(s) and possibly a third, independent team to conduct the actual tests.
This ensures that the System Testing is sufficiently thorough and that the expectations and
preconceptions of software designers will not impact upon the conducting of the tests.
This is not always possible on smaller projects, but the preparation of good quality System
Test Specification(s) will minimise any negative impact from using the same developers/
engineers to both develop and test the functional design.
5.6 Relationships with Other Life Cycle Phases and Documents (Inputs and Outputs)
Figure 5.1 shows the relationship between the various validation and development life cycle
phases and documents. Where appropriate the Test Strategy should clarify the project specific
relationships. The various system Test Specifications are related to other documents in the life
cycle and either use information from those documents as input (reference) data, or are in turn
referred to by other documents and therefore provide output (result) data.
These related phases and documents are:
5.6.1 Validation Plan and Project Quality Plan
Where they include the Test Strategy, the Validation Plan or Project Quality Plan should
explicitly indicate which specifications and corresponding Test Specifications should be
produced (and tests conducted).
The Test Strategy
23
As a minimum, the Validation Plan should refer to the process of auditing the supplier to
ensure that supplier tests are properly conducted and may also reference a supplier audit report
that indicates the general nature and scope of these tests.
However, at the time of writing the Validation Plan for a large or complex system it is unlikely
that the user will have sufficient idea of the system to be used to be able to define the tests in
much detail (unless the user is implementing the system themselves).
The detailed requirements of the Test Specifications will more usually be deferred to the
Project Quality Plan. In the case of large or complex projects a separate Test Strategy document
may be produced, or the content may be included in the IQ, OQ and PQ Protocols.
5.6.2 Design Specification(s)
The various Design Specification(s) provide an increasing level of detail regarding the function
and design of the system.
Design Specifications should be written in a structured manner, so that each separate function
of the system is clearly described and can be individually tested. They should contain explicit,
concrete details of functionality and design which can be tested and pass/fail criteria should be
clearly identified rather than implied).
As an example The system will interface to a Schmidt Model 32X bar code scanner, capable
of scanning and identifying 15 pallets per minute on a continuous basis rather than The system
shall be capable of identifying a large number of pallets.
When writing Design Specifications it is useful if:
The corresponding Test Specification is written in parallel, almost certainly by a different

person or team.
A member of the test team reviews the Design Specification.
Both of these steps help ensure that functional and detailed design requirements are testable.
The relationship between a Design Specification, the corresponding Test Specification, and
the Test Scripts should be identified as linked configuration items in whatever Configuration
Management system is used. This ensures that if one document is changed the other(s) will be
identified as part of any change control impact analysis and noted as requiring review and
possible modification.
5.6.3 Tested Software and Hardware
Once tested at a low level (hardware and software module), the system hardware and software
are a direct input to subsequent System Tests.
The tested software modules are outputs from the System Module Tests and inputs to the
System Integration Testing.
The hardware is an output from the hardware testing and is an input to the System Acceptance
Testing, along with the tested software.
Prior to conducting the System Acceptance Tests the system software should have
successfully completed a thorough and challenging Software Integration Test. Likewise, the
system hardware should have successfully completed Hardware Acceptance Testing. The
24
Figure 5.3 Output Tested Hardware and Software as Inputs to Subsequent Tests.
purpose of the System Acceptance Test is to bring the actual system software and hardware
together, and prove the overall functioning of the system in line with the requirements of the
Functional Specification.
Since the previous Software Module and Software Integration Testing will have been
conducted on the same (or functionally equivalent) hardware, the System Acceptance Testing
should be a formal opportunity to demonstrate and document the overall functionality rather
than conducting rigorous challenge tests (Figure 5.3).
It must be stressed that System Acceptance Tests should only be conducted once the
underlying software and hardware have been tested and approved.
5.6.4 System Test Specification(s)
The System Test Specifications are used as inputs (reference documents) during the actual
System Acceptance Testing.
They contain important information (as described below) and the System Test Specifications
are therefore a mandatory document for conducting the System Tests. No tests can proceed until
the relevant Test Specification document is reviewed and approved.
5.6.5 Factory/Site Acceptance Test Results and IQ, OQ and PQ
There is an increasing tendency to acknowledge that IQ, IQ and PQ Protocols and Reports
(which have been adopted from process/equipment qualification) may not be best structured to
report on a complex set of interrelated computer systems tests.
The focus should always be on documenting the rationale for the scope, nature and level of
testing and on interpreting the test results. In this context Test Policies, Test Strategies and
supplier FATs, SATs and User Acceptance Testing may serve a more useful purpose than more
traditional IQ, OQ and PQs.
Installation Qualification protocols and reports are still a useful way of documenting the
installed system and of bringing the system under Configuration Management. OQ and PQ are
less useful and may be omitted if they serve no useful purpose (and if the validation policy
allows this).
The Test Strategy
25
Where an organisation still requires a formal IQ, OQ and PQ to be conducted and reported
upon, the emphasis should be on reducing the testing required solely to produce such
documents. As described above, there may be a clear relationship between Factory and Site
Acceptance Test results and the formal IQ, OQ and PQ Reports. Wherever possible, the IQ, OQ
and PQ protocols should simply reference Test Strategies and Test Cases and IQ, OQ and PQ
reports should reference the documented results of FATs and SATs.
CHAPTER 6
The Development Life Cycle of a Test Specification
As with any formal process, there is a logical sequence to be followed when developing the
System Test Specifications and when conducting the System Tests and there are recommended
activities that should be included in order to assure successful completion of the testing phase.
These are described in the various sections in this chapter:
6.1 Recommended Phasing; Interfaces between and the Dependencies of Activities
It is recommended that the activities associated with developing the System Test Specifications
and performing the System Tests be conducted in the order shown in Figure 6.1 in order to:
Develop the System Test Specifications and conduct the System Tests in the most efficient
manner.
Provide sufficient traceability to ensure successful validation.
Certain of these activities have dependencies that require that they be carried out in a specific
order. Where this is the case the two activities are shown in Figure 6.1 as being linked with a
bold arrow and can be summarised as follows:
The Validation Plan, Project Quality Plan and Test Strategy (Test Plan) must be completed
before any other activity.
The Functional or Design Specification must be completed before the associated Test
Specification.
The Test Specification must be completed before actual Tests take place.
This implies that any changes or updates to the prior activity must be reviewed to ensure that
the impact upon all dependant activities is understood and any subsequent revisions carried out.
Many of the activities listed will have a formal or informal interface.
Formal interfaces may be required in the case of dependant activities, where one activity must
be completed before a subsequent activity starts. This is usually the case when the sequence of
the related activities is important in building a robust validation case.
In these cases the output from one activity is used as the input to the subsequent activity and
this interface should be acknowledged and documented by referring to the prior activity in the
documentation of the dependant activity.
It should also be remembered that some of these interfaces might be two-way. If problems are
encountered in a subsequent activity it may be necessary to review some of the prior activities to
see if any change is required. If this is the case any changes to the prior activity should always be
reviewed for any impact upon ALL dependant activities, not just the one that initiated the change.
Good Configuration Management will support this process. This is shown in Figure 6.1.
27
28
Figure 6.1 The Dependencies: Various Life Cycle Documents and Activities.
6.2 Milestones in the Process

The major milestones in the development of a Test Specification and the conduct of the System
Tests are:
29
Completion of the first draft of the Functional or Design Specification (until which the
development of the associated Test Specification can not start).
Completion of the Functional or Design Specification (until which the associated Test
Specification can not be completed).
Development, review and approval of the individual Test Scripts (until which the associated
System Tests can not start).
6.3 Inputs to the Development of a Test Specification
There are several inputs to the development of a Test Specification and checks should be made
that all of the required input (reference) material is complete, approved and available before
work on the related section of the Test Specification starts.
These inputs are listed in Section 5.2 but in summary come from:
Validation Plan/Project Quality Plan

Test Strategy
Functional Specification
6.4 Document Evolution

The development of a Test Specification is an evolutionary process and, although sometimes
overlooked by those responsible for their development, these documents are subject to review
and approval.
This will be followed by the actual System Testing and the detailed sequence of this evolution
is given in Figure 6.2. Note that although only two individual system tests are shown (1 to n),
there may be any number of system functions under test at each stage of testing.
6.4.1 The Review Process
Each section of the document will be subject to review, as determined in the suppliers Project
Quality Plan and possibly by the users Validation Plan. Internal (suppliers) review may take the
form of:
Review by the author alone (not recommended)

Review by peer(s)
Open review, for instance, by a walkthrough of the document by the author and peer(s)
Review by a separate Quality Assurance function.
Depending upon the requirements of the Validation Plan, the user may be required to conduct a
formal review of the System Acceptance Test Specification. This will be after the final internal
reviews have been completed and may be limited to any, or all, of the following:
Review of the general Test Specification only (to check general test principles and
methodology).
Review of a random number of individual Test Scripts (to check the quality of these).
Review of all of the individual Test Scripts.
Note that it is unusual for the user to review and approve Test Specification and Test Scripts
other that those for System Acceptance Testing.
30
Figure 6.2 The Evolutionary Development of Test Specification and Associated Test Scripts.
When conducting reviews the following points should be considered:
Reviews should be conducted against the relevant input (reference) documents to check
that the contents of the Test Specification meet the requirements for thorough testing of the
system functions under test.
Any assumptions that are not based upon the input documents should be challenged and
either justified in the Test Specification or omitted.
31
Where the Test Specification relies on information contained in the input (reference)
documents full cross references are required (by page and paragraph or specific section
number).
It may be necessary to conduct several iterations of the review process, possibly using more
stringent review procedures for the final review. For instance, peer review could be used for
initial and interim reviews, but open reviews could be used for the final review and approval.
Once the Test Specification has passed a review its status should be updated in the document
register to reflect this.
6.5 Constraints on the Development of a Test Specification
There may be constraints on the development of a Test Specification and these are usually a lack
of qualified personnel or the prerequisite input documents. This may either constrain the
development of the entire document, or individual Test Scripts.
The effect of these constraints should be considered before commencing on any particular
section of the document and unless the constraints can be overcome, work on the section must
be placed on hold. Emphasis should always be on proper planning to ensure that the required
resource and inputs are in place at the proper stage in the project.
Where a constraint is identified and a justifiable work-around is reached, it is suggested that
the reasons for the initial constraint are documented in addition to the solution. This will provide
the necessary level of traceability and it is suggested that the solution be discussed between the
supplier and the user, so that the impact on the overall validation life cycle can be considered
and the agreement of the user recorded in the project minutes.
6.6 Constraints on the Testing
There may also be several constraints on conducting the actual Software Integration Tests.
These may either constrain all of the tests or just individual tests. These constraints can be
anticipated, managed and overcome and some of these constraints (and solutions, if any) are
defined in Table 6.1:
Table 6.1 Constraints on Testing

Constraint
Solution
The Test Script is not available (or

finally approved).
Testing proceeds at risk, clearly stating any assumptions that

are made in lieu of clear input (reference) material and with all
test methods and results being fully documented. When the
appropriate Test Script becomes available a full review of the
Test must be held to determine whether or not the test and its
results are valid. No test should be signed off until such a review
is held. Note it is extremely unlikely that this will save any time!
Appropriate test equipment must be sourced from an alternative
location, with full records being made of the test equipment
used.
Final System Acceptance Testing may be deferred until the
system is installed on site (Site Acceptance Testing)
Appropriate Test Equipment is not

readily available.
It is not possible to functionally test
the system because of an inability to
simulate input criteria at the suppliers factory.
32

6.7 Conducting the Tests
The individual functions of the system are part of the foundation of a quality installation, and
properly conducted and documented System Tests are an integral part of successful validation.
Careful thought needs to be given as to how the tests can most efficiently be conducted and
the methods, tools, equipment and personnel used. Further guidance on this is given below,
throughout Section 6.7 as follows.
6.7.1 Test Methods
There are several different methods of conducting System Tests. Each method will define a
series of steps, conducted in a repeatable sequence in order to demonstrate that the functional
objectives of the test(s) are met.
All of the individual tests may be conducted using the same methodology, or a variety of test
methods may be used as appropriate to each test.
All tests consist of entering agreed test data as inputs to the system function and recording
the resultant output data. This can be achieved in a variety of ways using a number of different
test methods. If it is not possible to conduct such a predictable and repeatable test, the function
will have to be verified instead of tested (see Section 3.4).
Each of the main test methods is described below, but it should be noted that this only
provides an overview of these methods. The exact test methodology will still need to be
described in the actual Test Specification, either in the General section, or in the individual Test
Scripts.
The detail in which the test is described may vary according to the test methodology
described in the General section of the Test Specification. It may be necessary to provide
detailed step-by-step instructions (including providing full details of individual keystrokes for
the test engineer to follow) or it may be acceptable to provide general instructions that rely on
the test engineer using an overall knowledge of the system under test.
The basic principle to follow is that the level of detail included should allow the test to be
repeated in exactly the same manner. For final System Acceptance tests it may be acceptable to
allow the user (who should conduct the test) to follow the Standard Operating Procedure (SOP)
to perform the actions required in the test. This has the added advantage that it tests not only the
system, but also that the instructions in the SOP are aligned with the current version of the
system and that the users training is sufficient to allow the task at hand to be executed.
At many levels, the system functions under test are deemed to be unique and it is unlikely that
any general training or documentation will refer to the function under test. This places specific
emphasis on providing clear and unambiguous instructions, more so than for standard devices
or manufacturing tests, which may rely on standard documentation and training for standard
software and hardware modules.
Where full step-by-step details of the test method are not provided, the General section of the
System Acceptance Test Specification should provide details of the necessary level of training
required of the test engineer.
In this latter case it is acceptable for the individual Test Script to reference a General section
of the Test Specifications. For example Select box count function, check that the control
mode is in Auto: and enter the number of boxes to count refer to Section 2.1.1 of this
specification: Entering demand values for specific instructions.
When defining the test methods to be used the following issues must be considered:
33
6.7.2 Manual Data Input

Data can be entered into the system function manually by a test engineer entering data into the
system. The data to be entered will be specified in the individual Test Script and this will also
describe how the information is to be entered (e.g. what order the data should be entered in, how
long a wait period should be employed, etc.).
6.7.2.1 Manual recording of results
Manual data input often requires the results of the test to be recorded manually, sometimes by
observing a value on a screen or by observing an actual physical event (such as a pallet being
delivered, or a new order being entered into a system). Since the purpose of the System
Acceptance Testing is to check the actual functional performance of the system it is quite likely
that system functions or actual physical actions will need to be observed and recorded.
In this case the individual Test Scripts should provide details of where and how the resulting
data can be accessed or what is to be observed and should provide a place for the actual results
to be recorded.
Note that results may be recorded qualitatively (pass or fail) or quantitatively (numerically,
or any other quantifiable result such as a written word, error message, etc.). Where the test run
will be subject to later review, results should be recorded quantitatively so that the independent
reviewer can assess whether or not the test objective was met by achieving the recorded result.
Where the item or function under test has no direct impact upon product quality or data
integrity, good testing practice may be followed and test results may be recorded in the form of
a simple checklist (see Section 7.9).
6.7.2.2 Physical simulation
It may be acceptable or preferable for input data to be physically simulated. This might be
appropriate where multiple values need to be changed simultaneously and can be more easily
achieved by the manipulation of physical devices (e.g. switches, potentiometers, etc.) connected
via physical inputs to the system function under test. This may also be true where the
functionality under test operates in the real world. In these cases it may be necessary to
physically simulate or operate a real piece of equipment or plant in a defined and repeatable
manner (e.g. manually feed a pallet in an automatic palletiser or deliberately misfeed a label).
Where this is the case, detailed step-by-step instructions may also need to be given in the
individual section of the Test Specification, or instructions in the General section of the System
Acceptance Test Specification may be referenced as described above. For example Select bar
code reader number 1 and scan the label on the tablet box refer to Section 3.2.1 of this
specification: Using bar code scanners for specific instructions.
As above, this method of testing usually requires the results of the test to be recorded
manually, either by observing a value on a screen, by observing an output reading or status on
an indicating device connected to the software function outputs or by observing an actual event.
In this case the individual Test Scripts should provide details of where the resulting data can be
accessed or observed and should provide a place for the actual results to be recorded.
6.7.2.3 Software simulation/emulation (Test Harness)
For more complex system functions it may be acceptable or preferable to simulate or emulate
the input data by using software. This may be the case where:
The input data needs to be changed faster and with more accuracy than can be achieved
manually.
34

Where multiple inputs need to be changed in a precise pattern or sequence which cannot
be achieved manually.
Where it is not possible to enter values in any other way.
Simulation is usually used to refer to the process whereby multiple inputs to the system under
test are calculated in real-time by a software model that simulates the process or function the
system is controlling.
Emulation is usually used to refer to the process where the output from an external system
(which is interfaced to the system under test) is replicated by a simpler system. Such emulation
software can usually be controlled manually, allowing a wide range of interface conditions and
data to be tested without relying upon setting up a large or complex external system.
The use of software test harnesses is more usual when testing more complex system
functions (as opposed to the simpler underlying software modules).
A well designed system will allow input and output parameters to be passed to and from the
system function under test without requiring special test hooks (this ability is inherited from
the underlying software modules and is often a function of a well designed system). In order to
test the function a software test harness is developed to pass test parameters to and from the
function under test (this is the software equivalent of a hardware test harness, which is
traditionally connected to the system under test).
In this case the individual Test Scripts will reference the details of the test data (values,
sequence, pattern) and the details of the simulation software and dataset(s) that will be used.
This should include details of software version numbers.
Where a standard simulation routine is employed, its use can be described in the General
section of the Test Specification and referenced by the individual Test Scripts. Where a specific
piece of simulation software is used (which is more often the case for System Acceptance
Testing) its development should be controlled and reviewed as part of the evolutionary process
of developing the individual Test Specification.
The development of test datasets should also be controlled and reviewed in a similar manner
and should be documented as part of the test results.
The recording of results from tests conducted using simulation software may either be done
manually or automatically.
6.7.2.4 Test datasets
Where test datasets are used they are also subject to careful design and control.
The data included in such datasets should be carefully chosen to properly exercise the system
function throughout its normal operating range. Where more that one set of input values is
included in a test dataset the effect of the interaction between the input values needs to be
carefully considered.
For example, it may be necessary to maintain one input at a constant value while changing
another variable. A complete test dataset may include data which varies only one value at a time
while maintaining all others constant and which may then move on to vary two or more values
simultaneously.
A complete test dataset will exercise a system function at various points in its normal range.
It is unusual to perform challenge testing as part of a System Acceptance Test, the purpose of
which is to demonstrate the correct functioning of the system under test. Challenge (or stress)
testing (outside of normal operating limits and ranges, using illegal values) is more properly
conducted as part of the Software Module or Software Integration Testing.
Test datasets should be included as part of the overall Test Specification review process and
should be subject to change control when being developed and revised. Note that there are
35
special considerations to bear in mind when the test data consists of data objects (this is covered
in Section 7.17).
6.7.2.5 Combined test methods
Since the purpose of the System Integration and System Acceptance Testing is to test the
functional performance of the system under test, it is often required to use a combination of test
methods in a single test (manually input data, physically simulated data or data simulated by a
software test harness). This allows a wider range of functionality to be tested, and may include
physical simulation of parts of the plant or process, interfaces to external systems, etc. Where
this is the case the appropriate test methodology should be clearly documented.
Specific attention should be paid to the sequencing of various input conditions generated by
different methods of data entry. For example, attention should be paid to the timing of manually
entered data when it is inserted amongst values generated by a test harness (software
simulation). This will also be a consideration where test data is generated by different test
harnesses that operate simultaneously.
6.7.2.6 Automatic recording of results
It may be advantageous to automate the recording of test results. This may be particularly
appropriate where:
There is a large amount of data to be recorded

