How Emergency Medicine Resembles a French Fry

The Lord of the Fries

It is a critical axiom in solving problems that we 1). Make sure we are asking the right questions, and 2). Make sure we are measuring the right things.

My favorite learning moment in this regard is a business case study involving not the Cleveland Clinic, but Burger King Restaurants. In the 1980’s Burger King was getting their competitive clock cleaned by other fast food operators and one of the primary drivers of dis-satisfaction with their product was their french fries. So in good old fashioned quality improvement fashion, Burger King set out, through focus groups and surveys, to “measure” what traits constituted a “good” french fry. When all was said and done, they came up with a list of “attributes” that constituted a “quality” french fry, framed these in the form of a “specification” and sent out these requirements to potato farmers and french fry manufacturers (if there is such a thing). And it came to be, that all of the vendors could meet all of the specs – except for one.

This spec was a requirement that a “quality” french fry had to be of such a consistency that it could to be chewed something like 21 times before being swallowed. So they measured and they measured and only one vendor could provide a fry that met all the specifications (including the requisite number of mastications). Burger King happily verified the compliance of the product with the specifications and happily announced to the world that they had developed the perfect, high quality, french fry. With much fanfare it was rolled out all over the world – at which time one small problem was discovered.

Although it met all the measurements, and every minute specification was met – the french fries tasted like crap.

So it becomes critical that we understand our environments and our systems sufficiently to know what the right questions are, and what are the appropriate measures. As the re-defining of health care evolves, it is crucial that we all take a step back from the madness and make sure our view of the world allows us to focus on the right things and measure what is important.

I mention this because (for a variety of reasons) the delivery of health care has devolved to a segmented, task driven process. Each of us is responsible for our own little corner of the world (or of the patient) and as long as all of “our boxes” are checked, it is just assumed that we are doing what is right for the patient and delivering “quality care”. What this has resulted in is our increasingly defining a “successful outcome” for the patient only by how successful we are in passing the patient through our little spot on the assembly line of treating their medical event.

WHY we have a Problem

This was brought to mind as I watched an EMS professional being interviewed on TV in the wake of Michael Jackson’s death. He had been dutifully presented to a local news outlet to illustrate to the community the efficacy of the AutoPulse™ device which his service had recently purchased. Now in the interest of full disclosure – I own no stock in Zoll or any other manufacturer of medical devices or equipment. I have absolutely nothing against Zoll and I think the marketing people that I have met from Zoll are absolutely the nicest people in the world (with the exception of Lauren Angrest at Alvarado Hospital in San Diego – it’s true... look it up). So I’m not trying to pick on Zoll – it’s just a handy example of my point. So my apologies to begin with.

In any event, the television audience and I were informed by the uniformed EMS professional that since utilizing the AutoPulse™, the “save rate of their service had gone from 6% to over 68%”. Wow.... that’s impressive! It’s a better “save” rate than witnessed arrests in a hospital environment.... hmmm.

Well I guessed, and subsequent research confirmed, that our EMS professional was defining a “save” as a patient who, sometime during the EMS encounter, experienced a Return to Spontaneous Circulation (ROSC). Oh.. and did I mention the population on which this was based was 8 patients – but that’s another story.

Alright, I get it. This is a tremendously sucky environment (that’s a technical term only to be used by trained economists, so don’t try that at home) and we are all competing for scant resources so any chance to create some PR buzz is a good thing – right?

Well, I contend that that approach would be correct if we were all aligned and in agreement about what the goal of our health care system is. And clearly we are not aligned. Ideally, I think we could all agree that a system that focused on, and rewarded, health and a high quality of independent life, would be preferable to our current system. However, that’s just not the case. Our incentives (i.e. reimbursement) have created a system of tests and treatments rather than real health attainment. So given that, I think most knowledgeable people recognize that our goal should be to attain the best value for the patient, in obtaining a mutually agreed to outcome, across the entire spectrum of a medical event.

