Children's Hospital Regional Medical Center, Seattle, Washington
John: Yes. I’m John Neff from Children’s Hospital in Seattle, and this is Ginny Sharp here on my right, and Jean Popalisky. And we’ll do it tag team. It will be fairly evident as we move along. This is the topic of our presentation: Identification and Tracking of Children with Special Health Care Needs in a Managed Care [Setting], using the NACHRI classification of congenital and chronic conditions, which is that has been adapted by 3M.
And the next slide shows our project partners--so it’s fairly broad--our center at Children’s Hospital and Regional Medical Center, the Department of Health, the Northwest Medical Bureau, which was a plan up in the northwest counties of the state and that now has been bought up or been merged with Regents Blue Shield but is still a separate--during the whole grant here, this was a separate entity representing these four northwest counties. In that were 11 pediatric and family practice clinics, which were part of our group, and then the CIG means the Coding Investigation Committee. Those are the groups that we met from all the practices to go over in terms of how do you identify children with special needs using administrative data.And then the PHIG stands for Public Health Improvement Group, which was a broader group--which we’ll describe--to try to integrate this into public health systems. And NACHRI is the National Association of Children’s Hospital and Regional Center, which was a partner in this whole grant.
The project staff are the three of us that we’ve just introduced and who’ll talk; the Northwest Medical Bureau, who has been--Jeff Graham, who has been at some of these meetings in the past; and then the Washington State Department of Health is Jan Hicks-Thompson. And there are many, many others, but these were the key individuals.
The project goals were to demonstrate the effectiveness of a diagnostic-based chronic disease classification system and to accurately and consistently identify children with special needs, and to apply this in a managed care setting. The objectives, there were several: to improve the pediatric coding by providers; to increase recording of secondary conditions, not just one but get all the secondary conditions that you would expect; to improve the accuracy of the diagnosis and the severity; and actually, to develop a sustainable system for tracking these children through a period of time; and then, to facilitate information on what we’ve gathered on children with special health care needs for management and long-term care. And with that, I’ll turn it over to Ginny.
Ginny: Okay. Here our original project flow diagram that shows the process of the project. You notice there are three very separate streams diagramed on here; that Public Health Integration Group, which is sort of a broader, statewide dissemination and get it into public health idea; the billing--the analytical part, which is my baby. You know, let’s collect that data, baseline, let’s give them some training, and then let’s collect the impact data, you know, after the fact and see if the training worked and analyze it; and then, of course, there’s cross--fertilization here with the other groups; and then the coding investigation group, which was this group of, basically, the pediatricians in the largest clinics providing guidance every step of the way.
Now, one of the things we were asked to report on, I think is the changes that occurred over the four-year period. As you notice, I mean, our project staff stayed constant the whole time, which is pretty amazing. But there were a few little minor things that happened that affected the project. To start with, between the time we submitted the proposal and the thing started, the Skagit Medical Bureau and Whatcom County Medical Bureau, who are our original partners, merged. So before we even started, we had a merger. And then at the end of the project, as of January 1, 2001, Northwest Medical Bureau merged into Regents. Now, any of you who have worked with health plans and health plan data, know that if you change who’s in charge, chances are the data systems and how you access them are going to change. So, these were a few little minor changes here. They installed a whole new computer system and billing software for northwest Washington Medical bureau in 1999, which meant that we could not get any 1998 data. Period. And then the change in the Regents Blue Shield system that occurred last October.
Significant changes in the lines of business that these various health plans had were serving. In Washington, most of our Medicaid--our Medicaid managed care program is called Healthy Options. Skagit Medical Bureau had a large Healthy Options contract to start, so they served the Medicaid kids in their county. Whatcom did not. When they merged, they got--they picked up--Northwest Washington Medical Bureau picked up Healthy Options in all four counties for 1999 and 2000. So the population jumped, the population of kids, from who we’d had at the start of the program. And then in 2001, they dropped the Healthy Options again in part of the area because it was too expensive. So, the population of kids went from about 18,000 up to almost 45,000 and then down again to 28,000. Okay. So, if you’re trying to track kids across time and they drop out of your system, it makes it sort of difficult. The little minor things.
As you notice, our title slide is the original title of the project. In fact, the NACHRI classification system got subsumed and merged into this 3M Clinical Risk Group system partway through the project, starting in 2000 basically. So that’s a whole different kind of software than what we originally thought we were going to be using, although we did do both, actually, at various times in the project.
Our evaluation design involved, basically, trying to measure the change in the quantity of the ICD-9 coding. And the ICD-9 coding, as probably all of you know, is what’s used for billing through the health plans. We are looking at specific coding practices of the individual clinics, and we conferred on some of that in the past, some sort of baseline stuff. We wanted to evaluate year-to-year changes in the diagnoses and how well we can identify chronic kids, and evaluate changes in the clinic and individual providers. This is sort of hot off the press here, some of our more interesting findings here.
This is, I know, rather complicated. We ended up using data for calendar years 1999, which was before we provided the training. Okay. We basically were working with our coding investigation group through calendar year `99 to come up with a way of improving coding practices. We started doing the training in December `99, January 2000, and so the impact year is calendar year 2000. And what this shows is for different groups of clinics--of course, not all pediatric clinics accept training and participate in training at the same rate. We had a couple of clinics who not only adopted what we suggested, which is to use a little Post-It note that just says "no chronic condition," "chronic condition," and then a place to fill in the underlying chronic condition for the child so it gets coded. Very simple. And some clinics accepted it the way we proposed it.
