July 22, 2009: Afternoon Session (continued)

Female Voice: Right. So for this measure we have a—it is a 6, so it does not make it currently, and there are five people who put it in the bottom four, but you can see there are 11 people—oh, I'm sorry, not quite that many. There are seven people who put it in the passing range.

Female Voice: I had a question about this measure because it is being collected which is, do you want it to be high or low? What does it mean, and when will you know you have an improvement? I'm not sure what to make of this information.

Male Voice: If I can comment, this is a measure where, you know, there is no absolute—you know exactly what the right number ought to be. I think for the other two components of the CMS-416 that relate to dental, the percent of kids that have any service within a year, you want that one to be as high as possible. The percent of kids that have preventive services within 12 months, you want that to be as high as possible.

This one, you cannot tell for sure because it is going to depend on the variability of epidemiology of dental disease from one place to another. I can tell you from practical experience that typically Medicaid programs tend to run about half of the kids that get any dental services. The A part of the measure in many programs, 50 percent of them or so will get treatment. And that seems pretty reasonable, you know, given what we know about the epidemiology, but there is no sort of fixed definitive target number for this.

Female Voice: So I mean I guess from a State, I would ask our State Medicaid program directors here. I mean if the numbers are going down, does that mean you have treated all the disease, and so you can not worry, or you are not identifying the disease enough anymore and you need to worry? I mean what can you do with it?

Male Voice: Right. Let me tell you—

Female Voice: Or is this one of the ones we can recommend to the Centers for Medicare & Medicaid Services (CMS) you do not need to do anymore?

Male Voice: No, I do not think it is [cross-talking].

Female Voice: Oh, CMS is not going to—do you think? What, are you going to stop looking?

Male Voice: In States that use multiple plans to implement their Medicaid program, you can look across plans and look at sort of to what extent the plans are comparable in terms of their performance on these measures. And you can also look geographically within States, and States have done that to see whether or not there are regions there.

Female Voice: So the eligible population for this is not kids with a dental problem [cross-talking]—

Male Voice: These will be children enrolled in CHIP.

Jeffrey Schiff: Right. That is why [cross-talking]

Female Voice: Okay. So it would be helpful if it were that children in need of treatment actually got treatment.

Rita Mangione-Smith: Yes, identified as having a problem.

Male Voice: Right.

Female Voice: Okay, any other comments on this one?

Mary McIntyre: Well, I was just going to comment on the fact that—because we have worked with Jim for years in looking at this, and we actually capture all of it. We basically went in and put a dental statistics sheet together that over time has kind of evolved to look at preventive versus diagnostic versus treatment. And we even have like extractions in 6 and under, but the reality is what we are doing is until—you know you asked for what is the great number, it is really going to be very dependent on where you are. Some States are actually ahead of the game when it comes down to overall utilization. And with us, you know, our whole goal was to increase utilization while increasing the preventive services that are being received, the population getting met above. And then we had a target with the children that are receiving care, that at least 80 percent of them are getting some preventive services. And then of that, how many of them—then we look at diagnostic and treatment services.

So I think it is some very—and you could actually put in and say, here are some goals we'll try to get to. Because the reality is, we are not there, but we are improving or have been trying to reach a specific goal that has been set.

Male Voice: One additional comment, and the reason for sort of looking at the treatment piece as well is because now that many States are actually encouraging primary care medical providers to get involved in providing oral health services and the ways, my understanding according to a conference call or a telecast that CMS put on, is that currently what gets counted as a dental service is a service that is billed under one of these Current Dental Terminology (CDT) codes, regardless of whether a dentist is doing it or a physician is doing it or whatever.

So I think there is definitely value in having the treatment piece in as well because there are a lot of possibilities for getting at least the preventive services piece in there. But we know these kids have a high prevalence of dental disease and if the treatment services are not being provided, that is another level of assessment about how well a plan or a program is performing.

Jeffrey Schiff: Okay, we have Barbara, and then Ann, and then Lisa, and then let's be done.

Female Voice: And just a quick comment to that, and I think the intent of this measure really was to assess the children that needed additional care were receiving it. But if you are raising what are the specifications, or how it is actually requested in terms of how it is being reported, we may need to look at that, and that is one of the opportunities that we are looking at.

Rita Mangione-Smith: Exactly, yes, that is great.

Male Voice: And there is one overarching piece of this that will remind you, is that there is no—use of diagnostic codes, there is no accepted set of diagnostic codes that are actually used within the realm of dental services right now. So you do not have that ability to identify a child who needs restorative treatment and gets restorative treatment. You only can know whether they got—

Female Voice: Ann?

