regulation, part 2: health care

Peter Orszag had some very interesting thoughts on health care reform in recent issues of the New York Times. In his most recent piece, "To save money, save the health care act" published last week, Orszag freely admits that the plan has flaws, but also makes the legitimate point that fundamental reform of a health care system devoid of any form cost control (or for lack of better terminology, health care rationing) is going to be messy. I think the piece has some flaws, but I think this statement is an important one to note:

There are four ways to contain health care costs: by reducing payments to providers and suppliers; by rationing services; by having consumers pay a greater share; and by giving providers incentives to be more efficient.

So, lets consider each of those methods individually.
1) Reducing payments: Despite the shooting myself in the footishness of this, its certainly necessary. Part of the equation is that we are spending too much money on health care, but reducing costs means reducing compensation. So what should we paying for and what shouldn't we be paying for? Well, the fact of the matter is that we should be decreasing compensations for interventions that are not proven to improve overall prognosis (like: vertebroplasties, knee arthrocenteses, 4th line chemotherpeutic agents, renal replacement therapy in patients with hepatorenal syndrome). Conversely, we should be creating and/or increasing compensations for primary care interventions that reduce cost-ineffective, but necessary procedures (aka, effective smoking cessation counseling, vaccinations, obesity management).

2) Rationing services: Now, part and parcel of reducing payments is going to include reducing the number of health care services that should be provided. How do we decide what to ration? With QALYs, of course. See Peter Singer's NYT article "Why we must ration health care" for elaboration.

3) Having consumers pay: Certainly, increasing consumer contributions to the health care pot will help reduce the incurrence of health care debt. But high premiums seems to be at the core of people's discontent with the health care system. And while health care rationing will improve premiums, I don't think that people will be willing to offset those savings with increased payments. Rather, the answer probably involves insuring some of the 47 million Americans who are currently uninsured, perhaps with catastrophic and preventative insurance coverage that will reduce the significant amounts of debt they incur. This would be the same population in which reduction of health care interventions not shown to improve mortality would first be implemented.

4) Improve provider incentives: Finally, the issue of how providers are motivated. This issue has been broached by Atul Gawande, and is addressed by both Orszag and the health care bill, by tracking and ultimately attempting to reward and penalize providers based on the health of their patients rather than a fee-for-service model that rewards procedures rather than outcomes. There is a clear precedent for this: the No Child Left Behind Act, which provides a negative incentive for poor student performance. This seems alright, except for the fact that it may discriminate against school districts where students are more difficult to motivate. Similarly, infusing accountability into a compensation system for providers might penalize the providers who face the most daunting sets of patients (many comorbidities, poor compliance).

Certainly, one clear negative incentive that can be removed is the practice of "defensive medicine" in which doctors prescribe additional tests and/or interventions simply to avoid being sued for for being insufficiently aggressive. This practice, as delineated by JAMA, is widely pervasive, and seems to primarily take the form of unnecessary imaging to rule out highly unlikely pathology. Tort reform will doubtlessly reduce the practice of defensive medicine, though the extent of cost savings is difficult to estimate. And for those of you fuming over Obama and Orszag's socialist tendencies, let there be no mistake that both of them agree with tort reform.

Other (or even undiscovered) methods of cost reduction

One of the final aspects of the health care plan, one of the least mentioned, and probably my favorite, is the pool of money (the so-called Innovation Center) dedicated to supporting cost-effective initiatives. It's essentially a fund to say "If you have an idea to save money, we'll give you a little money to try it, and if it works, we'll give you more money. So go to town." Sounds like typical government waste? Well, as Atul Gawande noted, this was pretty much the model that led to agricultural reform. As he states, the answer involves engaging local providers to institute reform from the ground up:

At this point, we can’t afford any illusions: the system won’t fix itself, and there’s no piece of legislation that will have all the answers, either. The task will require dedicated and talented people in government agencies and in communities who recognize that the country’s future depends on their sidestepping the ideological battles, encouraging local change, and following the results. But if we’re willing to accept an arduous, messy, and continuous process we can come to grips with a problem even of this immensity. We’ve done it before.

Was it worth it?

I'll close with an extreeeemely interesting piece by William Saletan at Slate. In it, he vociferously defends the Pelosi-led congress, and refutes the idea that the resetting of congressional balance reflects a failure of governance by Pelosi. Rather, he states, she did what she ought to have done: used her Democratic majority to institute legislation that she felt supported the common good.