Two clinics, two larger clinics, actually, went beyond that and actually fed the information on underlying chronic conditions for each child into their clinical system that prints out the face sheets when you see a doctor. So when the patient came in, any underlying chronic conditions that had been noted in the past were already printed there for the doctor to see, and they could just circle if it had an impact on that visit, okay? So those were the ones who sort of went beyond our training. They, you know, fully implemented something in their practice. There was one group that took the training, didn’t make a change in their super bill itself but used the sticky notes, or at least to some point. And there are others who sort of half-heartedly took the training. You know, one or two docs showed up, maybe their clinic manager did or didn’t, and, you know, they weren’t interested in any follow-up visits. They, you know, were not really interested in the project all that much. So, that’s how we sort of divvied up the analysis here.
So this shows changes in the percentage of claims coming from each of these clinic groups, actually each clinic, that had multiple diagnosis codes. So that, obviously, all of them have the primary diagnosis because you can’t submit it without it, but this shows how many had a second, third, fourth, and fifth and the change over the two-year period. Now, this is all claims coming out of each clinic. What’s more interesting is the next one. This is looking just at children who were identified by the Clinical Risk Group software in either 1999 or 2000 as having a chronic condition. Okay. Now, the Clinical Risk Group software looks at all the data, not just the pediatrician’s visits but it also looks at any in-patient hospitalizations, ER visits, specialists--all that stuff that goes into a child’s health care. This is looking at--for the physician clinics that I already described, the change in the coding for just their office visits that were for evaluation and management, which is a certain group of procedure codes. So we’re not--I’m excluding the visits that were for well child, totally well child, or just immunizations and that kind of thing. But this is when they’re actually managing the kid, okay? So they’re not just zipping in to see the nurse and getting a shot, okay? You notice that the folks who actually changed their face sheet, or their super bill, had a much greater improvement in the average number of I-9 codes carried for these chronic kids. And this is statistically significant at the .01 level. We are very pleased with this result, needless to say. This is exactly what we had hoped for. You know, the more they accepted the change, you know, the more actually happened. Next.
Audience Member: So, some of that--
Ginny: What--wait--
Audience Member: Oh, I got it.
Ginny: Oh, you got it?
Lynda: Yeah. It might be easier if we got.
Audience Member: Here, John.
Lynda: Can everyone hear?
Audience Member: Do you have any idea what kind of diagnosis appeared...?
Ginny: No, I haven’t. John will show some of that, but--
Lynda: Do we want to stop for a second? I mean would it be helpful for them? What’s your thoughts?
John: Sure. We’d be glad to answer questions. Probably easiest would be clarification questions about what had been presented, and then we’ll have a full-blown discussion at the end.
Lynda: Does anyone have--because I have one.
John: Sure.
Lynda: I’m kind of--in terms of going back to that--yeah. Going back to that--because this is interesting that actually the group that wasn’t trained had, I guess, a higher--prior to the training or lack of it, had a higher number of codes.
Ginny: Well, yeah. But that could be a lot of specialists, too.
Lynda: Oh, okay.
Ginny: You know, that’s not just the pediatricians. That’s everybody, all clinic providers.
Lynda: I mean, I’m trying to think if for the group that wasn’t--this wasn’t a group that wasn’t trained but didn’t want to--they were asked--
Ginny: That’s if there was some training.
Lynda: The last one. Not trained.
Ginny: No. No.
Lynda: They didn’t participate at all. Were they asked to participate and refused?
Ginny: No. No.
Lynda: Okay. This was kind of like--
Ginny: These are other kinds of providers. These might be family practice. They might be specialists. They might--you know, could be the asthma and allergy guys.
Lynda: Okay. So it’s not a group that says, basically, "I don’t need training because I’m already up to speed."
Ginny: No. I thought that was curious, too. But it also--it’s--well, I won’t say that because I'm not sure. It should just all be clinic visits.
Jean: Well, some of the other key results that happened within the health plan as a result of this project were significant. They developed a unique ID for each child, and that was a major task because some children, even with multiple coverage, they were able to have a unique ID, which is critical for being able to pull out unique kids.
They developed a data queue for quarterly tracking of the chronically ill children, and this allowed them to connect with their server to get updated encounter data with the NACHRI codes so that they could pull up monthly reports and quarterly reports. Unfortunately, when they went from Northwest Medical Bureau to Regents, we lost that. And they’re going to, hopefully, try to reconnect that. But when they had the data system changed, they couldn’t reconnect it to the server. But they still had the development of it, and we were still just beginning to use it when the merger happened.
The other thing is they really streamlined their care management process, which really was unique because they had a lot of nurses doing non-nursing activities. And they realized, as a result of this project, that they really need to think about their whole care coordination process within the health plan.
They also developed some additional tools to look at outreach to adults based on some of the work that we did with the children in this grant, and I’ll talk about that in a minute.
And they also had groundbreaking compliance with the Medicaid contracts. This is also significant because in the Medicaid contracts, they talk about the requirement for care coordination for children with special health care needs. However, they don’t really talk about identification, and so this project allowed them to really be able to identify these children with special health care needs and then really comply. And they got high, high marks, and we’re very excited as a result of this.
We talked about the Coding Investigation Group. This was really a critical group throughout the project because we had full participation of four of the highest volume pediatricians throughout the project. And they were significant in giving us, you know, advice throughout the project. They corrected a lot of the errors in the face sheets. There were actually some providers that had codes that were either outdated or made-up codes. I mean there were all kinds of ambiguity in the face sheets. And we provided throughout the project--and Ginny’s reported on some of this in the past--some of the provider’s specific feedback about their kids within this particular health plan throughout the project. And this was something that they had never had before. This was information to them about their population that they otherwise weren’t pulling up on their own.
