Note Bloat in EHR Systems and How AI Scribes Fix It

Half your note was probably copied from somewhere else

Heres a stat that should bother you: 50.1% of text in clinical notes is duplicated from text previously written about the same patient. That's not a typo. Half the words in a typical progress note already existed in a prior note. They were copied forward, sometimes through several visits, accumulating like sediment.

Epic Research confirmed the trend with a study of 1.7 billion clinical notes. Note bloat increased by 8% over their study period. Across healthcare, note length grew 60% between 2009 and 2018, while note redundancy climbed 11%.

The result is notes that are long, repetitive and hard to scan. The AMA found that US clinical notes are four times longer than those in other countries. Four times. That gap isn't because American patients are sicker. It's because American documentation practices reward volume over clarity.

How note bloat happens

Note bloat isn't one problem. It's several problems wearing the same trench coat.

Copy-paste culture: Between 60% and 90% of clinicians use some form of copy-paste, templates, macros, autotext or autofill. A study of over 23,000 provider notes found that only 18% of text was entered manually. The rest was copied (46%) or auto-imported (36%).

Defensive documentation: Fear of audits and malpractice pushes physicians to document more rather than less. The logic goes: if something isn't in the chart, it didn't happen. So everything goes in the chart, whether it's clinically relevant or not.

Billing requirements: E/M coding guidelines historically rewarded longer, more detailed notes with higher reimbursement. Even though CMS revised these guidelines in 2021 to focus on medical decision making, old documentation habits persist.

EHR design: Most EHR systems make it easier to copy a note forward and edit it than to write a new one. The path of least resistance produces bloated notes by default.

Why bloated notes are actually dangerous

Note bloat isn't just an annoyance. It creates real clinical risks.

When a hospitalist inherits a patient with 47 progress notes, each one containing five pages of copied text, finding the actual clinical story becomes a scavenger hunt. The critical detail - the medication change on day three, the new allergy documented on day five - gets buried in redundant text.

Specific risks include:

• Outdated information persists: Copied text carries forward details that are no longer accurate. A "no known drug allergies" from three visits ago might still appear even after an allergy was documented
• Contradictory information: When old text isn't updated, notes can contain conflicting statements about the same patient
• Longer review time: Other providers spend more time reading bloated notes to find relevant information, which delays clinical decisions
• Audit and legal exposure: Notes filled with copied text can raise questions about whether the documented exam and assessment actually occurred during this visit

How AI scribes produce cleaner notes

AI scribes approach documentation from a fundamentally different direction than copy-paste workflows. Instead of starting with yesterdays note and modifying it, the AI starts with todays conversation and builds a new note from scratch.

This architectural difference eliminates the primary source of note bloat. Every note reflects what actually happened during the current encounter. Nothing is carried forward automatically.

Well-designed AI scribes also produce concise notes because they're trained to extract clinically relevant information, not to pad notes with boilerplate. The AI captures:

• What the patient reported (subjective)
• What the physician examined and found (objective)
• The clinical assessment and reasoning
• The treatment plan

It doesn't add filler. It doesn't copy forward a three-paragraph past medical history that hasn't changed. It documents this visit, completely and concisely.

Finding the balance between complete and concise

Shorter notes aren't automatically better notes. A two-line progress note that omits the clinical reasoning is worse than a bloated one. The goal is documentation that's complete without being redundant.

AI scribes hit this balance by separating what needs to be documented from what needs to be referenced. The patients chronic problem list, medication history and allergies live in structured EHR fields that don't need to be repeated in every progress note. The AI note focuses on what changed during this visit.

A few principles for leaner documentation:

• Document changes, not the unchanged - if medications haven't changed, say "medications reviewed, no changes" instead of listing all twelve
• Reference, don't repeat - point to the problem list instead of restating it
• Write for the next reader - ask yourself what another physician needs to know about this specific visit
• Let structured data do its job - vitals, labs and imaging results belong in their dedicated EHR fields, not pasted into the note body

Transcribe Health generates focused clinical notes from your patient conversations. No copied text. No bloat. Just documentation that reflects what happened in the room, structured for easy review by you and every provider who reads it after you.