Discussion
PARTICIPANT: I would like to ask for an encore from all three of our speakers. I think they had terrific remarks to make, and I have specific questions for each of them.
Dr. Fisher, you let us know that we have a lot of measures of the numerator of health care—the kinds of services provided, the numbers of services provided—but not very sensitive measures of the denominator. We have to do a lot of things wrong before we statistically significantly increase the mortality rate, but indeed we do them. What can we measure that is more sensitive than mortality rates, that will give us a clue before things get that bad, when we can still do things better?
Dr. Isham, you told us that we need to apply these things to the targeted populations. Could you give some specific examples of things we could be looking at in targeted populations that would help us significantly improve the quality of health care for lots of people and decrease the costs or at least keep them from going up so fast?
Finally, Dr. Leape, you told us that we really need to be thinking about how to improve the safety of health care in lots of ways. If you reject any of the premises that you hear before your turn, then tell us. And tell us the specific things we could be doing to assess what is happening before it gets too bad or maybe even to detect things that are good and alter our practices in the interest of patient safety and better quality for all of us.
DR. FINEBERG: That is what I call an equal opportunity interrogation, and Elliott, you are first.
DR. FISHER: That is a great question. I think the question of whether mortality is a good measure depends a lot on the population. In the population of seriously ill patients, mortality is actually pretty frequent. If you look at survival following heart attacks, or if you look at hip fracture patients or patients with cancer, mortality is not that infrequent. So I would encourage us to think about adopting measures that let us characterize the care of seriously ill patients. In this population, long-term mortality and costs will provide important insights into how delivery systems perform.
Most of my comments focused on the category of care we refer to as supply-sensitive services, that is, how much time patients spend in the hospital and how many physician visits they experience. These events are very frequent and very easily measured. You can see huge differences in utilization rates with almost imperceptible differences in outcomes across delivery systems, suggesting that there are real opportunities to improve efficiency by reducing utilization, and yet sensitive indicators that would let us know when we’ve cut too close to the bone.
When we get electronic health records in place, perhaps in 10 years, I look forward to having valid measures of functional status to complement the measures of mortality. You could then measure when patients have returned to work or when they have returned to the health status they enjoyed before they came to the hospital.
If we look at the models of health system evaluation that have been implemented in recent years, such as the Medical Outcome Study or the evaluation of the implementation of the DRG system, patient-reported health status has been used as a key measure. So I would look at mortality and functional outcomes to get a more sensitive indicator. And once Lucian has our electronic health records in place, this will all be possible.
I’d like to follow up with a question for Lucian. How many hospital beds and ICU beds closed in Michigan after they reduced all those hospital stays?
DR. ISHAM: I think my question was about applying improvement to targeted populations. How can you improve care for a lot of people in different target populations? We have been doing goal setting at HealthPartners since 1994. When we first started we picked eight areas that were a mix of things, some of which we thought we could do, some of which we knew would be tough, and some of which would be real challenges. Some were chronic disease. Some were preventive services. In that go-round we had some successes and some failures.
The second time around in 2000, when we did our next iteration, we picked the top three drivers for actual cause of death by going back to the McGinnis and Foege article on the actual causes of death published in
JAMA in 1993. Instead of heart disease and cancer and so forth—the causes of death that are typically listed—McGinnis and Foege identified the drivers of those deaths as factors like smoking, lack of exercise, and poor nutrition, the real underlying causes of death in their view. We actually set those drivers of death as our priorities and established targets for them. We also selected three chronic disease; because we felt we really did need to learn how to do chronic disease well. We selected heart disease, diabetes, and cancer. We carried diabetes over from our first set of priorities. We have been working on establishing priorities for our health care system for 10 years.
As a result of 10 years of effort on the tobacco priority, for example, tobacco prevalence in our population of 630,000 has fallen from 26 percent to 15 percent, and exposure to second-hand smoke for children has fallen from 23 percent to 8.6 percent. I think that has long-term implications for future burden of illness that I can’t really measure or tell you about today, but I am working on trying to figure it out because it makes a compelling story.
In our inner-city population in St. Paul, we are grappling with the issue again of these different preventive service standards and how you deploy effective interventions to different risk populations. So, for example, some of our efforts there have to do with environmental issues such as lead screening. We have a sense that different standards ought to apply to different risk populations, but I don’t really have as much success to talk about on this topic because we haven’t had as much coherence or consistency around the guidelines or as much thought applied to this scientific problem as we need. As I mentioned in my talk, this results in confusion and less effectiveness than we ought to have.
