Ninth Annual Maternal and Child Health Epidemiology Conference / December 10-12, 2003
STEPHEN E. SAUNDERS: Well, I think you’ll agree those were both excellent presentations, very informative. My role now is to kind of be a respondent to some of this…these presentations…and kind of answer the question as to why this might be helpful for a maternal and child health director, for an MCH director for the State of Illinois, what are some of the limitations we see in this kind of economic analysis/work in epidemiology and how we’ve actually used these kinds of analyses in the past, so let me start with that.
All right. Some of you may recognize this data use triangle, this is actually a concept, I think, that Bill Sappenfield and Magda Peck have promulgated in various sessions. I just put it up there because I wanted to point out that data, or in this case, economic analysis is really only one part of a data use triangle, or of a three legged stool, whatever analogy you want to use, in other words, to effect change, or policy, or program, at the State level, you need good data, good economic analysis, and other kinds of epidemiologic data, but you also have to have some sort of strategy that you can put this data or epidemiology into. And probably equal or even more importantly, you need to get the politics right. Sometimes, just having the data by itself won’t affect change, and sometimes the politic by itself will affect change without the data, so, what you need to keep in perspective in economic analysis and epidemeologic data, I hate to say that in a group like this, is not the only issue we have to deal with at the State and local level.
Another way to look at this, which is really almost the exact same diagram, it just makes it a little more fluid…is that the knowledge base, or in this case, the economic analysis or the epidemiology, needs to influence social strategy, needs to influence a political will, but it’s really sort of a circular process that one of these by itself doesn’t really make a change, that they all work together to make change, that it is an iterative process. So having some of the examples you gave…having the information or data at the beginning…in fact there’s some some social strategy…and in fact having some political will, going back and reevaluating it, making further changes and so on and so and so, I think we need to keep it in mind that although these technical analyses are terribly important, we need to put them in the context of the social strategy, programatic strategy, and of course the political environment that we all live and work in.
So, I think the basic issue, of course, is who doesn’t want to save money. Obviously, I think we’ve heard that economic arguments are persuasive, money is often much easier for policymakers to understand, a dollar saved, one dollar saved four dollars saved…even if it’s not exactly 100% accurate, I hate to say that, but in fact those kinds of arguments are much easier to make than to say ‘well, this program is going to reduce very low birth weight’ and then they say ‘well, what is very low birth weight?’ ‘Well, It’s going to reduce prematury.’ Well, is it really prematurity, is it low birth weight, is it this or that? Policy makers and members of the general assembly have a great understanding of dollars, and so clearly sometimes economic analyses are quite persuasive.
And I think they can really be more persuasive, perhaps, in our current economic environment. I know many states are suffering from economic shortfalls in the state of Illinois we’re suffering about a 2 billion dollar deficit…I guess that doesn’t hold a candle with the State of California, I think they are something like 25 or 30 billion dollar deficit, or something like that…but clearly, I think if there ever was a time when economic analyses is going to influence political will, we’re probably living in it right now. Let me give you a couple of examples from Illinois where we’ve actually used economic analysis, perhaps I have to admit we aren’t as sophisticated as some of the examples that Scott used, but at least we’ve made some kind of an attempt to use dollars and cents, or some kind of financial analysis to impact program policy.
This first example is a program called ‘Targeted Intensive Prenatal Case Management’, and this is actually an example where we, in conjunction with some advocates and some lobbyists, actually used economic data to actually get this program created. This is a program that targets intensive social support or intensive case management to Medicaid pregnant women living in certain high risk communities, and in fact, those communities were selected on the basis of Medicaid costs, so we used, instead of looking at infant mortality, which is what we’ve done in the past, the rates of low birth weight, they found, they found, or the advocates thought that they could sell this in the general assembly if they just went right to the bottom line, which was this is where the expensive babies are being born, they costing 60 thousand, 80 thousand 120 thousand each, and therefore we need to target our resources in those communities, and if we do so we can save X amount of dollars. In fact, this was quite successful and about two or three years ago, before the economic downturn, at least, the General Assembly in Illinois put in about 3 million dollars into this Targeted Prenatal Case Management program in the communities. So clearly, an example of where we used economic analysis to help us get a new program, although we didn’t get as fancy as some of your slides, Scott, but the dollars are what sold the program.
