Saturday, December 24, 2011

Back to the Risks of Prescription Drugs

The latest issue of the Medical Letter (Dec. 12/26, 2011; subscription required) is instructive in reminding us of an important theme behind the ethics of the medicine-Pharma interface. The biweekly newsletter for physicians, supported only by subscriptions and with no industry advertising, reports on two recently approved drugs (dabigatran and dronedarone) that turned out to have greater risks of serious bleeding than had initially been suggested by clinical trials. The newsletter then reviewed another new drug, rivaroxaban, a clotting inhibitor, which in early trials was shown to be somewhat effective (when added to other drugs) in decreasing the risk of future cardiovascular events, but only at the cost of increasing the risk of bleeding into the brain. The Medical Letter experts commented that this puts rivaroxaban in the same category with dabigatran and dronedarone, making one suspect that as more experience is gained, more will come to light about the actual risks of serious bleeding.

These reviews of new drugs reminded me of the fact I have previously blogged about (, courtesy Dr. Donald Light’s important book, The Risks of Prescription Drugs. Dr. Light has reviewed data that indicate that each year, more than 100,000 Americans die of the adverse reactions from prescription drugs, properly prescribed in hospitals. That is, deaths from overdoses and due to errors in prescribing are not included in this total. If accurate, that would make adverse reactions from prescription drugs roughly the 4th largest cause of death in America. And those data exclude all deaths that might occur outside of hospitals.

I have to admit that when I first read that figure, I did not believe it, and I imagined that critics would come forward with different calculations to show that Dr. Light’s results were too high. So I consider it important that no such critics have, to my knowledge, come forward; and Dr. Light informs me that to the extent that he can double-check these calculations, they still appear to be valid, and have been buttressed by more recent studies. If anything, alternative methods of estimating the numbers (there are no direct data available that would give a precise figure) would lead to even higher totals. He also reminded me that the AARP featured these figures in their September magazine, warning about risks of prescription drugs: prescription drugs#pg14

Bear with me for some comments. We already put the American public at high risk of bad things happening to them when we prescribe drugs. Yet the Medical Letter shows that we persist in rolling out new drugs that add even more to the risks.

Pharmapologists will immediately protest that these risks are nothing compared to the wonderful benefits of all these new wonder drugs. So let’s think about those benefits for a minute. Let’s consider the new kid on the block, rivaroxaban, as one example. This drug (brand name Xarelto, manufactured by Janssen/Johnson & Johnson) is an anti-platelet agent (like low-dose aspirin) and is indicated for patients at high risk for blood clots that could cause strokes, heart attacks, and other bad stuff. Patients at that level of risk are typically now treated with two drugs, aspirin and clopidogrel (Plavix). So the research studies on the new drug compared patients taking those other two agents, plus placebo, with those taking the other two agents plus rivaroxaban.

When that comparison was made, the combined bad outcomes (death, heart attack, or stroke) happened in 9.1 % of those taking the three-drug combo (at the lower tested dose of rivaroxaban) compared to 10.7% of those on the two older drugs only. That calculates as best as I can tell to a number-needed-to-treat of 62. That is, you’d have to give 62 people the 3-drug combo for 2 years to prevent one case of death, heart attack, or stroke, in excess of what would happen with just two drugs only. At that dose of rivaroxaban, for every one person who avoided a bad event, another one would suffer a serious (but in this study, non-fatal) brain-bleed-type stroke.

Keep two things in mind. First, the advocates of rivaroxaban are proposing that we take what’s now a 2-drug combo and turn it into a 3-drug combo for treatment of the indicated conditions. We know that the more different drugs you’re taking, the greater the chance of adverse reactions. Second, these are the early data on the efficacy and safety of rivaroxaban. Recall John Ioannidis’s important work showing that we can bet dollars to donuts that the earliest studies of a new drug will show a substantially higher success rate, and a lower adverse-event rate, than the totality of all studies performed over a number of years (see previous post,

To summarize, the US pharmaceutical industry, its pipeline running dry of really good drugs that are truly a major advance over existing drugs, is now busy flooding the market with drugs that offer very slight advantages and only at the cost of serious risk of harm. The total public health impact on the American public is increasingly negative, as Dr. Light’s figures indicate.

I am reminded of the old Harry and Louise ads, funded by the US private health insurance lobby to kill the Clinton health reform plan back in 1993-4. The couple sitting at their kitchen table, after listing all the terrible things that would happen if health reform was enacted, shook their heads and intoned, “There has to be a better way.” I think the same might be said of the way we discover, approve, and market new drugs today.