Politicians have tried and failed for decades to enact universal health care. This time, they succeeded. In 2008, Democrats won the presidency and both houses of Congress, and by the thinnest of margins, they rammed a bill through. They weren't going to get another opportunity for a very long time. It cost them their majority, and it was worth it. And that's not counting financial regulation, economic stimulus, college lending reform, and all the other bills that became law under Pelosi. So spare me the tears and gloating about her so-called failure. If John Boehner is speaker of the House for the next 20 years, he'll be lucky to match her achievements.

Awesome. Coming soon: GABA in stroke, chemokines in asthma!

The Purple Rose of Cairo

I've seen almost every movie Woody Allen's made but, for some reason, I always held out on seeing this one. I think something about the name reminded me of "The Curse of the Jade Scorpion" which is almost definitely his WORST movie.

Anyway, I just finished watching it for the first time and I am absolutely stunned. He beautifully weaves silly romanticism with devastating realism in a way that is in keeping with his style but he was never so on the mark. I think he gets closest with "Hannah and Her Sisters" but this movie has more of a resemblance to Fellini's "Nights of Cabiria." I'm glad I waited to see this one.



It's nice to discover new things.

regulation, part 1

Well, that's the best way I can make four things that I'm going to write about seem related, when they're not related.

1. Autoregulation of salt-water balance by the kidney. This has been a topic of interest for me recently, because I'm rotating on the Nephrology service at Bellevue. As a result of this, I've been trying my best to understand the kidney.

So, there are two major functions of the kidney. The first is to filter and excrete toxins and toxic metabolites (important, but not that interesting). The second is to sense, and carefully regulate, total body fluid and electrolyte balance. This is a difficult, and essential job: too little fluid and the body can't deliver oxygen to vital tissues; too much fluid and it starts to accumulate in the lungs, abdomen, and extremities. In healthy patients, the kidney responds quite acutely to changes in volume (changes so small that they may be beyond the limit of our measurable detection). This appears to be via a set of cells known in aggregate as the juxtaglomerular apparatus, or JGA. The JGA is composed of 2 groups of cells. First, the granular cells, which line the afferent and efferent arterioles supplying and draining each glomerulus, are thought to act as 'pressure sensors' that react to miniscule changes in arteriolar pressure and respond with appropriate vasodilation or vasoconstriction. This transmits small increases or decreases in blood volume to small changes in glomerular filtration in the kidney. Second, the 'macula densa' cells within the distal tubule sense changes in delivery of tubular fluid to the distal convoluted tuble and adjust GFR and proximal reabsorption (via renin-angiotensin-aldosterone system activation).

Of course, there are well known situations in which renal autoregulation goes wrong. The most common of these diseases is CHF, in which decreased cardiac output results in reduced renal artery perfusion; this is pathologically interpreted as a sign of low volume by the renal JGA, resulting in sodium retention, pulmonary vascular congestion, and peripheral edema.

Another common syndrome of renal dysregulation is advanced liver cirrhosis. Fluid balance is markedly altered in cirrhosis, as demonstrated by significant accumulation of fluid in the peritoneal cavity (cirrhosis) - this is thought to be due to both hypoalbuminemia that results from decreased hepatic synthetic function and pathologic sodium and water retention by the kidney. This is evidenced by chronic high renin, aldosterone, and vasopressin levels seen in cirrhotic patients, and low urinary sodium characteristic of increased proximal tubular reabsorption. However, the stimulus for sodium retention by the kidney in cirrhosis is unknown. A number of hypotheses have been put forward to explain this phenomenon:

the low plasma volume hypothesis: cirrhosis-induced portal hypertension (increased hydrostatic pressure) + hypoalbuminemia (low oncotic pressure) leads to fluid transudation, low plasma volume, and RAAS activation. Sounds good, but doesn't work, as direct measurements of plasma volume in 1967 found that patients with advance cirrhosis have high, not low plasma volume. Furthermore, an elegant temporal analysis of fluid overload in a dog with cirrhosis in the American Journal of Physiology in 1977 demonstrated that high plasma volume in the splanchnic circulation as well as sodium retention preceded portal hypertension.

the underfilling hypothesis: popularized by Robert Schrier at the University of Colorado, this suggests that systemic vasodilation leads to decreased ‘effective’ arteriolar blood volume and leads to volume expansion. This is supported by the use of various therapies to improve arteriolar filling, including vasoconstrictors and head-up water immersion (to be discussed).

