Clinical Brain Training

ClinicalBrainTraining.com

Archive for category Discussion

Recognition primed decision making

Sep 16

Posted by Jeff Young, MD in Clinical Teaching, Discussion, EMS, Hospital | No Comments

In the paper I posted by Gary Klein, they show the evidence, and discuss the background to their decision making model known as “recognition primed decision making”. I know many of you are not going to take the time to read the 100 page manuscript (but I highly recommend it, I’m rereading it now and its great), so I’m going to go over some if it. Sorry  I haven’t my podcast on this, but there have been some scheduling issues.

First, Klein says (and I agree) that when an experienced operator, in whatever high-stakes field, encounters a situation (fire, patient, etc.) they do the following:

– Size-up the situation in their mind, assigning it a “schema” based on their previous experience

Explanation: In your mind you record experiences you have had in your profession. With time you categorize those experiences into “boxes” or “schema” of similar situations. In other words, when I see a 80 year old trauma patients on inhalers at home who has 8 right sided rib fractures, I don’t make my clinical decisions based on each of those datapoints individually (elderly+/COPD+/-rib fractures+/- etc.), do it by matching the case to schema I have already created. For this it would be: “old patient with lung problems and big non-lung chest injury”. Then my mind automatically jumps to what I have done for “old patient with lung problems and big non-lung chest injury” in the past and how things have worked out. My plan is based on those past plans and results.

–Make a plan based on how patients with this schema have done in the past

Thus my reaction is: “ICU, pulmonary toilet, check if on steroids, try to find previous assessment of lung function (and get ABG to assess current), pain control (with epidural), get out of bed, incentive spirometry, check for wheezing, give inhalers, DVT prophylaxis, watch out for increasing oxygen requirements, check chest x-ray”. That may sound like a lot, but its pretty automatic and doesnt require much thinking, because I’ve taken care of so many of these patients. Why do I have “watch out for increasing oxygen requirements”? Because in my past, we didnt do that and its worked out badly. “Give inhalers” – same thing, at one point we probably forgot to do that and the patient got intubated. Etc.

–Keep an eye out for data that conflicts with schema, or which indicate plan isn’t working

This is the sign of an experienced provider, inexperienced people are so happy they came up with a plan that they tend to fixate on it, and don’t accept new data that says their plan is wrong. That’s why new officers and residents need to be watched like hawks. They will tend to push a plan through, even though new data is indicating the plan is not working. It’s not just experience, but ego. Experienced leaders know that things go wrong, and you have to adapt. Inexperienced leaders think changing the plan is an insult to their ego (which is fragile as a new leader anyway) and try to  make reality adjust to the plan, not adjust the plan to reflect reality.

This is just an intro to these concepts. Think about them and we’ll discuss examples from Klein and real life in the next post.

JY

 

 

Cardiac Arrest from Trauma

Aug 27

Posted by Jeff Young, MD in Clinical Pearls, Clinical Teaching, Discussion, EMS, Hospital | No Comments

The right thing to do with trauma patients that come in arrest, or near arrest can be difficult to figure out. The generally accepted, evidence based rule for this is that for victims of penetrating trauma, they must have had vital signs when EMS arrived (found dead, stays dead), and should not have been in full arrest for more than ~10 minutes (with CPR). For blunt, the patient must have had vital signs of some sort (a pulse at least) in the ED, or within a minute or two of the ED. Even with loss of vital signs at the entrance to the ED, the recovery to discharge rate is abysmally low.

But, EMS often breaks their backs with these patients, doing rapid extrication, ALS, CPR, and priority red transport, and it always feels wrong to just pronounce them on arrival when we stop CPR and find no pulse and fixed dilated pupils. I hate to admit it, but I have done resuscitative thoracotomies in these patients, when I know very well there is no chance of survival. Often, making sure the airway is in, putting in bilateral chest tubes, and a shot of Epi is enough to make everyone feel we’ve tried in the bay, but it’s really just going through the motions, since none of that will resuscitate a trauma patient that has bled out.

