Sunday, August 21, 2011

Rashomon, M.D.

Apropos of a case, which looks different based upon your perspective.

I was several years ago, working within my Emergency Medicine Avatar when I overheard the ED scanner call out the approach of a Code III chest pain. I turned to my partner and told him I would see the patient as she entered the ED. I was standing next to a work space finishing out a 'T-sheet' when the patient, a 72-year old female who had undergone angioplasty the week previously was brought into a cardiac bed for evaluation. As the EMS gurney passed my vantage point to the patient I noted that she was literally writhing in pain on the gurney. I ordered a 12-lead EKG, reviewed the tracing when done (no infarction, no ischemia), walked into the room, gently pushed on the L-3rd rib-sternal joint while listening to her heart. No murmurs or other auscultatory abnormalities, but the patient complained even more bitterly of pain. I then injected 10 cc of a 50% by volume mixture of KENALOG™ for inter-articular injection and 1% lidocaine into the patient's easily defined trigger point. 5-minutes later she was free of pain and released from the ED. Total time within the ED was approximately 15 minutes, during a peak time interval for the facility and for which all were grateful (patient's/nursing staff/my partner) for the prompt flow evaluation, timely therapy, and rapid disposition.

As you may know, Rashomon, the brillant film by Akira Kurosawa explores the nature of subjective truth and perception. It is our theme today, in the analysis of my mundane clinical vignette. The problem in front of us relates once again to the current notions of quality in health care. Was this a high quality clinical encounter? or simply efficient? or simply cavalier and reckless care?

As we warp and deform the encounter into a series of Rashomonian perspectives let us look at the case from the standpoint of the quality manager. This patient with known coronary artery disease who had undergone a recent angioplasty failed to undergo conventional cardiac risk stratification with serial cardiac biomarkers and serial EKG tracings. The nearly mandatory overnight hospitalization was not done and the patient released prior to absolute exclusion of AMI or unstable angina had occurred as directed by all relevant cardiovascular disease protocols and practice guidelines. (See the following pdf document for a relatively standard chest pain protocol- http://www.mssm.edu/static_files/MSSM/Files/Departments/Emergency%20Medicine/Manual/SUB52.pdf)

Seen from the perspective of the clinical performance measures and through the eyes of the clinical performance measures monitor, the case falls out. I am once again required to document and justify my deviation from the quality measures which are mandated by the patient's chief complaint of chest pain. No ASA, no beta-blockers, etc, etc.


If I had a medical educator look at the case, I was guilty of inadequate and substandard care. I literally asked the patient one question, 'does it hurt when I push here?' It is well known, and I agree with this dictum that history suggests or makes the diagnosis in over 80% of cases. Therefore, my encounter with this patient is suspect due to its abbreviated nature.

I coded the case as chest pain with a sub diagnosis of costochondritis. If the case had been reviewed by medicare, I would have been guilty of medicare fraud, since the diagnosis of undifferentiated chest pain is a high complexity type diagnosis that would justify with appropriate documentation of medical data elements a high complexity payment. However, if I coded the diagnosis with costochondritis as the initial diagnosis I would have been paid for a low or moderate complexity diagnosis. Furthermore, if I recall the case properly I probably coded the case as high complexity and probably failed to capture all data elements required for this charge.

If the case was reviewed by the hospitalizations utilization manager, I might well have been subtly or not too subtly harassed since this was a medicare beneficiary, i.e., an adequately insured patient and I deprived the hospital of the DRG attendant to medical observation for a chest pain rule out protocol while obtaining for the hospital only the initial payment for the Emergency Department evaluation of the patient. Furthermore, since the patient was in the ED settings, all advanced imaging charges would have been born by the payor, unlike imaging accomplished after hospitalization which results in reduced profit margin for the patient. Therefore, failure to do a Chest CT with IV contrast to 'rule out' a pulmonary embolism deprived the hospital of some not insignificant revenue. Apropos of this point, I am aware of an ED medicine group who are literally tracked by their administration to keep medicare beneficiary ED admissions at a given rate with respect to numbers of ED presentations. What I mean by this, is if 80 medicare beneficiaries are seen in 24 hours than X% of those patients are required to be admitted. This is done to preserve hospital revenue streams and the percentage is directed by administration to the ED group. The not too subtle implication is that if they can't find a reason to admit at this rate, another ED group can be found that will be compliant to this informal guidance.

