Tuesday, June 10, 2014

Stop!... Shovel time!



 While catching up on reading last weekend I came across this in a medical journal:


"The goal of the project is to supply an interactive system that translates vast amounts of data and scientific literature into insights that professionals can consult to inform their treatment decisions.

"A database, in this case big data, provides the foundation for the potential to use state-of-the-art analytics to generate truly actionable insights."


WTF does that mean?

This is the problem with modern buzzword bullshit. It is, like Macbeth said, "full of sound and fury, signifying nothing." Yet, medical journals are full of similar crap that tells you zilch.

EHR today is the same way. 5-page notes that automatically fill in what medications someone is on, what their allergies are, what their blood pressure is... yet only rarely do you find anything comprehensible telling you what the physician's impression and plan are- WHICH IS THE MOST IMPORTANT PART OF THE NOTE. Usually it's hidden in the ICD numeric codes, and oftentimes the "plan" says something like "see orders database." Which doesn't help me at all.

Quantity has replaced quality in medical writing, and the problem shows no sign of getting better.

The above collection of horseshit, BTW, was from an article about epilepsy treatment.

20 comments:

  1. The first sentence that you quoted is in English, more or less. The second one is an example of management aphasia.

    Management aphasia is the use of words without attaching meanings to them. Dilbert's boss might ask for an "object-oriented database" and even get one, but he doesn't know what one is, and he doesn't know that he doesn't know.

    ReplyDelete
  2. Sounding board: medical obfuscation: structure and function.

    Authors
    Crichton M.
    Journal
    N Engl J Med. 1975 Dec 11;293(24):1257-9.

    Affiliation
    PMID 1186808 [PubMed - indexed for MEDLINE]

    This article was written by the author michael Crichton as a resident.

    ReplyDelete
  3. "...article about epilepsy treatment". The horse in the foreground must be the real horse's aura.

    ReplyDelete
  4. yep yep. I always scroll down through the BS in the electronic chart and look for the section titled "Plan of Action". That is where all the good stuff is.

    ReplyDelete
  5. ". . .truly actionable insights."

    Different profession, but the above screams lawsuit to me.

    ReplyDelete
  6. I agree completely. The whole narrative is lost. You can't describe a person in checkboxes. And then it's impossible to find the action plan. There's no flow. It's an inefficient way of telling a story or sending a message. Frustrating!

    ReplyDelete
  7. I'm a software engineer who works on clinical trial software, so maybe I can shed a little light on the objectionable line.

    Basically, the worst possible thing from an information processing perspective is unstructured data (like notes). We don't really have any good techniques for handling it, so software developers will make users do really stupid things to avoid having to deal with it. That said, structured data (like stuff that uses ICD codes) means that we can apply a whole raft of tools that we've developed in other fields to take a look at very large sets of data (whole populations worth) and find correlations that might otherwise take a long time to show up (like say certain sub-populations really, really shouldn't take Vioxx).

    Certainly the current situation where everyone is using their own artisanal EHR (none of which can talk to each other) won't get us there, but at least that's the idea behind the buzzwords.

    ReplyDelete
  8. Same thing in vet med. When I ask for a record to be faxed over for a new patient, I get an itemized list of medications and procedures, but no actual physical exam or client communication. It's ridiculous.

    ReplyDelete
  9. You need to synergize a plan with your team members, for your team leaders feel that you are insufficiently engaged in the paradigm shift that presents a patient-centric holistic approach. If you are not invested you may be downsized , you must be proactive lest you miss the sea change that will leverage our resources and allow for organic growth of the profession. It will be a win win.

    Sorry Doc, the entire world speaks bullshit, that is why plain talkers are refreshing. Just listen to the DC folk.

    ReplyDelete
  10. I think that the plan, most of the time, is to collect more data.

    ReplyDelete
  11. To any physician out there who would love to hire an old-fashioned medical transcriptionist, here I am!! :o)

    It's despicable that EHR is both not helpful to you - but is killing the career that I dearly love.

    Dr. Grumpy, I love your blog!!!

    ReplyDelete
  12. the purpose of the medical record is to help in patient care, not to analyze large data sets.

    ReplyDelete
  13. I HATE EHR. I used to work for a busy 8 doc Opthalmology practice with EHR and I can't tell you how many patients were on "pred" year after year. When I brought it up at a staff meeting everyone just stared at me. "It's just the computer" was the excuse. And, as you said, you have to dig to find Dr's findings, impressions, etc. Old hand written chart notes used to be primarily that.

    ReplyDelete
  14. Yes, unfortunately, if it doesn't sound "sexy" they figure it won't sell.

    My favorite part of our EHR is that it has few different search methods and it's divided on the front, so you can see the appointments, click and there is the note attached.

    You can deviate around some of the stuff, which is the pain, especially with our allied health staff (they have appointments for their workload, but put the notes elsewhere).

    So much for interdisciplinary communication! We (MDs, RNs, etc.) write notes EVERY day, and they do one once a week on certain therapy patients.

    Just not right...

    ReplyDelete
  15. "Anonymous Anonymous said...
    the purpose of the medical record is to help in patient care, not to analyze large data sets."

    Not anymore. Data sets provide meaningful use...it will allow them to decide what drugs are too expensive for the 20% who don't fit the mold...who gets what, if any, treatment, based on the data sets...outcomes....it's beyond socialized medicine...ultimately...

    ReplyDelete
  16. When I was first married, I wanted to send letters and notes to my mother-in-law who doesn't speak nor understand English, and I neither speak nor understand (nor can read) Japanese. Nowadays, I use Google translator. I enter my text in for a direct translation, then I reverse the translated text from Japanese to English, and find some cheap hilarity. (I think someone might say 'you get what you pay for' with that.) After each back-and-forth translation attempt, I find myself resorting to simpler and simpler pidgin English two-word sentences using present tenses and rudimentary descriptions of emotions, and lots of photographs.

    ReplyDelete
  17. Notes in the chart used to be the way that the various medical professionals involved in the care of the patient communicated with one another. Now the chart note has two functions: one, justify the billing you have submitted, and two, cover your ass in case of a lawsuit. If you want to try to communicate, fine; just don't get in the way of billing or ass-covering to do so.

    ReplyDelete
  18. Doc Rugrat: The irony of CYA in medical records is that, in my experience, the more people try to write things up with the main focus of avoiding a lawsuit, the more they shoot themselves in the foot if there is a lawsuit. And I say this a lawyer who worked EMS briefly in undergrad. The smart thing is to document what actually happened in a way that would remind you enough to properly explain what you did and to defend it. Instead, you have someone with little or no legal knowledge coming up with policies based on anecdotes from other people who also know nothing about the law.


    ReplyDelete
  19. It's cool to be able to search a database of your patients (or patients in a healthcare system) and pull up the entire population who has a certain diagnosis or is taking a certain drug. You'd better have a damn good reason for doing it for HIPAA reasons, but it's cool for whoever does that research and analysis.

    But yes, we absolutely need people actually leaving notes and explanations of how the patient presented, how the visit went, and what your professional opinions are.

    Also, trying to have a computer system come up with "truly actionable insights" leans way too close to dangerous IMO. This can be skewed by a stingy insurance company or hospital to punish doctors who prescribe expensive medications or therapies, or it can make providers lazy, since all the information is right there and they don't have to think critically about the patient's presentation.

    E-prescribing, OTOH, is a godsend and should definitely stay.

    ReplyDelete

So wadda you think?