Friday, December 23, 2011

Painful to Report: ProPublica Skewers American Pain Foundation

The most recent investigative reporting by ProPublica, who have done great work in exposing industry funding and related conflicts of interest, puts me personally in a bind because I have considerable sympathy with both sides:

Charles Ornstein and Tracy Weber detail the degree to which the American Pain Foundation has been funded by manufacturers of narcotic painkillers, including the notorious Purdue Pharma, maker of OxyContin, and how it then proceeds to stick up for widespread use of narcotics for treatment of chronic pain--even in the face of increasing evidence showing, first, that opiate drugs may not be very effective for chronic pain; and second, that overdoses and addiction from these drugs are growing to epidemic proportions.

So let me see if I can add any balance to the discussion.

First, let me speak in defense of the position taken by the American Pain Foundation. One of the big problems in pain research has been the lack of Federal grant dollars, since there is no National Institute of Pain at NIH, and the NIH has until very recently shown very little interest in funding pain research, despite estimates that around 75-100 million Americans suffer from some degree of chronic pain. In this funding vacuum, the analgesic manufacturers have often been the only possible source. As I detailed in HOOKED, this has led to the undesirable state of affairs, of many of the most prominent experts in pain management in the US being "on the take" with drug company money.

The critics of the Foundation that ProPublica spoke with all note the current dearth of solid evidence that opiates are good for treating chronic, non-cancer pain. It is true that the available studies to date have been discouraging. I think today, any physician who goes to narcotics and narcotics alone as the first line approach to treating chronic pain would be off base. But two important facts need to be put into perspective alongside these data from controlled trials. First, after you try all the various other modalities for pain management, virtually all physicians in this business would agree that you end up with at least some--more than just a few--patients who are still miserable and unable to function despite it all. These patients often get some relief, occasionally substantial relief, with opiates, and with proper management and monitoring the rate of addition or serious side effects is very low. Somebody needs to speak up on behalf of this group of patients and make sure that they don't lose access to the one thing that substantially helps them.

Second, the ideal way to treat chronic pain is through a multidisciplinary pain clinic that offers physical therapy, pain psychology, and generally a team approach alongside a willingness to prescribe appropriate analgesic drugs. Now, go try to find such a place. Not unusually, there's none within 250 miles. (Too many so-called "pain clinics" are basically procedure mills where anesthesiologists do mostly worthless injections for big bucks, then turn patients away as soon as their insurance benefits run out.) If you can find a truly multidiscioplinarty pain clinic, then ask if it accepts Medicaid. Chronic pain, not surprisingly, is concentrated in lower income groups, who often lack the good insurance that's the ticket into most of the really good clinics. So we are talking about an ideal mode of treating chronic pain that is simply out of the reach of way too many sufferers.

Are these clinical realities about treating pain reflected in today's media? All we read about is the epidemic of drug diversion, overdoses, illegitimate and illegal "pill mills" selling narcotics to addicts, and so on. Somebody needs to speak up for the legitimate pain patients who may need opiates as a part of a broader approach to pain management.

OK, that's the one side of the issue. Now let me speak with my usual hat on, opposed to conflicts of interest. Here we see the chickens that come home to roost when you feather your nest with industry cash. (Sorry for the fowl cliches.) There's that little question of public trust. It seems that whatever the American Pain Foundation set out to do, it has seriously compromised its ability to do it by getting in bed with the likes of Purdue Pharma.

Let me talk a minute about Purdue, whose record of wrongdoing is now legendary. They make OxyContin. Addicts found out about 5 minutes after the drug hit the market that the capsule had an unfortunate property--if you crush it, you got an immediate rush of the drug, rather than the slow release that the capsule was designed to give, and that it does give if you take it properly, without crushing. This set up the popularity of OxyContin as the infamous "hillbilly heroin." By contrast, another slow-release form of opiate, MS Contin, works just fine--even if you crush the capsule, the matrix that holds the active drug won't release it all at once. That's great for legitimate pain patients and bad for abusers.

So what would a decent drug company have done? I say--obviously, take OxyContin off the market, and don't bring it back on until you have redesigned the capsule to work in a way that doesn't help addicts get high. Instead Purdue did everything they could to keep selling OxyContin as originally formulated and also did everything they could to whitewash the problems. And the American Pain Foundation helped them out.

I strongly agree with my bioethics colleague, Dr. Ben Rich, who has written extensively about medicine's ethical obligation to do a far better job of treating pain. Sadly, well-intentioned organizations that take cash from Pharma and create serious conflicts of interest for themselves make this ethical job harder to do.

