Grand Rounds Recap 04.15.20
/Attending Case Followup with Drs. Lang, Baez, and Paulsen - R4 Case Followup with Dr. Habib - R1 Clinical Knowledge - Food Impaction and Esophageal Foreign Bodies with Dr. Winslow - R1 Clinical Diagnostics - Alvarado Score with Drs. Meigh and Klaszky - R1 Clinical Diagnostics - D-dimer with Drs. Nagle and Comiskey - R4 Capstone - Evidence-Based Medicine with Dr. Nagle
Attending Case Followup WITH Drs. Lang, Baez, and Paulsen
Dr. Lang presents lessons learned from an encounter with a medically-complex young woman with critical illness
Fentanyl or ketamine are better analgesics for patients in shock or peri-shock. It may also signal to your nurses that you’re concerned for higher acuity.
Oral contrast usually just delays your CT scan without much benefit. Some expert opinion suggests that it may help localize certain pathology like bowel perf, but data shows that it does not significantly increase sensitivity.
Anchoring bias is prematurely settling on a diagnosis and failing to adjust as new info becomes available. This is different from confirmation bias in which you form an opinion and only notice information that supports it.
“Battle your bias”
Recognize it exists
Use “type 2 thinking”
Type 1 (your gut feeling, instinct) actually relies on biases
Type 2 (slower, methodical, analytical) reduces bias
Challenge your diagnosis
Important things to beware:
Premature closure
Diagnostic momentum
Vital sign abnormalities
High-risk patients
Return visits
Ask yourself:
What else could this be?
What will kill this patient in the next 24 hours?
Hickum’s Dictum - patients can have as many disease as they want (opposite of Occam’s razor).
Dr. Baez reflected on a very difficult encounter she had while responding to an interfacility transport call on Air Care.
In breaking down the discussion by class year we consider different aspects in managing PEA arrest.
What procedures are needed in the peri-arrest patient? What ancillary testing could you get at an outside hospital?
In this case, Dr. Baez was still in the transferring facility ED and acting as a consultant in another provider’s ED reflects on “leading when you are not in charge.”
Be patient-centered - it’s hard to argue with someone who is coming from the standpoint of patient advocacy.
Have the courage to act - your actions are in the patient’s interest.
Be a servant - attitude of service to others may get a better response.
Tap into others’ sense of purpose - get everyone on the same page.
Communicate clearly.
Clay Scroggins has an excellent book on the subject.
Finally, what about the phenomenon of CPR-induced consciousness?
Many of us have anecdotes. What does the literature say?
Not a lot, mostly case reports.
In a study looking at over 2000 cardiac arrests, in 16% of those that survived to discharge 39% of those had some kind of recall of their resuscitation ranging from hearing or seeing things going on around them up to out-of-body experiences.
In a prehospital study CPR-induced awareness was positively correlated with both ROSC and surival to discharge.
What about medications to help manage this unsettling process?
In one study, consciousness-altering medications were actually shown to reduce survival rates by a lot - 21 vs 58%. Importantly, the study did not include ketamine.
What about mechanical compression devices like the Lucas device? We don’t have any scientific literature on this question specifically, but anecdotally we seem to have witnessed CPR-induced consciousness more frequently with these devices.
There are many unique challenges presented by these cases involving logistics, personnel, medications, and psychological impact.
In summary, Dr. Baez recommends that in your approach
Be conscientious
Be encouraged
Engage with the patient (they can probably hear you!)
Consider ketamine
Dr. Paulsen used a case of a subarachnoid hemorrhage as a point of growth to overcome professional failure and become a better clinician.
Approach to professional failure
Rumination, self flagellation, shame
Subarachnoid hemorrhage is a frightening disease because it is both highly deadly and difficult to diagnose.
ACEP endorses using the Ottawa SAH rule in their updated acute headache policy
Ottawa SAH rule can r/o SAH if negative and the patient presents within one hour of HA onset and has normal neuro exam.
6 hour non-contrasted modern head CT can rule out SAH if the patient presents within 6 hours of headache onset and read by a neuroradiologist.
