On the evening of February 6, 2003, Carolina Donor Services called Duke University Hospital in Durham, North Carolina, offering a heart and a pair of lungs from a recent donor. Jesica Santillan, seventeen, had spent years with cardiomyopathy, a condition that had left her own heart and lungs failing, and doctors accepted the organs for her. Surgery began the next afternoon. By ten that night, with the new organs already implanted, Duke's own Clinical Transplant Immunology Laboratory reported what nobody had checked before the first incision: the donor's blood type did not match hers.

The call that offered the organs the night before covered the donor's height, weight, organ function, and cause of death. Blood type was not part of it. James Jaggers, the surgeon who accepted the organs for Jesica, later told Duke's own investigators that he had assumed the coordinating agency would not have called back with the offer if the types did not match. Nothing in the call required either side to say the two blood types out loud, and nobody did.

Thirteen days later, for the second transplant, nothing was left to an assumption. Compatible organs were located on February 19; a Duke transplant coordinator confirmed the blood types with Jaggers, then confirmed them again with the donor agency's own coordinator. Duke separately credits three physicians with ensuring organ compatibility before surgery began on February 20. The new heart and lungs matched her blood type. Less than a day after the operation, her brain function worsened significantly, and a scan showed bleeding and swelling. On February 22, after tests showed she met the criteria for brain death, she was pronounced dead at 1:25 in the afternoon.

On June 27, 2003, the United Network for Organ Sharing announced new rules for every transplant center in the country, the outcome of a review of the nation's organ transplant system that had begun after Jesica's case. The changes called for redundant checks throughout the process of matching and delivering an organ, so that blood-type compatibility no longer rested on one surgeon's assumption or one coordinator's unstated confidence. What Duke had already built for one patient in February became, that summer, a requirement for every transplant center in the network.

A blood type is a fact one laboratory can state and another can check. Only the second step, checking, is a safeguard; the first step, stating, is just an assertion that happens to be true most of the time. On the day Jesica died, Duke's chief executive, William Fulkerson, said the hospital had taken action to make sure nothing like it would happen again. The action that mattered was not a warning to be more careful. It was a rule that made someone else check.