Data Harmonization: Creating Order from Chaos

Elvis Pacelat asks why the prospect of harmonized content is so challenging, and what can pharma companies learn from other sectors about more optimal, modern ways of working and sharing data?

Elvis Pacelat

In an industry so mired in regulatory compliance requirements, life sciences organizations might be expected to be adept at managing information. Any omission in data recording and reporting is fraught with risk, so companies must follow agency specifications to the letter, or provide a complete audit trail for every step of every process - i.e. those involved in developing products, bringing them to market, and tracking them across their lifespan.

It is a heavy and daunting burden of responsibility. In pharma, every tweak to a product or the way it is supplied must be verified and validated - a huge and rigorous undertaking in terms of documentation and labour. Vigilance is all. One wrong move - incorrect dosage, mislabelling, inaccurate translations for new markets, or incomplete advice - could have such profound implications. It could mean direct risk to customers’ health, catastrophic fines and reputational damage. As a consequence, companies can become paralysed and afraid of change. This in turn can stand in the way of ambition and innovation, harming companies’ competitive prospects.

No organization is an island

Few other industries face the same demands and constraints, other than perhaps the energy sector (particularly nuclear energy). One of the many factors contributing to the complexity experienced in life sciences is the vast number of variants of each product - from the point of initial formulation in the lab, to its numerous manifestations in each market. Long before a product exists in its marketable form, it has a development history that needs to be documented. For administration purposes, it is assigned an ‘identifier’ so it can be tracked from its earliest compound stage.

In other markets, such as healthcare, it is logical to follow the subject (e.g. a patient) right through their case lifecycle so that there is a continuous record of everything that has happened, where and when, and with what result - from the moment the case opened to the moment of discharge and beyond. Increasingly this is actually happening too, via joined-up patient medical records that can be accessed and updated electronically across all aspects of treatment.

As a result of the requirement for Trusts in England, for example, to deliver discharge summaries to general medical practitioners within 24 hours of a patient leaving hospital, many Trusts have extended their electronic patient record systems to enable them to create and submit discharge summaries electronically (another benefit bring the built-in audit trail). In advanced scenarios, hospitals capture everything from diagnoses, co-morbidities, investigation and results, to procedures and treatments, medications and any follow-up arrangements, and issue a summary to the patient’s GP electronically. Extended systems can also be used to transmit an approved dispensing sheet to pharmacy, allowing medication to be dispensed efficiently. Episode coding is more reliable too in a joined-up digital system, because the recipient no longer has to decipher handwriting when interpreting a colleague’s notes.

As a result of cross-discipline content management, where this has been deployed, healthcare organisations not only meet their regulatory obligations, but benefit from slicker internal processes, reduced administration costs, and the ability to deliver improved patient care. Information sharing with other organizations – primary care givers, other hospitals and specialist healthcare practitioners – is also made easier once content is stored electronically in a format that can be accessed by authorized partners.

Breaking down silos

In life sciences this fluidity of information collation, movement and sharing doesn’t yet exist. Or it is the exception rather than the norm. More usually, data is recorded and stored locally in a format that suits the immediate task and supporting system. This may be manual, or a proprietary application-specific IT system. But this does not serve the wider needs of the organization, or help the company deliver against increasingly strict information compliance demands.

What needs to happen now, for a whole raft of regulatory and commercial reasons, is for pharma to think beyond the immediate purpose of operational data, towards the bigger picture - and broader usage.

Take that earliest product identifier – the code that defines the compound chosen as the basis for a new drug. Its existence should allow the business to track the development process as the selected combination of active ingredients is developed and brought on – from the lab and as it fulfils its market potential.

Having a unique, consistent identifier throughout the research and development process allows an information trail to form, which will play an important role in the marketing authorization process. The entire testing process and each round of results needs to be easily traceable too, from pre-clinical to clinical trials, covering the overall safety of the new compound, to its efficacy and optimum administration route and dosage.

