What’s Wrong With Prof. Light’s Argument That New Drugs Are More Dangerous Than Old Ones?

August 11, 2008

Pharmagossip recently highlighted a new study by Prof. Donald Light of UMDNJ which states that “new drugs are twice as likely to harm patients as to provide them with benefits superior to existing drugs.”

That’s a bold claim. Is it true?

I’ve relied on Light as a source in the past and found his arguments to be powerful. However, he also tends to overstate his case. Here’s an old story about his criticisms of the Gates Foundation effort to cure malaria. He’s yet to be proven right or wrong, but almost no one has taken up his position that the effort is misguided.

So I decided to look through Light’s PowerPoint presentation on his data. (Does absolutely everything have to be in PPT? Please?) There appears to be some statistical wiggle room, to say the least.

Look at slide 2, where he notes that the FDA recorded 32,000 cardiac events related to Vioxx. (Side thought: see here for a brief critique of why relying on FDA adverse event reports isn’t the wisest thing).

Then, on slide 3, we find that according to Light’s calculations cardiac events have jumped to 1.9 million. That’s a big slippage. Either the FDA missed 98% of all cardiac events — despite most people thinking that the FDA tends to overstate adverse events — or this number is wrong.

Put another way, if you took Vioxx, the odds of experiencing a cardiac event were either 5% or 0.09%, depending on whether you go with the Light or the FDA, respectively.

Knowing Prof. Light, I’m about to receive a lengthy email that will require several writethrus of this post.

Let me just say this: Most people think Michael Moore is the biggest threat to Big Pharma. They’re wrong. It’s probably Light. Unlike his bearded colleague, Light actually musters the stats in favor of his points. One of these days he’s going to do some damage with them …

Advertisements

5 Responses to “What’s Wrong With Prof. Light’s Argument That New Drugs Are More Dangerous Than Old Ones?”

  1. Merrill Says:

    The larger point that could be taken from Light’s study is the need to conduct larger pre-approval clinical trials. That was addressed in two studies posted online last week in Health Affairs.

  2. Scott Says:

    There is a presumption in medicine (and biotechnology drugs in particular) that the manufacturers are innocent until proven guilty meaning that until someone with the cash to collect such data (which in many cases is not available in the public domain), even though the manufacturers’ track records are littered with cases of abuse, misrepresentation of data and selective disclosure of the facts. We need more people like Professor Light actually working for the FDA!

  3. jimedwardsnrx Says:

    Merrill,
    I think Light’s most powerful argument is that approvals are given for tests in which the drug is compared to a placebo and not head-to-head vs. existing drugs.
    Jim

  4. Beth Says:

    The AEs reported to the FDA are typically understated, as reporting is voluntary for physicians. Thus, the actual incidence would be greater as Light has suggested. (It is not that the FDA ‘missed’ these AEs.)

  5. Robyn Says:

    “The extent of under-reporting is unknown, and depends on the severity of the adverse event, among other factors. One group of researchers estimated that the FDA receives reports of less than 1% of serious adverse events, whereas another group gave this estimate as between 8% and 13%.”

    Ahmad SR.
    Adverse drug event monitoring at the Food and Drug Administration.
    J Gen Intern Med. 2003 Jan;18(1):57-60.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: