Thursday, March 31, 2011

Prostate profits

A recent decision by Medicare to cover a new vaccine by Dendreon (Provenge) for advanced prostate cancer (development story here) was anticipated by lots of speculation on the reimbursement price.  I wanted to reflect on this as kind of a general case in medical 'science.'

Certainly, advanced prostate cancer is no picnic. But it is likely vastly over-treated, especially in older (> 70 yo) men.  Nevertheless, goals to extend life through more treatment continue to drive research and folks, like Dendreon, occasionally can persevere and bring new technologies to market that demonstrate some advantages.  In the case of Dendreon, the advantages are an extended mean life expectancy for roughly 21 months to 25 months.  Four months.  Two notions I'd like to ponder here.  The first is simple, just about the mean or average, where it is just that, an average and some folks in both groups had higher and lower life expectancies.  One quirk of these kinds of statistics is that the 'average' cannot be applied to individual cases.  Huh?  A clinician could recommend this vaccine to patients saying something like, "In a published study, folks in the group taking the active drug survived, on average, 4 months longer than those who did not."  If the patient hears, "If I take this drug, I will survive an extra four months," they are wrong.  Some of the patients taking the drug may have survived 4 months longer, maybe not.  Certainly, some patients survived more that an extra four months, maybe a lot more.  And then some died before they reached four months.  Oh, and some died, even though they were taking the drug, sooner than the average person NOT taking the drug.  So any individual patient could have any of these individual outcomes.  That is now a bitch to figure out.  Sorry.

Okay.  Gamble with the drug.  Just have to have someone pay the freight.   $93,000.   You heard me.  $93 K, for 3 treatments in a month.  Of course, as a monopoly, what's a payer to do?  Either deny coverage (generate outrage) or suck it up.  The beauty of this example is myriad.  Since Medicare only pays 80%, there is an $18,600 co-pay.  Some folks might have that, or part of it, covered by supplemental insurance, but it still will be a  chunk of change. Think of the bank of dunners at the call center, haranguing elderly cancer patients for the co-pay.  Better yet, the mark-up that will be charged by clinicians, but embedded within the $93 K.  that works because Dendreon will give discounts to big purchasers, who then can mark it back to the $93 K and pocket the difference.  Of course, that would increase the likelihood of prescribing, one would think....

Lastly, this fits into the idea of a 'prefect' medical advance.  The therapy isn't curative, so the patient stays in the pool of revenue generating cases and the therapy does not require a reduction in any other intervention. The cost simply gets added to all the existing costs.  And if the average patient does get the 4 months, that is 4 more months of production, in terms of opportunities to render care.  Every one wins.  Except for the payers.  Oh, and the poor slob who would have survived much longer if no one would have diagnosed him with prostate cancer in the first place. 

Thursday, March 24, 2011

The history in medicine




No, not the history of medicine (which is fascinating, but better covered by lots smarter people than me.  Sites like the National Library of Medicine serve up some nice treats in that vein, especially the images like this one.


But I am more thinking about the nature of history that is less tangible, the things that persist in our practices (of course, the practice of medicine, in my case), which are subverted into the present without shape or form, but real in that they sort of line the basic parameters of what we perceive, and more importantly, how we perceive.  No, not in some conspiratorial manner, as there is no 'agent' if you will, no governing body or authority that deems it so, but more of an opaque background and lens that shapes the way we act, without a conscious presence.  And yet, it is pervasive and inescapable within the current activities of medicine and medical science (and equally so in virtually any other domain, no matter the 'realness' of the observations.  For an example from the more purportedly 'hard sciences,' see the nice piece by Traweek on the culture of physics, for example: Beamtimes and Lifetimes).


Dali's representation, "The persistence of memory" is about as useful an image to convey this, the way these misshapen events linger, as I have found.  Plus it gets Dali into a blog.