Ann E.K. Page: Yes, a couple of things. One, remember the CMS-416 is only Medicaid. It does not include CHIP. It is a Medicaid-only reporting system so you get a hefty population reporting if you do not have the other. And the fact that oral health right now to the Medicaid directors, again, critical issue but an issue that is facing a real shortage of providers out there which is why they are looking to the extenders, to the primary care docs to provide this oral health, so I think getting the preventive stuff is critical, too. I mean shall we collect that as we struggle to go try and find a dentist that will treat a client?

Jeffrey Schiff: Okay. Lisa, did you have a comment or not?

Lisa Simpson: Oh, no.

Rita Mangione-Smith: Okay, I'm going to move us on to the next one which is another dental measure, and I'm going to ask—it is Jim, right? Could you tell us what the real wording of this measure should be exactly? Is it again present eligibles who received any dental service over 1-, 2-, and 3-year periods?

Male Voice: It is the members who received would be number of?

Rita Mangione-Smith: Percent of eligibles who received.

Male Voice: [inaudible]

Rita Mangione-Smith: All right, so this one is worded pretty accurately as it stands. It was a 6, so it just missed being included, although as you can see there are a fair number of subcommittee members who put it in the lower four categories with six people putting it down there. Lisa?

Lisa Simpson: Sorry, this one raises a larger issue that is across all the measures, but it specifies it here which is the length of continuous enrollment. And a lot of this, as we all know, and this is the—we heard from Cindy and we know the issues around continuous coverage and producing churning and all that is really important in this legislation, and we are all concerned about it. And so, we will not resolve it right now, but I think for any measure, we need to be also thinking about whether for some measures we want to recommend that you collect it for both the kids without gaps, and you collect for the kids with any gap, and you see because that gives you different answers. One is an answer about the gaps in the system and the churning that is going on and how much that is affecting their access to services versus the performance of the health plans and providers you are relying on for delivering that service, so it is—

Jeffrey Schiff: Jim?

James Crall: I'll comment on that particular issue. It is the difference between the CMS-416 measure around dental and the HEDIS [Health Plan Employer Data and Information Set] measure. The HEDIS measure says that they had to be enrolled for at least 11 months of the 12 months within the 12-month preceding period. I'm a co-author of the paper that actually looked at the differences you get using Iowa data, whether or not you use the CMS-416 measures we set as a kid is enrolled for any period of time during a 12-month period versus 11 months, or I think 6 or 7 months we picked as sort of a midpoint so—

Rita Mangione-Smith: What did you find?

James Crall: Obviously, you get higher levels of utilization for kids who are enrolled for 11 months, and I think this measure which is—it is part of a group that California just developed this. This measure has really never been used to any extent that I'm aware of—the one we are talking about right now.

Rita Mangione-Smith: So maybe if this one had a tweak in its specifications, it might be a measure that we would want?

Male Voice: I want to second Lisa's comment that we think about what is conceptually for a lot of measures because this a way that health plan can game a system. They can say, "Okay, I want to keep my numbers high. I'll just disenroll a lot of my patients, and boom, my numbers will go up, and I would be very concerned about that." That is a quality issue in and of itself, so I think it is very important that we think about that.

Rita Mangione-Smith: As we look at the EPSDT specifications for some of these measures, it might be interesting to see how much tweaking would it take as a subcommittee to feel like it was a measure worth keeping because right now it would not make it. Okay?

Jeffrey Schiff: Right.

Rita Mangione-Smith: Okay, next one. This one I think here is the place we were talking about earlier. If you look at children with chronic conditions, this starts the CAHPS® [Consumer Assessment of Healthcare Providers & Systems] measures.

So for children with chronic conditions—Denise, can we note the next time we put this through for Delphi we should really do the rolled up measures as the ones we scored; not the individual items. So for instance, under care coordination, you would just do care coordination; you would not score the individual parts of care coordination, but interestingly one of those was considered valid and—

Female Voice: Yes, I mean one of the issues is if we know what the rollup is, then that will tell us whether that is a valid measure of care coordination so—

Rita Mangione-Smith: Yes, I think it is fine to have that information there, but let's make it one single bar so that you are scoring the whole composite as being valid or not.

Female Voice: I had a question on this. In the medical home measure that is reported out of the National Survey of Children's Health Insurance Special Healthcare Needs, is care coordination that piece of the composite, both of those items because I'm only familiar with the second one of those two items being rolled up into medical home?

Rita Mangione-Smith: Oh, it is not here, is it?

Female Voice: No, but the coordination component I thought it was about services. For those who needed specialty services, where is the care coordinated? I have never seen the one on day care in schools before which—and you use it all the time so I do not know where that came from.

