Room 440D
Hubert H. Humphrey Building
200
Independence Avenue, S.W.
Washington, D.C. 20201
Lisa I. Iezzoni, M.D., M.S., Chair
Kathryn L. Coltin, M.P.H.
Paul
Newacheck, Dr.P.H.
Barbara Starfield, M.D.
David T. Takeuchi,
Ph.D.
M. Elizabeth Ward, M.N.
Carolyn Rimes
Olivia Carter-Pokras, Ph.D.
Gerry E. Hendershot,
Ph.D.
Dale C. Hitchcock
Michael L. Millman, Ph.D., M.P.H.
Beatrice
Rouse, Ph.D.
James P. Summe
Opening Remarks and Introduction - Dr. Iezzoni
Status of GW Project - Ms. Rosenbaum, Ms. Markus
Work Group on Quality Charge and Work Plan
DR. IEZZONI: Why don't we get started, because Sara and Anne have graciously waited for us, and I want to make sure that we don't keep them too much longer than they were scheduled to be here for.
So this is the Subcommittee on Populations. I'm Lisa Iezzoni from Beth Israel Deaconess Medical Center in Boston. I'm going to chair this afternoon's session.
I would like to just go around the room, so everybody knows who is in the room. People can introduce themselves, say their affiliation, and we'll get started with the agenda, which hopefully, everybody has.
[Introductions were made.]
DR. IEZZONI: Sara and Anne, why don't you orchestra your presentation for us. And because there are a number of new people to our subcommittee, if you could just briefly review what you are doing, and the concept.
MS. ROSENBAUM: For the past four years now or so, the Center for Health Policy Research has spent a considerable amount of its time on a series of projects related Medicaid managed care, and particularly to the legal framework in which Medicaid managed care is operating. As with any enormous systemic change such as managed care, the implications for every way in which we frame society are huge, and the changes for the law are no less small.
Actually, it was certain major legal developments, most notably ERISA and the application of the Sherman Act to doctors -- the Sherman Act of course prohibits restraints on trade, and promotes market competition -- it was those two laws that triggered this revolution to begin with about 25 years ago now. Everybody is feeling the effects of it. As with a lot of things, the law not only triggered this revolution, but then has been much changed in response.
One of the biggest changes that occurred had to do with the way in which we frame the legal environment in which this new health system operates. ERISA and the Sherman Act had the effect of deregulating most of health care, because it allowed systems to grow outside of the regulatory framework in which we had operated the fee-for-service system all these years.
Now what we are engaged in of course, whether it is the state government or at the federal level is sort of an epic struggle to get a new legal framework functioning for managed care again. This is exactly what we went through in our service system.
For the time being, and again, this is very much part of the normal evolution of legal frameworks, most of the framework that governs the system -- what it covers, how it operates, the rules it plays by, what its contents are -- is found in the large contracts between the buyers and the sellers of the benefits, which in this case is what we call a managed care product.
What we set out to do several years ago was to study these large purchasing agreements. It happens that in the case of Medicaid, and to a lesser extent, certain other public agency contracts -- DOD, DA, HCFA -- the contracts are public documents, and the procurement is a public procurement process. It has to be, because it is public funds.
So we used Medicaid as the base of study, but many of the things that you see in our studies would be issues no matter who the buyer is or who the seller is. We have now produced two rounds of study findings. The study, which has been undertaken with considerable time and funding support has been financed in great part by the Centers for Disease Control and Prevention, the Substance Abuse and Mental Health Services Administration, the Health Resources and Services Administration, and a number of private funders.
In the first year or so of the study we produced all of these results, and showed people how these contracts are structured, and what they cover, and what they do. And particularly in Medicaid, the tension between what managed care systems buy and sell, and what the federal Medicaid program requires -- there is definitely a challenge in getting managed care to work in Medicaid, because Medicaid is not typical commercial insurance. So this has been, I think a bigger issue than even managed care purchasing generally.
One of the earliest things we were asked about when we started producing these vast tables with thousands of entries showing how different states were making different purchasing decisions, and what decisions they memorialized in their contracts was, well, okay, there is all this language. What is good language? Help us write a contract that's good.
We said, well, we really can't do that, because there are certain provisions that are clearer than others. If you want to use "good" in the technical, legal sense. A provision that unequivocally states what the director wants to achieve is better than one that is filled with ambiguities.
But there is no good or bad. Part of it is we don't know enough about these systems to know at this point what we even mean we saw what's good or bad. Number two, what is good or bad -- this is a memorialization of a huge set of policy decisions. So it's going to be as good or bad as it is an approximation of your policy choices. I might make one set; you might make another set. We can show you different policy choices underlying different provisions.
What we did feel we could do though, was with the help of a lot of experts, draft a series of what we call sample purchasing specifications. They are not meant to function as a guide to what is the correct way to put a contract together. What they are meant to do is to give people sample language that is well drafted, of one possible policy course, and sometimes alternative courses.
On one provision we may draft two or three alternative sets of language. Any provision is of course, an optional provision. These things are privately done. They don't have any force and effect of law. And they are meant to get people to have some template to work off in putting an agreement together, or putting a procurement together. And they are also meant to help a purchaser think about the decisions that the purchaser ought to make.
So right now we have sample specifications in about 18 different issue areas, spanning many, many of the key areas of the CDC, communicable diseases, specifically, tuberculosis, HIV, sexually transmitted diseases, and vaccine preventable diseases.
DR. IEZZONI: I'm passing around the ones you prepared for us.
MS. ROSENBAUM: Yes. We have specifications on a number of the chronic illness areas of the CDC. We have specifications underway for HRSA on issues of access, pediatrics, reproductive health in conjunction with the CDC. We have specifications covering, or we are about to start working on specifications in the area of persons with mental illness and addiction disorders. Altogether, there are about 18 different units.
It has been a very long process. Before we turn to our findings, and a walk through on this set of specifications, it is worth talking about the process a little bit, because my job as your drafting counsel basically is to give you some options. Now you are, for all purposes, my clients. What I have done is a first pass through at an effort, and now you all need to sort of think about how you want to proceed from here.
Generally, what happens in these specifications is that a group of people who are knowledgeable in an area come together for purposes of getting the ball rolling. We are lawyers. While Anne has added qualifications that I don't have -- she is also getting a Ph.D. in health policy -- so she knows things that I don't know about research methods, but for this kind of work we are basically a tabula rasa.
I know a fair amount about health care financing from my work as a lawyer, and I know some subpopulation diagnostic issues, but I don't have any expertise in a substantive area of health care particularly. But we trying and bring together responsible people who do have that expertise. So when we did immunization specifications, we began with two experts from the CDC in immunizations. With them, we built a draft specification covering the major domains that get addressed in the managed care contract -- enrollment, benefits, the manner in which services get delivered, how quality is to be measured.
Once we got the specification in sort of a working format for the client, we then went through a vetting process. The specification was vetted within the CDC. It was then taken to a group of outside experts consisting of people from state Medicaid programs, health plans, consumers, public health people, a range of people, and they went at it for a while.
Mostly what we stressed at each stage of the vetting is don't forget, this is a wonderful, liberating process, because there is no right or wrong answer. You don't have to have a consensus here. We don't know enough to know what a consensus looks like. What we are trying to do is get good legal language down for what you all think is a realistic range of responses to a particular issue.
So to give an example from immunizations, on the issue of networks for example, there are some people who believe strongly that a child ought to be able to go get an immunization from wherever, and that if that child gets an immunization from a health department, and that was the most accessible and useable spot for that family, the health department ought to be paid as if were a network provider.
There are some people who think that that's very wrong. That in fact one of the great advantages of managed care is that you would rationalize care, and have the child getting care from one provider. So when you look at the sample specification, which is now on the website, under networks there are two or three alternatives, each drafted hopefully equally well, each giving essentially a version of an approach to that issue.
So that we can hopefully get people maybe seeing some of the policy issues they have to decide, seeing good language under the various approaches, and perhaps having a greater discussion then among themselves about what kind of immunization policy they want to foster.
We are in the process now of completing the unit on communicable diseases, and as I say, there is a huge unit on pediatrics, basic pediatric, special needs; and then other units to follow.
Now one of the things that came up in all of this, there are several very cross-cutting issues. One had to do with memoranda of agreement between managed care companies and other agencies that are absolute stakeholders in managed care, but aren't the purchaser. Specifically, health agencies -- the public health units, the MCH part of an agency. They are all profoundly affected by managed care, and there are certain ways in which through an MOU process, you can foster a relationship. So we are drafting a unit on MOUs.
DR. IEZZONI: That's memo of understanding.
MS. ROSENBAUM: Memorandum of understanding, memoranda of agreements are called MOAs. We call them MOUs. It is designed to be a generic specification that a state could start with, or a private purchaser could start with, or HCFA could start with in writing the Medicaid contract to say, in your service area we want you to have the following working relationships with the local health departments in your areas. Go sit down, and negotiate them through.
Another big cross cutting issue was data. It doesn't matter what the subpopulation is, what the diagnosis is, what the service benefit is, people were very concerned about what information would be available: (a) to the purchaser; and (b) to the stakeholders in the process.
I don't need to tell you all the issues that come up with data availability, but we were asked by this subcommittee to tackle this issue for you all. What we did was tackle it in basically three pieces. The first piece was a background analysis for you of the data collection requirements in the law.
Not surprisingly of course, there are very, very few specific data collection requirements in the law. You can infer certain data collection requirements. The department may well have the authority to add addition data collection requirements, and I'm sure that's a huge issue here now, throughout the department. But there are very few places where Congress has said to states under Medicaid, you will provide the following data in the following format, even in accordance with specifications set up by the secretary.
The second step was the assessment that was done by Anne, which was to look at what was happening in selected states around data collection, and that is what Anne is going to present to you; the results of this assessment.
DR. IEZZONI: Sara, have folks already seen the first report that came out?
MS. ROSENBAUM: I don't know the answer to that.
DR. IEZZONI: Presumably you all have seen that. It was an excellent report. Has everybody seen that?
MS. ROSENBAUM: So the second step was the assessment of what selected states are doing. In the end, we'll talk about the findings from that.
The third step was to take the results of Anne's and Lee's work, and start the ball rolling. Now I have not done this the way we normally do this. Normally, we would have had a couple of being the clients, back and forth, back and forth. But given the nature of this particular client, that you are an advisory group that meets infrequently, we decided we would just take a crack at getting going with the document, and talk to you, which I will do when I take you through the document, about how you want to proceed from here, and what we would suggest to you about this document, and how to get it reviewed and vetting in a way that would be most helpful for the subcommittee.
Maybe the best thing would be to let Anne take you through the findings. If you have any questions about the overall project?
DR. STARFIELD: If I could just a question about this?
DR. IEZZONI: That's what we are going to be talking about after Anne presents.
DR. STARFIELD: I just want to ask a question about what it is. We've got some paragraphs from the law, is that what that is?
MS. ROSENBAUM: This is all the law.
DR. STARFIELD: Even this stuff in the back is in the law too? The three categories, Part I, that's from the law?
MS. ROSENBAUM: No. This is as if I drafted a piece of a contract. I'm going to take you through this, walk you through, and show you what's here. There are some pieces that are absolutely driven by legal requirements, and some that are quite honestly a policy judgment call. That's what I'll show you.
DR. STARFIELD: The reason I am driven to ask that question is because this committee has put out core data elements. It seems to me that if you were going to look at the extent to which managed care or anyone else that you are interested in was providing data, you would at least start with core data elements.
MS. ROSENBAUM: One of the things that we should do when I walk you through this is figure out. We have fashioned this based on what states appear to be asking for, and then we pushed a little in areas that have come to our attention from the answers we got to this. But there may be whole other ways to come at this project that you might want to have us do. We can draft from any originating document at all. The core data elements are not unlike the tuberculosis specifications that we used to draft the tuberculosis unit.
MS. MARKUS: I'm going to present some of the preliminary findings that we collected from a number of interviews we conducted over the summer for the subcommittee. I will be pulling from some of the previous report we referred to earlier, the one on federal law requirements, as well as those that were found in the state contracts to the extent that was done.
I will first be talking briefly about the study described, the purpose the study, and the methodology used. Then I will move on to describing some of the major findings we have from the study in three main areas. The first one is how states actually make decisions about which things that want to have reported to the state.
The second one is which types of data they are actually collecting in the states, and they are requiring in their contracts. I will be focusing on three main areas, which are encounter data, access, and quality data, and with regard to these three areas, I will be referring back to some of the federal and contractual requirements, to show how what is required by the federal government of the contracts.
Finally, I will be talking a little bit about public health data of interest to the subcommittee, and how the contracts are dealing with the whole issue.
Finally, I want to just lay out some of the recommendations that came out through the interviews as we were talking to states and, that might be of interest to the committee.
The objectives of the study were three-fold. First, we wanted to describe how state Medicaid agencies decide which data they organizations to collect and import to the state. We wanted to know what the basis for their choices were. The second objective was to describe the types of data that managed care organizations submitted to the state as the result of these choices. The third objective was to access or describe states' capacity to analyze the data in a way that would inform the purchasing process.
As you can see, I have underlined full risk managed care organizations in this slide, and this is to underscore the fact that we did specifically focus on those requirements that are applicable to managed care organizations that contract on a prepaid capitated basis with the states to deliver a comprehensive set of services, which is being defined as three or more services. We are focusing on needs for this project, particularly because most states are now enrolling their Medicaid populations into these arrangements.
PARTICIPANT: What do you mean by three or more services?
MS. MARKUS: Inpatient care, physician services, labs.
MS. ROSENBAUM: Under the law, you are not a managed care organization unless you're doing inpatient and outpatient.
MS. MARKUS: Just last year, there were about 10 million Medicaid enrollees in pure risk managed care. That represents about two-thirds of the total Medicaid managed care enrollment. That is about one-third of the total Medicaid population. So this is the amount of lives we are talking about.
You will notice that I have underlined submitted, because we were focusing specifically on data submission duty. That is, explicit duties in the law for managed care plans to submit information to the state or the federal government for that matter.
The methodology we use is the qualitative design. The type of design we used was semi-structured interviews with state officials. We obviously got information about the managed care contracting process. As I said, when appropriate, we did review some of the federal and contractual language to complement the findings we got from the interviews.
We aggregated the results across the states, and looked for patterns, and saw if there was any similar responses across states that would be of interest to the various areas we are looking at.
In terms of the states we selected, we had a total number of 11 states. Two of those states were pilot states in which we piloted an interview guide, and corrected the questions in accordance with some of the problems we had when we asked the questions. The two pilot states were Arizona and Massachusetts, and I believe these were selected because the subcommittee had done previous site visits in those two states.
Then we selected nine additional states based on two main selection criteria, which you can actually see in the following slide, which was all the states we actually interviewed, except for Rhode Island, because we were still waiting to interview that state this week actually.
