Reusing clinical data to accelerate pharmaceutical R&D

In April 2023, we had the opportunity to speak at the Clinical Data Interchange Standards Consortium (CDISC) European Interchange conference on the opportunities and challenges for reusing historical clinical data for pharmaceutical (pharma) research and development (R&D). CDISC is an organisation that has pioneered the development and exchange of clinical data standards to facilitate the submission of clinical trial data to regulatory bodies like the Food and Drug Administration. Since then, biopharma companies have been exploring the use of these data standards for wider applications beyond submission to regulatory agencies. In this blog we discuss the opportunities and the challenges that organisations face in reusing their clinical data.

Drivers for innovation in pharmaceutical R&D

During a clinical trial, sponsors spend billions of dollars to collect a wide variety of data for operational and compliance purposes. Leveraging standards like CDISC to organise and reuse this data can enable sponsors to be more insightful about the next target molecule they pursue, or the next clinical trial they run. It can also be used to develop commercial and market access strategies for their target molecule. This potential application of historic data is largely untapped due to several challenges including the lack of access to the data, lack of governance processes on the reuse of data and outdated internal platforms to collate and analyse the data.

The success of a future-focused pharmaceutical R&D organisation is dependent on their ability to perform advanced analytics on a wide range of multi-modal data. Pharma organisations therefore need to invest in data standardisation and analytics infrastructure to keep pace with innovative R&D practices and o extract value from the internal R&D efforts. As described in Figure 1, there is evidence of investment in data driven R&D across several pharma organisations, technology companies, and regulators such as the FDA and the EMA. Large sets of good quality, standardised data from multiple sources is foundational to enable AI and generative AI initiatives within the organisation.

Figure 1: Overview of the internal aspirations and external trends that are driving life-sciences organisations to be data-driven

Source: Deloitte research from publicly available sources, 2023

Signature issues in clinical data reuse

From an industry perspective, organisations are struggling to recruit the skills and capabilities to bring together data, analytics, and insight to capitalise on opportunities for data-driven research and development. The signature challenges that pharmaceutical organisation face are:

• Operational - enabling data access for reuse is a complex process that should be made clear and transparent to involved parties. Organisations struggle to enable a governance framework that is transparent to business users and is supported by processes to provide standardised data for consistency in decision making. Transparency and standardisation help to build trust and confidence to use the data.
• Functional - the siloed nature of organisations limits the cross-functional thinking and collaboration for data-driven R&D. Data collected for every trial is tied to the trial outcome and data experts need to understand ‘how’ and ‘why’ the data was collected before reusing the data for a new analysis. Success in this area requires the creation of a cross-functional team of clinical data standards experts, delivery specialists, data engineers, and scientists.
• Technical - approaching clinical data management as a purely technical challenge drives organisations to pick platforms and tools that shoehorn end users instead of enabling them with the flexibility that they need to create value.

Core features of an effective solution for enabling clinical data reuse

A good clinical data management solution needs to deliver three core features: transparent data harmonisation (data products), flexible data analytics (analytics products), and scalable security and governance (business products). The core capabilities and principles required to successfully deliver these features are described in Figure 2.

Figure 2: Core capabilities driving R&D data management

Source: Deloitte analysis, 2023

Recommended approach for clinical data reuse

Recognising that every organisation will have a different level of maturity in their clinical data management journey, our recommended approach is to start by identifying a set of foundational use cases that resolve immediate business issues. These use cases could be:

• addressing findability and searchability of specific clinical data assets within your organisation
• pooling data to support publications and eminence for a disease area
• cohort identification for trial design
• evidence generation to expand indications for a drug program
• piloting AI/Generative AI initiatives for drug discovery.

The key is to be specific around a disease area or question that can be addressed within a reasonable time frame (10-12 weeks) as this will help focus the team on the outcomes in delivering the foundational use cases. These most common use cases can be classified into four areas: Find, Access, Solve, or Transform. Figure 3 describes some examples.

Figure 3: Find, Access, Solve, and Transform examples for clinical data use cases

Source: Deloitte analysis, 2023

Once the use cases are identified, organisations should bring together cross-functional teams that include members from their data standards, security and governance groups, therapeutic area medical team, and engineering & delivery team. This team needs to work in an agile and collaborative manner to scope out the data and infrastructure needs to address the foundational use cases.

From a technology standpoint, organisations can assess their internal technology infrastructure and make decisions on investment based on the answers to the following questions:

• What use case are we trying to enable?
• How much data is needed to address this use case?
• What are the types of data required to address the use case? (For example, clinical data, genomics data, imaging data, real world data)
• Do we have existing infrastructure that can be reused as a starting point?
• What are the tools and technologies that our data scientists and medical experts using for their analysis?

The technology and architecture choices for secondary use of data should be driven by analytics and insight generation needs. The core principles of transparent data harmonisation, flexible data analytics, and scalable security constructs should drive the platform roadmap. This approach keeps the focus on end-user and business value delivery thereby enabling organisations to create the traction needed to scale internal data reuse efforts. Figure 4, provides a summary of the key takeaways.

Figure 4: Key takeaways for organisations to enable the reuse of clinical data

Source: Deloitte analysis, 2023

Concluding remarks

The R&D organisation of the future is data-driven. The changing industry and ecosystem trends are pushing companies to capitalise on opportunities for data-driven research and development to remain competitive. Standardisation of historical clinical trial data is foundational for R&D organisations to generate actionable insights, leverage latest technologies like generative AI, and develop their future R&D strategies. In order to succeed, leaders need to bring together clinical standards experts, engaged medical teams, and delivery specialists and empower them with a data platform that is scalable, flexible and drives automation.