And we talked about some of the coding changes that were done by the providers with the face sheets. And then parents were also a part of this group, and they really provided a huge contribution when we got to the case management portion of this project because they really helped us to think about what is it that parents want. What are they getting from a health plan? What are they getting from other care managers throughout the community? So one of the things that we did in this last year is we looked at care coordination, and we developed an application of the CRG status and severity to really look at those kids that we thought might need case management. And then we developed the telephone survey that we used for this subgroup to further look at whether or not they needed care management services. So we implemented that survey with the focus group based on calendar year 2000 CRG classification.
And in this slide, we pulled out a theoretical group that we thought might need care management services. And we just pulled the higher severity, higher complexity kids. And that’s basically what this slide shows. You can pull it out any way that you want, but we wanted to start with a smaller group because we knew that the resources were limited for the care coordination. With this slide, we came up with an initial focus group of 339 kids. But when we matched it for enrollment a year later, we only got 197 kids in the focus group.
Ginny: And that’s the impact of dropping the Healthy Options.
Lynda: Can you go back and maybe for certain people who aren’t familiar, just kind of quickly talk about the different categories in the CRG.
John: Sure.
Lynda: I don’t if you’re doing--
John: No. No. No.
Lynda: We’re you going to do that later?
Jean: No, we--
John: I can do it. I can do it.
Jean: No, we don’t so go ahead.
John: Healthy children are children that have none of the above. The significant or acute are a series of conditions that are identified by the CRG, which in a sense are the children at risk. They’re the children that have significant trauma, meningitis, that have had child abuse, have had significant acute illnesses, and prematures. Single minor are really minor and now for--but the group that comes out in that classification group has the highest percentages of Attention Deficit Disorder. So one might consider actually looking at that group and moving them up in severity, but that’s well over half of that group.
Multiple minor means you just have minor conditions of more than one body system, like in your brain or your heart or other systems. Single dominant or moderate chronic is the next level up in severity, and the biggest group within that is asthma. And as you move out into the severity--that is, the severity levels, which are one through six, those are determined by co-morbidities and utilization of healthcare services. And as you move up into the severity of that group, it’s conditions of children that have asthma plus other physical conditions. I might add that most of the mental health conditions fall into the single dominant or moderate chronic group, and I’ll talk about it later. They’re really a very high percentage of children with special needs.
Pairs and triplets are children that have either the single dominant or moderate chronic occurring in more than one body systems. They’re relatively small numbers, representing only .42 percent of the population. But this is to be used for adults, and for adults, that becomes a very, very significant number. Malignancies are the metastatic malignancies, not the benign malignancies, and catastrophic are those conditions that you know are going to last probably for a lifetime; they’re quadriplegic, muscular dystrophy. They’re really the severe conditions. Yeah?
Audience Member: John, where would development disabilities fit in this?
John: Mostly into the single dominant or moderate chronic.
Audience Member: Okay.
John: If they’re mild developmental disabilities, they would be more like in the minor, but I can talk about that later on. And a lot of those conditions, at least as they come out in the CRGs, are kids that have developmental problems plus something else. So, it’s the plus something else that would probably drive them to a higher level of severity. So, Downs would be identified as Downs and so forth. But I can talk about that later.
As you can see from this, this has been pretty well duplicated in Betsy’s work and other people that have worked it. About 10 percent of the population, using the CRGs, comes out as having a chronic condition. And that’s less than you would expect from primary data sets, which are from the children with special health care needs survey, or the QuICCC-R, which is going to identify more closely to 18 or 19 percent. And I’ll talk about that later on. So, it’s both by conditions going this way and by severity going to the right. And I’d say by this five percent or more, generally around five percent, are ones that children that have more than minor conditions. They’re moderate and above. Is that--any questions on that?
Lynda: And, John, did you say that is a--are these same categories also used for the classification of adults?
John: Yes, they are.
Lynda: Okay. Because I think, you know, in terms of when--and when we have our discussions in terms of getting managed care that to think about identifying children with special health care needs, obviously that’s a big advantage to a plan to be able to use one system that covers both--covers their entire population base, their entire membership.
Auduience Member: So this was not the MCHB definition of children with special health care needs? This was the IC-9?
John: Yeah. That’s correct. And I’ll talk about that later on. Those of us who helped develop this, through NACHRI on up, used the MCHB definition as a template to describe what was a chronic condition. And we used that as a template as you moved out into the severity so that that was before us all the time. And that’s why we have an at-risk group in there and that’s why we have levels of severity.
Audience Member: This is known on 0medical bills for services in the last year--
John: Absolutely.
Audience Member: --not (inaudible) population group.
John: Exactly. And I’ll talk about some of that as we--in my part. Yeah.
Jean: Lynda, you also made a really good point about--
Lynda: You have to turn it on.
Jean: Where is it?
Lynda: It might be easier that way.
Jean: It’s on.
Lynda: I don’t people in the back are having trouble, if there’s noise--okay.
Jean: Anyway, you made a really good point about the managed care environment, and this is one of the carrots that we’ve really used to work with our managed care partners throughout the state is that oftentimes they say, "well, children with special health care needs are really a small population." But we have been able to work with them, saying if you can develop a system with them, then you can apply that to the adult population, which is exactly, exactly what’s happening within this health plan which covers kids all over the state. So I think that’s significant.
