Anatomy of a Progress Note

What's Actually in a Progress Note

I took a typical acute care progress note and mapped every line back to its source — every symptom, every sign, every lab interpretation, every vital sign trend, every piece of analysis. What I found surprised even me: less than 5% of the content came from direct conversation with the patient. Ambient AI scribes are built around a specific model: a physician walks into a quiet room, has a fifteen-minute conversation, and walks out with a note nearly finished. That model assumes the conversation is the note. In acute care, it almost never is.

Where the Rest of the Note Comes From

The other 95% comes from everywhere else. Lab results and how you read them, imaging reads, vital-sign trends over hours or days, a patient's response to therapy. A call with family. Overnight sign-out from nursing, a call to the nursing home about baseline function, a word with the outpatient doctor who knows the history.

And it goes beyond patient data. A CDI query you've answered. The medical necessity documentation a bed status decision requires. None of that is a conversation with the patient either.

The narrative, the assessment, the plan: all of it is synthesized from a dozen sources like these, most of which have nothing to do with a conversation at the bedside.

The Acute Care Reality

The bedside conversation itself is also harder to lean on in acute care. Our patients are often at their most vulnerable: hearing deficits, the breathlessness during acute heart failure, the confusion of sepsis, the physical limits of advanced age. Almost by definition, these aren't patients at their communicative best. That's part of why they're admitted.

The environment doesn't cooperate either. Acute care spaces aren't quiet exam rooms. Care happens in shared rooms, in hallways during a capacity surge, in procedure bays. That's rarely the controlled acoustic environment ambient tools need to work well.

The "quiet room, one conversation" model ambient scribes are built for just doesn't describe most of what happens on an acute care floor.

Built for How Acute Care Actually Works

This is why we built SutureNote differently. We didn't assume that capturing conversation would solve documentation burden. We started from where the note actually comes from. Upload the prior notes, labs, and records already scattered across the chart to give the note its skeleton. Then dictate your own clinical reasoning on top of it as it happens: a lab interpretation, a response to therapy, a family update. You're building the note the same way you actually built the day — not recording a conversation and hoping it adds up to the whole picture afterward.

That's the gap ambient scribes were never built to close, and it's the actual target: the labs, the sign-out, the CDI query, the medical necessity documentation, handled as well as the conversation itself. Do that and the note gets more accurate. Less gets redone. Less gets re-asked. It holds up downstream, for CDI, for coding, for medical necessity review. And the physician gets back time currently spent rebuilding all of it by hand.

That's why SutureNote exists.