If this patient with the known presence of CAD had sustained sudden cardiac death or an acute myocardial infarction anytime in the 6 to 12 months of her release from the ED, the grieving family of this patient would have been able to find probably 1,000 'domain experts' who would have been overjoyed to testify that my practice was substandard, reckless, and cavalier resulting in avoidable death or injury to my patient. The case is many years distant now, so I think I can say with some certainty that the patient did not sustain a myocardial infarction anytime over the period of reasonable observation since my own ED evaluation of the patient. However, from the perspective of the retrospectoscope I would have been guilty of medical malpractice due to my failure to do what other similarly trained and experienced practitioners would have done.

Now as we face the question, was this enlightened practice or bad practice which of the Rashamon-esque perspectives should we adopt, which is the real situation.

From my perspective, the thought process and the situation is as follows. I was heavily influenced in my clinical evaluation of chest pain during my own internal medical training by the following now classical article:

Ruling Out Acute Myocardial Infarction: A Prospective Multicenter Validation of a 12-Hour Strategy For Patients at Low Risk. Lee et al. NEJM, 34(18) 1239-1246,1991.



The flow chart is difficult to see at this magnification but the solitary physical examination finding predictive of low probability for underlying ischemia is a defined costo-sternal trigger point that exactly reproduces or amplifies the patient's pain. Furthermore, while 20% of anterior wall MI's are recognized to have accompanying chest wall pains, total resolution of the pain with local injection of Lidocaine reliably distinguishes the chest wall pain associated with AMI from costochondral pain.

When the patient passed me at the nursing station en route to her room I noted the colicky nature of her pain, i.e., she could not get comfortable and she writhed and twisted on the EMS gurney trying to find a spot where she was comfortable.

True ischemic chest pain and infarction pain is never colicky, the patient in this setting acts like an insect preserved on a board with a needle stuck in the middle of its chest. There is no movement, because the patient feels as if they are being crushed into the gurney. The sole exception to this rule of thumb is the patient with myocardial rupture which is also an intense pain with a colicky character but for which time rapidly triages the patient into the critically ill group.

So,from my perspective of the case, the patient by Gestalt had costochondritis, the exquisite sensitivity of the chest to the light palpitation pressure resulting from auscultation of the left sternal border, the presence of a non-ischemic EKG, and the rapid and complete extermination of her severe 10/10 chest pain with 10-cc of 1% Lidocaine all made for a defensible diagnosis of costochondritis.

The patient was happy, I was happy, my partner was happy, the nursing team was happy, the biller/coder for the ED group was happy as I could bill for 2 or 3 more high complexity patient's over this hour.

Was this a high quality encounter, or not? Was it sensitive to the goals of EBM practice or not? Am I or should I be paid for the quality of my documentation or the quality of my care, or more heretically for being right or at a minimum being prudent?

What is wrong with health care today, that this trivial patient encounter can engender so many potentially negative viewpoints and reviews? In fact with a case base of over 60,000 unique patient encounters my clinical expertise is and should be well in advance of EBM. Why do I personally feel harassed and hassled by a health care system designed to educate others in things I already know? Why am I subjected on a daily fashion and why are my practice habits being subverted by perverse incentives? Why is it that as someone always in the pursuit of excellence and growth in my skills and my knowledge do I find the new Galenism so repugnant and offensive? Why do I get the feeling that my goals and professional needs and more importantly the needs and goals of my patient are irrelevant in the land of health bureaucracy and health care mediocrity? Why am I not paid for expertise rather than slavish adherence to guidelines that often times do not fit the clinical scenario? Or rather why does my expertise not de facto trump the guidelines? Why should I have to work harder to do the right thing rather than mindlessly follow a guideline which as I have noted often times does not suit the clinical situation? Why should I follow guidelines targeted to chief complaints rather than validated diagnoses? Most irritatingly, why should someone with less training and experience and judgement be able to trump my own clinical impressions based on their position or title, i.e., clinical practice monitor/coordinator?