Defending small data

I read with interest a recent editorial that opined on the poor evidence for screening in cancer trials. The evidence was judged poor because apparently no screening trial has demonstrated a clear reduction in all-cause mortality, only disease-specific mortality.  One example discussed in the analysis reviews the data for colon cancer screening and notes that, while there were a statistically significant lower number of deaths related to colon cancer in the screened group, the total mortality in the two groups was no different.  The authors posit that the study is either underpowered for total mortality or that the screened patients may have more deaths due to the 'downstream effects' of screening.  The provocative conclusion by many a tweet and retweet is that cancer screening has not been shown to save lives. Apparently the path to progress in medicine now must be paved by studies with millions of patients.  I understand the desire for more and more data, but I see danger in the sanctimonious protestations of those who can only find truth within the confines of a million-person, randomized control trial.  This approach ignores the history of advances in clinical medicine, most of which live far outside of the boundaries of million-strong randomized clinical trials.

The history of interventional cardiology is particularly instructive.  The first catheterization of the human heart was performed by a surgical resident named Werner Forsmann in 1929 - on himself.  He thought it was a good idea to catheterize his own left antecubital vein, pass a ureteral tube into his right atrium, and then walk to the radiology department to document his accomplishment.  Forsman was fired for his troubles.  His work was rejected by the scientific establishment of the time and he languished for two decades - at times having difficulty finding gainful employment, and ultimately giving up research- until 1956, when he was finally recognized for his felicitous discovery.  The technique did not find widespread use and was mostly the provenance of physiology labs until a German physician named Andreas Grunzig decided to make an attempt at opening chronically narrowed coronary arteries using balloons and catheters.

The first coronary angioplasty was performed by Grunzig September 16, 1977 using a home-made catheter on a 38 year old with disabling angina.  The procedure was remarkably successful.  The patient remained free of angina and had no residual stenosis on a repeat catheterization at the 10 year anniversary of the seminal event. Cynics would say that Grunzig ushered in an era that saw the rapid rise in unnecessary coronary interventions, and they would be right.  But it was techniques developed by people like Grunzig who made it possible for a Seattle cardiologist named Marcus Dewood to take patients in the throes of a heart attack and inject contrast into the coronary tree to show that heart attacks were a result of blood clots that formed in arteries that fed the heart.  Soon after, cardiologists began blowing up balloons to open up thrombosed arteries in patients having heart attacks and primary angioplasty - arguably the single biggest leap forward in the treatment of cardiovascular disease - was born.

The epidemiology of cardiovascular disease, as a result, has seen dramatic change over the last half century.  Interestingly, the rate of acute myocardial infarctions has been fairly constant over time, but cardiovascular mortality has dropped significantly.
While the decline in cardiovascular mortality is at times attributed to the success of primary prevention and medical therapy, the data for primary prevention in this arena is relatively weak. That does not mean that tobacco cessation and aspirin do not work, it's just hard to show what exactly the size of these effects are.  A study in the New England Journal tried to adjudicate the number of deaths prevented by specific risk factors using assumptions that would kindly be called shaky.  Excerpted below is the example given in the paper.

The paper assumes from a metanalysis that a 20mmHg drop in blood pressure will reduce deaths by 50%. This is not a statement that can be made broadly for a variety of reasons. The starting blood pressure matters, it would be wrong to suggest all deaths prevented were cardiovascular deaths, and it is quite possible that aggressive blood pressure control in certain populations (diabetics) resulted in poorer outcomes.  The faulty assumptions are rescued, of course, with the use of an exponential function, because Euler's number fixes everything.  This same paper also attributed a reduction in deaths to interventions (revascularization for chronic stable angina) that have never been shown to reduce mortality!

I realize it has become dangerous to use one's clinical experience to inform one's views.  While I have no quarrel with evidence, the reality is that the longer I practice, the more I realize that clinical scenarios rarely fit even the best designed clinical trials.  In light of the underwhelming evidence with regards to primary prevention and cardiovascular mortality, my thoughts on the impressive reduction in cardiovascular mortality are driven by a group of patients that I cannot forget.  These were patients who had large 'missed' myocardial infarctions.  I can still vividly remember racing to place a central line in a middle-aged Cambodian man who had arrived in the CCU more than 24 hours after the onset of chest pain.  He had suffered a proximal LAD infarct  - an occlusion in an artery that fed a very large part of his heart.  As I rapidly added medications in an attempt to support a failing heart, a cardiac marker was called from the lab - a cardiac troponin that was over 400.  Cardiac troponins are released into the blood stream as cardiac muscle dies.  The higher the troponin, the more cardiac muscle that has died.  The vast majority of troponins that I get called for are under 1.  A troponin of 400 was inconceivable to me at the time.  My patient didn't have a chance - he was dead within 12 hours.

