The largest community of pharma leaders

Why Too Many Clinical Trials Fail

By Julien Moussalli Market Expansion Lead at IQVIA (LinkedIn:

A makeover of the $65 billion clinical trials market (1) is well overdue. According to the National Cancer Institute, in the U.S only, 1.7 million people have been newly diagnosed with cancer in 2018 (2). In the meantime, pharma companies are trying to recruit thousands of patients for the 10,000 clinical trials hoping to test new life-saving cancer drugs. Sadly, according to the FDA Global Participation in Clinical Trials report, less than 5% of cancer patients will end up taking part in a trial (3).

During the past 3 years, Julien got the chance to meet with patients, patient associations and organizations to understand why the patient participation rate in clinical trials is so low. From those discussions, he outlined a combination of two factors that, in his opinion, are the main reason why the current clinical trial model is failing:

  1. Worldwide, patients willing to get involved in research are having a hard time in understandingif their profile is matching the eligibility criteria for a given clinical trials;
  2. Also, time and cost of participatingin a trial creates a high burden for patients and their families.

Unfortunately, there are many more barriers to participation which results in an industry in dire need of a makeover. However, for the purpose of this article (and because he wanted to keep it short and concise, so you read the article until the end), Julien focused on the two factors listed above.

Julien believes Artificial Intelligence (AI) holds the potential to make drug development more patient-centric, resulting in a higher quality of research. As a e-health and start-ups enthusiast, he compiled a list of start-ups actively working on disrupting the current clinical trial stages keeping in mind that collaboration between those new comers and large companies will be key to foster innovation in the clinical trial space.

Leverage new technologies to raise patients’ interest 

Pharma companies rely on technology to discover and develop breakthrough treatments that can turn deadly diseases into manageable chronic conditions – or sometimes cure a disease altogether. The costs of developing a new therapy and bringing it to market can top $2 billion (4), and the research and development (R&D) process often relies on a clinical-trial model that remain unchanged since the 1990s! “Tomorrow, we will be asking those millennial (an over-connected group of individuals who don’t feel comfortable when they are far from screens) to participate in an old-school trial using paper to collect data? Forget it!” said Julien.

Artificial Intelligence (AI) has the potential to disrupt every stage of the clinical trial process – from matching eligible patients to studies to monitoring adherence leveraging the Internet of Things (IoT).

The first of many challenges that patients are facing is finding a clinical trial. According to Tufts Center for Drug Development, 80% of trials are delayed often related to patient recruitment – and 48% of sites miss their enrollment targets (5).

Let’s face the reality, patients rarely get trial recommendations from their physicians.

Julien had the opportunity to discuss with patients during conferences and at Hospital, he understands they often refer to patient communities to find the appropriate trial for their conditions. Alternatively, they courageously navigate some trial registration platforms, such as (6), where it can be difficult for them to understand the medical jargon – just imagine someone telling you:


You didn’t even read the whole sentence so just imagine how Gerard is feeling on the picture below.

According to the author, that is the ultimate paradox: “Although we have the majority of patients willing to participate in a clinical trial (for potential personal benefits and altruistic reasons), we end up with that majority desperately looking for a trial, and sponsors and CROs looking for patients to participates in their clinical trials. These recruitment delays are costly to pharma.” According to Industry Standard Research (ISR) Reports, when a trial goes beyond its intended deadline, sponsors lose millions of dollars in sales (7).

Here AI can help both sponsor companies and patients by extracting pertinent information from medical records and compare it to the requirements of ongoing trials.

Thanks to a simple matching algorithm, the right clinical trials could then be recommended to patients. It’s not without significance that we see many Patient Recruitment Services start-ups on the rise.

Julien had the chance to meet with few of them, starting with Pablo Graiver, CEO and co-founder at Antidote. Antidote’s trial-matching technology does what Kayak did for travel. The company has so far connected tens of thousands of patients to 120+ clinical research studies, across 26 countries and 14 languages.

Another start-up called Deep 6 AI analyses structured data, such as ICD-10 codes, and unstructured clinical data, including doctor’s notes, pathology reports and operating notes to find more, better-matching patients for clinical trials in minutes, rather than months. The software uses AI and natural language processing to extract tens of thousands of new clinical data points – symptoms, diagnoses, treatments, genomics, lifestyle data, and more – turning fragmented medical documents into unified patient graphs that contain all the information needed to match complex clinical trial criteria.

Large service companies have also taken patient enrollment to the next level. Research shows to starting-up an oncology site typically takes up to 378 days (8) – thanks to the Precision Enrollment approach, IQVIA can do it within 21 days by opening sites only when they have identified a patient, thus reducing protocol amendments due to unexpected problems with planning, feasibility and enrollment.

Treatment Adherence & Improved Data Quality

Once a patient is admitted to a trial, they need to adhere to the medication regimen.

Information on a patient’s drug usage is generally maintained in paper diaries, which can be inaccurate due to human error and is also inefficient. Also, travel to the clinic, as well as out-of-pocket costs, increases the likelihood of patients dropping out of a study.

Again, are we seriously going to ask millennial to drive two hours for their monthly on-site visit? Are we seriously going to ask them to complete a paper diary twice a day so that they come prepared for their visit?

They are always on their smartphones, let’s just leverage the tool part of our clinical trials.

AI-enabled trial management systems can help keep patients engaged. Technologies such as digital reporting apps, as well as wearables, allow for real-time engagement and communication, and support patient-centric trials. Patients can send feedback of treatment symptoms, share information with researchers, reducing or eliminating the need for patients to travel to sites, thus increasing patient adherence and compliance.

Let’s take a concrete example. For people with memory-degrading conditions like Alzheimer, mobile technologies can remind them to take their medicines. But those innovations are not limited to one specific disease – have you heard about Abilify MyCite? A pill prescribed for treatment of schizophrenia. It has a digital ingestion tracking system and has been approved by the FDA back in November 2017. Ingestible sensors and wireless pill bottles are being used to track drug intake, but in the meantime, we use paper diaries to collect data: our industry is fascinating!

AI can even go a step further to track adherence using visual confirmationAi Cure Technologies, a New York based start-up, uses an artificial intelligence platform to automatically confirm medication ingestion. The app is available on every smartphone and real-time patient adherence data are encrypted. Unlike Facetime or Skype, the system relies on computer vision algorithms to confirm the process of medication administration, no human review is necessary (9).

Another start-up in this space, Catalia Health, is developing a healthcare companion and coach that uses AI to tailor conversations to patients, set reminders, and ask questions. The robot assistant will use a touchscreen or voice activation feature to communicate with patients.

Where Do We Go Next? 

Technology will clearly be a necessary component to streamline trials. For some researchers, this presents another challenge: The needed technologies are relatively new, and pharma has always been slow to adopt new solutions.

“There may be a degree of truth to that but when I see the FDA providing its green light to commercialize a schizophrenia drug with an ingestible sensor and patients willing to welcome an AI based solution to support treatment adherence at home, I am comfortable to say that our industry stakeholders are well aware of the need to innovate and are already thinking out of the box.” said Julien.









Note from the Author: These opinions expressed here are my own and may not reflect that of the organizations I’m connected to.




Recent Articles