IDBS Data Management Blog

Amongst the companies providing fodder for booth trophy hunters, national dancers, alphorns and buckets of beer, what was one of the big buzzes at BIO 2012? Bioprocessing. Last year this area was just a zone, but it has fermented into a healthy biomass of hundreds of large molecule therapeutics, antibody and vaccine CMOs. Not just the Lonza, DSM and Xcellerons but many, many more providers of bespoke biologics manufacturing were there. This was no Pittcon but the amount of stainless steel and bio manufacturing expertize was the largest ever.

So, how come this growth-based industry is growing so fast? The FDA has fed the growth by its Quality by Design (QbD) requirements, backed up by some recent high profile fines for major pharma batch failures. Coupled with the rise in large molecule therapeutics, vaccines and biosimilars the momentum is unstoppable.

But there is still a big problem brewing in fermenters around the world….data overload. Bioprocess development and QA produce mountains of process and scientific data. This data is vital in building recipes that provide consistency of production and to compare today’s yield against historical data, to take decisions and share them with others. Amazingly many of these processes are still paper driven and laborious, so the impact for a low margin CMO, seeking to attract business and keep the FDA happy,  is clear. Go digital and get efficient. This is not a transactional ‘BI-like’ aggregation of dispersed data but an integrated, managed data approach.

When developing a process and undertaking experimental design, each step needs to be captured, compared to historical data, and integrated with other data to secure IP and provide process insight.  Into this data-rich environment step GMP-compliant process ELNs such as IDBS’ E-WorkBook which is both a research ELN and Process Execution System, able to cross the boundary of R&D. Coupled with Design of Experiments systems such as uMetrics, and internal historian systems, these platforms are able to capture, compute, compare and secure process data; then integrate upstream to Manufacturing Execution Systems (MES) and Enterprise Requirements Planning Systems such as SAP.

The bioprocess development process has been very document-driven and this is hugely inefficient. Documents should be generated just in time rather than generating data from multiple documents. This is where data management takes over from legacy document/paper management. Fast-moving Bioprocess organizations’ process engineers  now inhabit a flexible, cheaper data-driven world with documents generated on demand, rather than living in a document heavy environment.

There was plenty more at BIO 2012 but for sheer biomass at BIO in Mass the Bioprocess team wins. With improved data management and associated process insight these companies will expand further to become major league players providing cheaper, better therapeutics in the decades to come.

In July 2011 we worked with Scientific Computing, R&D, and Drug Discovery & Development magazines to survey almost 700 R&D professionals.

The survey aimed to uncover the top challenges R&D centric organizations face in the quest for R&D collaboration, IP protection, and effective data management. It also explored the key initiatives planned to support R&D Collaboration and Data Management in the coming year.

EXPLORING THE SURVEY FINDINGS

88% of R&D organizations lack adequate systems or practices to automatically collect data for reporting, analysis and decision making. An essential part of collaboration is how effectively individuals and organizations share data. The survey found that today’s researchers still rely on manual processes and non-scientific applications, such as Microsoft Office applications, for report production. The task is particularly difficult in the Defense & Aerospace, Manufacturing & Engineering, Energy & Utility, and Academia and Government sectors.

Comment: The R&D world is moving to a data centric environment rather than a document centric environment. Report writing is on average taking 25% of researchers’ time. It’s not about reporting it is about developing insights and understanding from other’s data, enabling everyone to be more productive.

As Jay Galeota, SVP Strategy & Business Development, Global Human Health, Merck explains in a recent Ernst & Young report*: “The most important thing is what you can actually do with the data. It’s one thing to have interesting information, but it’s the insights that are important to guide smarter, better decisions…”

6 out of 10 respondents relied on manual compiling and searching of data. The survey explored how researchers shared data and found that 60% are unable to compile relevant R&D data without manually searching through documents and reports. This results in a static document-centric view of the data.

Comment: This results in a document rather than data centric approach. Each respondent is both a consumer and generator of data – the collaboration challenge is as much about internal person-to-person collaboration, as external business-to-business sharing of data. It also results in significant  time lost/taken away from actual research because people are busy searching through documents for info. For streamlined research, people need to put hands on relevant data immediately, not spend time recreating it and certainly not spend hours looking for it.

57% R&D organizations are relying on in-house systems to manage R&D data. Survey respondents across all sectors are predominantly using legacy in-house, home grown/built, solutions.

Comment: We know that change occurs in R&D organizations on a 12 month basis, but it is fair to say that few in-house systems can effectively evolve or be upgraded in synch with this timeframe. This causes an ongoing burden to businesses and their IT support systems. In talking with industry insiders we hear that this is leaving an unsustainable graveyard of systems. Where software development in-house does have a role is in configuring best in class systems to confer a competitive advantage. Does any R&D organization have building software as part of their vision?

As Chris Thoen, MD Global Innovation Office, Proctor & Gamble commented*: “Only do what only you can do.”

Less than 25% of responding organizations have deployed foundation technologies such as ELNs, LIMS, KM, and PLM systems. This leads to a R&D data ecosystem – data flowing from idea to product – that is fragmented and siloed. The survey reflects the proliferation of third party process applications.

Comment: Many organizations are using a combination of multiple data and knowledge management applications across their business. This makes management of IP/data from early research through to commercialization challenging. And while point solutions have been implemented to solve particular issues, there is a resulting duplication within the IT landscape (e.g. data analysis and reporting tools). In many organizations, multiple similar category products are used by different parts of the enterprise.The result is that there is no end to end solution for data management that integrates into existing data silos

Discover more at IDBS seminars “R&D Informatics: Strategies for 2012” during November in Paris, Frankfurt & Boston, or attend our webinar on November 17.

Watch out for the next blog post in this series – I will explore what the R&D Collaboration survey tells us about IT complexity, collaboration & intellectual property protection.

* Source: Progressions: Building Pharma 3.0 – Global Pharmaceutical; Industry Report 2011 from Ernst & Young.

The switch from paper-based notebooks to Electronic Lab Notebooks has proved as profound as the move from 78rpm records to iTunes. To stay competitive in this increasingly data driven world, investing in ELNs is becoming more and more crucial.

Nowhere is that more true than in preclinical drug development.

In just five years the use of ELNs throughout late stage research, development and manufacturing, (dare I say, spearheaded by IDBS’ E-WorkBook), has led them to become enterprise systems offering everyone an easily mined and searchable rich data source.

Today’s advanced Enterprise ELNs integrate with existing LIMS to offer highly sophisticated and flexible data capture. Combining these systems removes bottlenecks typically found between disparate data sources and provides much faster access to raw scientific data and its context.

The business benefits are evident and proven – more effective collaboration, fast reporting and easier compliance with initiatives such as ‘Quality by Design’. That’s what I call a significant commercial advantage.

If ever there was a time to re-evaluate whether an ELN is the right system for you, it’s now.