Ordering Tylenol in Epic is a whole operation. I have to log onto Epic, open the patient’s chart, find the order, select the dose and route, and work through the remaining screens before I can move on.
Then there’s the time I spend searching the chart itself. As a pulmonology fellow, I might need to find a patient’s last pulmonary function tests, chest CT, prior medications, and the plan for a lung nodule. The information is usually there… somewhere in the chart. But finding it takes work.
How did we get here? Where might we go next? My prediction is that the EHR will remain the central place we store patient information, while specialty-specific platforms increasingly become the places where physicians work.
How the EHR Became Our System of Work
The shift to electronic records brought meaningful benefits. Information that once lived in paper charts became easier to search, share, and aggregate. Electronic ordering and clinical decision support also created opportunities to build safeguards into care.
Federal policy helped accelerate the transition. The 2009 HITECH Act supported health IT adoption, and programs beginning in 2011 offered incentives for using certified EHRs. Certain eligible clinicians later faced Medicare payment adjustments if they didn’t demonstrate meaningful use.
By 2024, 95% of office-based physicians reported using an EHR.
At a basic level, the EHR is a system of record: a place to store notes, results, medications, and other patient information.
But it has also become our system of work. We use it to write notes, order tests and medications, message patients, review results, and manage our inboxes. Much of our day happens inside the chart.
The challenge is that different specialties need different information at different moments. A cardiologist, an oncologist, and a pulmonologist may all use the same EHR, but the information each needs at the forefront of a visit can be very different.
What a Specialty-Specific System of Work Could Look Like
Imagine a platform that sits on top of the EHR and organizes the chart around the work a pulmonologist needs to do.
Before I see a patient, I could open a view that gives me a concise pulmonary history: prior visits, diagnoses, medications tried, the latest PFTs, and recent chest CT findings. If there’s a nodule that needs follow-up, I’d see the relevant imaging and prior plan.
For a patient undergoing lung transplant evaluation, the same view could show the workup completed so far, outstanding items, and the specialists involved. If pulmonary hypertension or heart failure is relevant, I could quickly find the last right heart catheterization or cardiology assessment.
That would save me from hunting through separate parts of the chart to assemble the story myself. And, I swear to you, the “hunting” part of pre-charting is cognitively exhausting.
I’ve written before about wanting a clearer, condition-by-condition picture of a patient’s history. This takes that idea a step further: organize the information around the specialty’s workflow, then help the physician act on it.
For example, the platform could flag that a patient’s PFTs were completed more than a year ago or suggest a follow-up chest CT based on the documented plan. It could pend an order for me to review, modify, or dismiss.
The platform could also help draft my note, including through an ambient scribe, and send the final documentation back to the EHR.
Cardiology would have its own view, with echocardiograms, catheterizations, and stress tests more prominent. Oncology might prioritize the treatment timeline, chemotherapy cycles, adverse reactions, and trial history.
The underlying record would still connect the care team. The workspace could change depending on who’s using it.
Who Will Build It?
I think companies like OpenEvidence and Doximity could be well positioned to build toward this future. Physicians already use their tools for medical questions, and that gives them a potential starting point for building products that fit into clinical workflows.
But answering a medical question and managing a patient’s longitudinal care are different jobs. To become a system of work, a platform would need reliable access to the right patient information and a way to send notes and orders back into the EHR. It would also need to show where its information came from so clinicians can verify it.
And an AI-generated summary is only as useful as the information it captures. If it misses a prior adverse reaction, pulls in an outdated plan, or presents an uncertain detail as fact, it could make care harder rather than easier.
Epic and Oracle also have an advantage: they already sit inside health systems and are building AI tools of their own. As I wrote in my breakdown of Epic’s AI features, specialty depth may be one area where other companies can differentiate—but they still need to deliver a workflow that physicians trust and can use without adding another layer of friction.
I don’t know which companies will get there first, or whether the best specialty-specific workspace will come from an EHR vendor or a company layered on top. I do think the opportunity is clear, though.
In summary, we’ve spent years digitizing the medical record. The next step is making that information easier to use in the context of the care we’re providing. If specialty-specific tools can do that reliably, physicians could spend less time searching the chart and more time understanding and caring for the patient.
Wanted to voice your thoughts on healthcare? Explore how to submit an Op-ed to Healthcare Huddle here.