We wanted the states that were not only enrolling a high proportion of their own Medicaid population in full risk managed care, but we wanted also a selection of states that combined, would present a high proportion of the total core risk managed care enrollment at the national level. As you can see here, we had a total of about 44 percent of the people enrolled in the full risk Medicaid managed care.
We did face about three refusals, or we had trouble contacting the states for several reasons. Some of them didn't have the time to actually do the interview. Others had to pool resources from various departments, and had trouble coordinating people to get back to us. So as we conducted the research, we dropped three, and replaced those with three additional ones.
MS. MOIEN: And the three you dropped?
MS. MARKUS: New Jersey, Hawaii, and California.
DR. NEWACHECK: Can you tell us a little bit about whether you think these states are fairly representative of what we might find out there in the world of Medicaid contracts? I know they are not necessarily too representative of the nation, but do you think they cover the sort of gamut of differences?
MS. MARKUS: They probably do to a certain extent in terms of contract language used, and the extent to which the provisions related to data in the contract, yes.
MS. ROSENBAUM: Interestingly, in this group you probably have at least from a technical point of view, the finest contract represented, and the worst contract represented. It's interesting. It is quite a range, if you think about what makes good or bad being the amount of policy thinking that went into the contract, which you can see from the contract provisions themselves, whether the answers are answers you would choose. The point is, they addressed a lot of issues, versus a contract that addresses very few, and contracts that are literally drafted poorly or drafted well.
PARTICIPANT: These 11 states have 44 percent of the nation's Medicaid.
MR. HITCHCOCK: California would have probably what?
MS. MARKUS: About 20 percent. As you can see, Tennessee alone is about 20 percent.
The next slide shows the states we have included in the study in terms of the types of Medicaid managed care we really have, because it has some implications for the types of federal requirements that might be applicable to them. As you can see, about four states -- Connecticut, Utah, Washington, Wisconsin -- have what we call freedom of choice waivers.
These are based on Section 1915-B provision in the Medicaid statute, which basically allows the state to grant a waiver of the freedom of choice provision. So that states can enroll their Medicaid population into managed care on a mandatory basis, however, all the other requirements in the Medicaid statute still apply. So for example, the federal and state regulations still apply to these 1915-B waiver programs.
Most of the states, however, had Section 1115 research and demonstration waiver program implemented when we interviewed them. This provision is not part of the Medicaid statute per se by the Social Security Act, but it allows the secretary to waive requirements of the act, including I think the Medicaid statute, as long as it furthers the objectives of the program.
So states can waive the freedom of choice provision as well, to enroll people into managed care on a mandatory, but they can also ask to waive other provisions of the act.
MS. ROSENBAUM: Benefits, eligibility, and for purposes of this, the qualifications of the managed care companies themselves. These states can run MCOs that look somewhat different from what you could in 1915-B state. Now some of this has rendered slightly moot by the new balanced budget act. Once the regulations are issued, we'll be able to see the extent to which if you went the new state option route versus one of these wavier routes, it might change the rules of the game.
But basically most states are -- to the extent that this is not a representative group of states, it is not representative in the sense that there are a lot of states here that technically have the authority to change the functional rules for managed care organizations.
But the secretary has been quite restrictive in: (a) what she is willing to waive about those functional rules, like a lot of states may ask for a waiver of an external review requirement, but that doesn't get waived; (b) and the secretary has separately imposed certain obligations on states that are data-driven obligations in the 1115 states. So it cuts both ways.
In all honesty in terms of grading the contracts, while it is important to know what states we interviewed, if you look at the contracts, you don't find big differences by whether it's an 1115 demonstration state, or a 1915-B waiver state.
MS. MARKUS: I was actually going to mention the newer state option, something that I will be referring to later, the section on 232 state options, because the secretary can impose additional requirements as part of the approval part.
Moving onto the first set of findings of how states actually make decisions about which types of data they will include in their managed care contracts. We asked questions about how they already make decisions to inform process; how the state characterized that process; who made the decisions; et cetera, et cetera.
And we found that the majority of the states -- 7 states in our study -- did characterize their process as a somewhat formal process for which they made decisions concerning data, however, three of them, half of them, said that it's not a process of specific data, but it is part of the general purchasing process they use for contracting of which data are a part.
In terms of the decision makers, what was interesting actually is that those states that responded that they had some kind of a formal process -- what they characterized as a formal process -- could pinpoint who the decision makers were actually. It went from one individual to a group of individuals, with the final approval of the Medicaid director. But generally speaking, it was either the Medicaid director, the managed care program director in consultation with the Medicaid director, or with other program directors within the agency.
Most states relied on sort of the team concept to make decisions. Although the exact composition of the group varied depending on the data issue at hand, they would have two basically structures. Basically, either the Medicaid director was part of the meeting, and making the decisions with the team, or the team would make a decision on its own, and then refer the decision back to the Medicaid director for approval.
Just to mention on those states that had an informal process, as you can see, that was pretty big; a group of people, anyone involved in managed care.
We asked states if, when they were determining which data they wanted MCOs to report back to them, whether they actually involved MCOs in their discussions. Interestingly, most states do, and that's in the pre-contract phase, before they actually sign a contract with the managed care organization.
Those states that didn't, mentioned the fact that that's because they have a competitive approach to procurement, so there was an obvious conflict of interest there, that they cannot talk to managed care organizations about what they want. They basically set the requirements, and contractors have to bid on those requirements.
MR. HITCHCOCK: So the other seven don't have that?
MS. MARKUS: They might have, but those that didn't cite it as a reason that that was in the procurement process.
MS. ROSENBAUM: A competitive procurement is still, I would say -- I don't know what the HCFA folks here would say -- but my sense is from my own work, it is not the rule, because the legal stakes go up so high in a competitive procurement. If you have any willing plan situation, you can be much more open in how you are setting the standards. You can be much more interactive in how you develop the criteria you use. You can do a lot more concentration with your providers, precisely because you are not screening providers out.
And so in some ways the discussion flow may be better. In other ways of course, the plans may have a greater opportunity to press for limitations of data that otherwise someone might want to collect. But the downside of that may be in fact less than the downside of taking on a competitive procurement.
MS. MARKUS: When we asked states that involved managed care organizations in their discussions about data what their experience had been negotiating with those managed care organizations, the answers varied pretty much from receptive, collaborative, basically good relationships, to a mixed experience, to an initially negative experience.
Those states that said that they had a mixed experience mentioned reasons such as the fact that they had a different leverage or a different power to impose their conditions on MCOs, depending on whether they were Medicaid only MCOs, or whether they enrolled both Medicaid and commercial enrollees.
Another reason was the encounter data part, which is always problematic. They had trouble negotiating those data. There was one state that said it was actually initially negative. They basically said they were overwhelmed, but as things progressed, and they gained more experience with the implementation part of the contract, things improved.
Interestingly, regardless of the approaching to contracting, most of these states do involve the managed care organizations in the post-contract phase, where they actually have ongoing meetings with the managed care organizations to talk about their concerns, and talk about areas for improvement. Actually, it was mentioned as one key factor for a successful implementation of a managed care program.
MS. ROSENBAUM: One strange thing about managed care, which is very, very different from fee-for-service Medicaid, which definitely has huge implications for what this committee is interested in, as well as other areas, is that under the law while a provider in a Medicaid program was seen as a provider in a Medicaid program, a managed care company is not. It is viewed as an agent of the state, or as an agent of HCFA.
As a result, anything the managed care company does is imputed to the state. Which means that if the managed care company does something bad, it's the state's problem from the beneficiary's point of view. It's not the company's problem per se. The state and the company may have words later, but it is the state's problem.
So while most states tend to see managed care companies as providers from whom they purchased services at an arm's length, which they certainly are, the courts would say, oh, no, they're simply an extension of the state itself. Which argues for sort of a rethinking of how you approach the development of the working agreement with a company, because you have to develop some sort of a hybrid relationship that will let you get the best out of the agency without giving up anything else, and that's not easy to do.
In the area of data, which is a very burdensome issue, or can be certainly, you may see some of these challenges; new relationships having the biggest challenges to work through.
MS. MARKUS: We also asked the states we studied if they actually involved other formal stakeholders in their discussions, and not surprisingly those involved that involved managed care organizations, also tend to involve other participants in the process. Which one actually depends on the type of data they are talking about. The examples of stakeholders were anywhere from state legislature to consumers to providers to the fiscal contractors.
DR. NEWACHECK: Did any involve other state health agencies like maternal and child health programs?
MS. MARKUS: Well, not specifically. Some states did refer to discussion with other departments, agencies, things like that.
MS. ROSENBAUM: One of the most notable parts about what we found was the absolute absence of health agencies.
MR. HITCHCOCK: Behavioral health agencies?
MS. ROSENBAUM: Health agencies -- anybody under the umbrella of people who were concerned about populations, as opposed to enrolled beneficiaries. This is a big theme here, and you will see it in the specifications.
DR. NEWACHECK: It doesn't sound surprising based on my knowledge of the other side of the coin.
DR. TAKEUCHI: Anne, did anyone on their own raise the issue of concerns for either a growing minority population, or minority composition in their states, and say we're interested in multiculturalism?
MS. MARKUS: I don't believe so, no.
MR. HITCHCOCK: We heard that they are interested only a little.
DR. IEZZONI: In Boston we did. We heard it more from the research community. We had a variety of different types of people talking to us.
MR. HITCHCOCK: We also asked leading questions too.
DR. IEZZONI: I would tend to agree.
MS. ROSENBAUM: What they're thinking about, and Anne will show you, is the financial and operational side of this huge business venture here. If someone is going to draw their attention to the human implications of that in the context of data, it is going to be as the result of a more concerted effort.
DR. IEZZONI: In fact, one of the things that we found in our site visits, and that I'm finding using data from Harvard Pilgrim and Blue Cross right now in Massachusetts is that the managed care companies don't even collect race information; information on race and ethnicity. They don't. It's a big issue for investigators who want to look at historic patterns of underservice. But from their point of view, it's not a business need of their organization.
MS. ROSENBAUM: It's not, and nor should they probably be in the enrollee database. That's a whole separate problem. But the question is, on something like that, should a provider -- there you have a huge policy issue, which we can come back to when we get to the specifications. Should a provider ever be the point at which you collect certain kinds of data?
DR. TAKEUCHI: Some managed care companies in California will raise the issue not necessarily of collecting race or ethnic data, but they will say, gee, we have to recruit more from the Asian community, because they are low utilizers. They will come in, but they won't use a lot of the services.
MS. MARKUS: When we asked our states what were the reasons for collecting the data, the three main reasons as listed on the next slide were compliance with federal requirements, which is not a surprise; compliance with the state requirements, or contract specifications; and measurement of quality. Half the states also mentioned performance measurement and rate setting and risk adjustment. One state did mention national public health goals.
DR. IEZZONI: Can you tell us who they were?
MS. MARKUS: Wisconsin.
MS. ROSENBAUM: They have a lot of reporting around national public health goals too in Wisconsin.
DR. IEZZONI: As a basis of policy when you write this report up, Anne, will you be identifying who --
MS. MARKUS: No.
DR. IEZZONI: You kind of promised confidentiality to these places? Okay, that's interesting.
MS. MARKUS: We'll list the states as I've shown.
MS. ROSENBAUM: In the contract study the states are named. So you can take a look at that.
DR. IEZZONI: All right, well, that's helpful to know.
MS. MARKUS: In terms of selection criteria to select specific types of elements of data that they want selected, again, federal requirements come first. And the general direction of HEDIS, which is not too surprising, and finally, the national standard format, HCFA 1500 and UB93. Additional selection criteria were also mentioned, such as added financial burden for both the administrative agency and the plan, and also consistency of standardization across programs, which was interesting I thought.
DR. IEZZONI: Is that UB93 or is UB92?
MS. MARKUS: No, I'm sorry, that's UB92.
MS. MARKUS: Based on all these items states selected under data that want collected, and they have also gotten into their contract. Then we'll talk about now the findings that we have collected on the encounter data, access data, and quality data. For each of these areas I will mention the federal requirements that apply, since the federal requirements were a pretty big part of what they collected in the first place.
In terms of encounter data, federal requirements were, as Sara mentioned earlier, were fairly limited prior to the BBA. Let me change with the BBA, because it gives a broader legal basis for the secretary or HCFA to require across the board submission of data by states through their MMIS.
Before that, MCOs generally had to maintain sufficient patient encounter data so they could identify the provider providing services to the Medicaid patient. Under Section 1915-B waivers that requirement also applied, but HCFA also started requiring about a couple of years ago, that the plans collect data for the purpose of focused clinical studies.
Then Section 1115 waivers, which is where the most requirements are made, HCFA started asking as a condition of approval that the states collect 100 percent of encounter data or complete encounter data of all services provided.
I also included as an illustration, as an example of the state contract negotiations. It basically states the data, the submission duty for the plan for encounter data. It states health plans must prepare and submit encounter data to the state department of health services or its fiscal agent as specified. The state instructions and format for encounter data are included in the bidders' library. Failure to comply with these standards may result in sanctions as defined in Section 2.18.5.
This is actually fairly typical of a provision that you can find in contracts. All states we included in our study did have a provision in their contract requiring the submission of encounter data.
PARTICIPANT: Excuse me. On the encounter data, I know one of the issues before about linking up the data was the determination was that it would have race and ethnicity in many cases. Was the fact that the identifiers that were unique to the files weren't available? They are required in encounter data, are they now provided with that same identifier?
MS. ROSENBAUM: Unless it is in the bidders library, none of the contracts themselves that I can think, but we can go back and look, addresses the issue of the identifier as part of the encounter data.
PARTICIPANT: One of the things that we have heard from the states is that they receiving individual level care data from the managed care plans, but it wasn't with the same identifier, so they couldn't match it to enrollment.
MS. MARKUS: We did ask states about how whether they were able to link their Medicaid data to the enrollment file, the provider files, to the fee-for-service encounter data, and most states said that in theory they can do this to a certain extent. Some states said that they don't have a patient identifier that lets them do that.
PARTICIPANT: So it sounds like the news is the same that we heard a while back. Are some of the states mentioned that get that feedback?
MS. MARKUS: It's very limited. It sounded like they could for example link fee-for-service provider data with their encounter provider data, because they were from two different databases through Lotus, I believe. But that was basically the extent of it, and it had to be a fairly limited amount of data.
PARTICIPANT: Who generally is collecting the enrollment eligibility data? Is it the state or the MCOs? It's the state.
MS. ROSENBAUM: Actually, most states do require their MCOs also to turn in parallel enrollment and disenrollment data. You will see that when we get to the sample specifications. I don't know whether anybody indicated where they might have a parallel requirement on the plans.
For payment purposes, I think they are looking for a reconciliation between their records on who is enrolled in a plan, and the plan's records on who is enrolled in a plan. I can tell you as the board member of a plan, that one of our biggest headaches is that our sponsors' very often, and our records are not compatible on who is enrolled. So I would assume that this is done as a cross check.
