In advance of the renewal deadline for PDUFA (Prescription Drug User Fee Act) in 2007, critics and supporters came out of the woodwork to do public battle. Once PDUFA was renewed the hubbub died back down. PDUFA, enacted first in 1992, is a program in which companies submitting drugs for approval to the FDA pay a user’s fee. In exchange, the FDA promises faster evaluations and approvals.
Supporters claimed that PDUFA brought life-saving drugs to the public more quickly and efficiently than a wholly government-funded FDA. Critics argued, among other things, that PDUFA created a pressure at the agency to hurry past safety considerations in favor of rapid approval. As Marcia Angell put it, PDUFA created “a disproportionate emphasis on approving brand-name drugs in a hurry. Consequently, the part of the agency that reviews new drugs gets more than half its money from user fees, and it has grown rapidly. Meanwhile, the parts that monitor safety, ensure manufacturing standards, and check ads for accuracy have languished or even shrunk.”
In this context, a working paper by Daniel Carpenter and his colleagues that studied the possible effects of PDUFA on drug safety was frequently cited. The paper was published in the New England Journal of Medicine in 2008. The research question was: Has the accelerated drug approval process stipulated by PDUFA resulted in carelessness in the safety evaluation of new drugs? The answer—by no means simple to demonstrate—was a qualified ‘yes’. The authors conclude:
“Taken together, these findings suggest potential adverse effects of the deadlines governing FDA drug review. There are many ways to accelerate regulatory processes, most notably by providing more staff… Deadlines may offer a blunt tool with which to accelerate review… [O]ur findings suggest that deadlines may also cause drugs approved under these constraints to have a higher likelihood of unanticipated safety problems once they are in widespread use. A plausible hypothesis is that relying more on staffing and less on deadlines could result in the same degree of review efficiency without increasing the risk (and resulting greater cost) of unanticipated drug-safety problems.” [2008:1361]
The paper limited itself to studying whether PDUFA deadlines were associated with postmarketing safety problems. In keeping with the research question, it evaluated the scene before/after PDUFA to determine if PDUFA was responsible for any adverse changes. Although the paper analysed specific submeasures, its basic approach was to use PDUFA as the independent variable, safety issues as the dependent variable, and the flow of drug company applications into the FDA between 1950 and 2005 as the controlled or constant variable.
In this post, I wish to comment on the paper’s findings, with the aim of situating them within a broader view of contemporary pharmaceutical industry concerns and practices. Specifically, I wish to identify changes in pharmaceutical industry practices in regards to drug trials, reporting, FDA applications, and postmarketing activities as being the most appropriate independent variable for this research. My inferences complement rather than contradict those of Carpenter et al. However, by locating the most significant source of safety lapses outside the FDA regulatory process itself, my framework suggests different policy implications.
I must make short business of this—historical research is never as compact, neat or provable as logistic regression. My central claim is that the very date selected by Carpenter et al. as the starting point for evaluating the incidence of safety problems (1993, the year following the adoption of PDUFA) corresponds roughly to a watershed point in the pharmaceutical industry when R&D and marketing were effectively combined, and the latter’s objectives began to take precedence to those of all others in the drug development process, including those that might put a brake to recklessness in safety matters. This new situation, when marketing became a total institutional fact in the industry, is traceable to the early 1990s.
This was the start of what Michael Oldani calls “the blockbuster era”. Blockbusters are drugs with revenues of $1 billion or greater per year. Going only by the numbers, blockbusters accounted for 6% of the overall pharmaceutical market in 1991. This figure tripled to 18% by 1997, and in 2001 occupied fully 45% of the market. The marketing activities that yielded this concentrative explosion of profits were not independent of clinical trials (including postmarketing studies designed to find new potential indications for drugs), FDA negotiations and other drug application procedures, and can by no means be considered secondary, much less irrelevant to the analysis of post-1993 drug safety findings.
Since Merck Corporation is often cited in connection with PDUFA debates (because of the Vioxx scandal), let’s make a quick case study out of them.
Until about 1990, Merck was considered the most research-driven company in the industry. They had a 70% drug approval rate at the FDA, as against the industry average of about 50%. They were not strangers to aggressive sales, however, marching orders came from Merck Research Labs (MRL), and the divisions were separate, with sales and marketing where they belonged: in the field.
Circumstances changed for Merck and everyone else at around that time. In 1984 the Hatch-Waxman Act permitted generics to cut into patented drug profits. Competition, including from smaller start-ups, cut into the amount of time Merck’s drugs could enjoy cash cow status. Managed care organizations got wiser at restricting their formularies and bargaining over the price of drugs purchased in volume. And most importantly, competitors such as Pfizer were going completely ballistic on marketing. To keep profits high Merck had to do the same.