Another example is using economic analysis to actually save a program. This is an example…this is a more general case…MCH support program, we’ve had it in place in Illinois since the late 80s, and actually we receive Medicaid money, and we match that money and all that kind of good stuff, that many states do now, to locate, to do outreach, to locate as many Medicaid pregnant women as we can find, and we actually do a pretty good job, we actually locate about almost 90 % of them so it really is quite effective in that way and then actually provide their risk assessment, a care plan, they’re referred to other support services and so on.
About 8 or 9 years ago in Illinois, as part of this program, we thought we could make it better, or some of us in the state thought we could make it better, by combining it with a mandatory HMO or a mandatory managed care Medicaid program for pregnant women, like in many states there was some controversy with this, but nonetheless, our state Medicaid agency contracted with, contracted out with an outside firm, whose name I probably won’t mention, because some of you may actually still be using them, but this firm took on the charge of organizing all the managed care issues in the city of Chicago…which as you can appreciate has a large number of Medicaid recipients, so it’s an area that has an interest for policy makers. Well, this program went along for about a year and it ended up being a financial and programmatic disaster. It blew up, basically the firm ended being in a lawsuit with the state, I think they are still trying to litigate it, I think the state eventually settled the program, but in the course of this managed care issue are case management program…case management, managed care…they sound a lot alike, especially if you’re a member of the General Assembly and don’t really exactly know what all of this is.
Anyway, so they said since it didn’t work, lets just get rid of this program as well, so we pulled out some of our outcome data and one of our outcome data was looking at costs of Medicaid infants in their first 8 years of life, and we were able to demonstrate that in fact costs of the infants whose mother had been in this program prenatally was actually much less. I know for all the purists out there, there was some selection bias, some other issues going on here, but we didn’t bother to mention that part (laughter) with our friends in the General Assembly, we just said ‘Hey look! These babies cost 5 thousand on average, these infants have 5 thousand in Medicaid costs over the course of the first 12 months of life, whereas the infants that weren’t in the program cost about 10 thousand dollars on average.’ Then, you know, they said ‘that sounds pretty impressive’ and we were able to save the program about 8 or 9 years ago.
Similarly here last year, like many states we were in this 2 billion a year shortfall, and the governors’ office sent down, as they are wont to do on a Friday afternoon at around 4 o’clock and said ‘what programs do you think you can cut for next year’s budget?’ ‘Well,’ we said, and we have a few sacrificial lambs we always throw out, but we said ‘no, we cant cut this program, it’s an effective program, because we have all this data.’ And in the end they agreed, and it was one of the few programs in this area, and this is a 44 million dollar program, so it was fairly significant in terms of financial scope, it didn’t sustain any cuts in the General Assembly last year, or in the governor’s budget, and it was, basically, because we had some cost data and we had some infant mortality type data. I do think that economic analysis can be very helpful, however, I don’t think money is everything.
I know Scott talked a lot about the issue of prenatal care, but there may be some services, I would suggest, such as prenatal care, well child care, that are good…and helpful for the population, helpful for families, helpful for women, helpful for children…just on the basis of what makes sense. It’s intuitive, there’s professional consensus, if you go to the Academy of Pediatrics you’ll find almost universal acceptance of the importance of well child visits. I don’t believe they have ever been tested to be cost effective, have they? No, I don’t think so. There is this whole controversy on prenatal care, is it cost effective, I think Scott has talked about that, if you go to obstetricians, you’ll find very few, if any, that will say ‘oh no, it’s not important’, and most of us in Public Health think they are important. Sometimes, cost benefit, or cost effectiveness is not the only issues, sometimes professional consensus and other issues need to be taken into account.
