Clinical Document “Redaction” is a hot topic in our industry today, and I was able to see this first hand at the annual DIA meeting in Philadelphia. DIA (Drug Information Association) is the largest multidisciplinary forum for knowledge exchange that fosters innovation to raise the level of health and well-being worldwide. Being a global organization, DIA is dedicated to bringing health care product development professionals together in a learning environment to improve health and well-being throughout the world. As a Regulatory Operations Specialist at MMS who has spent a lot of time recently working on clinical document redactions, I thoroughly enjoyed the opportunity to attend the DIA 2016 Annual Meeting, June 26th – 30th in Philadelphia to share some of our learnings in this important area.
I presented a poster on “Comparison of Manual vs. Automated Redaction Techniques for Clinical Submission Documents”. The poster session was scheduled for Wednesday, June 29, along with a short podium presentation on the same. Redaction is the simplest form of anonymization, whereby individuals with a scientific background read through clinical documents and mask all sensitive information like CCI (Company Confidential Information) and PPD (Protected Personal Data). Anonymization in the form of redactions is driven by the new EMA Policy 0070 that mandates the publication of anonymized clinical submission documents on a public portal, following conclusion of the regulatory decision making process on the MAA (Marketing Authorization Application). Looking at the amount of time and effort spent on manually redacting a few initial dossiers, we realized an underlying need to automate part of the process to ease the burden on scientific reviewers and at the same time improve the efficiency and quality of redacted documents. My poster detailed a comparative analysis of efficiency and quality parameters like processing time, number of redactions, false hits, corrections, and others between the manual and automated redaction methods. See below for the link to the poster abstract.
The first thing I noticed at the conference was a dearth of sessions and presentations on this new EMA Policy. The lone “Engage and Exchange” session on this topic on Monday, July 26, saw over 100 participants trying to find a spot in an already overcrowded space that was made to fit an audience of 25-30 people. The overwhelming response to this session could be attributed to the complete lack of information and sponsor experience with this Policy, driving people to gain as much information on this elusive topic as possible. I had a chance to network with a number of sponsor companies at this session, and that is when I realized almost everyone had already made a point to attend the redaction poster presentation on Wednesday. When they realized I was presenting that poster, I received a number of questions from companies trying to navigate their way through their anonymized submissions. “I am the only one at our company trying to make sense of this policy and it would be great to have some of your expertise and guidance! I just feel like I am throwing arrows in the dark,” said an extremely dejected sponsor representative. I also heard a lot of comments on the amount of time and resources some companies were spending on this daunting task and how interested they were in exploring options that could help them reduce the time it took to obtain quality results.
The amount of interest everyone seemed to have in the poster was exhilarating and intimidating at the same time! I was prepared for a lot of traffic during the hours allotted for my poster presentation, but the response surpassed all my expectations. People stopped by in the morning to enquire about the podium presentation as they wanted to make sure they could attend. As expected, a number of people showed up for my 10 minute podium presentation and I could see that everyone was really engaged and were taking notes throughout. At the end, I received lot of questions on the automated tool’s capabilities, limitations, and pricing. A few members of the audience seemed interested in knowing which anonymization criteria the tool handled better - CCI or PPD, and if it was possible to obtain complete automation at this stage. Some even enquired if the tool could automatically create justification tables, which just goes to show how the sponsors are looking at finding alternate ways of managing the multitude of steps involved in the anonymization process under Policy 0070, and how they are interested in reducing the tedium of it all. Everyone seemed to be amazed by the drastic improvement in efficiency we obtained by incorporating the automated tool in our redaction process - “I can’t believe your team only took 4 hours to redact a 900 page report with the use of partial automation! Even the 12 hours you spent doing the whole thing manually seems much less as opposed to the 25+ hours I have spent on reports of similar length without being completely sure of what I was doing!” More people stopped by the poster after the podium presentation and towards the end I ran out of the printed copies of my poster for distribution! I also met a few active members of the DIA community on disclosures and was invited to join and share my experiences on this topic.
Apart from my poster, I did get a chance to attend some of the highly informative sessions that were offered at DIA; a DIAmond session on ICMRA and a few innovation theater sessions on Transparency and eTMF. DIA is a great opportunity to pick up some insights on important new regulations and apply the knowledge to the important work we do every day. It also offers a great chance to learn about other emerging hot topics in our industry. I was excited to share the MMS experience in clinical redactions, and I look forward to continuing the dialogue within the industry on best practices in this evolving area.