We are certainly not the first to suggest that paradigms from the paper era were wholesale adopted into the EHR age, at great cost in documentation burden and clinician burnout. Hopefully the general argument so far is convincing. But you don’t have to take our word for it. We can examine the dominant conceptual frameworks in play at the moment of transition to digital records, and see for ourselves whether the note concept received significant scrutiny — or passed through unquestioned.
A note on intent: we are not here to criticize decisions made in a completely different technological and medical culture during a period of massive transition. We only want to establish whether the idea of the note was ever thoroughly examined by the institutions best positioned to examine it. This chapter dives into the fine details of documentation standards and how they have evolved since the inception of the EHR. If you share our soft spot for data standards, read on; if not, skim to the summary at the end and proceed to Chapter 5 with our blessing.
4.1 Meaningful Use: notes as an afterthought
Perhaps the most important place to look is the Meaningful Use (MU) program, which tied specific EHR adoption measures to Medicare reimbursement in an effort to accelerate the digital transition. MU was divided into three stages, each containing requirements that organizations had to meet through EHR usage.
Exactly one Meaningful Use criterion relates to notes: the requirement “to record electronic notes in patient records” — an optional “menu” objective in MU Stage 2. Meanwhile, the majority of the first-stage MU objectives focused on recording highly specific structured data elements: orders, allergies, medications, demographics, vital signs, laboratory results, and problems.
Two things follow from this. First, undifferentiated chunks of free text — notes — were a distant second priority during the era that defined what an EHR had to be. The note was neither questioned nor championed; it was simply assumed, carried along as the default container for everything the structured elements didn’t cover. Second — and this is worth pausing on — the somewhat-arbitrary division we now take for granted between “structured” data (entered outside the note) and “unstructured” data (entered inside it, templated or not) may be largely a consequence of Meaningful Use requirements rather than any clinical or informatic logic. The regulation drew a line through the patient’s information, and the software industry built walls along the line.
4.2 HL7 and C-CDA: flexible blocks, frozen templates
Initiatives to develop international standards for health information systems represent another venue where dominant documentation paradigms might have been critically examined. Such initiatives — including Health Level 7 (HL7) and Fast Healthcare Interoperability Resources (FHIR) — aim to promote interoperability by establishing agreed-upon standards for healthcare data transfer: a laudable goal in a world of fragmented systems. These standards exert significant influence on the design of present and future EHRs — not just on data transfer, but upstream on data organization and even user input workflows, particularly when interoperability metrics are tied to payment incentives.
HL7’s Consolidated Clinical Document Architecture (C-CDA) provides a set of flexible building blocks and document templates for common clinical communication scenarios: continuity-of-care documents, procedure notes, discharge summaries. The building blocks themselves are flexible and can construct arbitrary documents. But the majority of the templates conform to the dominant paradigms of the atomic note — progress notes, discharge summaries, H&Ps. (Not all C-CDA documents function as notes in the fact–note–chart hierarchy; the Continuity of Care Document acts more like a chart, summarizing the whole course of care up to a transfer.) In short, the standard concerns itself with communicating free-text documents from one system to another, and is technically flexible enough to permit different organizations of information — but the templates it provides conform to, and therefore enforce, a particular mindset about how clinical data should be sliced, stored, displayed, and transmitted.
4.3 FHIR: the note, standardized but unexamined
The majority of FHIR guidance focuses on transmitting structured data, but there is a small informative section on “clinical notes.” The US Core implementation guide defines a minimum set of note types that a compliant system must support: three diagnostic report types (cardiology, pathology, radiology) and five general-purpose clinical note types (history & physical, progress note, discharge summary, consult note, procedure note). Beyond these sits a much more comprehensive catalog of “document reference types” adapted from the LOINC ontology, which breaks notes into hyper-specific categories — “Deprecated Speech therapy Initial assessment note at First encounter” is a real entry — that compliant systems are encouraged to support in transmission.