PARTICIPANT: Have you seen any effective mechanisms for tracking migrant populations?
MS. GREENBERG: When they ask for parallel enrollment data, that doesn't include race and ethnicity?
MS. ROSENBAUM: A plan probably would never get it. They would simply be sent a monthly or weekly or daily list of the names that have come in, period. That's all they know. They get a list in the mail or electronically saying here are your members for today, and here are the people who have been disenrolled for today.
The characteristics of those members in fact is coming up a huge amount. I'm sure this is something that Kathy is well aware of. Where you are enrolling people -- we just finished a huge enrollment study for the department on managed care enrollment. Where this becomes of course quite crucial is less around race or ethnicity per se. When people with specific health care needs are auto-enrolled into a plan, and since they were auto-enrolled, there isn't even the preliminary flag from a voluntary selection that somebody with a major life impairment is coming into the plan.
It can be weeks and weeks before the plan -- ranging from a simple pregnancy to someone with a severe mental illness, which was a horrific problem in one state. Whole groups of people with profound mental illness were auto-enrolled and nobody knew. The plans they were enrolled in didn't have a proper network.
So you don't get any personal characteristics. Most states will tell you they can't convey. At that point, they don't know either. They just know them as a beneficiary.
MS. MARKUS: Well, we found out that all states collect encounter data, but the amount of data they collect varies. About half of the states said they collect complete encounter data, which means they collect also on labs and pharmacies. A couple of states said that they were missing pharmacy data. A lot of the states are actually going to move towards putting that in their contracts, so they will be collecting pharmacy data.
The variables that they collect are usually specified in terms of contracts. Oftentimes it is attached in the back of the contract, and you refer back to that. They list all the data elements, which tend to resemble any billing element that we would find in fee-for-service, procedure code, basic treatment provider, et cetera.
As I said earlier, I was mentioning the selection criteria. Three states said that they actually phased out data elements on the NSF/HCFA form.
Most states get that information electronically from the plan to the state, although that is still in the works for a couple of states.
PARTICIPANT: And most of the states get the enrollment data?
MS. MARKUS: No, the encounter data.
PARTICIPANT: Most states get the encounter data from the plan electronically?
MS. MARKUS: Yes, seven states.
PARTICIPANT: And we're getting individual level encounter data for all the states, but not complete for pharmacy, but it's getting most physicians and hospital visits?
MS. MARKUS: Right.
PARTICIPANT: Are you going to tabulate the characteristics of the encounter data that is collected?
MS. MARKUS: We haven't done that.
PARTICIPANT: I don't have a good sense of -- encounter data can mean there was a hospitalization, so there is a discharge.
MS. ROSENBAUM: That's the core data element.
PARTICIPANT: It really ranges from totally useless to detailed.
MS. MARKUS: And some states have two pages, and others have seven pages of justifications.
MS. ROSENBAUM: So what you would like to see is more detail around each of the respondent states' encounter data format? What goes into their format.
DR. STARFIELD: Enrollment too.
MS. ROSENBAUM: What elements are in the enrollment data they get back from the plan.
PARTICIPANT: There also can be an issue of the degree that they are following other standards, like some of the electronic standards for health care transactions.
DR. IEZZONI: That's a good point, comparing what they are doing to the rules that HHS may be about to promulgate.
DR. STARFIELD: I guess we still don't expect them to do that yet. That would be baseline, because there are no standards.
PARTICIPANT: Some of them will be using the equivalent of an electronic bill, although probably not using the 8037.
PARTICIPANT: For the enrollment data, will there be a question, when somebody is enrolled into a managed care plan, do they file a form with state?
MS. ROSENBAUM: Enrolling into Medicaid; not managed care, but becoming eligible for Medicaid.
PARTICIPANT: They fill out a form for the state?
MS. ROSENBAUM: Right. An application is completed.
PARTICIPANT: Then when they are making the decision about the managed care plan, then they fill out an application for the managed care plan. So that is essentially the enrollment?
MS. ROSENBAUM: Yes, in most states it's a two stage process, your eligibility determination, and the information in that process may or may not be the same. When I was referring before to the enrollee data, it's not the enrollee form where you would get the data. It's the underlying eligibility file where the data are, because the enrollment form, depending on the state, sometimes is synonymous, especially in the demonstration states, in Oregon and Tennessee, enrollment is synonymous with eligibility determinations, all at the same time.
But in most states it is actually broken into two stages. The only thing that would happen at the enrollment point is that the enrollee, the beneficiary, would be counseled by somebody, either through the mail or face-to-face, to choose a plan; would fill out something that has your name, maybe your ID number, your card number, and say I choose Acme, and then would go into Acme.
Then that form itself might have almost no information on it whatsoever, except the name and the identifier of the person, and the name of the plan chosen.
MS. MARKUS: What uses do states have with all that encounter data that they are collecting? We did ask them about current uses of what they were planning to do in the future with all that information. About half of the states said that they were using it to measure quality, although it was still very limited. A third of the states use it to set rates and adjust rates, and the remaining uses are measuring plan compliance with performance, and comparison of utilization rates within the plan. Only a couple of states are doing that at this point.
DR. NEWACHECK: Did you get any evidence that states actually made policy decisions based on that information? For example, refusing to renew a contract with a plan?
MS. MARKUS: Not really. They do talk about monitoring the performance of the program, improving baseline, what they do with the information, but we didn't ask them specifically.
DR. NEWACHECK: I think one of the tensions that we see and we heard in Massachusetts is that a lot of data are collected, but not really analyzed or used. That puts a burden on the plan or somebody. There is the burden of collecting that information and providing it, but it was not used for policy purposes, so why bother?
MS. ROSENBAUM: Well, also, until the methods get better, I think most departmental lawyers would advise their clients if they understood the issue completely, that it would be very hard to base a legal sanction on the data that is collected today, because the data are not good enough. They are not complete enough. They are not accurate enough.
This has been a tremendous problem. It's one of the reasons why so many states -- and when you get down to drafting a contract and thinking hard about this -- it is hard to get off of things like show us your network. If you don't have enough doctors in certain places, we will sanction you, because it's an easier thing than tying sanctions to encounter data or utilization data where everything from the elements that go onto the measure are in despite, the collection methods.
I can imagine what would happen in a cross examination over data where millions of dollars in sanctions are at issue. So your system has to be air tight enough to withstand a legal challenge to its sufficiency and its validity.
DR. NEWACHECK: So then if we don't have that sufficiency and validity now, what's the point of collecting the data?
MS. ROSENBAUM: Hopefully, at some point you might develop sufficiency and validity. I think the most important issue is -- and I was actually quite struck myself by the answer on capitation rate setting, because the big, big use right now I would say for the data, as imperfect as it is, is to begin to start connecting certain use patterns, with certain broad populations.
If you are enrolling an SSI population, and you are enrolling an AFDC-linked population, you want to be able to at some point, see where the big divergences come, and especially as we now enter this post-welfare reform era, where the restructuring of Medicaid is going to probably dramatically change the health profile of the typical Medicaid enrollee, because they enroll later. They will enroll at point of service, rather than as part of their welfare benefits potentially.
You need to know a lot more about the frequency of use. If I were in this position, while some of the other issues are very important for overall policy and planning purposes, this issue of whether the payments into the plans are even remotely sufficient for who is being enrolled is first order.
DR. IEZZONI: Well, keep in mind that Medicare, through the balanced budget act, has mandated that managed care organizations will risk adjust starting in January 1, 2000. Once Medicare starts doing that, based on encounter data, they are proposing using something called a TIP ACGs(?), that will I think probably drag a lot of other similar situations into the risk adjustment.
DR. STARFIELD: I'm having a hard time understanding what it is they could get from the encounter data that would help them with capitation rate setting? All these people are using just age and sex.
MS. ROSENBAUM: If they could tie it back to the eligibility files, they could get a lot.
DR. STARFIELD: Yes, but what are they? Three states are using encounter data to do capitation rate setting. What are they getting off of the encounter data to do it?
MS. COLTIN: They can run the procedure codes against the standard fee schedule.
DR. STARFIELD: Are they doing that? Are they actually doing that? Basically, a fee-for-service method of setting capitation rates. The more you do, the more you get.
DR. IEZZONI: Barbara, they are using the ACGs in Minnesota I think.
DR. STARFIELD: Well, Minnesota is not one of the states.
DR. IEZZONI: No, my point is that you should know about that.
DR. STARFIELD: I'm thinking of these states. None of these states are.
DR. IEZZONI: Washington state is about to start doing risk adjustment.
DR. STARFIELD: So they are basically using a fee-for-service method; in other words, the more procedures you do, the higher capitation you get.
MS. ROSENBAUM: The encounter data form is simply a shadow 1500. They are using a claims form as the model for developing their encounter data form.
DR. STARFIELD: It's incredible.
DR. NEWACHECK: They may also just be aggregating all the data they get from the plans to study capitation rates.
DR. STARFIELD: This is capitation rate setting.
MS. MARKUS: The sophistication at which they are doing that, we can't tell from the interviews.
MS. ROSENBAUM: It's something that we could certainly go back, if you would like to know more about precisely how they use the data, which of the elements they find helpful, how they have applied it. Are they aggregating? Are they using individual elements off the forms? Certainly, in three states, it's an easy thing for us go back and see this tomorrow.
MS. WARD: If they think it's something different than --
MS. ROSENBAUM: What they've been doing anyway. Yes, before, my guess is that the fee schedule was even more artificial in a fee-for-service program, and just completely tied to whether it was time to do another adjustment, because you had no dentists left in the program.
MS. COLTIN: It's interesting how states represent themselves too, because Massachusetts has not collected encounter data as of yet, and you're saying all states are collecting it. It's in the new contract, and the first time they will receive any is June 30 or within 60 days after the close in 1999.
DR. NEWACHECK: Sara, on the point about getting more information. I think it would be very useful to have some specific examples of where data has been useful to states, not just in capitation, but in any of these areas. If we're going to argue as a committee -- I'm assuming that one of our aims is going to be that we do need to collect more data to have some evidence that it makes a difference and is useful to be important, and the more specific, the better I think it is.
DR. TAKEUCHI: I would support that, especially with quality; what people thought quality was, and how they measured it.
PARTICIPANT: One of the questions we had, are any of the states or the managed care plans taking a close look at their high end users to design outreach plans?
MS. ROSENBAUM: There are states that collect selected utilization indicators. They will collect aggregated data. It shows up in their contracts. We have a flag in the sample purchasing specifications. There are two that say they can collect for use among plans. Were there states that said that they collected to look at use patterns within the populations? We can go back and ask them additional questions. What specifically were they looking for, and how did they do it?
PARTICIPANT: I think there was also some discussion with HCFA about HCFA's support for modifying the MMIS systems to support enhanced data analysis.
MS. MARKUS: I think they are funding some of it, or helping states set them up so they can actually manipulate encounter data and use it for analysis. That has come up actually in our interviews. Actually on the next page, if you will flip to that, it was one of the concerns some of the states had.
PARTICIPANT: States might be interested in taking it a step further, but at least up to a point they didn't have the technical capabilities.
DR. IEZZONI: We heard that strongly in our site visits. They don't have analysts. They don't even put together the information that they get.
MS. MARKUS: The states say we get more than enough data. It's a struggle to know what to do with it. I was actually in one state that said they thought they had adequate staffing and financial resources. So it wasn't we don't have enough people to deal with this. So that is an issue across states. Interestingly though, when we asked them whether they needed any technical assistance, most of them said no.
Concerns -- quality of the data, Sara alluded to earlier. The timeliness and accuracy of the data. Implementing the data collection process. Things cannot happen all at once at this time. At the plan level and provider level there are a number of issues. It is important for the provider to plan, and a plan should state that the plans won't have the capacity to report the data. They don't have enough resources. They have a high turnover rate of programmers, et cetera, et cetera.
Another issue that was raised is also the uniform interpretation of what an encounter is, and how plans actually report that information. For examples, labs, one plan might report that as two encounters, whereas another plan might report that as one encounter. So that was raised as an issue.
MS. COLTIN: With regard to the analytic capability, I can understand them saying that they don't want technical assistance. Did they talk any more about the difficulty in recruiting people who have the right kind of skills? Because that's an area where the federal government historically has provided public health traineeships and different types of programs that have helped develop people with those kinds of analytic skills. We are finding that we just can't recruit them.
Now that we have integrated delivery systems competing for those same analysts, as well as the managed care plans at the universities, and the states who are starting to get more into doing these kinds of analyses, there just aren't enough people around that have that kind of training and those skills. It seems to me that's an important role that the government could play.
MS. WARD: And whether any of them turn to their health departments to contract for analytic.
MS. MARKUS: I don't think contract, but they do consult with staff there, because they have the resources and certain skills to do that.
PARTICIPANT: But interestingly, they don't consult with them at the point of making their policy decisions about what to collect. They do consult with them after they have this mountain of data in-house that they are having problems accommodating. One of the reasons to pay a lot of attention in the purchasing specifications to this issue is to try and get states to re-think the front end part of this, as well as the back end.
PARTICIPANT: Are they investing in the capacity to understand or monitor the quality of the data? Do they really know, or is it impressionistic? Or do they have systems?
MS. MARKUS: A couple of states have used a system of validating the DNO. So they have plans submit the data on a monthly basis, and they check. Then they send it back to the plan. If they don't send it back clean, then they get sanctioned. There is this back and forth process. Some states said some plans don't require it at all. They have plans that do this and plans that don't.
DR. NEWACHECK: Do they actually invoke the sanctions?
MS. MARKUS: They haven't used them yet.
MS. ROSENBAUM: The typical contract says we can sanction you for anything that involves a violation of this contract. So on the question of are there sanctions, as a legal matter, yes, there is the potential for sanctions. Do they invoke the sanctions? That's where I think the variation would be great. Of course as the premiums drop, there are only so many times you can invoke a financial sanction before you are really jeopardizing a whole lot. Financial sanctions are sort of taught -- oh, yes, we'll just withhold capitation until we get it, except that you have all these people needing health care.
PARTICIPANT: The other point I wanted to ask was in terms of the relationship with the public health and integrating data from other -- there is another report that the Public Health Foundation is working out for the department that has gone into some of the same states looking for the integrated data. The draft report seems to indicate that there isn't much of that actually.
MS. MARKUS: They can't even do it internally with their own Medicaid data. So to the extent that they have to actually link that to other statewide information, only a couple of states are really starting to do something like that.
Finally, the HCFA requirements were a concern in terms of costs and the uncertainty around whether you can use that encounter to measure quality.
MS. COLTIN: Do you find that any of the states sort of enforce the data submission and other kinds of reporting requirements differently, depending on the size of the plans? Because a lot of these reporting requirements are fixed costs. If you have a large Medicaid enrollment, you can spread those costs over that large population. It may not be terribly burdensome for the plan to comply.