For example, I think folks would universally recognize that Alzheimer's disease is a significant and growing challenge to the medical system, from the patients and caregivers through the whole therapeutic and health care system.  But the history of Alzheimer's is fraught with indecision and difficulty, from the first 'recognition' through current research and interventions (see the Atlas, first chapter on history, but all the detail you like in a long download).  The simple history is that cognitive decline was historically an expected outcome of long lives, such that the period of life after 60 (when lifespan was less than 50) was called the "senile period."  It was when folks presented with 'senile' symptoms presented earlier in life that they were seen as some novel pathology (the index case was a 50 yo woman).  There was some interesting debate about the cause of the disease and it is likely that many readers know that there are abnormal formations or 'plaques' that from in the brain and appear to cause the symptoms.  Of course, other kinds of dementias also have these plaques (in even 'normal aging, as well in specific neurological disorders, such as Pick's disease). But we can leave the details and think about the simple history of the first, or 'index' case.  A woman, not old, shows up in the office and has lots of trouble with memory, personality changes and activities of daily life.  The 'naming' of the disorder takes some years, and that naming identifies a set of gross symptoms, a collection now called a disease, Alzheimer's. 

Fast forward.  With advanced brain imaging techniques, functional MRIs, PET scans and others, the actual activity and state of the brain can be observed and used to make this kind of diagnosis.  Of course, this means that some folks who might not have had Alzheimer's based on clinical presentation do meet the criteria for the disease and folks who look like they must have Alzheimer's do not meet the diagnostic criteria.

Now identified, the hunt for treatment (and of course, the "Cure") is engaged enthusiastically.  Current R & D spending for treatment is approaching $1 Billion (costs of care, for contrast, are estimated to be around $168 Billion dollars).   The focus of that research is my interest here, and the place where the persistence of memory both lurks and drives activity.

The vast majority of that expenditure, both public and private, is aimed at developing drug interventions that mitigate some of the cellular processes that are seen in the disease and its progression, from inhibiting plaque development to improving glucose utilization in the brain.  But the disease was not recognized because of tangles in the brain, impaired glucose metabolism or increased baseline inflammation.  Alzheimer's was a disease of behavior, not biochemistry (mostly because we did not have access to things like "biochemistry" when most clinical diseases were 'recognized.').  The legacy of the 'clinical' entity haunts the research practices today, where the focus on treatment is always viewed through the notion that the basic biology must have some direct connection to the clinical entity.  So, like all the 'research' programs that were born out of the recognition of clinical diseases, the biochemical correlates are always somehow related to the clinical manifestations in some causal, meaning linear, way.

The analogy is that by looking at the weather today (the current end stage outcome of all weather up to this point), we can find some 'knowledge' of the weather a week ago, or a month ago (biochemical activity).  But that is what we are assume, and then invert that, to work from our assumptions in the biochemistry to the disease.

This commitment to maintaining a connection between the end-stage clinical outcomes and basic biological processes comes at a great cost.  The clinical entities become constituencies, competing for limited resources.  It isn't about understanding the commonalities in chronic disease and addressing those, so the benefits of that therapy accrue to any of the end-stages.  The researcher in Alzheimer's and cancer, in kidney disease and in lupus are engaged in similar and even parallel research, as the basic processes in biology are similar across the body.  But rather than jointly strive to understand and address a basic issue, like low-level, persistent inflammation, or errant protein folding that is common to may diseases, the historical context of the 'clinical' disease restricts and balkanizes understanding.  Of course, from an industry and funder perspective, this is a good thing, called 'focus.'  No one seems to mind that the object of that focus is a mirage.








                  

Thursday, March 10, 2011

Remystifying data



I know, oxymoron.  The esoteric and mysterious is preferable to the mundane for aesthetic reasons, but, for practical action, the mundane in the domain.


