Your Next Appointment Starts Before You Walk In

Better follow-up begins with understanding what changed between visits—not just what can be remembered once the visit starts.

A follow-up appointment often begins with a deceptively simple question:

“How have things been since I last saw you?”

The patient tries to compress several weeks of meals, symptoms, sleep, activity, medication experience, good days and difficult days into a few minutes. The clinician has the previous encounter, perhaps some labs or messages, and whatever information has made its way into the chart.

The problem is not that nothing happened between visits.

It is that much of what happened was never organized into a useful representation of what changed.

That time between appointments is where daily life happens. Appetite shifts. Sleep improves and deteriorates. Activity changes. Symptoms appear and resolve. Food tolerance changes. Medication context changes. A patient may follow a plan closely for several days and struggle with it during another week.

By the time the next appointment arrives, the care team may be looking backward through a narrow window.

The instinctive solution is often more data.

More wearable readings. More food logs. More questionnaires. More alerts. More dashboards.

But more data is not necessarily the missing layer.

The missing layer may be a structured representation of what appears to matter now—and how that changed over time.

From a stream of events to a picture of state

Consider someone receiving GLP-1 or metabolic care.

Over several weeks, hundreds of small events might occur: meals, workouts, nights of sleep, symptoms, changes in appetite, hydration patterns, wearable signals and everyday decisions.

A clinician should not need to reconstruct every one of them.

The more useful question is whether those events can be organized into a smaller number of meaningful changes.

Was gastrointestinal tolerance becoming more relevant during a particular period?

Was the patient consistently having difficulty meeting protein needs?

Did recovery context change as activity and sleep changed?

Did hydration become a greater priority?

If qualified glucose information is available, did there appear to be a reason to give more weight to glycemic-supportive choices?

At DrunR, we think of these as dimensions of state: structured need or watch priorities designed to summarize what may matter for the person at a given time.

They are not diagnoses. They are not treatment decisions. And they are not a substitute for clinical judgment.

The point is to turn fragmented daily evidence into a more useful picture before asking what guidance or what clinical conversation should come next.

That is what we mean by State First.

Instead of starting with, “What should this person do?”, the system first asks:

“What does the available evidence suggest may be happening with this person now, and how confident are we?”

That distinction matters between visits because state is not fixed.

A patient may be relatively steady for several days, experience a period of lower appetite and increased GI sensitivity, later return to normal tolerance, and then enter a period where recovery or protein needs become more relevant.

A static profile does not capture that movement.

A pile of raw data does not necessarily explain it either.

The useful abstraction is the changing state between the two.

One state, two different conversations

This also changes how we think about the relationship between patient support and provider follow-up.

During everyday life, a patient may need help answering a very practical question:

“What makes sense for me right now?”

That might involve a meal, hydration, recovery, or another daily decision.

At the next follow-up, the clinician has a different question:

“What changed since I last saw this person, and what is worth discussing?”

Those are different decision contexts, but they should not require two disconnected versions of the patient.

The same underlying state model can support both.

For the patient, state can provide context for a decision in the moment.

For the provider, the history of those states can help organize what happened between appointments: where patterns changed, what evidence was available, what remained unknown, and which periods may be worth discussing.

This is an important design principle for DrunR.

We do not want generative AI inventing a story from a collection of disconnected signals. The State Engine is intended to produce the structured representation first. Confidence, missing information and the logic behind that state remain part of the result.

Generative AI comes afterward, helping communicate that structured result differently depending on who needs it.

A patient may need a simple explanation.

A provider may need a traceable view of what changed and why.

One state. Two decision contexts.

That does not mean every daily signal should reach the clinician.

In fact, the goal should probably be the opposite.

A care team does not need another stream of noise. It needs compression: the ability to move from thousands of daily observations toward a smaller number of meaningful patterns and changes.

This is where between-visit intelligence becomes different from continuous monitoring.

The objective is not to turn every variation into an alert.

It is not to claim that every change should trigger an intervention.

And it is not to promise that technology will identify problems earlier or improve outcomes simply because more information exists.

The more disciplined goal is narrower:

Preserve enough structured context that the next conversation can begin from a better understanding of what happened.

Imagine a patient returning for a follow-up after several weeks.

Instead of beginning entirely from memory, the conversation might have context showing that food tolerance became more difficult for part of the period, that protein priority increased during another stretch, and that sleep and activity changed alongside recovery needs.

The clinician still decides what those patterns mean.

They still ask the questions.

They still apply medical judgment.

But they no longer have to begin with an empty space between the previous visit and the current one.

That is the opportunity we see in the invisible middle of care.

Healthcare is organized around encounters, but patients do not become static when the appointment ends. Their state continues to change.

Better follow-up does not necessarily require more alerts, more dashboards, or promises of earlier intervention.

It may simply require us to stop losing so much of the context that develops between one encounter and the next.

Because the next appointment does not really start when the patient walks through the door.

The story has been developing the entire time.

‍