Female Voice: I do not know either.

Rita Mangione-Smith: So it sounds like we need to do a little investigation, and when you go through and rank these the next time, I think we probably should just because we did not do it quite right on this first round, we should redo this just—I mean it looks like by the score as you can see that they are all going to need it, right? But okay, Jeff, the next we were going to look at was preventive, is that—?

Jeffrey Schiff: We were going back up to the—not the access but the preventive one so we would go back.

Rita Mangione-Smith: Just before we leave this, did anybody have any comments they wanted to make on this "got help from doctors of"—I'm sorry, "got help from doctors as it helps the providers?"

Male Voice: Must be "or health providers."

Rita Mangione-Smith: Or, thank you. "Got help from doctors or health providers in contacting child support of day care." I mean it is just that there is a fair number of people that are on the bottom in there.

Cathy Caldwell: I actually wanted to comment on dental before we go too far away.

Jeffrey Schiff: Sure.

Cathy Caldwell: From a practical State use of data perspective, I would hate to see us say no to all the dental measures. They are very, very important. They may not be perfect measures. And CHIP does not report to the same, you know, like Medicaid does. Now, as Jim was saying, it is maybe right now impossible to say which children needed dental treatment and which got it just because of the whole data issue. The way I personally use that information when we run it for our plan, and I do not know what the gold standard is, but I look at the proportion of number of kids that got preventive versus restorative. Those numbers are useful for me, and a minimum number of time and of enrollment is important, too, because preventive is only recommended twice a year. If you have kids in your denominator that have 1 month, they really may not have needed a service during that period of time, so you need to define your universe as those who needed that service and got it. And so I look if the number who have preventive visits, the number that have restorative and even major, and I love to see it when we have a lot of children getting preventive services, and a greater number of children getting preventive than restorative. So I do not know what the gold standard is, but those numbers are actually very important.

Jeffrey Schiff: So what I'm hearing, though, is that is a different measure? I mean it is not?

James Crall: No, there is this cluster of three that are now fragmented out as separate measures, but they are all part of the CMS-416 reporting. Present the kids that get or total number of percent that got any dental service, the number of percent that got any preventive service, and the number of percent that got any treatment service defined as something beyond diagnostic and preventive services. At what sort of a cluster, so I would argue that we would hopefully—we would promote the use of all of those because they are used across all Medicaid programs right now, and if you are talking about accountability and comparability, that is almost a no-brainer.

Jeffrey Schiff: So then they need to be re-clustered in a way because it will give us more data for—

Rita Mangione-Smith: Not necessarily. I mean not necessarily; I think it is legitimate for people to look at each of those measures and make a decision. I do not know that we have to look at them as a composite unless everybody feels that that it is the way it should be looked at because that is the way they are normally collected.

Female Voice: We get to cluster it from positive [indiscernible].

Rita Mangione-Smith: Yes.

Female Voice: The cluster needs lots of measures that come in different pieces of the same thing; whereas, the composite you are going to have just one number, and I think that—

Male Voice: Cluster.

Rita Mangione-Smith: Can we just see a show of hands as to who would like this at the second round to come back as a cluster of measures, and we will rank the validity of the whole cluster, not pieces of it?

Female Voice: These clusters are the family of members around the topic. The composite means one measure that rolls up other measures.

Rita Mangione-Smith: Well, no, no, I understand that. I understand that, but it sounds like what Jim is saying is that if we are going to have one of them, we should have all of them.

Jeffrey Schiff: I think that was my point is that if you—the useful data is what you just said, Cathy, is the comparison of the preventive services to any service or preventive service to restorative service is. I understand your distinction. I guess the question is how do we want to score this going forward?

Rita Mangione-Smith: Are we comfortable as a group saying in our core set we mean just take one of those three, or do we want to say have to either have all or nothing?

Jeffrey Schiff: I think we have to be careful how we ask the question here.

So let's not call this a composite because we need the reporting of all three elements. Okay? If we do not have the reporting of all three, in order for it to be effective so I guess what I'm going to suggest is that when we put this back together for the purpose of reprioritizing, we put all three of them in there together and if, for some reason, I guess we will have to decide next time or via E-mail if they all do not meet validity and—I think what we have to know, and that is the purpose of this conversation—is what makes them valid or feasible or valid specifically is their association with each other. Does that make sense?

Female Voice: Mm-hmm.

Male Voice: Well, just the—one more round of Delphi voting. I think it would be useful to have any of these kinds of measures like this separated out because it is useful to see whether one of six in fact looks pretty dubious to people. It might be that in the end, you would say, "Well, we still want all six," but there is information there in variability, in validity or feasibility scores across individual items, if you will, and I would vote for keeping them separate and rating separately.

