Now we're going to tackle quality. Around the year 2,000 a book was published by what was then called the Institute of Medicine, that organization is now called the National Academy of Medicine, they're the body of medical experts that the federal government looks to for expertise and advice. When they put out a report, generally, it's a consensus of experts about a problem that the US government, or US society needs to deal with. This book about Crossing the Quality Chasm pointing out how US care and health care was of lower quality than they are, was a real call to the community to improve quality, to pay attention to it, measure it, fix it. More recently, a report came out of the same organization called vital directions for health and health care, notice that the word IT informatics do not appear, but what is interesting here, is in the quality Chasm Book, the role of health IT was highlighted as part of the solution for these issues, so they identified four action priorities, pay for value, engage people, activate communities, and connect care, besides the four action priorities, there are four essential priorities, and so you see measure what matters most, we already referred to that, modernize skills, and some of those skills are IT related, accelerate real-world evidence, we saw that in that Learning Health System, it calls you discussed previously, and advancing signs again, collecting evidence from practice to improve care. The real quality problem that we have in the United States, is demonstrated on this slide that plots per capita expenditure on health care, in other words, how much a government or society spends per person, against life expectancy on the y-axis. Now, if you spend a lot of money, you should expect a high life expectancy, but in fact you would love it if you're spending less money, and still getting a high life expectancy. The upper left-hand corner is terrific. Low-cost, high life expectancy, the right-hand side, high cost, low life expectancy is bad. Look where the US is, the US is the closest to that lower right-hand side, it spends like twice as much as anybody else, and yet they're not getting the same life expectancy. Japan, which is at the top and the middle, they're spending half as much as US does, and they get 10-years of more life expectancy. We can talk about the variability in Japanese society versus the US, and we could talk about many other things, but still, it's pretty amazing that we're not getting the bang for the buck that we would expect. The hopes that we have with the interaction between health IT informatics and quality is that, by better management of parentheses through infomatics, we'll have better outcomes that we'll be able to identify the high risk individual, which we saw talking about population health, and again, by being able to predict quantitatively through health IT informatics, maybe we'll be able to focus on our care of the people who need it the most, and ensuring patient safety, again through health IT, if a patient has a heart attack, which is generally due to a clot an important artery in the heart, then very rapidly, that patient should get medication that's going to dissolve the clot, and get good blood flow going back to the heart, and limiting the damage that you have from that heart attack. A rule that says that a patient with acute myocardial infarction or heart attack who comes into the hospital should be treated with fibrinolysis, which is clot-busting within 30 minutes. Sounds like a great rule, let's see what's involved. Well, first of all, when do you know that patient has had a myocardial infarction? Is it the ambulance drivers who tell you that they have it? Is it the triage nurse when the ambulance is coming? Is it the first doctor who examines a patient or maybe it's the senior doctor after reviewing laboratories have come in after about an hour or so? When you know, they have acute MI and whose judgment counts, and then where is that judgment recorded and how's the record it and when is it recorded? Fibrinolysis, what counts is clot-busting, which medications? Is that medication written down on paper, is it typed into a record and misspelled, or maybe they using some code, that tells machine what drug it is, so it can deal with it very quickly, which is what we mean by is a computable. Timing, when does the time start? Does it start when the patient called, within that 30 minutes, is that when the patient is called, when they come, when they arrive, when they're first seen and is it when the medication is ordered, or when it's delivered to the floor, or when the injection starts, or when the injection ends? How do I compare two institutions rates of timing, how do I compare two institutions timing to five analysis? We may not be actually measuring what we want to be measuring.