We have just completed our third round of trying to set goals, and we now have two types of goals. We have an innovative and simplified list of goals that doesn’t sound at all like public health or medical kinds of things. It has to do with transitions in care. It has to do with communication to patients and issues like that. It is focused on care as patients experience it. Then we have the health and care list, which does sound like public health and acute and chronic care. because we have got to relate to the priorities of improving the care that doctors give.
DR. LEAPE: I think the question from Dr. Fisher about the Michigan hospitals is a fascinating one, and maybe he or I should write the Michigan Hospital Association and ask them to do a study to find out. My fears are as yours are. I suspect very few have been closed.
With regard to the question about how to stave off things getting worse or identify accidents about to happen, that, of course, is the objective we have in creating a culture of safety. A culture of safety is the ideal.
What they have achieved in aviation and nuclear power and a few other industries is an environment where people are constantly on the lookout for hazards and are identifying them. People feel not just empowered to do something about hazards, but a genuine internal feeling of responsibility to do something about them, and that is exactly the kind of environment we ought to create. We are light years from that in most institutions, and the reasons are complex.
The first thing is that health care is much more complicated than flying an airplane or even running a nuclear powerplant because it is primarily interactions of people. It is not just a bunch of equipment that we make sure works right. So the challenge is a lot greater. Right now what most people are trying to do is to put in place better systems and better practices to address the obvious problems. But what we also want is to have caregivers feel empowered to do something. A desire for empowerment is one of the things that has led to interest in reporting near misses. One thing that keeps people from reporting adverse events and errors is concern about consequences, that somebody will get in trouble. One way around that is to report actions that didn’t happen, a near miss, the hazard that somebody intercepted or recognized. That is a very positive thing, and you can talk about it safely. I think it is very worthwhile and helps change the culture. But our major job right now is doing the obvious work of implementing known safe practices and figuring out how to get the will to make that happen.
PARTICIPANT: Everyone got a laugh out of Elliott’s comment that we could do without one-third of the workforce, as if it would be a silly thing to consider, but we did have the wonderful opportunity at the conjunction of the aging of America and this excess capacity. It would take some moving around among geographies and among specialties, but it still is the case that if one-third of our health care is delivered to us in the last few years of life and we are going to double the number of people who are passing through that stage at the same time within the next 20 years, we have a wonderful conjunction of an excess capacity and a growing need where we actually could conceivably hold the line—not that we know how to do that, but we wouldn’t have to put anybody out of work, so to speak. We would make a dermatologist be a primary care physician or make somebody actually take hands-on care of a patient instead of running a lab, but the numbers of people are just about right for what we need at the current technology in 20 years. So if we just stop growing we could achieve an awful lot of this, and that seems to be a wonderful opportunity that we won’t have on the downside of the population curve. Fifty years later the numbers are starting to have to contract some, and this may be tougher.
PARTICIPANT: A general question for all three speakers. A recent IOM report found that mental and substance abuse conditions were the leading cause of death and disability for women and the second leading cause for men and that any quality health care can’t separate the mind and body. They have to be somehow coordinated. My questions to you are, if we are talking about the issues of quality of care and safety, then (1) why no mention of mental and substance abuse conditions and (2) how do we set an agenda that begins to change that and integrates mind and body within the overall system of health care?
DR. FISHER: George Isham can tell you what actually can be done. I will tell you how I would think about it. I think we clearly need performance measures that reflect the quality of the mental health services that patients are receiving. Our Performance Measures Subcommittee has been thinking a lot about it, and the current availability of actual performance measures is relatively thin. It is clearly an area where we need to invest in learning about how to measure performance more carefully.
George can tell you how it actually can be done on a statewide level.
DR. ISHAM: I remind the questioner that the priority report identified a number of mental illnesses and substance abuse problems that were important. In fact, there was criticism of the recommendation for establishing priority conditions that was made in Crossing the Quality Chasm. There wasn’t enough emphasis on mental health and substance abuse, and there was a lack of emphasis in terms of prevention. Both of these deficits were addressed, I think, by the makeup of the committee and then in the report itself.
So, for example, there is a priority area dealing with the seriously mentally ill and one dealing with depression. The other way to look at this, however, is that once you begin to look at conditions you quickly get into the fact that people don’t have just a single condition—they actually have multiple conditions like heart disease and diabetes together. In fact, a significant co-morbidity for both of these chronic diseases is depression. I think it is absolutely critical that mental health and substance abuse issues be addressed in the national priorities.
