For any business, in any industry, it’s not how much data you have but what you’re able to do with it that determines its value. And few organizations have as much data to manage as life sciences companies, which means they have the most to gain by honing their data discipline.
To cope with increasingly demanding and complex regulatory requirements, major players in the industry have now accepted the need for a more centralized, consolidated approach to data management. The next evolution of this strategy for master data management – MDM 2.0 – involves using amalgamated product information to transform the way firms create essential everyday documentation. Process automation rates of 90%+ can lead to at least a 10-fold acceleration in the preparation of regulatory submissions as well as other routine content that is critical to get right - such as labelling and patient information.
Life sciences organizations are already acutely aware of the need to catch up with peers in other industries in the way they handle and exploit information. The burden of regulation and red tape, so necessary to protect patients, can slow companies right down - to the point that they can hardly imagine a time when they’ll achieve the level of agility, responsiveness and innovation they need to thrive in the digital age.
It is in this light that technology vendors have treated the latest regulatory requirements - in particular forthcoming ISO IDMP standards and new FDA Standardization of PQ/CMC data elements and terminologies guidance – as a reason to overhaul the way companies manage product information. The latest authority mandates and recommendations are so extensive and so demanding that life sciences organizations are left with no choice but to review the way they currently do things. Our recommendations, delivered through a range of comment articles over recent years, has been to move towards a centralized, ‘master data’ approach to product information.
But companies should not stop there.
From Data Hygiene to Process Transformation: The Case for MDM 2.0
Harnessing new industry standards as the basis for reorganizing product data offers companies a number of advantages already. It will make compliance much easier, and offers companies new visibility and control over everything that happens to key operational data, ensuring its quality and accuracy. Yet, once they have achieved this new definitive ‘master data’ position (an agreed, single version of product truth which informs numerous use cases), companies also have a chance to exploit this to make their operations more agile, cost-efficient and creative.
Next-generation master data management – or MDM 2.0 - involves taking companies’ investment in consolidated, centrallymanaged data, and turning it into tangible value – through automation of processes that currently take an inordinate amount of time, hampering firms’ competitiveness and ability to move quickly.
In PwC’s most recent global survey into CEO’s strategic priorities, life sciences CEOs emerged as more likely than any other industry sector CEO to count innovation as their top priority when it came to building or strengthening their organization’s capability. Over a third (35 per cent) of pharma, medical device and biotech CEOs said this was their primary preoccupation, compared with a crossindustry average of 23 per cent.1 Interestingly, the second area in which life sciences’ CEO priorities were noticeably higher than CEOs in other industries was in relation to human capital. Here, 19 per cent of CEOs expressed an intention to strengthen their capability to take advantage of emerging opportunities, compared to a cross-industry CEO average of 15 per cent. So, not only do life sciences organizations aspire to be more innovative in how they do things; they also need to be more clever and strategic in how they make use of their talent. MDM 2.0 offers a way forward in relation to both of these priorities.
The impact of adopting an MDM 2.0-based approach to data exploitation is potentially very significant, even – or especially – in relation to quite mundane, routine activities. One of the most laborious yet critical of these is regulatory submissions preparation and processing.
Currently, companies create regulatory submission documents, fill in forms, generate labels, packaging and patient information more or less from scratch each time there is a new requirement. This involves having to call up different systems, and look through various tables and spreadsheets to find data to manually copy and paste into the new output.
This is a hugely labor-intensive process which is fraught with risk of getting some detail wrong, using out of date information, or failing to conform to a market’s particular requirements. But with easy, confident access to the correct content components, organizations could be populating new documents at the touch of a button – automating at least 90 per cent of the process of content generation, so that all that remains is for someone to add any finishing touches and check everything over.
Eliminating Redundancy – and Risk
Automated content creation relies on two things: good, definitive master data; and the ability to pull in and mix and match approved data components according to the given context.
If existing content exists primarily in monolithic form, in previous documents for instance, it is of little value for future use – unless someone checks and re-enters the information each time. If the latest version of that content exists in more granular form, in a central data bank – as a series of searchable and easily extractable content assets - not only is it easy to repurpose again and again, but this core content only has to be updated or amended once, in one place. Those edits can then be applied across all new use cases, with a few simple clicks. Crucially, everything can be viewed and monitored in one place, too.
This is the kind of process that happens as standard in other markets where there is a lot of live content to keep track of across sprawling operations. And, at last, proof-of-concept projects are beginning to take shape in life sciences. Here, companies are starting to create templates for common document creation, based on master data. In this kind of ‘structured authoring’ scenario, output is generated with minimal effort. Once the context has been indicated (the product, the type of content needed), the correct data assets can be automatically pulled together to form the target content. In the case of a standard application form, where no customized tweaking is required, 100 per cent of the document compilation could be automated, accurately matched to the given market and target language.
Speeding Up Content Production - from Weeks to Days
The payback is still being calculated through these early trials, but the expectation is that the time savings will be at least 10-fold: so where new content preparation has previously taken 50 days, it will now take just five. These are phenomenal efficiency gains, offering to significantly accelerate companies’ speed to market while freeing up experts to focus their time more on higher purposes.
Assuming the chosen content management system is able to take care of document creation and approved local translations simultaneously, there should be no need to create each local version of documents separately. Structured content templates will be able to pull in the correct, pre-verified text fragments in each language, meaning there is no need to re-translate content each time. That’s because approved translations of existing wording and text extracts already exist in the master database.
A Smarter Approach
For the majority of life sciences organizations that still rely on very manual, decentralized processes for putting together product-related content, the transformation presented by master data management (MDM) and its next-generation manifestation, MDM 2.0, is huge. On top of the time and efficiency gains, it offers companies much greater confidence and oversight of the content being put out across global operations – minimizing the risk of product recalls resulting from inaccurate or incomplete information being submitted, or the wrong phrasing being used.
The vision for MDM 2.0 isn’t confined to the structured authoring of content, either. It’s about boosting what companies can do with data to improve their operations and business impact.
While initial projects might focus on internal operational data about their own products and processes, there is great scope to enhance this with external intelligence – for instance, data about market conditions, or evolving regulatory requirements in different regions and countries. The more complete and rounded the data that is input into central systems, the easier it becomes to plan for and manage new requirements – and improve success rates.
Improving Hit Rates: Machine Learning
There is much to be excited about, particularly as artificial intelligence and machine learning enter the picture, helping systems to ‘learn’ how to produce better output, or the conditions most likely to result in a new marketing submission being accepted first time.2 IDC recently predicted that, by 2021, 40 per cent of life science organizations will have achieved 15–20 per cent productivity gains through the adoption of cognitive/AI technology.3
KPMG discussed the scope for accelerating regulatory compliance using AI in its 2016 market report Intelligent Augmentation,4 which explores the potential parameters for ‘digital labor’ in life sciences. The paper assesses AI’s role in simplifying the submissions process, through the use of technologies that have compliance built in - with all actions traceable and auditable. Projecting forward a few years, life sciences organizations will be able to use cognitive technologies to keep pace with ever-changing regulations, it predicts.
As companies move towards automatically-generated documents, machine learning could learn to recognize and adapt to common edits that users are making to complete or finesse given document output. Instead of admin staff conducting periodic reviews and restructuring the templates accordingly, an AI-enabled system would anticipate and propose improvements based on frequent changes that users have had to make. And, when building a submission, the system might suggest which documents to include; which contributors to involve in the authoring/review/approval process; how to set up the timelines - perhaps even anticipating questions that are likely to come from the authorities based on points raised previously for related or similar submissions. The scope is probably much bigger than we’re even able to imagine at this early stage.
Let Technology Take the Strain
In the meantime, the aim should be to automate all of the routine activities that take away time users could be allocating to other, more demanding tasks.
The enabler for this is the creation of a comprehensive master data model – one that also includes active relationships and dependencies between the data, in a way that can drive new efficiencies and increased impact through proactive process automation.
The vision companies must work towards, and which is encapsulated by MDM 2.0, is one in which teams will simply tell a system what type of documents they need, for which product, and for what purpose (country/region, type of submission, and so on), leaving the technology to do the rest. That could be generating new documents from the master data and appropriate structured templates, or directing users to existing documents and even proposing updates, corrections, improvements based on previous use of the system or newly entered data (for example about the latest regulatory requirements).
Until now these easy-to-appreciate benefits have eluded life sciences companies, because of the not insignificant groundwork involved in getting to this point. As the industry starts its IDMP preparation in earnest, as it must over the year ahead, the need to return to the drawing board may present new ideas and opportunities, opening the door to MDM 2.0 and all that this represents.
References
- Key findings from the pharmaceuticals & life sciences industry: 20th CEO Survey, PwC, 2017: https://www.pwc.com/gx/en/ceo-agenda/ceosurvey/2017/gx/industries/pharmaceuticalsand-life-sciences.html
- How Artificial Intelligence is impacting life sciences, AMPLEXOR blog, September 2017: https://blog.amplexor.com/lifesciences/en/how-artificial-intelligence-is-impacting-lifesciences
- IDC FutureScape: Worldwide Health Industry 2018 Predictions, IDC, October 2017: https:// www.idc.com/research/viewtoc.jsp?containerId=US41114417
- Intelligent Augmentation: Life sciences companies are a natural fit for digital labor, from robotics to cognitive, KPMG, December 2016: https://assets.kpmg.com/content/dam/ kpmg/pl/pdf/2016/12/pl-intelligent-augmentation.pdf
Author Biography
Siniša Belina is senior life sciences consultant at AMPLEXOR Life Sciences. He started his professional career at Pliva (now a member of the TEVA Group), where in addition to his responsibilities in manufacturing, he also engaged in successful EDMS implementation projects. Belina later joined KRKA’s Regulatory Affairs Department, and finally moved to AMPLEXOR. He applies his detailed knowledge of pharmaceutical documentation and processes to areas of business process analysis and EDMS optimisation. [email protected]