Love them or hate them, electronic medical records (EMR) are here to stay. And that’s not necessarily a bad thing. They need not be painful but in some instances they have become so. I installed my first EMR in 2002 and have installed several since then. As I like to say, even if you “feel” like your EMR deployment was 100% successful, that really means it was about 70 – 80%. In lieu of being built by clinicians, many EMR and practice management (PM) systems (the IT piece that essentially handles the billing/coding, scheduling, etc.) tended to, in the early 2000s and before, be built by IT folks who created the product(s) for what they thought the clinicians and practice administration needed vs. what they wanted (and actually needed). And, there seemed to exist a lot of extraneous stuff. My aeronatucial engineer buddy uses the technical term “over engineering.”
In any event, a good EMR deployment is quite a project requiring a team effort but EMRs can be effectively rolled out to assist clinicians in the delivery of care. Structured appropriately, and for clinician needs and workflow, an EMR can assist with care delivery, patient flow, care quality management, optimization of schedule, care follow up; the list is endless vis-à-vis a medical practice. However, over time and/or due to sheer intransigence, many EMRs have devolved into voluminous repositories of clinical “noise” vs. useful chunks of manageable data. EMR functionality (impacted, generally, by the install) can force clinicians to seek the path of least resistance by deploying work-arounds to accomplish what, in the days of paper charts, used to take them seconds.
The reason? Well, I’m sure there are many. But clinicians, or their medical assistants, have been trained to “pull forward” old notes and data which, while some of it is useful apropos of patient history, some information simply piles on and serves no clinical or adjunctive benefit.
While I neither prostheletyze nor ruminate much about EMRs, I received an email push from Eric Bricker, MD, CEO of AhealthcareZ (ahealthcarez.com) who submitted to his email list an article suggesting that the “….University of Pennsylvania recently published a study in the Journal of the American Medical Association found that 50% of EMR text is copied and pasted.” (Dr. Bricker has a nice video on this topic on YouTube.) As someone who’s been on the operations side of the business for 30+ years, I am not at all surprised. When I think about my personal IM doc, whom I only see annually for my physical, I can only imagine what is layered into my medical record and what gets pulled forward year after year. Had I been a longtime patient of his (longer than my 16 years), with co-morbidities and owing to the capture of old paper charts scanned/abstracted into the EMR, the volume of useable information might get lost amongst the sheer mass of scanned paper.
Aside from the care management risk of a “noisy” chart that a clinician only has time to skim, there exists a significant med/mal risk in terms of shotty pre-visit workup. There are instances where inaccurate information that should otherwise have been deleted is towed forward visit to visit; there have actually been malpractice cases where this has been elucidated. And inaccurate coding based on data that was haphazardly pulled forward to document an E&M visit can subject clinicians to coding “situations” (think fraud and abuse). A documented, coded and signed chart is a legal document and (as we said when I worked Medicare fraud 800 years ago), “….if it wasn’t documented it wasn’t done…”). Given the data pulled forward, you might also ask “…if it was documented, was it done?”
Example: you have a clinician, or roster of clinicians, who are on a productivity (work relative value unit [wRVU]) comp model. What are the dangers with “overuse” of the EMR’s coding function and “pre-populating” charts? I had a client reach out a while back who suggested all of his specialists (in a given specialty), who were on a wRVU comp model, were achieving north of 95th% for comp relative to their peers. Additionally, all had, consistently, passed their coding audits. In the eyes of the health system, these clinicians were productive, were highly compensated, and were passing chart audits – bullet proof. However, the client was concerned that something was amiss – too good to be true. That got me to thinking: I took a random sample of their schedules examining lab and office days. What I divined is that the sampled schedule seemed (excuse the alliteration) to indicate that the clinicians were, essentially, billing more “time” for patient-facing care than they had available in the clinic day. E.g. a visit that took 10 minutes was billed at a 45 minute visit CPT level. What this meant was that the providers were able to use the EMR to “document” the work and meet all of the documentation requirements to bill higher levels of care while not really spending the visit-required length of time with patients.
As you can see, and I’m sure this is neither the exception nor the rule, these types of things can happen, whether by design or accident. The medical record must be curated and carefully managed to ensure only relevant clinical data is pulled forward for each visit. I wonder, at times, if in this age of speed and production we’ve somehow left behind the accuracy of our documentation.
I think it’s worth stating I’m not suggesting that clinicians are inflating codes or falsifying documentation wantonly or with malice. But EMRs can empower clinicians to obtain the most out of a CPT code.
EMRs can be very useful. Healthcare entities, generally, have been loathe to manage the myriad data points accumulated over time. The data that IT systems and EMRs contain can be utilized in a variety of avenues to treat conditions, manage populations, or address chronic disease states (e.g. ChF patients, managing patients w/diabetes, etc.). They must be used in a manner with which they were (theoretically) designed.