Validating Clinical Trial Data Reporting with SAS

by Carol I Matthews, Brian C Shilling

Overview

Validation is a critical component to programming clinical trial analysis. Essential to effective validation is the programmer's understanding of the data with which they'll be working. If you don't understand how the data is arranged, the values that are reasonable for each variable, and the way the data should behave, you cannot ensure that the final result of your programming effort is complete or even appropriate. Therefore, to be a successful programmer in the pharmaceutical industry, you need to understand validation requirements and to learn how to make the code do the bulk of the work so that your programs are efficient as well as accurate.

Product Details

ISBN-13:	9781599941288
Publisher:	SAS Institute
Publication date:	04/01/2008
Series:	SAS Press
Pages:	224
Product dimensions:	7.47(w) x 8.94(h) x 0.44(d)

Table of Contents

Preface     ix
Acknowledgments     xi
Pharmaceutical Industry Overview     1
Introduction     2
Regulations     2
Health Insurance Portability and Accountability Act     2
The Code of Federal Regulations     3
Guidance for Industry     4
International Conference on Harmonisation of Technical Requirements     5
Clinical Data Interchange Standards Consortium     6
Documentation     7
Standard Operating Procedures     7
Companywide Standard Operating Procedures     7
Department Standard Operating Procedures     8
Task Standard Operating Procedures     8
SAS Programming Guidelines     9
Quality Control versus Quality Assurance     9
Patient versus Subject     10
Conclusion     10
Validation Overview     11
Introduction     12
Validation versus Verification     12
Why Is Validation Needed?     13
Presenting Correct Information     13
Validating Early Saves Time     13
Developing a Positive Relationship     14
How Do You Approach Validation?     14
Start with All the Information     14
Have a Plan     15
Make the Code Do the Work     16
Ask Questions     16
Be Proactive     16
Validation Must Come First     17
Validation Methods     17
Independent Programming     18
Peer Review     19
Validation Checklists     20
Software Development Life Cycle     21
Conclusion     22
Documentation and Maintenance     23
Introduction     24
Starting the Process     25
Study Protocol     25
Annotated Case Report Form     26
Statistical Analysis Plan     29
Meeting Minutes     31
Internal Program Documentation     32
Program Header     32
Body Comments     34
Output Titles     35
External Documentation     35
Data Definition Tables     35
Program Directory     36
Validation Files     37
Make Programs Maintainable     38
Create and Follow Naming Conventions     38
Make It Easy to Read     38
One Program, One Purpose      42
Comments, Comments, Comments     43
Use Macros Judiciously     44
Make Data Maintainable     44
Order Your Data     44
Label Everything     49
Attach Formats Sparingly     50
Consistency Is Key     51
Good Housekeeping     51
Look-but Don't Touch     53
Conclusion     56
General Techniques to Facilitate Validation     57
Introduction     58
Validation Tools     58
Procedures     58
SAS Options and Language Elements     67
Using Macros Effectively     72
Techniques That Facilitate Validation     80
Start with a Clean Log     80
Print Only What You Need-When You Need It     81
Tracking Problems     82
Using PROC TRANSPOSE or an Alternative Solution     85
Tracking Dropped Data     89
Conclusion     93
Data Import and Export     95
Introduction     96
Validating the Import Process     96
Validating the Export Process     98
General Items to Watch For When Transferring Data     99
Working with SAS Files      100
SAS Data Sets     100
SAS Transport Files     101
Working with Other File Types     102
Microsoft Excel Files     102
Flat Files     103
Common Procedures Used for Validating Data Transfers     104
Proc Contents     104
Proc Compare     108
Conclusion     112
Common Data Types     113
Introduction     114
Study Populations     114
Safety     115
Intent-to-Treat     115
Efficacy     116
Common Data Domains     116
Subject Demographics     116
Inclusion/Exclusion Criteria     117
Subject Disposition     118
Medical History     118
Physical Examination     120
Vital Signs     120
Treatment Exposure     122
Concomitant Medications     123
Adverse Events     124
Clinical Laboratory Data     126
Conclusion     137
Reporting and Statistics     139
Introduction     140
Pre-Output Validation Steps     140
Code Review      140
Log Review     141
Output Validation Steps     142
Understanding the Data     142
Understanding the Output     143
Checking the Result     143
Cross-Checking Related Output     146
Checking the Cosmetics     153
Updating the Specifications     157
Keeping What Is Important     157
Final QC Steps     158
Conclusion     158
Sample Quality Control Checklists     159
Sample Statistical Analysis Plan     163
Glossary     181
References     195
Index     197