The output data needs to be recorded at a rate which is not possible by manual means
There is a real possibility of errors being made recording complex data manually
Results from an externally connected system need to be included in the test records
This is again more likely for Software Integration and System Acceptance Tests than for the
simpler Software Module, Package Configuration or Hardware Acceptance Tests.
Wherever possible, the automatic recording of data should be accomplished using standard
system facilities (data logging, alarm logging, trend data recording, etc.), which can be described
in the General section of the Test Specification and referenced by the individual Test Scripts. This
may also include the recording of outputs from the software function by a recording device
attached to any physical outputs from the system.
If it is necessary to develop specific data recording applications these should be controlled
and reviewed as part of the evolutionary process of developing the individual Test
Specifications.
Some automatic recording systems or externally interfaced systems may not produce
sufficient levels of documentation to provide traceability of the tests. For example, the time and
date may not be included, nor the name of the variable(s) being recorded. Where this is the case
additional documentation (usually in the form of manual notes) should be appended to the
output (see Section 7.12).
Many computerised test management tools are able to automatically record results as part of
their automated testing functionality.
6.7.2.7 Automated testing
Conducting the System Acceptance Tests can be a time consuming and expensive process given
the need to fully test every system function. This task can be eased by the use of automated
testing, performed by a computerised test management tool.
This basically combines the functions of software and process simulation and automatic data
recording to fully automate the task of system function testing. Although such facilities lend
36
themselves to the testing of many similar functions they can still be used to either conduct a
single test at a time or can conduct many tests one after the other.
If it is necessary to develop specific automated testing applications these should be controlled
and reviewed as part of the Test Specification(s).
6.7.2.8 Control and validation of test methods and test harnesses
It is important that appropriate test methods (sequence of actions) are selected in order to:
Conduct the tests in the most efficient manner and thereby minimise project costs.
Conduct tests of sufficient rigour to prove the proper functioning of the system under test
and thereby ease the subsequent task of Operational and Performance Qualification and
add value to the process of validation.
Most supplier organisations will have common test methods (sequences) for conducting similar
tests to ensure that these objectives are met and these will often include defining the structure
and order in which test harnesses will be used.
Wherever possible standard test harnesses (i.e. software functions for testing and recording)
should be used, although this may not be possible for testing unique system functions. Where
these test harnesses are standard features of the system under test (or of a dedicated external test
system) and where these features of the system have previously been validated it will not usually
be necessary to validate the methods used. Previous validation of such methods may, for
example, have been on a similar project or by general use in the pharmaceutical industry.
Where special test harnesses have been developed, either for the project or for an individual
system function, these methods must be validated in the same manner as the system under
test.
It is important that a supplier has a Quality Assurance system that will record the
development and status of such software test harnesses between projects. Such a system should
record the development of the software, where the test harnesses have previously been used and
validated, and any subsequent version changes. Without this system it is probable that project
specific validation of the test harnesses will be required.
6.7.3 Formal Acceptance of Test Results
All tests must be formally accepted (signed off) upon successful completion. Who signs these
will be specified in the suppliers Project Quality Plan or the users Validation Plan. There may
be two levels of acceptance as follows:
6.7.3.1 Signing off the system acceptance test (sign off by tester)
It is essential that the person responsible for conducting the test should also sign off the actual
test. This person must be a properly trained/experienced and qualified test engineer, and he/she
must conduct the test from start to finish, as being responsible for following documented test
procedure, and ensuring that actual test results are properly recorded, manually or automatically.
They should ensure that the recorded results comply with the acceptance criteria (details of
acceptable and unacceptable test results should form part of the individual sections of the
System Acceptance Test Specification).
By signing off the actual System Acceptance Test the responsible engineer is certifying
that:
The test was conducted in accordance with the documented test methods
37
The test engineer conducted or supervised all parts of the test (supervising either another
engineer or an automated test system)
The test results were properly recorded and are accurate
The test results were in compliance with the acceptable test results documented in the Test
Scripts
When the test engineer signs off the test record, he/she must include his/her name and signature,
and the date on which the test was conducted. It may also be necessary to record the test run
number when a test has been repeated more than once.
Where the individual System Acceptance Tests are to be witnessed (either internally by the
suppliers personnel or externally by the users personnel) the witness should also sign
(including his name, signature and date) to confirm:
That they have witnessed or reviewed the results

That the results are consistent with the documented acceptance criteria
More guidance on the requirements for signing executed Test Scripts is given in Sections 7.10
and 7.11.
6.7.3.2 Approval of the tests
As well as signing off the individual Test Scripts it may also be necessary to review and sign off
the complete set of tests (or a subset).
This may be the case where:
A specific stage of testing is complete (Hardware, Software Module, Package

Configuration, System Integration or System Acceptance), or a subset thereof.
The supplier wants to ensure that all of the Factory Acceptance Tests have been conducted
properly before moving the system to site. This is to ensure that the site work and
Installation Qualification are not started prematurely and that the Factory Acceptance Test
documentation is of sufficient quality and consistency to facilitate the users task of
validation.
The user wants to ensure that the test documentation is of sufficient quality and consistency
to facilitate his own task of validation before proceeding to the next stage of testing.
This review may take several forms including:
A review of some or all of the individual Test Scripts against the General section of the
System Acceptance Test Specification. This is to ensure adequacy of cross referencing and
consistency between tests.
A review of some or all of the executed Test Scripts to ensure that the tests have been
conducted properly, all results recorded properly, that the results were within expected
ranges and that the tests have been properly signed off by the test engineer.
The review of the actual test procedure and results may be conducted in parallel with the
conducting of the actual tests (the tester and the reviewer sitting side by side). In this case the
results sheets can be signed off by the reviewer just after the test engineer signs off the test
records. More usually this can also take the form of a post-test review, where some or all of the
test records and documentation is reviewed as a separate exercise (possibly by the users as part
of the IQ).
38
Regardless of when the reviews of the test records and documentation are conducted, the
reviewer should add their name, signature and date to the individual Test Scripts (usually when
all tests are reviewed), or to a separate document which records which tests were reviewed
(useful when a representative sample is reviewed).
6.8 Outputs from the Testing
The deliverables from a given phase of testing are:
The approved Test Specification

The executed Test Scripts, including results, raw data and any test evidence
Incident reports documenting any deviations
Change control records relating to any subsequent modifications, plus justifications and
approvals
CHAPTER 7
Recommended Content for System Test Specification(s)
The following sections describe the typical structure and content of a System Test
Specification(s). This assumes that a given Test Specification consists of a General section and
individual Test Cases.
In some instances (for large or complex systems) it may be preferable to develop Test Cases
as separately reviewed and approved documents, which reference a common Test Specification.
An example of this would be where many Software Module Test Scripts reference a single
Software Module Test Specification. If this were the case the Software Module Test
Specification would have to be approved prior to approving and executing individual Software
Module Test Cases.
7.1 Overview
As the name suggests, the overview section of the document provides a brief summary of the
document, including who wrote/developed the document, the review process by which the
document is approved, the scope of the document and its relationship with other documents in
the validation life cycle.
7.1.1 Front Page/Title Block
The title block of the document should include the following information as a minimum:
Project name
Document title
Revision number (subject to formal document release/control)
Author(s)
Approval signatures and context of approval (technical, quality/compliance etc)
7.1.2 QA Review Process

A section of the document will refer to the quality control/review/acceptance procedures which
apply to the document. These procedures will usually refer to the users Validation Plan and/or
the suppliers Project Quality Plan and will not usually describe these procedures in detail.
7.1.3 Scope of Document
The scope and purpose of the document should be clearly defined, so that those with only a
peripheral interest in the document can easily understand what the purpose of the document is,
what it covers and what it does not. Where relevant this should refer to the Test Strategy.
39
40
This should include the following sections:

7.1.3.1 Reference section
The purpose of the Test Specification should be described, with appropriate reference to the
other documents in the validation life cycle. This section may also refer to a general description
of the life cycle given in the Validation Plan.
7.1.3.2 Grouping and ordering of tests
In a project with multiple system functions it is likely that tests will be ordered in a specific
manner and may be grouped together. This implies that the tests themselves will be conducted
in a particular order. The reasons for the particular ordering and grouping should be described,
and may include:
Conducting Factory Acceptance Tests before Installation Qualification and Site

Acceptance Tests.
Conducting simpler tests first and grouping similar tests together, thereby building
familiarity with test methodology before moving on to more complex system functions.
Test sequencing, where the output from one test is the input to the next test. This is
important when the functional hierarchy means that more complex system functions are
based upon simpler underlying system functions, which are in turn built using the base
level software and hardware modules.
Making most efficient use of common test resources (test equipment, personnel, etc.),
prerequisites and set-up.
7.1.3.3 Areas not tested and why

It is possible that particular system functions, or some aspects of a particular function, may not be
included in the System Tests (for instance, where a function or design feature is verified rather than
tested see Section 3.4). Where this is the case the reasons for not including a specific test in the
series of System Tests should clearly be documented.
One reason for not testing individual hardware or software functions may be that the
particular system function has previously been tested and validated on a previous project (either
with the same or different users). Where this is the case the System Tests may be limited to:
Providing full details of where and when the system function has been tested and validated
before (referring to the users, Project, Test Specification and System Test records).
Confirming that it is the same version of the system function and underlying software and
hardware modules, running under the same operating system as previously tested/validated.
This is often the case with Commercial Off the Shelf (COTS) systems of known pedigree.
Even where this is the case, details of any system functions not tested should be listed, along
with the reasons for omitting the test.
7.1.3.4 Bespoke test methods and harnesses
Wherever possible, the testing of the system functions should follow a standard set of test
methods that sequence standard test harnesses, or use standard methods for screen navigation,
all of which are described in the System Test Specifications.
It is usual that the testing of many system functions will use different methods and harnesses,
especially where the functional requirement or design feature tested is unique to the application
or site. Where this is the case, the detailed methods and harnesses will be described in the
individual Test Cases.
41
Possible reasons for using bespoke test methods and harnesses may include:
It is impossible to test the system function using standard test methods. This may be
because the function under test is particularly complex.
The testing of the system function combines several standard test methods (for instance,
where some inputs are simulated by software test harnesses, but where a few inputs are
easier to simulate manually).
Where actual physical simulation is needed, requiring part of the physical process or plant
to be simulated or reproduced.
Where interfaces to external systems are a large part of the functions under test.
7.2 General Section
In order to reduce the size and complexity of the individual sections of a System Test Specification
(Test Cases) it is recommended that a General section of the System Test Specification be used to
document common terms, principles, methods and prerequisites.
The following sections may be included in the General section of the System Test Specification
and should be referred to by the individual sections of the System Test Specification (Test Cases)
whenever relevant.
7.2.1 Glossary
The glossary should list and define all of the common terms that are used in the System Test
Specifications and the test records. It is possible that this section refers to a common testing
glossary, or one maintained as part of another project document, or to a standard glossary
maintained by the supplier and user.
However, it is most likely that a project specific glossary will be used (or referenced) since it
is likely that this will combine the agreed terminology of the user and supplier.
7.2.2 General Principles and Test Methodology
Because many of the System Tests will test similar system functions of similar complexity, these
functions can be tested using a small set of principles using defined test methods and harnesses.
These common principles should be clearly described.
7.2.2.1 Principle of testing
The general test principles will document the overall nature and objectives of the tests and will
describe how challenging the tests will be. This will be specific to a given level of testing, or
may describe how a standard test strategy (see Section 3.3) is applied to a particular system.
For instance, in the Software Module Test Specification this section may state that the
principle is to conduct stress testing for all bespoke software modules with a high or medium
GxP criticality (i.e. to try to break the modules by using out of range data). It may also state
that this will not be performed for bespoke software modules of low or no GxP criticality.
For the Software Integration Test Specification this section may state that the principle is to
conduct stress testing for all software functions with a high or medium GxP criticality (i.e. to
try to break the modules by attempting to conduct illegal process in the work flow). It may also
state that this will not be performed for software functions of low or no GxP criticality and that
only normal process workflow will be demonstrated for these functions.
Where applicable, this may reference a relevant section of the Test Strategy.
42
7.2.2.2 Standard test methods and harnesses

The general description of the test methods to be used should be included in the General section
of the System Test Specifications. This may include a general description of each of the test
methods and test harnesses employed during the complete series of tests.
Including (or referring to) detailed instructions in the General section of the System Test
Specification means that detailed instructions need not be given in every individual Test Case.
This allows the individual Test Cases to focus on the functions under test and the expected
(acceptable) results.
The individual Test Cases can, therefore, refer to the General section of the Test Specification
and need not provide detailed instructions for every part of the test.
Descriptions should be provided for each type of test that will need to be conducted. As
described, these may include methods for:
Manual data input

Physical simulation (including process or plant)
Software simulation (including details of which software test harnesses to be used and how
they will be used)
Combined test methods
Automated testing
As well as describing the different test methods this section should include references to:
Any common documents such as operating manuals relating to the system under test. This
should explicitly list references to the separate sections referring to operations such as:
Entering (simulating) analogue input values
Entering (simulating) digital input values
Enter (simulating or emulating) string (text) data
Changing set points
Changing the mode of control loops (auto, manual etc)
Etc.
Any common documents providing instructions for:
Operating test equipment required to simulate physical inputs and display the status
of physical outputs
Setting up and running software required to simulate inputs
Setting up and running automated test software
Operating process or plant required to conduct the System Acceptance Tests
If standard documents are not available for reference purposes then detailed instructions for
conducting each type of test should be given. Although the General section only provides
generic descriptions for each method of testing (which can then be referenced by simply
including a tag number, point or channel number, function name, etc.), sufficient details
should be included to allow any suitably qualified, trained or experienced engineer to conduct
the test.
It is more usual to have to provide specific test descriptions and instructions for the type of
Functional Testing included in the System Acceptance Tests. This is because the individual tests
tend to be unique to the application and site. When it is required to describe the detailed test
methodology to be used it may be necessary to include specific, step-by-step instructions
including:
43
The actions the test engineer has to take, such as:
Step-by-step keystrokes to be entered into the system under test
Step-by-step actions to be taken using physical input simulation equipment (test

rigs)
Step-by-step instructions describing the physical actions external to the system (to do
with associated process or plant equipment)
Step-by-step keystrokes to be entered into the system simulating inputs or executing

automated tests
The acceptance criteria for the tests (see Section 7.2.2.3)
How the results should be recorded (see Section 7.2.2.4)
The test methods and harnesses described in this section will be applicable to the majority of
the individual tests at a given level of testing and all instructions should be clear, unambiguous
and use language which is not subject to differing interpretations. Any ambiguity should be
picked up as part of the review process.
7.2.2.3 Acceptance criteria
The basic purpose of testing is to ensure that the system functions in accordance with the
applicable Design Specification. This means that the function under test will produce repeatable
and predictable outputs for defined sets of input criteria.
The individual Test Scripts will clearly define the input criteria to be used in the test and these
should produce a predictable set of output criteria. The predictable outputs are defined as
acceptance criteria and the purpose of the test is to ensure that the output of the system
function matches the expected acceptance criteria.
In order for a test to serve a useful purpose, every test should have a defined set of
unambiguous acceptance criteria for each set of defined input conditions. In the case of
Functional Testing the acceptance criteria may be described (at least in part) in terms of the
actual physical process, plant or equipment controlled or manipulated by the system under test.
Examples include the emptying of a reactor, the stacking of pallets or even data being
transferred between two external systems.
The way in which the acceptance criteria are defined and documented should be described in
the General section of the System Test Specifications, and the actual acceptance criteria should
be clearly documented for each individual test in the Test Script.
7.2.2.4 Recording of results
As well as providing a description of the test methods to be used it is important that details are
provided of how results are to be recorded. This may include a description of the following
recording methods:
Manual recording of results

Automatic recording of results.
Where results are recorded manually, a description should be provided which details:
Where the data are recorded (usually on a standard test sheet, or possibly in a test database)
To what accuracy data should be recorded (generally to the same accuracy as is given for
the expected results).
Where data are recorded automatically information should be provided which details:
44

The version of the recording software used (there may be more than one recording
application used).
The instruments to be used to record the status of physical outputs or external parameters
(including the calibration details, model number, version number and connection details).
Where the results will be stored (file name and format in the case of software records, or
where a physical record may be stored when not attached to the test record).
How the results will be passed by the test engineer (by comparison to the defined
acceptance criteria and by recording on a physical record sheet referring to the data, by an
electronic signature attached to the data file, etc.).
How the data can be played back or recalled, including full operating instructions and
details of the software application to be used (name, description, version number, etc.).
This may be necessary in order to allow the tests to be reviewed or for the user to audit the
tests.
7.2.2.5 Test acceptance and sign off (approval)

The General section of the System Test Specification should also provide details of how the
individual Tests Scripts should be signed off. This may require some or all of the following:
Tests are signed off by a nominated test engineer.

Tests are reviewed at the same time the test is carried out (witnessed) by a separate
member of the suppliers test team
Tests are reviewed at the same time the test is carried out (witnessed) by a member of the
users project or validation team.
Tests are subject to separate review by a member of the suppliers quality team.
Tests are subject to separate review by a member of the users validation team (possibly as
part of the Operational Qualification).
7.2.3 General Test Prerequisites

Most tests will have some prerequisites that need to be in place before the test can be conducted.
While some tests will have prerequisites that are specific and unique to the individual test,
general prerequisites should be described in the General section of the System Acceptance Test
Specification.
These may include
7.2.3.1 Required system hardware and software
Details of the system hardware and software required to conduct the System Acceptance Tests
should be given. For System Acceptance Tests, this should always be the hardware on which the
Hardware Acceptance Tests were conducted. This may also be applicable to Software Module
and Software Integration Tests that are hardware dependent.
The General section of the System Test Specification should list:
Hardware that the tests should run on (usually just serial numbers at this stage). These
should be checked against the versions recorded as part of the Hardware Acceptance Test.
Software version numbers of all standard system software components (GAMP level 1 and
2 software) installed on the system that the tests should run on. Where applicable, in cases
where hardware is defined as including firmware and/or operating systems, this should
be checked against the versions recorded as part of the Hardware Acceptance Test.
Software version numbers of all system software modules (GAMP level 3, 4 and 5
45
software) installed on the system that the tests should run on. These should be checked
against the versions recorded as part of the Software Integration Tests.
Connection details of an appropriate test system configuration, including any data loggers,
recorders, system printers, consoles, etc.
Where appropriate, references to specific operating instructions and the applicable level of
experience, training or qualifications required to set up and use the equipment should be given.
7.2.3.2 Required test equipment
Details of the test equipment required to conduct the test should be listed. This will usually detail
general equipment required to conduct a large number (or all) of the tests and may include:
Equipment required to physically simulate inputs to the system under test (so called testrigs).
Equipment required to show the status of physical outputs from the system under test (so
called test-rigs).
Individual test instruments (meters, signal generators, oscilloscopes, logic probes, protocol
emulators, etc.).
Separate systems required to simulate software inputs to the system.
Where appropriate, details of the model and version numbers should be provided, along with
references to specific operating instructions and the applicable level of experience, training or
qualifications required to use the equipment.
Details should also be provided on how the equipment should be tested and how it should be
set up for each specific type of test. Where reference cannot be made to specific operating
instructions detailed instructions should be provided.
7.2.3.3 Required test software
Where tests are conducted using simulation software or automated test software, details of the
software to be used should be given. This will include:
Details of specific application software to be used, including application name, functional

description, version numbers etc.
Details of the hardware that this test software should execute on.
Details of the individual Test Script, including unique name and version number.
The change control procedures relating to the Test Scripts, including the author, version
control and review procedures.
Where appropriate, references to specific operating instructions and the applicable level of
experience, training or qualifications required to set up and use the software should be given.
7.2.3.4 Required test datasets
It is possible that multiple tests can be carried out using common sets of data. Where this is the
case, the details of the test datasets should be provided in the General section of the System Test
Specification.
These details may include:
Which type of tests the individual datasets are applicable to
46

