Introduction
Talk to any healthcare provider, administrator, or healthcare worker, and you’ll confirm that the flow of healthcare documents contributes significantly to the crushing burden of healthcare administration in their practice. While the promise of interoperability tells of a future where all healthcare information is well-structured and flows smoothly from system to system, it has been promised for decades, and healthcare workers are losing hope. The rise of intelligent document processing healthcare solutions reflects how teams are bridging the gap between that vision and today’s realities. The reality on the ground tells us that unstructured and semi-structured documents continue to slow down important healthcare workflows, stealing precious time that clinicians could use to treat patients.
Enter Intelligent Document Processing, or IDP. Much like how you can throw any image or document at ChatGPT and ask it for information about that material, we can use AI and machine learning to create structured data from the unstructured documents that clog up your offices. We can speed up the manual processing that plagues every medical office, large or small. This need is exactly what intelligent document processing healthcare solutions are built to address.
The question remains: How should IDP manifest itself within our healthcare systems? That’s where a framework like Jobs to Be Done (JTBD) helps focus our work as a digital healthcare product company.
Table of Contents
What is JTBD?
The JTBD framework is a twist on the user-centric approach to designing products. JTBD analysis helps organizations innovate by identifying the core needs a product fulfills, rather than everyone’s impression of what would work. Software product teams start by asking themselves, “What is our customer ‘hiring’ our software to do for them?”
Hint: OCR, Classification, and Data Extraction may be necessary features, but they do not illustrate the core need fulfilled by intelligent document processing in healthcare.
A semi-flawed example of JTBD in action is the apocryphal statement attributed to Henry Ford, “If I asked my customers what they wanted, they would have said a faster horse.” While there is no evidence that Ford said this, it illustrates the point that focusing solely on what customers ask for provides only one view of the situation. Creators of disruptive technology often zoom out from the immediate problems at hand.
Clayton Christensen, one of the originators of JTBD, was the first to suggest thinking of a product as something you are hiring to do a job for you. His classic examples are:
- You are “hiring” a drill to make a hole in a wall to fulfill your core need of hanging a picture. In other words, if you created some other way to hang the picture, the drill is no longer needed.
- Christensen discovered that McDonald’s customers aren’t always “hiring” milkshakes to be chocolate and delicious, they’re hiring them to be filling and keep them entertained during long commutes. In other words, if you created a product to fill and entertain commuters, you can disrupt the milkshake industry (you have to watch the video to understand this one, but Christensen makes a lot of sense).
I’ve often used disruptive digital consumer products like rideshare apps to illustrate the same point. Imagine it’s 2007 and you want to disrupt the taxi industry. You may choose to survey taxi riders and drivers and ask them about their experience and how it could be better. You would get a whole host of very popular suggestions around improving safety, cleanliness, navigation, and payment methods – none of which scream “Let’s build a ride sharing app!” The fundamental jobs-to-be-done in ridesharing serve two users and can be summed up as:
- Riders need to get from point A to point B conveniently
- People need to generate income from their vehicles
It’s important to understand that conducting JTBD analysis does not mean you ignore end users. To do JTBD analysis correctly, you zoom out from their current experiences and ask the right questions. Applying this framework to healthcare information, we can break down the primary “jobs” into specific tasks businesses seek to complete through document automation.
The JTBD of Intelligent Document Processing in Healthcare
Documo’s product team has spent the past two years asking healthcare practitioners, administrators, and staff about healthcare documents. We’ve heard them vent and watched them demonstrate a mishmosh of workarounds to get critical healthcare information from point A to point B.
When I first started at Documo, I asked my primary care physician, “…tell me how you feel about faxes…” She rolled her eyes, turned her EHR terminal around, and said, “This is going to take a while, I hope you have a few minutes…”
“I’m on the clock, I have all day,” and away she went. She showed me her view into my patient record (no HIPAA violations there), including a dizzying array of documents sent to her via fax, paper, and secure electronic messages. I’ll spare you the minutiae, but I was impressed with the volume of information my fairly healthy body generated over the years she has been my provider. When she was done, I asked her one question – “What do you do with all of these documents?”
Her response was very telling. I’m paraphrasing here, but two points resonated with me: “I need to see your documents when they arrive to make sure I know what’s going on with you. After that, they’re a reference to your history if and when I need them.” These kinds of real-world needs highlight where intelligent document processing healthcare tools can help streamline workflows and reduce admin friction.
Both of those statements capture the JTBD of healthcare documents in electronic medical records systems. Practitioners need to be informed about their patients, and they need an enduring reference to their patients’ histories. But, as we saw from the rideshare example, there are usually more constituents in a system than the obvious user. When we speak with all the people who handle healthcare documents and remove the notion of “processing documents” from their minds, they express their needs very clearly. Much like the JTBD of rideshare provides a simplistic view and says nothing about taxicabs, their needs say very little about documents:
Healthcare document workers: “I need to get healthcare information to the right people.”
Healthcare administrators: “I need to make appointments, process referrals, authorize treatments, and send bills.”
Lab workers: “I need to run tests and communicate results.”
Payers: “I need to review cases and approve treatments.”
Healthcare information technologists: “I need to keep systems running smoothly and compliantly.”
All of these folks (and more) are members of a team whose shared goal can be paraphrased from the Hippocratic oath:
I will help the sick, and I will abstain from all intentional wrongdoing and harm
The oath is too broad to be a JTBD for healthcare documents, it is a reminder that our products are there to help the sick without doing harm. It’s an important distinction because one of the greatest harms our systems can do is distract the clinician’s focus away from their patients as they deal with the crushing burden of healthcare administration.
While each user’s JTBD says very little about intelligent document processing healthcare solutions, there is a huge opportunity for them to “hire” these platforms to achieve those goals. Much like the drill is just a tool to hang that picture on the wall, IDP is a tool to help them get relevant healthcare information into the hands of clinicians and staff so they can focus on the care of their patients.
In an ideal world, they won’t even know what we’re using to achieve that goal for them. While we’re building an intelligent document processing healthcare solution at Documo, my goal is that our users never know or care how we’re helping them move healthcare information around. They’ll just know it’s happening so they can focus on the tasks that require their expertise and improve their patients’ lives. That’s the goal of intelligent document processing healthcare solutions – removing barriers so care comes first.
