It’s 7:40 on a Tuesday night and you’re still in the chart.
Ask yourself an uncomfortable question: in the last twenty minutes, what did you actually decide?
Probably not much. You reconciled a med list against a discharge summary written by someone who didn’t know the patient. You scrolled a consult note looking for the one line that mattered. You clicked into a task module to log a follow-up you’d already written in your assessment. You tried to remember whether you’d sent the referral or only meant to.
None of that is medicine. All of it is required before medicine can happen. And if you are like most clinicians, it consumes the majority of your working life.
As clinicians, we spend an enormous share of our time with the electronic health record, both getting data in — documenting information about our patients — and getting data out — searching through the chart to find information. The documented information serves routine clinical, medicolegal, and billing purposes, along with research, quality improvement, and population health. Each of these purposes, to varying degrees, influences what, when, and how we document.
The specific ways we document and retrieve information shape our day-to-day efficiency, our satisfaction with our work, and our ability to provide high-quality care. Poorly designed systems lead to the proliferation of out-of-date and incorrect information, more time spent searching the chart, medical error, and clinician burnout. When Sinsky and colleagues shadowed ambulatory physicians, they found that for every hour of direct clinical face time, physicians spent nearly two additional hours on EHR and desk work. When Arndt and colleagues instrumented the EHR event logs of 142 family physicians, they found clinicians spent roughly 5.9 hours of an 11.4-hour workday inside the EHR — with clerical and administrative work accounting for about 44% of that time and inbox management another 24%. More than half the workday, spent in the machine. Our own research found that half — literally half — of all the words in a large academic health system’s clinical notes were exact duplicates of text written earlier in the same patient’s record.
Numbers like these get cited constantly, usually in service of an argument about staffing, or reimbursement, or software vendors. This book makes a different argument.
The question nobody asks
You probably already have an opinion about the specific EHR software used at your institution — the navigation, the decision support, the search. You might have a strong opinion about electronic health records as a whole. But do you have an opinion about the underlying paradigms of information organization and storage?
For instance: what do you think about the idea of the note as the primary mode of documentation? Is the note the best way to bundle related clinical information together? Could the idea of the note itself be outdated — a relic of the era of paper charts that contributes to so many clinicians’ frustrations with the EHR? What other paradigms are possible, and how could they improve patient care and clinician satisfaction with the records of the future?
At first these may seem like strange questions. How else could we document our patient encounters? What would an alternative structure even look like? To be clear, we are not arguing that clinical documentation should be abandoned — only that we shouldn’t limit ourselves to the idea of the single-user, static, indivisible note. We will argue that a set of “notes,” as traditionally conceived, is not the most effective way to structure clinical documentation; in fact, it is a significant contributor to the problems that plague the EHR. Even if there remain use cases where the note is the appropriate concept in our design toolbox, there are many cases where alternative paradigms have decisive advantages.
This claim tends to provoke one of two reactions. Clinicians who have spent years watching the same lab value get retyped into note after note tend to nod. Everyone else asks the reasonable question: if the note is so broken, why is it everywhere? The answer — which takes the first part of this book to develop — is that the note is everywhere for historical reasons, not functional ones. It was the right technology for paper. It was carried into the digital era wholesale, without re-examination, and then locked in place by regulation, billing practice, interoperability standards, and habit. The note is not a law of nature. It is a design decision that nobody remembers making.
The plan of this book
The argument runs in three parts.
Part I dissects the note. Chapter 2 asks what a note actually is — and answers that it is a bundle of clinical facts grouped along three axes at once: by time, by clinical thread, and by responsibility. Understanding these three bundling functions precisely is what lets us evaluate the note against alternatives, rather than simply venting at it. Chapter 3 traces the note’s origins to the physical constraints of paper — single-reader access, atomicity, unlinked copies, no version history — and shows how those constraints were digitized rather than removed. Chapter 4 examines the standards era: Meaningful Use, HL7’s C-CDA, FHIR — and finds that at no point during the transition from paper to digital was the note concept seriously questioned. Chapter 5 turns to the templates — SOAP and its descendants — that govern what goes inside the note, and the peculiar dangers of notes that look perfect.
Part II names the damage. Chapter 6 introduces the framework of information chaos — Beasley and colleagues’ five documentation pathologies: overload, underload, scatter, conflict, and erroneous information — which gives us a precise vocabulary for the difference between a well-organized chart and a harmful one. Chapter 7 is the analytic core of the book: seven specific mechanisms by which the note paradigm directly produces information chaos. Chapter 8 re-examines the most-blamed behavior in documentation — copy-paste — and argues from our duplication research that it is a rational adaptation to a system that forgets, not a moral failing of clinicians. Chapter 9 tallies the human cost: the articulation work, the reconstruction tax, the moral injury of spending your evenings as a human indexing service.
Part III builds the alternative. Chapter 10 describes the chart as a single, dynamically updating, fully collaborative workspace — and reports the feasibility study in which we actually built one. Chapter 11 argues that the medical problem, not the note, is the natural unit of clinical meaning, and works through what should carry forward across visits and what should stay behind. Chapter 12 resolves the apparent conflict between consistency and clinician autonomy: normalize the artifact, personalize the projection. Chapter 13 takes up artificial intelligence — why faster note generation deepens the underlying problem, and where AI genuinely changes what is possible. Chapter 14 is the honest chapter about the terrain between here and there: what cannot ship today, why, and the pragmatic bridge worth building in the meantime. Chapter 15 concludes.
Who this book is for
We wrote this book physician-to-physician, but not only for physicians. If you are a clinician, we hope to give you a vocabulary for problems you have felt for years but perhaps never seen named — and to convince you that they are tractable, which is a different thing from easy. If you build health software, we hope to persuade you that the most valuable thing you can do is not to generate the old artifacts faster but to question the artifacts themselves. If you regulate, purchase, or administer these systems, we hope to show you where the leverage actually is: not in the interface, but in the paradigm underneath it.
It may seem strange to devote a whole book to clinical documentation, which can feel so ancillary to the reason any of us joined medicine — taking care of patients. But that is precisely the point. The time we spend fighting the record is taken directly from patients, from learning, from research, and from our lives outside work. Documentation is not a side issue. For most clinicians it is the single largest consumer of the working day, and it is a leading contributor to the epidemic of burnout — or, more precisely, moral injury — hollowing out the profession.
Most physicians we know aren’t asking to work less. They’re asking for their work to be the work. That is what this book is about.