From a patient and regulatory perspective, and indeed in the interests of risk management, all of this constitutes a duty of development. Marketing approvals are very specific too, based on agreed dosages and methods of application, reinforcing the need to get the detail right and keep it consistent. But the specifications, along with the eventual product name, can vary from country to country. This is where data capture and management becomes particularly complex in a life sciences context. Should companies define and track products by name and country, or by the chemical compound/active ingredient?

Consistency brings calm when chaos is all around

Because of the implications of information governance for customer safety and market confidence, national and international authorities are continually tightening the controls and increasing their demands for data collection and reporting, so that the onus is on life sciences companies to have all of the answers at their fingertips and ready to submit at short notice. Without a single, joined-up view across operations and products, this is difficult to achieve.

The emerging ISO Identification of Medicinal Products (IDMP) standard involves a highly involved set of information requirements. Even in the initial phase of the evolving standard, more than 30 different pieces of information are required - a figure that will rise to over 90 and then to hundreds of different elements, once the standard is more advanced. That’s a considerable administrative commitment.

So there are multiple, compelling reasons for pharma organizations to take control of their information landscape. Although some momentum around IDMP preparation has been lost by shifting deadlines, these requirements will be enforced, and aligning information management systems and processes will take time to get right. Preparation should not be delayed then, particularly if companies want to avoid a last-minute panic - which may prevent them delivering a more holistic, and broader-reach solution that could bring the business a host of other benefits.

IDMP paves the way

A positive side effect of IDMP is that it fosters a consistency in data management that hasn’t existed before in life sciences – across research and development, manufacturing and quality assurance, and beyond. It marks a move away from a very fragmented approach to information governance, dictated by departmental silos.

The practical way forward, companies are beginning to realise, is to move towards a holistic repository – a master resource, where all product-related content is amalgamated (either physically or virtually), ensuring that all departments and parties along the supply chain are referring to the same data. A resource that meets the needs of IDMP and new marketing authorization demands simultaneously.

This requires a facilitating infrastructure, something that is typically lacking but which is increasingly prevalent in other industries.

Emerging data and document management models such as the DIA TMF Reference Model offer organizations a good starting point, and gradually organisations will find that there is a choice of good, configurable solutions to choose from that will help them to collate, store and process data and documentation cohesively, in the same standard format from one end of the product lifecycle to the other. Using rules-based workflow and data sharing, such systems will help life sciences organisations accelerate deployment because they already provide the main building blocks that most companies need, allowing for minimal customisation to suit each company’s unique needs.

Assigning responsibility

The next consideration is who will champion and drive the harmonisation efforts. While regulatory affairs has traditionally acted as a collection point for compliance related data, organizations should not assume this will continue. Holistic product data management is much bigger than a single department, and needs its own champion – a responsible party who will oversee the whole flow of data from R&D to eventual product withdrawal, and can help enforce consistency end to end. This is the approach commonly taken in other markets, and has become a model for best practice.

As in other industries, responsiveness will be a key capability that organisations will need in future, for reasons of competitive advantage as well as compliance. In the case of pharma, the first draft of guidance on IDMP specifies that companies will have just 30 days to meet their IDMP data obligations once a marketing authorisation application has been submitted - a challenging requirement that cannot be met in an ad-hoc, cobbled-together way.

IDMP is just one of many new or improved regulations coming down the line, too - another reason that firms can’t afford to be too fixed and niche in their information management plans. Bespoke, single-purpose systems lock companies in and undermine any aspiration towards greater agility. They are expensive to own and maintain and cannot adapt easily to new requirements.

Having a uniform architecture and centralized master data resource, and configurable/adaptable systems, is the only real way organizations can prepare themselves for whatever the future brings. It’s a lesson other industries have learnt, and if life sciences stands apart it is only in that its needs for order are greater - not because they are fundamentally different.

About the author

Elvis Pacelat, is VP of Compliance Solutions at AMPLEXOR Life Sciences (www.amplexor.com).