I'd never heard of Healwell before and took a look over their offerings. Has anyone used the products? Beyond the…
Curbside Consult with Dr. Jayne 6/30/14
I spent most of this weekend doing a special project. Our coding and compliance officers approached me about how some of our providers’ notes look in EHR. They had seen some notes that were “really awful” and naturally assumed that something was going on with the EHR to cause them to be that way.
Our ambulatory vendor offers checkbox-style documentation templates, so I figured the complaints were about how their documentation was being output now that we’re dealing with SNOMED and other factors.
I asked my team to pull a sampling of notes from each of our specialties so that I could look at them myself. We’ve seen issues where the behind-the-scenes verbiage engine generates some subject/verb disagreements. Additionally, when a large number of positive and negative symptoms are documented, sometimes that can get a little strange.
Since our analysts are not clinical, I know that I can’t exclusively use their review to identify good vs. bad notes. Sometimes the documentation might be technically accurate, but would actually be something a receiving physician would laugh at.
We have a lot of subspecialists who do a lot of procedures, so I had the team pull a variety of those notes as well. They’ve been problematic in the past, especially when multiple procedures are documented. Most of those issues have been easy fixes. Still, considering the variety of specialties and all the different kinds of documentation, I had well over 100 visit notes to review.
By the time I was done, I could barely contain my aggravation. The largest subset of “awful” notes came from our providers who are heavy users of voice recognition. Some of the notes were downright incoherent. The problem however wasn’t with the technology – it was with subspecialists dictating sheer nonsense that normal humans (even those with medical degrees) would have difficulty comprehending.
The next subset of bad notes came from providers who have created their own documentation macros. The idea of providers having their own saved text blocks is generally a good one. We all know that there are some parts of the note that are the same over and over again: “regular rate and rhythm, no murmurs, rubs, or gallops, lungs clear to auscultation bilaterally, abdomen soft non-tender and non-distended with normal active bowel sounds.” From years of dictation it just rolls off the tongue, so it would make sense to save it as a block for EHR.
The problem comes when providers save text that either doesn’t make sense or has gender-specific findings that winds up being reused on the opposite gender. The point of saved text is to be able to quickly add documentation with little work. Some of our providers take the idea of efficiency too far, with so many acronyms and abbreviations it’s impossible to figure out what is going on with the patient.
Even with the subject/verb disagreement and some of the typical template issues, the group that most heavily uses check-box powered documentation did the best. They were easy to sort out due to the way the history blocks format and I was surprised at how much clearer their notes were compared to those done via other methods. Those that used the templates, however, had a much higher propensity to document Review of Systems items that I’m sure they didn’t actually perform.
For your amusement, I’ll share some of the highlights:
- General surgeon sees a patient to remove a skin cyst. She documents a gynecological review of systems with seven negative elements. I confirmed that it wasn’t from a paper form the patient completed and staff keyed in. She also documented the procedure as “EXC TR-EXT B9+MARG 2.1-3cm.” What does that even mean? I could extrapolate “benign” and “margins” from that, but it makes no sense for the type of cyst excised.
- The same surgeon documented a 21(!) point male urinary review of systems for a similar visit. The procedure document was the same except it was 0.6-1cm. At least she’s consistent. And apparently thorough, since she documented that she examined all 12 cranial nerves and the cyst was on the shin.
- Orthopedic surgeon documents a physical exam that includes a normal fundoscopic exam. I’d pretty much bank that the last time an orthopedic surgeon touched the instrument needed to look at the back of the eye, it was in medical school.
- Chief complaint of “bx results” which was saved to a provider custom list. Could we not have spared the extra characters to have it read “biopsy results” so that when the patient receives the note on our patient portal it makes sense?
- Not capitalizing the names of other physicians on the team. Nothing says “thanks for the referral” like addressing the letter to “dear dr jayne.”
- A “follow up back pain” visit with a (no kidding) 91-point review of systems including “changes in shape/size of moles” and “breast lumps.” I can’t wait until that one gets pulled for a CMS audit.
- Detailed discussions of radiologic studies pulled into the note from other practices. I guess in addition to being “one patient, one chart” the EHR also lets us time travel because the same CD with the MRI results that the patient hand-carried from shoulder surgeon was simultaneously imported to the orthopedic consultant’s May visit note and also to the nephrologist’s note with a date stamp two months prior to the visit.
I could go on, but it would just make me frustrated and likely make you angry. More than anything, it just makes me sad, especially since the providers electronically signed all of them and indicated that they were read and reviewed.
