Dr. Krumholz’s fine post hoc Vioxx study data analysis, published in an Archives of Internal Medicine article this afternoon, put me in mind of this highly-impertinent question:
What did the FDA rules, as the same were in effect in 2000 through 2004, require Merck to track, tabulate and report, in the way of post-market approval adverse event data?
This question is interesting to me — in view of Dr. Krumholz’s work — primarily because I am rather concerned that (1) either Merck did not collect and forward to FDA all the FDA-required adverse event data, or (2) it did collect it, but did not model it appropriately, to sniff-out emerging adverse event trends — trends like those that the article points out were in “hiding, in plain sight.”
So — from the then-applicable, relevant FDA rules:
. . . .Serious, unexpected ADEs must be submitted to the agency within 15 working days. [This requirement becomes 15 calendar days as of April 6, 1998.]
The 15-working day time frame begins on the date the manufacturer or any of its affiliates (including foreign) received the information. Sometimes an ADE report initially not classified as a 15-day report is reclassified as “serious and unexpected” after the firm’s follow-up investigation. In this case the 15-day time frame begins on the date the firm received the information that caused the report to be reclassified. If the applicant receives new follow-up information on a 15-day report, then the G4 date is the date of this new information. A follow-up 15-day report has to be submitted within 15 working days after receipt of the information. This requirement becomes 15 calendar days as of April 6, 1998.
An establishment’s report may include a causality assessment of whether the ADE was related to the drug. The regulations do not permit submission delays for 15-day reports pending completion of a causality assessment.
Spontaneous (not derived from a study nor literature), domestic ADEs that do not meet the criteria of a 15-day report are required to be submitted to the agency in periodic reports under 21 CFR 314.80 only. Periodic reports must be submitted quarterly for the first three years after application approval and annually thereafter. The periodic report submission date is usually based on the date the NDA/ANDA was approved. If a date other than the approval date is used to calculate the due date, determine if the firm received the required written FDA approval. . . .
So all of this should have been collected every 15 calendar days, and aggregated, and forwarded, along with statistical analysis, every calendar quarter, from 2000 thorugh late 2003. Did that happen? What sorts of statistical analysis packages were run on the quarterly, and cumulative data aggregations? Who at Whitehouse Station reviewed them, before forwarding the same to FDA (if they were in fact prepared, and forwarded, at all)?
Of the two, who should we most expect had the resources to ferret-out the emerging cardiovascular event pattern1? Based on Dr. Krumholz’s analysis — even though hindsight is 20-20 — by 2002, a tripling of cardiac events was appearing, with a “p value” below 0.05, or less than a 5 in 100 chance that the events could be assigned to random chance.
Certainly, as to all plaintiffs who’ve opted out of the global Vioxx settlement (and are still pursuing individual claims), this will be available as evidence.
Most impertinently though, I wonder tonight — from the snowy mountain-tops of Colorado — whether a showing might now be made that some of the basic assumptions upon which the $4.5 billion settlement itself rests — were in error. Will it reopen the settlement? That would seem rather unlikely — but as ever, we will need to wait and see — when the other shoe falls.
1. On that score, at page 25 of this March 2006 GAO report — on ways to improve timely FDA recognition of adverse event trends in previously approved drugs — we run across this startling observation:
. . . .In addition, it can be difficult to attribute adverse events to particular drugs when there is a relatively high incidence rate in the population for the medical condition. For example, ODS staff analyzed adverse event reports of serious cardiovascular events among users of the anti-inflammatory drug Vioxx in a 2001 consult. However, because Vioxx was used to treat arthritis, which occurs more frequently among older adults, and because of the relatively high rate of cardiovascular events among the elderly, ODS staff concluded that the postmarket data available at that time were not sufficient to establish that Vioxx was causally related to serious cardiovascular adverse events. With AERS data it is also difficult to attribute adverse events to the use of particular drugs because the AERS reports may be confounded by other factors, such as other drug exposures. . . .