Tuesday, August 6, 2013

If It's Not Good Information, It's Bad Information: Improving the Signal to Noise Ratio in ICU Communication

I learned over a decade ago a lesson that can be condensed into the following adage:  If the patient's family knows the creatinine level and the white blood cell (WBC) count, somebody is letting the intern do the talking.  And, (sorry, interns everywhere) that's not a good thing.

You see, interns don't know very much, just enough to be dangerous.  This derives from the fact that they have little to no meaningful experience.  They know what they read in Med I and Med II, and they can parrot their handful of attendings from key rotations during Med III and Med IV, but after that, silence.

What they mostly lack is experience which allows them to see the big picture and to know what the general course of a patient is likely to be.  When a patient such as the one in a previous post comes in, they can wax prolific about the FENA (fractional excretion of sodium) and pre-renal, intrinsic renal, post-renal, Bartlett's, Gittleman's, etc., but they probably don't know that renal failure requiring dialysis carries a mortality in the ICU of 60% and they certainly don't understand the contribution of poor functional status to prognosis in critical illness.  Because they don't teach those big picture things in medical school.  They teach biochemistry and physiology.  (Medical Educators everywhere, take note.)

An analogy would be:  you take your 1982 Honda Civic to the mechanic and the apprentice comes out and tells you that they're going to torque the head bolts to 80 Newton-Meters and fill the transmission with 750cc of whatever weight synthetic oil.  Who cares?  We want to know if you can fix it, how it's going to run afterwards, and how much it will cost.


Sadly, some physicians never graduate from the intern/apprentice stage of communication.  They will approach the family of a dying patient such as the one in the prior post, and after the conversation, the family knows how the creatinine is trending and what the white blood cell count is doing, but they have no idea about the overall prognosis.  This is a tragic failure of realistic and obligatory ICU communication.  There are several reasons why actual laboratory values and biochemical and physiological parameters should not be reported to patients' families, and should definitely not be the focus of any discussion about the patient's course:

  1. The creatinine level and the WBC count (as examples - there are countless others) are distractions from the big picture, as mentioned above.  If you're resorting to "objective" minutiae such as this, you probably are not telling the "big picture" story of four organ failures, or "advanced age, poor functional status, multiple comorbidities".  Or you don't understand the big picture yourself.
  2. There are the values that any literate person can report from the computer, and there is their interpretation.  If the creatinine is 4, it should be described as "his kidneys are still failing" rather than the reporting of a number such as a scribe or a reporter would do.
  3. People don't understand what creatinine and WBC mean, what their relevance in the grand scheme of things is.  And to try to explain them is frankly a fool's errand.  You cannot distill Med I and Med II and a nephrology fellowship into a 30 minute tutorial in the family meeting room and have it be useful.  Stop trying to do that.  You are creating a false sense of understanding. And it's going to backfire later, when the creatinine is improved but the patient is still moribund.
  4. The family, lured into believing (or allowed to believe) that these numeric values are the passkeys to the crystal ball into the future, will begin to obsess over them.  They will ask you and the nurses at every opportunity what the values are and when the next measurement is due.  They will be transformed into day traders during a stock market boom, and their emotions will be just as ragged.  But they will have no better understanding of finance or general market trends or the patient's prognosis than they would if they did not have this information.  If the information is just noise, it's bad information.  Don't give out bad information.
But even if you do communicate a realistic prognostic big picture, you will often be asked to report specific values to laypersons.  Resist the urge to comply.  Especially because the interval of prognostication should be 48-72 hours.  So the daily labs and vital signs do not add to the prognostication until that time period has elapsed.  (Which begs the question - why do we do daily labs anyway?)

Your job is to communicate what the family ultimately wants to know - namely, what is the outcome going to be?  If you have done this, they will be satisfied.  Or, they will not.  If they are not, but you really have communicated the big picture, it is because they don't like the big picture prognosis you have presented, and:
  1. They will mine for data that they think contradicts the big picture, such as the creatinine falling from 6.2 to 4.8 (after dialysis).  If this information were prognostically important, you would have already delivered a summary message of it.  Resist the urge to deliver this information and refocus the conversation on the big picture.  And the big picture is that dialysis is still being administered, and the kidneys are still failing, regardless of what the creatinine value at last measurement was.
  2. They believe, explicitly or implicitly, that frequent assessments of vital signs and laboratory values are prognostically helpful and that they, because of their vigilance at the bedside, with the help of a nurse who wants to interpret things in an encouraging light because this lightens her workload during the night shift, can identify important (and potentially unexpected and contradictory) trends before you do.  Dismiss this notion politely.  Frequent lab and vital sign monitoring is again like watching stock tickers during day trading.  It does not tell you the direction of the market.  It just serves to confuse.  So, preempt it by stating up front that the epoch of meaningful improvement in the ICU is 72 hours, not 2 hours.
The bottom line is that the laboratory values and physiological parameters are mostly meaningless.  They need context and interpretation.  If you have provided the context and interpretation, families will not know or want to know the values.  If they are not satisfied with your interpretation, they are groping for a different one, and they mistakenly believe that they can crunch the numbers themselves and find something you missed.  Complying with their request just serves to confuse them, and makes your job harder in the ensuing days.  So don't comply.  Explain what you (hopefully) know - that lab values and vitals trends over short periods are not prognostically useful.  And that they can have confidence that once a meaningful trend has been identified in 72 hours, you will sit down with them again and discuss the ramifications.

A physician's charge it to deliver signal, not noise.  If you're more suited for delivering the latter, you're in the wrong profession.  And there are several local used car dealerships that could use your services.

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.