I wanted to revisit the placebo, as some folks commented on it and there is more useful information than that critique (if that was at all useful).  In the clinical trial, we look at outcomes from the treatment (or active) group and the sham (placebo) group.  It is pretty common, especially when looking at things like pain, that the placebo group has positive result (which I already described).  If the active group has a positive result that enough better than the placebo group, then one can say the results are ‘significantly better.’  Significance is a technical term here, commonly meaning that the number of times we find these results by chance are less that 5 time in 100 similar studies.  Significance is a tricky notion, in that the amount of difference needed for something to be significant depends, in part, on how effective the treatment is and how many people are in the study.  Really effective treatments need less people.  Less effective treatments need more.  So saying that there is a ‘statistically significant’ difference in a treatment over a placebo doesn’t give you lots of information.  A better way to think about it is to see how many patients have to be treated before the effect of the treatment is seen in one patient.  This is called ‘numbers needed to treat’ or NNT.  

Here is an example:  If a 40% of folks in the placebo group respond to a treatment and 50% of the active respond, then we know that 4 out of 10 responders in the active group   (where 5 out of 10) are likely to be responding to the ‘placebo,’ so, after we subtract those from the active, we have 1 in 10.  That means that 10 people have to be treated for 1 to show a response to the treatment.  So, a treatment may be ‘significantly’ better than the placebo, but the number of patients who actually benefit may not be all that high.  Here is some data on pain meds.  Pretty clear that the best ones only provide the level of benefit to 2 out of 3 patients.:

Table 1: The Oxford league table of analgesic efficacy (commonly used and newer analgesic doses)

The 2007 Oxford league table of analgesic efficacy
(at least 3 trials or 200 patients)
Numbers needed to treat are calculated for the proportion of patients with at least 50% pain relief over 4-6 hours compared with placebo in randomised, double-blind, single-dose studies in patients with moderate to severe pain. Drugs were oral, unless specified, and doses are milligrams. Shaded rows are intramuscular administration
Analgesic and dose (mg)Number of patients in comparisonPercent with at least 50% pain reliefNNTLower confidence intervalHigher confidence interval
Etoricoxib 180/240248771.51.31.7
Etoricoxib 120500701.61.51.8
Diclofenac 100545691.81.62.1
Celecoxib 400298522.11.82.5
Paracetamol 1000 + Codeine 60197572.21.72.9
Rofecoxib 50675542.32.02.6
Aspirin 1200279612.41.93.2
Ibuprofen 4005456552.52.42.7
Oxycodone IR 10 + Paracetamol 650315662.62.03.5
Diclofenac 25502532.62.23.3
Ketorolac 10790502.62.33.1
Naproxen 400/440197512.72.14.0
Piroxicam 20280632.72.13.8
Lumiracoxib 400370482.72.23.5
Naproxen 500/550784522.72.33.3
Diclofenac 501296572.72.43.1
Ibuprofen 2003248482.72.52.9
Pethidine 100 (intramuscular)364542.92.33.9
Tramadol 150561482.92.43.6
Morphine 10 (intramuscular)946502.92.63.6
Naproxen 200/220202453.42.45.8
Ketorolac 30 (intramuscular)359533.42.54.9
Paracetamol 500561613.52.213.3
Celecoxib 200805403.52.94.4
Ibuprofen 100495363.72.94.9
Paracetamol 10002759463.83.44.4
Paracetamol 600/650 + Codeine 601123424.23.45.3
Paracetamol 650 + Dextropropoxyphene (65 mg hydrochloride or 100 mg napsylate)963384.43.55.6
Aspirin 600/6505061384.44.04.9
Paracetamol 600/6501886384.63.95.5
Ibuprofen 50316324.73.38.0
Tramadol 100882304.83.86.1
Tramadol 75563325.33.98.2
Aspirin 650 + Codeine 60598255.34.17.4
Paracetamol 300 + Codeine 30379265.74.09.8
Tramadol 50770198.36.013.0
Codeine 6013051516.711.048.0
Placebo>10,00018N/AN/AN/A

What is really interesting is when you start adding in the negative effects of treatment and then comparing the ‘benefit’ to the harm.  The folks at www.thennt.com are kind enough to do that for us in some areas.  May be surprising that mammograms are NOT beneficial and PSA (prostate cancer screening) does more harm than good.  But that never stopped us from pursuing them.  Oh, and if your really want to have some fun, next time your health care provider wants you to start a drug, ask then what the NNT is, so you can make a proper determination.