Why the Drug Pipeline is Dry: Another View

In previous posts, such as:, I have offered my opinions as to what's wrong with the research and development model favored by the drug companies over the past couple of decades, that has led to the notable dearth of important new drugs. As I myself know squat about drug research, I always offered these comments with trepidation. The post referenced above was reassuring in adding the opinion of two reporters; but now we have word from a medical expert who at least knows enough about the subject to get his stuff published in the New England Journal--specifically, Dr. Robert Schwartz of Tufts in Boston:

Dr. Schwartz is focused on how best to find promising new cancer drugs, but I'll let him speak for himself:

“Targeted treatment” became the shibboleth of the pharmaceutical industry, spurring on a multibillion-dollar search for targets in other cancers. The quest began with an enterprise based primarily on industrial-strength methods of gene sequencing, gene arrays, and gene profiling, which allow rapid examination of thousands of genes in a cancer cell, potentially revealing the genetic profile (or signature) of a malignant cell. These profiles could, in principle, enable the design of a specific inhibitor of an aberrant cancer gene or its product.

But the idea of finding magic bullets by open-ended genetic screening that deliberately avoided any prior hypothesis was considered dubious by some. And ultimately, critics would complain that more than a decade of investigation with powerful technical methods and huge investments by the National Institutes of Health (NIH) and the pharmaceutical industry had yielded few useful drugs....The molecular labyrinth of the cancer cell guarantees that the odds of identifying a single useful and specific therapeutic target by mass screening are very low.

...[W]e [should] reconsider the way we organize cancer research. In Science-Mart: Privatizing American Science, Philip Mirowski concludes that today's system of big-time, industrialized scientific research is deeply flawed because of burdensome intellectual property rights and investors' management of research. Moreover, international pharmaceutical companies are cutting back their drug-development research owing to economic constraints, and the federal government probably can't make up the difference in these times of deficits and cutbacks.

To summarize, treating potential drug molecules like an assembly line making Model T Fords is maybe a model for drug R&D that is as out of date as the Model T. Maybe you actually have to have a clue as to what makes disease happen--the way academics have traditionally studied disease in the laboratory--before you can find drugs that actually show promise.

Tuesday, December 13, 2011

Here We Go Again, with the Form

According to several press accounts such as:
--we can go back to our standard form and fill in the blanks (except to substitute devices for the usual 'drugs'):

Company: Medtronic
Devices: defibrillators and pacemakers
Settled Federal charges of: paying kickbacks to docs for using their brand of device ($1-2K per implantation)
Fine paid: $23.5M
Fine as % of Annual Sales during Peak Year: 0.4%
Company Admits Guilt?: Of course not

I calculated the percentage of sales based on last year's sales figures reported by the company of $16B, and their statement that 31% of their business involves cardiac rhythm devices.

As we have seen recently (e.g., if anything, the trend within the drug industry has been for more substantial fines, so the pittance the Feds collected in this case seems a major step backward.

Friday, December 9, 2011

"Intellectual Conflict of Interest" Rides Again

Dr. Roy Poses posted on his Health Care Renewal blog about Pharma critic Dr. Sidney Wolfe being denied a place on an FDA advisory committee--
--which prompted a response by a "Melissa Raven" and a reply from Dr. Poses (see same URL).

Dr. Poses was kind enough to reference two former posts on this site about "intellectual conflict of interest," of which the more pertinent and complete is:

I was about to add my proverbial 2 cents in the form of a comment on Health Care Renewal, but decided it would become too wordy, and I save wordy for this blog and the few long-suffering readers thereof, so here goes.

Mr. Raven sees value in excluding people with "intellectual conflicts of interest" from advisory panels and similar groups, and instead looking for methods experts who understand what counts as good and bad evidence, but who have not previously taken stands on the particular issue under review. Dr. Poses quotes objectors to the FDA action in the Wolfe case who say that eliminating "intellectual conflict of interest" amounts to excluding anyone who has previously thought carefully about an issue, which seems to be overkill on a scientific matter.

It seems to me that underlying this discussion is a possible confusion among three concepts-- disinterested; conflict of interest; and disclosure of conflict of interest.

Ms. Raven invokes the goal of the "disinterested" advisory panelist as her argument against seating somebody like Dr. Wolfe (whose specific sin, by the way, was having previously issued warnings, on behalf of the consumer organization Public Citizen, against the type of birth control pill that the panel was supposed to review). Now, there may be something to be said for this goal, and even taking it to the extreme of excluding academic types who have formed opinions on a matter previously. After all, a reason to exclude potential jurors is that they have preconceived ideas about the guilt or innocence of the defendant. But my point right now is that "disinterested" and "conflict of interest" are rather different concepts.