I had other patients with extensive heart attacks that could not be revascularized - they all did poorly, driving it home to to me in a way that reading about it never could - that the time to revascularization was the single most important metric that predicted a good outcome.  So when I look at the graphs that show a marked drop off in cardiovascular mortality that accelerated in 1980, I do not believe that it was coincidence that this also marked the beginning of the era of coronary intervention for heart attacks. Again, this is not to suggest primary prevention does not matter, rather there is little that clinically has seemed to have as direct of a clinical impact on patient outcomes than revascularization to limit the size of an infarct.

I wonder how Forsmann and Grunzig would have survived in the current climate that looks down on any clinical data that has fewer than a million patients.  Grunzig demonstrated results of coronary angioplasty in his first four patients at the American Heart Association meeting in 1977.  He was interrupted by applause, and was so stunned he was unsure if he could proceed.  I suspect today he would have been rewarded by denigrating tweets from the 'elites' among us asking for more data.
These apostates worshiping at the temple of data see themselves as gladiators championing the interests of patients.  They are right and wrong.  The current messy path medicine forges ahead on involves hurting patients.  Grunzig used primitive catheters he made in his kitchen and patients in those early years had complications that sometimes required emergent bypass, and at other times, resulted in death.  But here's the thing, I don't think Grunzig did anything wrong - he simply did the best he could with the tools and the knowledge available to him.  Without Grunzig, there would have been no DeWood, and no progress to a world of interventional cardiologists on call to open up an occluded artery within 90 minutes of a heart attack.

While well intentioned, the purists that seek to use big data to protect us flirt with therapeutic nihilism.  Forcing all progress through the funnel of a million strong clinical trial is a bar too high, a bar that protects us from medicine itself.  No one benefits from that.

  

Treating chest pain with a cup of tea

It is very early.  I am running to the 'clinical decision unit' (CDU) to see a patient of mine sent in the night before from a local skilled nusing facility.  Also known as clinical observation units,  'obs' units, or short stay observation units, these units were designed to help decompress busy emergency rooms and divert unnecessary, expensive inpatient admissions.  The units are typically adjacent to emergency departments, and usually are run by emergency physicians.  My particular patient was admitted due to an episode of chest pain at her facility.  A brief conversation the prior night with the emergency room staff revealed chest pain that clinically was not typical for any of the feared diagnoses of a heart attack, pulmonary embolism or an aortic dissection.  An electrocardiogram and cardiac enzymes were also initially unremarkable.  Regardless, the patient was elderly and had multiple other comorbidities, and was somewhat confused.  I recommended a short stay to allow anything malignant to declare itself.

And so, here I was, at the observation unit, digging through pages upon pages of printed gibberish that clearly had achieved the nirvana stage of meaningful use (for those wondering, that's after stage 3).  Ironically, the most useful piece of information lay in a handwritten progress note describing the episode. I could see why the patient had been brought here to be further evaluated, but after 18 hours of negative biomarkers, electrocardiograms, and no recurrence of symptoms, I felt comfortable letting her go back to where she came from.  I told the ER staff... who cancelled her stress test.  A stress test? Yes, a stress test had been ordered prophylactically.  We practice in a climate where every bad outcome has the potential for litigation - malpractice lawyers would have a field day with the case of anyone going to the ER, being discharged without some type of cardiac imaging study, and having a heart attack.  My recommendation to discharge the patient shifts the liability of an adverse outcome from the ER squarely on to my shoulders, and thus, poof goes the stress test.

As I get ready to move to the next patient on my list, I'm stopped by one of the ER physicians. There's a 34 year old woman with chest pain in the CDU as well.  She also had a brief episode of chest pain the day prior.  She had no other medical history, her electrocardiogram was benign, multiple biomarkers were negative.  Her symptoms had resolved.  She was scheduled to have a CT scan of her coronaries done.  Could I take a look?

6 million patients arrive at emergency departments complaining of chest pain.  About seventeen percent of them have a real coronary event.  Discovering which one of these patients actually has a real coronary event is one of the great clinical challenges that faces physicians.  Traditionally, doctors have used the clinical history, an electrocardiogram, and cardiac biomarkers to make a decision about hospitalization and further cardiac testing.  The teaching when I was in training was to categorize patients: low risk patients needed reassurance, intermediate risk patients needed some type of a stress test, and high risk patients needed a cardiac catheterization.  The hard part, of course, was drilling down to the patient in front of you in the emergency department and getting that label right. The data with this approach was  not perfect.  The chances of missing a heart attack were 1 in a 1000 (0.1%).  Review of these cases by an expert, experienced cardiologist in the majority of these missed diagnoses revealed no errors; the patients just had very subtle findings that the traditional approach was unable to resolve.