The principle behind the test datasets (i.e. testing under and over range values, values
across the valid range)
The details of the test data (format, filename and location, number of values and possibly
the actual values)
The change control procedures relating to the datasets, including the author, version
control and review procedures.
Because some or all of the software functions under test may be unique it is sometimes useful
to develop specific datasets for individual tests. Where this is the case all of the details above
will apply.
7.2.3.5 Required process, plant or equipment
Because the System Acceptance Testing is a functional test of the system, it is sometimes
necessary to test the operation of the system by reproducing part of the process, plant or
equipment that the system is connected to.
Some types of process, plant or equipment can be simulated by software, using a scaled down
or simplified physical model, or by an approximation of the real world items being controlled
or manipulated. Other items are more critical and lend themselves to actual testing in the
suppliers factory. Examples may include a packaging line being fed with real product, labels,
containers, etc. or an automated robot identifying, picking and transferring real product as part
of the test.
Where process, plant or equipment are being reproduced or simulated, details of how this is
achieved should be given. Where the test does not fully reproduce the functioning of the real
site installation, this should also be described so that these shortcomings can be picked up
during Site Acceptance Testing and/or Operational Qualification.
7.2.3.6 Required test documentation
Tests should always be conducted with the proper test documentation available and the standard
prerequisite documents can be listed in this section. Common documents required to conduct
the individual System Tests are:
The approved System Test Specification

A copy of the individual Test Script relating to the system function under test (part of the
Test Specification, or as a separate document)
Any related documentation referred to in either the General section of the System Test
Specification or the individual Test Script (e.g. operating manuals)
Test Record sheets to allow the results of the test to be recorded, reviewed and signed off
(if separate from the Test Script)
Incident Reports, for any prior deviations which may impact in the test
7.2.3.7 Test sequencing

Although different system functions should be independent of each other, it may be desirable to
conduct some tests in a preferred order in order to maintain testing efficiency.
Where general prerequisites exist for conducting prior tests these may be documented in the
General section of the System Test Specification (see Section 5.6.3).
7.2.4 Appendices
Finally, the System Test Specification may contain any necessary appendices.
47
These will usually be information that does not bear directly upon specifying or conducting
the general or individual tests, but will be useful background or reference material.
Example of information contained in an Appendix could be:
A standard Units of Measure table for the system under test

A bibliography of useful reference documents (such as various GAMP documents and
guidelines)
A full listing of the other related life cycle documents, including title, the latest version
number and status
A reference to the suppliers relevant Quality Assurance standards
Etc.
7.3 Individual Test Cases
As well as a General section, the System Acceptance Test Specification will also contain details
for all the individual System Tests that need to be carried out. These are often referred to as Test
Cases.
The Test Cases will provide additional information, building on the General section of the
Test Specification to give the final details on the purpose of the individual test, the functions
that are to be tested, the expected results, etc.
The level of detail included in the Test Case should be relevant to the GxP criticality of the
item under test. For items with no direct impact upon product quality or data integrity the Test
Scripts may follow general industry good testing practice.
The following sections document the level of detail generally required for testing items,
which has a direct impact on product quality or data integrity in systems of high or medium GxP
criticality. In all cases the level of detail included in the Test Script should be appropriate to the
GxP criticality. For each item of hardware, software, or system function the following items may
be included in the corresponding Test Case:
7.3.1 Unique Test Reference
Each test should be given a unique reference, which identifies it as a separate entity. This may
be the name of the system function under test (i.e. Centrifuge Operation) or may be some sort
of code (i.e. CFG001).
This unique reference should be used whenever reference is made to the test in order to avoid
any possible ambiguity.
7.3.2 Name of Hardware Item, Software Module or Function under Test
For Hardware Tests or Software Module Tests, the item of hardware or software module under
test should always be clearly identified.
For Package Configuration, Software Integration or System Acceptance Tests, if the system
function under test has a unique reference that does not explicitly identify the system function
under test (i.e. CFG001) the name of the system function should always be included as an
additional point of reference (i.e. Test CFG001 Centrifuge Operation). Note that to avoid
ambiguity the name of the function under test should always refer to the function as defined and
referenced in the appropriate Design Specification.
48
7.3.3 Cross Reference to Functional Description or Design Detail

The functional description or design detail under test should always be identified as an
individual section of the relevant Design Specification. This section should be explicitly
referred to in the Test Script, including document name, revision number, section and possibly
page and paragraph number.
It may also be desirable to include excerpts from the individual section of the Functional
Specification if this eases the task of explaining the functions that are under test although this
requires particular attention to be paid to change control procedures to ensure that the excerpt
is maintained and up-to-date.
7.3.4 Specific Prerequisites
It is possible or even likely that some or all of the individual Test Scripts will have specific
prerequisites that are not covered under the General section of the Test Specification. As
described above, these may be categorised as follows:
System Hardware and Software

Test Equipment
Test Software
Test Datasets
Required Process, Plant or Equipment
Test Documentation
Prior Tests
For the sake of efficiency, test prerequisites should be included in the General section of the
System Test Specification wherever possible. The prerequisite section of the individual Test
Script should only include:
Specific details which instance the general prerequisites (for example, a reference to the
General section on test dataset prerequisites and then the name of a particular test dataset
file from a common directory).
Specific details which are unique to the individual test (for example, the use of a particular
piece of equipment that is used for a single test).
7.3.5 Particular Test Methods and Test Harnesses

Specific details of the actual test methods and test harnesses used for the individual test must
be given. This should include the following.
7.3.5.1 Test objectives
A brief description of the test methods should be given, for example:
This test will check that the module rejects out-of-range input data, alerts the operator by
providing a error message and logs the error.
This test will check the capability of the system to recognise when product boxes are
incorrectly loaded into the feed mechanism or
This test will examine the correct functioning of Centrifuge Emergency Stop routines.
49
7.3.5.2 Methods used

A detailed method (sequence) for conducting the individual test should be given, either by:
Instancing a test method described in the General section of the System Test
Specification (for example, By using the test methods described in Section 2.3.7
Entering Manual Instructions, instruct the robot to transfer 6 pallets from the inflow stack
to the return stack.)
Describing in detail the methods used to conduct a unique test.
7.3.5.3 Test harness used

Details of the software test harness used to simulate inputs and read outputs should be given,
including the revision number of the test harness. This may either:
Refer to a standard test harness (for example Use the Batch Record Data Test Harness to
simulate finished product values and confirm that the system should create a completed
batch record see Section 3.1[a] for details of using the batch record system.)
Describe in detail the use of a specific test harness developed for an individual test.
7.3.5.4 Detailed instructions

Where detailed instructions for conducting a unique test need to be provided, these should
provide complete, step-by-step instructions in order to ensure repeatability in the way the test is
performed. This will often be necessary to avoid ambiguity in the case of unique functional
tests.
These are as detailed above for the general methodology section and, as a reminder, may
include:
The Principle of Testing

The method(s) of testing employed:
Manual data input

Physical simulation
Software simulation (the use of a unique test harness)
Combined test methods
Automated testing
Detailed reference to applicable documentation

Step-by-step actions taken by the test engineer
Equipment to be used
What feedback or results to expect
How the results should be recorded
7.3.6 Acceptance Criteria

For each test, details need to be provided on what constitutes a pass and what constitutes a fail.
This is done by providing a list or description of expected (acceptable) results. It is most
useful to define only the acceptance criteria and to define any other result as a failure.
However, there may be occasions where there is a wide range of acceptable results, and a
relatively small number of unacceptable results. On these occasions it is sometimes useful to
provide a smaller list or description of unacceptable results so that a failure can clearly be
identified.
50
There should be no ambiguity between what constitutes a pass or a fail and there should
be no middle ground between the two.
This places the emphasis on identifying concrete, observable parameters than can be
observed by the test engineer and upon using precise language to describe them. The
appropriate Design Specification should be the basis for defined acceptance criteria. This
should describe the User Functional Requirements or Design Details clearly enough to allow
acceptance criteria to be identified.
Acceptance criteria should be clearly and unambiguously defined. For Software Module,
Hardware and some System Integration Testing this can often be done by providing a specific
acceptable value or list of acceptable values. In the case of a list of acceptable values, it is useful
to document the acceptance criteria against the applicable input parameters in the form of a table.
Some System Integration, Package Configuration and all System Acceptance Tests are
functional tests and the acceptance criteria may be given by providing a description of the
expected results.
Note that the location where the results will be displayed should be described as part of the
test methodology (which may be a physical observation) and that the acceptance criteria is
usually a simple value, result list or description under a suitable header.
It may also be that the definition of acceptable performance may be clearly described in the
Functional Specification. In this case a written description of function may form part of the
acceptance criteria.
Examples of various acceptance criteria are given below:
List of Expected Results Hardware Test

Input Values (mA)
Acceptable Results (mA)
0
4
12
20
24
List of Expected Results Acceptance Test

Input Values
# Input Pallets
1
2
4
6
8
4
4
8
20
20
# Output stack pallets

0
8
4
4
0
Acceptable Results
# Pallets transferred
1
0
4
4
8
The system will automatically increase the centrifuge speed from rest to a minimum speed
of 3,000 rpm over a period of between 1 minute 30 seconds and 2 minutes. The centrifuge
will run at a minimum speed of 3,000 rpm for a period not less than two minutes and not
exceeding 2 minutes fifteen seconds. The centrifuge shall then decelerate, reaching rest in
not more than 30 seconds.
51
Acceleration Time (> 1m: 30s, < 2m: 00s)

Speed achieved
Run Time (> 2m: 00s, < 2m: 15s)
Minimum speed sustained
Deceleration time (< 30s)
7.3.7 Data Recording

As well as defining the test methodology and the acceptance/failure criteria, it may also be
necessary to detail how the results of the individual test should be recorded. This is usually the
case where results are not simply written on (or entered into) a copy of the Test Script.
As with other sections, the details of how the data is recorded for the individual tests can
either be by:
Instancing a recording method described in the General section of the System Test
Specification (for example, by manually recording the test results as described in Section
4.3.4 Recording the results of a compared floating point values, the results of the test
will be recorded on the unique test sheet, signed and witnessed).
Describing in detail the recording methods used to record the results of a unique test.
Where it is required to describe in detail the recording method to be used for a unique test, full
details of how the information is to be recorded should be given (as described above in the
paragraphs relating to the general data recording section).
7.3.8 Further Actions
It is possible that additional actions may be performed once the actual test is completed and
signed off. Although not all of these are absolutely necessary from a validation point of view,
additional information may be included in the individual Test Script. For instance, it may be
necessary to make a piece of equipment or process plant safe following the execution of a test.
7.3.8.1 Repeating failed tests
It may be useful to describe how a test would be repeated if it does not pass the first time. This
is useful in cases where it is not necessary to repeat all of the set up steps if the test is repeated
immediately. It should be emphasised however that the purpose of the System Acceptance Test
is to demonstrate the correct functioning of the system, and failed tests should be rare at this
stage in the project life cycle.
It may also be useful to place a maximum on the number of repeats that can be performed
before a system function is deemed to have failed the test. For instance, it may be acceptable to
repeat a test if the cause of the failure was a simple set-up error by the test engineer (a software
module was left in manual when a complex control loop was tested, for example).
However, there is little benefit in repeating the same test numerous times if the problem is not
immediately obvious, and it is useful to limit the number of tests that may be performed before
a review of the underlying software and hardware test status and code is carried out.
52
It will also be necessary to specify that details on the cause of the failure must be recorded
and where and how they should be recorded (incident reports). These can again instance the
General section of the System Acceptance Test Specification or provide unique, step-by-step
details.
7.3.8.2 Reset actions
It may be necessary to perform certain reset actions to return the system to normal conditions
following the execution of certain tests.
These reset actions can be defined in one of two ways:
As a check to make before the individual test is conducted (to check the prerequisite status
of the system/software).
As a series of reset steps or actions to be conducted after each test.
Either of these methods may again either instance the General section of the System Test
Specification or provide unique, step-by-step details of the tests or actions to be performed.
7.3.8.3 Preparation for next test(s)
It may be desirable to detail any actions that could usefully be carried out to prepare the
system/software for any following tests. This is, however, unusual since it is more flexible to
include any specific set up instructions as part each individual test.
If this is the case, it is again possible to either instance the General section of the System
Acceptance Test Specification or provide unique, step-by-step details of the actions to be performed.
7.3.9 The Use of Separate Test Record Sheets
In many instances, a complete copy of the Test Script will be used to record the results of an
executed test. However, there may be instances where the test may be complex and the specific
instructions, prerequisites, etc. take many pages. In this instance it may be useful to have a
separate Test Record Sheet. This makes the paper easier to handle during testing. The Test
Record Sheet should have a header page or section to document the test run number, the start
time and date and the name of the tester. There should also be a footer section with space to
indicate the completion time and date, a clear indication of the pass/fail status of the test and
space to sign the test.
The main section of a Test Record Sheet usually is comprised of a table containing:
The Test Step number

The specific instruction to carry out (including details of any test evidence required to be
taken)
The expected result
A space to write the actual result
A space to write any comments or references about attached evidence, test incident sheets,
etc.
A space to initial the test step (including a witness if required)
The Test Record Sheets can either be sets of pages, bound in the main System Acceptance Test
Specification, or may be separate sets of pages, usually kept in a loose leaf binder. The
advantage of this is that the sheets for each test may be removed for ease of use and returned
once the test is complete.
53
Where separate Test Record Sheets are provided, these should clearly reference the main Test
Script and should be subject to full change control. Each page should contain details of the test
(by unique test reference and name) as well as page numbers in the form of page N of M. This
is important since the individual Test Record Sheets may become physically separated from the
main System Acceptance Test Specification and will need to be returned under proper control.
CHAPTER 8
Good Testing Practices
Having established why we test, and what we should include in our tests (the Test Strategy), it
is important to understand how to test, and what constitutes good testing practice in the
healthcare industries. In previous sections we looked at some of the principles of testing and
some of the techniques that can be used. There is, however, a human element to testing that also
needs to be considered.
8.1 Prepare for Success
A successful testing programme begins well before testing commences. In many cases the
testing is expected to make up for deficiencies elsewhere in the development programme; as the
saying goes you cant test quality into a product. Thought needs to be given to the testing
programme from the moment the project starts, otherwise problems may well occur during the
test execution.
8.2 Common Problems
It is possible to identify some common problems, to examine their causes and to consider what
can be done to prevent them.
8.2.1 Untestable Requirements
In many cases the testing fails because the User Requirements, Functional or Design
Specification details were ambiguous, or were simply not testable. The obvious example is a
User Requirements statement such as The system should be user friendly. How is it possible
to objectively test such a requirement?
Requirements should be:
Unambiguous it should be clear what the requirement is and it should not be liable to
differing interpretation.
Testable from reading the requirement it should be clear what can be tested to prove that
the requirement has been met and what the test objective should be.
Single stated each requirement statement should contain only one requirement. In this
way requirements can be proven in single Test Scripts, or in separate sections of a Test
Script. Consequently, if a single requirement is not met during testing, the retest can be
limited to that single statement. If long, complex requirements statements are produced
(which is often the case from some techniques used to ascertain user requirements, such as
question and answer sessions) and all of these are tested in a single long Test Script, the
failure of any single point will require the complete test to be rerun.
55
56

Itemised each requirement should have a unique reference number or section number that
can be referenced by the Test Script and cross referenced to the Test Script in a Requirements
Traceability Matrix (RTM).
It is a failing with many accelerated development methodologies (Rapid Development

Approach) that User Requirements do not meet these criteria because they are not sufficiently
well documented. Even when accelerated methods are used to capture User Requirements
(question and answer sessions, conference room pilots, etc.) these should still be documented.
It is also very valuable to have an experienced tester review User and Functional
Requirements, to ensure that they are sufficiently testable.
8.2.2 Start Early
The writing of Test Specifications should be started early. As soon as a first draft of the relevant
User, Functional or Design Specification is available, work should commence on developing the
Test Strategy and the associated Test Specifications. This effectively means that the test team
works in parallel with the development team.
Unless this is done, it is very likely that the test team will not be ready for testing once the
development team has completed the building of the system.
This approach allows the structure of the Test Specification to follow the structure of the
Functional or Design Specification, thereby introducing structure and logic into the testing
programme. Note that the interrelationship between the test documents and the design documents
should be included in the Configuration Management of the system, so that any changes in the
Functional or Design Specifications will trigger a review of the relevant Test Specifications.
8.2.3 Plan for Complete Test Coverage
In many instances, the test team prepares a Test Programme to cover the items they feel are
important, with little or no regard to the items that actually need testing. When designing the
Test Programme, complete coverage of the User and Functional Requirements, and the Design
Details should drive:
What needs to be tested

What the test objectives should be
What Test Cases will cover each of the test objectives
How the Test Cases will be organised and scheduled
Unless this approach is taken it is likely that requirements or details will be left untested, and
that this will only be realised once testing is almost complete. This often leads to additional Test
Scripts being developed at the last minute, which should be avoided if at all possible.
In accelerated development methodologies the standard test coverage usually only covers
Functional Testing (System Acceptance Testing and some System Integration Testing at best).
In these cases additional Test Scripts must be developed to conduct stress testing, challenge
testing, user profile testing and so on.
Basically, Test coverage must meet the needs of the defined Test Strategy.
8.2.4 Insufficient Detail in the Test Scripts
Many accelerated development methodologies have standard off-the-shelf test scripts but
57
many of these have been designed for generic use across all industries. They are often unsuitable
for use in the healthcare industries because they have:
No place to record the reference to the requirement or detail being tested

No place to record the test objective
Insufficiently detailed instructions to set up the test
Insufficiently detailed instructions to execute the test
Ambiguous expected results, or no expected results at all
No place to record results or the Test Script requires just a qualitative indication of test
status rather than a quantitative record of the result
No place to sign for each step or even for the complete test
Insufficient room to record comments or references to attached evidence
These issues are covered in further detail later in the text, but the important issue is that a Test
Script template must be produced which is suitable for supporting testing in a regulated Life
Sciences environment.
8.2.5 Design Qualification Start When You Are Ready
Testing is often seen as the catch-all at the end of the development process, which will identify
and resolve all sorts of errors in the software or system. Although one of the main objectives of
testing is to do just that, this doesnt mean that the software or system should enter testing before
there is a high degree of confidence in the quality of the software.
This can be achieved by following good development practices and by conducting a formal
Design Review or Design Qualification, usually prior to the build commencing, but certainly
before testing starts.
The criteria for releasing the design for testing should be established prior to the Design
Review or Design Qualification commencing. The criteria can be included in a Design
Qualification Protocol. Establishing the criteria prior to starting the review means that a
baseline for acceptable design quality is established. It should be agreed that unless the
acceptance criteria for the Design Qualification are met, testing would not commence. This is
important since a risk-based Test Strategy will make certain assumptions on the development
process and the quality of the design and testing will only be appropriate if these assumptions
are met.
The Design Qualification may include requirements that:
All items are under established Configuration Management

Change Control is established for all Configuration Items
Source Code Reviews have been conducted for all GAMP software category 5 code
All source code is fully commented, including change history
All Requirements and Specifications are up-to-date and approved
The Requirements Traceability Matrix is fully populated and up-to-date
Where the development of the Test Programme has proceeded in parallel with the design, the
Design Qualification may also include a review of the Test Specifications, Test Scripts, the
traceability between Test Scripts and Requirements and test coverage traceability in the review.
While this is certainly not a mandatory part of Design Qualification, it does increase confidence
prior to testing.This ensures that testing is started with a high degree of confidence in the quality
of the design and that the nature of the testing is appropriate to the design.
58
8.2.6 Taking a Configuration Baseline

