Many of you recently participated in the SCDM Professional Development and Learning Preferences Survey. You may have noticed that a section of this survey was directly focused on our SCDM Publications. Below are some of the results from this survey. We invite your comments!

SCDM Publications, Data Basics, Data Connections and the Blog
• All things to all people
• More technical, CDISC, CDASH, understanding EDC
• A day in the life of
• Consistent topic through out year
• Controversial issue with SCDMs position – Position papers done by editors
• Extremes, hands on & strategic
• Higher level, trends and directions
• More modern layout, more white space

When asked how you use Data Connections: General Interest, Reference and Industry Updates.

Top 10 topics you would like to see in Data Basics:
Best practices/tips/tricks
Electronic Data Capture
CDISC
EDC/technology/database topics
Data Quality
Standards
Future of CDM/CDM Roles
Business, metrics
Managing Lab Data
CDASH

Top 5 topics you would like discussed in the Blog
Best practices/tips/tricks
Electronic Data Capture
CDISC
Data Quality
EDC/technology/database topics

It’s a new year and the Society is growing by leaps and bounds! The many committees of SCDM are looking for volunteers. This is the perfect opportunity to become involved, meet other members of the society and network with your colleagues. Committee opportunities involve membership, marketing, certification, publications and education. Information is available on the website or by emailing info@scdm.org. Act now!

The Fall Conference committee kicked off planning for the 2008 Conference on November 2nd. Ralph Russo and Derek Perrin convened with a handful of session chairs to brainstorm next year’s topics and recommendations for more session chairs. The team is excited to include a third management track at next year’s meeting in addition to hosting the second annual Data Driven Innovation Award. The SCDM administrative office reported that sponsorship for next year’s event is off to a great start with many slots sold already!

Another excellent SCDM Fall Conference draws to a close, and data managers and their friends and colleagues make their way back to their day-to-day lives, enriched perhaps by the thoughts and ideas they have heard over the last few days. 

Eight lucky folks who stayed bravely to the end of the last session were rewarded with raffle gifts generously donated by some of our conference sponsors, partners and exhibitors.  In no particular order they were:

·         Deborah Cole, from Quintiles, won a polo shirt donated by CDISC

·         Levent Ozcakir, from Allergan, won a 512MB USB drive also donated by CDISC

·         Yanalthe Ramirez, from Kendle, won a lovely pewter Scottish drinking bowl donated by Formedix (it doubles as a great nut bowl!)

·         Kirti Pandit, from Bausch & Lomb, won CDISC t-shirt

·         Nancy Cacciola, from BioStat International, won a $50 Visa gift card donated by Placemart Personnel Service

·         Demetris Zambas, from Schering-Plough, won a box of Fannie May chocolate truffles donated by The Cambridge Group (his colleagues will be mighty happy about this!)

·         Jameelah Aziz, from Cubist Pharmaceuticals, won a CDISC fleece jacket

·         Sally McCartney, from the Cleveland Clinic, won an iPod donated by PharmaVoice (she’ll be the one marching to her own tune)

Congratulations to the winners!  Many thanks to the donors!  And most especially thanks to all the hardworking folks from the SCDM administrative office, the SCDM Board of Trustees, the conference chairs, session chairs and presenters, and particularly thanks to all the attendees, without whose presence and participation the conference would not be.

See you in 2008, and, until then, may your data be clean, your studies positive, and your submissions approved!

The keynote address on Monday morning at the Fall Conference was delivered by Hugo Stephanson, of Strategic Research and Safety at Quintiles.  He provided very interesting insight into several future clinical research trends that will affect CDM.  Of these, I’d like to examine one in particular. 

Hugo pointed out that the nature of the molecules and interventions we are studying is changing in a fundamental way.  Traditionally, our test articles have been small molecules that were given to subjects, had some effect, were metabolized and excreted, and the effect ended with the elimination of the compound.  Today, more and more treatments have the potential to have biological activity long after they are stopped, and indeed many are deliberately designed to have a permanent effect.  We are already seeing the need to gather safety data for much longer than before (witness Avandia, Vioxx, and several others), and this need will intensify as we shift our research focus to these new compounds.  As Hugo pointed out, it will soon (perhaps already has?) become unacceptable to greet the request for 5-year safety data by starting a new study when we have subjects who were enrolled in trials five or more years ago.  Instead we will have to continue to follow up on all the subjects in our trials perhaps indefinitely. 

So what does this mean for us?  Beyond the obvious fact that we’ll never again be able to close a study?!  We’ll have to start treating each study analysis time point as an interim analysis, where we take a snapshot for the purpose of the analysis (perhaps an efficacy study report, or a traditional interim analysis, or a safety update) and have to manage the consistency issues between these snapshots.  Freezing and locking databases will come to have different meanings, as we’ll need to be able to add data on a perpetual basis.  Archiving and documentation will have to look different, and standardization will become paramount. 

