Whenever David J. Graham, M.D., M.P.H., HS ’81, wonders whether he made the right career move from Yale-New Haven Hospital resident to resident critic of the Food and Drug Administration (FDA), he recalls with remorse a patient who died under his care.

At the time, 1979-1981, he was doing a residency in internal medicine at Yale after graduating from the Johns Hopkins School of Medicine and planning to become an expert on viral infections of the central nervous system. But the patient died of an unexpected adverse drug reaction, and even though Graham was not at fault, “I was the person who had prescribed the drug,” Graham says. The incident “contributed to my interest in studying drug safety.”

He also discovered during his time at Yale that he “didn’t enjoy the day-to-day grind of patient care,” so he settled on epidemiology. Since then, he has had a 20-year career as an epidemiologist at the FDA, forgoing more lucrative offers in the private sector. He is now associate director for science and medicine in the agency’s Office of Drug Safety. Despite (and because of) that title, he has also become the FDA’s most vocal and listened-to critic.

For years, Graham, 51, has been the FDA’s equivalent of an ingrown toenail, causing his employer pain as he went public with allegations that the agency was allowing prescription drugs it approved to stay on the market despite evidence from its own researchers that the medications were harming and killing people by the thousands.

In Graham’s view, the drug safety problems began in 1992 with the passage of a law aimed at getting lifesaving drugs onto the market faster. To speed up approvals, the law forced pharmaceutical companies to foot most of the bill for the review process. That left the FDA “captured by industry,” says Graham. “He who pays the piper calls the tune.” In the meantime, the same officials who approved the drugs were being asked to monitor their safety after they’d hit the market. If there’s evidence of harm, “now they’ve got to do an about-face,” he says. “At its heart, that is an inherent structural conflict of interest.”

Graham said a former boss told him that the drug industry was the FDA’s client, but Graham, an assistant scoutmaster and devout Catholic with six children, didn’t buy it. If he sees that a drug may be harmful, he’ll investigate. If the evidence warrants it, he’ll challenge the FDA’s regulatory position internally, even if it earns him the enmity of his superiors, which it often has. The FDA labeled one of his studies “junk science,” and forbade him to publish it in a major medical journal. (Six weeks later the FDA changed its mind and the article was published, last February, in The Lancet.)

Graham has called for the withdrawal of a dozen prescription drugs, and almost all of them have since been removed from the market—often after a fight with his superiors—in some cases by the drugmakers themselves. One case made Graham a star witness before the United States Senate’s finance committee in 2004, when Merck & Co. suddenly removed Vioxx, its popular painkiller, after the company’s studies showed a higher risk of heart attacks and strokes among users of the drug. Graham had been warning about Vioxx for years, but the FDA refused to pull the drug on his recommendations. He testified that the agency also urged him to change the conclusions of another damning study about Vioxx just before Merck’s surprising announcement. The voluntary withdrawal placed the agency in an unwelcome spotlight over its alleged failures to protect Americans against unsafe drugs. Meanwhile, Graham sought whistleblower status, and telephone calls disparaging him to the Government Accountability Project, a nonprofit that protects maverick insiders, were traced back to the FDA. (FDA officials say they allow employees to speak their minds, and couldn’t explain the incident.)

“The FDA has let the American people down and, sadly, betrayed a public trust,” Graham told the Senate committee.

“I think I’ve had a substantial impact, as much as FDA officialdom wants to bad-mouth me,” he says. “I point to the evidence and I say, ‘Look, I’m almost always right.’ I don’t recklessly recommend the withdrawal of drugs.”

He says he hopes to finish his career at the agency, even if more difficulty lies ahead. He has two suggestions for improving drug safety. First, the government should create a drug safety center parallel to the FDA’s drug approval center. The new center would oversee postmarketing regulation and would be able to call on the FDA commissioner to pull a drug it deems unsafe. Second, this center should be provided with enough money to do its job. (A bipartisan Senate bill co-sponsored by Democrat Christopher J. Dodd of Connecticut and Republican Charles E. Grassley of Iowa would do just that.)

“I feel fairly certain that I’ve probably saved more lives taking the career path I have taken than I would have with another career path,” Graham says. “That’s something that colleagues both within government and in academia remind me of periodically when I tell them of my tales of woe at the FDA—they remind me that I’m doing good.”