If the worst happens, and the testing goes badly wrong, code gets corrupted or changed outside
of change control, you will have to start some, or all, of your testing again.
In some organisations, although code, documents and so on will be formally approved at
version 1.00 prior to testing starting, no formal baseline of the system configuration is taken
until after the testing is complete. It is much easier to recover any documentation or code if the
first formal baseline of the Configuration is taken prior to testing commencing. If there is a
problem it is relatively easy to restore the baseline configuration and start again.
It is also good practice to take another baseline after each phase of testing. This means that
the Configuration of the system can be saved at a known point in time, limiting the amount of
retesting needed if something does go wrong.
8.3 Testing in the Life Science Industries is Different
Many good Test Programmes, with an appropriate Test Strategy and well-written Test Cases,
have been ruined because of poor testing practices. When a large number of human errors creep
into a testing programme they call into question the outcome of individual tests and place the
whole Test Programme into doubt. At best, certain tests will have to be repeated again, at worst
the whole Test Programme may have to be repeated and in some cases projects have been
cancelled because of poor testing practices.
This section looks at the human aspects of testing, and what can be done to maximise the
chances of conducting a successful Test Programme in accordance with good testing practice.
One of the main problems when testing a large or complex system is that the size and
complexity of the Test Programme often requires additional testers to be used. In many large
healthcare companies there may be a core of people experienced in testing software based
systems that are GxP critical.
This is not always the case in smaller companies or with suppliers who supply many different
industries. Even in large healthcare companies contract resource may be used during the testing
phase.
Testing in the Healthcare Industries is different from many others. There are four basic
reasons for this:
The testing may be more thorough than in other industries. This is because of the need to
conduct challenge tests to prove the integrity of the application.
The content of the Test Scripts may be greater than in other industries (i.e. expected results
must be clearly included on the Test Script, the Test Objective must clearly be stated, etc.).
The level of proof required to document the test results may be higher than in many other
industries (i.e. quantitative results must be recorded as opposed to a simple pass or a tick
in a box, witness signatures may be required).
There is a need to assess the impact of any test failure or test anomaly with regards to the
GxP impact or any changes required and the extent of any regression testing.
If people conducting tests are not used to working in this environment it is likely that they will
make omissions or errors that will call the outcome of a given test into doubt. When this
happens on a large number of test cases, it calls the validity of the complete testing programme
into doubt.
Steps must, therefore, be taken to prevent this from happening.
59
8.4 Prerequisite Training
Unless the Test Programme is very short, or the entire team is familiar with testing GxP critical
systems, it is worth investing time in providing some training before the Test Programme
commences.
On projects that have slipped behind schedule there is often a tendency to rush into testing in
order to make up for lost time. This is, however, a false economy and the temptation should
be avoided.
There are many advantages to providing such training, including:
Training provides a useful breathing space for the team and allows them to ask any
questions they have about the testing programme.
It provides a basic introduction for those who have not tested in a GxP environment before.
It also provides a useful refresher for those who have tested in the GxP environment before
and who need to get into the right state of mind.
The training should cover:
For those who have not tested in a GxP environment before, a session on regulatory awareness (i.e. why systems are validated and why testing in the Healthcare Industries is different)
An overview of the Test Programme
Specific roles and responsibilities for those involved with the Test Programme
Checking out and checking in of Test Scripts
Conducting tests (executing Test Scripts)
What to do in the case of a test error (i.e. something clearly goes wrong)
What to do in the case of the test anomaly (i.e. something happens and you are not sure
whether it is an error or not)
The life cycle of a Test Script, from checking out, through execution, signing off, checking
in and subsequent review.
The life cycle of a test incident, including raising the incident, assessing the incident,
closing the incident and retesting
It may be useful for these last two items to be demonstrated by stepping through the process
using a real Test Script (although it is unlikely that this will constitute a real test execution).
A number of these topics are covered in outline in the following pages.
8.5 An Overview of the Test Programme
It is useful for the entire team to know what the overall Test Programme looks like. Everyone
should understand the timing and especially the interdependencies that exist between different
phases of testing and individual tests.
It should be made clear that although there will be target dates, the validity of testing must
not be compromised in order to meet such dates. If tests are poorly conducted it is likely that
they will have to be conducted again, thereby negating the benefit of meeting target dates.
8.6 Roles and Responsibilities
There are a number of roles that various people will play in a Test Programme. Everyone on the
team should know what the roles are and who is fulfilling each of those roles. These are
60
described briefly below. The title and scope of each role may vary between organisations, from
Test Programme to Test Programme and, in small Test Programmes, one person may fulfil
multiple roles.
8.6.1 Test Manager
The Test Manager is responsible for managing the overall Test Programme. This includes:
Assigning the execution of Test Cases to Testers (and witnesses)

Monitoring the progress of testing against the programme
Ensuring test errors and anomalies are resolved at an adequate rate.
Further detail on this role is given in Section 8.7.

8.6.2 Lead Tester
On a large Test Programme there may be a Lead Tester who supports the Test Manager from a
technical perspective. This person is usually an experienced tester, familiar with the process of
testing in a GxP environment. If other testers have queries on the process of testing they should
seek advice from the Lead Tester in the first instance.
The Lead Tester needs to maintain an overview on technical issues associated with the testing
process, such as maintenance of test datasets in the correct status (see Section 8.17), the correct
sequencing and execution of tests and the appropriate use of techniques and any tools that are
used to support the testing process.
8.6.3 Tester
As the name implies, a Tester is responsible for the execution of individual Test Scripts,
including the recording of the test results and capturing sufficient evidence. They should have
experience of testing in a GxP environment or should have been adequately trained.
8.6.4 Test Witness (or Reviewer)
Some tests will require the use of a Test Witness; typically where insufficient proof can be
captured. The witness may be a tester who witnesses other tests. Alternatively, key tests such as
System Acceptance Tests may be witnessed by the user or client. See Section 8.11 for further
information.
8.6.5 Quality/Compliance and Validation Representative
It is important that resources are available from either the suppliers Quality organisation, the
users Quality organisation or the users Compliance and Validation organisation. The specific
role that each will play will be determined on a project-by-project basis, by the contractual
relationship. This should be defined in the Validation Plan or the Test Strategy. Their availability
and involvement at the time of testing is important to ensure that issues can be resolved on-thespot and not be subject to interpretation after the event.
Their role is basically to:
Provide advice as to whether a test may proceed when a test anomaly has been raised (see
Section 8.13).
61
Review a percentage of test results after the tests have been conducted (this percentage will
have been defined as part of the Test Strategy). This is to ensure that results are adequately
recorded and that the results justify the decision as to whether the test objective has been
met and the test is a pass or fail.
Provide advice as to the GxP impact of any test failure, proposed corrective actions and the
scope of any retesting (see Section 8.15).
Regardless of whether or not someone from the Quality/Compliance function is available at the
time, all test results and evidence should be of sufficient standard to allow later assessment of
events to be made if required. However, having someone available at the time does reduce the
risk of tests being continued when they should have been aborted or being aborted when they
could continue.
8.6.6 Test Incident Manager
A person may be assigned to initially review and monitor all test incidents. This will include the
categorisation of Test Incidents (see Section 8.14), the assessment of incidents and/or the
assignment of Test Incidents for detailed assessment by others (see Section 8.15).
The Test Incident Manager should monitor the progress of all Test Incidents (often using a
Test Incident Register) and should ensure that the overall Test Manager is kept up to date with
regards to the status of the test incidents (also see Section 8.16).
8.7 Managing a Test Programme
The management of a large and complex Test Programme requires a Test Manager who is both
an experienced tester and a good project manager. A good understanding of the technical issues
involved is required, as is an ability to manage the interpersonal problems that can sometimes
surface. The role involves:
Assigning Testers and Test Witnesses to the execution of individual Test Scripts. This must
be done in a way that makes best use of the available resource and respects the test order
prerequisites (i.e. which tests must be conducted prior to others). Part of this process is an
ability to predict how long an individual Test Script may take to execute. This becomes
especially complex when tests have to be rerun and where it is more efficient to use the
Tester who conducted the previous run (but may not be available, or may not be the best
person to rerun a test due to tester errors in the first run).
Monitoring and reporting on the progress of the Test Programme. This can often be
facilitated by the use of a computerised test management tool (see Section 8.16). This will
often include keeping the overall Project Manager informed of progress and completion
dates of various phases of the testing.
Helping resolving any issues between the Test team and the Quality/ Compliance
organisation (who often see themselves with a different agenda).
Balancing the need to meet test deadlines while still maintaining the integrity of the Test
Programme. This is one of the hardest jobs to achieve. A good Test Manager will
understand that it is better to deliver a properly tested system late than deliver a system that
has not been properly tested on time.
Ensuring that good test practice is observed throughout the Test Programme. This is
achieved by monitoring the test statistics, liaison with the Quality/ Compliance organisation
and by setting a good example with regard to tidiness, adherence to procedures, etc.
62

8.8 Checking Test Scripts In and Out
One of the most common problems is that of Test Scripts getting lost. It is embarrassing when
any Test Script goes missing but doubly so when the Test Script has been executed and the loss
of the test results requires the test to be conducted again.
There should be a process where all Test Scripts are signed out at the start of each test session
and signed back in at the end of the session. In this way it will be known who is responsible for
a Test Script at any given time. Test Scripts should be filed and secured between sessions.
All Test Script pages should be clearly identified (including any attachments) so that if they
are mislaid and subsequently found, they can be returned to the correct Test Script. This should
include the page number, the Test Script name or reference and the run number.
These problems can largely be overcome by the use of a computerised test management tool
that automatically checks Test Scripts in and out.
8.9 Recording Test Results
It is a Regulatory expectation that Test Results are recorded quantitatively not qualitatively. In
simple terms this means recording the actual value or number instead of just writing pass or
fail or ticking a box.
The rationale behind this is to ensure that Test Results can be fully reviewed at a later date. If
the actual result is recorded it is possible to review after the event whether the actual test
objectives have been met.
This is of most critical concern for:
Bespoke (customised) systems (GAMP software category 5)

Highly GxP critical, GAMP category 4 or category 3 software
Customised hardware (GAMP hardware category 2)
For these types of system, the quantitative recording of results should be treated as mandatory,
especially during System Module Tests, Hardware Tests and Software Integration Tests. This
does, however, impose an overhead on testing which may not be appropriate in all circumstances.
In some cases it may be appropriate to make a qualitative record, for example, when:
Separate evidence is provided (see Section 8.12) this is applicable for all testing,
although the test step should make clear reference to the evidence provided.
The Test Script includes a clear range of acceptable results this may be applicable for
systems of low GxP criticality. In this case a mark should be made immediately alongside
the stated range (i.e. ticking in a box alongside text stating 9.997 <= result<= 10.003, or
displayed colour = red).
The test step is not critical to proving that the test objective has been met (i.e. a set-up step).
In these cases it is acceptable to provide a qualitative mark that the step has been executed
(tick a box marked executed).
The testing is part of a standard software test for a standard device (usually GAMP
software category 2 and 3), with a clearly defined testing process.
The testing is part of the normal manufacturing tests for standard hardware (GAMP
hardware category 2).
In the latter two cases, where the test is conducted manually it is applicable to record a
qualitative mark to indicate that a step has been executed and that the expected result is within
63
a range clearly defined in the Test Script. Under these circumstances it is acceptable to write
pass (or tick a box marked pass for each step).
For some automated manufacturing tests the quantitative value or the pass or fail
qualitative status may be recorded automatically.
8.10 To Sign or Not to Sign
Should a Tester sign each step? This is one of the questions that the industry has been pondering
for a long time. This is a mandatory requirement for testing in some healthcare companies but
not in others.
This again imposes an overhead on testing and the question must be answered on a case-bycase basis by considering the following points:
What is the GxP criticality of the test? Systems or functions with a low GxP criticality may
justifiably demand a lower level of signed corroboration.
What is the purpose of the step? If the step is required in order to set up the system for
subsequent steps that actually demonstrate the test objective, a signature may not be
required for the set up step or steps. This is especially the case where the correct set-up is
clearly demonstrated by evidence such as a screen shot or where a witness observes the setup prior to executing the key step.
Where there is doubt about the requirement to provide a full signature, a step may be initialled
as a compromise measure.
Regardless of which steps are signed (or initialled) by the tester, the completed test must
always be signed off and the context of the signature must always be clear. An example of such
a statement could be:
Signed to confirm that all of the above test steps were conducted in a contiguous manner, in
accordance with the test instructions for each step and that the recorded results are an accurate
record of the test execution.
The Test Script should also include a clear statement of whether the test objective was met or
not and whether the test was passed.
8.11 The Use of Test Witnesses
Given the potential GxP impact of testing, it is important that test results cannot be falsified. To
meet the expectations of certain of the Regulatory authorities this includes taking adequate steps
to ensure that test records cannot be falsified.
This may be achieved by attaching test evidence (see Section 8.12) or by the use of Test
Witnesses. The use of Test Witnesses should be decided upon on a pragmatic basis since the use
of a witness adds to the cost of testing and often slows down the testing process.
The guiding principle should be that for GxP critical tests, where no adequate proof of testing
can be obtained by other means (see Section 8.12), a Test Witness will be required to ensure that
the Test Objective has been met (see Section 8.11).
For some tests this may require that the witness observes every step. This is likely where the
test is designed to demonstrate a sequence of events, such as a functional work flow. In this case
the witness must observe every step. Signing each step is not mandatory so long as the Witness
signs at the completion of the test. This signature should include a statement that the context of
the signature is such that the Witness has observed every step and concurs that the recorded
results and/or evidence and any comments or notes that the Tester has made are an accurate
64
record of the test run. Unless such a statement is included, it will be necessary for the Witness
to sign every step.
For other tests, where the initial test steps are only required to set up data or to place the
system in a specific state to prove the test objective, it may not be necessary for the Witness to
observe all of the preliminary set-up steps. In this case the Witness may observe and sign at the
specific test step that proves the test objective or may sign at the completion of the test, with a
suitable statement as to the context of the signature.
A Test Witness must always understand what he/she is observing and must understand the
objective of the test. This may require adequate knowledge of the system or application under
test and may require prior training or briefing.
If a Test Witness does not understand the purpose of the test or any specific step, he/she
should seek clarification before that test or step proceeds. If he/she is not happy that he/she has
understood the objective of the test, if he/she is unsure whether the test objective has been met,
or if he/she believes that the test was not properly conducted, he/she should either refuse to sign
or more properly raise a Test Anomaly to review the situation.
The use of Test Witnesses may also be required for contractual reasons, where a User wishes
to ensure that the Functional Testing does adequately prove that the system meets its business
needs.
8.12 Capturing Test Evidence (Raw Data)
It is sometimes easier to capture test evidence than to record such evidence by hand. This test
evidence may be produced by:
An automated test system or test software, which produces a summary report of the test
steps, input and output data
Reports or other output produced as part of a test
Screen shots of screen data or status
All test evidence should be securely attached to the completed Test Script. Reference should be
made to the associated test evidence at the appropriate step in the Test (in a comments column
or a column specially included to refer to attached evidence).
The creative design of screen shots can be used to record a good deal of evidence that
minimises the need for adding additional data manually. Using standard display windows it is
often possible to compose a screen shot that contains the date and time, the actual evidence in
the form of one or more applications windows and even additional data such as the ID or
network address (IP address, MAC address, etc.) of the workstation the test was conducted on.
When computerised test tools are used the screen shot can be captured electronically and
pasted into a document, which can be electronically linked to the executed Test Script, thereby
doing away with any paper evidence.
For standard manufacturing test reports it is acceptable to sign these and file them without
the report being attached to a specific Test Script.
Depending upon the data contained in the output it may be necessary to manually annotate
the evidence with details such as date/time, serial number of unit under test, test run number,
number of pages, etc. When designing such tests it is always useful to include such data in the
report header.
65
8.13 Proceed or Abort? (Test Incident Management)

If Test Script instructions are not clearly written or if the system under test behaves in an
unexpected manner, it is sometimes difficult for an inexperienced tester to know what to do.
They may choose to continue to execute a test to completion when the test should have been
aborted. This may waste minutes, hours or even days conducting a test that may have to be
performed again.
In other cases they may abort a test when that test could have been continued through to
completion. This may be because the language used in the instructions is not clear to the Tester
or if a system dialogue box contains a message that is slightly different to the expected result.
If this occurs towards the end of a test it can again waste a great deal of time.
If the Tester (or Witness) is in any doubt they should always seek advice from the Lead Tester,
possibly the Test Incident Manager or the Quality/Compliance representative. This can be raised
as an anomaly and investigated informally.
Where the anomaly is minor (a typographical error on the Test Script) and the intent of the
test and the result is reasonably clear, it may be possible to continue the test and raise a Change
Note to revise the Test Script. The justification for this should be recorded as a comment on the
completed Test Script and signed and dated by those involved in resolving the anomaly.
If the resolution is not immediate there are two options:
Record the anomaly as a Test Incident and continue the test to completion, in the
anticipation that the anomaly can be resolved and the results of the test will still be valid.
Abort the test and raise a Test Incident.
It should be noted that if two or three anomalies are recorded during a Test Script execution the
outcome of the test would be very difficult to evaluate and justify after the event. A test should
usually be stopped once a second anomaly has occurred.
Note also that Test Scripts should be reviewed for recorded anomalies. A trend may indicate
a problem with either a specific Tester or the author of specific Test Scripts and a need for
retraining may be indicated.
8.14 Categorising Test Incidents
If a test is aborted a Test Incident should be raised. This can also be raised in the case of an
anomaly that cannot be immediately resolved.
All Test Incidents should be assessed and part of this assessment process includes assigning
a Test Incident Category. Tests Incidents can be categorised as follows:
Test Script error there is an error in the Test Script which caused the test to fail, or which
means that either the instructions were not clear, or could not be executed.
Tester Error the error was as a result of the Tester either making a mistake (mis-keying
data) or not understanding clear instructions or results.
Set-up Error the test set up was incorrectly performed causing the test to fail or meaning
that the test execution could not continue.
Test Data Error there was an error in the test dataset(s), causing the test to fail or meaning
that the test execution could not continue.
External Failure there was an unforeseen external event such as a power failure, causing
the test to fail, or meaning that the test execution could not continue.
Configuration Error there was a problem with the configuration of the system under test,
causing the test to fail or meaning that the test execution could not continue.
66

Application Error there is a fault with the software under test, causing the test to failor
meaning that the test execution could not continue.
Procedural Failure there was a problem with an operating procedure that was being
followed as part of the test, which caused the test to fail or which meant that the test
execution could not continue.
Note that this list is not exhaustive and that other categories may be defined depending upon the
nature of the system under test.
An analysis of the Test Incident Categories will provide an indication of the validity and value
of the testing programme. If most of the Test Incidents are as a result of configuration,
application or procedural errors then this demonstrates the validity of the testing by finding such
errors.
If most of the problems are due to Test Script errors, set-up errors, Tester errors or test data
errors this calls into question the validity of the testing. During a Test Programme the number
of such errors should decrease over time. If an upward trend is detected, or if the number of such
errors exceeds 40% after the first week of testing, a halt should be called to testing. This may
require Test Scripts and datasets to be reviewed again or Testers to be retrained.
Computerised test management tools allow such categories to be assigned to Test Incidents
and statistical data generated.
Note that the capture and analysis of test performance data is one of the key components in
defining and rationalising a risk-based approach to testing (see Section 3.3).
8.15 Impact Assessment
All Test Incidents must be assessed. If there is any doubt as to the GxP impact of the incident,
this assessment should include a representative of the Quality or Compliance department. In
some organisations a representative of the Compliance department will review and assess all Test
Incidents. This may not be necessary where the Test Incident Manager has a sufficient degree of
regulatory awareness to decide whether or not to involve the Compliance representative.
The impact assessment must determine:
Whether or not the test must be repeated (in the case of an anomaly raised as an Incident
during Test Script execution).
Changes that will have to be made to correct the error.
The extent of any re-testing that will be required as a result of the changes made. Note that
this assessment is greatly facilitated by using Configuration Management records to determine the interrelationships between the changed items and any other configuration items.
There must be a formal process for managing Test Incidents. This must define the stages that an
Incident will go through (typically raised, initial assessment, allocation (to a person for detailed
assessment), detailed assessment, approval to fix, fix, approval to close and closed see
Section 8.6 for a typical Incident life cycle). This must also define the roles and responsibilities
of those involved in managing the Test Incident.
Note that the fix may just require the raising of a Change Note and once this has been done
the Incident may be closed. Alternatively, changes may be made under the governance of a Test
Incident (following an approved Test Incident Procedure), so long as all changes are referenced
to the Test Incident.
67
8.16 Test Execution Status
In order to monitor the progress of the Test Programme, it is useful to have an overview of the
test execution status. Figures can be maintained manually or a good computerised test
management tool will provide statistics such as:
Total number of Test Scripts

