Photo: Tom Stewart/Getty Images

ScientificAmerican.com

In 2018, the American Cancer Society released an alarming report noting that 20 percent of clinical trials fail because of insufficient patient enrollment and that 56 percent of patients seeking care will not have a clinical trial in their immediate area and another 17 percent will not meet the eligibility requirements. A year later, the Improving Patient Access to Cancer Clinical Trials (IMPACT) study noted that fewer than 5 percent of adult patients participate in clinical trials nationally.

Unfortunately, neither report led to significant change in the system.

But the COVID-19 pandemic has forced our system to start addressing significant flaws in the current clinical trial infrastructure and find ways to evolve to make sure the “process” no longer interferes with discovery and helps patients when treatments may be working.

First, we need to remove archaic geographic barriers that prevent patient participation in trials. After COVID-19 struck, we had several patients concerned about their ability to travel to our trial site to receive care. One patient—Hollie—is participating in a breast cancer trial, and lives across the country on the West Coast. Instead of discontinuing her trial, we were able to transition her study tests to a local hospital and conduct her regularly scheduled visits via telemedicine.

Fortunately, at the national level, the Food and Drug Association (FDA) noticed this flaw and essentially changed its rules on the fly by developing a comprehensive guide to assist clinical trial sites with guidance on how to protect patient safety while maintaining trial integrity. The guidance provides flexibility for the trial and states it will not disqualify the patient from the trial if the treatment is demonstrating positive results. This is such a refreshing approach for the medical community because, under normal circumstances, Hollie might have been disqualified from her trial if she was no longer able to travel. We need to make this common-sense provision permanent.

A second step is removing insurance barriers that prevent participation in trials. Prior to COVID-19, we had a clinical trial patient from Oklahoma who was forced to travel to our Chicago hospital because her insurance company did not recognize our Tulsa hospital as an in-network provider. When COVID-19 struck, our team tried to convince the provider to change its policy and allow this patient to resume treatment in Tulsa instead of making her drive nearly 12 hours to our Chicago facility. We were successful; however, a patient should never have to endure this type of stress, even under normal circumstances. Patients who are out of traditional options should not have to jump through hoops for insurance coverage when they finally find a potentially lifesaving clinical trial.

A third (but certainly not final) step is improving how we connect patients with clinical trials. Historically speaking, clinical trials place a massive amount of pressure on physicians to identify qualified patients. The process can be slow and inefficient and may result in more misses than hits when it comes to finding the right patients for a study. In the case of Leonard—a retired firefighter who spent his life running into burning buildings—his primary physician gave him six months to live. He decided to seek a second opinion and that physician, Dr. Bonilla, found a clinical trial that aligned with Leonard’s condition. Unfortunately, there are too many stories where patients like Leonard never seek that second opinion, which begs the question, how can we remove “chance” from the equation and become more efficient in helping patients find the right trials?

Enter artificial intelligence (AI). A published study in BMC Medical Informatics and Decision Making concluded that using an artificial intelligence algorithm can “increase trial screening efficiency of oncologists and enable participation of small practices, which are often left out from trial enrollment.” So, AI can improve the patient identification timeline by nearly 90 percent, and virtually eliminates human disruption and potential misses caused by pandemics or other variables. The time has come to lean into AI to ensure we are getting the right patients enrolled quickly and more effectively than the archaic, tedious, and inefficient ways of the past.