[MUSIC] Now I'd like to talk about harm. Harm is psychological or physical injury or damage. And there is a number of different subtypes of harm that I'll address. The first one I want to talk about is inevitable harm. And you can have inevitable harm associated with a procedure, a therapy, or an action. And the only way to prevent this kind of harm is not to do, or not to administer that therapy, or not to conduct that procedure. Some examples of inevitable harm are an allergy that a patient develops from a appropriate therapy where the patient has had no previous history of allergy. That's inevitable. Also, the physical harm associated with surgery is another type of inevitable harm. Certainly when we do surgery, it's with the intent of doing more good than harm. But when you cut into somebody, you're causing some form of inevitable harm. The next type I want to talk about, or the next principle I want to talk about, is that an inevitable harm today may not be inevitable tomorrow. So, as an example, maybe we can develop a practical, affordable test that we can administer to a patient before we administer the drug to prevent an adverse event from happening. We actually do have a really good example of that type of test. There's a drug called abacavir that's used to treat HIV. It turns out that in about 3% of the patients, they'll develope a potentially fatal allergic reaction to it abacavir. But fortunately, there's a genetic test that can be administered to identify those patients susceptible to this type of reaction, and then avoid using this drug in that patient subset. Here's another really good example, and that is a central line associated blood stream infection, or CLABSI for short. It turned out that in the past, people pretty much said, well, patients are going to get, occasionally they're going to get a CLASBI, and there's not much we can do about it. Well, Dr. Pronovost and his colleagues, in the reference that I've got on your slide, actually published a study that shows that with the appropriate work flow adjustments, with the appropriate bundles and behaviors and system design, we can eliminate CLABSI. So what was thought to be inevitable in the past is now preventable. And actually you're going to hear more about CLABSI in the lesson on high-reliability organizations. The next kind of harm I want to talk about is preventable harm, harm that's caused by an error. And if that harm is very severe, it would be called a sentinel event. And I'll talk more about sentinel events in a later lesson.