Old Incident Reports, Part One

I was going through an old folder found in a desk drawer in my office and found a bunch of old incident reports. When I say old, I mean at least 10 years old. I opened the folder and blew dust off it like Indiana Jones. Old. Reading some of the reports put a smile on my face and reading others made me embarrassed. Deeply embarrassed, as a matter of fact. Reading some of those reports now, I feel like I’m lucky to still have a job.

One of the reports is my description of an ambulance crash. If I remember correctly, it is my most recent crash. I can't think of another, more recent one. Let’s pause a second and acknowledge that it’s pretty impressive, considering my last grinder was in 2003 or so, huh? I used to think occasional crashes were “just part of the job.” I was involved in a crash per year or thereabouts. I mean, when a medic drives x-thousand of miles per year, sometimes the dice come up with an unlucky number. Impossible to avoid, right? Nowadays I think occasional crashes may have been related to me driving like an asshat. But then I came across the report for this crash. It doesn’t seem to have been especially avoidable, so I don't know what to think. Maybe I have been lucky for the last decade.

Way, way, way back in the day, I was stoked to be working with one of my favorite partners. This was in 2003 or so. Early in the shift, we were assigned to a report of chest pain. I was driving and I flipped on the lights and siren. We were rolling on a huge four-lane road (four lanes northbound and four lanes southbound). I approached an intersection where I wanted to turn left. The light was red and there were cars in the number 1, 2, and 4 lanes. The three lane was open. I came up to the intersection, stopped (mostly), ensured the intersection was clear, and made my left turn through the red light from the three lane.

Crunch.

Yep. Ambulance crash. But it is not what you think. What happened was that the guy in the number one lane waited until I was in front of him before running the red to crash into the side of my bus. He up and plum rammed me. For no reason. Nobody has any idea why he did it. Including him – he said so to the police officer and to my supervisor. I have to admit – he was cool about it. It is nice when the other dude says the crash was his fault, in those words. He took all the blame. I mean, it was his fault and all, but a ton of other drivers would try to make excuses or stretch the truth. So good on him.

The worst part of the crash wasn’t the crash itself. I wasn’t at fault, the damage was minor, and nobody was hurt. The worst part is that they put my partner on another car while I was getting the crash investigated. When all of the investigation and paperwork got done and I swapped out ambulances into an uncrashed bus, I got a new partner! No! Crap! My dream of a cool shift with the partner I liked was dead! The partner I wanted to work with was across the city in another unit. Sadness…

At least crashing meant I could attend the rest of the shift. 

So it is possible that some accidents are unavoidable. It could be that all of my crashes are secretly, quietly unavoidable. Even the one where I hit a curb going about 40mph in the snow and got all the wheels to point in different directions. Even the one where I backed into a fence – completely unavoidable, right? After pondering it, I think many “unavoidable” crashes are due to poor choices and aggressive driving. There are certainly some ambulance accidents like the one described above that are accidental, but there are others resulting from a string of suboptimal choices with a crash at the end. 

I don’t know. I need to think about this some more.

Another Problem, Part II

In yesterday's case, I told you the story of how my partner and I were running an unresponsive gent with a believable report of seizure, obvious evidence of frank GI bleeding, hypotension, tachypnea, and hyperglycemia. I ran a 12-lead before departing to the hospital:

By ArcGenus41 (CC BY-SA 3.0, via Wikimedia Commons). Click to enlarge

When I saw it, I said aloud: “Oh, that’s good. Just what he needs. Another problem.” My partner glanced up from taping down the IV line and immediately saw the relevant ECG finding, as well. Did you see what we saw?

Hypothermia.