What you will not find, anywhere in this guidance, is a discussion of alternative types of fact bundling — any consideration of whether “progress notes” in particular, or notes in general, are the right way to “communicate the current status of a patient.” We understand this may fall outside FHIR’s mission; presumably the standard must facilitate the data types already in widespread use in order to secure vendor buy-in. But it does seem a missed opportunity. If these large multinational bodies — composed of exactly the expert stakeholders one would convene to define the future structure of medical documents — are not positioned to ask whether the note is the right structure, then who exactly is?
4.4 Paradigms all the way down
We will examine the usefulness of specific note types elsewhere; suffice it to say that some of them no longer make sense as fact bundles in the digital age, because by their very nature they incentivize the creation of duplicated, out-of-date, and erroneous information. What matters most to recognize here is the extent to which standards shape future systems — and thereby shape our own cognitive models of what documentation can be. Standards are how a temporary consensus becomes a permanent assumption.
The mechanism is worth spelling out. If every EHR ships a template called “progress note,” every insurance company requires a daily “progress note” for billing, and every clinician is trained to produce a “progress note” each day, then it becomes nearly impossible for anyone — vendor, payer, or physician — to imagine, let alone enact, a different paradigm, no matter how sensible. Each institution’s requirement is justified by the others’ existence. The note is load-bearing not because anyone examined it and found it strong, but because everything was stacked on top of it before anyone thought to look.
And as far as we can tell — having looked — at no point during the transition from paper records to digital records was the concept of the note in general, or of specific note types like the progress note, ever seriously questioned, nor were alternatives seriously considered.
4.5 The counterarguments, and why they fail
Two objections to moving beyond the note come up so reliably that they deserve answers here, before we go further.
“The note has to be the atomic unit for insurance and medicolegal reasons.” The claim is that we need note-level bundles to assign specific clinicians responsibility for specific actions. This argument lacks imagination. Many software systems we use daily keep meticulous, effectively infallible track of which users made which edits to which documents, down to the individual word. Most EHRs already track which clinicians changed which notes. Exactly the same granular responsibility-tracking is possible in a non-note paradigm: we can build systems that trace any piece of text to a particular user, even if that text has since been edited or deleted by another, and good interface design can make the provenance of every fact instantly visible. Responsibility assignment is a solved problem in software. It is not a reason to prefer the atomic note — either paradigm supports it, and the fact-level version is frankly stronger.
“Attestation requires documentation.” As established in Chapter 2, stating agreement with, and taking responsibility for, a set of facts does not require re-writing those facts. Attestation-based mechanisms already exist in most EHRs for medication reconciliation and problem-list review — a single action that says “I reviewed this; nothing has changed.” The concept simply has never been extended to unstructured clinical information, because the note’s atomicity makes it all-or-nothing: you’d have to agree with 100% of a colleague’s note to attest to it as a unit. Break the atomicity, and attestation without redocumentation becomes available everywhere — with every attestation tracked, and responsibility assigned at least as precisely as today.
There have, in fairness, been limited pushes against the note’s monopoly. From the beginning, EHRs have made some information individually editable at the fact level — demographics, orders, lab results, vital signs — living in non-note interfaces that require no bundling. More EHRs are adopting collaborative problem-list interfaces, editable at the level of the individual problem, with change tracking. These systems hold real promise. But notice the pattern: to actually cut into the note’s monopoly, they would have to replace the need to document the same findings inside a note. In current workflows they don’t — we document problems, medications, and results both inside and outside the note, which is to say the “structured data” innovation gave us a second documentation burden rather than a first solution. And even these inroads leave the vast majority of important unstructured information — symptoms, exam findings, diagnostic and therapeutic reasoning, the patient’s history in free text — completely trapped inside the atomic note.
4.6 Summary
The historical record shows a clear pattern. The technical constraints that created the note dissolved with digitization (Chapter 3), but the note survived — codified by Meaningful Use as an unexamined menu item, templated into permanence by C-CDA, given a standardized taxonomy by FHIR, demanded daily by payers, and drilled into every trainee. The standard objections to alternatives — responsibility and attestation — dissolve on inspection. What remains is a paradigm sustained by mutual institutional reference and habit.
That would be merely an interesting historical observation if the note were harmless. The next part of this book is about why it isn’t. To get there, we need one more piece of equipment: a precise vocabulary for how documentation systems fail. But first — a brief word of remembrance for the deceased.