Payables, interest rates, and Avastin

Okay, I hear you asking what in the hell do those three topics have to do with one another? Ah, welcome to the mysterious world of Medicare Part B drug reimbursement. As some of you have no doubt read, the FDA recently approved Lucentis, a VEGF-antibody fragment, to treat macular degeneration (AMD). Based on everything I've read, the drug is pretty darn amazing, but what is far more interesting is that doctors have been using Avastin for months to treat AMD. Since Lucentis is merely a fragment of Avastin, this seems to make a fair amount of sense. I would suspect that the only real difference between the two will be any properties relating to the Avastin antibody that is not included in the fragment that is Lucentis and any affects due to the different molecular weights. Thus, one can make an argument that the two drugs should be equivalent.

Now, in their brilliance, Genentech has priced Lucentis at $1,950 per monthly doses while Avastin costs only $55 per monthly dose--this massive difference in price is driven by the two disparate target markets--Avastin needs to be priced so that Genentech executives don't blush too much when they are asked about the monthly cost to treat someone with cancer. So far so good, right?

Now is where things get complicated. CMS covers both drugs and should reimburse both of them with a modest markup over ASP (actually, Lucentis might be reimbursed at a discount to AWP, but I haven't seen anything from CMS on this). The problem is that CMS will reimburse in, say, 30 days after they receive the claim. Thus your doctor has to carry the cost of buying Lucentis for 30 days... and not just once, but many, many times, since your doctor could be treating many patients with Lucentis. This isn't a problem form oncologists since their business model presupposes the existence of large receivables and, hopefully, large payables to fund the receivables. For opthalmologists, this can be a big deal if they don't have enough working capital to support this sudden outlay for Lucentis that might not be matched by inflows from CMS.

So, there is an incentive for doctors to use Avastin over Lucentis to treat AMD. This gets more complicated if CMS reimburses for Lucentis in such a way that doctors can make a greater profit on buying Lucentis and reselling it to CMS than they can with Avastin, net of the working capital concerns. Thus, this should be very, very interesting.

Conflict of interest week

It seems to have been conflict of interest in the health care community. First, Roy over at Health Care Renewal hit the nail on the head (repeatedly) on conflicts of interest at Stanford here and here. Then JAMA had to publish a correction to a major article because of inadequate conflict of interest disclosure that was only brought to light because of a letter to the editor, here, and the authors' reply, here, and a new conflict of interest policy, here. Then it is back to Roy, who linked to the San Jose Mercury News' response from Dean Pizzo of Stanford Medical School, here.

Personally, I think that the JAMA incident is the most revealing because of the defense raised by Lee Cohen. The charge, as summarized by Dr. Cohen is that:

Dr Urato suggests that the financial associations of the 9 physician authors constituted a conflict of interest and that the associations of all the authors should have been disclosed. His comments imply that the authors' financial associations could have influenced our interpretation of the data.

So far, so good. Clearly Dr. Cohen understands how the rest of the world interprets conflict. The rest of the paragraph, though, is the kicker. According to Dr. Cohen there was no conflict because his study was 1) federally funded, 2) prospective and observational, 3) they distributed audiotapes to convey the risks of discontinuing antidepressants, and 4) the data were analyzed by two doctors with no connection to the pharmaceutical industry. As a final rejoinder, Dr. Cohen states that becuase the authors did not evaluate any particular drug or treatment algorithm there could not be a conflict of interest.

So, lets think about these assertions for a bit. I don't think federal funding is a relevant reason not to disclose conflict of interest--indeed that Dr. Cohen could even consider federal funding as a reason not to disclose conflicts of interest indicates the extent to which conflicts have become endemic in the medical community since the issue isn't just that the sponsor might direct the research but that the investigator might be biased and federal funding does not address that concern. Being a "prospective, observational" study is a nice thing... except when that is a misleading description since the article states that women were enrolled after conception and could have ceased antidepressant treatment before conception. In the British studies establishing the value of folic acid in averting neural tube defects, the investigators recruited a cohort of women who wanted to get pregnant and tracked events; there is no reason for these investigators not to have followed that approach. Their third defense is also a bit of a throwaway because one could distribute identical, biased audiotapes, thus this is relevant to internal validity, but not to external validity. The fourth defense, that data analysis was conducted by two independent doctors does not address issues in the design of the study that could influence outcomes.

I leave the last defense to the end since it is so laughable. It only takes following the SSRI manufacturers through one storm of clinical data to realize the Glaxo's shareprice depends on Zoloft safety data since investors presume class effects for SSRIs.

Genetic testing, moral hazard, and adverse selection

I didn't set up this blog intending to write much about ethical issues in health policy, but there is one issue that I feel quite strongly about: genetic testing. CNN reported about 11 cousins each of whom had his or her stomach removed following evidence that he or she carried a gene for hereditary stomach cancer. I'm not going to get into the merits of prophylactic surgery at this time, primarly because I haven't seen the evidence for or against it. I want to follow-up on this statement:

Experts say that someday, doctors may do DNA tests as routinely as they check cholesterol levels now, spotting disease risks that can be lowered.

This is both the blessing and the curse of genetic testing.

Imagine a world with genetic testing for all disorders. What happens to people who test positive for a genetic disorder? It is not clear that everyone who tests positive for a genetic disorder will, necessarily, have the disorder in question, it simply changes the probabilities. Unfortunately, a rational insurer is going to charge people on the basis of their genetic risks, which will result in substantially higher insurance premiums for those people predisposed for cancer, diabetes, depression, or other expensive ailments. The upshot is that people who think they are predisposed for certain diseases are not going to get tested becuase of concerns over increased insurance premiums; even in the context of employer-based insurance many employers (remember, most people are employed by small businesses, not big businesses) are still subject to medical underwriting that will make insurance prohibitive for the business, or force the business to exclude coverage for genetically predisposed persons.

