Clinical Trials Reimagined: The Impact of Secure Data Sharing on Pharma R&D

Life sciences organizations have traditionally been early adopters of data science and analytics practices, which have revolutionized how they conduct research, develop life-saving therapies, deliver personalized medicine, and optimize their operations. Central to these advances is the ability to share data among healthcare providers, researchers, regulators, and more.

Data sharing fosters collaboration, informs decision-making, and improves patient outcomes. However, in such a highly regulated industry and with such sensitive data involved, the benefits of data sharing are accompanied by significant challenges. An increasingly disparate data landscape and myriad of cybersecurity risks make it difficult to manage secure, compliant data access without sacrificing speed and scale.

These challenges necessitate the adoption of evolving security measures, smarter technologies, and programs to safeguard sensitive information, uphold the company’s reputation, and maintain customer trust.

The clinical trial process in particular is an area that can benefit hugely from both internal and external data sharing. However, the majority of studies are not designed with data sharing in mind, making the data de-identification process complicated and expensive. Streamlining this requires an understanding of how data is used and shared in clinical trials, and how data products will help enable collaborative, secure research.

What is a Clinical Trial?

Before new drugs are introduced to the market, they must undergo a rigorous procedure – called a clinical trial – to ensure the drug meets certain standards. A clinical trial is a research study conducted on human beings to determine whether new drugs, diagnostics, or treatments are both safe and effective.

As patients, we trust that the medications and treatments that our doctors prescribe to us have passed through the clinical trial process, but we don’t necessarily know what exactly that process entails.

A clinical trial has four distinct phases:

How Data Is Used in Clinical Trials

Data is collected at every phase of a Clinical Trial, from the pre-clinical stage to the post-approval Phase IV. For example, there is data from protocols, consent forms, case report forms, clinical study reports, individual patient-level data, and market research data.

Clinical trials are increasingly using digital products. This includes moving from paper-based records to electronic records, which can provide vast amounts of raw data. The interpretation and analysis of the clinical trial data can also be in itself a form of data.

It is widely recognized that clinical trial data sharing is vital for fostering transparency, quality, and scientific advancement; reducing research waste and duplication; and sustaining confidence in the pharmaceutical industry.

For instance, sharing participant-level data, raw digital product data, and protocols, could lead to novel secondary analysis and improve decisions around future trials. Sharing clinical study reports supports meta-analysis and supplements clinical care. And, sharing data with the trial participants and the public increases the transparency, supports clinical care, and improves overall confidence in the sector.

Since 2013, a large proportion of the industry in both the US and Europe has committed to:

• Share participant-level data, study-level data, and protocols with qualified researchers
• Provide public access to clinical study reports synopses for all clinical trials reaching Phase IV
• Share summary result reports with clinical trial participants
• Publish results from all Phase III clinical trials and any clinical trial of significant medical importance

Since 2016, organizations have also recognized the need to follow the FAIR guidelines: clinical trial data must be Findable, Accessible, Interoperable, and Reusable. Many steps have been taken over the past decade to meet these principles, but there is still a long way to go in terms of meeting the commitments set out above.

How is Clinical Trial Data Shared?

The urgent need to share data is clear. What is less clear is the background of strict regulations, standards, and protocols that researchers and companies must follow in doing so.

Establishing patient confidentiality measures and procedures for identifying who should have access to what data is non-negotiable. This commonly leads to two data sharing models:

1. A centralized data model, where an independent body holds the data and determines who has access to it.
2. A decentralized model, whereby the researchers or research group retain control of their data and grant access to data requests at their discretion.

However, both of these models can be time consuming for the data consumer and the data owner, leading to overly restrictive and even all-or-nothing data access.

There has been some shift in thinking and new exploration of how to share data more openly. With the development of new data products, there is huge potential to move away from the slow and potentially expensive centralized models, and toward an approach that offers more instant access while still adhering to regulatory and confidentiality requirements.

Reimagining Data Sharing in Clinical Trials with Data Products

Data products have enormous potential to improve data sharing across clinical trials. The first step in operationalizing them is to understand the stakeholders involved.

Data Consumers

To understand more fully how data products help, it is important to identify the data consumers. This will tell you what they need from the data/data product, at what stage they need the data, and what access restrictions they should have.

