Clinical development is approaching a breaking point. The accelerating pace of innovation that enables life-changing, targeted therapeutics for smaller, more specific populations of unmet need is also heightening the financial risk profile of drug development.

In this context and against the continued rise of artificial intelligence, it has become clear that the systems that enable clinical data to flow remain stuck in the past. Centuries-old processes have been digitized, not yet transformed; “digital paper” has taken over the physical, but the underlying workflows remaining manual, fragmented, and reactive.

Innovators in other regulated industries have shown that this doesn’t need to be the case. Disruptive financial technology firms and neobanks, such as UK’s Revolut and Canada’s Wealthsimple, show the opportunity: real-time, fluid, and secure data with stringent privacy measures served to customers to meet their specific needs.

The potential for pharma: insights served to whoever needs to interrogate the data, in the way they like it via customizable analytics platforms, and with it the opportunity to become more efficient and better aligned with the demands of modern science, regulation, and patient expectations.

The problem is not new. It is, however, becoming ever more urgent to address as the healthcare sector faces ballooning expenditures even as budgets are set to stabilize.

Adding to this are intensifying measures to manage drug pricing, e.g. through centralized procurement in China, and an increasingly vocal policy environment in the United States, the leading global market for pharmaceutical products, all set against a precipitous patent cliff of more than $180 billion in sales at risk through 2030. What you get is a new priority: bringing productivity and process innovation to the next level, across both drug development and clinical practice.

Waiting and seeing how these trends play out would carry the cost of foregone opportunity, slowing down the path toward a better future for drug development:

• Faster, less expensive, and higher quality trials enabled by data-driven trial design and execution that can contain inflating drug development costs and total cost of care.
• Streamlined regulatory process as talent-strapped agencies like the FDA are enabled with more tailored insights on the data and can reach accelerated times to approval with greater efficiency.
• Strengthening patient trust as individuals are well-aware of how their data are collected, used, and shared, driving widespread societal acceptance and long-term participation in trials.
• Increased investments in new therapies as investors deploy capital in a world where the pace of innovation is well-supported by data and other infrastructure.

To bring this future to life, the industry must reimagine how clinical data flows end-to-end, from study planning to regulatory submission. This also means moving beyond isolated fixes and toward a connected, intelligent, and patient-centric approach to data.

The vision exists; the urgency is clear. What’s needed now is the alignment and execution.

Fragmentation is Slowing Us Down

Years of digital transformation, with acceleration during the COVID-19 pandemic, have brought newer tools, but the processes they support are still rooted in outdated assumptions. The challenges crop up across the length of the trial lifecycle, e.g.:

• Protocol design is often misaligned with execution. Clinical protocol complexity is growing; according to Tufts CSDD, data points have tripled compared to a decade ago, yet a quarter of it goes unused as they do not support core endpoints (i.e., primary, key, secondary, safety).
  • The true impact to trial quality is often poorly understood by sponsors but are clearly significant: time to last patient enrollment rose by 37% and both substantial protocol amendments and dropout rates rose by over 100%, alongside heightened data compliance challenges and slower reviews. (citation: Tufts CSDD; sources: https://link.springer.com/article/10.1007/s43441-023-00595-1, https://www.clinicaltrialvanguard.com/conference-coverage/tufts-csdd-new-insights-on-the-clinical-trial-industry/)
• Data capture remains cumbersome, and error prone. Study teams and sites are still burdened with repetitive data entry across disconnected systems such as study specific electronic data capture (EDC) systems and their own health records systems.
  • This not only increases the risk of human error and reduces the quality and consistency of the data collected for the trial, but also further strains good but overburdened healthcare systems as they continue to mitigate labor shortages, aging population, and public health crises.
• Digital tools introduce new operational burdens for sites. While patient-reported outcomes (PROs), electronic clinical outcome assessment (eCOA), wearables and devices offer richer and more real-time data, their integration often presents a challenge to sites and teams.
  • The heterogeneity of the solution landscape leads to constant demands on site staff to familiarize themselves with new technology, and sometimes juggle many of these in parallel across several studies as sites become de facto help desks for patients.
  • This adds strain to already resource-constrained environments and distracts from core clinical tasks. A recent survey by the Critical Path Institute highlighted that although most sites report positive sentiments towards eCOA, over a third of site staff most desired better training for deploying them and more responsive technical helpdesk. (source: https://www.sciencedirect.com/science/article/pii/S2451865423001874).
  • For trial teams, likewise, their integration implies greater need to shift attention to providing technical support.
• Patients are kept in the dark on their own data during trials. Even with advancements in transparency regulation, patients still rarely get access to their data at the conclusion of trials, with what data they could receive varying greatly (e.g., lab results, genetic data, imaging data).
  • A recent survey conducted among pharma firms by the PHUSE organization shows that some organizations still return no data at all, while others offer limited types, e.g., individual clinical results, aggregate study data, or occasionally genetic or imaging data, without consistency.
  • The method and timing of data return also vary widely, from secure portals to paper reports, and may depend on geography, regulatory constraints, or participant request. This lack of standardization in what, how, and when data are returned creates significant inconsistency for patients across studies and geographies, further impeding efforts to build engagement, especially in long-term trials.
• Data cleaning is still governed by human-led workflows. Most review processes rely heavily on manual checks and back-and-forth queries, making them slow, inconsistent, and difficult to scale, especially in high-volume trials with global regulatory submissions. These inefficiencies not only delay timelines but also introduce additional data quality risks.
• Regulatory engagement is sequential by design. Sponsors spend two to three months on average to prepare submissions, expending significant effort to prepare the data and document analyses towards differing requirements and areas of interests across countries, all the while not knowing how individual reviewers at competent authorities would ideally consume the information. They then wait for months-long review windows to get formal feedback.
  • This mandated waterfall structure stifles timely response to early insights and drags out the entire development cycle. Alternatives like rolling reviews/submissions have been successful in some cases, but remain the exception, with great potential remaining to further formalize data sharing arrangements, including through initiatives such as the FDA’s Federated Data Platform.

