Many cancer patients still carry stacks of paper records, and CDs with imaging between visits. Although companies are working to solve this problem, several challenges persist.
Anil Sethi learned this firsthand after his younger sister, Tanya, was diagnosed with stage IV breast cancer.
He created a previous startup, Gliimpse, with the goal of helping her after she was first diagnosed. The company was later acquired by Apple, where Sethi worked for two years as director of health records, until he got a call from Johns Hopkins saying Tanya would have just two more weeks to live.
He left and drove across the country with her as they sought out specialists who could help.
“We bought Tanya five more months of life as we criss-crossed the country trying to look for a hail mary,” he said.
When she died, Sethi promised he’d spend the rest of his career working to ensure people would no longer think of breast cancer as a death sentence. He founded his current startup, Ciitizen, in 2017 with this goal.
If his previous work with Apple was to make it easier for lots of people to bring in health records from different facilities, Ciitizen’s current approach is depth. Although the company is not exclusively for cancer patients, Sethi is focused on making the platform work for them first, given the vast amount of information they have to manage.
Many health records efforts are focused on pulling information using Fast Healthcare Interoperability Resources (FHIR), an interoperability standard that lets them pull certain healthcare information from health records systems, with patients’ consent. But that standard doesn’t currently cover everything, including pathology reports, information on a tumor’s size and grade, and discharge summaries, which Tanya needed.
“An API can’t pull out of a database… that which it doesn’t have in the database in the first place,” he said.
Because many of the built in EMR databases are still centered around billing, this information simply wasn’t there.
Instead, his solution is to use the patient’s right of access to pull all of their records, documents and genomics reports. Its users can tell the company where they received care so Ciitizen can follow the “breadcrumbs.”
“When Tanya was at Hopkins, they use Epic, so I used her login, and I got 30 or 40, maybe 50 records, a breakdown, no issue” he said. “But when I used her HIPAA right of access to put in a form, signed it, sent it to medical records, we got a dump of 2,200 pages. So we’re looking at two orders of magnitude different.”
Of course, this comes with a tradeoff. While information pulled in through an API is computable, those 2,200 pages that he received were all sent in a big, messy PDF.
Because of this, Ciitizen is training machine learning tools to turn this into standardized, usable information. It’s focused on fetching and codifying the 20 most important elements that oncologists would need.
The service is free to patients. It has a feature to help them identify clinical trials that they might be eligible for, and allows them to share their data with researchers, if they choose. Users can download their records or request for them to be deleted.
“I think we’ve all had a Tanya in our lives,” Sethi said. “You’ve got to know why you’re doing this work. This is really hard work.”