In a small plan, where you might have 5,000 Medicaid beneficiaries to spread that cost over, it becomes a really high administrative overhead, that in most cases, the Medicaid departments are not paying for in the rates that they are paying the plans. For the small plans, that can be, I think, a real disincentive to participate at all, but certainly to comply with some of the reporting requirements as well.
MS. ROSENBAUM: If there is variation, it will be in the practice. The contract will be pretty standard, because it is a public contract, and they can't bury it too much. But my guess is that precisely -- there they are not getting any data from certainly plans, it may be their weaker, smaller plans.
MS. COLTIN: They know they would push them out of business if they tried to enforce it.
MS. ROSENBAUM: Right, and then they are left with 30,000, and no place to put them.
DR. NEWACHECK: Aren't the contracts actually pretty uniform across plans within a state?
MS. ROSENBAUM: Yes, that's what I'm saying. The contract itself would read -- and it will say, you will have to collect the following data. Whether they then enforce those requirements equally is unknown. And of course the more they don't enforce them equally, the more when they do bring down the hammer, there are due process issues that get raised.
DR. NEWACHECK: Do any states use different contracts for different plans specifically?
MS. ROSENBAUM: It depends on the state. In a procurement during the best and final period, if it is a competitive procurement, and you have picked your bidder, you can invest in final, and make some final adjustments under some state procurement laws.
If it is an any willing plan situation, like an any willing provider law, you sign the agreement as if you would sign any participating provider agreement, and the question then is simply going to be how energetic is the state in enforcing the agreement. You get into this paradox that Kathy is raising, which is on the one hand, some of those requirements are simply not realistic for smaller plans.
On the other hand, once you start variable enforcement, it becomes much harder to enforce against anybody, because again, when you are in the world of public managed care, you have very important due process protections that would trigger for a plan, especially a plan that is in the program that is participating.
PARTICIPANT: Anne and Sara, before we leave HCFA requirements, are you going to talk about QSMIC(?) at all? I gather it just gotten started. Was that just not even on the agenda?
MS. MARKUS: A couple of states mentioned it as a future requirement; but they hadn't done that as of the interview.
In terms of access data, there are a number of federal requirements that MCOs have to comply with. One of them is the medical audit, the point of which is to assess the access to care. That system must provide for the collection of management data. So there is the limited reporting requirement there.
Again, for 1915-B requirement applies. They do have to do the medical audits. For Section 1115, the secretary could waive that requirement, but hasn't done that as of yet. Since the enactment of the BBA though, there is a clearer and more explicit requirement on the secretary to issue access standards, and to specify the data elements that need to be collected to measure compliance with these access standards.
In addition in the statute, MCOs would have to provide data to the state, which would have to be in compliance with a Medicare plus choice data set, which is used at this point, or an alternative data set that is specified by the secretary.
DR. STARFIELD: You are talking about access to data, and not data on access.
MS. MARKUS: No, data on access.
DR. STARFIELD: Can you get that from a medical audit?
PARTICIPANT: That's what the law says.
MS. MARKUS: Actually, some of the states who did rely on that, on the medical audit to assess access said that they are basically looking at various aspects of the care delivered such as turnover rates, combination of things that they do on site, rather than getting information transmitted to them directly, as some other states are doing. So I'm not sure the collect actual utilization data. They get various aspects of that.
MS. COLTIN: Massachusetts has collected actual data.
MS. ROSENBAUM: Exactly, and we pulled a couple of examples for you of the kinds of special studies the states will commission in their contracts. New York has availability studies.
MS. WARD: So do they cite those activities as "medical audit?"
MS. MARKUS: Actually, they would lay out access standards such as you've got to have 24 hours availability, and then will require a study to be done on that issue, submitted to the state. Actually, there are some states themselves doing the analysis.
DR. STARFIELD: Well, maybe the trouble I'm having is understanding what we mean by medical audit.
MS. ROSENBAUM: Well, I was going to say I think what happens is there is a clinical medical audit, and there is a way in which these folks may use health care audit as a more global concept than just a chart review, or the test review or whatever. They are thinking about it in terms of how is the plan functioning or operating.
DR. STARFIELD: So it's not a medical audit as we know it.
MS. ROSENBAUM: No, not as you know it.
DR. NEWACHECK: Anne, can you say a little bit more about the Medicaid HEDIS? You mentioned that as a post-BBA. Is it going to be required by all states to be used?
MS. MARKUS: I have been looking into that right now.
MS. ROSENBAUM: It could be in the BBA regulations. I would seriously surprised. In fact, most states, if I'm not mistaken, said in response to HEDIS that they use HEDIS as a general guidepost, which was a polite way of saying they sort of borrow liberally from it, but they see a lot of HEDIS as either not relevant to their populations, or more of a burden than they can put on, because they need all these other data. So they have already burdened people with the basic encounter data they need, and so then to ask for these other data, which have less relevance to them, it has not been uniformly audited.
DR. NEWACHECK: Well, I think that was true with the older versions of HEDIS. But the new CAP survey that is being --
MS. ROSENBAUM: The issue of it could be its satisfaction studies. I don't know whether states would, on their own, gravitate toward CAPS as their -- everybody asks for -- at this point, I think almost every state is asking for some sort of consumer satisfaction measure. I would suspect that if this turns out to be a better tool for measuring satisfaction, I think people are very eager to find the best available tool.
It's both the outcome measures and the input measures in HEDIS that I think states see as sometimes relevant, sometimes not. They are going to be guided by it, but they are not going to pick it up. It's just like our contract specifications. They might pick up pieces of it. They might be guided by it. It might help them formulate something they want to use.
DR. NEWACHECK: My understand was to get HEDIS accreditation, that you had to use -- at least with the commercial plans -- that you had to use the new HEDIS consumer survey. That was my question, whether they would or not.
DR. STARFIELD: The other thing is do you have to be a user to be in a CAP survey, or can you just be an enrollee? Do you have to have used it to get CAPS?
DR. NEWACHECK: No, just an enrollee.
MS. COLTIN: It's an enrollment-based sample. But the first question in the survey is have you used services in the last 12 months, and if not, you don't answer the rest of the questions.
DR. IEZZONI: Anne and Sara are probably going to have to leave earlier than you would want them to. This has been really fascinating, but we want to get to discussing the contract language. So Anne, do you want to just wrap up on this?
MS. MARKUS: Let's jump to --
MS. ROSENBAUM: I think you should mention on the public health database.
MS. MARKUS: Actually, we can wrap up with that, and just talk about public health data in general. I divided that into four main issues: direct reporting to medical agencies, direct reporting to public health agencies, encounter data, and the integration issue we talked about earlier with Medicaid data and public health data.
Not surprisingly, very few states require MCOs to report direct information to public health. One state had actually started collecting information on the deaths directly from the plan. They stopped doing that because they didn't get enough useful information about the circumstances of the death. So they actually stopped that.
In terms of direct reporting to public health agencies, this didn't come up in interviews, but in our review of the contracts, very few states incorporate a direct duty on the plan to report notifiable events to public health authority, which is interesting, since a lot of public health codes specifically do not include managed care organizations in their definition of the provider who has to report back to the state.
MS. ROSENBAUM: This is this whole tension about what is a managed care organization? It is a provider? Is it an agent? What is it? We deal with that in the purchasing specifications.
DR. IEZZONI: Did anybody raise concerns about that from a public health perspective?
PARTICIPANT: Public health was not a big issue.
MS. ROSENBAUM: The CDC has raised many concerns.
DR. IEZZONI: That's what I'm thinking. The CDC would be very interested in that.
PARTICIPANT: We raised the concern all the time, but it's always the same story. It is just a history of Medicaid and public health not speaking to each other, that we can sit in Atlanta and talk about all day and all night, but it doesn't it a reality at the state level.
DR. STARFIELD: Why is this any different than the commercial plans?
MS. ROSENBAUM: It's not. The CDC's interest in our work has not been confined to Medicaid. It is a universal issue.
MS. WARD: My experience with the plans is they see it as a statute that regulates the individual provider, and it's not their business. They don't want to get involved in it. They don't want to put it into their contracts. It isn't any different from child abuse reporting. There is already a statute that regulates individual providers.
MS. MARKUS: I mention also encounter data, since it is collecting a lot of information on pharmacies and labs, and it is an important source of information to public health data.
Finally, integration of Medicaid data. We have talked about it a little bit earlier. I'm not going to expand on that, but interestingly one state referred to the local process as a successful element to integration of data between the agency and the public health agency. It was a matter of trusting each other. As soon as they started doing that, they were able to actually get together and integrate the information.
DR. IEZZONI: This has really been great.
MS. MARKUS: I'll have a draft.
DR. IEZZONI: Super. So that, combined with the report that you sent us in the spring, is a wonderful body of work, but it's getting late. Any final question about this before we move on? When do you have to leave?
MS. ROSENBAUM: I should leave here -- actually, what I'm going to do is call the office, and have someone walk over to my classroom and my stuff, so I can leave here about 3:30 p.m.
DR. IEZZONI: Why don't aim to get Sara out of here by 3:20 p.m.? So we'll just defer our break until 3:20 p.m.
So Sara, do you want to begin to lead us through this?
MS. ROSENBAUM: Paul, standing here from the CDC, can tell you that any of these drafting sessions is an extremely iterative process. Any one of the items that we have drafted for any of our clients has gone through 8, 9, 10, 11 of these things before people are happy, and the public vetting starts.
Because you can't really imagine what you want until you see the document in front of you, and that's the way it should be. So this is essentially as if we were on the very first round. This has never been reviewed by any expert in data or data collect.
What I did to prepare this for you all to get the ball rolling was to take the results of Anne's and Lee's work, to take our big contract study that shows the current state of play regarding data collection expectations, data availability. It's really two different issues. One is what data you are going to collect, and what data you are going to report, which could be very different undertakings.
And then the third source of information for this draft was my own understanding from my many, many discussions with people at the CDC, and from other people about this very big issue that Anne just alluded to, which is what do we do about what we call public health data? The data that are related to vital events, notifiable diseases and conditions, and registries that are being maintained. There may be other kinds of public health events, but those are the ones that I am aware of.
So what I have done in this document is to address several different categories of data. This is by no means an exhaustive document. This does not address the data that you would be collecting to assure financial stability, solvency, and the other business aspects of your company. That's a state insurance department activity or the Medicaid agency borrows those databases to build its own financial rules.
These specifications address public health surveillance and community-wide disease tracking activities. It addresses data related to health care access and service utilization; data related to the measurement of health care outcomes or inputs; and data related to health care quality performance. So those are four areas that I flagged right away as appearing a lot in the contract specifications, and certainly showing up in Anne's review here.
So then essentially this document has two parts. Part one enumerates the classes of data that are covered under the agreement. What kinds of data does the purchaser have the right to expect will be collected and/or reported? Part two specifies procedures for contractors to follow in data collection. Now part two is very short, because one of the things that I noticed in all of these contracts is it says we want the following data.
In New York you saw it. You go to the bidders library to get our procedures. You will go to this manual that we're going to supply you to get our procedures. You're going to go to some other secondary source which is incorporated into the contract's data collection procedures.
Now the word "collection procedures" is pretty wide open. It could mean literally how we want you to sample, or what time frame we need you to use. But it could also mean the elements in the data. It could mean infinite. So you could decide that you actually want to, as a sample and sample language, address a whole lot more about procedures than I could even begin to come up with, because I have nothing to look at. I have no idea what a procedure would be here.
What I tried to focus on where the kinds of data that in theory, a buyer would want. So Section 101, which begins on page 4, simply says that this is the basic duty section that says you have to insure if you're a contractor, that the classes of data and information identified in part one are collected and reported in accordance with the specifications under part two. That's sort of Law 101.
The second is that the contractor must make the data and information covered by the agreement available to the purchaser, and to other individuals specified by the purchaser as authorized to receive such data, either in this document or subsequent written communications. I mean I don't need to tell you there are a whole of lot reasons why you can't simply give everybody who would like it, access to the data, but obviously the purchaser may have some interest in assuring access, and so that's what that's for.
Manner and form of collection reporting. The contractor is going to carry out the collection and the reporting in whatever way the purchaser says to carry out the collection and the reporting.
Finally, the flip side of B, which is you only give access to data in accordance with what we have allowed you to disclose. You cannot give anybody else access to the data. Now generally speaking, most states would have state confidentiality statutes, medical records laws, whatever, that would be incorporated by reference. You cannot write a contract that is ultra viras. You cannot write a contract that contributes in any way, a law that is out there someplace. An agency doesn't have that power.
So there is usually boilerplate in one of these contracts that says this is -- and I think somewhere in here I stuck the boilerplate in -- that this is in addition to any other data collection and reporting requirements under the law, and cannot be read as contrary to those.
The applicability -- this is very important. This is a big policy call, and I took a stab at the policy in this area. The terms of these contract provisions apply not only to the contractor, but to all providers and suppliers in the network, or who do business with the contractor.
The reason that this has become important goes back to actually both encounter and potentially public health data, because of the use of out-of-state labs, mail order pharmacies, you want to be sure that anybody who is doing business in your state, and from a legal point of view, it doesn't matter if the lab is in Oshkosh, if the lab has an agreement with someone who is doing business in the state, is analyzing specimens that were drawn in the data, it is doing business in the state.
Probably if you didn't specify this in the contract, the contractors and its subcontractors would be bound, but it's a good idea to make your intentions clear.
The suppliers of lab, pharmacy, and pharmaceutical services provision is here for the reason that Anne noted. We wanted to make crystal clear that for purposes of collection and reporting, that suppliers of these kinds of services are providers, because it is our understanding that a lot of the most important data relating to public health or encounters utilization patterns are going to come not from a physician's office or a hospital, but from a pharmacy or lab.
DR. IEZZONI: How about durable medical expenses, because for Medicaid with the disabled, DME is really huge.
MS. ROSENBAUM: It should be treated as a supplier. It certainly would be, and it can be explicitly put by the drug. And that's why what you do is with this, you would be drawing attention to something.
Of course in a contract, actually, the only thing that really matters is the definitions, so I gave you one definition to see what definitional activities you might want to think about. There is a definition here for a pharmaceutical service, to make clear that it covers not just prescribed drugs or biologicals and insulin.
A standard confidentiality provision. Here is the provision that says this is basically in addition to whatever other laws apply in this area. And then of course that's the summary. Behind that is the much more detailed version, part one, what data and information are required? So here you see data related to health status, health care access, and health care utilization.
So the first class is the data relating to the diagnosis or treatment of illnesses or conditions or other health events that either: (1) are classified under the law as notifiable diseases, diseases for which public health registries are maintained, or vital events; (2) are identified as such, because some states may not have laws on the subject, it may be that the state health agency identifies notifiable diseases, vital events, registry data; (3) data relating to enrollee-specific claim level encounters that cover medical, dental, diagnostic, and treatment encounters.
MR. HITCHCOCK: Where do you see the mental health and substance abuse fitting in here?
