Testing hemoglobin A1c (HbA1c) levels can determine glycemic control in people with diabetes. This is an essential test for guiding treatment and establishing treatment goals. In the early 1980s the first report appeared, indicating that the amount of glucose bound to hemoglobin was a good surrogate measure of the circulating glucose concentrations over the preceding 3 to 4 months. In 1993 the first well-controlled study was published showing that HbA1c was an excellent predictor of diabetes complications, and any reduction in HbA1c would reduce the likelihood of complications. In 1995 a performance measure for HbA1c was developed (patients with diabetes should receive at least one HbA1c measurement annually). A variety of tools were developed by the ADA and the National Diabetes Education Program to help promote the use of the HbA1c test and to use the HbA1c level as a treatment target. Performance improved (i.e., number of patients receiving at least one test annually) from around 60 percent in the mid-1990s to a current level of about 97 percent.

The final example provided was childhood vaccinations. Changes in law and public policy played a large role in vaccination rates when schools required all children to be vaccinated. State and county governments provided financial support for vaccines to be distributed to physicians or schools.

These four examples shared three common themes, Kahn said. First, they were all discrete and focused interventions. Second, the intervention and the desired outcome were closely linked. Third, guidelines and performance measures were developed by credible organizations. Nonetheless, how individual institutions actually implemented these interventions remains unclear because no literature has documented the exact steps or determined the most effective and efficient methods of implementation. Much like teaching a child to ride a bike, there is no exact science or literature that describes the best, most efficient learning process, but widespread success is eventually achieved.