01Why study this case — when an effective drug became a disaster through disclosure
The Vioxx (generic name rofecoxib) case was less about a defect in the drug itself than about how a company handled and disclosed the data on its risk. As a painkiller and anti-inflammatory, it genuinely worked. But because the data showing cardiovascular risk were not adequately or promptly shared, it is held to have contributed, in the end, to tens of thousands of heart attacks and deaths worldwide.
This is the symbolic case of "the ethics of data disclosure." What a pharmaceutical company holds is not only the drug. It also holds the data showing that drug's risk. How honestly and how promptly it brings that data into the open — this piece traces how an effective drug became a disaster, and the structure behind it.
02What Vioxx was — a painkiller "gentle on the stomach"
Vioxx was a COX-2 selective inhibitor developed by the pharmaceutical company Merck. Approved in the United States in 1999, it spread explosively as a treatment for arthritis and pain. Used by tens of millions worldwide, it was a true blockbuster.
Conventional painkillers (NSAIDs) carried the side effects of gastric ulcers and gastrointestinal bleeding. COX-2 inhibitors, by selectively suppressing the COX-2 involved in inflammation while sparing the COX-1 that protects the stomach, were promoted as "the same efficacy, gentle on the stomach." In the U.S., direct-to-consumer (DTC) advertising was also deployed, making it a highly recognized, popular drug.
03The VIGOR trial — the first sign of cardiovascular risk (2000)
In 2000, Merck published the results of a large clinical trial called VIGOR. Its aim was to show that Vioxx caused less gastrointestinal bleeding than a conventional drug (naproxen). That point was indeed shown. But within the same data lay a sign that could not be overlooked — cardiovascular events such as myocardial infarction were more frequent in the Vioxx group.
Merck explained this difference as follows: "naproxen has a heart-protective effect, so Vioxx only looked worse by comparison." In other words, not "Vioxx is dangerous" but "the comparator was too good." Whether this interpretation was right became, later, a major controversy. The sign of risk was there, yet it was absorbed by an alternative explanation.
04The APPROVe trial, and the 2004 voluntary withdrawal
The decisive evidence came from another clinical trial, APPROVe (2004). It was designed to study prevention of colon-polyp recurrence, but the result showed that using Vioxx for more than 18 months clearly raised the risk of heart attack and stroke. This time it was a risk of Vioxx itself, not something the "comparator" could explain.
On September 30, 2004, Merck voluntarily withdrew Vioxx from the world market. An FDA researcher testified to Congress that Vioxx may have caused, in the United States alone, tens of thousands of excess heart attacks and sudden deaths. A drug that genuinely worked had brought enormous harm through a delay in risk disclosure.
05What was questioned was "disclosure" — whose data is it?
What this case questioned most severely was neither the drug's efficacy nor the mere existence of a side effect. It was how honestly and how promptly the company disclosed the data showing the risk. The cardiovascular signal was there in 2000. It took four years to reach product withdrawal.
Lawsuits were filed around the world, and in 2007 Merck agreed to a settlement paying about $4.85 billion (for numerous U.S. claims). In the course of litigation, internal documents and the conduct of publication (the handling of unfavorable data, involvement in authorship) were scrutinized. "A drug's data is both a corporate asset and a public good directly tied to patient safety" — that tension lies at the core of the case.
06Four lessons that remain today
Lesson 1 — Disclose risk data promptly, even when it is unfavorable
The essence of the Vioxx case was not a defect in the drug but a delay in disclosure. The more unfavorable the data to one's own product, the more disclosure is delayed and interpretation drifts toward the gentle. How one handles unfavorable data is where a company's honesty shows. This connects directly to today's mandatory clinical-trial registration, full publication of results, and the prohibition of selective publication.
Lesson 2 — A convenient interpretation neutralizes the warning
VIGOR's cardiovascular signal was explained away as "the comparator was protective." Even when a warning is contained in the data, if it is gently absorbed at the interpretation stage, the warning does not function. When interested parties interpret their own data, unconscious bias creeps in. That is why independent verification is needed.
Lesson 3 — Even an effective drug becomes a disaster if risk disclosure is neglected
Vioxx genuinely worked as a painkiller. Drug disasters do not arise only from drugs that do not work. Even with an effective drug, if the risk is not honestly conveyed, patients make the wrong choice. Efficacy and safety together are what let a patient judge properly. The role of disclosure is to close the asymmetry of information.
Lesson 4 — The more widely a drug sells, the more a small risk becomes vast harm
Vioxx was used by tens of millions worldwide. Even if the per-person risk looks small, when the denominator is enormous, the excess harm reaches tens of thousands. Safety surveillance of a widely used drug weighs even more heavily than for a rare one. Not "sells well = safe," but "sells well = the harm can grow large."
07Connections to other chapters
The Vioxx case connects to other chapters of this site as follows.
- Drug Disasters 08 (Iressa) — a shared theme: balancing the expectation of efficacy against the handling of safety information
- Drug Disasters 01 (Sulfanilamide) — a modern version of the lesson that "not unlawful" and "ethical" are different
- Advertising Regulation — directly tied to the discipline of direct-to-consumer (DTC) advertising and efficacy-skewed information provision
- Ethics, Part 1 — the tension between profit and medicine: how far to treat data, an asset, as a public good
Vioxx was an effective painkiller. Gentle on the stomach, used by tens of millions worldwide. The problem lay not in the drug's efficacy but in how the data showing cardiovascular risk were handled, and when they were brought into the open. The sign was there in 2000. The withdrawal came in 2004. In between, tens of thousands of excess harms are said to have accumulated.
This case is called the symbol of "the ethics of data disclosure" because it showed clearly that drug disasters arise not only from drugs that do not work but from drugs that do, and that the scale of harm can be decided by a single matter — how a company handles the data it holds. How honestly and promptly it discloses data unfavorable to itself — there a company's honesty shows.
Anyone working in modern pharma learns the weight of those ordinary steps — registering trials, publishing results in full, hiding no unfavorable data — from the price of Vioxx's tens of thousands. Data is a corporate asset and, at the same time, a public good tied directly to patients' lives.