MS. ROSENBAUM: Well, they are either medical or dental. I'm just saying for now we can elaborate. I'm giving you only the most basic. You could list them absolutely as a specific, enumerated subcategory of data. You could assume --
DR. IEZZONI: We heard in our site visits that substance abuse and mental health issues are really very on the top of the burner, and of great concern.
MS. ROSENBAUM: Oh, yes.
DR. IEZZONI: The issue is also going to be what is carved out. I don't know whether your idea about data gathering also would focus on conditions that are carved out.
MS. ROSENBAUM: Well, that's right, because you may have two different kinds of contractors. In Massachusetts, in some situations you get one set of services from one plan, and then other services from another.
DR. IEZZONI: Right.
MS. ROSENBAUM: This is designed to be right now completely generic. So if you are writing a carve out contract or you're writing a general purpose contract it doesn't get into any specific diagnosis and naming the data. It simply names classes of data. But when we get around to how we want to work this document over, we can think about both a process, and then where you want to put more samples to highlight that a medical encounter means the following.
One question I have is whether given what some of what Anne and Lee heard, whether the notion of an encounter is even an accurate way of always describing the data that you want, because a lot of the data are not encounter. So that's a big thing to come back to.
So here is your encounter paragraph. Here actually is medical equipment -- pharmaceutical supplies, medical equipment. Then 4 is utilization of the following classes of services by individuals.
Now one thing that is flagged here is utilization by either voluntary enrollment or auto-enrollment status. This is an issue that right now, to the best of my knowledge, most states are unable to do in disaggregating data. For the very reason that Dr. Takeuchi just raised, this is the kind of thing which you would probably want to know.
You would want to know whether certain groups of people who sort of don't mentally connect with the plan, they are just placed in the plan, are users in the same way as people who voluntarily, knowingly, deliberately have chosen enrollment in a plan.
That is an example of a kind of sample where you are really pushing the edge of the envelope, because most states in our enrollment study told us that they do not communicate to the plan which enrollees are voluntary and which ones are auto, and therefore the plan couldn't possibly tell them. They wouldn't know unless it is coded somehow in the file. The state would have to do the matching itself.
The fifth category of data are data regarding the availability of access processes for services that promote physical and communications access. Specifically, this is actually drawn from three state contracts, Massachusetts, Minnesota, and I think some from Wisconsin, which expect you to provide data on how you equip your plan to be accessible. They want to see information on that. So it's not data in the sense of the other data.
Data information again on provider network. This is a big issue for most states. Many states will ask for -- they certainly want the list of the providers. Some states will go beyond that and want very specific information about the providers. So I've given you a menu of what shows up in the contract study.
DR. IEZZONI: The language piece was something we heard a lot about in Arizona. The match of language between the provider and the patient, the discordance is a problem. So again, we go back to the fact that we don't know about the language of the enrollees.
MS. ROSENBAUM: The purchaser would simply assume that the purchaser was enrolling people. Now Massachusetts, as I recall, has quite an elaborate system of informing its plans. As I read the contract, they literally inform a plan that you are serving an area where you have the following kinds of populations represented in your service area. We want to see essentially what you are doing. But they are unique. There is no other state --
PARTICIPANT: California does that.
MS. ROSENBAUM: They do it, but not quite at the same level of sophistication of saying: (a) we're going to give you a lot of data; then (b) we want you to come back to us with how you are doing it, and see on multilingual providers, Massachusetts as I recall is quite unusual. California will say you have to make your services accessible. You could have a translator. You could use the AT&T line. Massachusetts, as I recall in the most recent contract we have from the finished database, which is 1997, actually requires for certain areas, multilingual providers, which is singular in the country.
The seventh category is of course geographic and travel time proximity. These three, 5, 6, and 7 are literally hybrids from stuff in our database, where I flipped through and found things that were interesting, that states seemed to be asking for. Eight is something that is the kitchen sink clause that you usually find in a contract, which is anything that we haven't thought of. I defy any state to try and enforce this provision the way it's drafted.
It's so broad that it would be very difficult to enforce, but most states put it here: (a) because there are things they haven't thought of; and (b) because they may need data to verify data that they have received. That's often the reason why it's there, because that came up in some of the interviews too.
DR. NEWACHECK: Some of these things seem to be fairly general wording that I wouldn't expect to see in a contract, like for example putting in analyses of residential patterns within a contract. That could be interpreted a million different ways.
MS. ROSENBAUM: Absolutely.
DR. NEWACHECK: So would an actual contract have some detail about what that means.
MS. ROSENBAUM: This is the actual.
DR. NEWACHECK: So if this is the actual --
MS. ROSENBAUM: I literally cribbed from some state on this one.
DR. NEWACHECK: So why couldn't a provider just simply say, well, we surveyed two providers?
MS. ROSENBAUM: Well, this is why you should view this as your starting point, because not having the normal process for developing these, which was having two or three on the phone for six months to develop these and give you the focused study from the database, and going through all this. What you are seeing is my attempt to flag here, the big areas of data, because what we are going to need to do is figure out now how to take what is this basic first run document, and address in more detail.
Sometimes you may want to address it in clearer contract language. Sometimes you may want me to fashion for you that you have a couple of different ways to do this. One is that we write more elaborate contract language. One is that we do running commentary at the bottom of the page, which is very common with a legal document. You do contemporaneous notes that explain for the parties what is meant by residential. That we expect all residential areas, a substantial portion of your network, meaning greater than 50 percent. That's essentially parallel evidence to the basic contract term. Or you can write a lengthy contract term, with a lot of definitions underneath.
MR. HITCHCOCK: Or that it's competitive? Could you expect the offers to come back with proposals on how they would do these types of things?
MS. ROSENBAUM: It depends on the state. There are all these different ways. I mean for example, there are states that do an RFP that is quite detailed, and that says this is how we want it done. Submit your bid. It's a competitive state. Submit your bid, and then we'll pick some winners. In your bid we not only want to see of course you bid bid, the financials, but we want a full explication from you of how you are going to address these issues.
Some states may have an any willing plan situation, but then they are going to sit down prior to signing an agreement, and get a much better feel for how this participating provider is going to address their concerns.
There are some states that will write something called a RFI. They essentially go out and see what the horizon looks like. Tell us how you address this issue, this issue, this issue. And one thing they could say is tell us how you address what data do you collect. What data do you see is vital to your own operation.
But this is written not so much to give people all forms of the question, but to let them know what things, whether they are going to ask it as a question, or provide it as a statement, what they want to address, which is why you can go to town. You can add eight categories of data. You can decide to add 200 running footnotes. You can do all kinds of things. The more you want to bring to attention certain things, for example, on encounter data, which is just mentioned here as data back on the prior page, data related to enrollee-specific plan level encounters.
Well, we haven't defined an enrollee-specific claim level encounter. We haven't defined the elements of that encounter that have to be captured in the data. Do you want a definition? So you can begin to tie your thinking and recommendations in this field back to either a definition, because it's all sample. It's all for people to have examples of.
Then I will draft it into language that can be picked up and used, picked up and modified, whatever. But it would be in their format.
DR. IEZZONI: I can just ask Marjorie, is this an area where we should really try to think about the HIPAA standards? Because we have enrollment and encounter HIPAA data standards. Should be trying to think about that?
MS. GREENBERG: I wondered about that myself.
DR. IEZZONI: It would seem to me that we should.
DR. STARFIELD: Well, how much are you bound by the ABC?
MS. ROSENBAUM: How much am I bound by it? I'm bound by it how long would it take me to rip this up and write again. I've literally got you going with this to show you what goes into putting one of these things together.
DR. STARFIELD: For example, you've got lots of levels of analysis in here. You've got the plan, you've the individual enrollee, you've got the population.
MS. ROSENBAUM: Right. We can reorder it. I simply set it up the way the thinking flow works in the lay person's mind like mine, or the way you see it in the contract. Which is first I need to know what data you want me to collect. Then I need to know how you want it collected. Within the "what" data, there may be person level, there may be population level. You may have a topology you want me to use.
DR. IEZZONI: I think what we should do, because I'm looking at the clock, and I see that Sara is going to have to run in 10 minutes. We should figure out as a subcommittee how we want to interact with Sara and Anne. Because what we really need is we need a document that we can take to the full committee and say, look, this is what we're going to think about proposing these models.
It would be nice to be able to do that at the November meeting, although that might be too soon. That might be a little ambitious. So why don't we think about what might be the best way for us to work with this.
MS. GREENBERG: In doing this for a number of the CDC programs, you met with the CDC programs. You interacted on phone, personal, et cetera on the immunizations, et cetera. This is a little different where you are working with this advisory committee. I'm just wondering the extent to which you shouldn't be bringing HCFA Medicaid people into this discussion.
The fact is that you are going to be making recommendations to the department. If you haven't included people who will implement the recommendations, they won't have any saliency, I don't think.
MS. ROSENBAUM: The way this has worked each time is that some decision making person or body appoints two or three people who are going to meet early and often basically. Each version of this has been exactly the same. We take a stab in the dark, because we found that if we simply sit and talk to the client about what these things are, they don't get it, and we don't know how else to do it except to start.
Usually we are semi or significantly off base, and then they sort of say, no, no, this is the structure we want to use. This is how we want to lay it out. This is what the final pyramid to look like. These are the issues we need to address within each level of this structure. And here 14,000 pages of reading that you have to absorb and then draft into elements for us.
So what we really need at this point is either two or three people from this subcommittee, or we need designated people from the department who would be speaking for the subcommittee to work up the document. Now I have raised certainly the people who have worked on the specifications with us from our other walks of life, particularly the CDC folks, obviously have a great deal of knowledge in the area of if we were going to try and write a model public health specification.
On the encounter, on the utilization, I assume there are people who know what they need. The trick now is to get it captured in language that would be sample.
DR. GREENBERG: I'm sorry that Jason Goblar(?) isn't here. I know that he has been working the last several months within the Medicaid office for Rachel Block, looking at what kind of reports they are going to want from the MSIS. I think that kind of input would be useful to the committee. You may not agree with them.
DR. IEZZONI: And the mental health piece and substance abuse.
PARTICIPANT: If I could make a recommendation, having been in the middle of this from the CDC standpoint. I think what might be useful here, you've got to have sort of a small group, but it's really important that it not be just department staff doing this, even though you might want to do that, because you will not be invested. Even though the larger committee has got to be advised, some combination of maybe two members from this committee and a couple of staff, however that is parsed out, is what will make this work.
MS. GREENBERG: I think you need to be engaged, but I think you need both sides.
DR. IEZZONI: What I'm thinking is that we need somebody from a state. Now I don't know whether Elizabeth would have time or George. George van Amburg has rotated off the committee, but was the person that suggested that we do this project to begin with. I don't know Elizabeth, whether you would have time, but I think we need somebody from the state. Frankly, Kathy being somebody who knows a lot about a plan, but Kathy has got a gazillion things.
MS. GREENBERG: It's an interesting idea of maybe trying to engage George under contract.
DR. IEZZONI: If Elizabeth wants to do it, it would be great.
MS. WARD: I think George would be the one.
DR. IEZZONI: George also has a real health department perspective. I think George would probably do it, because he is interested.
MS. WARD: I would be glad to be sort of back up to George and be available through e-mail.
MS. GREENBERG: We need to cover how much time would be expected of him.
MS. ROSENBAUM: Generally, what happens is whoever is going to do this would take this thing home, make notes about issues they want to address at any of the levels we have been talking about here -- the scope of the document, the way it's set up, the contents of the document, whatever.
We would then schedule a -- usually about after 90 minutes, everybody is dead. So you do a 90 minute conference call, get some ground rules down about what has to be appear in the next round, and give me a bunch of stuff to do. We would then go back, write up what we were instructed to do, go off and draft another round.
My advice would be mostly in the shape of the table issues. What you want to show. The big thing to keep in mind here is that these issues are competing with so many issues in managed care contracting, and this is a legal document. This is not practice guidelines for data collection.
So what you have to capture here are the minimum legal statements that give you the right as a buyer, to get certain data in a format that is useful to you. The plan has the same set of issues, which is this is a legal document. It must end up with a set of expectations on it that it has a realistic change of meeting. Which means that you don't in here, write the collection method.
You don't in here, write down to the fourth order, but what we do need to grapple with is at that first and second order, how do we set this thing up to give a sensible portrayal of data, and what are the elements then of each data requirement? And what contractual things must go into the collection section?
So what we really need in this working group is a person with public health data knowledge. Someone with encounter and utilization data knowledge. Someone with plan capabilities data knowledge, whether it is Kathy or whether it's someone Kathy can think of. Then the policy input of the committee. Then we would have probably three or four 90 minute meetings. We'd get pretty close to done. The subcommittee could then review it.
DR. IEZZONI: Now how would we integrate people like Rachel Block and the QSMIC person and substance abuse and mental health? Because I think that they are going to be crucial as well.
MS. ROSENBAUM: SAMHSA has a person who is the liaison to this project.
MS. WARD: I think you need a state Medicaid person. You suggest from your interviewees, someone who is actually presenting a state Medicaid agency.
MS. ROSENBAUM: The other place where Medicaid comes into it is we assume that your vetting group for this will actually be the states who were interviewed for this process. So whether they want to spend a lot of time on the formation, or would rather react to it, it doesn't matter. It's always handy to have a Medicaid person in the initial go around.
DR. IEZZONI: Okay, why don't we do this, why don't we as a subcommittee talk among ourselves. By Thursday afternoon, when our subcommittee is all done with everything, we have Carolyn speak to you about who our representatives would be, or suggestions for how to go about doing this.
MS. ROSENBAUM: Right. I think it's something that can be done with no travel, just a lot of telephone calls.
DR. IEZZONI: Well, this has been wonderful. Sara, you've done a wonderful job. You have given us tremendous input in terms of workload.
MS. ROSENBAUM: Well, we have many happy months to go.
DR. NEWACHECK: How many more months do you have on the contract?
MS. ROSENBAUM: We have as many months as you need. Seriously, there is a contract that underwrites our work, but this is integral to everything that we do, that this will just go on until it's completed.
DR. IEZZONI: Well, this has been really wonderful. Why don't we take a break of our subcommittee, and come back in 15 minutes.
[Brief recess.]
DR. IEZZONI: We have a bunch of stuff we have to do as a subcommittee. Then the work group that Kathy is chairing, is going to be an important item for the latter part of the time.
Shirley, can you just make a list of things that our subcommittee has to do. We basically have to talk about how we are going to "staff" the interaction with Sara Rosenbaum and Anne Markus around coming up with data contract specifications. Maybe we're too tired to talk about that now, but we have to figure that out by the end of Thursday.
We need to talk about the process of coming up with a Medicaid managed care report, which we want to be really good. And we want it to be done soon, but it's going to have to wait obviously until the George Washington people are done, but maybe we should aim for January or something like that for the Medicaid managed care report.
