In American medicine, no physician had as distinctive a practice as Dr. William Haddon, M.D. He did not treat patients, perform operations or write prescriptions. He had no ties to a hospital. Yet he saved lives in vast numbers. He kept millions of people healthy. At his death at 58 last week, Haddon had fulfilled in amazingly large measurements both the highest ideals of public health and his personal commitment to medical compassion.

Since 1969, Haddon had been the president of the Insurance Institute for Highway Safety, and since 1972 the president also of an affiliated group, the Highway Loss Data Institute. Both are Washington organizations. Before that, he served three years in the Johnson administration as the first director of the National Highway Safety Bureau, now the National Highway Traffic Safety Administration.

As a physician with degrees from Harvard Medical School and the Massachusetts Institute of Technology, Haddon had no peer in developing ways to prevent death and injuries from motor-vehicle crashes. He abhorred the word "accident." It is no accident that people get killed or maimed in motor vehicles, he would say, when those vehicles are designed with few if any genuine safety protections. Haddon understood that everything that happens to people and property in highway crashes is fully predictable. Crashes are not accidents, they are fulfillments.

Haddon was angered endlessly by the figures of highway death and injury. He calculated that with some 13,000 people injured by motor vehicles every day, about 45 million people were injured each decade. That was one-fifth of the population. He refused to accept that as normal.

"In the face of such suffering and tragedy," Haddon wrote, "it is often suggested that the problem is overstated because the incidence of death and injury per mile traveled is low and declining. There is, however, a much more important standard than such mileage-weighted statistics -- namely, the number of deaths and injuries expected during the lifetime of each vehicle. This is the most fundamental measure of the effort and success of the manufacturer in designing and building vehicles to protect human life during their use."

Haddon concluded that the manufacturers' effort was small and the amount of success smaller: "During their lifetimes every 1,000 new vehicles will be involved in somewhere between 2,000 and 3,000 crashes; they will kill more than three people and they will injure more than 300."

Haddon's years of service in government and with the institute were the period in which manufacturers resisted nearly every safety innovation created, from passive restraints to crash-worthy bumpers. "It is not clear who fabricated that brilliantly negative slogan, 'safety doesn't sell,' wrote. "But whoever did was wrong. The truth is that 'safety isn't sold' by U.S. and many foreign manufacturers. Buyers have even been strongly discouraged from buying safer cars."

Along with Ralph Nader, Joan Claybrook, Clarence Ditlow and others who see motor vehicles as death machines, Haddon was perceived as a nuisance by the buccaneers of the auto industry. In many ways, Haddon was harder to combat. He could not be dismissed as "self-appointed," because he was in fact confirmed by the Senate as the government's first highway safety official, and he was then appointed by the insurance industry as its chief safety strategist.

Nor could he be accused of screaming and stomping. Haddon was a sophisticated reasoner whose arguments for passive restraints such as air bags have yet to be refuted by the auto makers who still preach speed and style over safety. Haddon said that "long experience in the prevention of diseases shows that, other things being equal, the less people must do to be protected, the more successful is the preventive measure."

He cited such common safety measures as pasteurizing milk and purifying water. Done at the source, these are more effective ways of preventing disease than asking each person each time he drinks to boil the milk or water. The examples, he argued, are everywhere: "We use insulation on lamp cords rather than trying to get everyone to put on gloves each time they must handle the cords."

Why not build similar protections into vehicles? Haddon estimated that for drivers and passengers to get the protection of seat belts requires 100 billion individual harnessings and unharnessings annually: "The alternative would be one-time decisions by a handful of public and private executives to install equivalent 'passive restraints' and ancillary passive protection."

Ten years ago, the Government Accounting Office said that federal safety standards begun under Haddon had saved 28,000 lives from 1966 to 1974. In the past decade, a similar saving has doubtless occurred. It was as though Haddon, as a physician, had been dispensing a vaccine to his patients. In truth, he was: the technological vaccine of safety as applied to motor vehicles, which, after war and hunger, are the worst pestilence of the 20th century.