From Documentation to Action: What EMRs Were Never Designed to Do

Most EMRs were built to answer a record-keeping question: what happened?

The more interesting question is what followed from that original design decision—and what primary care systems might look like if they were built around unresolved patient work instead of completed encounters.

That question mattered. It still does. A clinic needs a reliable chart. A physician needs a place to document the encounter, record the assessment, order the test, send the referral, bill appropriately, and preserve the clinical story over time. Compared with paper charts, digital records solved real problems. They made information easier to find, easier to share, and harder to lose in the literal sense.

But primary care has never been only a record of what happened.

It is also a living collection of things that still need to happen.

That is where the design assumption starts to show its age. The appointment ends, the note is signed, the referral is sent, the result is reviewed, and the chart looks, in many ways, complete. Yet the actual patient work may still be moving through the clinic in a less visible form. A referral may be sitting in a fax queue. A consult may be expected but not returned. A patient may need to repeat bloodwork in six weeks. A result may have been seen, but the next step depends on someone remembering to check whether the patient booked the ultrasound.

The system can show that the work was documented. It may not show whether the work was truly carried through.

The Encounter Became the Centre of Gravity

Most EMRs organize clinical life around the encounter. That is understandable. The encounter is where the medical record has traditionally been created. It is where the physician assesses, decides, documents, orders, and bills. The software reflects that history: the visit note is the container, the chart is the archive, and much of the system’s logic is built around capturing information at a point in time.

The trouble is that family medicine does not move in neat points of time.

A single visit can create a tail of work that stretches for weeks or months. The patient with abdominal pain leaves with bloodwork, imaging, a referral, a safety-netting plan, and a promise that someone will call if anything concerning comes back. None of that fits cleanly inside the visit once the door closes. The clinical responsibility continues, but the system’s centre of gravity often stays behind, attached to the encounter that generated the work.

That mismatch is easy to miss because the chart still looks busy. There are entries, tasks, messages, scanned documents, inbox items, timestamps, initials, and audit trails. Activity is visible. Completion is less obvious.

A referral can be marked as sent without the clinic knowing whether the receiving office accepted it. A lab result can be reviewed at 6:47 p.m. between two other inbox items, with the physician thinking, “I should recheck that after the repeat test,” while the actual reminder lives nowhere durable. A staff member can create a task that is technically present in the EMR but functionally buried three clicks deep under a patient-specific communication tab no one opens unless they are already looking for it.

Nothing is necessarily broken in the dramatic sense. The system is doing what it was built to do. It is storing the evidence that work occurred.

But primary care also needs help managing the work that remains unresolved.

The Quiet Gap Between Recorded and Resolved

Clinics often feel this gap before they name it.

It shows up when a physician keeps a small list of patients in the margin of a day sheet because they do not fully trust that the EMR will bring those loose ends back at the right time. It shows up when someone prints a requisition not because paper is better, but because paper sitting beside the keyboard has a kind of physical insistence that a closed digital task does not. It shows up when a doctor remembers a pending MRI while brushing their teeth, then sends themselves an email because the thought arrived outside the system that was supposed to hold it.

These are not signs of disorganization. They are signs of clinicians compensating for a design gap.

The gap is between recorded work and resolved work. The record can say that a referral was sent. Resolution means the referral was received, triaged, booked, attended, and returned with a consult note that someone reviewed and acted on. The record can say that a result was seen. Resolution means the necessary next step was assigned, completed, and not left suspended in one person’s memory. The record can say that a task exists. Resolution means the task stayed visible until it was actually done.

This distinction matters because documentation can create a false sense of closure. Once something is charted, signed, scanned, or initialed, the system may treat it as handled. In a legal or informational sense, it may be. In an operational sense, it may still be open.

Family physicians live inside that difference. They know that a clean chart does not always mean a closed loop.

Better Notes Do Not Automatically Create Better Follow-Through

A great deal of technology discussion still treats documentation as the main bottleneck in primary care. That is not wrong, exactly. Documentation is heavy. Notes take time. Inboxes spill over. Any tool that makes charting more accurate and less draining can be useful.

But there is a limit to what better documentation can solve.

