How AI and humans working together can make lung cancer a less deadly disease

What you probably already know: Modern medicine is increasingly focused on prevention, and new artificial intelligence models are opening up ways to detect some of the world’s deadliest diseases. But there is still plenty of work to do. Lung cancer, for example, is far more deadly than other cancers because symptoms can be similar to other respiratory illnesses like bronchitis or pneumonia. Approximately 20% of all cancer deaths in the U.S. last year were from lung cancer, and it’s the second-deadliest cancer worldwide. Lung cancer doesn’t need to be as deadly, though. The problem is that lung cancer screenings aren’t being done on patients who need them. Gena Cook wants to change that. She has recently launched Kaliper Health to ensure no one dies of a preventable disease.

Why? Kaliper Health is starting with lung cancer and lung health because it’s a big problem and a big market, Cook said. “About 85% of patients who are at risk and are eligible for a low-dose CT scan aren’t getting it, and because of that, we’re catching lung cancer later when it’s not as treatable,” she said. Part of that is because simple lung cancer screenings have only been available for about a decade — unlike breast cancer and colon cancer screenings, which have been in place for many decades. About 80% of people at risk for breast cancer get screened, 65% for colon cancer. For lung cancer, only 15% of people at risk are getting screened. “We’re trying to make lung cancer screening as ubiquitous as a colonoscopy or a mammogram,” Cook said.

What it means: About 12.5 million people are at risk of developing lung cancer, and Cook’s company has combined AI with a human-centered approach to make sure those people get the screenings they need. It’s really an issue of managed care: First Kaliper Health uses a technology solution to identify high-risk patients, then patient navigators call them, tell them about their insurance options, schedule screenings, follow up with the primary care physician, then do that again 11 months later. Without systems like this, Cook said, many patients get lost in the complicated process. There’s also a stigma associated with smoking that can make patients embarrassed or ashamed to seek care. “A lot of patients are scared,” Cook said. “We have lots of examples of patients that have canceled three times.” But her team doesn’t give up. And they’re saving lives.

What happens now? For Kaliper Health, lung cancer is just the beginning. This kind of tech-enabled service with a strong human component could be applied to many diseases. And as the system goes from thousands to hundreds of thousands of patients, the AI models can kick in to help track who is most at risk of developing disease. “Because the navigators work for Kaliper Health are the ones using the service, they know where patients are getting stuck,” Cook said. “We know where the technology isn’t working and we can automate around that process.” That combination of humans and technology working together to problem solve on behalf of patients is truly the future of health care.