Specifically, we noted Osborn waves in pretty much every lead. The end of the QRS (or the beginning of the ST segment, depending on how you look at it) gets a sharp waveform that some people compare to a fishhook. I don’t see the fishhook shape, but I can see Osborn waves. Here are some more examples that I found online*:

By Jer5150 (CC BY-SA 3.0, via Wikimedia Commons) Source

By WikiSysop (CC BY 3.0, via Wikimedia Commons) Source

The height of Osborn waves are supposedly proportional to the degree of hypothermia, but I've not seen a solid study to back that statement up, so I take that little factoid with a grain of salt. Life In the Fast Lane has some illustrations of worsening Osborn waves, though. According to a brief article in the journal Circulation, hypothermia is initially associated with sinus tachycardia. As the core temperature drops below 32.2°C (normal body temp being 37.0°C or thereabouts), sinus bradycardia is more common. In addition to the bradycardia, all intervals commonly lengthen. As the temperature continues to fall to the range of 30.0°C and below, atrial ectopy and atrial rhythms like AFib can appear. “At this level of hypothermia [<30.0°C], 80% of patients have Osborn waves…”  

Keep in mind that ECG findings aren’t the best way to measure a patient’s temperature, just like ECGs are sub-optimal for searching out electrolyte findings. The best way to find a temperature is with a thermometer. But recognizing ECG findings can put you onto a differential diagnosis that you may not have thought of otherwise. Like in the case we were discussing here…

My experience with hypothermia is usually generally similar to this case, in that this fellow wasn’t in a cold environment. The take-home point is that hypothermia is less common during raging blizzards than it is in settings in which you wouldn’t necessarily expect cold - like seizing indoors. Even the worst urban outdoorsman usually finds shelter when it gets hard-core cold. But I have seen hypothermia in early September when a college student got a little drunk and passed out in a park. The sprinklers turned on, he got wet, and then he got cold. Another guy tried to cross a river and got wet. The coldest patient I have ever been involved in resuscitating had a core temp in the teens (I want to say 17°C, but 14°C sounds right too). He had gotten intoxicated and fell asleep in an alley. In that case, the night was admittedly pretty cold but he was well-dressed for it. He lived, by the way. 

The patient in this case study had a core temperature of 28.4°C. He died in the hospital the next day. His medical issues were complex and interrelated. 


* You should also do an image search on the Google machine for Osborn wave and check out a ton of examples that I don't have permission to share.

Just What He Needs. Another Problem...

A few years ago, my partner and I were sent on a non-emergency response to a cheap residential hotel-apartment-type place on the report of a seizure. In my system, when a patient has a history of seizures but has stopped seizing it is not usually an emergency response. In any case, we arrived and were met by the building manager. On the way to the third floor of the other side of the building (because of course that is where the call is), the manager explained that he had let himself into our patient’s room to do some maintenance. He found the patient on the floor having a seizure. (This is the kind of place where one could reasonably expect the manager to be able to recognize a seizure. This wasn't his first rodeo.) The manager’s description of the activity certainly made it sound like a generalized seizure.

We arrived to find that the patient’s room was big enough for a bed, a television, and a hotplate. There were a pile of clothes in one corner and a pile of empty liquor bottles in another. Let's just say the maid hadn’t been to this room in a while. The patient was a male in his late thirties, but looked like he was older. He was supine on the wood floor, and he had indeed stopped seizing. His eyes were open, but he was unresponsive to any stimuli. His skin was pale but dry. There was circumoral blood and an old puddle of bloody vomit next to him. There was also bloody diarrhea in his drawers. After checking for other clues in the apartment (I found a half-full bottle of Seroquel and no other meds), I got to work setting up extrication (a probable hassle due to the previously mentioned frankly bloody diarrhea) and my partner palpated a quick set of vitals: 80/p, with a heart rate of 80 and a deep, sonorous respiratory rate of 22. We placed a nasal trumpet to stop the snoring (successful), moved the patient to the pram without getting anything on me (successful), and moved to the ambulance (also successful).

So where were we? Seizure, check. Gastrointestinal bleeding, check. Hypotensive, check. Unresponsive, check.

In the ambulance, my partner began working on an IV line while I placed a 12-lead and pulse oximetry. During the IV placement, my partner checked the patient’s blood sugar and found it to be 488. I didn’t find any sign in his room that dude was a diabetic. Another issue for the patient to overcome, check.