A more polished note does not tell the clinic whether the cardiology referral stalled. A faster dictation tool does not confirm that the patient completed the repeat creatinine. A better template may capture the plan beautifully while leaving the plan’s execution dependent on a physician’s memory, a medical office assistant’s vigilance, and whatever informal tracking habits the clinic has built over the years.

This is the part of the conversation that often gets flattened. The problem is not simply that doctors need less typing. It is that clinics need more reliable ways to carry patient work forward after the typing is done.

When systems focus mostly on capturing the encounter, they can make the record more complete without making the work more secure. That is why a clinic can become highly documented and still feel fragile. Everyone can be working hard, the EMR can be full of information, and the physician can still carry the background worry that something important has slipped out of view.

This is part of the problem Aeon keeps returning to: not how to help clinics produce more documentation, but how to make unresolved patient work stay visible long enough for someone to trust that it will not quietly disappear.

The Work After the Visit Is Not Secondary

One reason this issue persists is that post-visit work is often treated as an administrative shadow of the “real” clinical encounter. The appointment is seen as the main event. Everything after it becomes follow-up, paperwork, coordination, inbox management, or task handling.

But in primary care, the work after the visit is often where care either holds together or comes apart.

A referral is not meaningful simply because it was generated. A diagnostic plan is not complete simply because the requisition was printed. Reviewing a result is not the same as making sure the patient understood the next step. Continuity of care depends on the connective tissue between decisions, actions, confirmations, and returns of information.

That connective tissue is exactly where many EMRs are weakest.

They may contain the ingredients of follow-through, but they do not always provide a reliable operational view of what is still open. The clinic can see many pieces of information, but not necessarily the state of the work. Is this waiting on the specialist? Waiting on the patient? Waiting on staff? Waiting on the physician? Waiting on a result that should have arrived by now? Too often, answering those questions requires local knowledge, memory, or detective work.

And because physicians ultimately remain responsible, the uncertainty becomes mental load.

Not the loud kind. The low-grade kind that sits underneath the day. The kind that makes a doctor pause before leaving the clinic because there was one thing they meant to check, though they cannot immediately remember which patient it belonged to. The kind that turns the EMR from a trusted system into a place where information lives, but not always where accountability lives.

What Modern Clinical Systems Need to Carry

The next meaningful shift in EMR design is not just from paper to digital, or from typing to dictation, or from manual work to automation. It is from documentation as the centre of the system to patient work as something the system actively helps manage over time.

That does not mean every process should become rigid. Primary care is too human, too variable, and too context-dependent for that. Good systems should not pretend clinical judgment can be reduced to a checklist, and they should not flood physicians with alerts for every possible loose end. A system that creates more noise simply gives doctors a new form of work to manage.

The better question is quieter and more practical: what should the system keep visible until it is actually resolved?

A modern clinical system should help a clinic understand which loops remain open, who owns the next step, what is waiting on an outside party, and what has gone quiet for too long. It should make unresolved patient work harder to lose, without making physicians feel like they are supervising yet another inbox. It should support the real shape of Canadian primary care, where responsibility often extends across fragmented services, faxed referrals, delayed consults, provincial workflows, and limited clinic capacity.

This is not a feature argument so much as a design philosophy. If the system is only organized around what has been documented, then the physician remains the backstop for what has not yet been completed. If the system is organized around follow-through, the burden starts to shift. The work becomes more visible. The clinic can trust the process a little more. The physician does not have to hold quite as much in their head.

From a Place to Store Care to a System That Supports It

The original promise of the EMR was that the patient record would be digital, legible, accessible, and complete. That was a necessary step forward.

But it is no longer enough.

A chart can be complete and still leave the clinic uncertain. A note can be excellent and still leave the next step exposed. A task can exist and still fail to function as a reliable commitment to completion.

The deeper question is whether clinical systems can evolve from places that store information about care into systems that help care move forward.

That shift is subtle, but it changes everything. It asks less of the physician’s memory and more of the system’s design. It treats follow-through not as an afterthought around the chart, but as part of the core work of primary care. It recognizes that the visit may be where a decision is made, but the loop is only closed when the necessary work is actually done.

In practice, that means something simple: a physician locking the clinic at the end of the day without needing to keep a mental tab open for the referral that might never come back, or the repeat potassium they meant to recheck next Thursday if nobody else noticed first.