I have been reading a swath of economists and their analyses regarding the increases in health care costs. Mankiw has referred to several and the most obscure was that of Finkelstein at MIT who looks at the issue from the perspective of the microeconomist who is looking for correlations in time series data with no rational underlying model. As we have stated many times before, if one looks at the bottom up one can readily develop a model, and one which reflects reality.
Let us look at a simple model. That for analyzing ischemic heart disease. This is the heart disease which results from blockage of the arteries and is often reflected by people having Angina or shortness of breath. It is simply a plumbing problem.
From a cost perspective there are two questions; how do we diagnosis it and what does that cost, and how do we treat it and what does that cost? There is a third ancillary question and one quite important which is; does any of the treatments make any difference?
Let me answer the ancillary question first, yes the treatments reduce mortality by 50% over the past 40 years. Thus one does not have to be concerned with the issue of clinical effectiveness. Now one may ask is that a good thing, namely do we want to spend all that money on people who will just get old and die of something else anyway, well I leave that to the brilliant medical ethicist at such places as NIH, clearly the p;ace we would want to go to for moral decisions. But alas I digress. Back to the first two questions; diagnosis and treatment.
We can consider two scenarios, performing this diagnosis and treatment in 1970 when the costs were low, remember mortality was twice what it is now, and then doing it today, forty years later. To do this we look at Harrison's in 1970 and Harrison's today, since they will provide the procedures.
1970 Scenario
Diagnosis; we used an EKG, a simple chest X Ray, a stress test on a tread mill, and a stethoscope.
Treatment: we used nitrates, nitroglycerin tablets, recommended rest, and very little else. Eventually the patient dies.
2010 Scenario
Diagnosis: we use the EKG, stress test, electron beam CT, Nuclear imaging, echocardiography, positron emission tomography, exercise perfusion scintography and the like.
Treatment: It may start with medication such as nitrates and beta blockers and calcium channel blockers but soon progresses to catheter revascularization, or placing a stent, or even coronary artery bypass surgery. It may actually include several of these over time.
Now back to cost. The cost equation we have used can now be applied to ischemic disease. Namely the total health care costs for ischemic disease is:
Total (Ischemic Disease) = Population Size X Incidence (Ischemic Disease) X Procedure per Patient X Costs per Procedure
Thus we know the population, we know the incidence, neither of which we can do great deal to control, although we do have some influence over incidence by diet and the like, and we could control the costs, albeit they are not the problem, it is the procedures, diagnostic and treatment. Thus to look at and understand the costs of health care one MUST look from the bottom up and not from the top down. Looking at time series, doing regressions, and the like lets one see wonderful correlations but no causality and especially no way to control the system. One must look at this as a business and if you will as an engineer. How does it work at its very core and that what knobs can I twist to improve it?
As we have shown above the number of treatments have increased yet the mortality has decreased. Do we reduce the treatments and see an increase in mortality? This becomes a moral issue, the value of life, not just an economic issue.
There is another question which flows from this approach to the problem. Can health care ever exceed 100% of the GDP? This may be a self contradicting question but when one looks at the arguments about the exploding health care costs one seems to see that people look at health care as an ever increasing set of costs. You may just run out of physicians and surgeons before that ever happens. And a scenario one may truly have to worry about is that medical school grads may become like public school teachers in the 21st century, poorly educated and unionized! Just think of that horror story.
One need just look back to the 1930s and 1940s in New York City and Public School teachers were some of the best and brightest in the realm of education. Then they left and went to Wall Street. What filled their spots were in many ways what got public education where it is now. Will this also happen to medicine, will it be dumbed down so that the medical school grad of 2050 be the equivalent of a Practical Nurse of 2010, not disparaging a PN, but they are not asked to perform brain surgery. Then add making them unionized so that half-way through the operation they have to see the shop steward to decide if they will proceed since they are now on triple overtime. That will take care of all of the problems in health care.
Let us look at a simple model. That for analyzing ischemic heart disease. This is the heart disease which results from blockage of the arteries and is often reflected by people having Angina or shortness of breath. It is simply a plumbing problem.
From a cost perspective there are two questions; how do we diagnosis it and what does that cost, and how do we treat it and what does that cost? There is a third ancillary question and one quite important which is; does any of the treatments make any difference?
Let me answer the ancillary question first, yes the treatments reduce mortality by 50% over the past 40 years. Thus one does not have to be concerned with the issue of clinical effectiveness. Now one may ask is that a good thing, namely do we want to spend all that money on people who will just get old and die of something else anyway, well I leave that to the brilliant medical ethicist at such places as NIH, clearly the p;ace we would want to go to for moral decisions. But alas I digress. Back to the first two questions; diagnosis and treatment.
We can consider two scenarios, performing this diagnosis and treatment in 1970 when the costs were low, remember mortality was twice what it is now, and then doing it today, forty years later. To do this we look at Harrison's in 1970 and Harrison's today, since they will provide the procedures.
1970 Scenario
Diagnosis; we used an EKG, a simple chest X Ray, a stress test on a tread mill, and a stethoscope.
Treatment: we used nitrates, nitroglycerin tablets, recommended rest, and very little else. Eventually the patient dies.
2010 Scenario
Diagnosis: we use the EKG, stress test, electron beam CT, Nuclear imaging, echocardiography, positron emission tomography, exercise perfusion scintography and the like.
Treatment: It may start with medication such as nitrates and beta blockers and calcium channel blockers but soon progresses to catheter revascularization, or placing a stent, or even coronary artery bypass surgery. It may actually include several of these over time.
Now back to cost. The cost equation we have used can now be applied to ischemic disease. Namely the total health care costs for ischemic disease is:
Total (Ischemic Disease) = Population Size X Incidence (Ischemic Disease) X Procedure per Patient X Costs per Procedure
Thus we know the population, we know the incidence, neither of which we can do great deal to control, although we do have some influence over incidence by diet and the like, and we could control the costs, albeit they are not the problem, it is the procedures, diagnostic and treatment. Thus to look at and understand the costs of health care one MUST look from the bottom up and not from the top down. Looking at time series, doing regressions, and the like lets one see wonderful correlations but no causality and especially no way to control the system. One must look at this as a business and if you will as an engineer. How does it work at its very core and that what knobs can I twist to improve it?
As we have shown above the number of treatments have increased yet the mortality has decreased. Do we reduce the treatments and see an increase in mortality? This becomes a moral issue, the value of life, not just an economic issue.
There is another question which flows from this approach to the problem. Can health care ever exceed 100% of the GDP? This may be a self contradicting question but when one looks at the arguments about the exploding health care costs one seems to see that people look at health care as an ever increasing set of costs. You may just run out of physicians and surgeons before that ever happens. And a scenario one may truly have to worry about is that medical school grads may become like public school teachers in the 21st century, poorly educated and unionized! Just think of that horror story.
One need just look back to the 1930s and 1940s in New York City and Public School teachers were some of the best and brightest in the realm of education. Then they left and went to Wall Street. What filled their spots were in many ways what got public education where it is now. Will this also happen to medicine, will it be dumbed down so that the medical school grad of 2050 be the equivalent of a Practical Nurse of 2010, not disparaging a PN, but they are not asked to perform brain surgery. Then add making them unionized so that half-way through the operation they have to see the shop steward to decide if they will proceed since they are now on triple overtime. That will take care of all of the problems in health care.
Posts