Two developments coincided over the last ten years that started to change the landscape of the approach to chest pain patients.  The first development was the push by payers to reduce cost - for chest pain patients that meant avoiding admissions to the hospital.  Payers started to refuse payments for those patients deemed too low risk for admission to the hospital.  Spooked by this, observation units, and protocols to increase patient throughput through these units flourished.  It was within this Wild West-like atmosphere that John Wayne, in the form of the coronary computed tomography angiography (CCTA), arrived.  CCTA allowed high resolution images of the coronary tree within minutes, and promised an easy straight talkin' yes/no answer with regards to coronary plaque. A negative test effectively ruled out a coronary event, and as a side benefit, also could provide information about clots in the lung and abnormalities of the aorta.  One test that took minutes could obviate the need for a hospital admission, and shorten the amount of time spent in the ER because you didn't need to wait for the old cardiac stress test that was only done during daytime hours.  It also seemed to satisfy everyone's appetite for zero risk -  the promise was that a negative test meant you were absolutely not going to have a heart attack.  The clinical history in this brave new world was irrelevant.  After all, the traditional approach that focused on talking to patients missed one out of a 1000 patients.  Now we could be perfect, and it would save money!  The time for irrational exuberance was upon us.

Not surprisingly, coronary CTA's proliferated in ERs.  I noticed that patients who ordinarily would get admitted to the hospital, or would merit a cardiology consultation were instead being sent home from the ER after a negative study.  The calls from the ER didn't stop though; they changed.   Now the calls were for positive coronary CTAs, which some times forced additional testing.  After all, the if the CT scan suggested there was something wrong with your heart...it was exceedingly difficult to reassure the patient or myself with only a lowly stethoscope that everything was actually ok.

Clearly,there was no free lunch when it came to CCTA. There was improved efficiency in the ED, but there was an increase in downstream testing and radiation exposure with no decrease in the overall costs of care.  Even more troubling, I started noticing patients that I ordinarily would have suggested no testing were also being put in the imaging pathway.  The question no longer was if you needed a test, but what test you should get.

In a study (Foy et. al.) comparing no testing with multiple other cardiac imaging modalities that included stress tests and CCTA, low risk patients who received no further cardiac testing had no difference in outcomes.  Unfortunately, that wasn't even the worst part.  The unfortunate low risk chest pain cohort that did get additional cardiac testing ended up with additional, invasive testing.  It was estimated that for every 27 patients who underwent a myocardial perfusion imaging stress test, 1 patient got an unnecessary cardiac catheterization.  For every seventy-one patients who underwent a CCTA, 1 patient would require an unnecessary cardiac catheterization.  Since there was no difference in outcomes between the two groups, the additional invasive testing was all risk with no benefit.



As the figure above demonstrates, an approach that involved CCTA for all low risk patients with chest pain would result annually in 50,000 unnecessary cardiac catheterizations.  Cardiac catheterizations have become incredibly safe, but do involve snaking a catheter into the small arteries that come off of the aorta and feed the heart. Do this enough, and bad things are bound to happen.  In contemporary practice the mortality associated with cardiac catheterization is 1 in a 1000.  A rare event, to be sure, but that would mean an estimate of 50 healthy people dying annually due to having an unnecessary procedure.  To top it off with a giant maraschino cherry, there is nothing to suggest a CT scan-for-all approach would even reduce the rate of missed heart attacks.  Recall that Foy et. al., showed no difference in hospitalization for MI regardless of whether low risk patients got some type of cardiac stress/imaging test or not.

While Foy's paper is a retrospective review that is far from definitive, it does fit in with what I see in clinical practice.  The well-intentioned desire to cut costs by avoiding inpatient admissions, and improve safety by reducing missed coronary events has not lowered costs, and has not made patients any safer.  There certainly are trials that will point to the impressive negative predictive value of coronary CTAs, and the impressive close-to-zero event rate with these tests.  These studies are hopelessly flawed.  Many of the patients in these trials should have warranted no further testing, and would have been better served with my mother's masala chai - a much better tasting and soothing alternative.

I did take a look at the young woman with chest pain.  I could not bring myself to order a CCTA or a stress test.  I did recommend a nice cup of masala chai.  She remains alive and well to this day.

The problem with choosing Medicare Advantage plans: the abbreviated version

Mrs. Cassidy slowly walks into my office one busy afternoon.  Mrs. Cassidy has some serious style.  She is wearing a deep orange dress with a bright blue blazer.  There aren't too many folks that can pull that outfit off, but she can. She has a wide slow smile, and she speaks with a slow southern drawl that belies her southern roots. This was supposed to be a routine follow up visit for a 67 year old woman with a history of a mechanical mitral valve replacement and coronary disease.  Unfortunately, she tells me a story that is concerning for angina.  I think she needs a stress test. I quickly look at her insurance, and I let out a somewhat audible groan.  She has a Medicare Advantage (MA) plan.  I explain to Mrs. Cassidy that we will need to go through an extra step to pre-certify her stress test. She expresses surprise and asks me what she should do.  I will tell you what I told her, but first, let me tell you why.

Medicare Advantage Plans: A brief history

Medicare Advantage plans represent the federal government's experiment with delivering care to seniors through private insurance companies.  The experiment started as an alternative to the pay-for-whatever-the-doctor-orders model of traditional medicare.  Traditional medicare is fee-for-service, which means there are payments made for any and all services delivered.  If I think a patient needs a stress test, the patient just goes and gets the test..no questions asked.  This can lead to some perverse incentives and results in a lot of services being delivered.  Like the parent that abdicates the responsibility of disciplining their child to a reform school, the federal government hoped private insurance companies would mete out the tough love needed to reign in costs.

