Rob Woolson, Chief Strategist-Biostatistics and Data Standards for Regulatory Submissions,  has led SDTM/ADaM dataset conversion projects in multiple therapeutic areas. He has held a leadership role in six CDISC-compliant regulatory submissions, having guided the creation of ISS/ISE statistical analysis plans; integrated analysis dataset design and production; integrated display design and production; and submission-related documentation development.

In December 2014 the Food and Drug Administration released two final guidance documents and one technical conformance guide related to the content and format of electronic submissions.  These documents require the electronic submission of standardized non-clinical and clinical study data for nearly all submission to CDER and CBER, including INDs, NDAs, ANDAs, and BLAs by 2016.  In other words, study data must comply with a number of standards enumerated by the FDA, most notably CDISC data standards (for example, SDTM and ADaM).  These guidance documents are binding.  Submissions that do not follow technical data conformance standards will not be filed or received by the FDA, meaning if you don’t comply you could receive a refuse to file letter.  So, what do you need to do?  Ask yourself these five questions.

How much of my data currently conforms to these standards?

It may be the case that you’ve been working with CDISC standards from the start of development.  All of your clinical datasets are formatted to SDTM and ADaM formats and you plan to create your Integrated Summary of Safety (ISS) and Integrated Summary of Efficacy (ISE) datasets to conform as well.  It may also be the case that you know that none of your existing data conforms and that extensive remediation will be necessary.  In our experience, however, many situations are more complex.  For example,  some late stage development work may conform to CDISC standards with much of the earlier development work following a legacy standard (or no standard at all).  The first step is to figure out what you have.

For the datasets that don’t conform, do they all need to be converted to CDISC standards?

It is likely that if you are preparing for a submission now, at least some of your datasets will need to conform to CDISC standards.  However, that doesn’t necessarily mean that all of your legacy datasets will need to be converted.  By preparing a Data Standards Plan and submitting it to the FDA, you may be able to limit the amount of conversion and remediation work that is done, saving both time and money.  Getting feedback from the FDA at this stage is critical, so that there won’t be any surprises when you file.

Do I have the necessary expertise and capacity in house to make these assessments, perform needed remediation, and get the data submitted?

Moving quickly at this stage in development is critical.  Everyone wants to get their product to market as soon as possible, so once the final studies are finished you want the data submission work to be fast and accurate.  If you don’t have the internal capacity to move at the rate you want it may make sense to bring in an outside vendor.  Data submissions are also a pretty specialized area, and many companies may not have the internal expertise required.  This can be another reason you should consider an outside vendor.

If I need a vendor to help with these activities, what should I be looking for?

It’s likely that you’ve used one or more CROs at some stage of development, but the CRO that performed your clinical studies may or may not be the best choice when it comes to getting your data ready for submission.  Here are some things to consider:

• Experience
  • How many protocols have they mapped to SDTM or ADaM? Were any of them for products in the same or similar indication as yours?
  • How many marketing applications have they filed in the last five years?
  • What is their technical acceptance rate? Have any of their submissions resulted in a refusal to file?
  • What level of experience do the individuals assigned to your project have? You should expect a greater level of experience for statisticians and data standards staff working on a marketing application submission than you would need for working on a typical single study.
  • Does the vendor have experience communicating directly with the FDA on data standards plans?
  • Which standards do they have experience with? At a minimum, we recommend:
    • CDASH
    • SDTM
    • ADaM
    • define-XML
    • CDISC controlled terminology
  • Which version of the standards have they used?  Are they familiar with the versions you are using and plan to use?
• Tools and Technology
  • What tools are they using to assess compliance? Are they the same as those in use at FDA?
  • Do they have tools for metadata mapping, creating define.xml files, and performing QC on clinical databases?

What Types of Services Can I Expect a Vendor to Provide?

A high quality vendor should be able to look at your situation and make recommendations about the services you need.  Typical services include:

• Assessing and reporting on the current state of clinical databases relative to external and FDA standards and requirements. If you don’t already have one, they should be able to provide a plan for preparing databases for submission.
• Performing any needed data remediation.
• Preparing and submitting a Data Standards Plan to the FDA.
• Executing a test data transfer to obtain FDA reviewer feedback.

Be wary of vendors that provide a set package of services.  Each submission is unique, and these packages may include costly and time consuming services you don’t really need.