David Graham's field is the rarefied world of epidemiology, the study of diseases in populations. In his spare time he reads the New Testament in Biblical Greek. But this American scientist and self-confessed introvert has single-handedly shaken to its foundations the national body responsible for ensuring public health and safety in the US, along with several of the world's largest drugs companies.

David Graham's field is the rarefied world of epidemiology, the study of diseases in populations. In his spare time he reads the New Testament in Biblical Greek. But this American scientist and self-confessed introvert has single-handedly shaken to its foundations the national body responsible for ensuring public health and safety in the US, along with several of the world's largest drugs companies.

Dr Graham's devastating critique centres on the way the Food and Drug Administration (FDA) failed to act decisively on evidence that Vioxx, a blockbuster drug for arthritis which was taken by 20 million Americans, increased patients' risk of heart disease. According to Dr Graham, Vioxx caused heart attacks or death in up to 140,000 patients during the five years it was on the market.

When Vioxx's maker, the US pharmaceuticals giant Merck, eventually made its own decision to pull the drug last September, its shares plunged and patients launched an avalanche of class action lawsuits which will start to come to court in May.

"At least Merck withdrew the drug. I'm happy about that because it was a public service. The FDA would not have withdrawn it from the market, I'm convinced," Dr Graham said.

The extra twist in this story, which has caught the attention of Congress and catapulted Dr Graham on to the national stage, is that the FDA happens to be his employer of 20 years' standing.

Amid allegations that the agency has tried to oust Dr Graham from the department of drug safety that he works in, this unlikely troublemaker, who draws on his Catholic faith for inspiration and guidance, has hired a Washington lawyer who specialises in representing government whistleblowers.

He is also speaking out in public against what he describes as a culture of fear within the FDA. "The FDA's suppression and intimidation of scientists is a threat to public health. Unless it changes, and scientists can speak without fear, they cannot defend the public," he said.

The matter is a pressing one because the US is the world's largest and most lucrative drugs market. Due to a campaign spearheaded in large part by Dr Graham, serious doubts are now being raised on Capitol Hill about whether the FDA as it is currently configured is fit to act as the watchdog of the pharmaceuticals industry.

According to Dr Graham and other critics, the agency is not up to the task because it is too close to the companies it is supposed to be policing. "The FDA has become an agent of industry. I have been to many, many internal meetings and, as soon as a company says it is not going to do something, the FDA backs down. The way it talks about industry is 'our colleagues in industry'," he said.

That is not because FDA employees are "people with bad motivations", Dr Graham said. It is rather because the body is entirely geared towards concentrating on approving drugs, doing little once they are on the market.

"There has been a strong impetus in the FDA to say 'yes'. Clinical trials are not designed to address safety and people who handle the approval of a drug also handle post-marketing. But they are not able to make decisions because they have been too closely involved with the approval", he said.

There is also a problem because, under the Prescription Drug User Fee Act in 1992, the drug industry funds the applications process, and as a consequence puts pressure on the regulator to speed up approvals.

"There is a sense that you get what you pay for. The industry is paying the piper and calling the tune," Dr Graham said.

That was certainly the pattern with the FDA and Merck over Vioxx, a member of the so-called "Cox-2 inhibitor" class of drugs developed to improve on old painkillers such as aspirin by not causing ulcers or internal bleeding.

Barely months after being approved in 1999, studies started to show patients taking a high dose of Vioxx were five times more likely to have heart attacks or strokes than normal, with those on a low dose incurring twice the ordinary risk.

According to Dr Graham, the FDA did almost nothing about the mounting evidence until April 2002, when it ordered Merck to start putting warning labels on Vioxx bottles.

"It should not have taken two years for the FDA to do what it did, it should not have taken two months. For the FDA now to turn to the American people, Congress and the world community and say it did the responsible thing is inaccurate because it knows that labelling does not change behaviour. It used labelling as a subterfuge for keeping the drug on the market," Dr Graham said.

The FDA's seeming lack of ability to stand up to Merck is worrying given that Dr Graham believes there are five other major drugs displaying serious side effects which require further study, including Crestor, made by Britain's AstraZeneca.

The FDA, based in Rockville, a characterless town on the outskirts of Washington DC, will have a chance to prove it has some backbone next week, when it holds a three-day summit starting on Wednesday on whether all Cox-2s should be withdrawn, after concerns have been raised about other members of its class such as Pfizer's Celebrex.

Whatever the outcome of the meeting, Dr Graham believes there needs to be fundamental reform of the FDA, and he is leaning towards an idea put forward by several senators to split up the institution, so that an entirely independent agency to monitor drug safety could be established. "At the moment the office of drug safety within the FDA is outnumbered and out-financed by the office of new drugs. If there were a separate agency it could have a separate world view, and could put safety in the forefront," he said.

Reliance on clinical tests could also be replaced with the type of epidemiological approach that Dr Graham specialises in where large random samples of the population are tested.

Many argue this would be better than the current situation where problematic patients such as those with a high risk of getting certain diseases are screened out at the clinical stage to make it easier to establish the effect of the drug on its target area. The downside of this approach can be to warp expectations of what will happen when the drug is given to the population at large.

Dr Graham acknowledged it would not be possible to insulate patients against all risks. "All drugs carry risks but there is a moral and ethical need to have a good handle on the upper limit of the risk," he said.

An agency focused entirely on safety could also act as a brake on the drugs industry, which generates revenues by pushing a steady stream of new drugs on to the market irrespective of whether they double up with tried and tested treatments already available.

Wearing an ash mark on his forehead to denote the start of Lent when he gave this interview, Dr Graham admitted: "I don't have all the answers, but I haven't heard these issues being discussed within the FDA".

And in the meantime, this doctor who trained as a neurologist but switched to epidemiology so he could deal with patients in the millions has to be content with treading the lonely path of the awkward employee not prepared to go along with the status quo. "I didn't start out wanting to be a whistleblower. I am not considered part of the family any more. This whole thing has been quite an experience," he said.

The appliance of science

Age: 51.

Education: Johns Hopkins University School of Medicine. Trained in Internal Medicine at Yale and in adult Neurology at the University of Pennsylvania. Three-year fellowship in pharmacoepidemiology and a Masters in Public Health at Johns Hopkins, with a concentration in epidemiology and biostatistics.

Career: Twenty years at the Food and Drug Administration. Currently Associate Director for Science and Medicine in FDA's Office of Drug Safety.

Hobbies: Listening to classical music and jazz, reading, hiking.

Family: Married, with six children.