I think its important for all trauma surgery, ED, and EMS providers to know the real stats on trauma arrest. I believe the highest discharge from hospital alive rate in penetrating was 7% when arriving in full arrest (some studies are higher, but if you read those carefully, a number of patients were not in arrest, but were just hypotensive), and I have never seen a discharge from hospital rate for blunt arrest over 1%. We have had discharge alive in blunt patients who arrested in the ED, but these patients arrived with vital signs, and lost them in the ED in front of us, making return of circulation and aortic clamping almost immediate. Even with that, the vast number of patients with trauma who have arrested in front of me do not survive.

To make the right decision, you need to know the data.

JY

War Games Cases – Trauma/ED/ICU

Aug 23

Posted by Jeff Young, MD in Clinical Teaching, Discussion, Hospital, Podcast | No Comments

Here are three recorded war games cases presented with me as the student by Joe Sills, a medical student who worked with me. It allows you to hear what I do with unknowns, and how to work through the data to create plans. Thanks. JY

Podcast: Play in new window | Download

Gawande’s article on Cheesecake Factory

Aug 11

Posted by Jeff Young, MD in Discussion, Hospital | No Comments

Atul Gawande’s article asking why Cheesecake factory can provide outstanding products in diverse environments, at reasonable prices, and why we cant do this in hospitals has been sent to me by 5 people. The theme is “why cant we do this?”. I have several comments:

First, I respect Gawande tremendously, but there is a big knowledge gap between doctors who practice medicine, and doctors who attempt to control the way other doctors practice medicine. Gawande is in the former group, and I am in the latter. It is easy to identify the problems in medicine, but much harder to devise and implement solutions.

Second, the line cooks, waitresses, managers, and other employees at Cheesecake Factory accept their positions knowing they are joining an enterprise where their individual initiative is neither wanted, nor needed. I am fairly certain Cheesecake Factory isn’t looking to hire Thomas Keller as a chef. They want people that will easily fit into a system. They are rapidly indoctrinated to a system of processes that have already been figured out. Since, the processes are known and vetted, they are relatively easy to implement in individual restaurants, and in restaurants in different states, or countries. In addition, they have concrete data to represent their success and failure. Sales drops are easy to identify, as well as products that don’t sell, entrees with excessive production costs, and ingredients with high acquisition costs. This data can be operationalized immediately, without argument. That allows the restaurant to be nimble in response to market forces. In medicine, we have few if any vetted and conclusively proven care processes. Sometimes I wonder if there is anything we do that we can definitely say is the “best” way to do it. You only need to look at appendectomy (laparoscopic versus open), CPR (ABC now is CAB), angiography for vascular disease, CABG versus angiography, etc., etc., etc. Sometimes the lack of definitive data for so many things leads to the creation of care guidelines that turn out to be flat wrong, when new studies emerge. At Cheesecake Factory, if they sell a lot of something with a high margin, its awesome, if they dont, they stop making it. We dont have that option.

And finally, Doctors. We are all people at the top of our classes most of our lives, and then went into a profession where we were taught the ultimate responsibility for any adverse outcome was our fault, not the fault of the system. We question and object to data, refuse to follow guidelines because there “might” be a patient that would come along that would be hurt by the guideline, and don’t accept that there are some parts of medical care that others (NPs and PAs) do just as well as we do, if not better. Nevermind, the cottage industry nature of each medical practice where you are responsible for generating income to support you and your family. The more work you do, the higher your income (in general). A manager or chef at Cheesecake Factory does not possess these attitudes, or if they do, I am fairly certain they do not have the power to run their line or restaurant however they see fit. Doctors can, in general, run their clinical practice with a great deal of latitude.

Pointing out how far we have to go in medicine is no longer helpful. Anyone who doesnt realize that change is necessary has been asleep for the past 10 years. We need solutions that take into account the nature of medical care in it’s true state. Success in changing and controlling medical practice occurs when all involved realize they need to find a path that decreases variability, costs, and adverse outcomes.  First you need to identify care processes that are optimal (given the current state of knowledge), then you need to have information and control systems that make it easy for practitioners to comply with the recommended care processes, you need to provide feedback as close to real time as possible, and then you need to constantly look at your outcomes to make sure you are not hurting anyone. It’s a tall order, but it requires everyone to be on the same page.

