By 2025, clinical research and patient care converge: Data integration, validation, and evolving markets

A growing chorus of experts agree: 2025 signals a year when AI and its generative variants (genAI) move beyond early experimentation into a period of tangible, large-scale impact — including in the life sciences. While challenges remain as EY recently noted, the future will likely be one where the focus shifts from anticipating breakthrough genAI model advances to propel gains. Instead, it is more likely to center on optimizing mature deployments to deliver validated, reliable value in specific use cases.

Experts across the ecosystem—spanning data quality, patient-centricity, real-world evidence (RWE), and market access—align on this trajectory. “My prediction for 2025 is that we will see a shift away from talking about potential applications of AI and a shift towards real uses of AI in evidence generation,” noted Blythe Adamson, PhD, MPH, head of outcomes research and evidence generation, international, at Flatiron Health. “A key differentiator will be strong techniques for validation of AI to ensure its use is fit-for-purpose.”

This emphasis on validation mirrors the perspective of Noga Leviner, CEO of Picnic Health, who anticipates a convergence of data from clinical research and routine patient care: “In the coming years, the distinction between data collected for clinical research and data collected in real-world clinical practice will blur,” Leviner said. “Researchers will be able to tap into this rich, longitudinal data to gain deep, personalized insights about individual patients.”

On top of that, clinical trials are simultaneously growing more patient-friendly as sponsors increasingly prioritize patient needs, preferences, and experiences in the design and execution of studies.

Improving patient experience and embedding research in care delivery

Leviner’s patient-centric outlook highlights the need to tightly integrate research and routine care. Virtual clinics, remote monitoring, and digital tools will minimize patient burdens, ensuring their experiences are a “core focus, not an afterthought.”

This patient-centric approach aligns with insights from Shweta Maniar, Global Director of Healthcare and Life Sciences at Google Cloud, who recently highlighted how life science organizations are moving toward harmonizing data sources and scaling genAI deployments to accelerate personalized medicine. Organizations are fundamentally rethinking traditional clinical development timelines, she said. More firms are tapping genAI tools to efficiently mine extensive internal knowledge repositories. This underscores the emergence of a long-promised “n=1” patient-centric paradigm. “We no longer accept that it often takes 10 to 20 years for personalized medicine solutions to come to market,” Maniar said.

Data quality, validation, and adaptive design

Another factor in achieving trustworthy AI-driven insights is rigorous data quality. “In 2025, I see more fluid types and qualities of RWD coming to market,” explained Adamson of Flatiron. “More specifically, I expect there will be a more prominent element of adaptive design to ensure that the approach to data quality is reflective of its criticality to research, type of use case, and the complexity of curating it from source.” At Adamson’s employer, this strategy includes defining clear operational thresholds for when to use machine learning (ML)-based or human abstraction, or a combination of both. A decade of expertly curated data serves as a gold standard for validation.

Adamson also points to Flatiron’s Panoramic datasets as an example of scaling real-world evidence generation. By “unlocking access to 1.5 billion data points that can be longitudinally and frequently refreshed with longer patient follow-up time,” Flatiron can support research that dynamically assesses disease progression, treatment response, and patient outcomes over time. This data-rich environment, validated against expert curation, enables more precise, fit-for-purpose datasets that enhance AI’s reliability and relevance in evidence generation.

As AI deployments in life sciences mature, responsible application remains a priority. “Flatiron is fortunate to have access to a decade of expert curated data to validate the accuracy of AI/ML enabled data elements,” Adamson notes. Such grounding allows organizations to build trust in AI systems gradually. Leviner echoes this sentiment, affirming that AI should “enhance and augment the work of researchers, not replace them.”

Adamson predicts that in 2025, AI will unlock “speed and scale in RWD we haven’t seen before,” while Leviner anticipates a “continuously learning research ecosystem” emerging, with research deeply embedded in care delivery. These transformations dovetail with global RWD standardization. For instance, Flatiron’s FORUM (Fostering Oncology RWE Uses and Methods) consortium exemplifies ongoing efforts to ensure evidence transportability across borders. By 2025, says Adamson, “we will see more global standardization of RWE.”

Navigating market access, IRA, and biosimilar trends

Alongside advances in data science and patient engagement, the U.S. Inflation Reduction Act (IRA) and the expanding biosimilar market will shape 2025’s healthcare landscape. According to Kala Bala, SVP of Data Operations and EADE, and Andrew Rouff, Senior Consultant, Advisory Services, at Managed Markets Insight & Technology (MMIT)—a Norstella subsidiary delivering pharmaceutical market access analytics—manufacturers and payers will grapple with heightened uncertainty.

Strategic vigilance includes watching Medicaid redetermination, Medicare shifts owing to IRA provisions, and increases in HIX enrollment. Biosimilar trends are “becoming today what generics were previously,” Rouff adds.

They note that the coming year’s ambiguity stems from the changing political landscape, the next wave of IRA provisions, and the increased influence of biosimilars, particularly as pharmacy benefit managers (PBMs) begin white-labeling their own versions.

Impact of the IRA and ongoing pricing pressures

“One of the biggest challenges manufacturers are anticipating is the impact from the IRA—especially on formulary changes,” Bala says. “Payer behavior is expected to be very cautious due to the increased focus on being cost effective… The IRA’s price negotiation provisions for Medicare drugs could significantly lower the price of newly launched brands, forcing companies to carefully consider pricing strategies.”