The key regulator in the underfilling hypothesis is thought to be nitric oxide, which induces profound systemic vasodilation and would lead to renal underperfusion and hepatorenal syndrome; however, evidence for this is lacking, as is evidence for the utility of nitric oxide synthase inhibitors in preventing or treating HRS. However, there are physiological reasons to suspect that nitric oxide production by the liver in response to portal hypertension is possible given the fairly strong evidence that the liver, and in particular, the hepatic artery, serves as pressure sensors;

1) electrical stimulation of portal nerves has potent effects on renal blood flow and urine output
2) other acute causes of portal hypertension, such as thrombotic disease (Budd-Chiari syndrome) induce potent hepatic artery vasodilation and hepatorenal syndrome, while elimination of portal pressures with intrahepatic portocaval shunts improves HRS in patients with portal hypertension.

Treatment options for hepatorenal syndrome continue to be quite limited, but are based ultimately on the idea that systemic vasodilation secondary to portal hypertension leads to renal arteriolar underfilling and prerenal failure. The only current therapy with any basis for efficacy is a study published in the Journal of Clinical Gastroenterology in 2009, in which a combination of midodrine/octreotide (splanchnic vasoconstrictors which theoretically counteract systemic vasodilation) along with albumin (which exerts positive oncotic pressure to hold fluid within blood vessels). This combination was the first to demonstrate a mortality benefit in patients with hepatorenal syndrome, although it has been difficult to reproduce in follow-up studies.

In my quest to find weird, alternate, cost-effective therapies, however, I came across a somewhat old practice that has fallen out of favor for reasons that are not entirely clear to me. It has been proposed for centuries that water immersion leads to increased urination; this concept of so-called “immersion diuresis” is thought to be secondary to two properties: thermal heat loss and hydrostatic water pressure, both leading to systemic vasoconstriction, improved renal perfusion, suppression of ADH, and natriuresis. In case you’re wondering, this idea is what led ultimately to the clever prank in which you put a sleeping friend’s hand in water causing him or her (who are we kidding – its always him) to pee himself.

That wouldn’t much thrill Dr. HC Bazett, the physiologist from the University of Pennsylvania who first published the effect of water immersion on diuresis in humans. Dr. Bazett’s subsequent studies found that the pressure component was far more important than the temperature component; aka, total body immersion is required to achieve significant diuresis (take that, pranksters).

Subsequently, head-out water immersion (pictured above, looking like a torture machine, and so named, according to the illustrious Dr. David Golfarb, because “some idiot medical student probably drowned some poor cirrhotic, so they had to put ‘head-out’ in the title”), was shown to improve venous return, increase right atrial pressure, increase stroke volume, and improve cardiac output.

A number of nephrologists, including Dr. Schrier, applied this data to their particular field and organ of interest, demonstrating that head-out water immersion could indeed improve renal function by decreasing renin-angiotensin mediated vasoconstriction and improving renal perfusion and urine output (right, from a letter to the editor published in N Engl J Med in 1983). And while studies on the true efficacy of this practice are somewhat conflicting, one could easily say the same about the far more expensive therapy of octreotide, midodrine, and albumin. So why don’t people ever try this? Beats me.

Yikes, this post was long. Stay tuned for regulation part 2, featuring: asthma, strokes, and health care policy!

mr. arkadin

"I knew what I wanted. That's the difference between us. In this world there are those who give and those who ask. Those who do not care to give... those who do not dare to ask. You dared."

God, this movie sucked.

since nobody will post about lung cancer screening

I'm working on something about homeostatic control of volume status.

This is where the fish lives.

The fun starts around 7:30...

c-dif: all hands on poop deck

Clostridium difficile was first described over 75 years ago by Hall and O'Toole, and initially named 'Bacillus difficilis' not because of its virulence but rather because of the difficulty culturing the bacteria ex vivo. They found that this bacteria asymptomatically colonized 40-70% of neonates within the first month of birth. In fact, c-dif is a slow-growing, indolent bacteria whose morbidity has been unleashed exclusively by iatrogenic eradication of gut flora by (you guessed it) doctors. It turns out that c-dif's relative inactivity makes it insensitive to most standard antibiotics; therefore when a patient is treated broadly with a penicillin, cephalosporin, or fluoroquinolone for a pneumonia or cholecystitis, the coast becomes clear for c-dif to spread and do its damage. The damage is done primarily by thenefarious work of 2 toxins, appropriately named c-dif toxin A and toxin B (CDTa and CDTb, respectively). These toxins disrupt the actin cytoskeleton-stabilized tight junctions between intestinal epithelialcells via UDP-glucose-dependent glycosylation and inactivation of Rho and Ras protiens, leading to epithelial sloughing and subsequent inflammation.

