How AI makes European Health Data Space data usable
AI is making data accessed through the EHDS usable—turning fragmented access into secure, scalable workflows.
How to choose a medical research data platform (and why most pharma strategies fail)
Why medical research data platforms are not meeting research needs
A recent Frontiers in Digital Health review of medical research data platforms — coauthored by experts across academia and industry, including BC Platforms’ Chief Scientific Officer Timo Kanninen — highlights a growing gap between platform capabilities and the needs of modern pharmaceutical research.
Translational research is becoming increasingly collaborative and multimodal. In practice, this means your medical research data platform determines something simple: whether teams can combine data quickly — or spend weeks cleaning, aligning, and reconciling it every time.
Most medical research data platform strategies are built on outdated assumptions
Pharma organizations have spent the last decade investing in data infrastructure. Yet many still struggle to scale real-world evidence programs or move AI beyond pilots.
This shift toward more collaborative, multi-source research is already changing how data is used in practice. Clinical data, real-world data, and omics are no longer analyzed in isolation — they need to be combined across studies, partners, and geographies.
Most platforms, however, were designed for centralized storage and limited use cases. They do not reflect how research is actually conducted today. This misalignment creates real consequences. It slows study setup, limits data reuse across programs, and reduces the ability to translate insights into clinical impact.
The review highlights that organizations struggle to obtain “a sufficiently comprehensive market overview within a realistic timeframe,” making medical research data platform decisions inherently challenging.
Medical research data platforms are fragmented and hard to evaluate
The medical research data platform landscape described in the review is broad, heterogeneous, and constantly evolving, spanning academic and commercial systems with very different capabilities.
There is no standard model. No clear comparison framework. No universally accepted benchmark. As a result, medical research data platform selection is often based on partial visibility or inconsistent criteria. Organizations compare systems that are not designed to solve the same problem.
For pharma, this leads to fragmented architectures, duplicated effort, and increasing complexity when scaling multi-partner research.
There is no one-size-fits-all medical research data platform
Yet many organizations still approach the decision as if a single, optimal medical research data platform exists.
The review concludes that medical research data platform selection must be context-sensitive, based on a requirement-weighted approach rather than fixed rankings. Yet in practice, many organizations still rely on feature comparisons or vendor positioning.
This creates a mismatch between medical research data platform capabilities and research needs. For pharma teams, that mismatch affects real-world outcomes, including the ability to integrate external data, run multi-cohort studies, and scale analytics across programs. The study explicitly avoids “one-size-fits-all comparisons” and emphasizes aligning platform capabilities to specific use cases.
Medical research data platform capabilities are uneven — and many organizations operate below what they need
Even when a medical research data platform is selected, another issue quickly emerges: not all platforms operate at the same level.
The review introduces a five‑tier model of platform capabilities, from data hosting to federated learning. These tiers represent fundamentally different levels of functionality, not incremental improvements.
Many organizations still operate at the lower levels — focusing on storage and limited sharing — while modern research requires advanced analytics, multimodal integration, and cross-site collaboration. This creates a growing gap between medical research data platform capability and research demand.
Why interoperability matters more than infrastructure
This leads to a more fundamental question: what actually distinguishes a platform that supports research from one that does not? The review identifies the most important criteria for platform selection as security, interoperability, data quality, and support for multimodal data. Not infrastructure.
Storage and compute are necessary, but insufficient. Research depends on data that can be connected across systems, interpreted consistently, and reused across studies and programs.
Without interoperability, data remains siloed regardless of infrastructure.
Federated data platforms are becoming essential for pharma research
This becomes even more critical as data moves beyond a single organization. A key shift highlighted in the review is the move from centralized data models to distributed and federated approaches. In centralized models, data is moved into one system. In federated models, data remains local and analysis happens across sites.
The distinction reflects different “data-movement logics”: either bringing data together or moving computation to where data resides. This shift is driven by regulatory constraints, data ownership, and the rise of multi-partner research. For pharma, this is already a reality in global trials and real-world data programs.
Data usability, governance, and interoperability are the real bottlenecks
But even with the right architecture, another challenge remains.
The review highlights that the biggest challenges are not technical infrastructure, but the ability to manage data consistently across workflows. Data must be:
- harmonized
- standardized
- traceable
- governed
Without this, it cannot be reused or scaled.
This explains why many initiatives stall. Data exists, but it cannot be combined, compared, or used across studies.
Medical research data platform strategy is now a competitive advantage in pharma
This is where platform design becomes a business issue, not just a technical one. For pharma executives, platform decisions are now directly tied to business outcomes.
They determine:
- how quickly studies can be launched
- how efficiently real-world evidence can be generated
- whether AI initiatives scale
- how easily partners can collaborate
Organizations that remain infrastructure-led will continue to face friction. Those that adopt a data-first, interoperable, and federated approach gain a structural advantage.
This direction aligns with the approach highlighted in the review.
The real question to ask
Ultimately, this all leads to a much simpler question.
It’s not which medical research data platform to choose — it’s whether your teams can actually reuse and combine data across studies and partners, or whether they end up rebuilding everything every time.
As the review puts it, medical research data platform selection needs to be “context-sensitive” — it should reflect how data is actually used, not just what a system can store. If that isn’t happening, the platform is already the bottleneck.
Read the full Frontiers in Digital Health review: Selecting medical research data platforms for translational biomedical research: a five-tier overview and requirement-weighted assessment framework
Is your medical research data platform enabling scalable research — or creating hidden constraints?
Most platforms store data. Few make it reusable. BC Platforms enables secure, governed access to multimodal health data across a global network — so teams can analyze data in place and scale research across studies and partners.
FAQs
Most strategies fail because they focus on infrastructure instead of data usability. Medical research data platforms are selected based on storage or compute capabilities — not on whether data can be reused across studies, partners, and workflows. As a result, data accumulates but cannot be effectively used.
Typical signals include slow data access, repeated data preparation across studies, difficulty integrating external datasets, and AI programs that fail to scale beyond pilots. These indicate that the medical research data platform is not aligned with research needs.
Interoperability, governance, and data quality are the most important factors. These determine whether data can be combined, reused, and scaled across research programs — far more than infrastructure itself.
A federated model becomes necessary when data cannot be centralized due to regulatory, ownership, or collaboration constraints, but still needs to be accessed and analyzed across multiple sites or datasets.
Because it directly impacts speed, scalability, and cost. The right medical research data platform reduces time to insight, enables data reuse across programs, and supports AI and real-world evidence at scale.
Last updated