Number of Test Scripts not executed (and as a percentage of the total)
Number of Test Scripts passed (also as a %)
Number of Test Scripts passed first run (also as a % of the total and as a % of the total passed)
Average number of Test Runs per Test Script
Etc.
These statistics are useful for not just monitoring the progress of the Test Programme, but of the
validity and value of the Test Programme. If a low number of Test Scripts are being passed first
time, or if the average number of Test Runs per Test Scripts is more than 2, this may indicate a
problem with the testing. The Test Incident Category statistics should be reviewed to determine
the cause of the problem (see Section 7.14).
8.17 Test Data Status
Where simple numerical calculations are tested it is relatively easy to define test datasets with
pre-defined input criteria. These can be referenced by a number of tests.
In some complex systems the use of system test data requires significant thought. Where a
test modifies a data object in such a way that it cannot be changed back again (i.e. a document
is moved from draft to reviewed status or a batch is changed from quarantined to released)
that data object cannot be used for the same test. This is because the Test Script assumes that
the data object is in its initial state, which is no longer the case.
This is often made more complex when several tests may be using the same data object. As an
example, one test could change a document from draft to reviewed and a different test changes
it from reviewed to draft. This is even more complicated where the object does not follow a
known sequence of states but can be changed from any state to any state.
This is not a problem when Tests Scripts pass first time but if they fail and have to be rerun
the test data objects many be in the wrong status to repeat the test.
In these cases the management of test data objects needs careful consideration and some
solutions might be:
Sequence tests so that they use test data objects in the status they have been left in by
previously executed tests. This is only likely to work in small or simple systems with a
limited number of tests.
Define groups of test data objects that can be independently saved and restored prior to
tests being executed. The reload of such data object sets would be part of the set-up for
each test. This is only practical in systems which allow partial data object sets to be saved
and restored.
Write the Test Script in such a way as to use any suitable data object that is in the necessary
state to conduct the test. In this case the name of the object would have to be recorded as
part of the Test Script execution, and any expected results would have to refer to the
recorded object name rather than a fixed object name. This then allows any suitable data
object to be used (or created).
68

8.18 Test Log-On Accounts (User IDs)
Clarity of login accounts is crucial when testing systems with defined user profiles.
It is important that actions taken as a Tester and actions taken as a User are clearly
differentiated.
Specific user accounts should be created for the purposes of testing and these should be
granted appropriate privileges or assigned appropriate user profiles for the functions under test.
The allocation and use of these for testing should be recorded. These user IDs should be created
and maintained using appropriate user authorisation procedures. This tests not only the profilebased functionality of the system but also the user authorisation processes as well as the user
management functions within the system.
These login IDs should be used to execute tests. This is to ensure that profiles with
appropriate permissions are used to test the functionality and security of the system. Once
testing is complete these User IDs access rights should be removed from the system, although
a record of the User IDs should be maintained to ensure their uniqueness within the system (i.e.
to ensure that a duplicate ID is not assigned in the future).
Where Testers have their own user IDs these should not be used for testing user functionality.
This is because a Tester may have an inappropriate profile for the test. Where a Tester has to
reconfigure the functionality of the system as part of the test, this should be done using their
individual User ID (to differentiate actions conducted as part of the test script execution using
a defined user profile, and those conducted external to the test script execution as a tester).
There should be no group log-ins for the system and any Testers actions should be directly
attributable to them individually. Where it is necessary to modify any users authority or
permissions this should be done using the appropriate system access and authorisation
procedures.
Test Scripts should explicitly state the login ID to use during testing. Instructions such as
login to the system are ambiguous and should read log-in to the system as user gdg_tester_1.
CHAPTER 9
Supplier System Test Reports/Qualification Reports
Depending upon the nature of the testing, it may be usual for suppliers to produce System Test
Reports that summarise the results of the System Testing.
Depending upon the size and complexity of the testing, this may:
Be incorporated into a single report, detailing all of the system testing. This is most
appropriate for small or simple systems.
Require a separate Test Summary Report for each level of testing (i.e. Hardware Test
Summary Report, Software Module Test Summary Report).
Such Test Summary Reports should include an overview of the test results and may include
statistical data such as:
Total number of tests executed

Percentage passed first time
Total number of test reruns
This should also include a list of any critical test non-compliances and any outstanding issues
that remain, such as system functions that have not satisfactorily completed Acceptance Testing.
The report should include a clear indication of whether or not subsequent phases of testing
may proceed with justifications as to why.
Where the supplier does not produce such reports, this should be included in the relevant
Qualification Report (IQ, OQ or PQ Report, depending upon the level of testing). Where the
supplier does provide such reports, they should be referenced in the relevant Qualification
Report, thereby reducing unnecessary duplication.
Where the Test Report or summary only relates to a single test, or series of tests, it is often
useful to attach the actual test record sheets to the back of the Report. This makes it much easier
to reference the detail recorded in the actual test sheets.
69
CHAPTER 10
The Use of Electronic Test Management and Automated

Test Tools
10.1 The Need for Test Tools in the Pharmaceutical Industry

The Healthcare Industries have a need to reduce the cost of implementing IT systems and
solutions. This is even more important than in other industries because of the relatively high
price it pays for testing its computer systems in a highly regulated environment.
The use of computer based test management and automated test tools (computerised test
tools) can reduce the time taken to develop and execute Test Scripts and can significantly
reduce the number of human errors. These cost savings are accrued by the faster development
and execution of Test Scripts. Significant savings also result from the reduced number of test
execution errors and the resulting reduction in the need to analyse incidents and repeat tests due
to human error.
In addition to the cost savings available, the reduced number of test execution errors and the
ability to easily trace Test Scripts to individual Requirements helps build a strong validation
case. It is not uncommon for well-specified and well-designed systems to have their quality (and
subsequent validation status) badly compromised by a poor set of Test Scripts and test results.
The use of an automated test tool can greatly improve the control of Test Scripts and reduce the
number of test execution errors.
A more significant benefit can also be obtained when systems are being changed or
upgraded. This is especially the case in large or complex systems where many software patches
may need to be applied and the inter-relationship of these patches makes it very difficult to
rationally define the scope of any retesting. This is because not even the software developers
can track all the interdependencies in their complex software packages. In these cases, the only
safe solution is to retest the entire system, as appropriate to the GxP risk (justified in the original
Test Strategy).
The ability to quickly and easily repeat tests (for instance, user profile/security tests) using
automated Test Scripts means that upgrades are significantly faster and less expensive to
implement. This in turn means that the advantages of changes and upgrades are much more
likely to be fully realised and that the validated status is much more likely to be adequately
maintained.
The use of computerised test tools therefore offers significant cost savings and improved
support for validating computer systems and is of benefit to most pharmaceutical manufacturers
and healthcare companies.
10.2 Test Tool Functionality
Computerised test tools basically provide three useful features, namely:
71
72

The ability to manage Test Scripts and Test Incidents in an electronic (paperless)
environment while still executing the Test Scripts manually. Such systems also provide the
tools to plan test campaigns and report on the status of the test campaign, analyse test
results, allocate Test Scripts to Test Engineers and so on.
The ability to automatically execute Test Scripts, including the recording of Test Results.
Most test tool systems in widespread use have standard interfaces to the most popular
system interfaces and have the ability to interact with standard user interfaces written using
standard technologies such as JavaScript, ActiveX controls, Visual Basic, etc.
The ability to load test a system, by simulating the load of many hundreds of simultaneous
users. The ability to repeat such tests is a useful aid to on-going performance monitoring of
systems.
10.3 Electronic Records and Electronic Signature Compliance
Despite the obvious benefits to be gained from using such systems, they are not in widespread
use for testing GxP critical systems. There are two main reasons for this.
Firstly, as described above, such systems can provide significant cost savings and have an
attractive payback period. However, the purchase, implementation, rollout and validation costs
of such test tools are still significant. In order to achieve the anticipated cost savings and
payback it is necessary to initially use the automated test tool in support of a significant IT
project. Examples of this are large ERP systems, the development of standard products such as
Clinical Trial Management systems or global Intranet, Extranet and Internet applications.
The other reason for not adopting such systems has been the lack of regulatory compliance
(test record integrity) required when testing GxP critical systems in the Healthcare Industries.
In order to sell pharmaceutical products, biological products or medical devices into various
markets the safety of such products must be assured. This is facilitated by various Regulatory
agencies (the FDA, MHRA, etc) that enforce various legislation, standards and guidelines in a
number of areas.
So called Predicate Rules deal with issues directly impacting and controlling Good Clinical,
Laboratory, Manufacturing or Distribution Practice. Examples of these are Chapter 21 of the US
Code of Federal Regulations, Parts 210, 211, 606 and 820, EU Directive 91/356/EEC, etc.
The Predicate Rules are supported by additional regulations and guidelines that deal with
issues which may directly or indirectly affect product quality, data integrity (and possibly patient
safety). This includes the use of computer systems for the design, manufacture, testing and
distribution of such products and devices and is directly governed by regulations such as 21CFR
Part 11, Part 211.68, Part 820, 91/356/EEC Annex 11 and so on.
Inherent with these regulations is the use of records and signatures used to support GxP
critical processes. Various computer systems support or control processes that are covered under
the Predicate Rules and such systems are clearly GxP critical. In addition, the testing of such
systems is also covered under supporting regulations. Any system that is used to test a GxP
critical system is, therefore, also considered to be GxP critical (if only indirectly) and whilst test
records are considered to be GxP critical records, they are not specifically mentioned in the
majority of the predicate rules. Such test records are therefore not usually subject to the controls
of 21CFR Part 11, but the integrity of such records does need to be assured in support of the
validation of GxP critical systems.
If an automated test tool is used to test a computer system that supports a process covered
under the Predicate Rules such a system must have adequate security around the approval,
execution, post execution and long term storage of such testing records.
Most large computer systems which would economically justify the use of an automated test
Use of Electronic Test Management and Automated Test Tools
73
tool are GxP critical (at least in part) and any pharmaceutical manufacturer looking to use an
automated test tool should be looking for a tool which can assure the integrity of test records
(possibly supported by full audit trails and the use of electronic signatures).
10.4 The Availability of Suitable Test Tools
The developers of test tool applications were slow to realise the need to comply with industry
requirements for test record integrity and the result is that test tools that can assure test record
integrity are only just appearing on the market.
Even now, the choice of test tools that can be used in a compliant manner is extremely limited,
with only one mainstream manufacturer having considered the specific requirements for test
record integrity. It is expected that a number of other compliant solutions will appear on the
market in the next few years, widening the choice for healthcare companies wishing to use such
systems.
10.5 Test Script Life Cycle
The advantages of using an automated test tool for Test Script management are largely around
the processes (workflow) associated with Test Script authoring, reviewing, approval, execution
and post execution review and approval (Figure 10.1).
Figure 10.1 The Test Script Life Cycle.
Performing the testing in an automated, paperless environment allows it to be done in a more

efficient and more controlled manner, across one or more sites. This is especially useful when
74
Figure 10.2 The Test Incident Life Cycle.
the development and execution of Test Scripts is conducted by a team based on several sites or,
possibly, in different countries.
One useful feature that such systems support is the ability to enter which User or Functional
Requirements are tested by each Test Script. This helps to ensure that complete test coverage is
achieved (untested Requirements can be highlighted).
In order to meet the requirements for test record integrity Test Script may be treated as an
electronic record. Approval and execution signatures may be treated as electronic signatures,
and there should be adequate proof of a formal review (note that this is suggested, and is not
mandated by most predicate rules).
Defining the Test Script Life Cycle allows changes in the electronic record status and points
at which electronic signatures are applied to be clearly defined. This approach also facilitates the
determination of who can do what and when, which in turn makes it easier to define user
profiles (e.g. Test Script Author, Test Witness, Test Run Approver, etc.). This ensures that only
persons with defined access rights and privileges can perform various review, approval and
execution tasks.
10.6 Incident Life Cycle
In a similar way to Test Scripts, a life cycle approach can also be taken for Incident Management
(Figure 10.2). This will define the changes in status for an Incident (e.g. open, assessed,
approved, implemented, closed), where electronic signatures may be used and who is allowed to
sign (again based upon user profiles).
Unlike Test Scripts, Incidents have no draft status as soon as an Incident is raised it is
subject to formal management.
An efficient system should allow Test Script details to be passed directly to an Incident Record
(i.e. Test Script name and reference, step and run number, Requirements reference, tester ID,
etc.). It should also allow non-test related (ad-hoc) Incidents to be raised and managed.
A good computerised test tools system will allow a flexible life cycle to be applied to individual
projects, or even individual Test Scripts. This is useful for GxP risk-based testing where the
75
standard of compliance or quality review may be higher or lower than other areas, based upon the
principle defined in the Test Approaches and Test Strategy.
10.7 Flexibility for Non-GxP Use
In order to be most useful to Life Science companies, a computerised test tool should be capable
of supporting the testing of both GxP and non-GxP systems.
The testing of GxP systems imposes additional requirements in order to comply with specific
regulatory requirements and expectations. Examples of these are the need to capture proof of
test objectives being met (by electronically attaching screen copies for instance), the need for
witness signatures or the need for post execution review and approval.
Some of these are an unnecessary overhead when testing non-GxP systems. A cost effective
automated test tool should allow these additional requirements to be turned off when used to
test non-GxP systems.
However, when testing GxP critical systems it must be impossible for the user to bypass these
requirements (see Section 10.10).
10.8 Project and Compliance Approach
The selection, procurement, development, integration, testing and validation of a computerised
test tool is different from that of other computer based systems. This is because GAMP 4
defines a Tools category for certain types of system that support the implementation of GxP
critical systems or GxP functions.
A computerised test tool is not as critical as a manufacturing system that packages tablets or
sterilises medical devices. It is arguably less critical than a system that maintains training
records for individuals performing GxP critical processes against controlled procedures.
Because a computerised Test Tool does not have a direct impact on product quality or identity
(and thereby has no direct impact on patient safety) it does not require stringent validation.
However, if the system is not secure, the integrity of test records may be called into doubt,
because of the possibility of accidentally modifying or deliberately falsifying test records.
A computerised test tool should therefore be validated but the Validation Plan should assume
that this is a system with low GxP criticality.
10.9 Testing Test Tools
The approach to testing a computerised test tool should be to treat it as a development tool.
Where the tool is in widespread use in the Healthcare Industries testing may be limited to
Functional Testing only. If the system is not in widespread use in the Healthcare Industry,
challenge testing will be required to demonstrate the integrity of the Test Scripts, Test Incidents
and Test Results. This challenge testing should specifically demonstrate that the system is secure.
Good practice would suggest that a system that has not been tested should not be used in
operation, which infers that an automated test tool should not be used to test itself, at least, not
for an initial rollout. This means that an automated test tool should be tested manually in the
first instance. Consequently, testing the system is prone to the potential problems that the
system is designed to replace there is nothing that can be done about this except encouraging
and enforcing Good Testing Practice.
The life cycles defined in the development of the system can also be used as the basis for
developing manual test procedures and the necessary Test Scripts and Incident Management
processes and forms. The same life cycles can be used to manage the Test Scripts and Test
76
Incidents, albeit in a paper based environment. This will help to demonstrate the validity of the
defined life cycles.
Testing an automated test tool brings additional complications because of the terminology.
Clear definitions are required in order to avoid confusion around statements such as the
objective of this Test Script is to test the Test Script Approval process. The potential adverse
impact of this cannot be overstated (nor the potential for finding humour in the testing of a test
system).
If the computerised test tool has been customised, the system will need more comprehensive
testing. Such testing will include Software Module Testing. Additional negative testing or stress
testing should focus on areas of bespoke functional development (such as user profile testing,
audit trail testing, etc.) and the stress testing of the infrastructure.
10.10 Test Record Integrity
Any computerised test tool must be secure. As well as assuring basic test record integrity and
assuring that users cannot readily repudiate any electronic signatures, these security
requirements also include the need to:
Archive test records and test record audit trails, in line with the users record retention policy
Ensure that user editable on-line system documentation (i.e. help files, electronic user
manuals, etc.) are managed as secure files
As stated above, there are currently very few automated test tools that are capable of meeting
such security requirements and even these systems will need additional procedural controls to
be put in place.
A good computerised test tool will provide support for the procedural controls. These include:
The ability to produce lists of users and their assigned user profiles, so that these may be
reviewed on a regular basis.
The ability to require users to change their passwords on a regular basis, to define
minimum length passwords and the ability to prevent the re-use of recently used
passwords.
The ability to support a proceduralised approach to granting user access on a need to do
(user profile) basis.
An ability to manage users system specific training records, including revoking access
rights when retraining is required.
It should be noted that the list of technical and procedural requirements stated above are far
from comprehensive and need to be reviewed with respect to current regulations and guidance.
10.11 Features to Look Out For
Based upon the above, any healthcare company looking to purchase a computerised test tool
should base their decision upon the following criteria:
A proven track record of supplying such tools to the Healthcare Industries

An ability to manage Test Scripts and Test Incidents in a paperless environment
The ability to support traceability between User Requirements, Specifications and
individual Test Scripts or Tests
77
The ability to automatically execute Test Scripts

The ability to support a wide variety of system user interfaces
The ability to perform load and volume testing
The flexibility to efficiently test non-GxP systems, by turning off the electronic records
audit trail and electronic signature capability on a project-by-project basis
CHAPTER 11
Appendix A Hardware Test Specification and Testing
11.1 Defining the Hardware Test Strategy