We have a draft letter that Paul and Barbara worked on. Marjorie is not here. I was thinking about this, Paul, that if we talk about it as a subcommittee now, whether we could send it up in front of the full committee tomorrow, to have them vote on it. That would be great. That would allow Paul to feel that this information was getting up to the secretary, and Barbara, where it has to very soon.
And Mary, you're here, so we need to hear from you about the process of preparing the Medicare managed care report, and then Kathy's work group. So since Mary is here, why don't we start with hearing from Mary about what she has done to come up with this draft of the report.
A lot of us have not had a chance to read it through yet. So Carolyn and I kind of talked during the break. What we think we'll do is hear from Mary about what she did to prepare this. Then talk maybe on Thursday or maybe a little bit later on about how we might want to go about coming up with the final report based on what we heard from the George Washington people, and what we are going to hear from Mary now.
Then let's move to Paul's letter, and then let's move to Kathy.
DR. NEWACHECK: Before we start that process, can you talk a little bit about how you see all these pieces fitting together in Medicaid, that report, and the three items that GW is doing? Are the three GW items going to appendices?
DR. IEZZONI: Well, we need to talk about that, because frankly it is compelling information. I think the entire report has to be an appendix, because you can't have a report that is going to be hundreds of pages long, as it probably would be, if you looked at what Sara and her group have done so far. But we need to have that be a core portion of the report. How we integrate that is something that we should definitely talk through.
DR. NEWACHECK: So it would be included in the report, but it also might be a more detailed appendix as well?
DR. IEZZONI: The complete report from GW will be appendices to the report. I'm probably not getting the numbers exactly right, but we had talked about the report from our subcommittee maybe being 20 pages or so, with maybe a 5 page executive summary? Does that sound familiar? Twenty single spaced pages with double spaces between paragraphs, and then a five page executive summary. Then having possibly several appendices; the three reports from GW being several appendices, and maybe some other things.
DR. NEWACHECK: My other comment was the timing I don't think would probably work on those. It would be nice to have the development of the model contract provisions and items be based on something other than an ad hoc committee that got together and said here's my favorite six things, or whatever. But rather perhaps to follow some kind of a framework.
If the report were done and could serve that purpose, as a guide towards agendifying the types of indicators that we think are important to include in the model contract, it would give that contract a little bit more legs, as well as give us a way of sort thinking what we do want to include, and what we don't want to include.
DR. IEZZONI: I agree.
MS. COLTIN: I agree also. The way I was thinking of the pieces fitting together was sort of the way we do a gap analysis, where you look at what is it that people say they want and need. That's what we heard in our site visits and so forth.
What have we learned from Sara about what they are getting? What are some of the enabling things, like what do the laws enable them to ask for? Then you have the model contract, here is how you can do it.
DR. IEZZONI: Kathy, I think that we have to ask Sara and Mary to read the transcripts of our meetings, our two site visits, and we had a couple of hearings in Washington this fall. Because gaps were clearly identified during those. Mary has made a preliminary effort to begin to pull that together, but there was a clear sense of where we were heading in terms of gaps.
And frankly, for us to develop our own topology, that is what we were going to base our assessment on as well. And given Sara and Anne's huge experience in doing this -- I don't know if you know what they produced, but they have produced these report that are like this, that you spend $200 for, and you get like a three binder set.
That they should read through the detailed, unedited transcripts -- not the minutes, but the transcripts -- of this.
DR. STARFIELD: Is that part of their contract, to do that?
DR. IEZZONI: It's a lot of work, but we're going to have to think about their contract. I think maybe seeing whether that could be part. We could argue compellingly that that would be part of this process of having the conference calls.
DR. STARFIELD: But on the other hand, isn't it our role to synthesize, to put together the gaps?
MS. ARAKI: I do think that we did include some of that.
DR. IEZZONI: I think that they were supposed to read the transcripts.
MS. COLTIN: I know that they have read the Utah and parts of Boston.
MS. ARAKI: The other thing is what I heard them say, we are extending the terms of this contract already by extending the dates.
DR. STARFIELD: But no costs.
MS. ARAKI: If it's additional costs, we can see that they are.
DR. IEZZONI: The basic question is time from all of us, because we can talk about what would be the appropriate way to do it, but frankly, I had asked Carolyn to set up conference calls in August. Nothing ever happened, because everybody is away.
One of the ideal things would maybe to have another meeting in October, to just spend a full day on this. But is anybody going to have any time in October? Probably not. It's too soon to schedule it.
So frankly, we are kind of slightly between a rock and a hard place on this. We may not be able to do things in the way that might be the best way conceptually, but rely on our contractor, who has tremendous experience in this. Whatever process we can gerryrig together of people like George or Kathy's clone -- I don't want to say Kathy specifically, but Kathy's clone -- or whoever, maybe Paul you would be interested in being on these conference calls, as well to represent the case perspective.
Then folks from SAMHSA and somebody from HCFA on the committee, maybe Olivia, somebody from the Office of Minority Health.
DR. STARFIELD: I think the problem is the framework.
DR. IEZZONI: When are we going to do this?
MS. ARAKI: I budgeted a meeting between now and December, just so you know.
DR. STARFIELD: Around the time of the November meeting?
MS. ARAKI: I budgeted a meeting for you.
DR. CARTER-POKRAS: You budgeted a normal meeting, and then another meeting.
DR. IEZZONI: So you budgeted a second line event?
MS. ARAKI: Yes.
DR. IEZZONI: So a second line has been budgeted. We can fly in another time, but will people have the time in their schedules? That's the key thing. I mean I hear what you are saying, and as an academic, this is not the way that I would do it either.
DR. STARFIELD: Actually, we don't have a third day put on the November meeting.
DR. IEZZONI: No, I don't think there were additional dates put on the November meeting.
DR. NEWACHECK: I guess I wasn't suggesting that we needed to spend a lot of time developing a framework. I was hoping that our report, when we finished it, would provide that framework automatically in some sense, particularly if we thinking about it as we were drafting it. If the timing could work out, then that would be then the time to work with the GW folks to try to identify the measures and all that, based on that report. It would slow the process down of completing everything, because it would more sequenced.
DR. IEZZONI: And you are assuming that we have a process set up for producing our report, which we don't. And we need the time.
DR. NEWACHECK: Maybe that's what the extra day would be.
DR. IEZZONI: Right, that's frankly what it would be. I would see that day being the report frankly. I did not envision it being working with Sara that day.
DR. NEWACHECK: So if that day was working on the report, and coming up with some kind of minimum data set in some sense, but at least a set of things that we felt were important from our site visits and all that. Then that could be used by GW to identify the kinds of items we want in the model contract.
DR. STARFIELD: I wasn't here when -- I wasn't on the subcommittee when the plan for this study was done, so I don't really know what it is we are interested in doing for this report. Are we interested in thinking about it in the context of all managed care? Are we are we in the context of poor populations? What is it we are interested in? Just Medicaid, but you can't interpret that if you don't know what to compare it with.
DR. IEZZONI: Barbara that is what we were referring, what Medicaid managed care. The impetus really came from George van Amburg, representing the states, and Elizabeth presented a very compelling reason. Remember that we came up with this idea to do this in July 1997. So at the time this was a really big issue. It was before the state children's health insurance programs had been conceptualized.
It was also before the balanced budget act had been passed. Basically, what we were hearing from states, and what we were concerned about is that there wasn't a way -- it's exactly what you are concerned in this letter -- that there wasn't a way to track the impact on the health of the people that Medicaid is asked to oversee. There wasn't a way to track that given the minimum data that was going to become available from the states, and from their SCOs.
So it is basically a kind of do what you are doing here in this letter, to kind of raise the visibility of the issue. What we were planning to do was make specific recommendations about: (1) being concerned about the issue; and (2) thinking about who the stakeholders were, what's the different interests of different stakeholders might be. And what data we might recommend that states focus on, so that the federal government think about working with states to focus on, to be able to track the health of people who were enrolled in the Medicaid managed programs.
MS. COLTIN: Also, I think we would be remiss if in the report we didn't address the issue of the sort of federal/state. Clearly, there are reasons why particular states need to go a particular way, but maybe there should be some minimum set of information that you can use to compare, and to look at care across the country. I think right now we don't have that kind of framework at all.
DR. STARFIELD: It's interpreting what you've got. We have no normative standards.
DR. IEZZONI: That's exactly right. It was a time when this whole kind of state versus federal issue was a real kind of tug of war. And we felt that what we wanted to suggest was that there be some sense that there should be some common way to address these issues across the states.
MS. WARD: To me that was the problem statement. Everything that was done since then has continued to verify that there is no message from the federal level that says this is the data you are supposed to collect for the following reasons, and this is how we would expect you to be able to collect it. That's what we heard when we went on site visits, and that's what we just heard today, is that it's still irrational. There are still states who are collecting one thing, and some are collecting a million and one, with no business case for why they are collecting anything.
MS. COLTIN: Medicaid HEDIS was intended to actually provide -- criticize it as much as you like in that it doesn't go far enough, but for what it did do, it was intended to provide some common data across all the states that could measure some aspect of the kind of care that populations were getting. But what has happened is that every state has changed it here and there. They take it and they modify it.
They say, oh, we don't want that to be limited to kids who are continuously enrolled for a year. If they are in for six months, we want to look at it. Then you can't compare states anymore. Now you don't have any common platform.
So I think the federal government is going to have to take some position. If we are going to want to look at some core set of access measures, or quality measures or whatever across the country, it's going to have to be some set that gets measured the same way. Then if the states want to tweak it and look at other things, fine.
DR. IEZZONI: The perfect example is race. We heard repeatedly, especially in Boston, that there was not information on race of enrollees, and therefore, you really could not track access by different race and ethnic minorities, and yet, that is a key policy issue.
Back to the question I asked a few minutes ago, we really need to talk to the group about the content of this report. We don't have time to do it in the next few days. We are not going to have time to do in November. Are people willing to have somebody poll us for our calendars for October?
DR. NEWACHECK: One day?
DR. IEZZONI: One day.
MS. GREENBERG: We actually polled for October dates.
MS. ARAKI: That was for the third hearing.
DR. STARFIELD: Which is not going to be held, is that right?
MS. GREENBERG: No.
MS. ARAKI: Some of you may know you're going to come down here for APHA. That's November.
DR. IEZZONI: There is the November meeting, and then there is the November conference that you are doing.
MS. GREENBERG: The 2nd and 3rd.
DR. IEZZONI: So I would rather see if we could do it in October. I don't have my calendar right now. I think you're going to have to poll us.
DR. STARFIELD: What is the conference at the beginning of November?
MS. GREENBERG: November 2 and 3 is the HIPAA conference, the conference on implications of HIPAA for public health and health services research.
DR. IEZZONI: So, Carolyn, can you work on polling people ASAP, and let's see if we can get a meeting in October, the purpose of which will be to go through what we think should be the recommendations that we're going to want to make to the secretary coming from this Medicaid managed care project.
For those of you who weren't at the site visits, and don't have all the transcripts, and don't have the minutes, or got them and threw them away --
MS. RIMES: The minutes are up on the Web site. If you need the original transcripts, you'll have to yell at me.
MS. GREENBERG: I think the transcripts are up too. We would have put them up when we got those.
MS. RIMES: But I thought we took them down when we got the minutes. It's all up there guys, and it's many pages. We lost half the data on transcripts, and I thought I would die. They all fortunately had notes.
PARTICIPANT: Olivia, I have a question. Since I haven't been at every single meeting in the last few months, has the OPL that came out in April from HCFA to the managed care plans been disseminated to committee members?
DR. IEZZONI: What is an OPL?
MS. GREENBERG: Didn't they withdraw that?
PARTICIPANT: No, they are still working on that. It's internal discussions.
DR. IEZZONI: What is an OPL?
MS. GREENBERG: This was the one about telling people they should not ask race and ethnicity, ask enrollment.
DR. IEZZONI: Oh, that. That was a while back.
PARTICIPANT: That came out in April, but they had reissued a correction on that, and I thought perhaps if we are talking for October, we may want to get a briefing on that, because it may influence the kinds of suggestions you want to make to the Medicaid managed care plans. A lot of the discussions they are having internally certainly go across many of the same issues.
DR. IEZZONI: Well, we should find that out.
PARTICIPANT: When do we want the managed care plans to claim this information? How is it best used? How can we get at the data that we need for some of the reporting and monitoring?
DR. IEZZONI: Okay, well, we should find out about that.
MS. COLTIN: It may not be as big an issue for Medicaid if most of the Medicaid enrollees are actually being enrolled by the state, or by the state's agent, and then the data are transmitted to the managed care plan. Because the OPL only addressed what a managed care plan could ask of a prospective enrollee. So if the state Medicaid agency is asking that of a prospective enrollee, the OPL doesn't cover that.
DR. IEZZONI: What I'm hearing from Sara is that you can't link the information.
MS. COLTIN: I don't see anything in the OPL to the effect that a data set couldn't be passed from a state agency to a health plan. It was the health plan asking the enrollee, because it was focused on Medicare.
PARTICIPANT: We were suggesting that the language be strengthened to say that it encouraged the plans to work with the states or whatever, to analyze these data. The other piece was that the office had a lot of very good experience where they have actually collected information on applicants for particular programs on the race/ethnicity of the applicants, but on a separate piece of paper. So that was separated from the application, and so you could look at the racial/ethnic distribution of your applicant pool versus the race and ethnic distribution of your enrollees to see if there tends to be a big discrepancy.
If there is a big discrepancy, that could be a red flag then to do a further investigation. This is a lot of the discussions they are having now with the office, how to revise that OPL. That discussion may be helpful.
Olivia, why don't we work on that? That will be a really good thing, because that's real positive.
DR. IEZZONI: I'm looking at the clock. We have a lot that we have to do. We basically agree that we need to talk more about the Medicaid managed care report, which we have known all along. What we are going to try to do is come up with a date in October. Lynnette has just gone to try and get my calendar. So if other people have their calendars available.
I really apologize. I feel that it's so important to have both of you actually physically here, or else I would say we could maybe conference you in, or telephone conference you in. Would that make it easier for you?
MS. COLTIN: Could we meet somewhere else that is more central?
DR. IEZZONI: If George is still participating in this, since he was so integral to getting us going on this, it would be really great to have him be able to fly in.
So Mary has made a very first initial kind of cut. Mary, do you want to tell us what you did.
MS. MOIEN: I just was going to go back a little bit to the beginning, since there are a couple of new people, although I wasn't involved in the beginning, which as you all heard was last July, apparently into the beginning of the year, right before the Phoenix. You all decided it would be worthwhile I guess to have a small contract to write up a report talking about data needs from Medicaid managed care.
So I became involved immediately before the Phoenix meeting. So I was participating in the Phoenix meeting, and in the Boston meeting. Now after the Boston meeting we stayed for an extra half day to discuss and outline and design of the report, which I could tell already people are really thinking differently along.
My report really emphasizes describing and grouping stakeholders, and identifying and grouping stakeholder needs, as well as summarizing people who were at all of the different meetings, using the syntheses from the earlier meetings, et cetera, and to the synthesize individual people who made presentations.
