Tuesday, May 15, 2018

Root Cause Analysis: Dig Deeper, or the Weed Will Keep Growing Back

In a recent JAMA Performance Improvement piece, the authors describe the case of a man who presented to the emergency department with dizziness.  He was sedated for an MRI, his history of OSA (obstructive sleep apnea) may have been glossed over, and he arrested in the radiology department.  The subsequent "root cause analysis" traced the untoward outcome to a failure to recognize the OSA and the adverse effects that may follow sedation of a patient with this diagnosis.

The problem with this "root cause analysis" is that it assumed that the MRI, requested by a neurologist on-call, via telephone, was necessary.  It was not.  The root cause analysis got it wrong because it did not trace the roots to their deepest source:  glossing over the patient's chief complaint and considering it and its evaluation carefully and rationally.  Stroke is an uncommon cause of dizziness and the MRI was probably not indicated, especially in light of the other information provided in the case.

Here is the letter that I sent to JAMA which was not accepted/published.  It is a case of the distinction between rationality and intelligence.  Very intelligent people traced the "cause" or the "root" of the complication to a missed piece of information (OSA) and corollary ideas (he may have complications from sedation), but they failed to consider underlying assumptions:  namely that the MRI was necessary or would yield net benefit in the first place. 

Medicine is best played like chess, not like checkers.  "Intelligent people have superior performance when you tell them what to do."  A failure of a "root cause analysis" such as this will foment the regrowth of the weed.

Here is the letter:

I enjoyed the Performance Improvement case describing oversedation of a patient with obstructive sleep apnea1.  I posit that the most proximate possible root cause of the complications described was ordering an MRI with low clinical yield2, without pre-specifying what abnormality was being sought as well as its probability, and without delineating, a priori, how any resulting findings would change management3.  Presumably, the neurology consultant was looking for stroke.  What was its pre-test probability in a patient with dizziness?  Would management have changed if stroke were detected with imaging?  Were there contraindications to therapies for stroke?  Was the patient already receiving the indicated therapy for stroke?  What is the probability of a false positive finding (i.e., one that doesn’t explain the patients’ symptoms; an “incidentaloma”), and how might that finding lead to interventions which may yield net harm if stroke is not present?  What was the response to meclizine and odansetron, and how did this incremental information alter the prior probability of stroke?  Because decisions necessarily precede actions, they must always be considered as possible proximate causes of downstream complications.  Even if the other errors identified in the reported root cause analysis can be avoided in the future, injudicious testing may lead to other complications, including cascades of additional potentially harmful testing and intervention unguided by careful, rational, clinical decision making.

1. Blay E, Jr, Barnard C, et al. Oversedation of a patient with obstructive sleep apnea prior to imaging. JAMA 2018;319(5):495-96. doi: 10.1001/jama.2017.22004
2. Fakhran S, Alhilali L, Branstetter BFt. Yield of CT angiography and contrast-enhanced MR imaging in patients with dizziness. AJNR American journal of neuroradiology 2013;34(5):1077-81. doi: 10.3174/ajnr.A3325 [published Online First: 2012/10/27]
3. Pauker SG, Kassirer JP. The threshold approach to clinical decision making. The New England journal of medicine 1980;302(20):1109-17. doi: 10.1056/nejm198005153022003 [published Online First: 1980/05/15]




Confusion, Diaphoresis, and Hyperventilation Aboard a Private Airplane

This was intended to be a case report but the amount of work required to publish a case report is just too great to justify it.  The publishing landscape has been flooded with an attendant raft of predatory journals, so one must be very careful.

This will be an online interactive case report.  I will tweet this post, asking for comments and diagnoses in the comments below (preferable to twitter comments) and update with the answer and a discussion in 1-2 weeks.

A 68 year-old otherwise healthy male passenger was flying with his friend, a pilot, in a private plane from California to Montana for a fishing trip.  Within an hour after takeoff, he became confused and diaphoretic and was hyperventilating, then he lost consciousness for approximately 20 minutes.  The pilot applied supplemental oxygen and checked his passenger's arterial oxygenation via oximetry, finding it to be 95%.  The flight was diverted for a medical emergency.  During descent and landing, the passenger regained consciousness but remained confused.  In the emergency department of a nearby hospital, he had normal vital signs, and had a non-focal neurological examination, but remained confused.  Representative images from CTA of the chest are shown below (click on the images to expand).  A head CT was normal excepting for some age related atrophy.  What is the diagnosis?