But money isn’t everything. I think there may be some examples where just the example of the program itself may sell that program. The example I want to give is newborn hearing screening, as opposed to newborn metabolic screening. Now, as you know, many states over the last 3 or 4 years have actually gone to universal newborn hearing screening where there is usually some kind of a law or rule that requires hospitals to do screening of all newborns, before they’re discharged, for congenital hearing loss. My reading of this, at least in Illinois, and in the states that I’m familiar with, this program was not sold on an economic analysis basis, I don’t know if that’s your experience or not, Scott…at least WE didn’t know about the economic analysis or economic studies. This program was sold in our state, and many of the states that I’m aware of, not so much that it was going to save money, but that it was good for kids.
There were some epidemiological studies out that said that when you detect congenital hearing loss early, within the first 2 or 3 months of life and you get those kids in treatment, with hearing aids or whatever kind of treatment is appropriate, before 6 months of age, that actually those kids have near normal cognition and near normal speech, whereas if you wait and pick up the kids with congenital hearing loss in the usual way, which was before we did testing, when they’re 12 to 24 months of age, you’ve missed some opportunities in terms of brain development, and their cognition and their speech is never normal. That by itself, that evidence, if you will, which really didn’t have any costs associated with it, was enough to sell our General Assembly, and our advocates, and the folks that push this in Illinois, that newborn hearing screening, in fact, on the face of it, was a good idea, and in fact we did pass a law, and now we have universal screening, not to think that everything’s perfect in Illinois, we’re, of course, still working on the follow-up aspects of this issue, but, nonetheless this was sold not so much on cost, but that it programaticly made sense, that if you could have kids ending up with normal cognition and speech vs. not…that by itself was probably a good thing to do.
And sometimes, cost benefit or cost effectiveness, I get confused which this is…I guess this is cost effectiveness, right, is not enough? The example I’ll give you here is there have been a number of studies that say that family planning services save money and there is one here that’s quoted as saying that for every dollar spent on family planning, you’ll save $4.40. Well, that by itself, you would think, should convince the Federal Government and State Government to expand family planning all over the place, so that we could have all the women in need of family planning actually served, and those of you in family planning know the Guttenmaker projects numbers every year that says there’s X number of women in need, that’s low income women, who don’t have access to subsidized family planning services and those numbers are very large. For example, in the State of Illinois that women in need or WIN is estimated to be 4 or 5 hundred thousand. You would think we should use this cost benefit or cost effectiveness analysis to get this taken care of.
Well, it hasn’t been that easy. For example, many are aware that the federal government, through CMS, used to be, in prior days, called HCFA, does allow states to have a Medicaid waiver to provide family planning services to Medicaid recipients beyond the normal 2 months postpartum, or for other Medicaid eligible women, in fact there have been almost 20 states that have had this waiver to do family planning under Medicaid. But if, in fact, everybody believed this $4.40 argument why do we make states go through the waiver process? It’s been proven, and these are actually evaluation waivers…demonstration evaluation waivers, where the states that have done this have proved that it actually does save money to Medicaid…that when Medicaid pays for family planning services, they save money on prenatal care, and the associated costs that go with low birth weight, and so on and so on.
So, if in fact, it was just the economic argument, why is HCFA, or now CMS, making states go through this hoop of having a waiver? Why not just make it an optional benefit, like prenatal care? Well, you figure that out. And the other issue is, why don’t states, why haven’t states put their own state money into family planning if it was actually going to save money to the Medicaid program? There are only a handful of states that in fact put GRF, or state resources into family planning. I know in our state, we feel ourselves quite lucky to have a whole seven hundred thousand put into family planning, and you can put that in the whole scope of things for the budget of the State of Illinois.