We were going to meet in June to go over some of these issues, then the subcommittee got involved in the insular meeting. So we actually never met again. Carolyn and I have talked a couple of times, and kind of updated some outlines, and all based on the meeting in Boston. And then when nothing else happened, I just decided to start writing, because I had a early August deadline on my contract to get the first draft completed.
Carolyn and I met after I sent her the draft, and we talked about some additions and revisions. She has been away since then, so I haven't even gone over with her, the changes that I made, some of which we had agreed on, and some of which were kind of additions.
But I had tried to write it in the vein that I thought we had discussed after the Boston meeting. In fact, it seemed to me that at the time that we had meetings I think here in March -- in Washington in March I think we met -- and then in Boston, that at that time I think the idea was not to have any information in the report in terms of specific data items needed, but just to talk about the fact of who needed data, and what data was being used for.
I was hoping that since we gotten a e-mail fairly early last week, that people would have had a chance to kind of look it over. The main thing that we had not discussed at all in Boston was recommendations, which we thought we were going to be discussing in June, because I thought that some of the reading and the background had been kind of decided on.
So between Carolyn and me, and some of the stuff was already written up, we attempted some possible recommendations that were in here. I was looking for guidance in that direction. In addition, I had just put in some comments that the GW findings were not available yet. I think they were originally going to be available for me to finish it up in August.
I guess my gut feeling is I can go through the report, but if nobody has gone through the report, I don't know that there is any use to go through it.
DR. IEZZONI: No, I don't think we should go through it at this point, unless people have very specific things. My e-mail system had a melt down last week briefly, so I actually haven't looked at it either.
DR. STARFIELD: That's not why I didn't get it.
DR. IEZZONI: You didn't get it at all?
DR. STARFIELD: I didn't get it, and my e-mail was functioning fine.
DR. NEWACHECK: I read it, and I did have some comments. I'm just trying to think about a process so that we don't have to wait until October or some other date to continue moving along.
MS. MOIEN: I don't think I would be doing anything between now and then.
DR. IEZZONI: What I would like to do is remember Item 13 of this morning? We're not going to call it a meeting, we're going to call it a discussion. I would like to have a discussion off line about how we're going to staff and the timing, and the preparation of the final report, because I would like to involve Marjorie. There are just going to be issues about whether there is money and timing and all that kind of stuff.
So what I would like to do is defer a final kind of decision about the mechanics of getting this written until we've had number 13 discussions among ourselves. But what you should do is plan to hold me to by the end of Thursday that we kind of have a plan figured out.
DR. STARFIELD: This report, Mary, is mostly about the hearings, right?
MS. MOIEN: Yes.
DR. STARFIELD: So we've got the GW stuff. We've got the hearings, and now we've got to put them together.
MS. MOIEN: Although at that time the idea was just to take a few -- I thought, it was my impression -- general summary statements from the GW report, filter them in, and having them be attached as their own.
DR. STARFIELD: I think the subcommittee is going to have to do more than that. I think we're going to have to do a lot of synthesizing.
DR. IEZZONI: So why don't we --
MS. GREENBERG: We're clear that the GW thing is not going to be over real soon. All to the good that they want it to be a credible document and contribution, but obviously people are going to have to work with them.
DR. IEZZONI: Over the next two days let's all try to discuss off line, what people's availability might be. I have volunteered Kathy, and I volunteered Elizabeth, and we need to talk about Mary and about contracts and money and time and everything. So let's try by the end of Thursday to have a plan for the mechanics of preparing the report.
In terms of comments to feed back to Mary at this point, Paul, are your comments best communicated verbally? Do you have them in writing?
DR. NEWACHECK: I can write them down. I can send you an e-mail or something like that, if that makes more sense.
DR. STARFIELD: Do you want to send them to all of us?
DR. NEWACHECK: I could do that, just cc everybody.
DR. IEZZONI: All right, great. So, Mary, thank you very much for making this first call. What we will do is we will talk among ourselves and figure out how we are going to make this happen. Carolyn and Marjorie will be working with you about where we are going to be headed from here.
If people have specific comments on this version, as Paul does, e-mail them.
Does that sound like an okay plan for the next couple of days?
So Paul and Barbara -- does everybody have a copy of this draft letter? What we would like to do is make a decision about this during the next few minutes. Do you think Don will come to this committee meeting tomorrow?
MS. GREENBERG: Sure. When you give your subcommittee report, this can be part of your subcommittee report.
DR. IEZZONI: As an action item?
MS. GREENBERG: Right.
DR. NEWACHECK: Even if nobody has seen it, they would vote on it?
DR. IEZZONI: Circulate it.
MS. GREENBERG: What you might want to do is circulate it in the morning.
DR. IEZZONI: So, Paul.
DR. NEWACHECK: Maybe I can just say a few words, and then we can take two minutes to read it, because it's just a very short letter, and see if we want to agree do we want to do it as a group.
This is something that Barbara and I came up with in response to this new program, the State Children's Health Insurance Program, which was enacted as part of the balance budget act last year. It gives states an opportunity to expand either their Medicaid programs, or create new programs for low income children. There is a fair amount of federal money available, $20 billion over 5 years, and up to $40 billion over 10 years for this purpose.
The balanced budget act did not include any money to evaluate this program, or to collect data on how the program works, and states have a huge amount of leeway. There is tremendous discretion in eligibility, in benefits, in cost sharing, the works, but there is no data collection system in place to find out which programs work, to establish whether the states are being accountable in how they are using their money.
So what we did is we wrote this letter, especially making an argument for the need to have some kind of a system in place to collect information on the State Children's Health Insurance Program that would be useful both at the federal level and the state level, for accountability purposes, for program monitoring, evaluation purposes.
And we suggested one specific approach to doing that through using the SLAIT survey that has been developed by the National Center for Health Statistics. That's the State and Local Area Integrated Telephone Survey. NCHS is ready to go in the field with it almost. They have the mechanism in place. They have developed a draft survey instrument that can be used to collect baseline information in all the states, and then can be used perhaps later to collect follow-up information down the line.
So this letter basically makes the argument that that's a useful approach given the absence of other data at the state level, to get the information we need to assure accountability and performance of the plans.
So maybe everybody could take just a minute or two to read it.
DR. IEZZONI: Can you tell us a little bit more about the plans? Because you say that it includes immunization information, but don't they use laws a lot more?
DR. NEWACHECK: Well, no, but that's what it is being used for right now. It is being used for immunization surveillance, but it could be extended.
MS. ARAKI: Is that the one that is now being piloted in Texas?
DR. NEWACHECK: I'm not sure. It's being piloted in two states for other purposes. It is being used --
DR. STARFIELD: I think it will require some modification, but the basic survey is there.
MS. GREENBERG: You have to screen a large number of households in order to find two year old children.
DR. IEZZONI: I think that that needs to be made clearer, because when I see a survey, I think that there is a content.
DR. STARFIELD: We say with appropriate adaptation. That's why we put in that phrase.
DR. IEZZONI: I know, but for people who don't know anything about surveys, it would be important to say that this is a mechanism that allows you to screen and to identify households that are appropriate to ask these question to; to make it more mechanical.
DR. NEWACHECK: Now if we change it at this point --
[Discussion regarding how changes will be made to the document, and then dissemination of the revised document.]
MR. HITCHCOCK: This approach with this survey has been discussed a lot, and there have been issues that have been raised that are difficult ones. The issues of precision for instance, of the estimates that you would get given that the impact of the program is not supposed to be really huge. What size sample would you need to detect changes from year to year?
DR. STARFIELD: We've already done actually some work on that. It seems to me to make this at all efficient or workable, you would have to target areas that have a proportion of people likely to be eligible for the program. I don't know when we say SLAITS here, whether we mean the staff they've got already, which probably wouldn't be adequate for this.
MS. GREENBERG: There have been months of discussion. The thing that concerns me maybe, and obviously NCHS is very interested in SLAITS being used productively I know is that I don't think the subcommittee has gotten any kind of a report from the group that was set up to look at this problem.
DR. STARFIELD: An interagency group, is that what you mean?
MS. GREENBERG: You've been engaged in that.
DR. NEWACHECK: I've seen the report that came out of AHCPR that done with MKHAS staff.
MS. GREENBERG: The AHCPR report?
DR. NEWACHECK: Well, they did their own analysis of sample sizes and things like that.
MR. HITCHCOCK: Lisa Simpson I think chaired the group.
DR. NEWACHECK: Right, it was her staff.
MR. HITCHCOCK: I think the consensus of the group was that the SLAIT is the best mechanism.
DR. NEWACHECK: It's not cost effective.
DR. STARFIELD: That doesn't imply a sampling plan I don't think. I don't think we meant to imply a sampling plan.
MR. HITCHCOCK: You don't really need to necessarily either. Just understand that there are some methodological problems. I would add a sentence to the bottom that is one of my favorite things when it come to survey integration. I don't think we can do too much more to existing surveys that we have, but rather than do a brand new survey to measure something, I think it's completely within the spirit of survey integration to fold something into an existing survey.
So I would put at the bottom of that third page, a sentence to the effect of something like the use of an existing multi-purpose survey, rather than the development of a new costly survey with a single focus is entirely consistent with the department's survey integration plan. I'll give you those words. I think survey integration is a buzz word.
MS. GREENBERG: I think though that this is not a low cost option either.
MR. HITCHCOCK: It's much lower cost than --
MS. GREENBERG: De novo.
MR. HITCHCOCK: Yes.
DR. CARTER-POKRAS: Did they work out the funding situation yet?
MR. HITCHCOCK: No.
DR. CARTER-POKRAS: It's still on the table then.
MR. HITCHCOCK: That's the purpose of this letter really.
DR. NEWACHECK: To encourage that.
DR. CARTER-POKRAS: They've had so many data counts and presentations, I thought it was included in that $35 million.
MR. HITCHCOCK: There is some money in the 1999 budget for it, but not in 1998. I don't remember how it works.
DR. CARTER-POKRAS: There is $35 million in health statistics initiative that NCHS has gotten approved. I thought that was included in that.
DR. NEWACHECK: Is this recent?
MS. GREENBERG: That's for 2000.
DR. IEZZONI: This would be something that the states would pay for out of their -- it's funding that the federal government give?
MR. HITCHCOCK: Right.
DR. CARTER-POKRAS: That was part of the point when they were going through this, because they discussed perhaps the behavioral risk factor survey system that in many cases the state helps to pay for. And they thought that you would have greater consistency if the federal government was responsible for the data collection.
MS. GREENBERG: Of course it would be too late for tomorrow, but you are going to have Ed Hunter at your meeting on Thursday. He could certainly tell you what the status of this is.
DR. IEZZONI: Do we feel conceptually that this is the right way to go? I like the fact that it was so specific. Usually when we write these kind of letters to the secretary from our committee -- other subcommittees are more specific than we are. So that's why I really like the last paragraph. So are there any other words that people would want to change before we bring it in front of the full committee?
Who is going to be making these edits? Paul?
DR. NEWACHECK: Well, I have this one.
DR. IEZZONI: Okay, because Olivia is about to make a couple of suggestions.
DR. CARTER-POKRAS: I just had a couple of suggestions. It's hard for me to visualize how long this is going to be single spaced. But I don't know if you want to have a recommendation out or some way that it jumps out at you if you just picked up the piece of paper and glanced at it.
DR. NEWACHECK: Like a subheading or something?
DR. CARTER-POKRAS: Some people do that. They say recommendation in bold letters, and then they spell it out, and you know. Then you have the rationale following it. It's a little easier to find.
MR. HITCHCOCK: We can take care of the formatting. That doesn't need to be decided tomorrow, right, the formatting? It's the content.
DR. STARFIELD: Well, we ought to identify the recommendation at least.
DR. CARTER-POKRAS: One thing I was going to suggest regarding the SLAIT, I know it's a concern of those of us who have used some of the telephone interview data, even though it is going to be supplemented with estimates to kind of get at those folks who don't have telephones, we may want to specify that in particular.
Because when we are talking about eligible kids for Medicaid who are currently not covered by Medicaid, who are not in the Medicaid program, but are eligible for it, we've got a lot of kids who are like Hispanic, low income groups that are hard to reach, because they don't have telephones. So we may just want to just spell that out. Make that a statement when we talk about to adapt the place, we may be want to say what we mean by appropriate adaptations.
DR. NEWACHECK: The way I wrote this, with Barbara's input was more to make a general point that the committee endorses the concept, but not get into specifics, because once we get into specifics, there are all kinds of different views.
MR. HITCHCOCK: HHS has developed the questionnaire and the study, and some of the methodological issues associated with this.
DR. NEWACHECK: I think that's an important issue for implementation. There are lots of other issues, like the sample size issue, which we have not touched in here, because they sort of raise issues that I think go beyond what this committee would be good at.
I mean if this committee is saying well, look, we are people who are experts in the data field. We feel that there is an important gap here. We are recommending that you fill it in this kind of way without going into a lot of detail about the specifics.
MS. WARD: I would have something in there that talks about the exact implementation issues that would need to be addressed later or by group X.
DR. NEWACHECK: We could do that. We could also offer to provide that assistance too. Maybe that would be a better way of doing it.
MS. WARD: Once we sit down and think about all those kinds of implementation issues and the budgets that go along with that.
MR. HITCHCOCK: A lot of it has been thought about.
DR. CARTER-POKRAS: It's been discussed at length, many places.
MS. WARD: It may be that just that we need to as a committee, know that the issue of paying for oversampling at local county areas has been addressed, and there is a plan. I don't need to talk about it any more it someone says to me, that's been thought of, and that is going to happen.
MR. HITCHCOCK: The committee needs to just lend its weight to the overall discussion and momentum to get this thing going.
PARTICIPANT: There is rather an ambitious list of objectives here at the bottom of page 2. Is there basically an implication here that SLAITS is going to meet all those objectives?
DR. NEWACHECK: Well, it kind of suggests that. I think it would help provide some information on all the things that are listed there, but not necessarily the level of detail that somebody might expect. You sort of have to read between the lines I think to decide.
DR. IEZZONI: Then you might want to rephrase things as SLAIT being a place to start.
DR. STARFIELD: I'd like to add the word access to appropriate care in there, instead of just access to care.
DR. IEZZONI: Okay, so, Dale, you're going to be getting this electronically on your e-mail system.
MR. HITCHCOCK: Which is not working at the moment. I'm the only one in my office where it's not working. Any minute I expect it to be up.
DR. IEZZONI: So you and Paul can have one of these Item 13 meetings as well.
DR. STARFIELD: Are you going to do that right after this meeting?
DR. IEZZONI: So are there final comments? Anybody around the room have any suggestions? Thank you, Paul and Barbara, for doing that. I think it will be great to have this to the secretary ASAP, because things are really starting to happen.
