What is the Big Deal with CDISC and Data Standards?

Data standards have existed long before CDISC was ever established. The goal of data standards is to enable different users to access and work with the data without having to re-inventing the wheel. This reduces the amount of time spent in software development and training required since all users would only need to be trained on the same set of data standards. Prior to CDISC, there was a plethora of data standards throughout the industry. Each company had their own set of standards and even each department within an organization had their own variation. At that time, it appeared that there was no compelling business need or driving force to motivate establish a global standard since the data used represented the intellectual asset of each organization, and this was hidden and guarded for competitive advantage between companies. Some inter company standards existed in rare moments when two organizations used the same CRO to perform the same task and the CRO recommended the same set of standards for ease of interoperability. However, in most instances, standards were modified to fit each company’s needs so the end result was a mix of different standards. The approach of having multiple standards within each organization ultimately defeats the purpose and goes against the benefits of having a uniform standard.

One of the motivations behind CDISC is to provide the FDA a more efficient way to review data across sponsors. This came to light when safety issues arise from drugs that are already in the market yet there were many deaths due to safety issues. Since the data submitted to the FDA were not in a standard structure, there was no way for the FDA to easily perform analysis spanning across sponsors. At moments where thousands of patients are at risk of heart attack due to a drug that the FDA has already approve, it became essential that a timely analysis be performed across large sets of data in order to decide if a recall of the drug was necessary. Without data standards, it was already difficult to analyze data from different studies coming from the same sponsor company, let alone comparing drug across different sponsors. This can only be successfully done if the data from various sponsors are stored in a uniform standard data such as the one established by CDISC in the format of the Janis data warehouse. This will allow the FDA to make a ruling in the event that a safety issue arises for a particular drug. This will allow for a timely analysis to be performed across different drugs that may span different companies without having to do extensive data transformations. It will act as a unifying force across all companies to adhere to one set of standards. Within the set of standards, there are many CDISC data standard models including models such as ODM, LAB, SDTM and ADaM. Each model is intended to be used for different purposes. This chapter will focus on the implementation of SDTM since it is the format in which the FDA will require companies to submit their data in this format. The guidelines are made available for download at the CDISC.ORG website. Rather than reviewing the guidelines section by section, this chapter will use it as guidance in an implementation. The implementations will use examples to demonstrate the challenges and rewards that are gained from using the standards.

Why Implementation of CDISC?
Implementation of the CDISC data models is no longer a theoretical academic exercise but is now entering the real world. This chapter will walk you through the steps and share lessons learned from implementations of CDISC SDTM version 3.1. It will cover both technical challenges along with methodologies and processes. Some of the topics covered include:

• Project Definition, Plan and Management


• Data Standard Analysis and Review


• Data Transformation Specification and Definition


• Performing Data Transformation to Standards


• Review and Validation of Transformations and Standards Domain Documentation for DEFINE.XML

Regulatory requirements are going to include CDISC in the near future. It will therefore be mandated that the submissions be stored in this format. It is therefore wise and prudent to establish procedures on how you would apply CDISC data standard techniques and processes. This would prepare your organization so when the regulations take affect, you are not starting from scratch and therefore delay your electronic submission and ultimately the scheduled drug approval.

CDISC standards have been in development for many years. There have been structural changes to the recommended standards going forward from version 2 to 3. It is an evolving process but is beginning to be more stable and has reached a point of critical mass that organizations are recognizing the benefits of taking the proposed standard data model out of the theoretical and putting it into real life applications. The complexity of clinical data coupled with technologies involved can make implementation of a new standard challenging. This chapter will explore the pitfalls and present methodologies and technologies that would make the transformation of nonstandard data into CDISC efficient and accurate.

It is important to have a clear vision of the processes for the project before you start. This provides the ability to resource and plan for all the processes. This is an important step since the projects can push deadlines and break budgets due to the resource intensive nature of this effort. The organization and planning for this undertaking can become an essential first step towards an effective implementation.

more to come in book I am working on "Becoming a SAS Programmer in Pharmaceutical Industry"....