Assuming that Hardware Category 1 hardware is genuinely standard hardware in widespread
use, specific testing of the hardware is not required. For Category 1 hardware it is sufficient to
demonstrate that the hardware has been manufactured in accordance with the requirements of a
recognised quality system (e.g. ISO 9000).
Where available, standard manufacturers test records may be used to support the validation
case for systems of medium or high GxP Criticality. Specific hardware acceptance testing is not
required since this will be implicitly tested as part of the System Acceptance Testing.
In cases where the hardware is Category 2 (customised), system specific testing will be
required and this may also be the case if Category 1 hardware is used outside the operating
conditions specified by the manufacturer.
The degree of testing of Category 2 hardware should be applicable to the GxP Criticality of
the system being tested. For example, it may be sufficient to test systems of low GxP Criticality
within the normal operating range, without testing outside the range, if there is a low probability
of this range being exceeded in operation.
For systems of high GxP Criticality it may be justified to test outside this range in order that
failure modes may be predicted and contingency plans made. Further examples of this are
discussed below.
It should also be noted that many systems that have some degree of customisation will also
contain many standard system components and that the Test Strategy would need to distinguish
between the two.
The objectives of the Hardware Tests should be clearly traceable to the appropriate sections
of the Hardware Specification. Meeting the objectives of the Package Configuration Test Scripts
should clearly demonstrate that the hardware is built and operates as specified.
Issues to consider for Hardware Testing follow.
11.2 Standard Test Methods
There are many different methods of conducting standard Hardware Acceptance Tests and these
vary according to system type and from manufacturer to manufacturer.
Each method will define a series of steps, conducted in a repeatable sequence in order to
demonstrate that the objectives of the test(s) are met.
It is likely that different test methods will be used for different types of test. For example, the
correct functioning of a standard printed circuit board can be tested using Automated Test
Equipment (ATE). Individual component parts may also be subject to Functional Testing such
as heat-soak tests and burn-in tests. These are usually part of standard manufacturing tests.
Other tests may be project specific and may require particular input conditions to be
physically created, simulated or manually entered.
79
80
Tests may be conducted using a combination of standard or specific test methods that provide
the appropriate challenges to verify hardware operation and performance.
Modern day hardware is often controlled by firmware. Such firmware provides control
functions that assure that the hardware components deliver the stated functionality. Tests on
firmware-controlled hardware must take into consideration GAMP validation categories.
A number of example test methods are described below. Specific test methods will be
described in the actual Hardware Acceptance Test Specification, either in the General section or
in the individual sections of the Hardware Acceptance Test Specification.
The detail in which the test is described may vary according to the general test methodology
described in the General section of the Hardware Acceptance Test Specification. It may be
necessary to provide detailed step-by-step instructions (including providing full details of set-up
instruction or individual keystrokes for the test engineer to follow) or it may be acceptable to
provide general instructions that rely on the test engineer using an overall knowledge of the
hardware system under test.
Where the test method requires prerequisite knowledge of the hardware it should be clearly
stated in the General section of the Hardware Acceptance Test Specification.
Wherever possible, general test methods should be used and documented in the General
section of the Hardware Acceptance Test Specification. In this case it is acceptable for the
individual section of the Hardware Acceptance Test Specification to reference a General section
of the Hardware Acceptance Test Specifications. For example Select channel 1 on analogue
input card AI13 and set the input value to 4mA refer to Section 1.2 of test procedure Testing
Current Input Modules for specific instructions.
When defining the test methods to be used the following issues must be considered.
11.3 Manual Testing of Component Hardware
In some circumstances, printed circuit cards, populated printed circuit boards and system
components (electrical or mechanical) may be subject to manual testing and this will usually be
part of the manufacturers standard Quality Assurance procedures.
Manual tests may vary but usually include the use of equipment such as logic probes, multimeters, customised test rigs, etc.
Systems should be in place in order to ensure that:
Documentary evidence of the individual board or item under test (including art work
revision number, firmware version number, drawing number and serial number) is
generated
The test rig the test was performed on is referenced
Testers details are recorded
The name and version number of the test procedure used is documented
The date and time of the test is recorded
The test results are clear and accurate
Details of any failure are documented
11.3.1 Automated Test Equipment

A large proportion of testing of unpopulated printed circuit cards and populated printed circuit
boards is conducted using Automated Test Equipment (ATE). This is usually part of the
manufacturers standard Quality Assurance programme and has the advantage of being thorough
and efficient.
Hardware Test Specification and Testing
81
These systems can provide sufficient evidence of testing so long as supporting systems are in
place to provide:
Documentary evidence of the individual board under test (including art work revision
number, firmware version number and serial number)
The ATE machine the test was performed on
The ATE test jig used (including drawing revision or version number)
The revision of the ATE Test Programme being run
The date and time of the test
The test results
Details of any failure
Supplier audits should challenge the tests conducted using ATEs. Tests should sufficiently test
operating limits to ensure that the board is thoroughly tested, not only for correct operation but
also for items which could lead to subsequent failure, such as dry joints, short circuits, etc.
Should any individual board fail ATE testing, a system should be in place to record the details
of the fault, any rework and subsequent re-testing as described above for manual testing. A
maximum number of retests should also be defined before a card or board is scrapped as
unusable (before the cost of retesting and rework outweighs the value of the board).
11.3.2 Burn-In/Heat Soak Tests
Burn-in tests are often conducted to identify circuit boards that may pass a detailed test, but still
may be likely to premature failure due to manufacturing defects or individual component
failures.
Such tests are again likely to be part of the manufacturers standard Quality Assurance
procedures. These tests should provide documentary evidence of:
The boards or items under test (including art work revision number, firmware version
number and serial number)
The equipment the test was performed on
The revision of the Test Programme being run (a good burn-in test will usually sequence
through a series of temperature changes, possibly including a heat soak test at elevated
temperature)
The test results
Details of any failure(s)
The manufacturer should again have a system to record the details of any fault(s), any rework
and subsequent retesting following the heat soak test (for instance, how many failures will be
allowed? Will ATE testing be required again before another heat soak test? etc.).
11.3.3 Standard Integrated Hardware Tests
Testing of integrated hardware components is essential to ensure that they collectively meet
performance requirements and that hardware event synchronisation functions correctly.
These tests may, for instance, test such basic functions as:
Network communication
82

Field communication
Data storage and retrieval
System performance under normal and maximum loading
Error handling and recovery (including power outages)
The following information must be recorded:
Documentary evidence of the component hardware items under test (usually just serial
number at this stage)
Details of any support equipment used for the test
The revision number of the test procedure being tested
The test results
Details of any failure(s)
These tests are very important where standard hardware is being used in a non-standard
configuration or architecture and systems of medium or high GxP Criticality should stress test
points at which the hardware is integrated to ensure that interfaces are predictable and robust.
This is even more important where custom hardware from different manufacturers has been
integrated into a complete system.
This may also be useful for medium or high GxP Critical systems when standard hardware
(Hardware Category 1) from different manufacturers is integrated in a non-standard or unusual
manner. As a minimum, the compatibility of such components may be verified by comparing
equipment specifications, written confirmation from suppliers, etc.
11.3.4 Automated Testing
Conducting the Hardware Acceptance Tests can be a time consuming and expensive process
given the need to fully test every component piece of hardware. This task can be eased by the
use of automated testing and most manufacturers have invested in this technology to reduce
their manufacturing and QA costs.
Automated testing combines the functions of simulation and automatic data recording to fully
automate the task of hardware testing. Such facilities lend themselves to the testing of many
similar hardware devices and can either conduct a single test at a time or can conduct many tests
in sequence or parallel.
Automated testing is likely to be part of the standard manufacturing tests but if it is necessary
to develop project specific automated testing applications these should be controlled and
reviewed as part of the Hardware Acceptance Test Specification.
11.3.5 Hardware Acceptance Test Methods
The system hardware is an essential part of a complete validated solution and correctly
conducted and documented Hardware Acceptance Tests are an integral part of the validation
documentation set.
Hardware is implicitly tested as part of a System Acceptance Test and standard hardware
(Hardware Category 1) does not require hardware acceptance testing. Where the hardware is
customised (Hardware Category 2) hardware acceptance testing is required.
The nature and extent of hardware acceptance testing must be applicable to the GxP Priority
of the system. Where it is possible to segregate hardware into different levels of GxP Priority
Hardware Test Specification and Testing
83
(which is usually only possible for certain types of manufacturing, control or automation
systems) different hardware components may be subject to different levels of testing.
As a minimum, Hardware Acceptance Tests must test the correct operation of the hardware
across the full operating range. For systems of medium or high GxP Priority it is also useful to
test the system outside the normal operating range in order to predict what may or will happen
(the failure mode). Specific emphasis must be placed on the areas of customisation.
An example of why this is necessary could be a customised control system, the individual
components of which are designed to operate between 0 and 50C. The actual environment in
which it is installed may be controlled to between 18 and 25C but if the HVAC system fails, what
is the minimum and maximum temperature that may be reached? If the temperature were likely
to exceed 50C (in certain enclosed areas of a production line for instance) what would the effect
be on the customised system? If the customised hardware repeatedly fails in a catastrophic and
unpredictable manner at just 42C, it may be necessary to require that all manufacturing
operations cease if the HVAC system fails. If the system monitors ambient temperature and
performs an orderly shutdown at 50C, with no critical data being lost, it may be possible to state
that manufacturing operations may continue up to 45C, subject to on-going temperature
monitoring. All of the above assumes that no personnel work in the area and that the product is
not subject to any adverse effects in such conditions.
When determining the extent of such tests it is, therefore, necessary to understand the nature
of the customisation and the associated risks, the GxP Criticality of the system (or function) and
the expected failure mode.
The complexity of such testing highlights the fact that standard (Category 1) hardware should
be used wherever possible.
Careful consideration must be given to the use of standard test methodologies, efficient test
sequencing, appropriate control and utilisation of tools, equipment and personnel.
11.4 Performance Baseline
In order to ensure that meaningful performance monitoring may be conducted during the
operational phase of the system, it is useful to establish a hardware performance baseline during
hardware testing.
This would include taking hardware performance measurements (or capturing records) under
controlled conditions that could also be established during the operation of the system. These may
include:
The number of hardware (device) errors over a specified period of time

Signal/noise ratio measurements
Verification of screen display clarity (under known light conditions, at defined distances,
etc.)
Number of network errors and retries with a known data loading
It is useful to measure performance in a known clean environment so that any problems with
the system hardware can be distinguished from installation problems. This baseline can be
established as part of the FAT and repeated as part of the SAT.
This is especially useful with performance that may degrade over time, such as system
grounding (earthing), screen degradation, etc.
CHAPTER 12
Appendix B Package Configuration Test Specifications

and Testing
12.1 Defining the Package Configuration Test Strategy

The overall combination of Package Configuration Verification, Functional Testing and stress
testing needs to be appropriate to:
The GxP Criticality

Any prior experience with the package (that is documented)
The size and complexity of the overall system
The level and nature of verification/testing of each package will be unique to each system,
should be well thought out by experienced personnel and documented and justified as part of
the Test Strategy.
The objectives of the Package Configuration Tests should be clearly traceable to the
appropriate sections of the Package Configuration Specification. Meeting the objectives of the
Package Configuration Test Scripts should clearly demonstrate that the system is configured as
specified.
12.2 Configurable Systems
Many software applications require a number of configuration settings to be made. These
involve the setting of various parameters that determine how the system (or software module)
will function.
The setting of these parameters may involve:
The entry of numerical settings in a field (such as a constant, a multiplication factor, etc.)
The logical setting of a switch (on/off, true/false)
The selection of one of a number of options (from a pick list or one out of a series of radio
buttons in a dialogue box)
Many different types of system require configuration parameters to be set, including ERP
systems, LIMS systems, Distributed Control Systems, etc. Such systems are diverse in the way
specific functions will be performed and the setting of the configuration parameters will
determine exactly how the software module will function. Examples of the changes a
configuration switch could make include:
Whether or not a control loop is forward or reverse acting (positive or negative gain).
85
86

Whether or not a Purchase Order requires one or two signatures and the value of purchase
order for which a second approval signature will be required.
Whether or not data from a sample can be saved with or without a supervisors signature
being required.
Once User Requirements and Functional Requirements have been determined and documented
(in the User Requirements Specification and Functional Specification) the configuration
settings need to be determined and documented in the Package Configuration Specification.
There should also be clear traceability between the individual setting and the corresponding
Functional and/or User Requirement.
A large or complex system may contain tens or hundreds of different modules, each of which
may be of different GxP Criticality and a consistent method of determining the appropriate level
of verification and/or testing should be defined and justified. This can serve to focus the
verification/testing on the modules of medium or high GxP Criticality while reducing the
overall level of testing.
Depending upon the GxP Criticality of the system, module or function the Package
Configuration then needs to be verified and/or tested.
12.3 Verifying the Package Configuration
Verifying the Package Configuration is simply confirming that the configuration settings have
been made as specified in the Package Configuration. This may be accomplished in a number
of ways, dependent upon the GxP Criticality:
For systems or functions that are not GxP critical it is acceptable for the person configuring
the system to confirm the correct setting of the configuration parameters by completing a
simple checklist (tick-boxes, with a single confirmation signature at the bottom).
The package may be configured as defined in the Package Configuration and the person
configuring the system is required to confirm each step either by signature or, preferably,
by recording the actual configuration setting made. Since there is no independent
verification of the Package Configuration this is only suitable for packages, applications
or modules that have a low GxP Criticality. This would only be acceptable where an
independent test of the Package Configuration is to follow, and where there is a high
likelihood that any errors in the configuration would be detected in the subsequent testing.
The package may be configured as defined in the Package Configuration and the person
configuring the system is required to document each step by taking screen shots and by
signing and dating them (as required for other test evidence).
The package may be configured as defined in the Package Configuration and a witness
independently records the configuration settings at the time they are made.
After the system is configured an independent reviewer checks the system configuration
settings against the Package Configuration Specification. Proof of such a review would be
required, such as the completion of a summary sheet designed for the purpose. Dates and
signatures should be recorded for either method.
These last three methods are suitable for verifying the configuration of systems or functions of
medium or high GxP Criticality.
Any person involved in setting the configuration parameters should do so with a copy of the
approved Package Configuration Specification in front of them. Anyone making, witnessing or
reviewing such settings should also be sufficiently familiar with the package, application or
Package Configuration Test Specifications and Testing
87
module so that they can easily interpret the settings documented in the Package Configuration
Specification and can tell when any errors are made.
For packages or modules that are not GxP Critical (or of low GxP Criticality) configuration
verification is still a cost effective activity due to the reduction in subsequent errors that arise
when a system is incorrectly configured.
For simple systems that are widely used in the Healthcare Industries, which have few
configuration settings and where the GxP Criticality is low, it may not be necessary to conduct
any further testing although this decision would need to be documented and justified. Similarly,
where the use of a system and the resultant configuration settings are identical to other systems
that the organisation has formally tested before it may be sufficient to just verify the
configuration settings.
12.4 Functional Testing of the Package Configuration
As stated above, the configuration of a software module determines the exact way in which it
will function. Verifying the correct configuration is a useful and worthwhile exercise but
verification alone will only confirm that the settings have been made correctly and does not
prove that the system functions in accordance with the System Requirements. For most
applications this will still need to be formally tested.
In many cases it will be adequate to perform simple Functional Testing (i.e. to prove that the
system meets the Users and Functional Requirements and functions as specified) and not to
stress test the system (see Section 12.5).
This is justifiable in cases where:
The package is a relatively simple one (with limited variation in functionality) and in
widespread use in the regulated Healthcare Industries.
The package is functionally identical to a system that has already been stress tested (and
the configuration has been verified as such).
The GxP Criticality is low.
Such Functional Testing is conducted as part of the Software Integration Testing or System
Acceptance Testing and would not require additional Package Configuration Testing.
12.5 Stress Testing of the Package Configuration
In some systems it will be necessary to not only functionally test the configuration of the system
(to prove that it performs as specified during normal operations) but also to stress test the
system in order to determine the effects of abnormal operations.
This may be required where:
The system is large or complex

The package, application or module is not in common use in the Healthcare Industries (or
this is the first time a particular organisation has used the package)
The system is of medium or high GxP Criticality
In such cases the purpose of this additional testing is to demonstrate how the system will handle
errors, incorrect attempts at operation or illegal operations.
Examples of such stress testing may include:
88

Attempting to enter illegal configuration settings and conflicting combinations in order
to determine the effects of such settings (systems should, of course, reject illegal settings
when attempts are made to enter them).
Deliberately attempting to operate the system incorrectly, typically by attempting to
perform GxP critical functions when logged in as a user with insufficient privileges (so
called user profile testing).
Attempting to deviate from the normal operational sequence when logged in as a privileged
user, in order to confirm that the system correctly handles errors and/or enforces the
correct sequence of operations (a requirement of 21 CFR Part 11).
Such stress testing can be conducted either as part of the Software Integration Testing or,
perhaps more usefully, as a set of specific Package Configuration Tests conducted prior to the
Software Integration Testing.
12.6 Configuration Settings in Non-Configurable Systems
Even systems that are non-configurable (i.e. the function is fixed, as in the case of software
Category 2 or 3 or the function is determined solely by coding, as in the case of software
category 5) may still have a number of configuration settings. These may typically be made at
the operating system or database level and can impact on the security of the system, the integrity
of the data, etc. Typically these configuration settings are used to control security on folders, the
enabling of audit trails, etc.
Where they have a direct impact on product quality or data integrity, they should be defined
as part of a Package Configuration Specification, which may be combined with a more general
Software Design Specification. As a minimum, the setting of these should be verified as part of
the system Installation Qualification and subject to formal change control.
CHAPTER 13
Appendix C Software Module Test Specifications and Testing
13.1 Defining the Software Module Test Strategy

Software Module Testing is required for all systems that include software that has been
customised.
Where the entire system has been developed as a one-off solution, all of the software modules
will require detailed Software Module Testing. In the case of a system which has limited
customisation, only those software modules that are new or have been modified need to be
subject to Software Module Testing.
Software modules that have been formally tested previously need not be subject to Software
Module Testing on a project so long as:
The supplier (or possibly the user) can reference the previous test results.
The software module has been under formal change control as part of a recognised quality
system and is demonstrably the same version.
No other changes in the system (such as the hardware, application version, database
version, operating system version, etc.) invalidate the previous testing.
The Test Strategy should clearly identify which software modules are subject to Software
Module Testing, which are excluded and the rationale for this.
The objectives of the Software Module Tests should be clearly traceable to the appropriate
sections of the Software Module Design Specification(s). Meeting the objectives of the Software
Module Test Scripts should clearly demonstrate that the software is designed and operates as
specified.
13.2 Examples of Software Modules
While a full definition of modular software is outside the scope of this guideline, examples
could include:
C/C++
Visual Basic
Ladder Logic
Java Script
DOS batch files
13.3 Stress (Challenge) Testing of Software Modules
Prior to testing, all software modules subject to testing should have undergone a formal source
code review and should be under change control.
89
90
The nature and extent of the testing should be appropriate to the GxP Criticality of the
specific module. Modules which are non GxP-critical should be subject to good testing practice
and test results may be recorded by use of a simple checklist. Software modules that are GxPcritical require formal Test Scripts to be developed.
Modules with a low GxP criticality may be subject to Functional Testing within their normal
operating range, with no requirement for stress testing on the basis that there is a good
likelihood of any fault being discovered in a subsequent stage of testing.
Modules of medium and high GxP criticality should be subject to stress (or challenge) testing
across a wide range of illegal or invalid inputs:
Enter discrete inputs (illegal patterns such as device feedback signals from a valve or a
motor indicating that a valve is both open and closed or that a motor is both running and
stopped)
Analogue inputs (above or below normal ranges)
Operator inputs (out of range or illegal/random garbage inputs)
Illegal conditions (similar to discrete input patterns as above but with a wider range of
input types simultaneously tested)
The degree to which the tests attempt to break the module (i.e. cause the software to freeze,
enter a recursive loop, cause a hardware or operating system error, cause unexpected or illegal
output, etc.) should be applicable to the GxP criticality and complexity of the module. Modules
that are too complex to test adequately should be deconstructed into a greater number of
separate modules.
CHAPTER 14
Appendix D Software Integration Test Specifications