Over the years, the liberal use of antibiotics in hospitals has driven increasing virulence of c-dif, such that a hospital stay of 2 days or longer now results in a 10% risk of contracting c-dif colitis. The risk appears to be highest in the elderly, possibly because of their impaired immune response. This virulence began to progress to the point that c-dif outbreaks began to emerge in the community and in absence of recent antibiotic therapy, first in Hungary in 2004 but quickly in the US by 2006, and particularly in patients on gastric acid-suppressive therapies.

One of the main concerns with controlling c-dif infections is the frequency of recurrence (20-33% after the first infection, 65% or higher in those with more than 1 occurence). And while c-dif colitis is treatable with either metronidazole or vancomycin, antibiotic efficacy decreases in subsequent recurrences. Recurrent c-dif appears to be more virulent based on upregulation of CDTa and b toxins; yet this may also be a key to its treatment. One of the fun things about emerging therapies for c-dif is how 2 of the most promising therapies are about as different as can be: recombinant humanized monoclonal antibody therapy versus....someone else's poop.

The high-tech answer: passive and active immunization.

The first hint that the immune system might be of some help in fighting off recurrent c-dif came from a study in 2000 published in the New England Journal. In it, Kyne and colleagues showed that patients who had c-dif colonization without infection had high serum levels of antibodies against toxin A. They followed it up with a study published in Lancet one year later in which they showed that following acute c-dif infection, the patients who were 'immune' to recurrence made IgM antibodies against toxin A within 3 days. Similar results were later shown for c-dif toxin B. It took about 6 years for fully humanized monoclonal antibodies to be made against both toxins and shown to protect hamsters from recurrent c-dif (Infect Immun. 2006, 74(11): 6339-47).

4 years later, in 2010, a landmark study published in NEJM showed that administration of mAb against both toxins A and B resulted in a relative risk reduction of 70% in patients with 1 episode of c-dif colitis and 80% in patients with more than 1 prior episode. The inherent problem with this strategy is that it provides only passive immunity, and therefore patients would require frequent infusions. A more lasting solution would be to induce active immunity in patients with vaccination against target peptide antigens. In 2003, a vaccine composed of immunogenic portions of toxin A was able to induce significant titers of IgG in patients, and case reports of 3 patients with severe recurrent c-dif colitis showed that two out of three patients who were vaccinated generated high titers of antibodies to both CDTB and CDTA and induced remission without recurrence. A phase II trial in patients aged 18-85 with 1+ prior episode of c-dif colitis is currently underway.

The low-tech option: your poo, please.

The other potential solution to c-dif is simple, logical, fairly gross, and impressively efficacious. It goes by the name of "fecal flora reconstitution," which is a fancy way of saying "poop transfusion." It's based on two major ideas: first, that patients who succumb to c-dif infections have not only a poor immune response to c-dif, but commensal gut flora that don't protect them against c-dif overgrowth, and second, that the best way to overcome a deadly, but frankly wimpy bacteria like c-dif is to re-introduce a set of much more competitive gut flora. Thus: poop transfusions. It's simple; check the donor for communicable diseases as well as a history of c-dif exposure, liquefy their feces in non-bacteriostatic saline, dump it into a colonoscope, administer the donor feces at the level of the ileum, have the patient take immodium to maximize contact time between the new flora and the gut epithelium, and watch the magic happen.

And, for the most part, it does. Almost every study, despite variability, demonstrates sustained responses in refractory patients with only one dose in >90% of patients, and response to a 2nd dose in the majority of the 10% of patients who don't respond initially. The point? if you can get over slightly stomach-turning concept, its an incredible, cost-effective solution to an iatrogenic problem. And its sure to make one bond in your life stronger - after all, once you share poop, what can't you share?

***

I'll leave you this election night, in which 4/10 voters verbalized support for the "death panel" touting Tea Party, with the thoughts of David Meltzer of the University of Chicago's department of Medicine and Economics, on the need to accept rationing for what it is (a fact of life in the health care system) so that we can have some real debate, not over whether we should ration (a healthcare system without it is ludicrous) but rather, who should be in charge.