The patient was satting at 100% via the non-rebreather mask and the NPA and I got this 12-lead:

The quality is not good on this ECG, but you can still see what you need to see.  Click to enlarge

When I saw it, I said aloud: “Oh, that’s good. Just what he needs. Another problem.” My partner glanced up from taping down the IV line and recognized the significant finding, as well. 

Do you see it?

Hack Your Brains - In Advance

A few years ago, I read Blink: The Power of Thinking Without Thinking, by Malcolm Gladwell (Little, Brown & Company, 2005).  At the time, the book struck a chord with me.  The main concept explains how people, especially experts, can tell at a glance what is going on and make accurate decisions.  Gladwell shares examples of this, including art experts who immediately know a statue is a forgery, psychologists who can tell if a marriage will break up in three minutes, and another psychologist who can understand microexpressions.  Experts do this by “thin slicing” available information, using relevant data quickly and discarding the other, useless parts of the whole picture.  The ability is affected, Gladwell explains, by other brain hacks such as priming.  The main point is that clear thinking can be overwhelmed by irrelevant information, and experts learn to ignore irrelevant information and trust their intuition.

This appealed to me because I had seen this way of thinking in action.  I’ve seen paramedics glance at a patient and know what was wrong within 5 seconds.  I’ve seen firefighters subconsciously know a roof was unstable, without being able to specifically explain why in a reasonable way.  I’d seen doctors know when a patient was hiding relevant parts of their story.  As an expert in EMS, it appealed to my self-image to imagine I could consistently do this.

The problem with Blink is that expert-level knowledge is required for successful thin slicing.  John Gottman, the psychologist who could predict divorce at a 90% success rate after three minutes of watching a video of a couple’s interactions, used extensive science, methodology built up over the course of years, and reason to be able to do the trick.  Paul Eckman’s microexpression Facial Action Coding System doesn’t use thin slicing – the system fills a 500-page binder.  Another problem is that incorrect thin slicing happens all the time and leads to dreadful outcomes.  It could be argued that some police shootings of unarmed people are examples of horrifyingly incorrect thin slicing.  The third issue with the book is Gladwell doesn’t explain how this works or how to improve the reader’s ability to perform this skill.  How does an expert mind accomplish what a novice mind cannot? 

Finally, I have a “Moneyball problem” with the ability of experts to consistently make accurate intuitive decisions.  If experts were especially accurate, then baseball scouts (experts) wouldn’t need statisticians (non-experts relying on data) in baseball player evaluations.  Slower decisions based on reasoning and data can be as strong as blinking out an intuitive answer.  When I first read the book, I intuitively loved it.  After further reflection, I think Gladwell’s Blink is probably an example of reporting bias – include the hits and don’t mention the misses.

Blink popped back into my mind because I finished reading Thinking, Fast and Slow by Daniel Kahneman (Farmer, Straus and Girous, 2011).  Kahneman and his late collaborator, Amos Tversky, are/were psychologists who specialized in investigating why people don’t always act in ways that rationality would dictate they should.  There is an imaginary creature, Homo economicus, which is used as a model in economic discussions.  This creature always acts perfectly to enhance its interests and economists can model the behavior.  People don’t act perfectly in their own interests, however.  Studying why they didn’t led Kahneman to win a Nobel Prize in 2002.  In economics.  He is a psychologist.  Badass.  (Tversky would probably have won too, but you have to be alive to win a Nobel.)

Thinking, Fast and Slow explains how the brain works with two systems.  System 1* is the intuitive cognitive system.  It makes snap judgments.  System 1 is effortless and automatic.  2+2=___.  Black and ______.  System 1 gave you those answers.  System 1 is also affected by emotion and associating images together.  When you associate a song with a specific time in your life, or a smell with a specific location, it is probably System 1 at work.  System 1 is the most common use of the brain.  Expert use of System 1 is what Blink was talking about.

