What Clinical Trial Trends From 2024 Reveal About Cost Complexity, Evidence Gaps

Clinical trial design decisions are often made years before a therapy reaches the market, but their consequences are increasingly felt by payers at launch.¹

A 2025 Phesi report assessing data from 2024 offers a window into these trends and into how the clinical development choices sponsors make may shape downstream costs, access, and evidence gaps across the health care system in the years to come.²

The report draws on nearly 66,000 actively recruiting clinical trials from its artificial intelligence (AI) platform, revealing that breast cancer remains the most studied disease worldwide for the fifth consecutive year, followed by solid tumors, stroke, prostate cancer, and non–small cell lung cancer.¹ The analysis also highlights encouraging but still concerning trends in trial progression:

1. Phase 2 attrition rates have declined to about 1 in 4 trials terminated early, yet this level remains elevated compared with prepandemic norms
2. At the country level, the US continues to host the highest volume of recruiting investigator sites across these top indications, while China shows the fastest site growth among major markets
3. Beyond the top 5, obesity—buoyed by an expanding body of research around glucan-like peptide-1 therapies—narrowly missed the ranking, reflecting shifting research interest that may reshape future development priorities

Where Clinical Trial Activity Is Concentrated

Clinical trial activity in 2026 remains heavily concentrated in a familiar set of global markets, according to Gen Li, president and founder of Phesi, in an interview with The American Journal of Managed Care^® (AJMC^®). These patterns reflect both historical leadership and evolving regulatory expectations. The US continues to lead clinical development activity, a position reinforced by its advanced research infrastructure and deep integration of industry, academia, and health systems.

Other consistent contributors include Japan and several major European countries, which together round out the core regions driving clinical trial intensity worldwide. What is shifting, however, is not simply where trials occur, but why.

“It's the fact that the US started requiring the development programs to provide the adequate portion of the patient data coming out from the US, which is, from a data science point of view, the right move,” said Li.

Ensuring that medicines are developed using data from the populations they are intended to treat can strengthen the relevance of trial outcomes.

“You have to have the medicines be developed by the people and data from the patient you are trying to treat, but on the other hand, it's further intensified the competition in the US, because we are already very heavily loaded with so much of the clinical trial activities,” said Li.

Trial Feasibility, Early Termination, and the Cost of Getting It Wrong

Despite advances in trial technology and analytics, the fundamental challenges affecting trial feasibility and early termination have remained largely unchanged, explained Li. One persistent issue is protocol design, where misalignment between trial assumptions and real-world patient characteristics often leads sponsors to amend protocols midstream.

These amendments, Li explained, typically reflect a gap between what sponsors intend to measure and the realities of the patient populations they hope to enroll.

Although the COVID-19 pandemic introduced additional disruptions, Li noted that even as those pressures have receded, feasibility challenges have continued at similar levels.

Another major obstacle is patient enrollment within a reasonable timeframe, which Li linked closely to investigator site performance. Identifying sites with both the appropriate clinical expertise and access to eligible patients remains difficult, particularly as competition for participants intensifies. When these issues are not addressed early, trials face delays, underenrollment, or termination altogether.

“Those are the 2 kinds of error challenges facing us,” said Li. “And of course, as we will continue to talk about, there are ways we should be able to improve those situations and overcome some of those challenges in a pretty material kind of way because of the AI and other technologies.”

Increasing Trial Complexity and Its Impact on Evidence Generation

As medical knowledge advances, trial complexity is increasing, particularly in oncology and hematology, explained Li. Modern research increasingly recognizes that patients with the same disease can be highly heterogeneous at the cellular and molecular levels. For example, in lung cancer, it is no longer sufficient to identify patients with a KRAS mutation; sponsors now often target specific subtypes, such as G12C or G12D, dramatically narrowing the eligible population.

This growing specificity requires not only more precise patient identification but also advanced diagnostic capabilities at investigator sites, along with infrastructure to ensure patients are treated correctly according to their biomarker profile.

“All the experiences of the investigator sites needed to move along with it, and so we needed to build the infrastructure around those biomarkers to be able to treat those patients correctly,” said Li. “Those are the things all requiring basically revolutionary ways to better design our trial and better ways to identify the investigator size with a much higher level of precision than we can as of now for most of the sponsors.”

The result is that trial design must become far more sophisticated, requiring revolutionary approaches to site selection, enrollment strategies, and protocol execution. Increasing complexity, although scientifically necessary, further challenges evidence generation, as smaller, highly targeted populations can limit the generalizability of trial results and make demonstrating robust clinical outcomes more difficult.

The Role of AI and Real-World Data in Smarter Trial Design

One of the most transformative opportunities in clinical trial design lies in leveraging AI and massive real-world data sets to anticipate and solve challenges before a trial begins, explained Li. Phesi integrated more than 300 million patient records from around the world, contextualizing those data with information about who collected it, where and when it was collected, and under what protocols. This platform allows sponsors to virtually simulate trials before enrollment starts. This approach also allows researchers to proactively identify potential issues with protocol design, site performance, and patient access, reducing the need for midtrial amendments and improving trial feasibility.

“In that way, we can anticipate a lot of the challenges and pray and be proactive, and therefore, in a very material way, design those clinical trials to avoid amendments and have better investigator sites, which we know have that kind of access to the right patient population who [we] are inspired to treat,” said Li. “We are actually seeing those other things among our clients.”

He added that adoption of AI in clinical development is no longer optional; it is fundamentally transforming the industry, and sponsors who fail to embrace these tools risk falling behind.

In clinical trials, AI platforms like Phesi’s digital patient profiling help model control arms, minimizing patients’ exposure to placebos or inferior comparators.³ Additionally, AI-driven recruitment tools streamline enrollment by matching patients to relevant studies, addressing barriers related to awareness, eligibility, and site access.

These AI-enabled approaches not only enhance trial efficiency and precision but also expand access to innovative therapies, making clinical research more patient-centered.

As clinical trial development enters 2026, the combination of complex, targeted protocols and advanced AI-driven analytics is reshaping how therapies are tested and evaluated. Leveraging real-world data, digital patient profiling, and predictive modeling offers sponsors the tools to design smarter, more efficient trials.

For managed care, understanding these trends is increasingly essential as the quality, relevance, and efficiency of clinical trials today directly influence coverage decisions, treatment access, and health system costs. Ultimately, aligning innovation in trial design with patient-centered evidence generation promises not only faster development but also more reliable outcomes for patients and payers alike.

References

1. Global data analysis of clinical development: Top five studied diseases in 2025. Phesi. January 7, 2026. Accessed January 16, 2026. https://info.phesi.com/top-five-2025-studied-diseases

2. Revolutionizing clinical study design: The role of AI and analytics. IQVIA. June 12, 2025. Accessed January 16, 2026. https://www.iqvia.com/blogs/2025/06/revolutionizing-clinical-study-design

3. McCrear S. 5 AI and digital health advances transforming breast cancer care and research. AJMC^®. November 14, 2025. Accessed January 16, 2026. https://www.ajmc.com/view/5-ai-and-digital-health-advances-transforming-breast-cancer-care-and-research