We’ll have to make decisions about what safety data to collect after the efficacy analysis is complete.  Should it vary depending upon whether the NDA has been approved?  Should it continue to reside in the traditional clinical database or should a subset be ported into a long-term safety database?  What relationship should the clinical database and the safety database have with the ISS and ISE datasets and analyses?  To what extent should the various snapshots or copies be reconciled as additional information becomes available about earlier events?  The current (largely arbitrary and artificial) divisions between the phases and especially between pre- and post-submission will disappear, which will mean that post-marketing departments in companies will have to partner to a much greater degree than usually happens today.  Today safety databases contain primarily serious AEs, but more and more of the safety problems being identified today are “significant” AEs, which are non-serious events that occur in somewhat higher proportions in given subpopulations, or occur extremely infrequently and so are unlikely to be seen during the pre-submission studies.   This means that we may need to focus on different types of information for each event. 

The biggest challenge is likely to be keeping tabs on subjects after they finish the efficacy portion of the trial, as the investigators may well not be their primary care physicians, and so will not have a big incentive to continue to follow the subjects.  In fact, anyone who has tried to get a query resolved even a couple of months after a study completes knows what a nuisance that can be!!!  This suggests that we may need a whole new way of approaching this endeavor.  Indeed it turns out that if the subject is empowered to provide the updates, they see it as being in their own self-interest to do so and are much more compliant.   

In alignment with this, Quintiles has sponsored the development of a service called iGuard.org that is designed to follow subjects indefinitely.  In theory, as Hugo suggests, safety data should be fairly generic, and it shouldn’t be difficult to make it largely non-drug-specific.  In fact, this could be a very good opportunity for independent providers to step in and do this, as it is inefficient for each pharma company to have to do this independently.  In addition, in time, it is likely that reporting spontaneous AEs will move from being voluntary to being required, as it already is in Europe.  Given this, a centralized approach will make much more sense. 

Clearly the world of the data manager is going to change dramatically over the next several years.  As participants in the healthcare world we have a responsibility to look ahead and participate in charting the course to greater patient safety and more effective medicines delivered more quickly, cost-effectively and efficiently.  It is a worthy cause.

Congratulations to DSG for their first place award in the first annual SCDM-sponsored Data Driven Innovation Award competition.  Their solution was developed in partnership with Sirion Pharmaceuticals, and involved using eCaseLink / eSource to provide cost-effective randomization services. 

Congrats also to Nextrials, who were the runner-up winners for their collaboration with Z-TEch using Prism to develop an interface to capture device data directly into the study database. 

Your CCDM designation is valid for 3 years, and in order to maintain it you must acquire CEUs to show that you are continuing to grow professionally.  If you are a CCDM attending the Fall Conference, be sure to pick up the forms to submit for CEU credits.  You can find them at the Registration Desk, and at the SCDM Booth in the Exhibitors’ Hall.

In the Fall Conference Leadership Forum on Sunday there were spirited discussion around the past, present and future of data standards.  One particularly hot area, ably examined by Landon Bain and Clem McDonald, concerned the integration of Electronic Health Records (EHR) systems and clinical research data.  EHR systems hold the medical records that follow each person around as they interact with healthcare providers, so you and I and all of us have records in at least one (and often many) EHR systems.  The clinical research data capture happens via EDC as part of pharma-sponsored clinical trials, but also can result from academic research projects. 

There is a lot of debate about the degree to which clinical trials data can be extracted from EHR systems and what that will mean for the structure and conduct of the trials.  One extreme (which is pretty much where we are now) is that data for each purpose is collected completely separately, and multiple times if necessary.  The other extreme envisions a world where never again will we have to have data capture separate from the EHR, and all information can be mined from the patients’ records. 

Reality will probably be somewhere in the middle.  Data points, such as demography, can probably be extracted as they exist, and most epidemiological and observational research will be able to operate this way.  Research that relies upon significant clinical events, for example cardiac arrest, or death, will probably also be able to pull much of its information from EHRs.  On the other hand, assessments such as surveys or specialty lab tests, would never normally be conducted, and studies could never assume that specific information about study treatments would be appropriately recorded in patient charts without direct prompting.  Some data will therefore require special processes to ensure their correct definition and capture. 

There are, of course, many issues to consider.  For example, patient confidentiality must be addressed.  Additionally, while there are some organizations that have experimented with EHR/EDC combinations, this technology is in its infancy, and there are currently no generally available solutions for healthcare providers and hospitals to acquire.  Many will be hesitant to develop their own, and indeed, they shouldn’t develop their own because it will simply add to the proliferation of systems that will have to be integrated.  Another consideration is that the regulatory requirements for audit trails and for system development and implementation validation are more stringent for agency-bound data, and it is likely that most hospitals will not want to make their EHR systems 21CFR11-compliant.  