The intrusion and then market-share triumph of Pfizer’s me-too cholesterol drug, Lipitor, over Merck’s Zocor (Merck had pioneered the category with Mevacor), was a big signal to the company that marketing was king. Sales forces throughout the industry more than doubled during the 1990s. Also dating to the early 1990s, when PDUFA was getting underway, direct-to-consumer advertising was permitted and a direct channel of communication to consumers was opened, with all manners of consequence for the volume of drug use, safety reporting, and so on. The amount of reporting on health in the media also exploded during that time.
In 1994, Merck’s new CEO, Ray Gilmartin, scrapped the executive vice president of human health position and replaced it with three marketing presidents. Fond of brandable acronyms, Gilmartin introduced PACE, the Product and Cycle Time Excellence model for drug development. In this new structure, marketers were allotted dedicated budgets for Phase V, or postmarketing research. Marketers could now design and conduct their own trials, or “label change studies”, with more or less only advisory input from Merck Research Labs, the traditional R&D executives and firm leaders. Basic research is said to have fled the company. Like many of its competitors, Merck was reduced to being a commercializing agent for research conducted outside the firm. R&D was replaced by R$D.
What all this mans is that at around the same time PDUFA started, Merck changed its definition of who its chief customer was. The new customer was not the patient or even the doctor, but the FDA. Documentary evidence from Vioxx and the growing heap of other instances of safety violations in the industry dating to the 1990s demonstrate that the approval process—the label, if you will—was the marketing target of greatest significance, and the FDA an unwitting accomplice and victim. One can only wonder whether the FDA’s failure to warn the public about Vioxx during all that time when it knew of its dangers is traceable to the trust it had in Merck dating to earlier, less-marketed times.
Daniel Carpenter, Evan James Zucker, and Jerry Avorn. Drug-Review Deadlines and Safety Problems. (2008) New England Journal of Medicine, 358:1354-1361. Available at: http://content.nejm.org/cgi/content/full/358/13/1354.
- Warning Label Spin: Further Reflections on What the FDA is Up Against
- Pharmaceutical Marketing, Capitalism, and Medicine: A Primer (Part III/III)
- Cases for Overhauling Pharmaceutical Governance
- Four Good Questions About Pharmaceutical Industry Reform
- The Colonization of Pharmaceutical Science by Marketing
You may or may not also be interested in this article that appeared in the last year in the Journal of Health Econ, reaching a similar conclusion as Carpenter et al:
“The risk we bear: The effects of review speed and industry user fees on new drug safety”
Mary K. Olson
Journal of Health Economics
Volume 27, Issue 2, March 2008, Pages 175-200
Thank you! Olson’s paper is really sophisticated and thorough. There are many factors to account for, as both papers show. I’m sure the authors thought of more.
Macro-environmental effects of the sort I’m pointing out are difficult to measure and to demonstrate. The more recent time frames used in Olson’s analysis (1990–1992, 1993–1995, etc.) would mitigate some of my complaint. Another interesting paper she wrote
M.K. Olson, Are novel drugs more risky for patients than less novel drugs?, Journal of Health Economics 23 (6) (2004), pp. 1135–1158,
(her answer is yes) also has bearing, not necessarily positive, on my line of thinking.
Safety (ADRs) is an important measure, particularly in light of what the FDA can and cannot be expected to regulate. Carpenter et al. and Olson’s research should by themselves be the impetus for revision to PDUFA. Also needed is research into the claim made by pro-PDUFAers, namely, that speed to market saves lives. Given what most new drugs are about these days, one might guess that the answer is ‘no’. Olson or Carpenter et al. I’m sure would have an idea how to design such a study.
Still, the scale of these studies are not geared to address the questions raised by the kinds of observations available at the more general level: whether people are overmedicated in the first place (a long-term safety question), how much it all costs, whether innovation is also a casualty of the blockbuster phenomenon, the autonomy of the FDA and doctors, and so on. If the answer to all of these is bad, and it is traceable to the structure of industry and wholly privatized pharma, then the reform suggested is much deeper than relaxing approval deadlines under PDUFA.
So ultimately pharma is it’s own worst enemy, i.e. the marketing guys in the search for next quarter’s profit to satisfy the stock analysts, destroy the infrastructure of the company
Several great economic theorists tarnished their reputations for all time with the unfulfilled prediction of capitalism’s self-destruction. I would personally like to believe that our economy’s reliance on marketing-created value (as against scientific, in the case of pharma) has run its course and must inevitably, as you suggest, collapse on itself. But I wouldn’t bet my blog on it.
Pingback: Pharmaceutical marketing and the abstraction of consumer needs | Somatosphere
Pingback: The political economy and everyday life of pharmaceuticals | EDS Info (Ehlers-Danlos Syndrome)