With this care coordination, as I said, we came down to 197 kids in the focus group as a result of matching for current enrollment. And we’ve really only been able to contact, with the telephone survey, about, so far, 140 of these kids. We’re still in the process. We’re running it right up to the edge to the end of the grant. So this is sort of some interim information. But 97 of the families completed the survey. It was a phone survey. One of the care managers called the families on the phone, and 10 children were identified for a formal case management that otherwise had not been picked up by the plan. Generally, plans have a dollar amount that kicks in that they will then case manage those people that hit that dollar amount. Some of them will also talk about specific severity of services or intensity. And the care manager really was aware of family impact with this survey, and so 10 of these children are now being case managed. And in addition to that, 21 of the families have gotten additional information by phone, not put into a formal ongoing case management, but really given a lot of resources and referrals in the community with the health plan. And so we think that this is a significant group of kids that might need, what we are starting to call amongst us, sort of a level two case management. They don’t need all that intense ongoing, but they might need some help. They might need a jump-start. They might need referrals. They might need some additional resources or just knowing that now the health plan may have a different rule. They can talk to people. The have a contact. If they do get into that particular situation, they have somebody that can help work them through the system.
Faye: A question about the Medicaid rates in your state. Does your state limit the number of encounters for a child covered by Medicaid that can be billed in one day?
John: Oh, in one day?
Jean: I don’t think so.
John: I don’t think so.
Faye: Because in Illinois, this is something that we’ve been fighting on for years. So if a child comes to see the doctor in the clinic, and they have a case manager visit it with a nurse or a phone call, the facility will either have to eat the cost of one of those encounters or they may have to make the family come on separate days.
Jean: I don’t think so, particularly because most of the kids in managed care--I mean, they’re given a blanket amount per month, and then they just are supposed to provide the care for that amount.
Audience Member: I have a question about the 97--or the families that you’ve already conducted. In addition to the 10 that you were able to refer and enroll into a full case management program, did you encounter among those 97 other children who were already receiving case management services from the plan?
Jean: No. Not from--
Ginny: Oh, I thought there were two.
Jean: No. We already pulled--I pulled those out of that number.
Ginny: Okay. So--
Jean: They might have been receiving case management services outside in community, and we do have that documented, but I didn’t present it because it’s still so early.
Audience Member: Okay.
Audience Member: Was this a two-way pass that some families didn’t want the services as well? The 97....
Jean: We have--no. They were--
Audience Member: It shows that you offered it.
Jean: Correct. They could refuse it, but nobody had refused it, anybody that was offered it.
John: I think what the plan was excited about this is that a lot of people really had no idea what resources were out there to support them. And so that it really was a way an avenue for the plan to get people the resources that they needed without a lot of other expenses. And I just would like to emphasize what Jean said. This was a sample that we took, which we took a group that we could manage. This was in a major part of our grant and that you can expand this within this system depending on kind of how far up the severity level you want to do. So this is just a trial and a sample group.
Jean: Okay. The other part of our project was this Public Health Integration Group, and this group ended up going through for year one through three instead of all four years, and I’ll explain why. But that gave us the additional resources in the final year to really think about what do we want to do now that we can identify these children within a plan. And we had four meetings in the first three years.
One of the first things that we did is we did an in-depth data assessment at the local level; that means in the four counties figuring out who had what kind of data, who had sharing agreements, did anybody share, et cetera and so forth. And as a result of that, we really discovered that there needed to be a lot more interaction with the state agencies because they were really wanting information from different state agencies specific to their county, which was not always readily available. So what we did then is we pulled together state stakeholders from mental health, Department of Health, DSHS, multiple, multiple state agencies. And we pulled them all together to really think about how is it that you want to integrate you data, and what is it that can be done. And what we discovered is that this was a much larger portion than we could possibly take on in this project. However, there was a real readiness within the group and with all these different agencies. There really was a feeling that the time was right. It just wasn’t going to happen with our project. But as a result of that meeting, the Title V Department of Health representative that we’ve had on with our grant went on to apply for an MCHB grant and was--received awarding for Washington Integrated Services Enhancement for Children with Special Health Care Needs. And they’ve got several subgroups, and I’m a part of some of that work. So these are really very well integrated within the state. So we’re able to keep an eye on what’s going on. It just didn’t happen under the resources of this grant.
Lynda: Is that a CISS grant?
Jean: Integrated Services?
Lynda: Yeah.
Jean: Yeah.
Lynda: Yeah. Okay.
Jean: Okay. So this last piece is really our model of how can we summarize what we’ve done over the past four years? And this says side view, because we thought, really, we’re looking at the target population and these are--you can’t see very well, but there’s really circles. It’s sort of like, you know, a concentric circle with the target population being in the middle. And what you really need to do is you need to start out with your population. You want to think about who you’re going to include in your population: all members versus lines of business. You have to consider changes in coverage and provider network, and you have to consider population mobility, lapses in coverage--these have to do with some of the continuous enrollment requirements to run the CRGs.
And the other thing you want to consider is claims data. What’s your definition of paid claims? We ran into that under managed care or fee for service. It really can show up differently in the data.
System upgrades and conversion. We’ve had two or three of those throughout the project, you know, and you have to just be able to keep moving forward with those kinds of changes.
Kids with multiple coverage, we already talked about that in terms of making sure that you have a unique id for each child depending on the coverage.
And then what’s the quality of the diagnostic information from the providers? That was the big thrust of this grant, and we can see that that’s important.
And then for the analysis, what operational definition of children with special health care needs are you planning to use? What are the technical specifications to allow you to achieve that?
