Clinical Case 133: POCUS Perils [and the Debiasing Dance]
August 23, 2016
I am a mad point-of-care ultrasound [POCUS] fan. I scan pretty much every patient that I can. I do this to sharpen my diagnostic skills, practice and to explore new utilities of the humble ultrasound. However, as my POCUS practice has matured and I have made countless errors; I am learning very slowly to do POCUS in the right order. What is the right order? Well, POCUS is a great tool. But like any diagnostic “test” it needs to be interpreted in the fullness of the clinical context.
Too often lately I have been catching myself doing “spot POCUS” on request from my colleagues. Usually in the middle of a busy shift a mate will call me over to their patient and say: “Can you look at this?” At which point I wander into the cubicle, introduce myself #hellomynameisCasey and scan the troubled area… This is BAD MEDICINE! The myth is that if you are a super-keen, proficient sonologist that you will find the underlying problem and Presto! However, what often happens is an inadequate, limited scan followed by some head-scratching and then a retreat into history and clinical exam… or onto CT which may / may not be necessary.
So my new mantra is: “Always start with the story, feel with your fingers, look at the things that the probe cannot see and then scan.” So today’s case is a cautionary tale. One which I feel illustrates the problems and perils with “SPOT POCUS”. OK onto the case….
It is late Sunday afternoon and everyone is keen to get home, and onto the beach. That is why we live here in the tropics. Just as I am tidying up the paperwork and sending my last few patients home my mate asks me to pop up to the inpatient ward to “have a look”. Of course, this means – scan. The story is as follows:
Jerome Jenkins is a 35 year old local who is a frequent flyer in the ED. HE drinks to excess a lot of days and can often be found sleeping on the waiting room floor. Jerome tends to be tolerated as he rarely causes too much fuss and is a happy drunk. Over the years he has suffered his share of traumas and complications associated with his lifestyle including pancreatitis, retinal detachments and a few seizures. Yesterday – Saturday – he came in very drunk. Blood alcohol = 0.450 and unrousable. This is not unusual. However, he woke complaining of lower abdominal pain. Although he is usually “turfed” to the Sober-Up shed on waking, this time his Obs were a bit off and it was decided to keep him in for some observation and work up if he remained in pain.
Fast forward to the next morning – and Jerome is now clinically sober and complaining of lower abdo pain. His belly is mildly distended and he has soaked up a bit of morphine. He is usually pretty stoic and lives a hard life – so this is atypical. His Obs are normal aside from a BP of 95/60.
Of course, I know none of this story when I get a call to come up and “do a bladder scan”. Jerome hasn’t passed much urine all day, he has increasing spuprapubic pain and seems a bit distended. The automated “Bladder scanner” used by the RNs on the ward is reading his bladder volume as 2100 ml! That seems a bit odd. So I was asked to come and do a measure…. so here is something like what I saw:
So I was very proud of my interpretation when I saw this…
The bladder is moderate size [measured at 400 ml]and there is free fluid in the adjacent pelvis.
Clearly I am sooo much smarter than the “auto Bladder Scan” machine! It was confusing extravesicular fluid for urine… dumb machine.
Now, being a very average doctor who was not in anyway “invested” in Jerome’s care. I made a quick verbal report to my mate. “He has a heap of free fluid… not sure why… does he have ascites?” To which my colleague replied: “yeah, I think so. His liver function has gone right off since we last checked. His ALT is through the roof..” In retrospect these are a pair of reciprocating confirmation biases… a perfect storm for disaster.
So lets go back and look at this story the way I like to nowadays. Lets perform my little “Debiasing Dance” that I have learned to do in the few years. Hat tip to Dr Rob Orman [ER CAST] and friends for this.
The Debiasing Dance is best done in private. The first step is to get somewhere quiet, a hallway or tearoom is fine. In honesty it often happens in the loo. The key is to get out of the clinical area where there is too much data flying around for just a few minutes. Try and reduce the case down to these bare basic elements:
Who is the patient? Fit ‘n young or old and frail. Immunosuppressed or at risk for other problems. This is another way of estimating the “pretest prevalence” of a whole raft of possibilities. In this case Jerome is young but at risk for lots of badness, infections and plenty of “unknowns”. The story is woefully incomplete. The last 5 times I met him he had been assaulted – so… trauma ought to be up there on the list!
What is happening right now? What are you / or the other clinicians worried? Why were you called to see them? Often this is because of abnormal Obs, a new symptom or clinical deterioration. Try and block out all that has come before; stuff that might falsely ‘explain’ the current status. The classic here is “Oh don’t worry about that lowish BP, he is always like that...”
What do we know objectively? We often attempt weigh a heap of information in these situations. Some of it is hard, objective and irrefutable. However, some info is soft, subjective opinion or based on hearsay. The former needs to be explained, the latter can be a source of bias. Try and weigh these appropriately and deal in the objective. What features of the case stand out when examined through an impassive lens.
The retrospective triad. I am not sure if somebody smarter than I invented this concept. Please let me know if they did. The retrospective triad involves running the case backwards and thinking: “The patient had X, Y and Z… hence it was obvious that the diagnosis was THIS.” In Jerome’s case if we run through steps 1 – 3:
Jerome is an alcoholic who lives rough and gets assaulted frequently.
He has free fluid – which is new as far as we know
He has poor urine output and lowish BP (usually his “low BP” is measured when he is extremely intoxicated in ED)
The retrospective diagnosis seems obvious, but of course it is usually clouded by a heap of irrelevant, misleading and biased data. Try and focus on the facts that rise above the fog. Then one needs to be imaginative and ask yourself “what if…?” as sometimes the third spoke of the triad is hiding in the haze of available information. Often this is as simple as going back to the bedside and asking that one pertinent question or just looking where others have not. For example, does Jerome have a massive bruise in his perineal area as a result of a well placed boot?
The real trick to doing the debiasing dance is to be able to imagine yourself looking back in a week and doing an M&M presentation (or writing to your MDO lawyer) and then apply that imagined hindsight in a prospective manner.
Most intraperitoneal bladder ruptures result from trauma to the bladder (often blunt) when the bladder is full. So the working theory is that Jerome was passed out drunk with a bladder full of beer when he was kicked / punched etc to the lower belly…. POP.
If you want to read more and see some pretty pictures of this injury then pop over to Radiopedia and check out a few other cases.
OK. Hope you learned something from case 133. Big shout out to Dr Nick Gilbert – my super Resi who demanded more clinical cases on the blog. I always try to keep the locals happy! Get your groove on and try the Debiasing Dance, worse case is you waste a toilet break and look a little silly if you actually dance in the tearoom.
I am a GP working in Broome, NW of Western Australia. I work as a hospital DMO (District Med Officer) doing Emergency, Anaesthestics, some Obstetrics and a lot of miscellaneous primary care. Also on the web as @broomedocs | + Casey Parker | Contact