Within the clinical trial itself, there are many key players with different data consumption needs:

• The research team comprises the people setting out the procedures, recruiting the doctors and participants, and analyzing the results, among other tasks, across various sites and research teams. They need access to all the data in a timely manner and generally would have full access to the data. Data products could be beneficial at every phase of the trial, from resource management and quick identification of safety issues or anomalies, to pinpointing methods of identifying ideal teams or patients.
• The doctors, nurses, and staff implementing the trials require unlimited access to their own patients’ data. But, they could also benefit from data products that highlight updates from other patients (e.g. dosage, side effects, benefits, adapted protocols, etc.), while masking the data to protect patient confidentiality.
• The trial participants and, potentially, their families usually have very limited access to the research, and are given information only when the results are complete (e.g. at the end of the Phase in which they participate). It would, however, be interesting to explore if real-time data shared between patients enriches the data itself. Researchers and care teams could potentially identify safety issues more quickly if the patients themselves could compare their experience to others in the trial.
• The business as a whole needs data to drive decisions. Any data product that could shed light on the financial implications of the various trial components, early identification of potential failures, and market analysis of the drug or treatment could be hugely beneficial.

Beyond those with direct involvement in the clinical trial, there are additional external data consumers. As with the internal stakeholders, they have different needs and appetites for the data:

• Outside Researchers may require the full details of every aspect of the clinical trial, with only minimal data masking to ensure confidentiality.
• Healthcare Professionals are interested in the results and issues affecting patient care, as well as how the test product compares to other drugs or treatments.
• Other Pharmaceutical Companies require the data to aid their own research, so what data they have access to depends on the intent. For instance, they may seek information on anything from the drug composition or dosage, to the trial’s financial data and sales.
• The Target Audience (e.g. people who could benefit from the drug) want to know the results, side effects, etc., and how it relates to them. The data needs to be presented in an easily accessible and understandable manner.
• The General Public requires overall information, such as aggregated information from many different clinical trials. Again, this must be straightforward and easy to digest in order to build confidence and transparency.

Benefits of Data Products for Clinical Trial Data Sharing

It’s clear that data sharing in and across clinical trials has massive potential, but is complex. The current practice of reworking data per study has led to data often remaining in silos, preventing teams from collaborating toward analytics-based insights. Such an approach can be cost-intensive and slow down the study or in some cases even contribute to a failed study.

Treating data with a product mindset focuses the study with data sharing in mind right from the start. Having a clear understanding of the data product’s consumers and the granularity of data they need, together with baked-in data security and access control policies, are all data product characteristics that improves data sharing for everyone involved in the process.

Imagine a future where:

• All clinical trial data is readily available in differing degrees depending on your consumer status, even during the trial itself.
• Adopting a data product marketplace approach for consumers to access the data improves data findability and availability. By offering different levels of detail through governed consumer output ports, you enable multiple consumer status endpoints on the same underlying data. A data marketplace also fosters collaboration and trust between data product owners and consumers.
• Consumers can easily combine data from a study with other studies or even data from other domains within the organization to better identify trends, patterns, and relationships between different aspects of the trial.
  
  Data products ensure a level of interoperability (the ability to work with other systems or products without additional effort on the part of the user). A good example of this is to enforce the use of natural business keys for typical master data components such as patient, HCP, product, etc.
• Safety issues and concerns are highlighted as quickly as possible, and shared across participants.
  
  The speed and agility in which data products help to generate insights and serve data to a much broader audience. Cross-referencing the data with other relevant data products helps to quickly identify anomalies and safety concerns in a data-driven way.
• Users can make instant, well-informed decisions based on the information that accompanies the data.
  
  Trustworthiness is a key characteristic of a data product. Accompanying metadata such as freshness, completeness, service levels, ownership, and usage ensures that consumers have all the necessary information to act decisively.
• Based on consumer status, the relevant data or data analysis is accessed in a timely fashion, or even instantaneously.
  
  Data products by nature demand a high level of automation and accessibility. Built-in access control policies and consumer entitlements reduce the need for lengthy manual approval processes and federate governance closer to data teams.
• Audit reports provide on-demand monitoring of who has access to what data, as well as how they’ve used and shared specific data products. This helps achieve industry guidelines and support GxP compliance.

Looking Forward

Data sharing plays a vital role in fostering scientific progress and supporting well-informed decisions in clinical practice. Most organizations do recognize the importance of data sharing, and support initiatives to enhance clinical trial data transparency and promote scientific advancements. Despite this, recent investigations indicate significant scope for improved data sharing within the field of clinical practice.

Health data is indeed sensitive and not always easy to share in a responsible way due to the high level of industry regulations and data protection guidelines that must be followed. However, with a shift towards a data product mindset, coupled with the right tools and technologies to help with processes like computational governance, automated access control policies, and observability, safe and secure data sharing is possible.