- Clinical trials account for almost 40 per cent of the US pharmaceutical industry’s research budget
- Only 6 per cent of people with chronic illness enrol in a clinical trial
- Artificial intelligence helps patients find the right trial
- AI lowers trial dropout rates, as well
- AI also improves the data collection aspect of clinical trials
Clinical trials are essential for the development of new drugs. However, they’re very costly and time-consuming. Also, they’re rife with inefficiencies that are preventing potentially life-saving drugs from reaching the market. “The standard clinical trial is pretty much the only thing in medicine that hasn’t changed in the last 70 years,” says Don Berry, a professor in the Department of Biostatistics of the University of Texas MD Anderson Cancer Center and one of the world’s leading experts on clinical trials.
Lack of clarity, low patient enrolment and retention rates, poor patient adherence and compliance, and outdated data collection and verification methods with heavy reliance on subjective information supplied by the patients themselves are just some of the issues that plague today’s clinical trials, prolonging their timelines and increasing costs. Artificial intelligence could be the solution.
Clinical trials account for almost 40 per cent of the US pharmaceutical industry’s research budget
Clinical trials account for almost 40 per cent of the pharmaceutical industry’s research budget in the United States, which translates to approximately $7 billion per year. There are usually three phases in human drug trials. In Phase 1, researchers test the safety of a new drug on a small group of people and work out the most appropriate delivery method and the optimal dosage. In Phase 2, the drug is given to a larger number of people to study the efficacy. Then it moves to Phase 3, where the drugs are given to hundreds or thousands of patients, with researchers continuously testing safety and efficacy and comparing the results to a control group that was given a placebo.
Before a drug is approved by the FDA, it usually needs to pass two separate trials. According to a recent study of more than 7,000 clinical trials between 2006 and 2015, a new drug has less than 10 per cent chance of being approved by the FDA. Of the individual phases, Phase 2 proved to be the one with the lowest success rate, at 30 per cent. Phase 3, which accounts for 60 per cent of all clinical trial costs, has a success rate of 58 per cent. On average, a clinical trial process for a single drug lasts for 7.5 years and costs between $161 million and $2 billion.
Only 6 per cent of people who suffer from a chronic illness enrol in a clinical trial
There are currently almost 20,000 open clinical trials in the US, promising to deliver revolutionary new treatments for a wide variety of illnesses, including various types of cancer, HIV, and Alzheimer’s. However, most of them will fail to enrol enough people to conclude or even start the trial. Why is that? Why are people not signing up for trials that could potentially save their lives? In fact, only 6 per cent of people who suffer from a chronic illness enrol in a clinical trial, while more than 65 per cent of sites fail to meet their original patient enrolment, with nearly half of them enroling only one patient or none at all.
As it turns out, the main reason is that most people simply don’t know these trials even exist or that they’re eligible for them. Information about all current clinical drug trials is kept in a national registry that’s freely accessible to everyone. However, for an average person, the registry can be very difficult to make sense of, so most of them just give up before they find what they’re looking for. But that may be about to change, thanks to a company called Antidote.
Artificial intelligence helps patients find the right clinical trial
Antidote uses artificial intelligence to help patients find the right clinical trial for their condition. To make that possible, the company first gathered information about thousands of clinical trials from various sources, including ClinicalTrials.gov and the World Health Organization. This information is usually very disorganised, especially the eligibility criteria, so Antidote used clinical experts to structure it manually to make it easier to search. Then it created a platform called Match, which allows patients to look for a matching trial by entering their condition, location, age, and gender. Patients can then further narrow the search by entering how far they’re willing to travel or adding more details about their condition. Once they find matching trials, they register their email with Antidote, which forwards them information about trial organisers and explains what steps they need to take next to sign up. It will also keep them updated about any future trials that match their search.
The company also partnered with hundreds of health publishers and non-profit organisations, including the American Kidney Fund, the Lung Cancer Alliance, and the Muscular Dystrophy Association, which can now simply embed the Antidote widget into their websites to allow visitors easy access to the Match platform. To date, Antidote has gathered information about more than 55,000 clinical trials from nearly 170 countries. “It makes it less of a wild goose chase for patients”, says Esther Schorr, the COO of PatientPower, an online cancer news site and Antidote partner. “There’s just so much information for the common man or woman to get through. Technology can really make a patient’s journey easier”. Patients can use the service for free, while pharmaceutical companies and clinical research institutions need to pay a fee for limited access to the user database. While that does raise certain privacy concerns, it’s probably something most patients would gladly agree to, considering that these trials could potentially save their lives.
AI can help lower trial dropout rates, as well
Patient enrolment is the most time-consuming part of every clinical trial, taking up as much as 30 per cent of its timeline. It’s also responsible for nearly $6 billion of expenses annually. However, almost 80 per cent of all clinical trials don’t conclude on time, while 85 per cent of them fail to retain enough patients. And not every patient that enrols in a clinical trial will stay with it for the entire duration, with an average dropout rate of approximately 30 per cent. There are many reasons why patients choose to discontinue a trial, including simple logistics, fear, the invasiveness of the treatment and diagnostic procedures, as well as difficulties in adhering to the steps in the protocol. But artificial intelligence could offer a helping hand here as well, by sending patients reminders to take their medication, keeping them engaged, and monitoring their medication adherence.
AI can also improve the data collection aspect of clinical trials
One of the biggest issues with clinical trials is that they still rely heavily on subjective data supplied by the patients themselves to monitor their progress. While that information can be very valuable, it’s also extremely prone to inaccuracies that can jeopardise the success of the trial. AI-enabled IoT sensors and wearable technology would allow researchers to gather more accurate, objective, real-world information about patients’ heart rate, blood pressure, glucose levels, and sleep patterns, and monitor how the drugs impact their organism in real time. This would also reduce the risk of patients dropping out because it would eliminate the need to frequently travel to a trial site for a check-up.
For example, tech giant Apple is in the process of creating its own clinical study ecosystem that centres around the iPhone and Apple Watch. The company has also partnered with numerous health institutions and medical researchers to provide them with access to an unprecedented amount of real-time, user-generated health data. Furthermore, Apple released ResearchKit and CareKit, two open source frameworks that allow researchers and developers to create their own medical apps for the real-time monitoring of patient health. This software has already been used by more than 500 doctors and medical researchers for clinical studies encompassing more than 3 million participants.
Clinical trials are essential for the development of new drugs, but they’re also a very expensive and time-consuming process that’s ripe for disruption. Artificial intelligence has the potential to transform almost every stage of the clinical trial process, helping patients find the right trials for their conditions, improving patient enrolment and retention rates, and upgrading data collection methods.