Dealing with uncertainty in the middle of the night…

Aug 3

Posted by Jeff Young, MD in Clinical Pearls, Clinical Teaching, Discussion, Hospital, Podcast | No Comments

A discussion of the tension between cost containment and the need to prevent unexpected negative outcomes. Some patients do not have a sign on them saying “I’m sick, intubate me and take me to the ICU”. Sometimes the signs are much more subtle, and will not be appreciated without a high degree of suspicion and paranoia.

Podcast: Play in new window | Download

Q&A with Dr. Chris Porter on Onsurg.com

Jul 26

Posted by Jeff Young, MD in Clinical Teaching, Discussion, Hospital | No Comments

Check out a discussion on surgical education I had with Dr. Chris Porter, founder of the great website Onsurg.com.

http://onsurg.com/voice/onsurg-interviews/dr-young

What we are doing wrong in measuring hospital quality and safety..

Jul 25

Posted by Jeff Young, MD in Discussion, Hospital, Information | No Comments

I opened up the ACS news and saw my chief resident’s smiling face. He presented a paper demonstrating that procedural volume does not correlate with mortality, thus destroying a popular quality measure. In it he says that finding out what really does effect quality is the “billion dollar” question.

A year ago, after a great deal of frustration with the quality measures that were being propagated, I wrote the document I have attached. I sent it to various powers that be nationally, and naturally didn’t hear anything. I invite you to read it and think about it. Evidence is rare and weak for anything related to examination of processes, and I hate to put another unsubstantiated model for patient safety out there, but you can read it and decide.

Here is the link:

AnRealisticAssessmentSystemforHospitalSafetyandQuality doc

JY

Active Shooter Incidents

Jul 21

Posted by Jeff Young, MD in Discussion, EMS, Hospital | No Comments

I’ve participated in probably a few dozen disaster drills and mock disasters, as well as commanded the hospital side of five MCIs with up to 20 patients over the years, and after listening to the tapes of the fire response from the Aurora incident Thursday, I don’t believe we as a medical system in the US are adequately prepared to handle those incidents effectively. It sounds like the FD and the hospitals in that area did a fantastic job, but from listening to the tape, it wasn’t until the Battalion Chief arrived at the theater that any semblance of ICS was instituted. Fire depts have a great system to insure that trained Chief officers are available 24/7, hospitals, not so much. For a center like ours, which is the only trauma center for at least 50 miles in all directions, we would have to handle almost all the red priority patients. The trauma surgeon on call is likely not trained in ICS, and may be a year or two out of fellowship. I have confidence the processes for an MCI will work, but that may not necessarily handle this sort of incident effectively.

With an active shooter, the most important hospital resources are OR’s, surgeons, and OR teams (anesthesiologists and nurses/techs). We would also need blood, ICU nurses, and a lot of other help to take care of everyone else in the hospital while we’re dealing with this. Most disaster plans treat an MD as an MD, for this type of situation, general/thoracic/and vascular surgeons and anesthesiologists are the highest resource. After that, we cant do anything without an OR and and OR team. While in most disasters, EM physicians and ED staff are the most important, they cant take a patient to the OR and stop bleeding. Every trauma center (and hospital in a populous area) needs to learn more about and create an active shooter MCI protocol that specifically addresses these key resources.

For those readers (especially those overseas) who have been through these types of incidents, or people who have created these types of protocols, please add your comments so we can learn from your experience. Our hearts go out to all the victims and families affected by this tragedy.

JY

Keeping major problems from killing patients

Jul 12

Posted by Jeff Young, MD in Clinical Pearls, Clinical Teaching, Discussion, Hospital | No Comments

I’ve taken care of lots of patients with life threatening injuries and surgical problems. When I look back and try to figure out what was the key factor in survival, it is keeping the problems one-dimensional.

What I mean is that it is rare to find any single system problem (outside of the heart and the brain) that is immediately and uniformally fatal. What happens is that you get a major abdominal problem, you dont watch kidney function close enough (or react quick enough to changes) and you lose the kidneys, then your fluid balance goes to hell and the patient gets intubated and has worsening vent settings, and so on, and so on.

