By Michael Zaroukian, MD, PhD, FACP, FHIMSS and Peter Basch, MD, FACP
Incorrect Assumptions Lead to Incorrect Conclusions That Misinform Policy
Please refer to Part 1 of this post for our comments on the recent Wall Street Journal op-ed questioning the policy behind the EHR incentive program, and Part 2 for our comments on the extent to which EHR use was associated with higher short term charges, as published in the Center for Public Integrity report and recently summarized in the New York Times, and Part 3 for our comments on where one would expect EHR use to show short-term cost increases.
One of the inferences of the Center for Public Integrity report was that a finding of repetitive language (boilerplate documentation) in medical notes or similar appearing notes should raise suspicion of fraudulent behavior. Regarding the issue of extreme repetitive documentation, we agree that it is hard to imagine instances where appropriate documentation is also a complete and exact copy of a previous note, or the exact note (with the name of the patient changed) that can be found in another patient’s chart. That said, we absolutely expect that parts of many medical notes should include similar if not identically worded phrases and sentences that can be seen in many other notes because medical documentation uses a standard language and boilerplate templates for describing findings. For example, as medical students we were trained to use a single comprehensive Review of Systems and a single comprehensive physical examination template, and we were criticized for missing components of it.
Fortunately for patients, physicians are trained to use standard terms to describe and clearly communicate positive and negative findings to each other. While there can be regional variations, such variations are predictable and comparatively minor. For example, a patient’s normal lung examination may be documented as “clear” or “clear to auscultation” or “clear to percussion and auscultation” or “clear to P&A” depending on where doctors were trained and how concisely they want to document their normal findings. The same examination may also be expanded eloquently but unnecessarily to “chest/lung examination: inspection reveals no asymmetry, retractions or use of accessory muscles. Palpation reveals no tenderness, crepitance, or tactile or vocal fremitus. Percussion note is resonant with no areas of dullness and with normal diaphragmatic excursion. Auscultation reveals bronchovesicular breath sounds with normal I:E ratio, and no crackles, wheezes, bronchophony, egophony or whispered pectoriloquy.”
The medical trainee who did not learn to consistently use standard descriptions and become progressively more concise, or persisted in the use of inappropriate or unclear terminology for each patient (“lungs sound totally awesome but the heart just didn’t sound right!”) would likely flunk out of medical school. Having ourselves been students who initially used highly verbose descriptions, we were praised in the classroom for our excellent retention of the clinical skills course content and our detailed and thorough descriptions, followed shortly on the wards by corrective feedback or even criticism from residents and attending physicians that our notes were too long to read and took us too long to write, without adding benefits to quality and safety for the vast majority of patients. Note that “boilerplate language” is used for both normal and abnormal findings as well as much of medical history taking – and this is true whether or not documentation is handwritten, checked off on a paper documentation form, templated in an EHR, or “uniquely” dictated (“please insert my normal 14-part Review of Systems”). This process of self-discovery of unique verbiage for most encounters is a common EHR implementation practice. Where providers commonly state a limited number of descriptions of historical or exam findings, it is typical to construct EHR documentation templates with drop-down options to facilitate appropriate documentation using a limited set of most descriptive terms (chest pain could be described with drop-down options such as: dull, sharp, pressure, burning, stabbing, etc.).
So if we can agree that much of medical documentation has always looked more similar than dissimilar between patients and even within the same patient over time – we contend that both of us may well be repetitively and accurately described over the last several decades in our respective medical records as a “well-developed, well-nourished man in no acute distress” – the difference between handwritten paper notes and EHR documentation is that with the latter, one now has the ability to actually read the documentation, so perhaps the realization that we say mostly the same things much of the time is a “downside” of legibility. We would argue that when the lungs are clear to percussion and auscultation or a holosytolic, grade II/VI murmur is present and best heard at the left lower sternal border and apex, it is both appropriate and a best practice to describe these findings the same way each time they are documented, and they should be documented every time they are relevant to the visit. The patient’s health will not be served by creative writing that is potentially confusing or wastes time.
For the record, it is our view that ‘copy forward’ and ‘copy-and-paste’ EHR functionalities are just tools that can be used appropriately or inappropriately. Used appropriately and with modifications that bring the data they contain up-to-date, these functionalities can contribute to more efficient, consistent and complete documentation of important conditions and findings. Used inappropriately, they can be inaccurate, unsafe, and contribute to fraud and abuse. The reality is that this is potential exists for the entire EHR and for paper charting systems, which can be used inappropriately to enabling the authoring of ‘science fiction’. As such, we would argue that copy-and-paste and copy forward capabilities of EHRs, like scalpels that can also cut both ways, should be available to those trained and committed to using them correctly and to good purpose, and withheld from those who are unwilling or unable to use them according to expectations.
Seeing similar terminology across patients or within the same patient over time is NOT prima facie evidence of fraud; it is prima facie evidence that the author of the notes graduated medical school. Depending on the setting and scope of practice, notes for different patients with the same problem will look more alike than different and for the same patient over time, except where there are acute changes or illnesses, much of the normal, chronic or unchanged findings documented in the note should appear pretty much the same from the previous note. As stated in each of the prior parts of this post, we take intentional billing fraud seriously, and would join the call to develop software that helps to illuminate and address potential fraud. That said, such software that looks for common terminology, phrases, or even similarity in full notes would detect appropriate and accurate documentation far more often than intentional fraud.
Please continue to Part 5 for our discussion of coding software, and on view on the call to regulate its use.