and Testing
14.1 The Purpose and Scope of Software Integration Testing

The purpose of System Integration Testing is basically to ensure that the different components
of a system or solution function correctly together. This may be where different applications are
integrated into a single solution (for instance, a standard Laboratory Information System is
integrated with a standard mail server to send e-mails or where a standard weighing and
dispensing system is integrated into a standard ERP system). This is also required where
customised GAMP category 5 software is included as part of a solution.
Where the solution is mature, properly developed and well proven in the market place and the
user has taken appropriate steps to confirm this, there is usually no need to conduct System
Integration Testing; System Acceptance Testing alone will suffice. However, System Integration
Testing will be required when:
The system is a complex one, consisting of many software components and it is not a
mature product, well proven in the market place.
Any existing module is customised in any way. Its interfacing and integrating with other
modules must be tested but the extent of such testing may be limited to the interfaces with
the customised software. Note that this may also include the reconfiguration of GAMP
category 4 software (Package Configuration) where the new configuration is significantly
different from that which is commonly used in the industry.
Any new functionality is developed within a system.
Prior to System Integration Testing the individual applications or modules will have either been
defined as mature and standard in the market place or will have been subjected to Software
Module Testing at an appropriate level. The purpose of Software Integration Testing is,
therefore, to ensure that the various software components all function correctly together.
At a higher level, Software Integration Testing is also used to test the interfaces between
different systems. In order to evaluate the functioning of the interface it is necessary that the
individual systems (or at least those functions in the system responsible for handling the
interface) have been fully tested.
When defining Software Integration Tests the purpose is not to challenge test the individual
modules or systems, but to ensure that the modules or systems function correctly together in a
robust, reliable and predictable manner. Where appropriate, stress testing (or challenge testing)
may be part of this.
The objectives of the Software Integration Tests should be clearly traceable to the appropriate
sections of the Software Design Specification(s). Meeting the objectives of the Software Design
91
92
Test Scripts should clearly demonstrate that the overall system is designed and operates as
specified.
14.2 System Integration Tests
When designing appropriate tests the following questions should be asked:
What would happen if one of the modules or systems stopped responding or functioned in
an abnormal manner? Would the integrated modules or systems also stop functioning?
What would happen if the modules or systems become unsynchronised? Are all messages
appropriately identified to allow responses to be correctly attributed to queries? Is there
sufficient buffer capacity to store queries whilst waiting for outstanding responses?
Is the system infrastructure capable of handling the anticipated traffic? This may apply to
network components as well as the data bus in an embedded system?
Can the system detect data that has been corrupted by, or between, different software
modules or components? The ability of individual modules to reject out of range or corrupt
data should have been checked at the software module level, but this is worth checking for
critical functions.
Are there incompatibilities in data formats? This may apply to numerical formats (such as
the length of integer words or floating point numbers) or encoding standards, such as
binary coded decimal.
What would happen if a software module, component or device would not relinquish
control of a network or carried on transmitting continuously? What effect would this have
on network traffic and system performance?
Are appropriate checks on the validity of originating modules or devices made? Would it
be possible for another device or component to be attached to the system and would this
be detected?
What would happen if the user of one system attempted to gain access to a different
system? Are the user profiles recognised by all components in the system? Are users with
inappropriate levels of access rejected by all modules and functions? This is especially
pertinent if the application is reliant upon user authentication at the operating system or
network level.
Such questions help to determine the nature of the System Integration Tests and also how
challenging such tests should be. Performing a risk analysis will help to determine where the
critical risks lie and where test effort should be focused.
CHAPTER 15
Appendix E System Acceptance Test Specifications

and Testing
15.1 The Purpose of System Acceptance Testing

All challenge testing should have been conducted as part of the previous levels of testing. The
purpose of System Acceptance Testing is to demonstrate that the system functions in accordance
with the requirements of the Functional Specification. This basically means that it functions
how it was designed to perform.
As part of this, some of the User Requirements may be tested as well (it does what it is
supposed to do) and System Acceptance Tests can form part of both the Operational
Qualification and the Performance Qualification.
System Acceptance Testing usually has some contractual significance and is usually
witnessed by the user. Sign off of the System Acceptance Tests is usually defined as a milestone
in the project and may well be the point at which the system is handed over to the user.
15.2 The Nature of System Acceptance Testing
System acceptance testing is functional in nature. System Acceptance Tests are usually designed
around the normal functioning of the system and serve to demonstrate that the system functions
as specified in the Functional Specification. Part of the tests may be conducted as part of the
Factory Acceptance Tests and may be completed as part of the Site Acceptance Tests.
Normal operation and work flow will be demonstrated. This may be picking and stacking
pallets, the raising of a purchase order, the approval of a CAD drawing and so on. These
functions will be tested or witnessed by the actual users of the system (profile challenge testing,
to demonstrate that other users are not able to perform these tasks is conducted prior to the
System Acceptance Tests).
The objectives of the System Acceptance Tests should be clearly traceable to the appropriate
sections of the Functional Specification and possibly to the User Requirements Specification.
Meeting the objectives of the System Acceptance Test Scripts should clearly demonstrate that
the system is functioning as specified.
15.3 Establishing a Performance Monitoring Baseline
The test results taken during the System Acceptance Tests can be used to establish a baseline for
the future performance monitoring of the system. When the system is first installed the
operation of the system should be at its peak (hard discs are relatively empty, network traffic
should be within design limits, overheating should not be a problem and system grounding and
earthing should be optimum).
93
94
This may degrade over time, but unless a suitable set of baseline test data are established, it
will be impossible to make any meaningful comparisons. Although not formally part of the
System Acceptance tests (unless certain criteria are included in the User Requirements or
Functional Specification) it is nevertheless useful to include such testing.
This may be conducted during the Performance Qualification phase of the system testing,
where reliable data can be established over an appropriate time scale.
CHAPTER 16
Appendix F Risk-Based Testing
GxP Priority is just one risk factor to consider when determining the degree of risk associated
with a particular software module, application or function.
For instance, a software module may be of high GxP criticality but if it has been tested by the
supplier, used by hundreds of other pharmaceutical companies and has not been modified, the
risk of using such a module may be low and it may require little or no testing.
On the other hand a software module may have been customised by a development team
working under extreme time constraints, with no prior experience of conducting source code
review, with no coding standards, using a language which is relatively new to the market and
which they have never used before. If this module is of medium GxP Priority, does it really
require less testing that the module of high GxP criticality?
The test strategy must therefore take a number of factors into account and these can be given
a weighting factor. GxP Criticality and GAMP software and hardware categories are just two
factors that may be taken into consideration. The overall risk can then be calculated and used in
the development of an appropriate testing strategy.
The table below provides an example of such a weighting, with various risk factors included.
This is obviously a more complex approach than one solely based upon GxP Criticality and
software/hardware category and may only be applicable to large organisations involved in testing
a wide range of systems on a more-or-less continuous basis. If organisations decide to use such
an approach they may wish to discuss and agree risk factors, categories and weightings
appropriate to their own organisation (Table 16.1).
The different system components can be assessed against the risk factors and categories
above and a scoring given based on the sum of the weighting factors. Different modules can then
be assigned the different test approaches.
As an example, if the above weightings were used, the maximum score would be 100. For
software this may equate to (Table 16.2).
95
96
Table 16.1 Example of System Risk Factors

Risk Factor
Category
Weighting
GAMP Software Category

1
2
0
1
Risk Factor
Category
Weighting
GAMP Hardware Category

1
2
0
10
Risk Factor
Category
Weighting
GxP Priority
Low
5
Medium
10
High
20
Risk Factor
Category
Weighting
Business Criticality
Low
5
Medium
10
High
20
Risk Factor
Category
Weighting
Software Module Complexity

Low
1
Medium
3
High
5
Risk Factor
Category
Weighting
Established History of Module/Application

Mature
Established
1
3
New
5
Risk Factor
Category
Weighting
Speed of Development
Slow
5
Normal
3
Fast
10
Risk Factor
Category
Weighting
Supplier Track Record

Excellent
Good
0
3
Average
5
Poor
7
None
10
Risk Factor
Category
Weighting
History of Development Tools

Excellent
Good
1
2
Average
3
Poor
5
New
5
Risk Factor
Category
Weighting
Size of Development Team

1
2
3
1
34
3
59
4
10+
5
3
3
4
5
5
10
97
Table 16.2 Example of Test Approaches Based Upon Risk Factors

Rating
1 to 10
Test Approach
No specific testing required. Will be tested as part of overall System Acceptance Testing (Functional Testing).
25% of System Acceptance Test Specification and Results are subject to Quality function review and
approval.
11 to 20 Will be tested as part of overall System Acceptance Testing (Functional Testing). Testing outside standard
operating ranges is required in order to predict failure modes.
100% of System Acceptance Test Specification and Results are subject to Quality function review and
approval.
20 to 35 In additional to System Acceptance (Functional) Testing, the system must be subjected to stress testing
during normal operating conditions to challenge:
Network security
Quality function review and approval.
100% of all System Acceptance Test Specifications and Results are subject to independent Quality function
review and approval.
36 to 50 In additional to System Acceptance (Functional) Testing, the system must be subjected to comprehensive
stress testing across normal and abnormal operating conditions in order to challenge:
Network security
Quality function review.
review.
51 to 65 Software Module Testing mandated prior to System Integration Tests and System Acceptance Testing.
Only testing within standard operating range required
25% of Software Module Test Specifications and 10% of all Software Module Test Results are subject to
25% of all Software Integration Specification and related test Results subject to independent Quality
function review.
review.
modes.
50% of Software Module Test Specifications and 50% of all Software Module Test Results are subject to
66 to 80 50% of all Software Integration Specification and related test Results are subject to independent Quality
function review.
review.
modes.
function review.
review.
CHAPTER 17
Appendix G Traceability Matrices
The following tables summarise the traceability between the various activities and deliverables
associated with the development of Test Specifications and the associated testing. It should be
noted that the actual interrelationship of these would usually be defined in the overall Validation
Plan.
For each of the tables, input activities are listed on the left. Dependant activities to which they
may have an input are listed along the top of the table. This means activities across to the top
are traceable to activities down the left hand side only when there is a tick (3) in the intersecting
cell.
99
Test Strategy
Hardware Specification
Functional Specification
Software Design
Specification
Package Configuration
Specification
Project Quality Plan
Software Module
Specification(s)
Validation Plan
Software Module
Test Specification
Hardware Test
Specification
Activities along the top

are traceable to activities
listed below
Table 17.1 Test Specifications Traceability
Package
Configuration
Test Specification
17.1 The Development of the Test Specifications
Software
Integration Test
Specification
System Acceptance
Test Specification
100
Applicable section of Hardware

Specification
Applicable section of Functional

Specification
Applicable section of Software

Design Specification
System
Acceptance
Test Scripts
Applicable section of Package

Configuration Specification
Software
Integration
Test Scripts
Package
Configuration
Test Scripts
Test Specification
Applicable section of Software

Module Specification(s)
Test Strategy
Software Module Test Specification
Software Module
Test Scripts
Hardware Test
Scripts
Validation Plan

listed below
Table 17.2 Test Script Traceability
17.2 The Development of the Test Scripts
Traceability Matrices
101
Test Strategy
Hardware Specification, Test

Specification and Test Scripts
Software Design Specification,

Integration Test Specification
and Test Scripts
Package Configuration Specification,

Test Specification and Test Scripts
Tested Package Configuration
Tested Software Modules
Software Module Specification,

Test Specification and Test Scripts
3
Tested Hardware
Software Module
Testing
Hardware Testing
Validation Plan

listed below
Table 17.3 Test Execution Traceability
Package
Configuration
Testing
17.3 Test Execution
Software
Integration
Testing
System
Acceptance
Testing
102
Software
Integration
Test Report
System
Acceptance
Test Report
Overall
System
Test Report
3
Installation
Qualification
3
3
3
3
3
3
Overall System Test Report
Tested System
System Acceptance
Test Report
Software Integration
Test Results
System Acceptance
Test Results
Test Report
Tested Package Configuration
Software Integration
Test Report
Test Results
Software Module Test Report
Performance
Qualification
Operational
Qualification
Tested Software Modules
Software Module Test Results
Package
Configuration
Test Report
Hardware Test Report
Software
Module
Test Report
Hardware
Test Report
Tested Hardware
Hardware Test Results

listed below
Table 17.4 Test Reporting Traceability
17.4 Test Reporting and Qualification
Validation
Report
Traceability Matrices
103
CHAPTER 18
Appendix H Test Script Templates
The following example provides details of a Test Script Template and can be used as the basis
for developing project specific versions. An example sheet is also included.
18.1 Basic Template for a Test Script
Note that fields to be completed during the actual test or as part of the acceptance are shaded.

Individual System Acceptance Test Record Sheet
Project No:
Project Name:
Test Reference:
System Function under test:
Functional Description of System Function under test:
Specific Test Prerequisites (not included in General Section)
Specific Test Method:
Description of Test:
Test Methods to be used:
Detailed Instructions:
Page 1 of 2
105
106

Project No:
Project Name:
Test Reference:
Input
Expected Result
Unacceptable Result
Actual Result
Maximum number of Repeat Tests:

Test Run Number:
Test Result (tick ONE box below):
Pass:
Reset/Further Actions:
Test conducted by:
(Print Name):
Date:
Test Reviewed/Accepted by:
(Print Name):
Date:
Page 2 of 2
Fail:
Test Script Templates
107
18.2 Example of a Specific Test Script
Note that this Test Script does not contain detailed instructions and references a test method.
Fields to be completed during the actual test or as part of the acceptance are shaded.

Project No:
Project Name:
Test Reference:
AC95003
Acme Pharmaceuticals Building 30
CFG 034 (Revision 1.02)
Centrifuge Operation
Functional Description of System Function under test:

The system will automatically increase the speed of the centrifuge, from rest to a minimum
speed of 3,000 rpm over a period of between 1 minute 30 seconds and 2 minutes. The
centrifuge will run at a minimum speed of 3,000 rpm for a period not less than two minutes
and not exceeding 2 minutes fifteen seconds. The centrifuge shall then decelerate, achieving
rest in not more than 30 seconds.
Page 1 of 4
Specific Test Prerequisites (not included in General Section)

None
Specific Test Method:
Equipment Module Test See General Section, 1.2.4.
Description of Test:
The selected equipment module (Centrifuge 034) will be placed in manual mode and the status
set to manual default. The operation of the centrifuge connected to the test system will then
be observed.
Test Methods to be used:
The test will be initiated via manual entry of data into the control faceplate for Centrifuge 034.
The operation of the Centrifuge will then be observed.
Detailed Instructions:
1. Select the faceplate for Centrifuge 034. If not already in MAN mode, select MAN mode.
2. Set the Status of Centrifuge 034 to run (entering the Engineers password when prompted).
3. Observe and note the time taken to reach full running speed, the time at full speed and the
time to decelerate. Note the speed reached (displayed on the centrifuge local display) at the
end of the run up period and the lowest speed observed during the run.
4. Record the observed values displayed as the comparison data value.
5. Select the control faceplate for Centrifuge 034. Select AUTO mode.
Page 2 of 4
108

Project No:
Project Name:
Test Reference:
Software Module under test:
AC95003
Acme Pharmaceuticals Building 30
CFG 034 (Revision 1.02)
Centrifuge Operation
Input
Expected Result
Unacceptable Result
Run up time
>1m30s, < 2m00s
<1m30s, > 2m00s
Speed reached
3000 rpm +/ 50
<2950, > 3050
Run time
>2m00s, < 2m15s
<2m00s, > 2m15s
Minimum run speed
2950 rpm
< 2950rpm
Deceleration time
< 30s
> 30s
Maximum number of Repeat Tests:
Test Run Number:

Test Result (tick ONE box below):
Pass:
Page 3 of 4
Reset/Further Actions:
Test conducted by:

(Print Name):
Date:
Test Reviewed/Accepted by:
(Print Name):
Date:
Page 4 of 4
Fail:
Actual Result
109
18.3 Example of a Test Script with Detailed Instructions

The following test script provides detailed instructions for each step (in this case, for testing an
ERP system).
Integration Test Scenario <Test Scenario Reference>
This is a GxP Critical Test and the following test controls MUST be applied:
Pass Criteria
In order for a Process Step to pass the test, the following criteria must be fulfilled:
All test prerequisites and conditions must have been set-up prior to test execution.
All Test Scripts associated with the Process Step must have been successfully executed,
with test evidence provided as described below
The tester shall sign off each Test Script, to confirm that the Test Script was executed as
documented, that the actual results matched the expected results, that the test acceptance
criteria were met and that there were no test anomalies.
In order for a Business Requirement to pass the test the following criteria must be fulfilled:
All of the Process Steps associated with the Business Requirement shall have passed as
defined above and shall have been reviewed by the Test Stream Leader and the
implementation stream Validation Consultant.
All test incidents shall have been resolved.
The Business Process Test Review Section shall be signed by the Test Stream Leader and
the implementation stream Validation Consultant.
Test Evidence
Evidence shall be provided to demonstrate the successful execution of each individual test
script. Evidence is not required for each step. Evidence shall be provided either by:
Attaching a screen shot of the test outcome to the executed Test Script, which clearly
demonstrates that the acceptance criteria has been achieved. Attachments shall be signed
and dated by the Tester OR
A witness shall observe the execution of the entire Test Script, and shall sign and date the
witness section of the Test Script sign-off.
<reference> version (N.n)
VALUE/CODE
DESCRIPTION
Quality Inspection in Production <Test Scenario Reference>

Production Order with Quality Inspection using Inspection Points
Recording Quality Results for Inspection Point Processing
<Acceptance Criteria for Test>
4.
5.
6.
7.
R =:
r =:
S =:
s =:
E:
M:
0:
E:
M:
0:
<Mi Services>
<Draft>
Record is stored electronically

Manual (paper record) is stored
No record is stored (required)
Signature is applied electronically
Hand-written signature is applied (usually to a paper record)
No signature is used
Inspection type 03
COMMENTS AND NOTES
OWNER:
STATUS:
RUN NO.:
RUN
DATE:
User Profile identifies User Profile used for testing

Tester name must be entered in full for each step where the main Tester is not conducting the transaction test. The main Tester should initial each step.
Pass or Fail must be entered in full for each step.
Comment references should be recorded in the final column for all comments added on the final sheet. Where applicable, comments shall be used to record details of any
attachments.
Records required by the applicable Predicate Rules

Record required for other business purposes (default)
Signature required by the applicable Predicate Rules
Signature required for other business purposes (default)
NOTES and INSTRUCTIONS

1.
Run number and date shall be entered at the top right of this page for every test.
2.
Expected results reference set-up data unless otherwise stated.
3.
GxP Code identifies 21CFR Part 11 requirements in accordance with the specified Predicate Rules.
SETUP DATA
DATA OBJECT
Company Code
Plant
Storage Location
Material Master
Routing/Ref. Op. Set
Test Data Set (where used)
SCENARIO:
BUSINESS CASE:
DESCRIPTION:
ACCEPTANCE
CRITERIA:
110
BUSINESS PROCESS
STEPS/BPP NUMBER
Release production order.
Review Inspection Lot

created as a result of the
release of the production
order.
Results recording for

an inspection point
for characteristics
or specifications
for the inspection lot.
Defects recording for

characteristics/
specifications.
Usage decision for the

inspection lot.
No
QA11 or
QVM3
QE11 or
QE51 or
QE71
QE11 or
QE51 or
QE712
QA03
CO02
TRANS.
CODE
Inspection lot
number.
Inspection lot
number, operation
and inspection
point identification.
Inspection lot
number operation,
and inspection
point identification.
Material or
Production order
number
Production Order
Number.
INPUT DATA/
SPECIAL
INFORMATION
Usage Decision made.
Results will be evaluated

as accept or reject
according to the data
recorded for the
characteristics or
specifications for each
inspection point. Also a
confirmation will be posted
to the production order.
Inspection lot
number
Order status changed

to Released.
Inspection Lot created.
EXPECTED
RESULT
TRANSACTIONAL STEPS < Test Scenario Reference> <Run Number> < Run Date>
GxP
CODE
USER
PROFILE
TESTER
NAME or
INITIALS
RESULT
COMMENT
REF

111
Comment
Date (dd-mm-yy)
Date (dd-mm-yy)
Signed (Tester)
Witnessed (if required)
PASS/FAIL (Delete as appropriate)
Number
Comments: <Test Scenario Reference> <Run Number> < Run Date>
112
CHAPTER 19
Appendix I Checklists
The following checklists provide a list of useful questions to ask when preparing a System Test
Specification. By answering these questions the author of such a document will be able to
determine whether or not the System Test Specification can usefully be started and completed
and whether or not the actual System Acceptance Tests can be conducted.
Note that the some of the questions and answers will be specific to the type of Test
Specification being developed (Hardware, Software Module, Package Configuration, System
Integration or System Acceptance). Only when all the answers in a particular section have been
answered should the relevant work commence.
19.1 Checklist 1
Before Starting the Test Specification
Question
Answer
Validation Plan/Project Quality Plan

Is the Validation Plan/Project Quality Plan available?
If it is, is it signed-off and released?

Is the Validation Plan/Project Quality Plan clear in specifying
a need for this Test Specification?
If it states that one is not required, does the suppliers own

Project Quality Plan require one?
Does the Validation Plan/Project Quality Plan allow this Test
Specification to be combined with any other document?
If it does, which other documents?

