Real-World Evidence Europe at eyeforpharma Barcelona

Mar 12, 2019 - Mar 14, 2019, Barcelona

Become the catalyst of trust and quality

RWE: The Missing Link

Linking real-world datasets is essential to deliver true insights

Pharma’s investment in real-world evidence shows no sign of drying up. “Very few pharma companies today are not investing in RWE initiatives, especially if they are among the top 50,” says Bengt Anell, Associate Director in Market Access at Shire, which provides unrestricted research grants to researchers engaged in epidemiological studies using real-world data linked to multiple data sources, and sponsors patient registries tailored to rare disease patients.

Yet, current levels of investment in generating real-world data (RWD) will only get pharma so far. Greater investment is needed if companies are to effectively link data from disparate sources, a key step in enabling data to ‘speak’ and provide meaningful and actionable insights.

This article is from our latest 'Trends in Real-World Evidence' magazine. Download for free.

Data linkage challenges
Linking data is not an easy process; not only does it require the right technological tools and knowhow, it raises a range of concerns, from patient anonymity and data ownership, to how to transcend geographical barriers.

Anonymizing data is a fundamental challenge, says Nigel Hughes, Scientific Director at Janssen Clinical Innovation – Patient Data for Research. “Data linkage requires, via an underlying consent model, linking different data together – with the patient being the common linker.”

He says this is a challenge from a “socio-technical perspective”, especially in term of complying with the EU’s GDPR [General Data Protection Regulation] or HIPAA [Health Insurance Portability and Accountability Act of 1996] in the US.

“For instance, HIPAA comprises 18 identifying variables which, when removed, can ensure compliance with anonymized data, but it degrades our ability to link the data,” says Hughes.

Creating a “safe haven” where data can be linked before it is anonymized is one potential solution, says Hassan Chaudhury, Director and co-founder at Health iQ. A safe haven permits access to only a few data operators who replace all identifying information with unique coded identifiers, attached to the data permanently. This is already being employed by NHS Digital [in the UK], which plans to link datasets by default.

“This is only the beginning of the journey,” says Chaudhury. “We need everyone to pull in the same direction to make it work. It will be as rapid and as widespread as the industry wants.”

Data ownership
Another important issue to be tackled if data is to be effectively linked is the fact that patient-generated data belongs to the patient and so should not be used without permission.

This is a message driven intensively by the US start-up, – the world’s first organization to develop human rights around data ownership. While its CEO, Richie Etwaru, agrees that data linkage has the potential to solve some of the big problems in healthcare, “linking data is a lower-level problem. The higher-level problem is whether the data being linked was obtained and commercialized with appropriate levels of consent, authorization and permissions.”

He asks whether data is being de-identified “legally, ethically, and socially responsible” to soothe the growing divide between citizens and corporations on the topic of data ownership.

His solution is the world’s first global data consent and authorization blockchain. “We have anchored the consent and authorization around a title of data property ownership that is issued to consumers.”

This “new reality,” as Etwaru refers to it, has been coined Fair Trade Data, and his company is working with data collectors, aggregators and sellers to implement the consent and authorization blockchain.

Even if pharma can navigate around the Scylla and Charybdis of patient anonymity and ownership, it must avoid one last obstacle, the far-less-technological issue of national boundaries and policy.

Issues of data ownership may see national authorities employing policy that make it difficult to connect databases in different countries. For example, the separate datasets of a patient who emigrated may be impossible to connect.

The Digital Healthcare Data Federation project, which brings together partners across Europe, is destined to circumvent such barriers, says Chaudhury. “If you want to link data from Scotland and Italy, neither country will let the data leave due to fears that it will break the social contract. However, if the states run the query themselves and the results are sent to a central hub – and then on to researchers – this federated model means the data itself never leave its home State, but the data can link at the central hub and be changed to the right language.”

Where is pharma now?
Data linkage might be a hot topic for pharma right now, but other players in the health ecosystem may need more time to catch up.

“There are some hospital networks that are linking more to primary care records, rather than just summary data, but it’s small-scale and nascent today,” says Hughes.

However, pharma companies can help drive greater linkage, says Anell. “Shire is offering patient programs that are advancing data linkage between treatment of rare disease patients and quality of life and clinical outcomes.”

This includes a collaboration with the largest rare-disease patient community, PatientsLikeMe, to create a repository of patient health data that can be linked to genotype and physiological data, helping the company to track research outcomes and gain deeper insights into the needs of patients with rare diseases.

If solutions are to be found that overcome all the obstacles and deliver on the needs of all stakeholders, Hughes believes “an open discussion in society” is needed to highlight the benefits of sharing and linking data, because it needs to be “enshrined in policy, law and standard operating procedures”.

Some projects are already working to make ethical data linkage possible – programs pharma and its partners can definitely learn from. They are:

Fast Healthcare Interoperability Resources (FHIR)
A standard framework developed to facilitate the exchange of electronic healthcare data while maintaining data integrity. While a significant element of this framework is the standardization of data, it has a ‘Linked Data Module,’ designed to facilitate data linkage across datasets.

A Resource Description Framework (RDF), which is a standard for data interchange built around existing Web standards such as XML and URL, is used to achieve this. RDF allows for data integration from multiple sources, which can then be analyzed as one data source.

EMIS Health
A data extraction service designed to help the UK NHS access primary care data that could aid decision-making, improve patient care, and reduce costs. A key aspect of the resource is the OpenPseudonymiser algorithm, which can be applied to different datasets to pseudonymize data and link it without sharing identifiable information.

Clinical Practice Research Datalink (CPRD)
Jointly funded by the UK’s MHRA – Medicines and Healthcare products Regulatory Agency – and the National Institute for Health Research, de-identified primary care patient data is collected from a network of general practices and linked to secondary health-related datasets such as Hospital Episode Statistics, diagnostic imaging data, and cancer registries, among many others. In keeping with data privacy laws, patients can opt out of their data being used.

Secure Anonymized Information Linkage (SAIL) Databank
Much like the “safe haven” suggested by Chaudhury, the Wales-based SAIL Databank provides a governed infrastructure for linking anonymized health and population data. To access the databank, researchers must undergo a rigorous application process, complete Safe Researcher Training, and be assessed by the Information Governance Review Panel. In addition, all data undergoes a four-stage anonymization process that comprises a re-combining process and additional encryption safeguards after the initial encryption.

Accelerating value-based healthcare
The ability to link data across the value chain and patient journey could be transformative for the pharmaceutical industry, accelerating moves towards value-based healthcare. Yet companies are faced with some key challenges around technical capabilities, internal infrastructures, and governance and ethics.

To overcome these challenges, there is the need to learn from some of the robust data linkage projects currently underway and to combine these with advancing technology such as blockchain to take data linkage to next level.

This article was first published in Trends in Real-World Evidence in December 2018. To download a free copy, click here.

This article is from our latest 'Trends in Real-World Evidence' magazine. Download for free.

Since you're here...
... and value our content, you should sign-up to our newsletter. Sign up here

Real-World Evidence Europe at eyeforpharma Barcelona

Mar 12, 2019 - Mar 14, 2019, Barcelona

Become the catalyst of trust and quality