THE SURGEON WHO performed poor Jesica Santillan's botched transplant has already taken the blame. "I am ultimately responsible," Dr. James Jaggers of Duke University Hospital said in a statement. Thus he's become a figure both noble and detestable, the captain of a sunk ship, a confirmation that Americans mythologize doctors while deeply suspecting them of the capacity for great arrogance and harm.

This front-page tragedy is an important reminder of the enormous number of fatal medical errors -- from 44,000 to 98,000 a year, according to the now famous 1999 report from the Institute of Medicine. But so far it seems the fallout from the case is likely to teach the wrong lessons about how to reduce the number of such errors. Yes, the surgeon handbook probably instructs Dr. Jaggers to double-check that the blood types of the donor and the organ match. But in this case, as in most similar cases, it's likely that the system failed to work in several places; personnel at the hospital and the donor bank crossed signals, and no one ultimately coordinated the information. Patient safety advocates call it the Swiss cheese theory: If cancer is misdiagnosed it's almost always because the holes didn't line up that particular day -- the person who reads the charts was in the ICU on call and the nurse misread the handwriting, and so on, adding up to a dozen places where the error could have been stopped but wasn't.

Some hospitals have tried to address some of the simplest and most common systemic errors that can lead to tragic results -- a misread dosage, a wrong limb removed. The Veterans Administration hospitals have taken for their model airline safety, setting up a system of checks and balances. They standardize the ways nurses mark sites for surgery. The same nurse who marked the site always transports the patient to the operating room. The nurse will wait for the surgeon before he or she drapes the site. At Duke officials have already announced they will now have three people verbally confirm the matching blood type, just as a pilot and copilot parrot back flight orders. Another simple problem is bad doctor handwriting, which can cause nurses to administer medicine in fatal doses. A few hospitals have computerized drug orders; some have substituted checklists for a scrawled instructions.

The reforms seem simple, yet the medical culture resists. Doctors see them as just another erosion of their autonomy, more red tape from the penny-pinching HMOs. In its 1999 study the Institute of Medicine study proposed a non-punitive method of reporting at least fatal errors, encouraging hospitals to think of them as an information-gathering exercise. But hospitals balked, fearing the information would be used against them in lawsuits. A committee formed by former president Bill Clinton proposed a compromise, publicizing general data but avoiding specific information, such as the results of internal hospital investigations, that could invite lawsuits. Medical groups rejected even that compromise. Perhaps this latest tragedy will force hospitals to rethink.