Does the Validation Plan/Project Quality Plan place specific
requirements on the supplier for reviewing this Test Specification?
If not, what requirements are there in the suppliers own

Project Quality Plan?
Does the Validation Plan/Project Quality Plan require the user to
review and accept this Test Specification?
Yes/No
Yes/No
Yes/No
Yes/No
Yes/No
Yes/No
Yes/No
113
114
Question
Answer
Test Strategy
Is the Test Strategy
An integral part of the Validation Plan?
An integral part of the Project Quality Plan?
An integral part of a Qualification Protocol?
A separate document?
Is there a separate Test Strategy Document detailing
The testing rationale?
The relationship between the various types and phases of testing?
The relationship between the suppliers FAT, SAT and

the IQ, OQ and PQ?
Specification
Is the Appropriate Specification available
(Functional, Hardware, Software, Package Configuration
or Software Module)?
Yes/No
Yes/No
Yes/No
Yes/No
Yes/No
Yes/No
Yes/No
Yes/No
Yes/No
How does the Specification impact or influence the System

Acceptance Test Specification?
19.2 Checklist 2
The Development of the Test Specification
Question
Answer
Does the title block of the document include:
The project name?
The document title?
An up-to-date revision number (subject to proper change control)?
Details of the author(s)?
Space for approval signatures?
Yes/No
Yes/No
Yes/No
Yes/No
Yes/No
Are details of the document QA process referred to in the document?
Yes/No
Is the scope of the document clearly defined, including:
The relationship with other validation life cycle documents?
The reasons for the grouping and ordering of tests?
Details of functions not tested and why?
Allowable deviations from general test methodology?
Yes/No
Yes/No
Yes/No
Yes/No
Has a glossary been included or referenced?
Yes/No
Are general principles and test methodology clearly explained, including:
The principles of testing?
Standard test methods?
Yes/No
Yes/No
Checklists
115
How results should be recorded?

How tests are signed-off and accepted?
Yes/No
Yes/No
Is sufficient detail included or referenced for all of the above?
Yes/No
Are general test prerequisites clearly explained, including:
Required system hardware and software?
Required test equipment?
Required test software?
Required test datasets?
Required process, plant or equipment?
Required test documentation?
Prior tests?
Is sufficient detail included or referenced for all of the above?
Yes/No
Yes/No
Yes/No
Yes/No
Yes/No
Yes/No
Yes/No
Yes/No
19.3 Checklist 3
Before starting any Test Script
Question
Answer
Validation Plan/Project Quality Plan/Test Strategy

Is the Validation Plan/Project Quality Plan/Test Strategy available?
Yes/No
Yes/No
Does the Validation Plan/Project Quality Plan/Test Strategy provide

sufficient details about the format and quality of the Test Scripts?
Yes/No
Does the Validation Master Plan/Project Quality Plan/Test Strategy

specify who will be required to sign-off the Test Scripts?
Yes/No
Does the Validation Plan/Project Quality Plan/Test Strategy specify

whether or not the user will be required to approve individual
Test Scripts?
Yes/No
Does the Validation Plan/Project Quality Plan/Test Strategy specify

whether or not a user review of the Test Scripts will be part of the
Operational Qualification or not?
Yes/No
Functional/Design Specification
Is the appropriate Functional/Design Specification available?
Yes/No
Yes/No
How do the individual sections of the Functional/Design Specification

impact or influence the individual Test Scripts?
Test Specification
Is the appropriate Test Specification complete?
Yes/No
116
How much detail is included in the Test Specification and how much is
required to be written in the Individual Test Scripts? Is there any gap or
overlap?
19.4 Checklist 4
The Development of the Individual Test Scripts
Question
Does each test have:
A unique test reference?
A test name?
Details of, or reference to, a complete functional description

of the test item?
Answer
Yes/No
Yes/No
Yes/No
Are specific test prerequisites clearly explained, including:
Prior tests?
Yes/No
Yes/No
Yes/No
Yes/No
Yes/No
Yes/No
Yes/No
Is sufficient detail included or referenced for all of the above where

they differ from details in the General Section of the System Acceptance
Test Specification?
Yes/No
Are specific principles and test methodology clearly explained, including:
A description of the test?
Standard test methods?
Expected results (pass/fail criteria)?
How the test results are recorded?
How tests are signed-off and accepted?
How and when tests are repeated?
Reset actions?
Setting up for subsequent tests?
Yes/No
Yes/No
Yes/No
Yes/No
Yes/No
Yes/No
Yes/No
Yes/No
Is sufficient detail included or referenced for all of the above where they
differ from details in the Test Specification?
Yes/No
19.5 Checklist 5
Before Conducting any System Acceptance Test
Question
Is the full System Acceptance Test Specification to hand?
Answer
Yes/No
Yes/No
Checklists
117
Is the relevant individual section of the System Acceptance Test

Specification easily to hand?
Yes/No
Yes/No
Is the relevant Test Record Sheet available and to hand?

Are all the prerequisites items available and properly set up?
Prior tests?
Yes/No
Is any other necessary documentation to hand?
Yes/No
Are the Test Engineer and Witness suitably qualified/trained/

experienced to conduct and sign-off the test?
Yes/No
Yes/No
Yes/No
Yes/No
Yes/No
Yes/No
Yes/No
Yes/No
19.6 Checklist 6
Prior to Signing-Off a System Acceptance Test
Question
Answer
Were the step-by-step, documented instructions properly followed?
Did this include set-up, prerequisites, testing and reset?
Yes/No
Yes/No
Was the Test Engineer suitably qualified/trained/experienced to

conduct and sign-off the test?
Yes/No
Were the results as expected (as per documented acceptance criteria)?
Yes/No
Were the results properly recorded and documented?
Yes/No
Was the number of test failures within the allowable maximum?
Yes/No
Was the Test Script properly signed and dated by the Test Engineer?
Yes/No
CHAPTER 20
Appendix J References and Acknowledgments
20.1 References
[1] Good Automated Manufacturing Practice (GAMP) Guide, version 4. Published by ISPE
[2] Guidance for Industry, 21 CFR Part 11; Electronic Records; Electronic Signatures (draft
guidance from the FDA, published February 2003)
20.2 Acknowledgments
From a personal perspective, my thanks go to Kevin Francis of Mercury Interactive and to
Andrew Dickson of G4 Consulting for their insights into state-of-the-art computerised test
management tools.
To everyone at Mi Services Group; thanks for encouraging us all to get involved with
defining industry best practice, and especially to Rachel and Nichola for correcting my
grammar.
I should also mention all those colleagues who I have had the privilege (and sometimes the
misfortune) to test with over the years. They are too numerous to mention, but in their own way
they have all contributed to the development of this guide.
My special thanks go to Chris Reid of Integrity Solutions for his useful comments when
originally editing this guideline, and to Tim Cronin at Fluor, for his constructive review
comments quite a task (how they found the time is quite beyond me).
Thanks are also due to Sue Horwood, who remains dedicated to the task of producing and
publishing practical guidance for the industry at affordable prices.
And finally, from a personal perspective, to my family for all their encouragement and
patience and for sticking with this guide through its many incarnations.
Your support is all very much appreciated.
119
Index
A
Accelerated development methodologies, 56
Acceptance
criteria, 43, 49
test, expected results, 50
Activities
interfaces between, 27
life cycle, 28
Automated testing, 1, 35
B
Bar code scanner, 23, 33
Bespoke systems, 14, 62
C
Challenge testing, 34
Change Control, 57
Classified data, 12
Commercial Off the Shelf (COTS) systems,
40
Compliance and Validation (C & V), 20
department, 5
representative, 60
Computer systems
reason for testing, 6
testing, systematic approach to, 1
Configuration
baseline, 58
Error, 65
Management, 24, 27
Corrupted data, 11
COTS systems, see Commercial Off the Shelf
systems
C & V, see Compliance and Validation
D
Data
classified, 12
corrupted, 11
illegal, 11
input, manual, 33, 42
out-of-range, 11
played back, 44
raw, 64
recording, 51
status, 67
Datasets, required, 45
Department of Regulatory Affairs, 5
Design
Qualification, 57
Specifications, 17, 18, 23, 48, 55
Development life cycle, test specification,
2738
conducting tests, 3238
formal acceptance of test results, 3638
manual data input, 3336
test methods, 32
constraints on development of test
specification, 31
constraints on testing, 31
document evolution, 2931
inputs to development of test specification,
29
milestones in process, 2829
outputs from testing, 38
recommended phasing, 27
Development methodologies, accelerated, 56
Document(s)
evolution, 29
life cycle, 22, 28
reference, 47
scope of, 39
Test Strategy, 23
walkthrough of, 29
121
122
E
Enterprise Resource Planning (ERP), 6, 15
ERP, see Enterprise Resource Planning
Error
Configuration, 65
Test Script, 65, 66
typographical, 65
F
Factory Acceptance Testing (FAT), 16, 17,
37
results, 24, 25
Specifications, 3
Failed tests, repeated, 51
FAT, see Factory Acceptance Testing
G
Good testing practices, 5568
capturing test evidence, 64
categorising Test Incident, 6566
checking Test Scripts, 62
common problems, 5558
design qualification, 57
insufficient detail in test scripts, 5657
planning for complete test coverage, 56
starting early, 56
taking configuration baseline, 58
untestable requirements, 5556
impact assessment, 66
management of Test Programme, 61
overview of Test Programme, 59
preparation for success, 55
prerequisite training, 59
proceed vs. abort, 65
recording Test Results, 6263
roles and responsibilities, 5961
Quality/Compliance and Validation
representative, 6061
Tester, 60
Test Incident Manager, 61
Test Manager, 60
Test Witness, 60
signature, 63
test data status, 67
test execution status, 67
testing in life science industries, 58
test log-on accounts, 68
Test Witnesses, 6364
Group log-ins, 68

GxP Criticality, 8, 13, 63
data classified as, 12
system, 17
GxP Priority, determination of, 7
H
Hardware
customised, 62
required, 44
serial numbers, 16
testing, 23
criticality, 8, 910
expected results, 50
Test Specification, 3, 13, 15
Healthcare industries, 57
I
Illegal data, 11
Impact assessment, 66
Information Systems (IS), 5
Information Systems and Technology, 21
Input criteria, pre-defined, 67
Installation Qualification (IQ), 13, 17
protocols, 24
Report, 25
Instancing, 49, 51
IP address, 64
IQ, see Installation Qualification
IS, see Information Systems
L
Laboratory Information Systems (LIMS), 15
Lead Tester, 60
Life cycle
documents, 22, 28
management, 3
Model, 13
Life science
industries, testing in, 58
organisations, 6
LIMS, see Laboratory Information Systems
Log-on accounts, 68
M
MAC address, 64
Manufacturing test reports, 64
Master Validation Plan, 22
Model
Life Cycle, 13
Index
Standalone Systems Lifecycle Activities
and Documentation, 14
O
Operating System, 7, 12
Operational Qualification (OQ), 17, 18
OQ, see Operational Qualification
Organisation(s)
approaches to risk-based testing, 8
Life Science, 6
standard of testing in, 5
Out-of-range data, 11
P
Package Configuration Test Specification, 3,
13, 15, 17
Performance Qualification (PQ), 13, 18
Power failure, 65
PQ, see Performance Qualification
PQP, see Project Quality Plan
Prerequisite training, 59
Project Manager, 4, 21
Project Quality Plan (PQP), 4, 13, 22, 23, 27,
29
Purpose, 1
Q
QA, see Quality Assurance
Quality Assurance (QA), 5, 11, 20, 29, 36, 39
Quality versus Financial functions, 7
R
Raw data, 64
Reference documents, 47
Regulatory compliance, testing and, 5
Requirements Traceability Matrix (RTM), 56,
57
Results
automatic recording of, 35
manual recording of, 33
recording of, 43
Riskbased testing, 8
RTM, see Requirements Traceability Matrix
S
SAT, see Site Acceptance Testing
Scope, 34
applicability of guideline, 3
guideline audience, 4
123
what guidelines covers, 3
Server, new application installed on, 14
Site Acceptance Testing (SAT), 15, 16, 17, 46
results, 24, 25
signing off, 36
Specifications, 3
witnessed, 37
Software
bespoke, 14
design, testability of, 4
Integration Testing, 34
Integration Test Specification, 3, 13, 15
Module Test, 15, 17
Case, 39
Specification, 3, 13
required, 44, 45
simulation, 33
testing, 23
criticality, 8, 910
team, supplier, 22
SOP, see Standard Operating Procedure
Standalone Systems Lifecycle Activities and
Documentation Model, 14
Standard Operating Procedure (SOP), 32
Stress testing, 34
Supplier
Project Quality Plan, 22
Quality Assurance, 20, 36
responsibilities of, 19
software test team, 22
standard testing, 17
System Acceptance Test, 34
results of, 18
Specification, 3, 13, 15
witnessing of, 20
System GxP criticality, 17
System Test(s)
methods of conducting, 32
Specification(s), 3, 21, 24
integrating or omitting, 14
logical sequence to be followed when
developing, 27
System Test Specification, recommended
content for, 3953
general section, 4147
appendices, 4647
general principles and test methodology,
4144
general test prerequisites, 4446
124
glossary, 41
individual test cases, 4753
acceptance criteria, 4951
cross reference to functional description
or design detail, 48
data recording, 51
further actions, 5152
name of hardware item, software
module or function under test, 47
particular test methods and test
harnesses, 4849
specific prerequisites, 48
unique test reference, 47
use of separate test record sheets, 5253
overview, 3941
front page/title block, 39
QA review process, 39
scope of document, 3941
T
Test(s)
acceptance, 43, 44
approval of, 37
coverage, plan for complete, 56
datasets, 34, 45
data status, 67
evidence, capturing, 64
execution status, 67
grouping of, 40
harness, 33, 36, 48, 49
hooks, 34
Incident
categorising, 65
life cycle of, 59
management, 65
Procedure, 66
Manager, 60
methods, 32
Objective, 48, 58
preparation for next, 52
prerequisites, 44
Programme, 59, 60, 61, 67
rationale, 8
Record Sheet, 52, 53
reference, unique, 47
repeated failed, 51
reports, manufacturing, 64
results
formal acceptance of, 36
recording of, 62
-rigs, 45
Script(s), 32, 37
checking, 62
error, 65, 66
execution, 68
insufficient detail in, 56
sequencing, 46
software, required, 45
Witness, 60, 61, 63
Test, what to, 712
GxP Priority, 7
software/hardware category, 78
testing or verification, 1112
test rationale and test policies, 811
Testing, 11
automated, 35, 42
challenge, 34
constraints on, 31
efficiency, 46
life science industries, 58
minimised, 6
necessity of, 6
outputs from, 38
practices, poor, 58, see also Good testing
practices
principles of, 41
reason for, 56
cost savings, 6
Quality Assurance department
requirements, 5
regulator requirements, 5
standard, 6
risk-based, 8
site acceptance, 15
software module, 15
stress, 34
Test Specification(s)
constraints on development of, 31
evolutionary development of, 30
Factory Acceptance, 3
Hardware, 3, 13, 15
inputs to development of, 29
milestones in development of, 28
Package Configuration, 3, 13, 15
relationship between, 1314
Site Acceptance, 3
Software
Integration, 3, 13, 15
Index
Module, 3, 13
System, 21
Acceptance, 3, 13, 15
integrating or omitting, 14
types of, 13
Test Strategy, 1325, 27
document, 23
integrating or omitting system test
specifications, 1416
hardware acceptance test specification
and testing, 15
integrating test specifications and
testing, 16
package configuration test specification
and testing, 15
software integration test specification
and testing, 15
software module test specification and
testing, 15
system acceptance test specification and
testing, 1516
Matrices, 8
relationships with other life cycle phases
and documents, 2225
Design Specifications, 23
factory/site acceptance test results and
IQ, OQ and PQ, 2425
System Test Specifications, 24
tested software and hardware, 2324
Validation Plan and Project Quality
Plan, 2223
125
relationship between test specifications,
1314
riskbased rationale, 13
role of factory and site acceptance tests,
1618
roles and responsibilities, 1822
Information Systems and Technology,
2122
Project Manager, 21
supplier, 19
supplier Quality Assurance, 20
supplier software test team, 22
user Compliance and Validation, 2021
Training, prerequisite, 59
U
User
Acceptance Testing, 24
compliance and validation, 20
IDs, 68
Requirements, ambiguous, 55
V
Validation
life cycle, 1
Master Plan (VMP), 4, 13, 21
Plan (VP), 4, 19, 27
Verification, 11
VMP, see Validation Master Plan
VP, see Validation Plan

Testing Computers Systems For FDA Mhra Compliance Computer Systems Val - ebooKOID

Uploaded by

Copyright:

Available Formats

Testing Computers Systems For FDA Mhra Compliance Computer Systems Val - ebooKOID

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Testing Computers Systems For FDA Mhra Compliance Computer Systems Val - ebooKOID

Uploaded by

Copyright:

Available Formats

TESTING

Sue Horwood Publishing

This edition published in the Taylor & Francis e-Library, 2005.

The Test Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

Testing Computer Systems for FDA/MHRA Compliance

5.5.6 Information Systems and Technology . . . . . . . . . . . . . . . . . . . . . . . . . 21

The Development Life Cycle of a Test Specification . . . . . . . . . . . . . . . . . . . . . . . 27

Recommended Content for System Test Specification(s) . . . . . . . . . . . . . . . . . . . 39

Good Testing Practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

Testing Computer Systems for FDA/MHRA Compliance

8.2.1 Untestable Requirements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

Supplier System Test Reports/Qualification Reports . . . . . . . . . . . . . . . . . . . . . . 69

The Use of Electronic Test Management and Automated Test Tools. . . . . . . . . . . 71

Appendix A Hardware Test Specification and Testing . . . . . . . . . . . . . . . . . . . . 79

Testing Computer Systems for FDA/MHRA Compliance

11.3.2 Burn-In/Heat Soak Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

Appendix B Package Configuration Test Specifications and Testing. . . . . . . . . 85

Appendix C Software Module Test Specifications and Testing . . . . . . . . . . . . . 89

Appendix D Software Integration Test Specifications and Testing . . . . . . . . . . 91

Appendix E System Acceptance Test Specifications and Testing . . . . . . . . . . . . 93

Appendix F Risk-Based Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

Appendix G Traceability Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

Appendix H Test Script Templates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

Appendix I Checklists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113

Testing Computer Systems for FDA/MHRA Compliance

Appendix J References and Acknowledgments. . . . . . . . . . . . . . . . . . . . . . . . . 119

Example of Software Testing Criticality. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

The Relationship between Test Specifications and Test Activities . . . . . . . . . . 14

David Stokes, Spring 2003

The purpose of this guideline is to:

Demonstrate the value of a systematic approach to computer systems testing (why we

Provide a detailed guide to the recommended contents of computer systems test

Provide guidance on the use of automated test tools in a compliant environment.

2.1 What This Guideline Covers

The cost/benefits of conducting an appropriate degree of system testing.

A checklist of questions to be used when developing System Test Specification(s)

Hardware Test Specification

Testing Computer Systems for FDA/MHRA Compliance

This guideline is of value to:

Testing Computer Systems for FDA/MHRA Compliance

Reduce the effectiveness and efficiency of the system at go live.

How much testing to conduct?

Testing Computer Systems for FDA/MHRA Compliance

GAMP 4 GxP Criticality

Table 4.2 Example of Hardware Testing Criticality

GAMP 4 GxP Criticality

No specific hardware or software testing is required.

Basic system (log-in) access

All Test Specifications and Results are subject to peer review.

Basic system (log-in) access

All Test Specifications and Results are subject to peer review.

Testing Computer Systems for FDA/MHRA Compliance

Testing Computer Systems for FDA/MHRA Compliance

Checking the version of an installed operating system.

The Test Strategy

Hardware Test Specification

Further specific information on each of these types of testing is given in Appendices 1 to 5.

Testing Computer Systems for FDA/MHRA Compliance