SN 98.7, SP 99.9, NLR 0.010
Be aware that this presupposes that you have a third-gen CT scanner and trained neuroradiologist to read
Aneurysmal SAH has very high morb/mort, so if clinical suspicion remains high then further eval may be pursued
CT/CTA vs CT/LP
ACEP has a level C recommendation for CT/CTA
However, CTA leads to problems with false negatives or incidental aneurysms
CT/LP
Number needed to diagnose 116, number for intervention 290; up to 1 in 6 LPs are traumatic
LP can also diagnose other causes of HA like meningitis
LP is also more technically challenging, has risks and discomfort, and can be difficult to interpret
MRI/MRA
FLAIR sequence is specific but not sensitive, and not sensitive enough to discard LP when you’re still concerned.
Perimesencephalic SAH
5-15% of cases, 33-70% of nonaneurysmal bleeds
Generally better prognosis
Bouncebacks
What are the most common errors with ICH?
Insufficient assessment
Failure to order imaging
Inappropriate response to imaging
Rising strong - responding to failure
Brene Brown - failure at something should be thought of as just your “first draft”
The Reckoning
Recognize that you feel something, think about it, and reflect on how it connects to thoughts and behavior
The Rumble
Revisit, challenge, reality check, be vulnerable, change
The Revolution
Own your truth - write an honest narrative and move on
R4 Case followup - neonatal resuscitation WITH dr. Habib
For his R4 case follow up, Dr. Habib reflected on a case of a new mother who had experienced a precipitous delivery in the ambulance on the way to the hospital. Fortunately, the new infant was healthy and there were no complications, but what if the scenario were different? Neonates are a population that most of us (thankfully) don’t have much experience with resuscitating, so Dr. Habib decided to take the opportunity to give us a refresher on the Neonatal Resuscitation Program (NRP) algorithm.
The NRP algorithm breaks down into three basic blocks:
Team briefing and equipment check.
This step is straightforward. Know where your equipment is kept and have it readily accessible. If you have time for a briefing then talk with your staff and know who’ll be doing what.
The first 60 seconds after birth, “the Golden Minute.”
This focuses on the critical actions that need to happen in the first minute after the infant is born. The vast majority of children who need some kind of resuscitation will respond in this timeframe.
Initial assessment - are they term? How’s the tone? Are they breathing or crying?
If yes then proceed with usual care, child can stay with mother.
If no then warm, clear airway, dry, and stimulate. An important nuance of airway clearance: suction the mouth before the nose. If there are secretions in the oropharynx then the infant could aspirate them if startled by suctioning the nose first.
Is the child apneic or gasping? Is the heart rate below 100bpm?
If yes then provide positive-pressure ventilation, monitor SpO2 and consider ECG monitoring.
If not apneic, but labored or cyanotic then clear the airway, provide supplemental O2 as needed, and consider CPAP.
Ongoing resuscitation after the first 60 seconds.
Still having problems ventilating or oxygenating? Is the heart rate less than 100bpm?
Check the chest movement, consider pneumothorax, and secure the airway with an ETT or supraglottic device if you need to.
Is the heart rate below 60bpm?
Start chest compressions, intubate and provide 100% FiO2, and consider emergency umbilical vein catheterization.
Dr. Habib’s simplified algorithm for the irreverent souls amongst us:
First, recite the Moellman Prayer - Inhale deeply. Exhale. Unleash a string of choice expletives. Get to work.
Be prepared
Mental prep
Baby normal sized?
What’s the amniotic fluid look like?
Are there more babies in there?
Is there anything else you should think about?
Physical prep
Warm
Clear airway
Auscultate
Ventilate and oxygenate
Intubate if you have to
Medications - fluids, blood, epi
The golden minute - neonates are most likely to respond to interventions within this first minute
Make them breathe
Make them breathe
Make them breath some more
Ok, code them, I guess
R1 Clinical Diagnostics - Food Impactions and Esophageal Foreign Bodies WITH Dr. Winslow
Clinical history is the biggest part of diagnosis - acute onset dysphagia, retrosternal fullness, hypersalivation, refuses to eat
Anatomy
Three areas for narrowing - cervical (cricoid cartilage), mid-thoracic (aortic arch), distal thoracic (GE-junction)
Foreign bodies and food impactions are most common in the middle third and distal esophagus
Food impactions
Risk factors
Esophageal abnormalities (rings, strictures, etc)
Eosinophilic esophagitis is one of the most common causes, and this can be the first presentation (suspect when pt also has strong history of atopy)
Diagnosis
Largely clinical
Don’t delay for imaging, but there may be role for CT if you suspect perforation, sharp ingestion. May miss radiolucent material. Don’t use PO contrast, it can interfere with endoscopy if needed.