System 2, in contrast, is the conscious, thinking, rational mind.  It is the system used to figure out 24x37=_____.    System 2 can read a map.  System 2 solves big problems in a rational, logical way.  The problem is System 2 is slow and lazy.  If you don’t force System 2 to work, it won’t.  As proof, I offer the fact that most of you didn’t expend the effort to find out 24x37=888. 

Too much data!  System 2 is out! Source

Here is the problem.  System 1 is often wrong.  System 1 makes bad decisions.  System 2 is pretty much consistently right, but is lazy and hard to turn on consistently.  Think about buying a car: You can figure out the prices, spreadsheet payments, work out pros and cons of models, and work out which is the best car.  But System 1 deeply and truly likes the red one.  People end up buying the red one.  Corporations prey on System 1 all the time in advertising.  Las Vegas is built on System 1 thinking (“This machine is about to get HOT!”).

You can almost see the spot that is missing System 2... Source

Why is this on an EMS blog?  Because System 1 has problems providing consistently good medical care.  When a drip rate calculation is needed, you turn to System 2.  System 2 yawns at you and rolls its eyes.  System 1 waves its arms above its head and says, “Oooh, oooh!  I know!  I know!”  System 1’s idea is to simply turn the drip on for a while to get the desired effect and system 2 can figure out the drip rate later when we are writing the PCR. 

System 1 freaks out at MCIs.  System 1 causes us to work dead people who should be pronounced.  System 1 causes us to skip the secondary because the patient smells bad.  System 1 finds a homeless guy in an alley and “thin slices” that he is just drunk, rather than checking other reasons for altered mental status.  System 1 is where errors happen.  So what can we do to strengthen System 2 in our professional lives?  It doesn’t work to say: “I will try harder.”

To me, the trick to activating System 2 is to turn 2 into 1.  Make your rational, reasoned, calculated decisions into intuitive ones.  We do this by working them out in advance.  This is the strength behind preplans, SOPs, and protocols.  Exercises prime System 1 actions through advance System 2 work, as well.  Practicing rare events until our actions become instinctive is switching System 2 thinking into System 1 thinking. 

It is also important for protocols and SOPs to not rely on System 2 thinking.  Protocol authors should strive to get rid of the potential for error.  Require less thought.  Don’t let System 2 yawn and roll its lazy eyes when you need it to work.  For example, my treatment protocols don’t call for drip rate calculations anymore.  Take epi: my protocols call for 1 milligram of any concentration of epinephrine into a liter bag of saline, run wide open and titrated to effect.  In advance I know the concentration of epi in the liter bag is 1 microgram per milliliter (System 2).  But I don’t need to think through it when with a sick patient on a drip.  I need to inject the med and open the line (System 1). 

Airline pilots don’t rely on their memories when using checklists and emergency procedures.  They pull out a book.  Trying to rack your memory for the emergency procedure while in the emergency is calling on System 2.  Having your copilot read the pertinent list of actions is using System 1.  This is also the idea behind the Handtevy pediatric treatment system – figure it all out in advance and then look it up.  Don’t change your treatments between adult patients and pediatric patients; simply open a book and look up the smaller doses.  Switch System 2 thought processes to System 1 thought processes.  We know adult treatments so well that we are working in System 1 when we are treating adult patients.  Pediatric patients, especially sick ones, freak us out and we didn’t memorize all the dosages well enough to cut through the fog.  The need to recall pediatric treatments fall on System 2.  Don’t depend on System 2 for your success.  Switch 2 for 1 and get to work.

Take the time to recognize where system mistakes occur and fix them in advance.  Make your job simpler, so you don’t need to rely on System 2 while under pressure.  Don't get me wrong - System 2 is not the problem.  Don’t get me wrong.  System 2 is what allows us to calculate the trajectory of a probe we’re shooting at Pluto, after all.  But take the strengths of System 1, especially the speed of System 1, and combine that with the strengths of System 2.  You do that by working out your actions in advance.

*Yeah, I wish he had come up with more descriptive system names, too…