The important thing to recognize is that the point of the integration is not, for the most part, to eliminate the need for study-specific data generation and capture.  The point is to realize that if we can eliminate the need for capturing the same data in more than one place, or for having procedures repeated for the sake of entering information in more than one system, we have gone a significant way towards minimizing the cost, time and patient inconvenience inherent in running clinical trials.  Even better, if we can insert the trials data collection process seamlessly into the normal flow of patient care and information gathering, we have made the coordinator’s life easier and very probably increased the quality of the data, and this can’t be a bad thing.

In browsing the topics at this year’s Fall Conference, I noticed the session on Data Standards.  This is a particular passion of mine, and the three talks look very thought-provoking.  In the second one, Ms. Trollinger describes the challenges and conflicts between the requirements for collecting and storing data, vs. those for sharing data, especially across therapeutic categories.  While there are many similarities in the data needed for different therapy areas, there are also many differences in both content and in structure.  This makes creating a single data dictionary something of a challenge. 

Tackling this challenge would be a tall order in a company, but there everyone (at least in theory) reports into the same organization, and, provided management has the intestinal fortitude to do so, common approaches can be mandated.  The project outlined in the talk had an even greater challenge, in that it was sponsored by one of the National Institutes of Health, who has a significantly smaller degree of control over the structure of the studies it sponsors and thus must provide carrots rather than sticks (i.e., give investigators incentives to cooperate, rather than force them to do so). 

This project is an example of an increasingly popular trend towards sharing data in the academic and government research environments, and to a lesser extent in industry (the latter being driven mostly by the need to create consistently structured data for regulatory submissions).  CDISC is the most familiar one to those of us in industry, and it is aimed primarily at regulatory agencies.  Outside of industry, I know of HL-7, which is (among other things) a set of messaging standards for defining and exchanging data between healthcare providers.  There is also the National Cancer Institute with its caBIG project, which is a set of data dictionaries and processes for standardizing the definition, collection and storage of data from oncology trials sponsored by NCI.  There may well be other projects.

It is wonderful to see such awareness of and enthusiasm for standardization.  This can only help to increase efficiency, reduce costs, and enhance the safety and efficacy of medical treatments.  On the other hand, it is a concern that, although these organizations do have some awareness of each others’ work, they are still working mostly independently.  Not only is there some duplication of effort, but they will inevitably make different decisions and produce slightly different results.  Anyone involved in such a project understandably has a tremendous sense of ownership and will be committed to his or her own decisions.  At some point the various schemas will have to be harmonized, and this will require a great deal of work both to achieve the harmonization and then to apply the results back to the source systems.  Because of this, it is critical that the organizations already developing standards be open, transparent and vocal about their work so that any other organizations contemplating initiating a standards project will join an existing one and work toward ensuring their needs are met there.  By the same token, it is encumbent upon the existing projects to be open to the needs of others so that the number of projects is limited and eventual integration is not even more daunting. 

What sounds like an exaggeration might become reality. Safety considerations are writing speeding tickets for fast-track drug trials and flashing the green light for larger and lengthier ones.

In a Boston Globe article published last Monday, Diedtra Henderson quoted Dr. Clifford J. Rosen, director of the Maine Center for Osteoporosis Research and Education, requesting that “federal regulators should insist upon solid proof that diabetes remedies enhance long-term quality of life by reducing risks for such serious side effects as heart attacks, eye and kidney ailments, and amputations.” Dr. Rosen contributed his opinion to a recent New England Journal of Medicine commentary piece.

Henderson concludes: “For pharmaceutical manufacturers, changing the status quo could result in a dramatic increase in the number of patients they would need to enroll in clinical trials that would become longer and more expensive.”

Dr. Rosen’s clinical trial recommendations would have called for double the number of subjects randomized and at least quadruple the length of the longest trial for GlaxoSmithKline’s Type 2 Diabetes drug Avandia.

Avandia has received the latest attention, but it wasn’t long ago that the cardiac complications associated with use of pain relievers Vioxx and Celebrex were making news. Increased study enrollment and study length would not be limited to evaluating Type 2 Diabetes drugs; they could potentially involve any drug trial.

My role as a data manager does not invite the question whether paving the road for larger and lengthier clinical trials will pinpoint health risks that would otherwise be absent from smaller, shorter studies. But I can and will predict the following trends if Dr. Rosen’s guidelines are followed:

• Increased number of subjects screened and randomized.
• Lengthier enrollment periods.
• Greater scrutiny of medical history, including medication history.
• More substudies attached to the longer trials.
• Urgency of SAE reporting and reconciliation.
• Certainly lengthier studies, but not necessarily more visits or more CRF pages.
• Increased use of subject questionnaires and telephone interviews.
• Interim database locks more common.
• As trials become more expensive to maintain, expect closer accounting of incidental expenses (travel, software upgrades, perks, overtime pay, etc.)

None of these possible implications for data management makes me jump for joy. But, let’s consider it job security (for us) and safer healthcare (for all.)

Reference:

http://www.boston.com/business/healthcare/articles/2007/09/10/burden_of_proof/