Tough love meant I was going to have to get 'permission' from the insurance company to do a stress test on Mrs. Cassidy.  If my staff was unable to get the test authorized with the information I gave them, I would have to get on the phone with an insurance company physician to plead my case.  This seems like a win for the federal government from a cost standpoint.  Surprisingly, it's not.

The federal government pays private insurance companies a fixed amount for each enrollee in their plan.  In the beginning (prior to 1997), the federal government would pay 95% of what they would have paid had a patient enrolled in traditional Medicare.  This saved the government money, but resulted in MA plans making up only 13% of the total Medicare market.  In an effort to expand access and have more people enrolled in MA plans, the federal government increased reimbursement to MA plans in 2003 in a bill best known for introducing Medicare prescription drug coverage.

Enrollees in MA plans now cost the taxpayer $1000 more than if the patient had stayed in traditional Medicare, and enrollment in MA plans has risen rapidly to cover almost 30% of all Medicare eligible seniors.

The government has been active in trying to rectify the payment asymmetry by tying reimbursement to risk scores and value.  This has not proven to be without its problems.

Scoring risk: You are higher risk if they say you are

In 2004, the Medicare program began to tie payments to private plans to beneficiary risk score.  Risk scores are based on diagnoses coded during the year prior to the payment year.  Insurance companies responded by investing resources in ensuring 'appropriate' coding.  The patients didn't change.  But if the patient enrolled in a MA plan, they became higher risk.  The evidence clearly showed insurers were upcoding to raise payments.

Measuring value: Harder than it sounds.

Payments were also tied to value -based payments to reward MA plans that provided high quality care. Quality in the MA program is graded on a 5 star scale that is determined from a weighted scale comprised of variables that include adherence to best practice processes (adults should get flu shots) and outcomes.   An example of an outcome measure is blood pressure targets.  In 2010, the MEDPAC report on quality used blood pressures < 130/80 in diabetics as an outcome measure.  Unfortunately, 2010 was also the same year that the ACCORD trial examining aggressive blood pressure targets in diabetics was released: Targeting a systolic blood pressure target of < 120mmHg compared to < 140mmHg did not reduce the composite endpoint of fatal/non fatal cardiovascular events.  The world of hypertension was so spooked by this trial that the panel convened to create a national guideline using this trial as evidence that lower was not always better when it came to blood pressure.  So it is entirely possible, that in 2010, 5 star plans were those plans that did harm to diabetic patients by having lower blood pressure targets.

In brief, the federal government's experiment in risk-adjusted, value-based payments to private insurance companies has only served to shift health care dollars to insurance companies at increased cost to the taxpayer.

Between January 1st and February 14th each year, if you happen to be enrolled in a Medicare Advantage plan, you can leave your plan and return to original Medicare.  John F Kennedy famously said, "Ask not what your country can do for you, ask what you can do for your country".  Mrs. Cassidy, do what's right for your country.  Choose traditional Medicare.

What the Wall Street Journal tells us about complications after surgery: Not much.

The Wall Street Journal published an article on Christmas day that told the story of an 83 year old woman who suffered a heart attack after a joint replacement at a rural hospital.  The story serves as an introduction to a piece about the higher cost and poorer care delivered at rural hospitals.  There are certainly some very interesting points I was not aware of with regards to financial incentives provided by the government to do procedures at rural 'critical access' hospitals, as well as higher 30 day mortality after joint replacement surgery at these rural hospitals.

The Wall Street Journal article does provide this nugget from a Harvard public health researcher: “Patients are getting bad outcomes, probably because they are getting procedures at hospitals without the experience to do it well.”  This certainly may be true, but no data exists in the article to back-up this assertion.  Are there more infectious complications of the surgery?  Are there more re-operations? Are the surgeons that operate at these centers less experienced?

The aim of the initial CMS initiative to expand access to care for rural patients seems to have worked. More patients are getting surgery closer to home as a result.  It is troubling that mortality rates are higher at these hospitals.  Perhaps the answer is to take away the incentives and move surgeries back to the larger hospitals.  I don't know, and the article isn't particularly helpful in answering those important questions. I would hope the folks in our profession who help shape public policy, like the Harvard physician quoted in the study, would be a little more careful in implying causation when all that has been discovered is a hypothesis generating correlation.  My hope is that he isn't advocating for changes to public health policy based on simple correlations.