These are not isolated issues but interconnected barriers that reinforce one another. Without coordinated, strategic change, fragmentation will continue to erode efficiency, inflate costs, and delay access to life-changing therapies.

Envisioning a Seamless Clinical Data Future

Through conversations with pharmaceutical leaders representing 14 of the largest global pharmaceutical companies at the PHUSE Future Focus Workshop, and then at the SCDM 2025 EMEA Conference Industry Summit with added representation from major technology firms and clinical research organizations, a shared vision has emerged.

In this future, data runs with the speed and reliability of a Swiss train or Amazon same-day delivery. Transparency in near real-time enables faster decisions, smarter trials, new insights, more equitable outcomes, and builds mechanisms for a smarter healthcare system that learns from the wealth of data from interventional studies and standard care delivery to steadily improve care provision.

Imagine a clinical ecosystem in which data flows seamlessly and securely across sponsors, CROs, regulators, and clinical sites. Instead of relying on emails, spreadsheets, and PDFs, stakeholders interact through interoperable platforms that speak the same language, reducing friction and accelerating collaboration.

AI and automation are embedded at every layer of the data pipeline, from capture to submission, supporting quality reviews, curation, transformation and evaluation. In this future, development teams are no longer working in isolation.

They’re empowered by live, unified data environments that provide a single source of truth that enables faster iteration, adaptive trial designs, and more responsive decision-making. Standards are streamlined, and regulatory engagement is continuous rather than episodic.

Patients are also further informed and empowered throughout the trial, not just at the end. They have visibility and control over their own information as it is collected to understand how their participation actively contributes to bringing new drugs to markets for other patients.

Critically, this vision is global by design, and numerous organizations are already working toward facets of this vision, on their own and through organizations such as Accumulus Synergy.

Emerging data environments such as ICH M11 and CDISC are built to be interoperable from the start, not retrofitted after the fact. First steps have been taken on solutions for automated data cleansing and statistical analyses with many other use cases like AI-led submission preparation in reach.

By coming together to address the changes laid out next, we can accelerate progress and avoid fragmentation along the road.

The Way Forward Starts from What is in Our Grasp

The vision for a seamless, intelligent clinical data ecosystem is ambitious; it is also attainable, beginning with steps we can take today to challenge inefficiencies and unlock transformation. Across our conversations with industry leaders, three strategic imperatives emerged as the foundation for progress: evolve the foundation, transform the capabilities, and disrupt the model.

First, evolve the foundation

What’s needed is a complete infrastructure reset: modular, interoperable systems that support real-time data flow, automation, and quality at source.

• Implement digital-first data capture that circumvents double entry at sites, while building integration with devices and other endpoint technologies that patients are already familiar with (e.g., smartwatches, smart home hubs) to simplify their deployment. This will also support the integration of decentralized procedures with potential to reduce both patient and site burden given the mitigation of growing expectations for sites to be technical help desks.
• Gradually integrate electronic health record (EHR)-to-EDC pipelines as the clinical landscape matures and formats are harmonized.
• Replace fragmented workflows with interoperability via application programming interfaces (APIs), as seen in early success stories like HL7 FHIR pilots and Project Vulcan, that allows sponsor systems, CRO platforms, and site tools to exchange data seamlessly, reducing reliance on manual transfers and enabling faster, informed decision-making.
• Streamline data collection, cleaning, and validation through AI-driven automation with the aim of transforming data cleaning from a bottleneck into a background process. This carries the potential to reduce human error and free up valuable time of the study teams to focus on higher-value activities.

If we get this right, data will flow more seamlessly, captured once, with built-in quality checks, securely curated for accuracy, and made accessible to the right stakeholders at the right time. It’s a pragmatic leap forward, grounded in what’s feasible today and essential for what comes next, while respecting the guardrails needed to preserve scientific integrity.

Second, transform the capabilities

New tools demand new processes. As real-time data platforms and AI-driven analytics are embedded, organizations must evolve their operating models. Centralized, live data environments should become the norm, offering a single source of truth that spans the development lifecycle.

