Evolving Data Standards Implications for Statistical Analysis and

Evolving Data Standards: Implications for Statistical Analysis and Review Stephen J. Ruberg, Ph. D. Eli Lilly & Company Sept 2004

Outline Background The Language of Clinical Data CDISC Mission / Scope • The World of Standards • CDISC & HL 7 Goals • Computable, Semantic Interoperability • Implications for Statistics 23 Sept 2004 FDA/Industry Statistics Workshop Company Confidential Copyright © 2000 Eli Lilly and Company 2

In the beginning. . . And the Lord said “Behold, they have all one language…. Nothing that they now propose will be impossible for them. Come, let us go down and there confuse their language that they may not understand one another’s speech. ” Genesis 11 23 Sept 2004 FDA/Industry Statistics Workshop Company Confidential Copyright © 2000 Eli Lilly and Company 3

The World is Getting Smaller What enables you to travel the world and still be able to use your credit card at most ATMs? Standardization of the electronic banking world This creates value for the banks and the users 23 Sept 2004 FDA/Industry Statistics Workshop Company Confidential Copyright © 2000 Eli Lilly and Company 4

Pharma 2010 IBM Business Consulting Services analysis of the pharma industry in 2010 • New forms of data management, integrated databases and network of communication channels to share clinical data • Increased use of rolling dossiers • Pervasive connectivity with patients requires data standards and interoperable environments • 23 Sept 2004 FDA/Industry Statistics Workshop Company Confidential Copyright © 2000 Eli Lilly and Company 5

FDA News, Press Release July 21, 2004 FDA Draft Guideline on Study Data Specifications • Study Data Tabulation Model (SDTM) developed by CDISC can be used for Clinical Trial Submission Data • “. . . Having the data presented in a standard structure will improve FDA’s ability to evaluate the data and help speed new discoveries to the public. ” Lester M. Crawford, Acting FDA Commissioner. • FDA is currently exploring regulatory approaches to require the use of the STDM standard for regulatory submissions. 23 Sept 2004 FDA/Industry Statistics Workshop Company Confidential Copyright © 2000 Eli Lilly and Company 6

The Language of Clinical Trials Language has two components Structure: syntax, grammatical rules • Content: vocabulary • Vocabulary has attributes Well-defined words; spelling • Nouns, verbs, etc. • Definition, meaning • 23 Sept 2004 FDA/Industry Statistics Workshop Company Confidential Copyright © 2000 Eli Lilly and Company 7

Syntax vs. Semantics Time flies like an arrow. Fruit flies like a banana. The dog eats red meat. The dog eats blue trees. Give the patient pain medication. Give the patient medication for pain. 23 Sept 2004 FDA/Industry Statistics Workshop Company Confidential Copyright © 2000 Eli Lilly and Company 8

CDISC Vision - Simply Stated A future conversation on data interchange: anyone: How do you want me to send the data? everyone: We are using CDISC version 4. 3. anyone: I can do that. 23 Sept 2004 FDA/Industry Statistics Workshop Company Confidential Copyright © 2000 Eli Lilly and Company 9

CDISC Mission CDISC is an open, multidisciplinary, non-profit organization committed to the development of industry standards to support the electronic acquisition, exchange, submission and archiving of clinical trials data and metadata for medical and biopharmaceutical product development. The mission of CDISC is to lead the development of global, vendor-neutral, platform independent standards to improve data quality and accelerate product development in our industry. 23 Sept 2004 FDA/Industry Statistics Workshop Company Confidential Copyright © 2000 Eli Lilly and Company 10

CDISC Mission Extending to all aspect of pharmaceutical drug development All clinical trial data • Protocol elements • Stability studies • Toxicology data • 23 Sept 2004 FDA/Industry Statistics Workshop Company Confidential Copyright © 2000 Eli Lilly and Company 11

CDISC in the “World of Standards” International Conference on Harmonization (ICH) EFPIA JPMA EMEA MHLW KIKO U. S. Dept. of Health and Human Services (HHS) Ph. RMA U. S. FDA Protocol Std ADa. M SDS ODM Med. DRA 23 Sept 2004 FDA/Industry Statistics Workshop Health Level 7 (HL 7) Reference Information Model RIM Clinical Document Architecture e. CTD = Organization LAB = Dictionary, Codelist NIH/NCI NLM DICOM TC: RCRIM CDISC CDC = Standard Company Confidential Copyright © 2000 Eli Lilly and Company = Model ISO LOINC SNOMED = Document Standard, or Architecture 12

CDISC/HL 7 Goal Research data at the point of clinical care. Clinical care data at the point of research. Every physician is a potential investigator. Every patient is a potential trial participant. 23 Sept 2004 FDA/Industry Statistics Workshop Company Confidential Copyright © 2000 Eli Lilly and Company 13

Challenge for Statistics A world of computable, semantic interoperability • Access to electronic medical records Analysis of data from controlled trials and naturalistic settings More data • More unstructured data • A few patients from many investigators 23 Sept 2004 FDA/Industry Statistics Workshop Company Confidential Copyright © 2000 Eli Lilly and Company 14

Opportunities for Statistics Adaptive randomization designs Integratable data across Studies • Compounds • Companies (e. g. FDA) • Rolling submissions 23 Sept 2004 FDA/Industry Statistics Workshop Company Confidential Copyright © 2000 Eli Lilly and Company 15

Summary Knowledge Information Automated Tools Integrated Data Standards 23 Sept 2004 FDA/Industry Statistics Workshop Company Confidential Copyright © 2000 Eli Lilly and Company 16