I do take issue with the story used to make this point.  Unfortunately, a heart attack is a complication of any surgery.  It is not clear from the story what parts of care provided at this rural hospital were substandard.  Patients at teaching hospitals do have heart attacks post-operatively as well.  Differences in outcome may relate to delays in access to subspecialists you need in this situation.  Unfortunately, this article doesn't shed light on any of this.  It instead joins a laundry list of articles that leaves the distinct impression that something bad happened to a loved one at a hospital that was preventable. There are plenty of things I wish for.  I would love to have a light saber, I would love even more to travel at light speeds in the Millenium Falcon, and most of all I would love to live in a medical world free of any harm.  Elon Musk would lead me to believe we are a lot closer to flying in a Millenium Falcon than we are to working in a zero harm environment.  Surgery, especially, is not close to a zero risk endeavor.  While I have found estimates reported in the literature to be overstated at times, as many as 10 million patients (out of 200 million undergoing non-cardiac surgery) worldwide are estimated to suffer a major cardiac complication. There is a possibility that your 80 year old grandmother will have a complication of surgery even if it takes place at an ivory tower institution, and even if every single medical and surgical practice standard is met.

As we usher in 2016, I do have a solution that will definitely sate everyone's desire for zero risk.  In order to get to zero risk, I advocate we stop operating.  I feel safe in guaranteeing that there will be no complications of surgeries this coming year if there are no surgeries.

The quest for zero infections: A fool's mission?

Joyce is sick.  I am in the intensive care unit, peering at vital parameters that glow on the screen above her bed.  My eyes linger on those numbers because it is easier than looking at her.  A fever rages, her core temperature reads 103.4 degrees.  Her white hair is plastered on her forehead with sweat, and a tube to help her breathe emerges from her mouth and heads to a ventilator that angrily tweets a musical alarm every few minutes.  Her breathing is painfully obvious.  Her stomach moves paradoxically inward on every breath, and I can see the muscles in her neck tense with the effort of every breath.  Mercifully, her eyes are closed.  A nurse walks in and starts to change a bag of fluids that is hanging by her bed.  I follow the flexible plastic tubing that arises from the bag to an infusion pump, and then to a catheter that snakes under a see-through dressing underneath Joyce's left collarbone.  I ask the nurse about how long the catheter has been in place...'3 days'...I'm told.  I mutter about the possibility of a central line infection - the dreaded central line-associated blood stream infection (CLABSI).  The nurse shakes her head, and tells me - "we don't get those anymore".

CLABSIs are ground zero in the war on preventing patient harm.  The story entered the mainstream consciousness in the lyrical words of Atul Gawande in the New Yorker in 2007.  There he told a story of an unlikely Superman in the form of a critical care intensivist named Peter Pronovost. Dr. Pronovost was waging war against infections from these nefarious central lines that were saving and killing patients at the same time.  He published a landmark study in the New England Journal of Medicine that used an evidence-based intervention to dramatically reduce infection rates in the intensive care unit.  Some form of the implementation bundle that worked for Dr. Pronovost soon found itself in ICUs everywhere.  Dramatic reductions in CLABSI rates followed.

The remarkable part about this feel good story is that it seemed so easy.  The ICU I was in had seemingly reproduced this success as well.  I spoke to nurse after nurse who noted that CLABSI's were fairly rare events at this point.  Everyone had a different theory about how this happened.  Some attributed this to the full barrier sterile technique used for line insertion, others to the limitations put in place to draw blood from central line catheters, while others felt it was the maintenance of the catheters after they were placed.  Regardless, everyone had CLABSI on the brain.  The government had mandated that hospitals report their CLABSI rates, and with this, hospital resources to blot out this particular black spot were committed in earnest.  It should come as no surprise that infection rates would drop.  The size of the effect impressed me.  Hospitals, in my experience, are like large lumbering elephants.  It usually takes three forms, five signatures, and at the end of it all, I am told HIPAA won't allow it.  No matter what it is.

In 1958, Hawthorne Works, a Western Electric factory outside of Chicago commissioned a study to see if their workers would become more productive in higher or lower levels of light.  The workers' productivity improved when changes were made, but slumped when the study was over.  It was suggested the productivity gains were made because the workers were being observed. This has come to be known as the Hawthorne effect.  Not surprisingly the simple act of measuring, observing or testing has an effect on the performance of subjects.

There was no doubt that implementing a variety of maneuvers to reduce infections was having an impact, but this is not the whole story. In this particular ICU, for instance, I noticed that blood cultures were now rarely drawn from the central lines themselves.  They were now drawn peripherally.  This means hunting for a vein to thread a needle into to get a culture.  It was so much easier to take an empty syringe, hook it up to a catheter that was sitting in a vein and draw back 10 ml of blood.  The patient didn't have to be stuck with a needle and a nurse/physician didn't have to spend precious minutes hunting for a vein in a critically ill patient whose veins are either collapsed or buried in tissue laden with fluid.  Requiring peripheral cultures, though, does make plenty of clinical sense.