Systems alone aren’t enough. Unlocking their value requires multi-disciplinary teams with shared accountability, unified around clear outcomes.

• Create integrated working models in which roles evolve to support continuous insight generation, faster iteration, and real-time data responsiveness. This is underpinned by governance structures that move from static, phase-based oversight to more dynamic, cross-functional guidance aligned with real-time development needs.
• Establish future-ready capability frameworks that prioritize adaptability, technical fluency, and strategic problem-solving to ensure that teams are not just participating in transformation but driving it.
• Invest in upskilling to support new ways of working, as cross-functional teams evolve how they collaborate and apply expertise in increasingly digital, automated, and real-time trial environments already pushing to implement adaptive trial designs and decentralized protocols.

Building and strengthening new capabilities doesn’t just apply to the trial sponsors; critical interfaces with external parties, such as trial sites and regulators, must also be examined to ensure the whole system for clinical development can effectively collaborate and collectively capitalize on the benefits of new technologies.

Done right, this is not just a staffing upgrade, it’s an operating shift. The organizations that lead will be those that redefine who owns data, how insight is generated, and what execution excellence looks like in a real-time world.

Finally, disrupt the current model

Ultimately, the target is to unlock the true value of seamless data flow: a learning system that serves both drug development and the provision of healthcare. At the center of this shift is an AI-powered “translational layer” that evaluates data from disparate sources and consistently synthesizes it into analyses and insights in the format that best serves the individual user’s needs.

By surfacing what matters most in a manner that is faster and with greater precision, this layer elevates both the speed of execution and the relevance of decision-making, creating a continuous learning system that will serve to improve cross-sector outcomes and better serve patients.

• Deploy an AI-driven translational layer to generate fit-for-purpose, stakeholder-specific insights, while maintaining full data traceability and integrity, as well as preserving integrity of analysis with appropriate use of data both blinded and unblinded, to ensure reliability across clinical, regulatory, and commercial domains.
• Build a learning system that integrates real-time clinical data with external data sources such as EHRs, wearables, and community trends to continuously improve trial design, treatment strategies, and population targeting.
• Align and integrate diverse and disjointed data sources, including clinical, safety, pharmacovigilance, EHR, and non-traditional inputs, such as digital biomarkers and social media, into a harmonized data platform to contextualize a rich dataset and make it more easily consumable to generate richer analyses.
• Empower patients as active participants with a global standard for data transparency to unlock greater trust, equity, and engagement in research. With the right regulatory approach and privacy guardrails in place, patients may even be further engaged to participate in trials and the opportunity created to better integrate clinical trials as part of patients’ care pathways.
• Greater patient engagement and utilization of their trial data comes with an even greater need to establish appropriate bioethical guardrails amidst potentially new use cases and applications. For example, greater visibility of data for patients will need to be accompanied with appropriate patient-centric education to ensure non-maleficence and patient autonomy, and increasing application of AI systems will need to be thoroughly tuned for bias to ensure justice and fairness for all patients.

This is how clinical development becomes more agile, inclusive, and cost-effective. It’s not just about better data; it’s about a smarter system that continuously learns, serves its users, and delivers value across the entire healthcare ecosystem.

Rewiring the System with the Future in Mind

The future of clinical data isn’t waiting for a technological breakthrough; digital-first platforms, real-time data environments, and AI-powered insights are well within reach. What’s missing is an ecosystem-wide alignment around how to deploy them at scale and the drive from sponsors, regulatory authorities, and society-at-large to overcome the resistance of legacy thinking that are no longer optimal in serving the needs of medical science and patients.

Across our conversations with clinical data leaders, a clear consensus emerged. The barriers are real, but so is the opportunity.

What’s needed now is a coordinated push to modernize the infrastructure, capabilities, and governance models that underpin clinical research. This means evolving foundational systems to be more connected and interoperable.

It means transforming how teams work, embedding data science into clinical operations and enabling faster decision-making. And it means disrupting outdated assumptions about data ownership, access, and value.

It’s a call for focused, collective progress, anchored in what we can do today and open to what’s possible tomorrow. By modernizing the clinical data value chain, we can create a future where data moves with purpose: reliably, intelligently, and in service of both science and patients.

But lasting transformation depends on more than any one sector. It will take a broader coalition—industry, regulators, standard setters, healthcare systems—aligning around a shared vision.

The blueprint is in place. The momentum is real. Now is the time to deliver.

The authors would like to thank all participants of the PHUSE Future Focus Workshop and the SCDM 2025 EMEA Conference Industry Summit for contributing their valuable perspectives on the future of clinical data. The authors would like to further thank the extended leadership teams of PHUSE, SCDM, and Kearney for their review and feedback.

About the Authors

Stephen Bamford, Ward Lemaire, Martin Hodosi, Sascha Ahrweiler, Tonny Huang, Cindy McShea, Carol Schaffer, Peter Krusche, Rishi Goel, PHUSE Board of Directors, SCDM Executive Team.