The problem is that doctors “continue to be trained to and socialized to think in stories” but electronic documentation in EHRs – or so it is assumed – require that they be “glorified (or not) data entry clerks” checking off boxes and providing codes for billing and other data elements for quality metrics. (Indeed both the ACP and AMIA have both highlighted this as an urgent issue. See our recent blog post “Human Digestible Documentation Tops AMIA HER 2020 Task Force Recommendations.”)
The opportunity is vast: use natural language processing (NLP) to mine the information in the electronic health record to uncover evidence and risk factors that protect the health of individuals and populations. He cites a recent Stanford study linking the use of proton pump inhibitors to heart attack, made possible only by use of NLP to scour 16 million records.
Think about a world in which a patient’s electronic notes, going back many years, can be mined for key risk factors or other historical elements, without the need to constrain the search to structured data fields like prescription lists or billing records. (Conflict alert: I’m an advisor to a company, QPID Health, which builds such a tool.)
We are happy to be part of the solution with our NLP-based clinical reasoning platform.
In February the American College of Physicians (ACP) published a policy paper that said technology should preserve the “humanistic elements of the record” and that the goal of documentation should be “history-rich notes.”
They observed that today’s copy-and-paste-bloated medical records represent “coding and compliance trumping clarity and conciseness.”
Now, the EHR 2020 Task Force of the American Medical Informatics Association (AMIA) has issued a series of recommendations to improve EHRs, starting with “Simplify and Speed Documentation.” According to AMIA “important aspects of the patients’ stories can only be effectively captured by narratives” and that “with natural language processing we might have accurate and human-digestible narrative as the primary input with computer-understandable discrete data as a by-product.”
Compared to human narrative, purely coded templates do not distinguish the informational wheat from chaff nor do they capture the subtle special circumstances of each patient. Further, coded templates are a disservice to the communication needs of clinicians.
Do we detect a theme here? Yes!
Stop making clinicians enter rote and redundant documentation that doesn’t have any meaning for patient care! Let them tell the story and make technology do the work of making data re-usable for quality reporting and other administrative tasks.
Why do both organizations, and many other studies and articles, focus on documentation? Reimbursement rules, concerns about malpractice, and other factors have increased the documentation burden on clinicians. EHRs have added an extra 48 minutes per day for documentation, much of it checking boxes and filling out templates. Doctors are not happy, and neither are their patients when their doctors are preoccupied with data entry.
What’s the disconnect?
Here’s an example: Reimbursement is higher under Medicare Advantage when there is clear documentation for complications of chronic diseases. But to a doctor seeing that patient this year, it doesn’t seem necessary to document (again!) that the patient had a toe amputated last year because of neuropathy. To the coder responsible for making sure the hospital is appropriately reimbursed for the care they provide, it matters.
But the fact that the patient had an amputation exists in the record. Why shouldn’t technology help find that, and leave the nurses and doctors free from the burden of incessant documentation? (It can indeed. That’s where QPID comes in.)
Letting clinicians practice medicine, and easing the burden of incessant documentation, is what QPID is all about. It’s great to see the ACP and AMIA bringing the issue to the fore, and suggesting solutions that are in use today by our customers.