If I can give one bit of advice from 18 years of taking care of these patients, even if you miss something big and have to do an emergency operation in the middle of the night that should have been done the previous day, you can get the patient over it if you jump on protecting their other organ systems with two feet. That is not the time to diurese them if you are unsure of their fluid status, it is not the time to start narrow spectrum antibiotics, and it is not the time to hold of on intubating them because you think the ABG is not giving you an accurate picture.

 

The way to get these patients through these episodes is to be over aggressive in protecting the function of the remaining systems. Once you lose the kidneys,  go into ARDS, and become coagulapathic, things get much more complicated. Also, remember, do what you neeed to do to protect their remaining organ systems, and this means holding off on procedures and tests that are not central to the problem at hand. If a patient is looking septic, that is not the time to undergo a 5 hour plastics facial reconstruction, unless you think that is central to the infectious problem. Same with ortho procedures and all others. Remember, operations that do not correct your central problem are a physiologic hit to the patient, and it may be one hit too many. These are decisions for the most experienced member of the team, and that should be the attending.

Discussion from Trauma Resuscitation Podcast: The Positive Fast

Jul 4

Posted by Jeff Young, MD in Clinical Teaching, Discussion, Hospital | No Comments

This is a discussion of the algorithms and tactical decisions when you have a trauma patient with a positive FAST. In the podcast, we had a quasistable patient with a positive FAST and a severe pelvic fracture, head injury, and pulmonary contusion. How do you decide what decisions the FAST triggers and what should your reasoning be?

First, remember trauma patients have blood in their belly all the time and the vast majority of them do not go to the OR. In my practice, we admit around 75 blunt spleen injuries, and 90 blunt liver injuries a year, and only around 10% go to the operating immediately for hypotension. The remainder are treated with a non-operative protocol of bedrest and re-evaluation. Of those, around 93% are successfully managed without surgery. So just having fluid in your abdomen, which is what a positive FAST says, does not mean you need to get a laparotomy.

So who does or does not get a laparotomy? Keep in mind we are only talking blunt trauma, penetrating has different algorithms.

An unstable, or recurrently unstable patient with a positive FAST gets a laparotomy, in my opinion. These are patients who present with a SBP<90 who does not increase their BP with fluids (at least 2 liters) and/or blood. If that patient has a positive FAST, then you go to the OR hyperurgently (meaning right now). If that patient has focal neurologic signs, then you need to do your best to rule out a intracranial mass lesion with CT, but that needs to be an attending decision, and the patient needs to be alive (you have more options with a living patient). I once took a patient during residency in this situation to get a head CT with a clamp on their aorta. My attending was not amused. Having a clamp on your aorta is pretty much the definition of unstable, and they cant stop for a head CT. A recurrently unstable patient (or more commonly known as a transient responder) needs surgery. However, in that case, a head CT may be possible with vigorous blood resuscitation, but again, the patient needs to be alive enough to transport.

In this case, we have a quaistable patient (I know many of you have not heard these terms, and we may be the only ones using them). This is a patient with a stable BP, who requires ongoing resuscitation, but whose BP continues to rise and their perfusion improves. To us, this means the patient has bled considerably, but may have stopped, or has a bleed that is worsening when the BP exceeds a certain level. Thus, their perfusion never gets back to normal, but is not worsening. These patients usually have time for CT, but need to be watched like a hawk. We also feel better having big central access and an arterial line. If the patient begins to get unstable, they get pulled off the CT table and go to the OR.

The pelvic fracture adds a big wrinkle. We feel laparotomies are not good for a patient bleeding from pelvic fractures (the Denver group obviously disagrees). This could be a long argument, but I’ll skip it. read the literature and form your own opinion. In this case, you are quasistable with a head injury so you go to CT. You find hemoperitoneum, a minor liver injury, and a major LC3 pelvic fracture with an arterial blush. Angio can kill two birds with one stone in these patients, if you have good access to IR, and good interventionalists (which we have). They can control the pelvic bleeding, squirt theliver and take care of any arterial injuries there, and even squirt the aorta if you didnt have time for a CTA.

Once, again the Chief resident and the Attending should be no more than 20 meters from these patients at any time, and should always keep a close eye on the monitor. If they begin to crash, you may need a new plan. Remember, never let someone bleed to death when you could stop the bleeding in the OR, but also remember, some bleeding can only realistically be controlled by IR.

Add your opinions, nicely.

JY