Cultures from catheters have a high false positive rate: organisms may grow that are colonizing the catheter but not causing an infection.  Deciding between the two - colonization or infection - can be challenging and is many times a clinical decision.  The CDC criteria for CLABSI in this regard is monolothic: any blood stream infection with a pathogenic organism counts against you.  Secondly, the act of collecting blood from a catheter is thought to raise the risk of introducing micro-organisms into the blood stream, and causing an infection.  These two points have been known for some time, but it was the specter of public reporting of these infections that really pushed hospitals to move in a systematic fashion to develop rigid protocols about cultures being drawn from central lines.  I applaud this new found fastidiousness with regard to cultures, but it does, however, make comparisons of infection rates somewhat challenging.  How much of the decline in infection rates is secondary to a more restrictive policy related to ordering blood cultures?  It is hard to know, and even Pronovost's landmark study cannot escape from the bias inherent in performing a non-randomized study that in essence tracked infection rates over time.  Regardless of the size of the true effect, Pronovost's accomplishment cannot be understated.  I can think of little that has been as clinically impactful in as short of a time.  To get hospital CEOs, CMO's, infection control committees, nurses, technicians and physicians on the same page with regards to evidence based practices for line infections is a true tour de force.

Of course, a hard lesson I learned after a peanut better and jelly sandwich orgy applies here too - there can be too much of a good thing.  Pronovost started on a mission to reduce infections, but yet that is not what we try to do now.  Our current mission statement is to get to zero.  CLABSI's have become the poster child for 'preventable harm',  a phrase that does not just imply,  but unabashedly states that any infection associated with a central line can be prevented.  Patients and stakeholders have carried this torch forward. Medicare has pushed forward with policies that would withhold reimbursement for any CLABSI's, because of course, they are all deemed preventable.

The evidence for a zero is non-existent.  Pronovost's study reduced infection rates a remarkable 66%.
A 2012 study provocatively titled: Zero risk for central line-associated blood stream infection: Are we there yet? found that we were not there yet.  The study based in 37 ICUs in South Wales sought to identify the longest time a central line could stay free of infection with another 'insertion intervention bundle'.  The authors reported a significant reduction in rates of infection with their prevention bundle, but importantly found no catheter dwell time that was associated with zero risk.  It should not take a study to say this.  In today's non-Star Trek world, the reality of critical care medicine frequently involves the insertion of central line catheters that are life saving.  These catheters are placed, not in the vacuum of space into sterile objects, but rather within the milieu of a hospital teeming with pathogenic organisms, and into patients that are far from sterile.  Our attempts to sterilize the human environment is an exercise in reducing the burden of micro-organisms, not eliminating them.

Tying reimbursement to an impossible goal places inordinate pressures on hospitals, which have consequences that are unintended, but quite predictable.  The response of hospitals and their staff takes many forms, and I have already written about the new found parsimony in blood cultures being drawn from central lines.  Other maneuvers are less sanguine.  A study done to compare infection control practitioners  to a standardized computer algorithm had disturbing results.  The medical center that had the lowest rate of central line infections as judged by infection control practitioners (2.4/1000 central-line days) had the highest rate by the standardized computer algorithm (12.6/1000 central-line days).  The study authors concluded that the variability of infection rates seen when compared to a reference standard suggests significant variation in the application of standard CLABSI definitions.  It would appear that the CLABSI definition was being applied in a subjective fashion to report rates much lower than an objective standard.  Welcome to a world that tells you what you want to hear.

The narrative that surrounds the noble sounding 'prevention of harm' strips meaning from the very phrase when applied in such a liberal, incoherent fashion. The reality is that patients arrive at hospitals and critical care units in extremis.  In that hospital room we struggle to make Joyce whole again.  It is a struggle we may not win because the very interventions we use to save a life can also take a life. The fervent hope of everyone in Joyce's room is that on balance the benefits of the interventions taken are greater than the risks.  Over time, mortality rates have fallen in intensive care units - a fact that is remarkable given that patients in ICUs today are sicker than ever.  This fall in mortality is driven by progress in treatment of underlying disease states, as well as improvement of processes in medicine that predates ProPublica-style transparency and 'performance'-based reimbursement.
Remarkably, yet predictably, the current iteration of 'transparency' combined with unrealistic expectations may be doing more harm than good.

The magical world of funding for the Affordable Care Act

Congressional leaders just agreed to a budget that would keep the government open through September 2016.  I was happy to hear the government was not going to shut down.  I was much less happy to hear about the fate of provisions supposed to fund the Affordable Care Act (ACA).  The ACA - costing $1.2 trillion over 10 years - was supposed to 'mostly' pay for itself.  Revenue was to be generated (in large part) by a series of taxes on a variety of different sources.  These taxes did not fare so well in the current budget.

'Cadillac' Tax
The ACA took aim squarely at high cost employer-sponsored plans.  Economists believe that since employer health insurance is tax deductible, high cost plans proliferate as a mechanism to provide a tax free benefit to employees.  These expensive plans are expensive because they cover most of the cost of medical care, insulating the patient from the actual cost of medical care.  The ACA imposed an annual 40% tax on plans with annual premiums exceeding $10,200 for individuals and $27,500 for families to be paid by the insurers.  The results were to be two fold: One, create a disincentive for employers to offer 'cadillac' plans, and two, generate revenue to pay for the ACA.  A broad coalition composed of democrats and republicans lobbied to defeat this tax.

