January 9, 2009

The Chart Is Dead, Long Live The Chart

I. In the Annals of Emergency Medicine appears a short article called, The Chart Is Dead- Long Live The Chart.

The article explains how the chart has regressed from a place to write down thoughts and ideas ("writing as thinking") to a fee extraction device, simply templates specific to reimbursement guidelines, and not the delivering of care.  This means that the chart does not aid in the work of the clinician; I'd add that it actually becomes the work.  The chart is dead.

Summarizing some of the author's, and my own, observations:

  • Default terms and ideas which go/don't go into the chart-- so all charts look pretty much the same.  There is barely anything in the record that distinguishes one patient from another.   It's as if privacy laws are superfluous.
  • Unreliability of the information. On day 3 of a hospitalization, does "+SI" mean he really had SI, or is that just there for the insurance?   Does PERRLA mean P+E+R+R+L+A, or just "Patient had two Ps?"  This isn't a comment on clinical rigor.  Let's assume doctors are doing a thorough job regardless of documentation.  The question here is that since PERRLA doesn't actually mean P+E+R+R+L+A, why document it at all? Who are we really trying to fool? 
  • Research by retrospective chart review is therefore invalid.  It is unreliable, misleading, and worse than useless.  Please read that 100 times.
  • The shadow chart-- and any resident will know exactly what this means.  "Sign outs" or "the list," three lines long per patient but substantially more focused and useful than the entire actual chart.  Often, information is there that would never  dare be included in the regular record: "mother likely borderline as well."  Lawyers have no idea what they're missing out on; and, here's a secret-- the sign outs are usually typed on the same ward computer.  Let fly the subpoenas.
Since doctors believe the chart to be a billing form or evidence in a malpractice case, they are not using it as a diary of clinical care.  The irony, of course, is that both insurance companies and lawyers pretend that it still is the very thing that they have caused it not to be.

II.  How did it come to this?

Another article in the Annals of Internal Medicine, entitled "The misinformation era: The fall of the medical record"  also criticizes the uselessness and unreliability of the chart.  The author observes that clinicians are often afraid to document things that might upset the patient:

Physicians have become cautious of writing that a patient is paranoid, flirtatious, hypochondriacal, verbose, or homosexual. Possible demeaning comments about the patient's manner, dress, speech, level of education, and intelligence are also omitted. 
If flow-chart medicine is what we strive to practice, than let's imply dispense with charts-- and doctors-- fire up the Hippocratobot, and start billing Medicare.  Otherwise we're going to need to accept that, in order to be effective, a doctor is going to need to know even the things which are "irrelevant," precisely because they are not.  And he's even going to need to write them down somewhere.

The rest of the article will resonate with any clinician:

In view of the flagging integrity of medical record information, it is particularly galling that medical records have been accorded such authority in our society. Physicians may spend as much time with records as with patients. Medical records dictate whether and how much physicians and hospitals are paid... Records determine our patients' insurability, job qualifications, credit ratings, disability awards, and retirement. They are central in malpractice litigation...

...But to restore the integrity of the medical record is more difficult. The manipulation of chart information physicians go through to protect private information about patients from public view and to satisfy the cost-control regulations are demoralizing and degrading to medical professionalism. This subterfuge will continue until confidentiality is restored to medical records and the regulations are revised to accord with the realities of practice, rather than requiring clinical medicine to be a mechanistic clockwork technology. In the meantime, all medical record information should be regarded as suspect; much of it is fiction.
Boy, things are getting worse.  Actually, no: this article was written 20 years ago.

III.  So if things have been this bad for a while, what is the real problem?

The charts haven't gotten worse, our ability to think has gotten worse, the chart now simply reflects that.   I know we believe we chart differently than we practice or think, but the two are very connected; what you write causes you to think a certain way.  Since we are not writing down dating history, therefore we are not thinking about dating history, it becomes secondary.  We're instead thinking about about "mixed episodes."   Get it?  We're trying to reduce "dating history" into a symptom cluster.    What's on our minds first is Lipitor and LDL, and only secondarily quit smoking and eat better, because we document Lipitor and cholesterol, our charts contain places to write those down.

"But isn't LDL important?"  See? You've been contaminated also.  The number is meaningless beyond low, normal, and high; and those three have relevance only as they correlate to other pathologies.  But you have information bias, you feel like you need to know the actual number, you feel like it tells you something, you feel like "normal" isn't informative enough.  And I'm telling you "normal" is more informative because your mind is not cluttered with the uselessness of "one hundred and nineteen."

Knowing that certain info should go into the chart (Lipitor) causes you to think about that primarily.

