Exploring the Role of Clinical Documentation: a Step Toward EHRs for Learning

We need more clinicians weighing in on the design of the tools they use, so I was gratified to see a policy paper from the American College of Physicians about EHRs this week. In a sporadic and tentative manner, the paper recognizes that our digital tools for clinical documentation are part of a universal health care system that requires attention to workflow, care coordination, outcomes, and research needs.

The strong points of this paper include:

  • A critique of interfaces that hobble the natural thought processes of the clinician trying to record an encounter

  • A powerful call to direct record-keeping away from billing and regulatory requirements, toward better patient care

  • An endorsement of patient access to records (recommendation 6 under Clinical Documentation) and even more impressively, the incorporation of patient-generated data into clinical practice (recommendation 5 under EHR System Design)

  • A nod toward provenance (recommendation 3 under EHR System Design), which tells viewers who entered data and when, thus allowing them to judge its accuracy

Although the authors share my interests in data sharing and making data available for research, their overarching vision is of an electronic record that supports critical thinking. An EHR should permit the doctor to record ideas about a patient’s condition as naturally as they emerge from his or her head. And it should support other care-takers in making treatment decisions.

That’s a fine goal in itself, but I wish the authors also laid out a clearer vision of records within a learning health care system. Currently a popular buzzword, a learning health care system collects data from clinicians, patients, and the general population to look for evidence and correlations that can improve the delivery of health care. The learning system can determine the prevalence of health disorders in an area, pick out which people are most at risk, find out how well treatments work, etc. It is often called a “closed loop system” because it can draw on information generated from within the system to change course quickly.

So at the start of the policy paper I was disappointed to read, “The primary goal of EHR-generated documentation should be concise, history-rich notes that reflect the information gathered and are used to develop an impression, a diagnostic and/or treatment plan, and recommended follow-up.” What about supporting workflows? Facilitating continuous, integrated care such as in a patient-centered medical home? Mining data for new treatments and interventions? Interfacing with personal health and fitness devices?

Fortunately, the authors massage their initial claim by the time they reach their first policy recommendation under Clinical Documentation: “The primary purpose of clinical documentation should be to support patient care and improve clinical outcomes through enhanced communication.” The primary purpose gets even better later on: “As value-based care and accountable care models grow, the primary purpose of the EHR should remain the facilitation of seamless patient care to improve outcomes while contributing to data collection that supports necessary analyses.”

One benefit of reading this paper is its perspective on how medical records evolved to their current state. It notes a swelling over the decades in the length of notes and the time spent on them, “the increased documentation arguably not improving patient care.” Furthermore, it details how the demands of billing drove modern documentation, blaming this foremost on CMS’s “issuance of the evaluation and management (E&M) guidelines in 1995 and 1997.” I suspect that private insurers are just as culpable. In any case, the distortion of diagnosis in the pursuit of payments hasn’t worked well for either goal: 40% of diagnoses are wrongly coded.

The pressures of defensive medicine also reveal the excessively narrow view of the EHR currently as an archive rather than a resource.

The article calls for each discipline to set standards for its own documentation. I think this could help doctors use fields consistently in structured documentation. But although the authors endorse the use of macros, templates, and (with care) copy/forward, they are distinctly unfriendly toward structured data. Their distemper stems from the tendency of structured interfaces to disrupt the doctor’s thinking–the presevervation of which, remember, is their main concern–and to make him jump around from field to field in an unnatural way.

Yet the authors recognize that structured data is needed “for measurement of quality, public health reporting, research, and regulatory compliance” and state in their conclusion: “Vendors need to improve the ability of systems to capture and manage structured data.” We need structured data for our learning health care system, and we can’t wait for natural language processing to evolve to the point where it can reliably extract the necessary elements of a document. But a more generous vision could resolve the dilemma.

Certainly, current systems don’t handle structured data well. For instance, the article restates the well-known problem of redundant data entry, particularly to meet regulatory requirements, a problem that could be solved with minimally intelligent EHR processing engines. The interactive features available on modern mobile devices and web interfaces could also let the clinician enter data in any manner suited to her thinking, imposing structure as she goes, instead of forcing her into a rigid order of data entry chosen by the programmer.

Already, Modernizing Medicine claims to make structured data as easy to enter as writing in a paper chart. As I cover in another article, they are not yet a general solution, but work only with a few fields that deal with a distinct set of health conditions. The tool is a model for what we can do in the future, though.

The common problem of physicians copying observations from a previous encounter and pasting them into the current encounter is a trivial technical failure. On the web, when I want to cite material from a previous article, I don’t copy it and paste it in. I insert a hyperlink, I did in the previous paragraph. EHRs could similarly make reporting simple and accurate by linking to previous encounters where relevant.

The ACP recommendations are sensible and well-informed. If implemented by practitioners and EHR developers who keep the larger goals of health care in mind, they can help jump over the chasm between where EHRs and documentation are today, and where we need them to be.

Full Disclosure: Modernizing Medicine is an advertiser on this site.

About the author

Andy Oram

Andy Oram

Andy Oram writes and edits documents about many aspects of computing, ranging in size from blog postings to full-length books. Topics cover a wide range of computer technologies: data science and machine learning, programming languages, Web performance, Internet of Things, databases, free and open source software, and more. My editorial output at O'Reilly Media included the first books ever published commercially in the United States on Linux, the 2001 title Peer-to-Peer (frequently cited in connection with those technologies), and the 2007 title Beautiful Code. He is a regular correspondent on health IT and health policy for HealthcareScene.com. He also contributes to other publications about policy issues related to the Internet and about trends affecting technical innovation and its effects on society. Print publications where his work has appeared include The Economist, Communications of the ACM, Copyright World, the Journal of Information Technology & Politics, Vanguardia Dossier, and Internet Law and Business.