Three Ways to Help Cut Clinical Trial Timelines in Half: White Paper from McKinsey & Co
Advances in science must be accompanied by advances in clinical operations in order to bring cutting-edge medicine to more patients. This white paper from McKinsey & Co identifies how the life science industry can utilize innovative collaborations, information technology and new approaches to clinical operations workforce to do so.
Can clinical trials be faster and better? Yes. Here’s how.
Put six life science researchers in a room and it’s likely that there will be six different opinions on almost any subject except one: that the sector is on the cusp of an era of breakthrough innovation. Compounds in active development doubled from 2014-24, and the number of Phase 3 clinical trials in every major category has increased sharply since 2000 (see chart). And the potential of generative artificial intelligence (gen AI) is only beginning to be realized.
At the same time, clinical trials remain time-consuming and costly: Going from Phase I to launch often takes a decade and success rates are stuck at 10 to 12 percent. Although there has been considerable recent experimentation, such as adaptive randomization and seamless Phase 2/3 trials that implement pre-specified changes in real time, the science continues to run ahead of the system. Moreover, many patients who could benefit from participating in trials cannot do so. For example, only one in 10 of Americans with schizophrenia, and one in four heart patients, can join a trial at the same clinic where they get care.
By blending best practices with thoughtful innovation, we believe it is possible to cut clinical trial timelines in half while expanding access, reducing costs, and improving outcomes. Here are three principles that can help make that happen.
Foster collaborations that broaden the clinical trial patient population.
Clinical trials are typically based in academic medical centers, limiting the population that can take part. By developing a hub-and-spoke system, these institutions could link up with geographically dispersed community hospitals, regional health systems, clinics, and even pharmacies.
This would broaden patient reach and access, particularly for cohorts who are now underrepresented; maximize evidence generation; and extend the benefits of new medicines. In 2022, for example, Walgreens used its database of 130 million patients to identify potential participants living near its stores and notified them about trials they could join. It also formed agreements with 25 partners in biopharma, academia, non-profits, and government.
Use information technology to optimize trials.
Training AI and advanced analytics on existing data sets could help researchers refine criteria on whom to include, optimizing endpoint selection and better defining patient populations. AI has also proved excellent at site selection, which is critical to performance: Top-enrolling sites typically outperform median sites by two to four times.
Gen AI can be used to analyze protocols, identify historical trials with similar characteristics, and crunch trial- and site-level data and performance metrics. All this leads to more precise clinical conclusions, and improved R&D productivity, a longstanding issue for the industry.
As these examples indicate, stage by stage, AI is useful. The greatest potential, however, is in the integration of generative and agentic AI throughout the process. Reducing the clinical development timeline by a year can translate into more than $400 million in net present value across a sponsor’s portfolio. The rapid progress of agentic AI brings this promise closer to practice.
Because AI agents can collaborate, use tools, and learn from experience, they can complete complex workflows, efficiently and precisely. They can also do critical, time-consuming tasks like tracking documents and flagging pending items; agents that are embedded into electronic medical record workflows could identify potential trial candidates. Finally, an AI co-pilot can help managers in a variety of ways such as prioritizing critical issues, executing specific actions, and addressing delays. speeds up trials, while improving both accessibility and accuracy.
Define and recruit the talent needed for clinical operations.
If the way trials are done and analyzed evolves —and they will—so must the human skills that support them.
Over the last decade, McKinsey analysis has found that monitoring has become much more centralized, rather than site by site. Clinical research associates (CRAs) are managing several times as many protocol sites as in the past. That means they need expertise in overseeing site relationships. Other clinical operations roles are transitioning from executing tasks to managing strategic relationships. Interpersonal and engagement skills are therefore increasingly important; sponsors will need to build this expertise, whether by developing these skills among CRAs or assigning dedicated key account managers.
Based on the specific task, a range of specialized roles will be necessary. In ultra-rare disease trials, there may be a need for a patient liaison to work closely with the medical team. In chronic disease trials, the better option may be for a digital recruitment team. Across the board, sponsors will need to ensure competence in analytical skills and data-driven decision-making.
Doubling trial speed and patient participation by 2035 is a stretch goal, but a realistic one. Success would have many positive consequences: more accessible trials, better evidence, and most important of all, improved treatments for the unwell. Progress will require imagination, commitment, and a decision to put patients at the center of drug development.
Gaurav Agrawal is a senior partner in McKinsey and Company’s New York office. Piotr Pilarski is a partner in Washington DC. Valentina Sartori is a partner in Zurich. Kevin Webster is a partner in San Francisco. All are associated with McKinsey’s Life Sciences practice.
In this article
