The fireworks that accompanied the publication of To Err is Human by the Institute of Medicine in late 1999 generated some magical thinking about how easy it would be to fix the problem of medical errors. A few computer systems here, some standard processes there (double checks, read backs), and maybe just a sprinkling of culture change—and poof, patients would be safer.

We now know how naive this point of view was. The problem of medical errors is remarkably complex, and the solutions will need to be as varied as the problems. Do we need better information technology? Yes. Improved teamwork? Yes. Stronger rules and regulations? Yes. Checklists, simulation, decision support, forcing functions? Yes, yes, yes, and yes.