But it is highly unusual for any of us in healthcare to view the value of our actions across an entire medical event. Rather we gauge our value and our effectiveness only on the impact that we have in a given moment (or small segment of time). Historically, we have claimed that there are just too many other variables in the system for any one person or one treatment segment to be held accountable. But those arguments will no longer sustain the weight of scrutiny.

One thing is clear in the current debate about the reinvention of health care in the U.S. – regardless of what models are ultimately adopted – that which adds value will receive compensation – that which does not, will fall by the wayside.

The only way to effectively measure “value” is across the entire spectrum of a patient’s medical condition. In the absence of such a view, relatively expensive “short term” fixes, will always give way to less expensive “long term management” even if the expensive, up front costs are ultimately cheaper over the long run. As an example, let’s go back to the AutoPulse™ example.

These devices are available to the market for around $11 - $14,000. There have been 3 studies conducted with relevance to the efficacy. In one, the data supports that it enhances perfusion and hemo-dynamic stability in a porcine (pig) body. The largest randomized sample study on humans (Hallstrom, 2006) had a total of 1,071 enrollees. The primary end point for the study was survival to 4-hours and there was no difference between the AutoPulse™ population and manual CPR. On a much more meaningful dimension patients who received manual CPR had almost twice the chance of survival to discharge (9.9% vs. 5.8%). In the most meaningful measure, Cerebral Performance Category (a rating of brain function and performance) those who were patients discharged from the hospital after being treated with manual CPR far exceeded those treated with the AutoPulse™ (CPC Category 1 or 2 which maintains at minimum sufficient cerebral capacity to carry out independent activities of daily life - 7.5% for manual CPR and 3.1% for AutoPulse). In fact the Institutional Review Board, halted the research in March of 2005 because of the deleterious neurological outcomes associated with the AutoPulse™ device.

Also in 2006, Krep, et al. conducted a study of the AutoPulse efficacy. The design measures in this study, however, were ROSC and End-Tidal CO2 values. The sample size of this study was 46 (1/10^th the size of the Hallstrom study). It revealed that ROSC was achieved in 54.3% of patients and 21.8% of patients were discharged from ICU. Of these patients, however, 70% (7 of the 10 admitted to the ICU) were discharged with CPC Category 3 or higher meaning –

“Severe cerebral disability: conscious, dependent on others for daily support because of impaired brain function. Ranges from partially ambulatory state to severe dementia or paralysis”.

Six months after discharge, 5 of the 7 patients were still alive with no improvement in their neurological state, 3 patients had died and there was no information available for the remaining 2.

To put it another way, manual CPR yielded a human being capable of independent living in 7.5% of out-of-hospital arrests. The Load distributed device (AutoPulse™) yielded 3.1% in one study and 2.2% in the smaller study.

The point being – from the perspective of treating the patient in a manner consistent with quality across the entire spectrum of a medical event, here is a device that cannot yet be demonstrated as adding value. In fact, a case can be made that while successful in the mechanical perfusion of individuals, the ultimate outcomes are worse and vastly more expensive than if nothing had been applied at all.

So here we have the perfect of example of an answer in search of the right question. Clearly this device has the potential to enhance perfusion and ultimately improve etCO2 and perhaps even ROSC. And in a world where we measure success by our little corner of the world – that was all and good.

But we are entering a different world now, and those of us that ignore that reality do so at our own peril. Just as Burger King’s focus should have been on the ultimate taste of the french fry, our focus needs to be on the totality of the patient experience. Our world will shortly not tolerate nor reward treatments or expenditures that cannot prove a value - and our new world, value can only be viewed across the spectrum of an entire medical condition.

It’s way past time to start re-thinking the questions we ask and ensuring we are measuring what is important. More and more, we are forced to view our actions as a small part of a holistic scheme wherein we gauge our success not on our actions, but on the ultimate result from the perspective of the patient.

Epilogue

In a November, 2008 taste test, Burger King ranked Number 3 out of the top 5 fast food restaurants. I’m not quite sure what that means.

Blog Information Profile for drcrcarlson