Bedside ultrasound is an emerging modality.
Complications
Booerhave’s, fistula formation (aorto-esophageal or tracheo-esophageal) if delayed presentation.
What about glucagon?
Theoretical benefit as a smooth muscle relaxant, but induces severe nausea and vomiting.
Studies have shown no significant difference between glucagon and control.
Other pharmacology
Nitroglycerin, CCB, benzos all have roles to play.
Effervescent meds (carbonated beverages or similar) increase intraluminal pressure and help pass the food bolus
Lower median hospital costs compared to glucagon
Endoscopy
Emergent (within 2 hours) if complete and secretion intolerance or pointed ingestion
Esophageal foreign bodies
There are many different types of foreign bodies and many different types of patients.
80% are in children 6 mos - 3 years, but can be seen later with developmental delay
Plain films
Coins or disk batteries in the esophagus usually present the broad, round surface. In the trachea they will present in side profile because of the open posterior component of the cartilaginous tracheal rings.
Disk batteries
Cause harm from direct pressure and liquefactive necrosis.
Complications are bad - perforation and mediastinitis, fistula formation, stricture formation in long term.
Studies show damage as early as two hours and can continue to generate charge longer than 24 hours.
Identify the battery - look at the packaging or what device was it in if possible
Larger than 20mm are associated with the most severe sequelae
Call the National Battery Ingestion Hotline (800-498-8666) or poison control
Give something to coat the battery until it can be removed - honey or sucralfate
Cylindrical batteries if intact pose low threat, but may have difficulty passing through the stomach.
Management
Localize with xray
Get endoscopy if in esophagus
If in stomach either repeat imaging in 48 hours vs endoscopy
Magnets
Single magnets are lower risk, consider removal if in esophagus. They can attach to an external metal source like a belt buckle
Multiple magnets present higher risk for complications (obstruction, necrosis, fistulas)
Endoscopy if in the esophagus or stomach
Surgical removal if in small/large bowel and symptomatic
If asymptomatic in small or large bowel
Serial radiographs for progression/elimination
Surgery if no progression in 6-12 hours
Pill esophagitis
Direct mucosal injury from dissolving medication.
Culprits antibiotics especially tetracyclines like doxy, iron, and potassium.
Discontinue offending agents, endoscopy if symptoms severe or still symptomatic after one week, most will resolve in 7-10 days
R1 Clinical Diagnostics - Alvarado Score WITH DRs. Meigh and Klaszky
Check out Dr. Meigh’s excellent post on the Alvarado Score here.
Alvarado Score
Accounts for signs, symptoms, and labs
Risk stratifies and recommends intervention from rule out/look for other causes, observe, or just go to surgery
Score 1-4 99% SN, 7-10 82% SP
Less accurate for women or children
Note that score isn’t designed to help decide about CT scans
No necessarily better than gestalt
PAS
Developed later, addresses limitations of Alvarado score for peds
Rebound tenderness is replaced by tenderness with cough, percussion, or hopping
Some other elements scored higher or lower
Risk stratifies and recommends no imaging, US or MRI, or straight to surgery consult
Pearls
CT features - normal can be 2-20cm and in virtually any position around the cecum, most commonly retrocecal. For abnormals look for hyperintensity of the wall, dilation, fecalith
Low threshold for CT in elderly
Consider going straight to surgical consult if high risk and clinical suspicion is high
R1 Clinical Diagnostics - D-dimer WITH Drs. Comiskey and Nagle
Check out Dr. Comiskey’s detailed post here
Wells’ Criteria
Check out our Well’s Criteria post by Dr. Pulvino for a deeper dive.
We’re all familiar with the Wells’ Score as a method to risk stratify patients and determine the next diagnostic steps, but it’s worth reviewing because it’s the foundation on which most of the newer D-dimer risk stratification strategies are built.
The scoring system was validated first into a three-tiered low-medium-high risk model and then later into a two-tiered model.
The two-tiered model (Wells’ score < 4 or > 4) is better at ruling out PE using moderate sensitivity d-dimer assays, and is favored by most guidelines.