Monday, November 20, 2017

Sunk Kidney Bias: A Lethal Form of Sunk Cost Bias

Hal Arkes
The heuristics and biases program of Kahneman and Tversky, once an obscure niche of cognitive psychology, became recognized among lay persons with Kahneman's Nobel prize in economics in 2002.  The popularity of the program surged with Kahneman's book Thinking Fast and Slow several years ago and several among the scores of related books about behavioral economics became best-sellers.  This year, Richard Thaler was the Nobel laureate in economics for his work in behavioral economics.   I became aware of heuristics and biases just before Kahneman's Nobel and started looking for them in medicine in 2003.  We (Aberegg, Haponik, and Terry, Chest, 2005) indeed found evidence for omission bias, and have discovered other biases along the way, some which are very intriguing but we aren't even sure what to name them (Aberegg, Arkes, and Terry, Medical Decision Making, 2006).  My point here is that these biases are useful but difficult to identify as patterns systematically operating within medical practice in predictable ways - they pop up here and there only to recede and reappear years later, if they are recognized at all.

Then there are biases about the biases.  Highly cited expositions of biases in clinical care, such as those of the insightful emergency physician Pat Croskerry (Academic Medicine, 2003), among many others) very often surmise the presence of biases in clinical care, without the kind of empirical evidence that established the biases in the first place.  Sometimes, new and probably useful biases are proposed (such as "search satisfycing"), without any empirical evidence, at all in any domain, for their existence.  They are merely postulates.  (Granted, empirical evidence is very difficult to generate, this the reason I don't do this kind of research anymore.)  Finally, the descriptions of the biases applied to medicine are often strained, or just plain wrong.  My favorite is the bastardization of "anchoring and adjustment" into a description of any time a physician seizes upon a diagnosis and discounts disconfirming evidence or fails to consider alternatives.  This is not anchoring and adjustment.  Anchoring refers to a numerical anchor, and failure to adjust away from it when providing numerical estimates.  Here is a summary of the original descriptions, from the wikipedia entry on anchoring and adjustment:

Tuesday, October 31, 2017

Applied Respiratory Physiology Vlog. Parts 1,2,3,4: Respiratory Failure Explained as Workload Imbalance

The following embedded videos are parts 1-4 of a 5 part talk I've been giving and refining on Applied Respiratory Physiology for about 10 years now.  (It is split into 5 parts because of youtube size limitations and for digestible 10-15 minute segments.)  The principles herein derive from many sources, but special credit must go to Nunn's Textbook of Applied Respiratory Physiology and The University of Chicago critical care text edited by Hall, Schmidt, and Wood.  For the majority of the ideas and applied principles herein, I have never seen them discussed in any lecture in 20 years of attending pulmonary conferences, didactics, etc.  My interest in applied physiology and Nunn's textbook indeed originated because of my frustration with the esoterica of the basic and advanced physiology that I was taught from medical school through fellowship -  I determined that much or most of it was not applicable at the bedside.  This lecture series, I hope, will be far more clinically applicable, intuitively appealing, memorable, and useful than what has been traditionally taught.  Real life examples highlighting the extremes of human respiratory performance should, I hope, make this a memorable lecture seeries.  I welcome comments and criticisms below.  Enjoy!



Tuesday, September 26, 2017

DIPSHIS: Diprivan Induced Pseudo-Shock & Hypoxic Illness Syndrome

This would be a very informative case report (and it's true and unexaggerated), but I anticipate staunch editorial resistance (even sans puns), so I'll describe it here and have some fun with it.

Background:  The author has anecdotally observed for many years that so-called "septic shock" follows rather than precedes intubation and sedation.  This raises the possibility that some proportion of what we call septic (or other) shock is iatrogenic and induced by sedative agents rather than progression of the underlying disease process.

Methods:  Use of a case report as a counterfactual to the common presumption that shock occurring after intubation and sedation is consequent to the underlying disease process rather than associated medical interventions.

