Guest Column | June 28, 2019

Efficient Cancer Care Through Informatics: 3 Ways The Process Is Broken And A Path Forward

By Satish Sanan, Inspirata

Clinical Informatics Fellowship

In 2012, within a relatively short period of time, my wife and I lost three close family members to different types of cancer. Back then, I was spending most of my energy on a horse-racing business with two big farms and 350 horses. But after my sister-in-law died from cancer at 40, at the height of our grieving period, my wife asked me why I was wasting all my time and money on horses when I could be doing something about this. She knows I like complex problems. My first thought was – I’m not a physician, how am I going to impact cancer care? But the question nagged at me.

Eventually, I sold my farm in Kentucky and all my animals and resigned from the 21 company boards I was serving on. I decided I was going to see if there was something I could do to have a positive impact on cancer patients. As a serial entrepreneur in the tech space, I wanted to find an industry-disrupting technology, so I set out looking for it at every bioinformatics and oncology conference in the country. I ended up meeting and partnering with Inspirata’s founder, Dr. Mark Lloyd, a cancer researcher who at that time was working at the Moffitt Cancer Center.

At that time, I knew very little about the process of cancer diagnosis, but I knew that for patients and their families, it needed to happen as quickly, efficiently, and accurately as possible. We thought that by creating a database of cancer patients with the experiences of past patients, we could facilitate enhanced decision-making opportunities and create disruptive artificial intelligence algorithms to provide doctors with new insights. I asked Mark, “How many patients do you have now?” and he said, “About 9,000.” But I wanted to know – why not nine million?

That question became my north star as I started traveling the world to learn more about the cancer care continuum, starting with how it’s diagnosed and following its many possible paths along each patient’s experience. It took me a long time to really learn the business. I’m still learning it, but throughout my travels and my research, there are three issues that make it apparent how broken the cancer care continuum is and how informatics can help fix it:

  1. Patients are rarely fully engaged in their care. The patient journey starts with diagnosis but can end anywhere. They’re scared, they’re confused. They may not be truly engaged in the process of their care. Do they have the information they need and are they educated enough about it? Do they have access to the right physicians and clinicians? Those two questions are what we’re ultimately trying to solve for. We have to be able to make clinical and other information available to the right stakeholders to make the right diagnostic and financial decisions. You have to make sure the patient and the family are engaged in the decision-making process. It’s all about quality of life.
  2. Data is buried in reports and mostly unstructured. Up to 70 percent of data in healthcare is unstructured – essentially in a format where it can’t be easily analyzed. This presents a huge missed opportunity to create insights for patients, and to leverage individuals’ data at a higher level to identify trends in cancer and therapy types – information that can potentially save lives. The first challenge is making this unstructured data usable, and then, finding a way to develop insights that are actionable and meaningful. Right now, artificial intelligence (AI) and natural language processing (NLP) are two methods we’re using to surface insights from unstructured data, but it’s a huge industry-wide problem. You can’t put information in the hands of people who need it – when they need it – if it’s not in a searchable, accessible, understandable format.
  3. In the U.S. especially, cancer data is siloed by devices and systems that don’t talk to each other. Today people use smart phones to take billions of photos. Those photos can be shared, viewed, and managed across hundreds of different devices and platforms – it is an open system. In the U.S., we have difficulty tracking patient information from one hospital to another, and sometimes even struggle to move data within a single hospital system. Medical data simply isn’t portable enough. An open and interconnected approach is the only way we’ll be able to make cancer data more available and useable. To advance, the industry will need to become more interoperable. Harnessing information together as big data will better serve everyone by determining critical population trends in diagnosis and therapeutic effectiveness.

With the progress the industry has made with digital pathology, AI, NLP, and informatics, there’s reason to hope that, in five to 10 years, we will get to the point where data is a powerful tool placed directly in the hands of patients. Until then, we’ll continue pushing to get that information into a precise form, into the hands of clinicians so they can make the right diagnosis and design the right therapy. By making data accessible and actionable, we can help clinicians and their patients make every moment matter.

About The Author

Satish Sanan is Chairman and CEO of Inspirata, a global digital pathology company that pioneered the use of digital imaging for accurate cancer diagnoses, and now offers cancer registry abstraction and insight solutions to drive better cancer diagnosis, treatment, performance, and research.