group makes the discovered changes concrete and often provides a potent incentive to other teams. Pilot team members frequently become natural advocates and consultants, with high credibility at the peer-to-peer level.

Deployment and implementation methods, when used as part of an improvement strategy, differ from traditional clinical research in two ways. First, improvement often focuses more on clinical work flow and operational process than on patients’ clinical response to treatment. It aims to make the process do the right thing, the right way, the first time, every time (James, 1989) to achieve demonstrated excellent performance. Often, this means carefully designing care delivery systems so that health professionals find it easy to do it right (James, 2001). It involves building best care into standard work processes, with publication of new biomedical science as a secondary goal. Second, even though the pilot project may have identified key process factors that play important roles in implementation in other settings, most delivery settings, as noted above, include unique local factors. Therefore, successful deployment requires the ability to try change ideas locally and determine whether they do in fact produce better results in the particular setting. In other words, effective, standardized detection and reporting of injuries and near misses is a key part of deployment. Under a pilot and deploy strategy, ideas tried during the deployment phase have the advantage of having shown success in at least one previous setting. When such ideas are implemented with local testing in other settings, the rate of successful change accelerates.

The aim of improvement is to establish a new baseline, but achieving this aim often requires new work processes, support systems, and professional habits. These requirements feed back to the need for new case finding methods, evaluation procedures, and classification systems. An effective, standardized injury detection and reporting system therefore plays a fourth key role: once successful change has been implemented, it helps the care delivery team maintain the gains (Juran, 1989). Otherwise, processes and performance can drift back to their old baselines as attention shifts.

The application of the above ideas is illustrated through two case studies. One concerns ADEs and the other postoperative deep wound and organ space infections. Throughout these case studies, the key elements of the