And then you have to think about how you’re going to use the data. What’s your purpose in gathering this data? Are you going to use it for care management? Are you going to use it for some quality tracking? What is it that you want to do with it? And then to find the methodology.
So when you look at your data set, you really want to make sure that you only have children with special health care needs; that you have current, unique members; that you have the particular age group that you’re interested in looking at; and that all duplication has been eliminated.
And then the screening piece, that’s what we used is a telephone screening because it was really so critical for families to know that they had a resource available to them and they were very open in talking with the care manager and very interested to know that there was another resource available for them. And then it brings you to your target population and how you want to proceed.
John: Thank you, very much. I suppose I ought to say this at the end, but I think you can see what a great group this is and how much pleasure I’ve had working with these two professionals.
This is some of the key results, and I want to go over this a little bit slowly, just because this tells you roughly how CRGs do in identifying children with special needs, depending on certain major conditions. And these we picked out of the literature as the ones that we feel are the ones that are most prevalent as representing conditions of special health care needs. And these are the prevalences for each one that you would expect based on a literature review. And not surprising, the one that’s way at the top of mental health conditions, mostly from SAMSHA data, which is pretty good, say that within the school-age population between 6 and 12 percent of the children have significant mental health conditions. The six percent represent the severe; the 12 percent represents the whole. This shows you what it demonstrates by the CRGs. The CRGs did surprisingly well picking up close to three percent there, and I’ll go into that later on.
Asthma is another one that’s very prevalent, which occurs in about four to six percent of the population, and CRGs will pick up about two percent. And again, these are children that interact within the system.
Attention Deficit Disorder, which in some areas can be melded into here, come up with a two to four percent prevalence in the literature, and CRGs pick up about 2.3 percent. And that would make sense since these children go to the doctors for diagnosis and for medications.
Learning disabilities, it’s hard to find out what that exists in the literature. It’s a squishy number, but it’s certainly above five percent. And that, in the CRGs only pick up 0.3 percent. So that administrative data’s not going to do a good job picking that up. Mental retardation likewise; .4 is a severe; three is generally what’s the accepted number; and CRGs only pick up .2 percent. So it’s going to be low in these numbers.
Then Cerebral Palsy also is low, and that’s because a lot of these went to neurodevelopmental clinics for their care, not to the mental health.
But things that you would expect to interact with the health care system--diabetes, cystic fibrosis malignancy--and you could get a large number of these. They’re very, very close in the administrative data. So if you use administrative data, this is a pretty good framework in which you can kind of put your idea of what you’re picking up and what you’re not. Yeah?
Audience Member: Did the enhancement of coding, which was the result of the training, improve any of those numbers?
Ginny: No, we never looked at that.
John: We haven’t looked at that yet, because the others just relatively frequent--I imagine it will in the mental retardation because that’s where most of your secondary diagnoses come in and because people say, "Oh, I don’t record that."
Audience Member: Right. Right.
John: So I think that’s--
Audience Member: So you may--if was just...collect some of them.
John: Right. Right. And the other thing I’ll talk about in terms of chart reviews, whenever--which we’ve done in the chart reviews. Whenever you go down and take a cross-sectional view of the population, you’re only going to find out not only who are utilizing the services or not but children with special needs have a tremendous variation in terms of one year they use a lot of services and one year they don’t use a lot. So cross-sectional measurements are not going to give you the complete picture. If you look at a whole series of years, you can really get a much better picture. Let me go to the--
Lynda: And John--
John: Yeah?
Lynda: I mean, I think the important thing is that the CRG system is based on what’s in the claims system.
John: Exactly.
Lynda: And something like mental health, either if it’s a carve out and therefore outside of--
John: Exactly. Thank you.
Lynda: --managed care or some people may not, you know, access the system. So it’s only what’s in the system.
John: Yeah.
Ginny: Or provided by the schools or--
Lynda: Or it’s provided in the schools--
John: No. That’s really important. And this plan provides 12 visits for children with mental health so that you could--12 encounters or 12 consultations, not from the primary care doc but referrals to specialists.
Lynda: So it can only be picked up if the child with the mental health condition comes in for, let’s say, a preventative visit--
John: Exactly.
Lynda: --and their physical exam for the year, and the doctor’s been trained to say just don’t circle that preventative health, you know, code but also to circle the diagnosis that’s related to the mental health. But then you might have the parents, because of confidentiality reasons or just family reasons, not wanting that diagnosis.
John: That’s right.
Lynda: So it’s very specific to what’s in the data.
John: Exactly. Exactly. Next slide. This is a slide--we did some work that we presented at the Children’s Health Researchers Council on the charges. And I think this is where CRGs can really be very effective because they are designed originally as a risk adjustment system. And this shows you the percentage of patients as they are distributed, and the red is significant acute.
Basically, 10 percent of children have a chronic condition. They represent more than 30 percent of the encounters and around 50 percent of the charges. So 10 percent of the children--and that’s--you know, to me that you get back to, why is it important to a health plan? It’s because they’re big chargers. Now, I’m sure that children with special needs cost more than 50 percent of the dollars because these are just charges. It doesn’t represent out of pocket. It doesn’t represent what the states put into it. So that I think that when you look at this population, it’s generally accepted that that percentage of children that have special needs consumes probably up to 60 to 70 percent of the resources that are devoted for children. As you move up from the different CRG groups, and for an average cost for a child that was healthy, it’s less than $500 per child per year, which is really inexpensive. As you move up to the minor chronic, it becomes about six times as expensive; for the dominant, it’s 10 times as expensive; catastrophic and metastatic, you’re getting up to 20 times as expensive. And even though these are small, small numbers, the expense per individual child is just enormous.
The next slide. Now, one of the things that we’re really working on now, and this is hot, and I’ll end with this, is that we have done some fairly extensive chart reviews. And as a matter of fact, I’ve just completed looking at about 900, and there is what you get out is a quantitative and it’s a qualitative impression, and I’ll try to share that with you. But basically, what we have, we took a sample of 900 children, which is this group, which we did with FACCT, and that’s a super-sample. We took 500 that were randomly sampled and another 500 that was a stratified sample where we made sure we had a lot of children with chronic illnesses of a high level of severity. From that 900, we have gone through and asked this basic question: does this child have a physical, developmental, behavioral, or emotional condition, or is it at risk? Is it expected to last 12 months and then require related services? It’s the MCHB question.
Now, let me just tell you what charts we were able to do. Initially, because we did this comparing, both the QuICCC-R and the Children with Special Health Care Needs screener and the CRGs together, we chose out the 130 cases in which there was lack of agreement. And that was then 130 cases. We added to that. Went through all the charts--and it was very clear where you have literally four years of administrative data--`96, `97, `99, and 2000. You can, by looking at encounter claims, identify whether this child has a condition or not. Diabetics, children with spinal bifida and hydrocephalus, those are very clear. But then there were another 320 children that you really couldn’t be absolutely certain, so we chose those for chart reviews, and then we chose 50 randomly. So we came up with 500 charts for review from this administrative data and out of that we were able to find about 391. There were about 109 charts that were lost. And this is the results of this, comparing the QuICCC-R, the children with special health care screener, and the CRGs.
And before I do this, you’ll see that there are chart reviews that are done exactly the same way. You have to get a stratified sample to be able to do it. But when you do that, you build in a huge bias. And the bias is is that you’re choosing--the way to get a stratified sample is generally by using administrative data, and you go out and you say we want to find out people that have these encounters and select them. And that’s the way the HEDIS selection is done. So when you do that, you’re biasing it positively towards the administrative data system. Then when you go through and pick out the ambiguities, you generally don’t choose a large sample from the ones that have no health encounters; you’re choosing it from the ones that do. So you’re really moving this sample toward bias, toward the administrative data set. Until you begin to see it, you really begin to se that bias. That’s why it’s not surprising that in the sample the CRGs performed at a 75, 73, 76, you know, and 87 and 84 percent level. So don’t think that if you do a random sample of your population it’ll show that. It just won’t. Also you’ve got a significant number of false positives, which are probably mostly children that have a chronic condition but don’t require services, and that also will be heavily biased in this. That won’t be what you’ll see in a regular population. But despite that, there’s some interesting trends that, again, any physical condition here that is expected to last fairly much longer, the CRGs and the QuICCC-R performs better than the Children with Special Health Care Needs screener. When you’re dealing with developmental conditions, there’s just no doubt, even with all the biases built into it, the CRGs don’t do as well. And that’s shown in everything that we do. It’s just not a good way, at least currently, to identify developmental.
When you look at behavioral and emotional, it’s surprisingly how well the administrative system works there. And when you review the encounter data, you can see this. And again, this is dependent on what services that plan and the concern that the pediatricians--in this area, pediatricians work very closely with the school system, very highly concerned about the schools, put lots of information in the charts about children that were having emotional conditions. And as you can see here--and so that it performs pretty well and that would match with that data that I showed earlier in terms of it picks up maybe about a third of what you would expect of mental health conditions in the area. As far as the false positives, the thing that’s significant here is that the CHSCN screener performs better. So this is less sensitive in this area but probably more specific. So this is if you do your HEDIS screener, this is kind of the results that you would expect. Yeah, Ardis?
Ardis: Where do you chart reviews come from--those four pediatricians (inaudible)?
Jean: No.
John: No.
Jean: We actually visited 22 different sites to find the charts. Multiple--
John: Multiple sites. And that’s why we weren’t--that’s why there were 100 we weren’t able to find--you’re talking about a large rural area. We just couldn’t go to the small, small sites.
But there’s so much that I learned from this, and this is still preliminary data. This was done blindly, and there’ll be another pediatrician that will go through it and match it for concordance. I’ve become very, very much convinced that the at-risk group is a terribly important group to go after. And I’m so terribly convinced that mental health is, just as you would expect from the review of the literature, is way up there as the major cause of chronic conditions in children. And you’ll see--I mean, it’s almost heart rendering when see four years of administrative--five years of administrative data, or covering five years, you’ll see a preadolescent girl showing signs of depression, and they’ll be treating her or doing whatever it is. Then shortly after that, you’ll show signs of conflict and conduct disorders. And then shortly after that, you’ll begin to see orderings for contraceptives, and then pretty soon after that you’ll see several abortions coming up. I mean, or you can see patterns that go on that are really unbelievable and you have a feeling of oh, my heavens. If you can develop the right interventions for some of these children at an earlier period, you may well make a difference on down the line. If any of you ever have time to do this, to go through four years of administrative data on 500 kids, it’s very, very eye opening.
So, this is the publications. We have two that have come out in "Ambulatory Pediatrics." We have two more at least that will be in--one that’s ready for submission and at least two or three others that we’ll be able to develop for publication. We’ve presented at several national audiences and that we’ve presented earlier in this project at the tri-regional meetings.
And then these are the implications: that CRGs--of what we’ve learned--CRGs perform well at identifying children with special needs who interact regularly with a medical care system. I’m absolutely convinced that there’s no one system that can identify all children with special health care needs and you look at them for different ways. So, you use different systems. Don’t get tied into one particular system as the system you’re going to use. Try to figure out ways of combining them in different ways.
The other thing that is so remarkable when you review these charts that there is no clear line that you can say, "Now this person needs added services and now they don’t." Each condition has its broad spectrum, and it’s a bit arbitrary when you really choose that time to say, "Now this person meets the MCHB definition and now they don’t." So it’s a very broad spectrum.
I think that mental health conditions are a major, major issue, and we have to figure out ways of identifying them better. And then I also believe very strongly that the at-risk population has received relatively little attention and it warrants a lot of attention because that’s where we might be able to make a difference on down the line. And I want to thank my team and thank everybody else’s that’s worked. I want to thank Lynda, and it’s a pleasure to work with all of you. Yes?
Tom: Was your age population birth through 18?
John: Yes.
Lynda: So does that mean that for the youth--the 18 to 22 is--
Tom: In your state, the Title V program goes was 18? Is that not a factor?
John: Right. Yeah, right. I can never remember whether we included 18 or we’re just below it. I think we we’re just below it.
Ginny: It was less--17.9.
John: Just below it. Less than 18. Yeah, through 17. Oh, that’s the other thing, by the way. If you do surveys or chart surveys, look at the age group that you’re identifying because if you go from 12 to 17, you’re going to pick up most of the mental health. Below, it’s going to be other conditions.
Lynda: Let me also throw out in terms of implications for--at the national level, that one of the things that the CMS, which is Center for Medicare and Medicaid Services, The new name for HCFA, is--you know, in years past we’ve talked about this set of criteria that CMS uses to evaluate waivers and when children with special health care needs are enrolled into managed care, and under the new administration, they decided to look at that criteria, and actually, they’re really going to make some significant changes to it. And what they did is--and I was actually part of the review team--they went out and talked to several Medicaid agencies and some Medicaid plans. They did not do as broad a review as I would have liked, which would have included, obviously, family organizations and providers. But I think they went fishing, and they got the fish they wanted, which was that Medicaid agencies are having tremendous difficulty with defining this population and identifying them. And I think one of the issues--and, you know, I think you’ve really said it very well is that it’s really difficult, I think, to identify this population only using one system. And part of the problem with the way the criteria was written was it was forcing Medicaid agencies to identify kids based on these artificial--they had these artificial groups, like they had SSI, and they had five different categories, and CMS kept saying, "We want to know exactly how many kids are in each group." And it’s difficult to do at that broad, what I consider Medicaid at the policy level, where it’s something that should be done more at the operational level. So this is going to have, I think, huge implications for kids who are enrolled in Medicaid plans in terms of what protections are going to be offered by their Medicaid agencies based on how the states are going to end up defining this population, and I think it’s going to be very narrow.
Ginny: And actually, Lynda, one of the things we’ve been talking to the 3M developers about is, could we at in-take into a Medicaid system, have a questionnaire to identify these conditions and just force those into the CRG analysis and run it on the in-take information and classify those kids then so they immediately be identified as child with special needs needing the care coordination services, and they, [said] "Oh, yeah. Sure. That’d work."
Lynda: And the other issue that keeps coming up--and this happens not only with Medicaid but I think on the private side--is, you know, there’s a lot of numbers being thrown out there in terms of who is this populations. And you know, we go from 15 to 18 percent, 12 to 19. Whatever it is, it’s more than what the health plans think it should be. Whatever number they think it should be, it’s more than that. And what CMS is saying is that the plans are saying, "Look, we can’t provide case management, and we can’t provide this level of services if you’re talking about 15 percent of our enrolled population." And I think as we go through the discussions, we should really be thinking about the fact that we’re not asking you to do that. And, you know, I mean, of that 15 percent, not everybody needs case management. Not everyone needs these advanced services. Not everybody costs that much. It’s not that all 15 percent are all the same.
John: That’s right.
Lynda: And somehow that message, I think, is not getting out there strong enough--
John: And you’re absolutely right.
Lynda: --to the Medicaid agencies and certainly to CMS--
John: That’s absolutely right.
Lynda: --although I’ve been pretty loud about it.
John: No. That’s absolutely right.
Lynda: I think they’re starting to get it. We actually had, I think, a victory on Friday where they finally agreed to include the MCHB definition in their list of definitions.
John: That’s great.
Lynda: So, it still remains to be seen what’s going to happen. I can talk about it more as we go throughout the day.
John: And then with each individual condition group there, I mean, mental health is going to depend on a different case management than asthmatics and developmentally disabled. You’re right. And the other thing is, what I’ve learned out about this is because the spectrum is so imprecise in terms of measuring, avoid things like who our children represents 18.53 percent of the population.
Lynda: Right.
John: I mean, it doesn’t. It’s a range. It’s a big range. And parents are going to have one feeling the way they feel and the medical system is going to feel another feeling. Anyway, thank you very much, Lynda.
Lynda: Anybody else have any questions or comments? Faye?
Faye: Since we’re talking about identification of children with special health needs within the Medicaid paradigm and we’re all aware of changes in buffer reform that are creating more requirements for families to be in the workforce and fewer exemptions, what is being done so that the people who are administering the welfare to work programs are cognizant of children with special needs and the fact that the family involvement in their case management, their daily care, is in fact a form of work, and children need to be able to comply with their medical regime and their therapies and everything and that is only available during the business hours.
I know that a small amount of work has been done and that some states are more advanced than other states. I know that in Wisconsin their DD Council has a manual for families to document the daily care and the medical care of the child. Where, then again, back to Illinois, which is not a good example, there’s, you know, a 20 percent exemption and most of that is going to parents who have special needs, you know, substance abuse and mental health needs.
So I’d like to know, can this be used and is it going to be used in a proactive way to enable families to support their kids?
Yolanda: In Texas they passed legislation that exempted families that were under welfare reform work requirements. The problem that we’ve discovered is that it’s never been used. The policy exists, but people aren’t aware of it, so it’s never been used since the welfare reform legislation was passed.
Lynda: I think it’s a good comment. I mean, certainly--I mean, I think you know that I feel very strongly that the work of families is largely unknown, and I think it’s something that, you know, we need to work on.
John: I think we need to require a substantial change in policy in states to find this as acceptable work in the welfare reform and that given the new current focuses on welfare reform is going to be substantially harder for families. This sounds like an important policy study.
Faye: It is--it’s kind of--unrecognized. I think many of you know I also, you know, have another hat as a volunteer for Family Voices and you know pretty often we will get calls from case workers in the department of human services. And, you know, we’re sending subsidizing because childcare is an oxymoron, and they say "Look, I have this lady in my office and she said her child is 10 and he has Downs Syndrome and he has asthma, and he’s on oxygen." I said, "Listen, you know, I don’t want to hear this. You have to put your kid in day care," but there isn’t any place for her child to go for day care, and the reimbursement that they’re offering would not pay for a skilled provider to come to her home. And they said, you know, your clock is done. You know, you have to go to work. So it’s real frustrating. I think we need to look at that.
John: Well, we worked on that somewhere for the state of Washington. And there’s a group that sees and understands that there ought to be exemptions and there ought to be special services. It gets right back to helping us identify who those kids are which puts the onus back on us to be able to come up with some system to be able to say these are the areas where you can put in those kinds of services.
Tom: --And it’s state-by-state, right? The welfare reform movement is happening--
John: State by state.
Tom: --it's a very broad parameters from the federal.
Lynda: I think there are terms of it that--the division where we’ve got a multifaceted--in terms of trying to do things, Faye, at the national level. But there’s a limit to what we can do at the national level and then trying to work, you know, at the state level. And you know, I think this, you know, has implications also in terms of, you know, people are employed and having the support that they need to maintain that and maintain their health insurance with that. So, it’s both sides of the health. People who want to go to work and having, you know, the childcare and the provisions or, you know, people being forced to work when they don’t have the adequate support. Bobby?
Bobby: I think that, you know, another thing that’s important to recognize is that, like in Wisconsin, we have some fairly decent policies on the books but how are those policies interpreted locally? And what’s the enforcement mechanism? And, you know, I mean, the best one is the education and bringing the stakeholders together to help them meet and understand what the issues are. But you also have to have an opportunity to lay down the law a little bit sometimes and make sure that the policies are being enforced correctly. In situations in Wisconsin where we’ve had to intervene legally on behalf of families where they have a very medically fragile child that they’re taking care of in the home and they’re being punished because they’re a good parent; they’re active and involved. And their hours are cut because they’re available. And not taking a look at any of the other factors within the household of other children, other activities, jobs, things like that, that slides because of poor interpretations of the policies at the state and local level and needing to be reminded that they can’t do that.
Lynda: OK.
Faye: I just wanted to, you know, remind everybody that there’s a list of our web casts archived in the back, and one of them that we did on Medicaid and children with special health needs and interviewed a mother who was a TANF recipient, whose son is a teenager now who has cerebral palsy, autism, and very severe asthma and how she had sought out a pro bono legal group to assist her and she finally got her clock stopped. But you know, talking about what had happened along the way, several times that her child’s Medicaid card was revoked and she had to go back and what she’s been through.
Lynda: And I would say I get a lot of emails from parents. Somehow the word is out that Lynda Honberg in Rockville can provide health insurance... Please squash those rumors, because it’s heartbreaking for me--
John: We haven’t even started it!
Audience Member: Can we link to you off our web site?
Lynda: But one of the big issues is, you know, how can I go back to work? I’ve got a child who--you know, I can’t find childcare. You know, and actually the FICC is doing a--I think it’s the middle of June? The Federal Interagency Coordinating Council is doing a special session on childcare for children with chronic conditions because it is such a big issue.
Audience Member: I’m wondering--have a question question whether there is such an important opportunity here on the both national and state level between Family Voices and the kind of welfare rights advocacy movement to make a connecting point there. I don’t know if there’s anybody here from Family Voices, but I think what you’ve identified is a really potential collaboration between the kind of advocacy and effectiveness that Family Voices has achieved with respect to identifying those needs and lending that strength to the welfare groups in various states. It would be an important connection to try and facilitate I think.
Lynda: Faye, are you the only one from Family Voices? Oh, that’s right.
Audience Member: Has there been much conversation about that financial level with Family Voices in terms of that connection?
Audience Member: I’m only in touch with the state level really.
Faye: I have brought it up over the years.
Audience Member: Yeah.
Audience Member: Yeah, I’ll talk with the--
Audience Member: I think it’s an important leadership opportunity. .
Lynda: I’m going to suggest we--you know, again, this agenda is just a template that we’re going to go through the day, and I’m going to suggest that we take a 10 minute break until 10:00 and then we’ll start with our next panel which is Jim Perrin and Yolanda Cantu. Okay? So come back at 10:00. Thank you. Great discussion.
###
End of Transcript