Medical Device tax.
A tax was imposed on medical device companies, in part because expanded insurance coverage was believed to increase their revenue.  This tax has also been 'delayed'.  It is estimated that this will cost the federal government $2 billion dollars per year.  A quick review of stock prices of three dominant cardiac device companies show major increases since signing of the ACA into law in 2010.  For reference, the S&P 500 gained 50% in the same period of time.

Health Insurance Provider tax
The insurance companies that were supposed to see more revenue as a result of taxpayer subsidized coverage had an annual fee they had to pay.  This annual fee was related to the dollar amount of the net premiums written during the prior year.  In 2016 that fee would have amounted to $11 billion dollars. The insurance companies argued that the fee would 'have' to be passed on to the consumer in the form of higher premiums.  This is the same insurance industry that has seen significant stock gains and returns for their shareholders as well since the ACA was signed.

The White House obfuscates

No statement from the whitehouse has been forthcoming on how this revenue will be made up. I suspect the whitehouse team would point to healthcare savings because of the ACA. They would, of course, be wrong.

From the obamacare facts website:

"In 2015, due in part to the ACA, health care spending grew at the slowest rate on record (since 1960). Meanwhile, health care price inflation is at its lowest rate in 50 years." (accessed 12.16.2015)

This hyperlink references CMS data from 2013 that healthcare costs were growing at the slowest rate since 1960. The whitehouse took credit for bending down the cost curve despite the fact that the slowdown predated the signing of the ACA. Despite the claims on the obama website, the actual 2014 health care cost data reflects quite a different story. 

 In 2014, Health care care expenditure is projected to have increased 5.5% (year over year), the first time growth would be higher than 5% since 2007. As a consequence, in 2014 health care expenditures are now projected to be 17.7% of GDP, the first increase in the health care share of the economy since 2008. Why did health care costs go up so much? As per the economists at CMS (Health Affairs) the rise in healthcare costs are mostly related to the expansion of healthcare coverage under the Affordable Care Act (ACA). Not surprisingly, expanding coverage to millions of people costs money. To reiterate, the ACA is largely responsible for changing the trajectory of the health care cost curve in the wrong direction

Of course, this was before the most recent iteration of the budget that deep-sixed revenue supposed to 'self-fund' the ACA.  Don't even get me started on the revenue that was supposed to come from reductions in overpayments to private Medicare Advantage plans.

The ACA, of course, continues to be funded.  The White House quotes 2013 numbers when talking about 2015.  It's magical.  Maybe Santa Claus is for real.  

Risk scoring/value based payment models haven't worked: The MedicareAdvantage story

Mrs. Cassidy slowly walks into my office one busy afternoon.  I see her out of the corner of my eye because she is so hard to miss.  Mrs. Cassidy has some serious style.  She is wearing a deep orange dress with a bright blue blazer.  There aren't too many folks that can pull that outfit off, but she can.  She has a wide slow smile, and she speaks with a slow southern drawl that belies her southern roots.  This was supposed to be a routine follow up visit for a 67 year old woman with a history of a mechanical mitral valve replacement and coronary disease.  Unfortunately, she tells me a story that is concerning for angina.  I think she needs a stress test. I slide over to the insurance tab on the EMR and I let out a somewhat audible groan.  She has a Medicare Advantage (MA) plan.  I explain to Mrs. Cassidy that we will need to go through an extra step to pre-certify her stress test.  She expresses surprise and asks me what she should do.  I will tell you what I told her, but first, let me tell you why.

Medicare Advantage Plans: A brief history

The concept of receiving medicare benefits through private health plans has existed since the 1970's, providing an alternative to traditional fee for service (FFS) Medicare.  The federal government pays private insurance companies a capitated amount per enrollee, with the idea that the private market would be more cost effective than traditional Medicare.  Prior to the Balance Budget Act (BBA) of 1997, Medicare paid 95% of average traditional Medicare costs to private plans. The idea behind lower payments seemed sound.  Managed care had more flexibility with paying providers (i.e. using gatekeepers, demanding prior authorizations for tests, etc.), and managed care patients were, in general, healthier.  Under these rules, however, the private market was a relatively small slice of the total Medicare pie, covering only 13% of all Medicare enrollees.

The year everything changed was 2003.  The Medicare Prescription Drug, Improvement, and Modernization Act (MMA) is best known for introducing prescription drug coverage to Medicare patients (Medicare Part D).  Less well known is the provisions within the bill to provide a major increase in government contributions to private (Medicare Advantage) plans.  As a result, Medicare now pays more for private plans per enrollee (~$1,000) than the cost of care for beneficiaries in traditional Medicare. In 2014, payments to MA plans totaled ~$159 billion.

Not surprisingly, enrollment in MA plans has exploded since 2003.  MA plans went from enrolling 13% of Medicare beneficiaries in 2003 to 30% of beneficiaries in 2014.   To be fair, much of this extra payment is mandated to be used as benefits for MA enrollees.  This can take the form of extra benefits to plan members or premium relief.  This still sets up a payment scheme that favors those enrolling in MA.  Sign up for an Advantage Plan! Get free stuff that the 70% of folks enrolled in traditional Medicare will pay for!