You might ask who had the best documentation. Hands-down the most coherent, thorough, and clearly non-padded were the notes done by one cardiology group using a mix of voice recognition for the history and plan and template documentation for the physical exam and review of systems. I didn’t identify any gratuitous documentation and the notes were high quality. It probably takes them longer to document since they’re speaking most of the note vs. clicking. However, their documentation was so pretty I wish I could clone them. But CMS says cloning is bad, right?
Got documentation problems? Email me.
Email Dr. Jayne.
Much of your frustration derives from the long-standing notion that charts and chart notes are for the benefit of the provider for a multitude of reasons and only tangentially of benefit for the patient. The rest of your frustration is from the well-meaning but destructive rules and regs from CMS, and the do-gooders who believe things like MU are good for other more untrustworthy physician souls. It will consume your life to try to “fix” the wrong-doers, and it’s only going to get worse.
If we do not police ourselves, others will do it for us. So my question unfortunately puts you on the hot seat: Do you, Dr. Jayne, have the authority to demand bad-actors make changes? If not you, then who are you ceding control to?
Maybe you should create a macro that says “Signed but not read,” as I was asked to do once by a doctor who was surprised when I thought the request was an attempt at humor.
You were right about the angry part. There is little we can do with technology to keep physicians from being lazy about documentation or venal, essentially charging for things they have not done. CMS/MU, while requiring a level of detail to justify charges that may be annoying, never force us to document something we have not done. What may help will be the shift away from volume-based reimbursement, removing the perverse incentive for excessive documentation (to justify desired, excessive reimbursement).
It is disheartening to see the alacrity with which many of my fellow physicians have surrendered what once was a bastion of clinical thoughtfulness: The physician’s note. I respectfully disagree with “ThoseAreMyCharts” who attributes the sad decline of physician documentation to “the do-gooders who believe things like MU are good”. In fact, you can qualify as a “meaningful user” under current rules without ever creating a single patient note. The CMS documentation guidelines are partly to blame, but ultimately every physician is responsible for the notes to which she or he puts their name.
The quality of physician progress notes is an an endemic issue, exacerbated by EHR’s. Several points:
1. Some provider notes were terrible before EHR; they’re just terrible, but legible now.
2. Copy/paste saves time, but doesn’t necessarily save lives; sometimes the opposite, as you’ve illustrated.
3. Macro’s designed and used well save time and can be clear and helpful.
4. Narrative in the HPI and Assessment and Plan are key to understanding the patient and the physician’s reasoning, during direct patient care. .
5. Templates and coded data (problems, meds, allergies) are important to having computer-understandable data for later analysis.
6. Too many other customers get in the way of physician-to-physician communication in progress notes (billing, compliance, incentives).
We’ve designed an electronic readability index, as have others, looking to score physician progress notes for their ability to communicate important findings. I think if we can begin to bring pressure through peer review, that is comparable to Medical Records review for compliance, or Billing review by the finance folks, maybe we have a chance. Otherwise, we (the physician colleagues) will be the last ones to the trough, and notes will continue to slide into Meaningless Use.
I don’t think I’ve ever ready anything on Mr. HIStalk that disappointed me as much as this article. This is evidence of a complete fail of the use of tools (point-and-click, voice-to-text, templates, and EHRs) as well as incentives and penalties that are driving physicians to think fastest/easiest is any kind of a rational goal for their work. I get that the federal pig trough to buy Epic was a powerful incentive, but the original observations of physicians in their own words is the raw data for the science of medicine. We’ve sped up the process of sharing data without considering that the data is no longer worth sharing.
What do compliance officers think about this article?
My current favorite is a urologist who sends me letters titled:
“Dear Dr: No Referring Doctor”
They obviously know who referred the patient, since I’m receiving the referral note.
My narrative axiom: a genius can do it well – but 2 or more geniuses create an incoherent mess.
So I may know someone who designed a “Data Kidney” that recursively elevates context of text (much like a real kidney establishes gradients of “cleanliness” of fluids.) But the last I heard, seeAmess didn’t go for it, and the developer doesn’t document in public domain to avoid getting ripped-off by big companies. (the developer’s generally warped. I couldn’t work with him.)
Ultimately, narratives are the bane of information transfer and coding’s worse, but it’s the best we have got – or the best that we are aware is possible without involving warped jerks. Yet the hard part seems to be resolving narrative issues without backfiring the process by involving more geniuses.
Notes visibility to the patient via an EMR portal can be a positive force. One not necessarily involving organizational mandates, legislation, or professional rivalries. Social forces can be a surprisingly strong motivator.
Not a panacea of course, just another tool in the kit.
Hate to sound Luddite, but there is nothing like the self-correcting real time logic of ol’ fashioned pen and paper setting down narrative. I fail to see how point and click transformed into “blocks” helps anyone. Except the folks responsible for receving the 25% of our health care dollars in non-clinical administrative overhead.