As I discussed in HOOKED, and in a subsequent publication referenced below, 'conflict of interest' is hard to define succinctly, but a good approximation is, "One has become party to social arrangements that would tempt a person of normal psychological makeup to neglect a duty to protect the interests of a person or group in favor of a secondary interest, in such a manner as to create a risk of loss of trust in one's social role." Illustrative example:

  • Medical scientist undertakes clinical study of a drug.

  • The medical scientist also becames a paid speaker and consultant to the company.

  • The company wants to publish data that would promote the use of the drug and suppress data hinting at problems with the drug.

  • A medical scientist has a (primary) duty to publish the scientific truth for the benefit of future patients.

  • The scientist now has a (secondary) interest in serving the goals of the company, which conflict with the goals of good science and patient benefit.

  • The secondary interest is of the sort that would lead a person of normal human psychology to be tempted to ignore the primary duty in favor of the secondary interest.

  • This social arrangement (accepting speaker fees etc.) is of the sort that could lead reasonable onlookers to lose trust in medical science.

Notice that this describes a pretty specific set of circumstances. On the other hand. "disinterested" is a much broader term, which does not necessarily implicate any conflict of interest or threaten loss of trust in a social role. Because I know somebody is interested and not disinterested, I may distrust a particular opinion that person offers, but that is not the same as losing trust in a social role.

The sorts of financial conflicts of interest that us pharmascolds have focused on are relatively easily avoided. One does not have to fill one's own pockets with Pharma cash to practice good medicine or do good science. On the other hand, in an academic venue or in medical practice, being truly disinterested may be well-nigh impossible.

Suppose that as a medical practitioner I tell my male patients that they should have PSA tests routinely. Studies now come out showing that PSA testing may be more harmful than helpful. I am hardly disinterested when I read those studies. If their conclusions are true I have been exposing my patients to potential harm out of proportion to their benefits. Of course I am "interested" in how I interpret those studies, how hard I try to find flaws in their design, and so forth. So find me a "disinterested" medical practitioner, let alone scientist.

I have previously stated, ad nauseam in fact, that disclosure of conflicts of interest is a necessary but often hardly sufficient response to the ethical problem posed by COI. But it is in the nature of financial conflicts of interest not to be generally known unless disclosed. On the other hand, the fact that an academic has taken a certain point of view on a scientific topic is usually self-disclosing. (Just listen to the guy for about 5 minutes.) Anyone wondering whether I have an opinion on most matters can easily do a PubMed search and see what I have published on the topic. Dr. Wolfe certainly does not keep his Public Citizen advocacy activities secret.

So I suggest we keep these distinctions in mind as we discuss such matters as whether the FDA was silly to exclude Dr. Wolfe, which I vote that they were.

Brody H. Clarifying conflicts of interest. American Journal of Bioethics 11(1):23-28, January 2011.

Thursday, December 8, 2011

Supremes Look at Biotech Patents

The Minneapolis Star-Tribune reports:
--that the U.S. Supreme Court heard arguments yesterday in a patent dispute between Mayo Clinic and a firm called Prometheus Laboratories. The case has implications for potentially reining in the overly broad patents now being issued for biological and medical discoveries, especially related to the genome.

The article is unfortunately not very specific about the diagnostic tests that are at the root of the patent dispute. We learn from another press source--
--that the test involves correcting the dose of thiopurine administered for various autoimmune disorders, especially in the gastrointestinal tract. Mayo previously used the Prometheus test to find out how much thiopurine stayed in each patient's bloodstream, until its own scientists came up with what they say is a better test, and then Prometheus sued Mayo to prevent their test from reaching the market. The lower court ruled for Mayo, a higher court reversed, and now the case is before the Supremes.

Mayo's case, which is backed by the Feds, is that the Prometheus patent is too broad, and would patent the idea of doing the sort of test Mayo has actually developed, even though Prometheus never invented that specific test. As I wrote in HOOKED, the patent system, which is an interference with the free market designed to trade off monopoly privileges for promoting innovation, is actually today squelching innovation because the patent office is willing to patent most anything, whether it has been actually taken to a realistic stage of development or not. The Supreme Court was told in argument that if they found for Prometheus, they would basically be saying that a manufacturer of a bad product could use the patent system to prevent a competitor from marketing a better product--hardly the sort of promotion of innovation the patent system as designed for.

I previously blogged about patents:
--in relation to pay-for-delay to prevent generic drugs from cutting into brand-name profits. Anything the Supremes might do to reduce overly broad patents, and to stop industry from "patenting the sun" in Jonas Salk's famous phrase, would be welcome.