The second problem is this: the doctor is required to do two entirely contradictory things.  First, he is supposed to do what he thinks is best for the patient.  Second, he is asked to document the facts of his prosecution.

No one looks bad if they don't write, "told to eat carrots."  But you look sloppy if you write ""LDL high." 

In other words, the medical chart is a college application: here's what I did, to the best of my embellishment, knowing that you don't care about what I actually did, but rather about the things you think are important-- all of which I did only because that's what you want from me.  Can I get paid/dismissed now?

IV.  You're kidding, right?

No.  Maryland State Board of Physicians v. Eist, 2007.

A man-- not a patient-- complains to the Medical Board that Dr. Eist overmedicated his wife and son (Eist's actual patients), which caused (supposedly) his wife to become psychotic and the boy extremely anxious.

The Board, trying to get to the bottom of it all, told Eist, and I'm quoting,

deliver immediately upon service of process a copy of all medical records of [the wife and son]; treated at your facility; which materials are in your custody, possession or control.

His response was, and I'm paraphrasing,

bite me.

Slow down, Souter, Eist is right: the complaint was filed by the husband-- he's not the patient; and he was in the process of divorcing his wife and demanding custody of the kids.  Get it?  No?  Once the Board gets the records, the man would then be able to demand those records from the Board as evidence for his custody hearing: "my wife isn't fit to care for her kids."

Well, the wife eventually allowed Eist to release the records to the Board, so he did.  In thanks, the Board slapped him with a $5000 fine for not having complied immediately.


What do you think Eist wrote in that chart that he didn't want the husband to see that-- and this is important-- the husband couldn't have obtained from other sources? The husband has  empty prescription bottles in the house; he has insurance bills.  He knows she went to a shrink. Do you think Eist was doing dream work with her, and documenting it?  Of course not.

The state board, in its brief to the Court of Appeals, argues that the lower courts failed to weigh appropriately the state's need for patient records when investigating alleged medical misconduct and the doctor's ethical requirement to preserve patient confidentiality. The records sought from Eist concerned primarily the drugs and dosages prescribed by the physician, data which medical agencies routinely collect as part of their compelling interest in protecting the health of all residents who seek medical care, the board states.

The Board itself expected there'd be nothing useful in the record.

Eist's, and all doctors', dilemma is this: if he has written what he really wants to write in the chart, and the husband gets hold of it, he's damaged his patient and can even be sued for doing so.  So he shouldn't give the Board the records.  On the other hand, he's punished for not giving the records-- because the Board expects that he/you are smart enough not to put anything in those records, so why wouldn't you give them up?

So then why would he/you write something in the chart that, if it fell into the wrong hands, could damage your patient?  You wouldn't.  Are we supposed to honor the confidentiality of medical records but assume they can be read by anyone at any time?  Apparently yes.


Doctors have to document for billing and for malpractice, but at the same time not so much that it would damage the patient in the event of accidental disclosure.  If these are the things on his mind as he is writing the chart, it is impossible for these things not to be on his mind as he practices medicine.  To the exclusion of other things.   It's not just that it takes up time; it completely changes the way you think.

Example:  They'll write in a progress note, "Lipitor 10mg PO qd."  But they'll write that information also in the section called "Medication Record" and also photocopy the prescription.  But no doctor would ever write only "Lipitor" in the progress note, because that seems incomplete.  Try it, see how weird it feels.  Is triple documentation useful?  No.  But it triply reinforces that that information is very, very important, while things you don't write down-- "needs to eat two carrots a day and one less donut"  is less important.

How we chart makes us worse doctors.

VII.  "It maybe makes you a worse doctor, but you're a jerk anyway.  I'm not like that.

Let's all agree that I am not a person short on thoughts.  But I notice in my own practice that I don't think about things which are later obvious to me when I'm listening to music not writing in a chart, because my pattern is to write certain things down.

Life is one big Stroop test.  Don't read the words-- just say out loud the colors you see:


It is very hard to attend to what you think is important when you are faced with competing informaion.  Note that the only way to really succeed at this task is to avoid reading the words.

These are just stupid colors and words.  Do you really think you can block the effect of thrice documented Lipitor in your practice?

VII.  What's coming?

Easy answer that no one will like.  Since it is evident that chart serves no useful purpose to the clinician; and it only serves a negative, or limiting purpose vis a vis insurance and litigation, and it adversely affects the way we think about patients and the way we treat them, it will, inevitably, be abandoned.  What will replace it?  The only thing that is objective enough to serve both science, billing, and forensics:   video.

Which is another reason why I am retiring.