(2 of 3)
Even at the bedside, there was still a chance for someone to realize that something had gone desperately wrong. Lehman's blood tests showed abnormal readings, and her electrocardiogram indicated enormous stresses on her heart. "She was vomiting sheets of tissue," her husband Robert Distel told the Globe. "[The doctors] said this was the worst they had ever seen." But they considered Lehman's violent reaction to be normal for such an aggressive treatment. All told, at least five hospital staff members failed to figure out that their therapy was killing her.
Dana-Farber has acknowledged full responsibility for Lehman's death, as well as the permanent heart damage of another woman who managed to survive an identical fourfold overdose. In addition to ordering an exhaustive investigation, the hospital's top physician has mandated additional precautions against such egregious mistakes. Staff members have been reassigned, and the pharmacy's computer program has been modified so that drugs cannot be dispensed over preset limits.
Unfortunately, as long as doctors are human, treatment blunders can never be eliminated entirely. "With 4 million patients a day visiting physicians, it's inevitable if inexcusable that mistakes will be made," says Dr. James Todd of the American Medical Association. Three large studies over the past 30 years have documented a distressingly consistent rate of medical mishaps in the U.S. By one measure, such negligence in American hospitals may result in 80,000 deaths each year.
That toll is a sign to some critics that improvement is needed in the systems that hospitals use to catch errors and review doctors' performance. "You would not walk on an airplane if you did not know that there are safety checks and backups and backups of the backups," says Dr. Sidney Wolfe, head of the Health Research Group. Hospitals need just as many fail-safe mechanisms, he says, "so that even if one or two fail, the third one catches the mistake."
In some fatal cases, it is hard to know if malpractice was involved. Even the most routine procedures can go awry, whether or not doctors are at fault. New York City was stunned last week by the news that a four-year-old girl had died following a tonsillectomy, an operation that is almost never deadly. Like 100,000 other Americans each year, Desiree Wade was sent home a few hours after the surgery, which was performed at St. Luke's hospital in Manhattan. She developed a fever and became increasingly sick. Her coughs apparently tore open the surgical wounds in her throat, and she bled to death. There is no evidence that the surgeon did anything wrong, but state health officials are investigating whether the child received the proper follow-up care.
If a tonsillectomy can go bad, imagine all the things that can happen with a bone-marrow transplant, a coronary-bypass operation or an experiment in gene therapy. As medicine has become more complex and the pace of technological change has accelerated, the opportunities for error have multiplied. And when doctors are constantly testing new treatments, as they are at Dana-Farber, they cannot fall back on years of experience.