[MUSIC] Our next section is on the Failure to Sustain a project or an improvement. One of the basic enemies is the second law of thermodynamics and the fact that entropy happens. What is very ordered and planned has a natural tendency to decay. This is a slightly more targeted figure to explain this, where we begin with order at the point when we have completed our project. There is a tendency over time to descend from order to chaos. The amount of organization that we may achieve at peak is counterbalanced by that tendency toward entropy. And this is what we are constantly fighting. Here's another example, you may recognize the style, if not the specific painting, of the noted impressionist artist George Seurat. And as you may know, he constructed his pieces of art using a myriad of tiny points of color. That's all very well and good. However, if entropy were really to occur, you might wind up with simply a collection of dots, with a collection of colored dots, which might not be as effective or pleasing. There are many examples of projects that are not sustained. In fact, I would say the large majority of improving projects are not sustained for any significant amount of time. One example might be the use of a safe surgery checklist in outpatient surgery. The safe surgery checklist, as promulgated by the World Health Organization, has been demonstrated to reduce infections and injuries in patients when it is used routinely. However, checklist fatigue can occur, and this can reverse any progress that is made. As you can see here, with infection rates on the vertical axis, and time on the horizontal axis, infection rates, which might be at an unacceptably high level, begin to go down, and then go down substantially after the intervention, the checklist, is employed. However, over time, with fatigue and decay of the intervention and the measures that surround it, infection rates gradually creep up again to their baseline level. Of course, it’s not just fatigue that contributes. There are a number of factors that contribute to failure over time. To name just a few of them, one is the failure to educate staff, and particularly new staff, on the technical and the adaptive aspects of the innovation. What needs to be done and how can it be made to fit in to specific contents where you are working. The second is the failure to require continuing checks of confidency are staff educated and capable of executing those steps. A third, a failure to incorporate changes into the daily workflow. If this is something which is done as a one-off, as a project, but is not incorporated into daily work, it is very unlikely that those changes will persist. And, finally, failing to or ceasing to monitor progress on a routine basis on the goals or outcomes that are being measured. When feedback ceases, again, performance tends to decay back to its baseline levels. The IHI cleverly described what they described as the seven deadly sins of spread, or the seven spreadly sins. The first is starting with large pilots instead of small local pilots. A second, making one person do it all. If that person then is absent, things don't work. The third is relying on vigilance and hard work rather than systematizing improvements. The fourth, spreading the pilot without the ability to adapt it to local needs, where it may not fit as well as in a highly controlled or specific context. The fifth, requiring the person or team who drove the pilot to be responsible for system-wide spread. The same people are not necessarily going to be equally competent at all of those tasks. In fact, it's very likely that the expert who helps design the pilot may not be the best person for day to day management and sustainment of a project in practice. The sixth, measuring data infrequently. Perhaps measuring a whole pile of data but very rarely, as opposed to small bits of data, just enough to tell you what's going on more frequently. And the final is expecting marked improvements in outcomes at an early stage. We should celebrate our initial and early wins and continue to move ahead.