Jeffrey Schiff: I think this trouble I'm having—let me ask you a question. What if one part of the measure is a numerator and one is a denominator which is what I seem—it seems like we are talking about here, are they separate for validity purposes at that point? Do you understand my question?

Rita Mangione-Smith: So one measure is a count of what would be a numerator, and another what is a count will be in a [cross-talking]?

Jeffrey Schiff: Yes, exactly, so I'm having trouble, and there are three of them, so you can mix them up a little bit so they are as far as numerator—

Male Voice: Maybe I can clarify that, Cathy. You could tell me if this is how I would say. The measures as they are structured are what you said they all use the same denominator and they have various numerators. You get any service, you get a preventive or you get them all.

Rita Mangione-Smith: So then they are not interdependent?

Male Voice: But they apply—they are used in the field as what Cathy was commenting on is having the ability to look at ratios once you get those counts or percentages.

Jeffrey Schiff: Barbara had a comment.

Barbara Dailey: Yes. And my point was simply to say of those three, maybe we think two were really terrific and not the third, and that is information that is useful, so why not score them separately? It is that much more work for us.

Female Voice: I just want to highlight as I mentioned this morning, the Government Accountability Office (GAO) particularly focused on dental health at the very primary level in terms of just the preventive access—going to Ann's point about availability of providers is one of the challenges. So even though it is just a count, it is very important in terms of identifying some of the challenges of the programs, so by looking at them separately, you do glean different information.

Rita Mangione-Smith: So we will keep them separate, is what I'm hearing?

Female Voice: I recommend that.

Rita Mangione-Smith: And score them individually in the next round?

Female Voice: I would in the next round.

Jeffrey Schiff: And I think just to add to what Rita says, I think if we get different scores for different ones of this, we will have to come back and talk about what the meaning of that is in terms of—

Rita Mangione-Smith: Well, what you would have to decide is information from these as individual measures useful, or do you have to have all three for it to be used? [Cross-talking]

Jeffrey Schiff: Right, you can try to put a note on that. Try to note that on the Delphi.

Rita Mangione-Smith: So I think I'm going to move us on to the prevention and health promotion measures that are backed up towards the top probably around your first page.

Jeffrey Schiff: Right.

Rita Mangione-Smith: And the first we are going to look at is percentage of low birth weight infants. That one does not make it. There are seven subcommittee members who put it in the passing range, five who put it in the clearly failing range, and another six who put it in the uncertain validity range. Comments about this one? So right now this one does not make it the way it is scored? Low birth weight, yes [cross-talking] low birth weight infants.

Male Voice: I guess my question would be why did the people who did not vote for this not vote for it to inform the rest of us because I think this is an immensely important quality indicator? Now, of course, if it is stratified by I suppose some population characteristics, but I think it is incredibly important. I actually added up prematurity as a sort of associated or different quality measure but I'm just curious why this did not cut the mustard.

Female Voice: If I'm not mistaken, all of the States collect this measure. It is a population-based measure of the percentage of low birth weight infants in your State, so I guess one of the questions on the table because it is a population-based measure, it is a public health measure that is collected, and these measures are for Medicaid and CHIP plans. Since the State collects this, it seems to me that what we should be doing at a State level is trying to do data file matches between your Medicaid cohort and your CHIP cohort, not a plan level but at a population-based level. I'm not sure this is something a plan would necessarily be reporting.

Female Voice: So just looking back at the number of States using this, according to the survey that CMS had done, 21 out of 36, there are 36 managed care organizations, and 21 out of 36 said that they were using this measure, did that. It is used by a lot I mean—

Female Voice: It is used but they may not be reporting it. I guess that is the—

Female Voice: That we do not know.

Female Voice: I just have a question on—because this measure exists in all vital statistics, are we therefore saying that the Medicaid and CHIP programs should work with their vital statistics agency and get that measure and break it down by Medicaid, by insurance characteristics? Or are we saying that the Medicaid program has to recollect it somehow through the claims at low birth weight which I would—

Rita Mangione-Smith: Excellent question. I do not know the answer to that.

Jeffrey Schiff: I think that is for us to decide if we make this valid and important.

Lisa Simpson: Yes. I mean we are using this as a public health population-based measure and taking your Medicaid population or you taking the plan.

Rita Mangione-Smith: So let's understand how that will influence your vote. Enlighten us.

Lisa Simpson: Well, just personally, I mean—

Rita Mangione-Smith: Is that a feasibility issue the one that you are talking about, Lisa, or is it about—what would you say?

