“When you have a chronic condition—especially one that involves hospitalizations—you wind up spending a lot of time with healthcare providers,” says PicnicHealth’s Director of Data and Informatics, Gillian Hanson, “and they wind up being your mentors and the people you look up to.”
That’s one of the reasons why Gillian, who is also a naturopathic doctor by training, is such a good fit with PicnicHealth.
“I have a lot of experience from the patient side,” she says, “because I have a bleeding disorder. It’s very similar to hemophilia. It’s called Von Willebrand disease. Yes, it’s a terrible name!”
When she came on board to help create PicnicHealth, Gillian had real empathy for what patients go through when trying to obtain past medical records on their own. From personal experience, she completely understood that there had to be a better way.
“The last thing you want to do at that time in your life is go through this rigmarole and to track down your records,” she says.
The goal then, was to make the problem of searching for medical information go away. Once that’s achieved, patients can use their time more wisely, to focus instead on their health and their families. The PicnicHealth objective was to give patients a bird’s eye view of their entire history—to make it easy for medical records to be stored, accurate and easy to review.
A tall order, easier said than done.
“When you get sick,” Gillian says, “you don’t want to spend your free time on the phone with a medical records office, fighting to get copies of your medical records. And it really is a fight. It’s a struggle. There are different requirements and hoops you have to jump through in order to get your records. Some require payment. Some require a copy of your driver’s license or other forms of identification. Sometimes it gets lost in the mail or it gets lost in their system, and you have to start all over again.”
Fortunately, as PicnicHealth was becoming a reality, the digital age was helping the medical records mess to untangle a bit—but just a bit. The process of gathering medical records still wasn’t perfected.
By 2014, website portals were becoming more present in physician offices. The zeitgeist in the industry was to allow patients more access to their medical records through these portals. However, the information available was usually limited to lab procedures or medication lists from a patient’s recent visit. The big medical history picture was still not a reality.
“As we started the initial process of creating PicnicHealth, we were just pulling in information from portals,” Gillian says, “but what we realized really quickly was that only having a small subsection of information is not actually that useful when you are using your records to go to new providers or a new hospital, or to try to get an appointment with a specialist. The only way to do this is to have access to your entire record set. And the only way to get those is to actually request your records from the medical records office. The portal was just not sufficient.”
Thus began another mountain to climb: requesting records, one by one, office by office. Gillian certainly had a lot of experience doing this in her private practice, but not at a scale this big. It took a few years and a lot of contact with healthcare facilities and record offices, but the process was at last built out.
The next step was to standardize requests for medical records so that it would be a consistently easier process to acquire them.
Of course, this was a monumental task, but it was still only half the job. The next phase was to take the records and organize them into data sets. Having a stack of thousand pages of medical records will do nobody any good; searching for specific patient information would be like trying to find a needle in a haystack.
“What we started doing—and still do—is divide up this information into individual healthcare encounters,” Gillian says. “It’s what we call ‘structuring’ the information. So specific pieces of data about each visit can be displayed on a timeline for users so they can sort and filter the info in a more user-friendly way.”
What was needed was a training data set, in which the gathered information was split into different categories for easy access (for instance: doctor visits, providers, facilities, medications, vital signs, lab reports). Then the machine learning takes over, forming an algorithm that can predict where all kinds of information will show up in those stacks of records.
Sounds tiresome? Luckily, Gillian is a self-professed data nerd. She was clearly in her element.
“This is the kind of thing that I love doing,” she says of collecting and separating data. “It's kind of like reading your favorite book over and over and never getting tired of it.”
She also kept her eyes on the prize.
“It’s not just building a machine,” she says. “It’s genuinely impacting people’s lives.”
When it comes to using this data for research purposes, there was one constant that was non-negotiable: patient confidentiality and privacy. The extracted information never contains any identifying information. Each piece of data is assigned a code and stored separately from the original patient records.
As a result, a patient cannot be identified and associated with the information being studied. This allows the information to be combined together with the de-identified data of other patients. All of it is used securely and privately, for research purposes only. And not traced back to the patient.
“We don’t want to do anything secretively,” Gillian says. “And we want—and always wanted—the patient to be at the center of this experience. We started this company for patients to alleviate a problem that they were experiencing. The patient has to remain at the center.”
An unexpected result of these patient-centered goals: building communities.
Cohorts of people from different disease communities began signing up for PicnicHealth. This helped build the record collection, which in turned helped the research. As a result, what can be seen are some consistent themes among patient experiences, as well as some interesting differences.
Some people had as many of fifty years of medical records on their timeline, allowing Gillian to note how the disease progressed throughout a lifetime. This includes all the different drugs, treatments, and therapies that went along with the experience (and when). This “longitudinal health journey” allowed for a longer-term view of the disease and a better understanding of how it was handled. As a result, patterns emerged.
“People had this thirst for research to better hear their story,” Gillian says. “There now seemed to be a safe and secure way for research [to be gathered and studied], so that people could better understand their disease and their journey.”
It was like walking into a dark room and turning on a light.
For Gillian, her PicnicHealth experience has brought her around to her own beginnings, with her first diagnosis.
“We’re currently doing a lot of work with hemophilia, I’m excited to say, so it’s coming back full circle,” she says. “I’m able to give back to the community that has given so much to me, and who made me feel that I wasn’t alone, that it was possible to manage this disease. People are able to live full and rich lives by having [medical records] be something in the background as opposed to—when you’re diagnosed—front and center. It’s a daily experience, so these communities have been really helpful for me to help me navigate that journey of diagnosis. Now it’s just another condition that lives in the background for me.”
Another unexpected benefit from the gathering of communities:
“It’s a chance for more voices to be heard,” Gillian says. “Most people in my experience do want to participate in research. They do want to make sure that their voices and story are heard and yet they may not be able to participate in a research trial. This is a way for them to contribute and make sure that their story, their disease journey, is heard without putting such great demands on them.”
Visit PicnicHealth.com to sign up for a free PicnicHealth account if you too have hemophilia and would like to contribute to research.