Results:  A 20-something man was admitted with pharyngitis, multilobar pneumonia (presumed bacterial) and pneumomediastinum (presumed from coughing).  He met criteria for sepsis with RR=40, HR=120, T=39, BP 130/70.  He was treated with antibiotics and supportive care but remained markedly tachypneic with rapid shallow respirations, despite absence of subjective respiratory distress.  A dialectic between a trainee and the attending sought to predict whether he was "tiring out" and/or "going into ARDS", but yielded equipoise/a stalemate.  A decision was made to intubate the patient and re-evaluate the following day.  After intubation, he required high doses of propofol (Diprivan) for severe agitation, and soon had a wide pulse pressure hypotension, which led to administration of several liters of fluids and initiation of a noradrenaline infusion overnight.  He was said to have "gone into shock" and "progressed to ARDS", as his oxygen requirements doubled to 80% from 40% and PEEP had been increased from 8 to 16.  The next morning, out of concern that "shock" and "ARDS" were iatrogenic complications given considerations of temporality to other interventions, sedation and vasopressors were abruptly discontinued, diuresis of 2 liters achieved, and the patient was successfully extubated and discharged from the ICU a day later.

Conclusions:  This case provides anecdotal "proof of concept" for the counterfactual that is often unseen:  Patients "go into shock" and "progress to ARDS" not in spite of treatment, but because of it.  The author terms this syndrome, in the context of Diprivan (propofol) in the ICU setting, "DIPSHIS".  The incidence of DIPSHIS is unknown and many be underestimated because of difficulty in detection fostered by cultural biases in the care of critically ill medical patients.  Anesthesiologists have long recognized DIPSHIS but have not needed to name it, because they do not label as "shock" anesthetic-induced hypotension in the operating theater - they just give some ephedrine until the patient recovers.  DIPSHIS has implications for the epidemiological and therapeutic study of "septic shock" as well as for hospital coding and billing.

Sunday, August 27, 2017

The Number Needed Not To Treat To Harm (NNNTTTH): A Heuristic for Evaluating Trade-offs in Medical Decisions

A frequent conundrum of decision making that arises in medicine is when there is a generally indicated therapy, say, anticoagulation for atrial fibrillation, that poses unique risks in a particular patient.  CHADS2 and HAS-BLED scores are calculated, but don't quiet the hemming and hawing or quell the hand-wringing.  What is usually a simple dichotomous decision is now one laden with probabilities, risks and benefits, and compromise between competing objectives.  (See:  The Therapeutic Paradox:  What's Right for the Population May Not Be Right for the Patient.)  In order to restore nuance to the decision, we need to try to estimate the numerical values of the risks and benefits to determine if the net utility of anticoagulation is positive or negative, something the aforementioned calculators are intended to do in a semi-quantitative way.  But what if you opine that your patient has a specially enhanced risk of side effects and you're worried about falls or bleeding but ambivalent because of a concurrent fear of denying him of the benefit of stroke prophylaxis?  What if you think that he would have never been included in a trial of stroke prophylaxis and the results of those trials may have limited generalizability to him?  What if you think he has only a year to live?

The number needed not to treat to harm (NNNTTTH) is the number of patients whom you have to not treat with something beneficial in order to cause one harm from your omission.  It is numerically equivalent to the number needed to treat (NNT), but it reframes the decision from action to omission and from benefit to harm.  Ignoring bleeding altogether (because making relative utilities for bleeding and stroke is a fraught endeavor), you could ask yourself "how many patients can I withhold stroke prophylaxis from for one year before I statistically cause (or allow to happen, if you are prone to omission bias) a stroke?"  For most patients, withholding stroke prophylaxis has a NNNTTTH of about 25-30 per year (check the corresponding NNT from CHADS2 for a more "precise" estimate).  Reframing the question into "how much am I asking the patient to pay, in terms of statistical likelihood of stroke, to avoid anticoagulation and the particular side effects that cause me concern in his case?" can often provide some reassurance for the clinician and the patient alike.

Wednesday, July 19, 2017

Screening in Disguise: You Can't "Unknow" that Troponin, But You Can Dismiss It After Careful Thought

During MICU rounds last month, there were a lot of troponins ordered, and most of them should not have been.  Invariably when abnormal troponin values are reported on rounds, there is no mention of whether the patient had anginal chest pain, whether there were ischemic EKG changes, or whether this information was sought at the time the troponin was drawn.  This is because troponins are being used as a screening test, rather than as a diagnostic test.  "Not so!" exclaims the resident, eager to convince me that he has not engaged in the kind of mindless testing he knows I loathe.  I am told that because the first troponin was mildly elevated in a little old lady with cirrhosis, overdose, right heart failure and urinary tract infection, that we need to follow it to see where it "peaks".