People are more receptive to change and more willing to focus on what is best for the organization as a whole rather their individual unit or department's agenda, says Merryman.
Just because strategic priorities are identified, however, doesn't mean something won't come along and trump them. For UPMC, that happened on December 3, 2008 when an 89-year-old patient died of exposure on the roof of UPMC Montefiore Hospital. The health system immediately began implementing a new protocol for managing wandering patients called "Condition L." By February 28, UPMC had rolled out the new protocol to every hospital, educated all employees, and performed a drill at every hospital.
"No one here—beginning with myself—defends the failure," says Merryman. "We took that learning with a degree of seriousness." Implementing a new process across 20 hospitals with 20,000 employees in 60 days may not have been pleasant, says Merryman, "but it is possible." To accomplish the goal, some other programs were put on hold, including one of the organization's top strategic priorities—the "Ask Once" program. "That is an important issue," Merryman explains, "but it's not critical to a patient's outcome."
Decide what to test
When choosing what to test, organizations should focus on initiatives that are reasonable to accomplish in a certain time period that will also have the largest impact, Nielsen says. "You have to weigh each one. You look at it with your nurses and doctors—your multidisciplinary team—to really evaluate it."
For example, Cedars-Sinai recently evaluated changing its triage system to a two-tier process in which a nurse performs a quick assessment, gathers key information, and rapidly prioritizes patients. "We implemented it with several key triage nurses who tried it over a couple of weeks," says Nielsen. At the end of the testing period, Cedars-Sinai determined it was worth exploring because the process helped identify cardiac and neurology patients faster. By rapidly testing the improvement initiative you are minimizing the risk, says Nielsen.
Organizations may know within one or two weeks if the initiative is going to work, adds Merryman. Process changes that reach the patient may require longer to test, but behind the scenes support work can be tested very quickly. "We make testing way too hard," she says. "Healthcare in its process design loves to test to the exception and, in all honesty, we'll never have a process that meets all of the exceptions, which is why all of the processes are complicated, broken, and people go around them. It is better to build a process that works to the rule because, if you do that, you'll give people back time to manage the exception."
Time to Cut Your Losses
Here are two scenarios when it's better to pull the plug on a process improvement initiative rather than trying to make it work: