Why we need better, cheaper clinical trials

Clinical trials are the most important bottleneck to progress in medicine

This series is focused on biomedical progress. Given that, it’s only appropriate that I kick it off with a series of posts on clinical trials. Clinical trials are crucial to developing new treatments, yet they are probably the single most significant bottleneck to progress in medicine.

Organizations like the Institute For Progress, leaders like Rob Califf of the FDA, and scientists Martin Landray of the good clinical trials initiative have done an excellent job highlighting why clinical trials are so important and why we should be focusing our attention on making them faster and cheaper. But for those who have not yet had a chance to read their materials, I’ll briefly summarize why trials are so crucial and why they are so desperately in need of reform.

We need clinical trials to prove that treatments work

In recent years, there have been many who have questioned whether we should be so reliant on randomized controlled trials. I am sympathetic to those who are asking these questions; clinical trials are expensive and time-consuming. Meanwhile, technological advances have made alternatives to clinical trials more compelling. I’ll cover these technological advances in far more detail in later posts, but for the sake of simplicity I’ll highlight two key areas.

First, there have been efforts, driven in part by AI and the explosion of biomedical data available to researchers, to obtain a deeper understanding of medicine. Companies like Unlearn are seeking to leverage AI to eliminate trial and error from medicine. From these efforts, researchers hope that we can vastly shrink and simplify clinical trials – or eliminate them altogether.

Second, there is the proliferation of real-world data – that is, data obtained from routine clinical practice, now accessible through electronic health records. Alongside this data, researchers have developed of new techniques that allow us to draw causal conclusions from this data about which medicines and treatments work. Both the availability of the underlying data and our ability to work with it are expanding rapidly, and I plan to cover these developments in detail in later posts. I view our progress in causal inference methods in particular as an important and under-appreciated achievement.

But while these technologies hold enormous promise, they are far from being able to replace the central role of clinical trials today. The first barrier is regulatory: FDA and other drug regulators are already grappling with how to incorporate these technologies into their own regulations. In later posts I will describe in detail some ways that FDA could further embrace these advances. But I suspect progress in this area will be gradual, and clinical trials continue to play a critical role for the foreseeable future.

The second barrier is more fundamental: despite advances in basic science and the advent of real-world data, many – if not most – clinical questions of great importance cannot be answered without clinical trials. For example, real-world data collected from routine care can’t tell us about the impact of experimental treatments that are not yet widely available. And our understanding of basic biology, while expanding, is still insufficient to allow us to predict whether drugs will successfully treat a disease – even though I have high hopes that this will change over time.

Clinical trials are the principal bottleneck to biomedical progress

So for now, clinical trials remain an essential, inescapable part of biomedical product development. This illustration from phrma, the drug industry trade group, shows how clinical trials drive drug development costs:

Although this diagram is not quite drawn to scale, evidence shows that clinical trials take up the bulk of the time and money associated with drug development; a cost compounded by the cost of failed trials that don’t make it through the development process. And when it comes to investments like new drugs, time is money: The most widely-cited estimate of drug development cost attributes roughly half of the cost of drug development to the cost of capital – in other words, the opportunity cost of spending millions of dollars developing a drug that may not produce revenue for over a decade. When we see rates of return of pharmaceutical investments dipping below what is necessary for continued investment, it is reasonable to pin at least some of the blame on the cost of lengthy and costly clinical trials. This is not to mention the real human cost when lengthy clinical trials delay treatments for patients.

Moreover, the problem is getting worse. In an interview with Stat News, George Yancopoulos, the founder of Regeneron commented:

“Certainly the cost of clinical trials has become so outrageous we have to do something to change it,” ... He remembers that when he started it cost $10,000 per patient to conduct a clinical trial. Now it can cost $500,000, he said. “Just think how expensive that can be. It really limits what we can do.”

How do we fix clinical trials?

Those who follow progress and economics in other sectors will notice a familiar pattern here: There is enormous progress in the world of bits, coupled with stagnation in the world of atoms. On the “bits” side, we see tremendous progress in AI and a newfound ability to capture and harness health data at scale. On the “atoms” side, we see a stagnant clinical trial landscape in which costs appear to be rising.

In my next posts, I’ll cover, exactly how inefficient today’s clinical trials are. After we discuss the trials themselves, I’ll go deeper and explore exactly why the trial ecosystem is so inefficient. Why, despite the enormous financial incentives – not to mention their impact on human health – have we been unable to fix clinical trial costs? I’ll highlight some of the parallels between clinical trials and other sectors of the economy where we see similar stagnation.

After I discuss the state of trials today, I’ll talk about what we can do to improve them. Throughout my career, I’ve had the privilege to participate in conversations with some of the nation’s leading experts on clinical trial reform. I’ll share some of the best ideas I have heard, along with some important issues that I believe these conversations have missed. In keeping with the theme of this series – which is about the innovation ecosystem – I’ll focus on how we can address our institutions, incentives, and infrastructure to support more efficient and abundant clinical trials.

-Adam