So, sometimes, cost benefit is not enough to sell the issue, there must be other political issues or social strategy issues that are important. And sometimes, cost effectiveness doesn’t make any difference at all! Some services have such a strong political support that no matter what the data shows, they are going to be funded and expanded. The example I’m going to give you here is ‘abstinence only’ education. Now, I’m not going to stand here and tell you that abstinence only education is not good, but I think there’s really no evidence that it’s good OR bad…but if you look in the budget that was just passed by HHS, by Congress, it’s one of the few programs that got an increase this year, was increased money for abstinence education. So, clearly, there must be something else going on here, besides cost effectiveness or cost benefit. I’m not sure what that is, but I’ll leave that up to you to think about.
But clearly, go back to that data use triangle, go back to that circle, political will and social strategy, and so on…data is nice but it’s not everything.
What are some other issues that we should also consider? It’s difficult to stop spending the pound of cure in order to fund the ounce of prevention. I think we’ve all suffered with this problem. I gave you two examples here, one is adolescent mental health services. Now, I don’t know how it is in all of your states, but in our state, mental health services for children, or adolescents in specific, are woefully under funded. It’s very difficult to find mental health services for low income families, for folks without insurance, or even if you have insurance, it’s hard to get mental health services, and especially for children and for teens, and in fact the only ones we really serve in our mental health system are the ones that are severely involved…the kids at the far end of the spectrum, the kids you’ve got do something with because, well, at this point they’re at that far end. We would have a hard time taking money away from that, which everybody would agree is probably the best use of the money that we’re going to put in mental health and put it in prevention, or even put it in the treatment of mild behavioral problems or garden variety depression, if you will, or other issues that children and adolescents face. It’s very difficult to stop spending the money on the severely involved to put money into the ‘ounce of prevention’, despite how cost effective it might be.
The other example here is here is home visiting, I think many of you are familiar with some of the articles Olds has put out lately, about the cost effectiveness of the home visitation nursing home visitation programs to reduce costs in Medicaid and AFDC and other incarceration costs and so on…but, of course, the problem is that those costs are down the road, those savings are down the road. Maybe it’s 4 years away, maybe it’s 10 years away. Maybe they DO save money ten years out, but does it save any money this year? No, in fact it COSTS money this year. It’s often very difficult to convince people to do that.
Alright, let’s see…I think some interventions attract constituencies. The example I want to give you here is that it is very difficult to redirect resources from a good intervention, to perhaps a better one. This is the same example I was giving you before. There is this tension, this argument in the field between home visitation by paraprofessionals, the so called Healthy Families America model, anybody familiar with Healthy Families America? Many states have that, in fact, in our state, the advocates pushed this program a number of years ago, we actually almost got almost 12 million dollars of new strictly GRF new state resources into paraprofessional home visiting under the guise of Healthy Families America, we of course, called it Healthy Families Illinois, because we’re in Illinois.
But now the data is coming out from Olds and from others that in fact if you use a nurse instead of a paraprofessional you do save money whereas with just a paraprofessional you don’t save as much. That’s maybe an oversimplification of some of his research, but the point is, that once you have a program started and you have a constituency for that program, it’s very difficult to redirect that money, hypothetically, into maybe an even better program. So, even though cost benefit and cost effectiveness can be used to tweak programs along the way, it’s also hard to change a program once you’ve got a have a bunch of money out, a bunch of providers doing the program, and, of course, an entire advocacy network that thinks this is the best since sliced bread. And in fact it may be good, but would it have been as good or better to have used a different model.
This example is: Where’s the Beef? I can tell you that those of you that work with members of the General Assembly hear this argument all the time. Everybody that comes to them for any program says ‘well, this is going to save money, prevention saves money’. And in fact, it may save money, but when does it save money? But they ask, ‘if prevention is so good, and you’ve been doing all this great prevention, how come I keep spending more in health, how come the state keeps spending more in corrections, how come the state…and so on. Sometimes it’s difficult to convince policymakers and members of the General Assembly who have to deal with this issue on a day-to-day basis, that it’s really true…or it’s difficult to prove that things didn’t get worse with the intervention.