Thursday, December 1, 2011

Time Out for Some Drug Pricing Issues

I try to stay focused in this blog on the ethics of the medicine-Pharma relationship and so as a rule steer clear of drug pricing issues as such. Of course drug prices indirectly affect the ethics of medicine as physicians try to figure out how an increasing number of their patients can gain access to needed medications.

Linda A. Johnson of AP did a recent story--
--about Pfizer's struggle to keep making profits off its best-selling drug, Lipitor, which is about to go off patent. (Brand-name Lipitor sales at about $11B annually are now fully 20% of Pfizer's business.) Unfortunately her article obscures as much as it reveals. (Readers who understand the economics of drugs better than I do please feel free to chime in with comments.)

Quick recap of the generic weirdness according to the Hatch-Waxman act: As a method of incentivizing generic competition, Congress granted special privileges to the first generic manufacturer to come up with an FDA-approved bioequivalent product for the brand-name drug. The generic company that's first to the finish line gets a 6-month period of exclusivity before other generics can jump into the market.

That means that a brand-name drug goes from being a monopoly to being a duopoly for 6 months, then the real free market takes over. During the duopoly period, it is common for the generic competitor to be priced only a bit lower than the brand-name drug price. That means that the generic firm can make quite hefty profits off its drug for the 6 months, and the brand-name company can often extend its period of profits since it's not really being undercut all that much by the competition. Then 6 months later the price really falls as everyone and his proverbial duck can get into the game, and it's only then that consumers really benefit.

The company that's the lucky winner in the generic lottery, according to a press release just out, is Ranbaxy Pharmaceuticals, Inc. which is a wholly owned US subsidiary of the largest generic drug firm in India--itself a most interesting development, but that's for another post. Neither the press release nor the AP article says what Ranbaxy will charge for its generic atorvastatin (aka Lipitor) during its lucky 6 month window.

Now back to Johnson's report--she states that Pfizer is working hard to maintain brand loyalty among its customers, unlike the usual behavior of the brand name company which is to throw in the towel when the generic competition appears and profits drop. They are working as hard as they can to price brand-name Lipitor at lower than the generic price, at least during the 6-month window. That includes issuing copay cards for pharmacies allowing $4/month Lipitor prescriptions, and trying to lock pharmacies into special discounts on the condition that they don't substitute a different generic brand.

What's not clear to me at least is what the overall game plan is once the 6 month window is over. Does Pfizer figure it can lock in business during the time when the only generic competitor is relatively expensive, and then later reap the benefits when the generic competition becomes dirt-cheap? Or is it actually the case that Pfizer figures it can make decent money even selling its Lipitor at generic prices for the long haul?

What's also not clear is that it's hardly possible that this plan was Pfizer's Plan A. As we discussed in the past--
--the usual Plan A of the brand-name firm is "pay-for-delay," cutting a sweetheart deal with the first generic competitor to lie low and not challenge the brand-name drug for an extra 6 months, assuring that the big profits keep rolling in for that half-year. It's not clear whether Pfizer tried various ways to buy off Ranbaxy and came up empty-handed, or whether the current sales strategy is in fact the result of a secret deal that Pfizer managed to cut. Hopefully we'll hear more on this soon.

Addendum 12/1/11: Apparently three US senators had some concerns similar to mine:
Also in this news coverage by Duff Wilson at the NY Times, it's shown that I might have been in error above in saying that the first kid on the generic block was Ranbaxy, contrary to what their press release said. Watson Pharmaceuticals has FDA approval and is actually shipping generic atorvastatin as we speak. But Watson has a deal with Pfizer to make the "approved" generic. One press account says Ranbaxy has approval for its generic, another (Wilson-NYT) indicates that it's been held up, so I really don't know what the score is at this point. Stay tuned for further exciting adventures.

Addendum 12/8/11: Two more counties heard from... First, Merrill Goozner both on his Gooznews blog and in the Fiscal Times:
--adds a bit more detail, and clarifies that both the Watson and the Ranbaxy generics are approved and out there, so it's both-and not either-or. Then a group in the New England Journal of Medicine, including our old friends Drs. Joseph Ross and Harlan Krumholz:
--provide a bit more detail, in the context of trying to calculate how much savings the US public could look forward to in coming years as a result of atorvastatin being available generically--they figure about $4.5B annually by 2014, assuming that full savings are realized. They then discuss how Pfizer could mess up the picture. Their worst-case scenario is that the sweetheart deals Pfizer has cut might discourage some generic makers from getting into the competition and that would be the main way prices would be kept higher than otherwise. So bottom line, still more questions than answers as to how all this will play out.