Why Doctors Spend 2 Hours on Notes for Every 1 Hour of Care

A doctor finishes seeing patients.

The waiting room is empty.

The clinic is technically closed.

But the work isn't finished.

There are still notes to complete.

Emails to write.

Referrals to prepare.

Records to update.

And tomorrow's patients are already waiting.

This is one of the biggest problems with modern clinical documentation.

The Consultation Isn't the Whole Job

When a patient spends 30 minutes with a doctor, that doesn't mean the doctor's work ends after 30 minutes.

There can be significant documentation work before and after the consultation.

The doctor may need to record:

The patient's history

Symptoms

Examination findings

Assessment

Treatment plan

Medication information

Follow-up instructions

Referrals

Administrative details

And all of this needs to be documented accurately.

Why Does Documentation Take So Long?

One reason is that clinical notes aren't just normal writing.

They need to contain the right information in a format that is useful for future care.

A doctor has to remember what was discussed, decide what is clinically relevant, organise it properly, and enter it into the record.

Doing this for one patient might be manageable.

Doing it for dozens of patients every day is different.

The Keyboard Can Become Part of the Consultation

Think about the typical consultation.

The doctor is talking to the patient.

Then looks at the computer.

Types something.

Looks back at the patient.

Checks the record.

Types again.

The technology is supposed to help the doctor.

But sometimes it can become another thing competing for the doctor's attention.

This is one reason ambient AI documentation has become interesting.

Instead of requiring the doctor to manually document every part of the conversation while it happens, an AI scribe can capture the conversation and generate a draft note afterward.

AI Doesn't Remove Documentation

This distinction matters.

An AI scribe doesn't mean:

“The doctor no longer needs to document.”

It means:

“The doctor may not need to create every part of the first draft manually.”

The doctor still needs to review the generated documentation.

That human review is critical because AI-generated notes can contain omissions or inaccuracies. Research into clinical AI scribes has found meaningful variation in error frequency and severity across systems.

What Could the Workflow Look Like?

Without an AI scribe:

Patient Consultation

Doctor Remembers Key Details

Doctor Types Note

Doctor Edits

Doctor Finalises

With an AI-assisted workflow:

Patient Consultation

AI Captures Conversation

AI Creates Draft

Doctor Reviews

Doctor Edits

Doctor Finalises

The second workflow doesn't eliminate the doctor.

It changes where the doctor's time is spent.

What Happens to the Time Saved?

This is where the real benefit can appear.

Instead of spending as much time constructing notes, the doctor may have more time for:

Patients

Clinical review

Communication

Referrals

Follow-up

Family or personal time

Other clinical responsibilities

The exact time saved depends heavily on the doctor, specialty, documentation requirements, EHR, and AI system.

So claims like “AI will save every doctor two hours a day” should be treated carefully.

The better question is:

How much unnecessary documentation work can this particular clinic realistically remove?

The Goal Isn't Faster Medicine

This is important.

Healthcare shouldn't become a race to see more patients.

The goal should be to reduce unnecessary administrative burden while keeping documentation accurate and useful.

A doctor shouldn't have to choose between:

Good patient communication

and

Good documentation.

Technology should help support both.

This Is Where AI Becomes Useful

AI is particularly interesting when it handles work that is:

Repetitive

Structured

Time-consuming

Predictable

Easy for a human to review

Documentation is a good example.

A system can prepare the first draft.

The doctor checks it.

The doctor remains responsible for the final clinical record.

What About Errors?

This is where guardrails matter.

An AI system that saves time but introduces incorrect information isn't solving the problem.

It's creating another one.

The workflow therefore needs:

AI generation → Human review → Final approval

Current clinical AI documentation systems explicitly position generated notes as material for clinician review rather than something that should automatically become the final medical record.

How ELRQ Looks at the Problem

At ELRQ, we'd rather ask:

“Where is the clinic losing time?”

Maybe it's documentation.

Maybe it's appointment enquiries.

Maybe it's follow-up.

Maybe it's referral letters.

Maybe it's all of them.

The solution doesn't have to be one giant AI system.

It can be a connected workflow:

Consultation → AI Scribe → Doctor Review → EHR → Patient Follow-up

The point isn't to automate the doctor.

It's to automate the work around the doctor where it makes sense.