Scoring risk: You are higher risk if they say you are

The current mantra about cost in health care places a large portion of blame on a fee for service culture.  You will find little argument from me that paying per unit of healthcare delivered will result in a lot of health care being delivered.  However, the prevailing ideology that states bundled payment models will be cost effective is more faith-based than evidence-based.  The history of payment to MA plans is quite instructive in this regard.  The idea of capitated payments to private insurers started as an attempt to cut costs.  Capitated payments, however, encourage insurers to enroll a lower cost, healthier enrollee, and not surprisingly, that is exactly what insurance companies have done, and continue to do.  Offering free gym memberships (Silver Sneakers) is very nice, but it also does tend to attract beneficiaries that can go to a gym and workout.  The end result is that the federal government pays more for MA enrollees relative to their costs in traditional Medicare.

The solution seemed clear.  In 2004, the Medicare program began to tie payments to private plans to beneficiary risk score.  Medicare calculates payments to plans separately for each beneficiary, multiplying the the plan's payment rate by the beneficiary's risk score.  Risk scores are based on diagnoses coded during the year prior to the payment year.  Insurance companies responded by investing resources in ensuring 'appropriate' coding.  The appropriately titled "Upcoding: Evidence from Medicare on Squishy Risk Adjustment" by the National Bureau Of Economic Research (NBER) found that a 10% increase in MA market penetration lead to a 0.64 percentage point increase in the average risk score in a county.  MA plans generate risk scores for their enrollees that is on average 6.4% higher than what those same enrollees would have generated under traditional Medicare. Moreover, consumers choosing traditional Medicare have similar risk scores and diagnoses in their employer plans at age 64 and in traditional Medicare at age 65, but consumers choosing MA show a boost in risk scores and diagnoses the year after their transition.  Who would have imagined it?  Simply enrolling in a MA plan makes you higher risk.

The other nostrum to cure all ills of the current health care system proselytized by fast talking policy makers and politicians are value based payments.  Once again, the MA experience provides lessons aplenty.

Measuring value: Harder than it sounds.

Quality in the MA program is graded on a 5 star scale that is determined from a weighted scale comprised of variables that include adherence to best practice processes (adults should get flu shots) and outcomes (diabetics should have blood pressures < 130/80).   These specific parameters come from the 2010 MEDPAC report on comparing quality in medicare plans.  The problems are readily obvious.  2010 was also the same year that the ACCORD trial examining aggressive blood pressure targets in diabetics was released: Targeting a systolic blood pressure target of < 120mmHg compared to < 140mmHg did not reduce the composite endpoint of fatal/non fatal cardiovascular events.  The world of hypertension was so spooked by this trial that the panel convened to create a national guideline used this trial as evidence that lower was not always better when it came to blood pressure lowering.  That was until Nov 2015, when the SPRINT trial (that excluded diabetics but included patients at high risk of cardiac events) did show a mortality benefit of targeting systolic blood pressures less than 120.   So it is entirely possible that in 2010, plans were incentivized to do harm to diabetic patients by having lower blood pressure targets.  

Mired in this quicksand of higher cost and who-knows-what-value care, there was no damsel in greater need of being rescued than the health care system.  Riding to the rescue was the Affordable Care Act (ACA).  Unfortunately, the ACA has proved so far to be more Don Quixote than Sir Lancelot.

The Affordable Care Act (ACA)  and Medicare Advantage plans

The ACA sought to balance the scales by reducing the extra payments to MA plans.  The idea was to stop overpaying certain parties, and use the money saved to help pay for provisions of the ACA.  This seemingly reasonable course of action has been fairly bumpy.  In 2014 and 2015, CMS announced that Medicare Advantage payments were set to rise for the following years.  This reversed a decline proposed earlier in the year, and came after a major lobbying effort from the health insurance industry, as well as members of both political parties.  So much for Sir Lancelot.

In brief, the federal government's experiment in risk-adjusted, value-based payments to private insurance companies has been a failure.  It has only served to shift health care dollars to insurance companies at increased cost to the taxpayer.  Our senior citizens do, happily, have more 'free' gym memberships as a result. 

Adding insult to injury, the same insurance companies being paid more for each enrollee increase the degree of difficulty for physicians trying to take care of their patients.  Today, my staff has to spend time pre-certifying a test; tomorrow promises to be an even greater time-sink as payers demand even greater reporting of practice demographic and outcome data.  One would think the people in charge would have learned lessons from the MA misadventures.  After wading through meaningful use stages, PQRS initiatives, and queries from clinical documentation specialists about what type of respiratory failure my patient had, I can most definitely say: They have not.

Between January 1st and February 14th each year, if you are enrolled in a Medicare Advantage plan, you can leave your plan and return to original Medicare.  John F Kennedy famously said, "Ask not what your country can do for you, ask what you can do for your country".  Mrs. Cassidy, do what's right for your country.  Choose traditional Medicare.