There are also uncertainties about the IRA’s future. Citing its environmental provisions, President-elect Trump has vowed to repeal or roll back parts of the legislation. Yet drug pricing pressures will likely remain front and center—S&P Global recently noted bipartisan support for lowering drug spending, quoted by Fierce Pharma. In his first term, Trump previously expressed support for Medicare drug pricing negotiation though did not implement it. More recently, he has criticized pharmacy benefit managers, charging them with inflating prices.

Such conversations on reining in drug pricing are certainly not new, and Bala believes that brands facing lower price pressures and serving high unmet needs may still find launch success. “Indications with lower price pressures and unmet medical needs may witness an uptick in launch success because of the IRA’s incentive on innovation in such therapeutic areas,” she notes.

Biosimilars, contracting, and PBM strategies

Against this backdrop, biosimilar launches are also set to intensify. “We’ll see increased competition in many therapeutic areas in the form of biosimilars and/or novel mechanisms of action,” says MMIT’s Rouff. He anticipates that Medicaid redetermination, Medicare changes, and growth in health insurance exchange (HIX) enrollment will reshape payer mixes and channel exposures. These shifting channels echo Rouff’s broader guidance that healthcare organizations must track evolving coverage landscapes closely.

PBMs are using their position with white-labeled biosimilars to alter the status quo. “Humira biosimilars have had historically low uptake due to strong AbbVie contracting,” Rouff explains, “however, recent efforts by CVS and their subsidiary Cordavis have been successful in taking market share from Humira.” Meanwhile, Cigna plans on removing AbbVie’s Humira from some drug reimbursement lists next year, Reuters reported in August. With Stelara biosimilars on the horizon in 2025, similar strategies from Cigna’s Quallent and Optum’s Nuvaila may follow.

Key contracting trends are also emerging. “There is additionally a trend in warranties for brands like Hemgenix, Roctavian, and Beqvez,” Rouff notes. “It will be interesting to see how these innovative offerings will change in the future.” He also emphasizes the importance of interchangeability considerations, route-of-administration (RoA) factors, and therapy areas (TAs) in biosimilar contracting strategies.

Longer-term PBM reforms

While major PBM reforms may extend beyond 2025, Rouff highlights potential avenues for future change. These include adjustments to how PBMs reimburse their pharmacies versus independent ones. He also expects increased transparency around group purchasing organizations (GPOs), and regulatory incentives to foster greater PBM competition.

Making good on AI’s promise to bolster operational efficiency

As stakeholders navigate this shifting landscape, AI’s potential extends beyond clinical insights—encompassing operational efficiency and more personalized patient engagement. “Data privacy and cyber security concerns will need to be addressed in tandem to fully leverage these technologies for ultimate positive outcomes for patients and the industry as a whole,” Bala emphasizes. She notes the potential to streamline administrative processes: “While I do not believe prior authorization will ever go away entirely, there is much potential for greater efficiency from the status quo, which still involves fax machines and calls on landlines back and forth between the payer, patient, and provider.”

Though Rouff believes 2025 may be too soon for major AI-driven disruptions in payer behavior, payers are already eyeing AI to automate data retrieval for research, provide initial screenings for drug coverage, and support clinical decision-making. Over time, as these tools prove their reliability and transparency, AI will likely play an increasingly central role in both clinical and market access strategies.

Emerging therapeutics and market disruption

Beyond policy shifts and biosimilar competition, new therapeutic fronts are opening. “GLP-1s and their impact on obesity, diabetes, and potentially other diseases” represent a core area of innovation, Rouff says. Access will be bolstered by direct-to-consumer (DTC) programs: “DTC programs like LillyDirect and PfizerForAll will increase patient support in accessing therapies in TAs like obesity, diabetes, migraine, and vaccines.”

Therapeutic expansion is also underway in respiratory health. “Numerous biologics are expanding into COPD which will result in increased competition, new options for patients, and enhanced portfolio contracting opportunities for manufacturers,” Rouff notes.

These evolutions coincide with the rise of non-traditional players. “Industry disruptors like Mark Cuban’s Cost+ and Amazon One Medical are gaining traction and may continue to impact various TAs as they grow,” Rouff explains. That trend could potentially reshape patient access models, prompting payers and manufacturers to adapt more quickly.

Making promises an embedded reality

By 2025, the industry will transcend current boundaries. Instead of simply discussing AI’s potential, we’ll see it embedded across the value chain—guiding research, shaping patient experiences, informing pricing strategies, and influencing market access decisions. With robust validation standards, adaptive data quality frameworks, global data harmonization, dynamic payer-manufacturer relationships, biosimilar competition, and innovative market entrants, the life sciences ecosystem is poised to become more patient-centric, agile, and sustainably innovative.

“Rather than the traditional siloed, linear model, research will be deeply embedded into real-world care delivery,” Leviner said. “This will lead to faster insights, more personalized treatments, and better outcomes for patients.”

For healthcare organizations, staying agile and informed is critical. Rouff recommends carefully monitoring multiple evolving factors: “New developments across the IRA, biosimilars, disruptors, and new therapies all happen fast,” he says. “Timely insights can make the difference between success and failure!”

---

Filed Under: clinical trials, Data science, Drug Discovery, Regulatory affairs
Tagged With: AI validation in healthcare, biosimilar market access, data quality standards, IRA impact pharma, patient-centric trials, real-world evidence generation, therapeutic innovation 2025
 

---