There is a way around this, which is to realize that society has always been exposed to these risks. What is new is the ability to identify people at higher risk. This suggests that in a universal insurance system, one can get around some of the hurdles to genetic testing because there is less damage from documenting one's genetic predisposition for a disease than under individual or employer based systems.

Anyway, enjoy discussing this idea.

House and product placement

Tonight's House features a product placement of BiDil!!! Have drug companies no shame?

Polticians politicize, why is that surprising?

There has been a lot of commentary about Republicans' politicizing scientific and technical debates, but I think that in the fury of such hot button issues as Plan B, global warming, and others we have lost sight of something important: it is in the nature of politicians to politicize things. One should also acknowledge that Democratics are as guilty of this as Republicans, but typically on less hot button issues. One of the issues that really gets my gander up is the opposition by many Democrats (and Republicans) to the Cape Wind project off of Cape Cod. I haven't seen a principled explanation of why the project shouldn't go forward--though there have been legitimate complains about the adequacy of the enviromental impact statement. All that said, my feeling is that Democrats (I'm looking at you Kennedys) and Republicans (Stevens, R-Big Oil) are opposing the project for political reasons, but using technical excuses to cover what they are doing--the definition of politicizing science.

Weird science

Following Tara's lead, I wanted to refer everyone to a truly insightful discovery on the nature of carbonated beverages and breath mints.

Health Wonk Review is up

Here is round 8

What's the difference...

Joe has a great post over at Managed Care Matters discussing a lawsuit between UnitedHealth Group and Jamaica hospital in Queens, New York. The crux of the matter is that United wants Jamaica to help getting the anestethiologists at Jamaica to sign a contract with United. My question is this: what is the difference between having a few, very large, profit-oriented payers and having a single payer responsive to elected officials? In my view, the latter situation is probably better than the former. While one can expect to hear complaints that a government payer would set prices (and, indeed, Medicare does set prices), what do private payers do? How does the negotiation between private payers and physicians work?

It really wasn't big pharma's fault after all...

Get ready for some fireworks in the U.K.

Mr. Tonmoy Sharma (for lack of additional clarity on honorifics, I will use Mr. throughout) has been accused of "deceiving the NHS and some of the world's largest drug firms" according to the Observer. A quick pubmed search suggests that Mr. Sharma may be involved in almost 300 different journal articles (this is an upper bound, I didn't verify "Sharma T" in each article), several of which are reports of clinical trials. This is significant since one of the accusations against Mr. Sharma is that he attempted to falsify data. Further, most of Sharma's research relates to schizophrenia, where there are numerous lawsuits alleging undisclosed side effects and the like. I suspect that we will be hearing a lot more about Mr. Sharma...

Zostavax... not so fast honey

CNN needs to learn that there is more to a journal article than the abstract... today, following approval of Merck's Zostavax for shingles they wrote:

Merck studies have shown that the vaccine reduced the occurrence of shingles by 51 percent. In patients who took the vaccine but were not successfully inoculated against shingles, Zostavax still managed to reduce pain and discomfort by 61 percent and reduced persistent nerve pain, the most common symptom, by 67 percent.

Now, this got me a little bit suspicious, after all, how would a vaccine reduce the burden of illness once you have gotten ill? Well, lets go to the videotape (or NEJM). From the abstract:

The use of the zoster vaccine reduced the burden of illness due to herpes zoster by 61.1 percent (P<0.001),> the incidence of postherpetic neuralgia by 66.5 percent (P<0.001), and reduced the incidence of herpes zoster by 51.3 percent (P<0.001).

So the abstract can be read to support what CNN wrote, but let's look a little further, after all, the abstract refers to this obscure burden of illness score that isn't clearly defined. From the methods section, however, we can get a defintion:

The primary end point was the burden of illness due to herpes zoster, a severity-by-duration measure of the total pain and discomfort associated with herpes zoster in the population of study subjects. For each confirmed case of herpes zoster, responses to the "worst pain" question in the Zoster Brief Pain Inventory were used to calculate a herpes-zoster severity-of-illness score, defined as the area under the curve (AUC) of herpes-zoster pain plotted against time during the 182-day period after the onset of rash. Subjects in whom herpes zoster developed had severity-of-illness scores ranging from 0 to 1813. Increasing mean scores are highly correlated with a decrease in the health-related quality of life and in functional status among older adults. A score of 0 was recorded for subjects in whom herpes zoster did not develop during the study period.

The "herpes-zoster burden-of-illness score" represented the average severity of illness among all subjects in the vaccine or placebo groups; it was calculated as the sum of the herpes-zoster severity-of-illness scores of all members of a group divided by the total number of subjects in the group. The secondary end point was the incidence of postherpetic neuralgia, defined as pain associated with herpes zoster that was rated as 3 or more on a scale ranging from 0 ("no pain") to 10 ("pain as bad as you can imagine"), persisting or appearing more than 90 days after the onset of rash. Scores lower than 3 were not associated with significant decrements in the quality of life or the ability to carry out activities of daily living and were therefore not considered to represent postherpetic neuralgia.

This definition sheds a lot of light on how Merck can claim the reduction in burden of disease--fewer people developed shingles in the vaccine arm! Thus, surprisingly, the vaccine worked and reduced the total burden of illness in that arm, but not the burden of illness among people in the vaccine arm who developed shingles.

Shame on CNN for not reading the articles, shame on Merck for writing an abstract they should have known would be misinterpreted, shame on NEJM for publishing that abstract.