Monday, August 11, 2008

Can CME be both Biased and Accurate? Of Course!

I enjoy reading Notes from RW, the blog of Robert W. Donnell, MD, a hospitalist in Arkansas who writes frequently for Medscape. While the two of us disagree robustly about the propriety of industry-funded CME, he argues his points rationally. In his current blog entry, he takes me to task for my recent comments slamming Medscape's CME offerings. His main point is that I haven't demonstrated that the industry-funded CME articles in Medscape are inaccurate. To quote Dr. Donnell:

Concerning the
Invega article, cited in his post on Medscape’s “CME corruption”, Carlat wrote "Has Janssen--I mean, Medscape--said anything inaccurate in this puff piece? No." That, we are to believe, is not important. In Carlat’s way of thinking the fact that the CME article is favorable to a product of the sponsor is damning enough without regard to the accuracy of the piece.

It's not that I believe accuracy is "not important." In fact, I believe accuracy is vitally important for CME. Unfortunately, however, accuracy does not protect CME articles from commercial bias.
The clearest argument is to compare drug advertisements with CME. The FDA regulates all industry advertisements for accuracy. Drug ads are required to make accurate claims about both the efficacy and the side effects of the sponsors' products. When ads are inaccurate, the drug companies can count on receiving nasty letters from the FDA such as this one, which scolded Pfizer last year for presenting inaccurate information about their antipsychotic Geodon.

Medscape's CME article on Invega is accurate, but is nonetheless promotional and commercially biased, because it focuses on the one clinical situation in which Invega offers an advantage over its competitors. It does not discuss advantages of other antipsychotics, nor does it discuss the many potential disadvantages of using Invega.

Likewise, Janssen produces advertisements for Invega which are accurate, but commercially biased. For example, here is a link to an advertisement for Invega .

It reads, in part:

"Powerful Efficacy for Many Patients Experiencing Symptoms of Schizophrenia


Patients who may benefit from a change in therapy

Powerful symptom control
In all 3 pivotal studies, INVEGA demonstrated significant improvements in PANSS total scores at all doses (3 mg, 6 mg, 9 mg, 12 mg ) versus placebo (P<0.001)."

There is nothing inaccurate here. Invega, like all FDA-approved antipsychotics, is efficacious for symptoms of schizophrenia. It does, indeed, demonstrate improvement in PANSS vs. placebo in all 3 pivotal studies. Nonetheless, this accurate medical information is commercially biased, because it is organized around the advantages of the sponsor's product and does not discuss the features of its competitors. For example, the ad does not mention that Invega performed poorly relative to Zyprexa in some of its FDA trials.

This advertisement could not possibly qualify as a CME article, because it breaks the two content rules of the ACCME's Standards for Commercial Support. Rule 5.1 states:
"The content or format of a CME activity or its related materials must promote improvements
or quality in healthcare and not a specific proprietary business interest of a commercial interest." By contrast, advertisements are crafted specifically to serve a company's business interest.

Rule 5.2 states:
"Presentations must give a balanced view of therapeutic options." Again, Janssen's drug ads are accurate but they do not give a balanced view of therapeutic options for schizophrenia. Nobody would expect such a balance--after all, these are ads.

However, the Janssen-funded CME article is supposed to give a balanced view, but it does not. Not only is the entire article organized around Invega's clinical advantages, but it ignores its two most conspicuous disadvantages: hyperprolactinemia and extrapyramidal symptoms (EPS). Instead, the article highlights weight gain and diabetes--side effects of great relevance to Invega's competitor Zyprexa.

(As an aside, Dr. Donnell defends another Medscape CME article which I criticized for an exceedingly gentle description of Zyprexa's side effects. That one, of course, was sponsored by Zyprexa's manufacturer, Eli Lilly. In Janssen's Invega article, by contrast, Zyprexa's side effects are described with vigor: The recent CATIE [study] highlighted the increased risk and incidence of metabolic consequences in patients with schizophrenia, which was associated with the use of a variety of atypical antipsychotic medications as well as specific modifiable risk factors.[10] In particular, the study noted the effects of specific antipsychotic medications on weight gain/obesity and metabolic endpoints. For example, the superior efficacy of olanzapine was tempered by its propensity to cause metabolic problems and greater weight gain than the other treatments, with an average increase of 2 pounds per month, as well as an increased risk for other metabolic effects vs the other agents.[17] In fact, the American Diabetes Association (ADA), American Psychiatric Association (APA), American Association of Clinical Endocrinologists, and the North American Association for the Study of Obesity together issued a consensus statement highlighting the relationship between specific atypical antipsychotics and the development of weight gain, diabetes, and hyperlipidemia.)

The bottom line is that accuracy is a minimum requirement of both CME articles and drug ads. In my experience, it is rare for industry-sponsored CME to be blatantly inaccurate. Instead, accurate statements are strung together strategically in order to encourage the use of the sponsor's product--and to discourage use of the competitor's.


Sara said...

"Invega, like all FDA-approved antipsychotics, is efficacious for symptoms of schizophrenia." This may be "accurate" but only in a limited context. Maybe it's efficacious over the short term because of its sedating effect or maybe it's efficacious when someone has been taken off some other antipsychotic and is in withdrawal so it soothes the symptoms of withdrawal (rather than those of schizophrenia). But in terms of actually helping to heal schizophrenia no antipsychotic is "efficacious." Long term studies have demonstrated that those patients who never go on antipsychotics or just have minimal treatment (like John Nash for instance!) have a much higher rate of recovery than those who are put on them as a maintenance treatment. Real longitudinal studies are needed to demonstrate this and they are rarely done but nevertheless they are out there and this is the conclusion. Maintenance treatment on antipsychotics is disabling and harmful to physical well-being -- worse than that, it greatly increases the risk of rebound psychosis if treatment is interrupted or even if it isn't. Is this ever mentioned in CME? I don't think so and it's an all important aspect that would certainly contribute to the "accuracy" of any presentation on Invega or any other antipsychotic. The short term clinical trials that are referred to in CME courses leave out huge pieces of the puzzle and are rarely conducted with parameters that take into account how patients are really treated in clinical practice.

Anonymous said...

I actually find Medscape information routinely inaccurate, and always spun in favor of its sponsors. I am a medical professional working in preventive health, obesity and nutrition. Medscape is not a credible source of balanced, critical or accurate medical information.