By Damian Roland (@Damian_Roland)
Here is a challenge. Get your hands on something you can time yourself with: a watch, phone timer, etc. Time the task below…
https://twitter.com/Independent/status/679199711681601536/photo/1
How long did it take you? 30 seconds? A minute? Did you even find the panda?
Please feel free to post your time in the comments section at the bottom of this blog (and don’t look other peoples’ times first – that’s like looking at the radiology report and then finding the fracture!)
I am not really interested in the time taken, though. What I’d really like to know is how did you go about the task? Did you rapidly scan for things that didn’t look like a snowman? Did you systematically move through each row to find a panda? These are subtly different approaches. Sadly, I don’t know the best approach, as once you have seen the panda it is impossible to un-see it!
Being able to find specific information hidden among a large amount of redundant ‘noise’ is a challenge. Whether it is completing a newspaper quiz or diagnosing a febrile child who may have sepsis, understanding the process of decision making is important. Its important because it may lead to personal improvements in our own reasoning. Also you need to be able to teach others what is best practice and what leads to trouble.
This blog was ultimately inspired by my daughter who has just started playing Guess Who.
Guess Who is a classic game where you must ‘guess’ which person your opponent has selected from a sheet of 24 people by ruling out certain characteristics in their appearance. (Do they have moustache? Are they bald? etc). “Are they are man or a woman?” is a common first question When my daughter seemed to have understood the rules we started our first game. It went something like this.
Isla: “Is it Anne?”
Me: “errr. Yes”
Isla: “Yes!!!” (runs off to tell her sister about her victory)
Me: “But that’s not quite how it works…”
By getting the answer correct on the first guess she had bypassed all my carefully constructed advice and once again made a mockery of my alleged skills in education.
A colleague of mine Dr. Rachel Rowlands (note different surname, not a relation) uses a similar analogy of a word search. In the table below can you find the word sepsis? (any direction – horizontal/vertical and diagonal)
What process do you use? Personally I scan for ‘S’ and ‘E’ together in close proximity; you must have these letters to form SEpsis. But, do you go across rows or down columns to do this? Not sure. I do think though that trying to find the whole word in one go is probably not an effective strategy. You could start from the back and looking for ‘I’ and ‘S’ in close proximity, but this may increase your chance of a dead end compared to ‘S’ and ‘E.’
Let’s put this in a clinical context
Junior Doctor: “Dr. Roland.”
Me: “Damian, please.”
Junior Doctor: “Doctor Roland, I have this child who I think is really unwell. They have had a fever, have had a couple of vomits and have been a bit drowsy. I think they have severe sepsis. I am just going to put in a cannula, do some blood tests, and give some antibiotics.”
Me: “Can I see the child?”
Child playing happily and eating crisps.
Me: “This child does not have severe sepsis.”
(This is a made up scenario to make a point and doesn’t reflect my general feedback style before people write in to complain!)
Almost universally, well appearing children who are eating a packet of crisps do not have sepsis. My initial scan of the patient removes lots of possible diagnoses to consider, whereas the junior doctor quite correctly identified features which could be sepsis. I looked for the “S” and “E” and couldn’t find it, the junior doctor looked for the “I” and the “S” and did.
The hypothetico-deductive technique has served medicine for centuries. It will continue to do so. But we need to continue to examine and understand how and why decisions are made based on limited, available information. It is simply a matter of understanding how you approach a word search…
Guess who image courtesy of Jedudedek, via Wikimedia Commons