The last point I think is the most interesting. The savings may be a long ways away, I think I’ve already covered that, or, I would say, they may be in a different term…if you get my drift. Or, they may be in somebody else’s budget. So it may be just fine for the Department of Corrections if we could provide more parenting, more supportive services to at risk families, that we may actually be able to reduce future incarcerations in our state prison systems, or juvenile justice systems, but clearly those savings are going to be a long ways away, it’s going to be in somebody else’s gubernatorial or legislative term, and it’s going to be in someone else’s budget, in this case the savings might be in corrections, but the cost would be in health, or human services. All these issues are practical issues that, in fact, get in the way of using good economic analysis.
Well, the last point I want to make here, that actually both these speakers have already made, is that you need to know your audience. The kind of presentation you might make here about economic analysis, and qualities, and net savings, and utilizations, and all these terms, would not work at all in the General Assembly for sure, but they probably wouldn’t work well in the media. They may work well in certain levels of government…certain policymakers in government. You know, sometimes the farther you get up in the chain, the sophistication on individual issues is going to be less and less, so these kind of arguments need to be tempered.
This is not actually a cost benefit example, but it’s one that the staff found, and the first quote, and these are actually real quotes, the first quote is by the CDC. They said ‘The U.S. total fertility rate, TFR, moved above replacement during which a given generation can easily replace itself, for the first time in almost 30 years, in 2000.’ And that was released in February 2002. The next day the Wall Street Journal picked this up and said, ‘well, the U.S. birth rate rose in 2000 in the first time since 1971 the number of newborns exceeded deaths…the government said.’ Well that really wasn’t what the CDC said; the REAL fact was that the number of births, in fact, had exceeded the number of deaths roughly 2 to 1 every year from 1980 to 2000. So, once again, it just depends how you frame the argument. These kinds of frames, despite the limitations that we’ve all heard about, are the ones that are easiest to sell, that every dollar spent on prenatal care saves three dollars, maybe it’s not true, I don’t know but the IOM set it back in 1985, The Institute of Medicine sounds reliable, it sounds prestigious, made up of some very thoughtful and important people, and so this is something that you can sell to the director, to the members of the General Assembly, to your advocates who have lots of other issues on their plates, and so on.
The one about WIC, that every dollar on WIC saves about $3.13…I’m sure everybody has heard that statement…that WIC saves money? Well, I’m sure if you really critically looked at the evaluations they are fraught with issues of selection bias, and prematurity bias and all other kinds of biases, but is there any more successful program at getting federal money than WIC? I don’t think so. The current budget is over 4 billion; every year that I’ve been dealing with the WIC program for the last 20 years, every year they get more money out of Congress, so maybe $3.13 isn’t such a bad investment. And the one on immunizations probably is actually true but clearly it’s an easy thing to understand, especially with the issue of flu these days, I guess.
The last slide I want to show is this one that was actually already covered, about the methodology, the example I already gave you, is that a lot of the WIC evaluations that have been in place for the past 20 years are good evaluations, but none of them are randomized controlled trials, they all have the issue of selection bias, so perhaps some of the hardest to reach women never get into WIC, and therefore, when they’re in the control group their outcomes look worse and therefore WIC looks better, but nonetheless, WIC is a very popular program, I believe it’s a great program, an effective program, I think it’s a very important program, certainly Congress seems to think so, the USDA seems to think so, clearly methodology wasn’t important to some folks, lets leave it at that.
In conclusion, I think you’ve heard from all three of us that economic analysis is an important part of keeping the public trust. It’s an important tool for us to use in the management, in epidemiology, and at the federal and local level to impact programs, to modify programs, to get new programs started, to argue for the benefit of programs, but the cost and other kinds of information must be used along with good program design, and, of course, good political will to impact change, and if you recall back to that data use triangle, that circle, keep in mind that politics and social strategy are also important, as is data…and cost is not always the most important consideration in Public Health policy. We wish it was, but we all know it isn’t, but it certainly is one of the important things. Thank you.