Monday, March 7, 2011

Myth making, science style

One of the great things about myth making is that the ‘belief’ that supports the myth is rarely ever explicitly stated, certainly not in the myth, but usually not in the wider culture.  Anthropologist, at least of my ilk, are about investigating the assumptions, not to somehow ‘reveal’ the truth or the Marxist “consciousness raising,” but to understand how these assumptions drive cultural economics, in the sense of power and resistance, status and symbolic capital, and other stuff (more technical language).  

Because I traffic in ‘normal science,’ I am often involved in designing and running clinical research (sausage making with statistics).  One of the critical notions is that there is a ‘gold standard’ way to design trials, which usually means that there are two matched groups (cohorts), who either get a therapy or intervention that is supposed to work (active) or one that is indistinguishable for the first, but doesn’t work  (placebo).  And neither the patient NOR the clinician know which group they are in.  The so-called ‘double-blind (patient and clinician do not know which group they are in), randomized (meaning patients are assigned randomly), placebo-controlled (working and non-working intervention) trial (RCT, for short).

Because all of that is horribly boring, I’ll pick out my favorite part, the....Placebo.  

Lots of people have written, blogged and tweeted about the topic.  Part of why it is a great myth is that lots of folks know what a placebo is:
  1. (pharmacology) A dummy medicine containing no active ingredients; an inert treatment.
  2. (by extension) Anything of no real benefit which nevertheless makes people feel better

Huh?  How can a placebo have no REAL benefit, yet make people feel (do, act, etc.,) better?  Ain’t that a benefit?  And if so, where does it come from?  Hmmmm.

This is the way myths are constructed.  Let’s say we are a pharma company and we need to demonstrate the effectiveness of a drug.  So we give some folks the drug and some folks a pill that is identical, but no drug is in it.  We have already decided, before we even start, that the ‘drug’ is the meaningful part of this experiment, and the rest or it (just a ‘pill’) is meaningless.  

That is exactly the point.  We agree to pretend that the ‘pill’ itself either doesn’t have meaning or is less meaning full than the ‘drug.’  But why?  Find a pill. Look at it.  From my perspective the pill is a ‘sign,’ a physicality that has meaning.  Some of the meaning is highly specific, that if it is ricin, for example, it will cause very specific effects cellular that might  will kill me.  But other kinds of meaning are included within a ‘pill’ as well.  There is the whole notion, that goes back eons, of a ‘transformative substance,’ meaning ingesting something can ‘change’ you (think about religious acts, for instance).  There is transactional meaning, that the ‘substance’ comes from someone (an MD) of higher ‘status’ than the receiver, so it has some transactional value, and the larger social milieu of how ‘treatment,’ ‘medication,’ and other ‘signs’ are indications of our social status.  This semiotic web is too complex to dissect (but can be modeled, a tale for another day). But we give highest value to the ‘biochemical’ aspects of the drug, e.g., those above the ‘placebo’ (or should we say ‘meaning’) effect.  The common thinking (both scientific and lay) is that the ‘placebo’ effect is in the mind and the ‘drug’ effect is actually biochemical.  Pure crack smoking, from a scientific perspective.  Yes, I said scientific perspective.  What it the ‘mind’ but an emergent property of the biochemistry of the brain (I personally don’t localize the mind in the brain as the body is a continuous, complex and iterative process, but most folks do).  So if the ‘placebo’ effects the biochemistry of the brain in such as was as to produce a positive effect, what’s the difference between the placebo and the ‘intended’ effect?  They are both biochemical outcomes of pharmacotherapy.  Why is one ‘better’ than the other?  Wait a minute, one effect, the drug effect, is patentable, commercial and monetized.  The placebo is not.  

Myth making, profit-style.