The PERC rule should only be applied to patients who are low-risk by Wells’ and if negative means there is less than 2% chance of a PE.
In the two-tiered model, with Wells’ less than 4, if you can’t apply the PERC rule then consider getting a d-dimer. If greater than 4 you should probably be going straight to imaging.
D-dimer came into its own in terms of clinical utility in the ED with the Wells’ Criteria. However, there are two patient populations that we encounter frequently in whom the dimer was unreliable - the pregnant and the elderly. Fortunately, we now have ways around this.
The YEARS algorithm and pregnancy-adjusted YEARS
The YEARS algorithm was intended to provide a simpler, more efficient, more clinically practical approach to risk stratify patients suspected of having a PE. It consists of the three most predictive criteria of the Wells’ Score (one point each for clinical signs of DVT, hemoptysis, and whether PE is the most likely diagnosis) and incorporates variable D-dimer thresholds.
With zero points the dimer cutoff is 1000ng/ml, and with 1-3 points the cutoff is 500ng/ml.
So, how does this apply to pregnant patients? If there are signs of DVT then you can obtain compression venous ultrasound to evaluate for DVT and if negative then the remainder of the YEARS criteria remains the same.
Age-adjusted D-dimer
With age comes comorbid conditions that produce elevated D-dimer values for countless reasons that are not PEs. This makes the dimer much harder to use to risk stratify patients older than 50, but fortunately we now have good evidence and backing from ACEP to use age-adjusted cutoffs for D-dimer.
Basically, multiply the patient’s age by ten. This will then give you the age-adjusted cutoff in ng/ml, which you can then convert if needed to whatever units your lab reports. For example, a 56-year-old patient would have a cutoff of 560ng/ml, rather than the traditional cutoff of 500.
Although this decreases sensitivity by 1-2%, it increases the specificity by 15-20% in patients older than 50.
Importantly, this should only be applied to patients who are in the Wells’ < 4 group.
PEGeD
This is the next up and coming method to apply a probability-based D-dimer cutoff for increased specificity.
It starts with the Wells’ Criteria and focuses on the score < 4 group.
In that < 4 group the D-dimer cutoff can be set to less than 1000ng/ml, which successfully decreased imaging from 51.9% to 34.4%.
R4 Capstone - Evidence-Based Medicine WITH Dr. Nagle
Origins
Pierre Louis started collecting and analyzing data on patients with PNA who received bloodletting and correlated this with their survival
Started the link from descriptive statistics to inferential statistics - foundation of how we apply data sets on study populations to the general population
Frequentists inference
Draws conclusions from sample data by emphasizing the proportion of that data.
Estimates the probability that observed events occurred by chance and thus the null hypothesis is true
This method of analysis spread because it was relatively easy to apply.
However, it isn’t good at telling you if the observed difference in outcome is due to the intervention.
Test characteristics like SN, SP, PPV, NPV are all frequentist values
Things like disease prevalence are not always readily known
Bayesian inference
Enter Thomas Bayes, who was looking for a way to statistically prove the existence of god. Should be easy, right?
For our intents and purposes we should start with the idea of Bayes’ theorem - the probability of an event (disease) based on prior knowledge of conditions (age, comorbidity, disease prevalence, etc) that might be related to that event.
There are some important considerations for us as clinicians that we will always come back to:
How bad is the disease?
How good/bad is the treatment?
How bad is the test?
Will it change management?
This approach relies upon pretest probability, which in our realm equates to things like “gut feeling,” “gestalt,” experience, or decision tools.
If pretest probability is too low then the test isn’t worth getting.
If the probability is very high then a negative test doesn’t change management, and so is still not worth getting.
Likelihood ratios arose from Bayesian inference as a way to plot a line from pretest probability to posttest probability by way of sensitivity and specificity.
Higher likelihood ratios (e.g. 10 or greater) start to greatly affect the post-test probability
Nomograms can plot from pretest to post-test probability, but beware if you have a low tolerance for false negative - in that case you’d want more testing in any case. Subarachnoid hemorrhage is a prime example.
Dr. Nagle recommends the smartphone app Doc-Nomo to help you plot your own nomograms quickly.
Bayesian inference has been criticised for being difficult to apply and involving too much subjectivity.
Examples of clinical applications include D-dimer for PE and dissection, COVID-19 test, and PAS score for appendicitis, to name just a few.