Lisa Simpson: I mean my advice being a former maternal and child health director, I mean it is a wonderful opportunity to work with your Medicaid agency and make sure that vital statistics is contributing to information about the very large population that Medicaid and CHIP cover. But it is not reported at the plan level, but it creates—but I would not want to put an added burden other than the coordination on the Medicaid or CHIP programs.

Female Voice: But it could be reported at the plan level if we work toward that end, and these are voluntary measures, keep that mind. The other thing is going back to some of the discussion very much earlier this morning, prematurity—to Glenn's point—and also low birth weight are both very expensive babies. And it would seem to me that there is some value in keeping that in mind as we put these voluntary measures on the table in that if the States are looking to save some money, one of the places where they may be able to improve quality and save dollars is in that arena of prematurity and low birth weight children.

Female Voice: Well, there is a measure of repeated premature births as well, and this could figure in if you are having well-child checks, and I know that there are studies that look at the repeated prematurity within a family, so I think this is a very important measure. It is birth certificate data. It will be there, but now it should be tracked within as SCHIP.

Rita Mangione-Smith: Could I ask if there are any Medicaid or CHIP programs around the table that use this measure and what the specs are? What the sampling is? No?

Female Voice: [inaudible]

Rita Mangione-Smith: A Medicaid director reports that their managed care programs are using this measure; that is a misunderstanding of the result.

Female Voice: Yes, I was just going to go back again to at a plan level. What I would do because it is on the birth certificate, I mean that is where you are getting it from. It is vital records. So what I would do is do a data file match, for example, with my Medicaid agency if they can tease out my plan enrollment. The plan would not report a thing. I would basically do a bump against the data files for a plan with my vital record to find out what percentage of low birth weight infants are enrolled in plan A as opposed to plan B, but it is not reported by the plan; it is on your vital record.

Rita Mangione-Smith: Why would not it be reported by the plan? Do they not know the births in their plan and what the weight was?

Female Voice: Actually, in some States the baby does not even get enrolled in the plan. It is a sort of a fee-for-service environment until so many days post-birth. I mean it depends on your State.

Jeffrey Schiff: Mary?

Mary McIntyre: I just want to comment on the fact that as far as looking at the percentage of low birth weight infants, and we do not have a plan because our program's in fee-for-service in a PCCM (primary care case management), but we do actually look at the percentage of low birth weight infants. Although right now, other than information and trying to figure out how to address it to put in something we are currently looking at are maternity-specific measures that we would try to move and change from what we are currently looking at so then this would be consistent with that process.

And one of the things we actually had looked at in trying to pursue was what we call an inter-pregnancy program that will specifically address very low birth weight—the moms that had low birth weight babies, and try to get them into this program.

Female Voice: We use the information to [inaudible]

Mary McIntyre: And so they use it for actually quality improvement because right now we are collecting it, and we are not really doing anything to improve quality, so we are trying to change that focus to actually get to the other side of it.

Jeffrey Schiff: Great.

Female Voice: I just want to note on the feasibility side that State maternal and child health programs have to report annually on a number of things, and one of the things is whether they match their vitals data with their Medicaid data so the MCHP (maternal and child health program) bureau could tell you how many States are doing these matches. It is not going to tell you whether they can get down [indiscernible].

Jeffrey Schiff: Okay. If you have a comment on this, put your cards up, and then we will not take new ones after that. Okay, Paul.

Paul Melinkovich: The problem I have with this measure around Medicaid is not that is not an important measure, but often patients are not enrolled in Medicaid until they initiate prenatal care or until they deliver. I mean many of the patients that we take care of have never gotten into Medicaid until they deliver, so it is hard for me to fathom how you measure the processes of care that would impact low birth weight if they are not in Medicaid until they deliver.

Rita Mangione-Smith: That comes into specifications, you know, that you specify that the mother was actually cared for through Medicaid.

Jeffrey Schiff: Okay, Ann.

Ann E.K. Page: I wanted to follow up. Mary, I think it was an interesting point. At least it made the light go off in my head. Most Medicaid clients are in some kind of care management, but a lot of them are not in insurance company plans. But the PCCM model is very, very, very common, and in some States like Oklahoma and Vermont, the managed care plan the State has and they manage everybody's care, so you are not going to get information on a lot of Medicaid clients from plans.

Male Voice: Right.

Rita Mangione-Smith: So in the same vein, initiation of prenatal care got a 6.

Female Voice: Where are we?

Rita Mangione-Smith: We are in the next one down—just the next down. Got a 6 so it does not make it, but only two people have it in the bottom four and they are even out of four. So we also have an additional seven people who gave it a failing validity score by making it uncertain.

Return to Contents
Proceed to Next Section