David Northmore, vice president, EMEA, MarkLogic, explains how a data hub can resolve IDMP data integration challenges.
Information overload
Focused on improving patient safety, the new EU ISO IDMP standards are set to create a few IT headaches for pharmaceutical firms as they race to get their data in order. The first phase of IDMP came into force in July 2016, and any firm operating in the EU that fails to demonstrate IDMP compliance risks a fine of up to 5% of annual EU revenues. And while it is not yet legally binding, the USA, Canada and Switzerland have also signalled their support.
Implemented by the EMA, the goal is to simplify the exchange of data between all stakeholders by providing a universal system for identifying drugs. This includes how they should be used, consumed and packaged and is applicable to all stakeholders including pharmaceutical drug researchers, developers, manufacturers, distributors, registration authorities, product safety analysts, and quality control specialists.
But achieving this unified view of medication formula and dosage requires a data strategy rethink.
Much of this information currently lies in multiple, unconnected data silos such as billing, clinical, manufacturing and supply chain management systems. The flurry of pharma M&As over recent years presents additional data challenges, creating even more system fiefdoms and silos that have to be joined together to provide an integrated, consolidated view for reporting and regulatory purposes.
To add a further layer of complexity, around 70% of this drug data is unstructured or semi-structured – for example, clinical dossiers, protocols, package inserts and regional label submissions. Add to all of the above the fact that the IDMP standards are not yet finalised and the data challenges become even greater. Maintaining compliance against a moving target is tough!
For IDMP compliance all of these different types of information need to be stored, classified and made available in the same way as traditional transactional data.
The traditional approach to this data conundrum would be to use a rigid relational database linked to a Master Data Management system and a data lake to hold all the data. But the relational model is not designed to handle unstructured data and is too inflexible to accommodate changes without resorting to many man hours of data wrangling - extraction, transformation and loading (ETL). Costs go up, timelines increase and business agility is stymied. In some cases, ETL work can account for a staggering 60 per cent of a project’s overall costs.
Some early industry estimates suggest that the cost of transitioning from the current XEVMPD requirement to IDMP is conservatively forecast to exceed €38 million across a sample of 14 European Federation of Pharmaceutical Industries and Associations (EFPIA) member companies. It would take many thousands of man hours to manually curate a multinational pharmaceutical firm’s data, using a relational model, in readiness for IDMP compliance: identifying and collecting all the data sources and then manually linking the relationships between these data sources. Automating this process using an Enterprise NoSQL database with in-built semantics and the ability to learn as it goes is far more cost-effective. One proof of concept we have done for a customer suggests it would be possible to dramatically reduce the number of man hours needed to build such a system.
Building an Operational Data Hub, using an Enterprise-grade NoSQL database, to integrate all of these data silos delivers a solution that is specifically designed for rapidly changing, multi-structured data applications. A data hub is a virtual filing cabinet that can hold a single unified view of all drug-related data, both structured and unstructured. Effectively a complete drug lifecycle management system, every single ‘event’ that happens is recorded, meaning that firms and regulators have a system of record for all drug-related data and can easily understand what is going on with any drug, anywhere in the world, and at any point in time.
Once this 360-degree single view is enabled, it is easy to find the data needed for IDMP compliance - and to see commonality between data sets. This is important for producing safety summaries from data captured on drug trials, and actual data once the drugs are licensed, for example. Having this single source of truth makes it much easier to plan in the case of an adverse event: find out what a particular drug is approved for in a given region, review information on previously identified adverse events, identify any potentially counterfeit batches, and determine whether the batch may be incorrectly labelled. This data not only helps firms to identify other drugs with potentially similar effects, but also means they can react quickly to streamline recalls and perhaps even isolate the recall to a single country.
Choosing the right NoSQL database that comes with integrated search and semantics capabilities and full enterprise-grade ACID compliance is essential. In a database with ACID capability, even the largest datasets are processed consistently and reliably so none of the data is ever altered or lost. Importantly, the system can also be quickly adapted, extended and enhanced to meet changing business and regulatory requirements.
Because IDMP is all about looking at the relationships between data, all relevant information needs to be harmonised within and across enterprise boundaries and natural language processing/standard vocabularies supported. This is where integrated semantics and search come into play. Asking a question in a number of different ways and getting the same answer consistently is a must. Members of the pharmacovigilance (PV) and manufacturing departments, for example, need to be confident that they will surface the same answer from the system to the same basic question on a historical product recall, regardless of how they have framed the question: a single source of truth. The semantic capabilities also need to be flexible enough to handle the fact that drug vocabularies invariably change throughout the average product lifecycle of 5-10 years.
A database that supports bitemporal data management is also rapidly becoming a must-have for the pharma industry. With its roots in the financial services sector, bitemporal allows firms in heavily regulated industries to minimise risk through “tech time travel”—time stamping and rewinding documents to identify changes by looking at data as it was over the course of time. This capability can help pharmaceutical firms with reporting as well as helping them to demonstrate and maintain compliance with, for example, the IDMP standards as they evolve.
There is no time to lose. The pharma industry needs to act now to get its digital filing in order to meet the emerging IDMP standards - and ultimately to help enhance patient safety.
An operational data hub is the perfect prescription for curing the data pain points associated with IDMP compliance. Benefits include improved operational efficiencies, faster data retrieval, and more informed decision-making. And of course the ability to avoid large – and wholly avoidable - regulatory fines.