Kathy, we do have on the agenda, the work group on quality charge and work plan. But the work group is not this subcommittee alone. It's a bunch of other folks too, right?
MS. COLTIN: But most of the people who signed up are on this subcommittee also. That's where the great overlap is. It's you, Elizabeth, Barbara, Vince.
Why don't you see if John would be willing to come, and why don't we talk about calendars for October?
[A general discussion of meeting dates.]
Why don't we say a possible date would be the 20th. Is the 21st possible for people? No. Is the 22nd possible for people? No for Paul. So the only day -- I'm avoiding Mondays and Fridays. So the only day of the week is October 20th. Then the week of the 26th, is the 27th absolutely out?
DR. STARFIELD: Yes.
DR. IEZZONI: So the 28th, which is Wednesday. Paul cannot do that. And the 29th, which is Thursday.
DR. NEWACHECK: I don't think so. What about the following Tuesday?
DR. IEZZONI: That's when the HCFA conference is taking place. Friday, the 30th? Friday the 30th is a maybe for Paul, and Kathy doesn't know. It's okay for Barbara. It's okay for me. And Elizabeth doesn't know.
MS. WARD: I can bring my calendar tomorrow. I'll have my office fax it to me.
DR. IEZZONI: Either the 20th or the 30th of October are the two dates. Lynnette, Vince needs to be here.
Why don't we turn off the tape, just for a second.
[Whereupon, taping of the meeting was temporarily stopped.]
DR. IEZZONI: At this point we're back.
MS. GREENBERG: But also at this point we do not have lead staff to replace the other lead staff person.
MS. COLTIN: So I guess that's the first issue that we need to raise.
MS. ARAKI: I don't think it would hurt to mention it in your report.
DR. IEZZONI: All right, Kathy, do you want to take the lead on this?
MS. COLTIN: I have a question. Has the subcommittee charge and work plan been fully decided?
DR. IEZZONI: We're talking about that Thursday.
MS. COLTIN: There is a portion of the full subcommittee charge and work plan that includes the work that is going to be done under the quality. I didn't know whether you actually wanted a separate charge, because I think what is written here really covers what I said at the last meeting.
This is the 29th one that was e-mailed to us. I can read it to you. It's very brief.
DR. IEZZONI: Why don't we have copies of that for tomorrow and for Thursday.
PARTICIPANT: It's also the issue, Kathy, with the year 2010 workshops too, because they were all muddled in together, because I didn't know whether to separate them or not.
MS. COLTIN: Well, I think that's going to be the first order, is to review the goals and objectives for the work group, and make sure that we are all in agreement about what those are. Whether we do that as a separate charge for the work group, or whether it is part of the overall charge for the subcommittee under its subheading of quality of care matters not to me, but I think either way we need to resolve that.
MS. GREENBERG: As a work group it does need some charge. That doesn't mean it can't be identical to a piece that's in your overall charge.
MS. COLTIN: There are essentially two goals and objectives that I had identified at the last meeting. Both of them grow out of an analysis of the recommendations that were made by the president's advisory commission. Basically taking each of those recommendations, identifying what the data issues are underlying the ability to achieve whatever is being recommended.
Then assessing whether or not our current data systems are up to the challenge. Whether or not in fact we have the data, and if not, what we would need to do to improve, whether it's national information infrastructure, whether it's state level data, whether it is private sector data, but what it is that might have to be done in order to produce whatever information would be necessary to try to realize the objectives as they are laid out in the advisory commission's recommendations. So that was one thing.
There are a lot of recommendations that read like improving the ability to measure and monitor errors in care. Now there is an awful lot of data issues that underlie that kind of a recommendation, so it really means talking about what kind of data would you need to be able to monitor and improve error rates? What kinds of error rates are you talking about? What kinds of data are available now? Where do they come from? What are the strengths? What are the weaknesses? What might need to be done? So that's an example of just what I'm thinking about.
MS. GREENBERG: I should say too in fairness to AHCPR, although there has been the staff turnover, these are critical issues I know to AHCPR, and very high on the AHCPR agenda. So I'm sure we'll get the support that we need.
DR. IEZZONI: It's about to be changed to the agency for health care quality or something like that.
DR. CARTER-POKRAS: If it passes.
DR. IEZZONI: What is the name supposed to be, Jim?
MR. SCANLON: The last I heard is what you indicated, agency for health care quality.
DR. IEZZONI: And they are spelling health care as one word these days?
MR. SCANLON: There is an editorial debate over that.
MS. COLTIN: Then the other objective had to do with taking a look at the informational systems road map that had been developed by NCQA, and taking a look at sort of melding together what we came up in reviewing the data needs with respect to the advisory committee's recommendations, and then seeing whether the informational systems road map recommendations go far enough, and how would we modify those in terms of what needs to happen with the private sector in particular.
Because those recommendations were aimed mostly at managed care plans, but managed care plans get their data from other providers, so I think that there is a ripple effect. So it would be a question of looking at that road map from the standpoint of the private sector, and what kinds of data improvements might need to be made within the private sector to help.
So I think we would be addressing the public sector issues in more detail in the discussion of the recommendations of the advisory commission, but where we identify that the data need to come from the private sector, I think we would try to do it within the model of the road map, and say how might that be changed, because I think that provides a nice model to start with and work from, because it conceptually kind of lays out I think, what the key issues are.
I don't think the key issues in managed care plans are that terribly different from what they are in other provider settings. So that was the way I was envisioning the work of that work group proceeding.
Discussion?
PARTICIPANT: It seems to me that there is an issue that is not addressed in that. First of all, I kind of kept on hearing you use the word "information systems" in the sense of automated information systems.
MS. COLTIN: Not necessarily.
PARTICIPANT: A lot of quality stuff that is collected -- well, first of all, one point is that a fair bit of the quality information that is collected, is collected based upon surveys. So not only looking at how those systems are put together, but also standardization. Particularly if you have some of the rarer diseases, with standardization and methodology, your risks of being surveyed every year go up significantly, particularly in smaller managed care plans.
The second has to do with data collection at the provider site, which a provider who may be related to three or four managed care plans may give three or four different forms, i.e. different formats of collecting the same information. And so looking at some ways of standardizing the data collection. I don't know if that is included within those two charges?
MS. COLTIN: I think that within the information systems road map there are a lot of points that address that, that deal with things like lab results data and so forth. I think there is going to be a lot of overlap when we start talking about data coming from the private sector, with the work that is going on in Simon and Jeff's around the computerized patient record.
There will be some overlap with what is going on around the K2 standards, and how some of the data elements and definitions that are part of those administrative transactions might also apply to other types of transactional data, whether it's automated birth certificates, or other kinds of data that use those same data elements.
Or even surveys in terms of whether -- is payer ID going to become the field that is used in some of those surveys of managed care plans, or collecting information about what kind of insurance people have. Or is the employer ID going to be used in the surveys of employers and so forth? So I think some of those issues will come up, and there will be those overlaps.
Hopefully, the work of the different groups is progressing at different paces, and by the time we really to dealing with the private sector issues, those groups will be far enough along that we can draw on a lot of what has already been recommended.
DR. IEZZONI: So, Kathy, it sounds like a great idea to me to build off of those two documents. I'm just thinking through how hard it has been for us to find the time to --
MS. COLTIN: To work. That's a concern for me too.
DR. IEZZONI: On Thursday we are going to be as a subcommittee, coming up with our next big initiative. The Medicaid managed care initiative is still in progress, but for all intents and purposes it is winding down. We are going to need to start to think about a new, big initiative starting in the winter or next spring.
MS. COLTIN: Can this be it?
DR. IEZZONI: I still think that our subcommittee is charged with looking at vulnerable populations and minority issues, and kids. The post-acute care sector, which has --
MS. COLTIN: All of those issues are captured in the president's advisory commission recommendations. There is a set of recommendations around vulnerable populations.
DR. IEZZONI: Maybe we could say that we would start there, because it's going to be a lot easier to come up with quality data for the commercially insured work class adult.
MS. COLTIN: But are you seeing the charge of this formal work group as having to relate only to special populations?
DR. IEZZONI: Not at all. My point was going to be though that I think we are going to need to have an initiative that is specific to kind of the more population-based focus of our subcommittee. Our work plan cannot be "just" quality.
DR. STARFIELD: I guess I don't see that. I agree about the population focus, but why can't we look at quality from a population focus, not from a managed care?
DR. IEZZONI: Well, let's talk about it on Thursday. There are a lot of competing issues I think that different members of the subcommittee -- like David Takeuchi isn't here right now, Vince Mor isn't here right now. So what I hear in my one-on-one type of discussions is that there are other people on this subcommittee who might want to focus on slightly different things.
If a broad rubric is quality, within quality they would want to focus on things that might be more specific to where what's near and dear to their hearts. So the basic point is that on Thursday we need to really figure this out, and come up with a work plan, realizing that John is not on our subcommittee. He has important responsibilities in other places. So how do we deal with a work group that -- are you the only person that is not overlapping?
MS. GREENBERG: Yes. All the others are on the population subcommittee.
PARTICIPANT: I'll try to give it quality time though.
DR. IEZZONI: It's the charge, as well as the work plan.
DR. NEWACHECK: I'm concerned that as a relatively new member to the subcommittee, that we have so many things going on, to do them well is very difficult. I think we have seen that with Medicaid. Doing that well is very hard, and even with site visits and staffing and all that, it's still been a difficult task. I think if we put too much on our plate, we are going to be in trouble.
I'm not commenting on what we should put on our plate, but I think we need to be careful about the quantity.
DR. IEZZONI: We need to on Thursday, talk this through.
MS. COLTIN: I think we can look for synergies. It seems to me as I was listening to the conversation about the Medicaid managed care report, and what a lot of the issues are that the stakeholders did identify in our hearings, a lot of them revolved around access and other types of quality issues. When I think of quality, I don't just think of clinical quality, although I certainly include that, but I think of access issues as quality issues, and certainly satisfaction or acceptability issues as being part of that as well.
So I think that there is actually a nice opportunity to transition from the work of the Medicaid manage care around vulnerable populations, and what the quality issues are, try and link that work and what we have learned there with some of the recommendations from the president's advisory commission about vulnerable populations, and use that as a springboard to then go on and look at some of the other issues.
Because there are recommendations there about what the quality issues are in long-term care, and other kinds of care settings that might be of interest to others who have ideas about other agendas. I think there are ways we can piggy-back among this work.
DR. IEZZONI: Let's talk about this on Thursday. And keep in mind that the way that our subcommittees kind of operate is that we have a great, big project. Last year it was Medicaid managed care. But we also have to monitor certain things that are happening. Like we have to monitor what is happening with the change in the definitions of race and ethnicity. We have to monitor the Social Security Administration's change in definitions of disabilities.
So the way that we have tended to function in the past, just for those of you who are new is that typically in our breakout sessions we will hear from issue specific presenters. That is not true at this one, but in the past it has worked very nicely. We had a very nice couple of presentations back at the June meeting, both from PRO from New York and from the Bureau of the Census and OMB about the race and ethnicity.
But then we have these separate days that we block off for hearings or meetings about the big topics. I think we are going to have to think about how we are going to schedule.
MS. GREENBERG: This may be a dumb question, because although we are talking about two lists, we are talking about all the same people except for John. There is the rub.
DR. IEZZONI: Is Dan also on the quality work group?
MS. GREENBERG: No, Dan is not on that. He's on the 21st century one. I think what you need to come back to the full committee with is, there actually is a mandate for -- there was an agreement that there would be a work group on quality. Maybe I'm not hearing you right, but I hear you saying that whatever that work group on quality wants to do, has got to be a piece of the subcommittee's specific agenda.
DR. IEZZONI: No, I didn't say that. I said that on Thursday we need to talk about it. We need to talk through how we are going to integrate everything, and thinking about the staffing and so on. I don't have any preconceptions. All I'm saying is that because there are certain items that our subcommittee does like monitoring the race and ethnicity, when we have breakout sessions at the full committee meetings where John doesn't need to be there, because he's elsewhere in other breakouts, that we should continue to monitor the things like substance abuse issues, that have been a traditional focus of this subcommittee.
And then think about other dates for meetings where we do something where John might come in, and would not come in.
MS. GREENBERG: You may need to have either separate working sessions and separate meetings that would not require all the members of the subcommittee, just the members of the work group.
MS. COLTIN: I think we're going to have to recognize and try to balance the work load, because we are not going to be able to mount any kind of an effort around the issues that the quality work group will be addressing if people's time and effort is totally absorbed with subcommittee work in another area.
MS. GREENBERG: You're right.
MS. COLTIN: So as I'm looking at the next couple of months and saying, okay, you are scheduling a meeting in October around the Medicaid managed care. It looks like the November meeting is also going to have to focus in large part around that issue, as well as planning.
It seems to me that realistically, we couldn't even think about really starting much of an effort in quality of care until December, or maybe even 1999. So we need to be realistic in setting up a work plan, because it is the same the individuals.
DR. STARFIELD: There are people on the subcommittee who are not in the quality work group. To the extent that they are part of some other work group, we could meet in parallel actually.
MS. GREENBERG: It's complicated though. As Don keeps telling me, that's life. He's not overly sympathetic to people having conflicting demands on their time.
DR. STARFIELD: I don't think Hortensia, for example, signed up for the quality and Dan didn't, and I guess you didn't either.
MS. GREENBERG: I know that we tentatively had on the agenda for the November meeting, a session to follow-up on the session that we had on quality of data for looking at quality of care, to actually have this work group help us put together a session on quality of care. This might be premature is what I'm hearing.
MS. COLTIN: Well, I don't know, because I don't know when we are going to have the time to even do the planning for that, the way the schedule is shaping up.
DR. NEWACHECK: It seems like we have probably enough to do for two or three meetings at least, just to catch up to where we would like to be on Medicaid and the territories and some of these other items.
DR. IEZZONI: I need to talk to you about the territory staffing.
DR. NEWACHECK: Does it make sense to kind of defer a decision on this for a while until we sort of catch up with the back log?
DR. IEZZONI: Well, I think that it makes sense to defer beginning work. I don't think it makes sense to defer a decision, because we need to start scheduling time. It's a miracle that we found a date that most of us could come in October. That is truly miraculous. Somebody is looking out for us here.
But I think that the notion of beginning to put calendar dates for 1999 for subcommittee work, and beginning to plan, because just talk to Carolyn about what we do -- you haven't been involved in the hearings of our subcommittee except the Boston one. But you have to plan for those months in advance. So it we wanted to have a meeting that was a day and a half of hearings, then half a day of subcommittee deliberations, we really need to give staff at minimum, three months notice, to be able to do that. So I think it's very reasonable to start planning.
DR. CARTER-POKRAS: I think it will help if there is some discussion about a final charge and work plan, so there is some kind of framework to work from too. And the specifics I think will fill in when some of the other details get cleaned up a little bit.
DR. IEZZONI: I detect tremendous fatigue around the room. We have worked hard this afternoon. It feels like that to me. Does it feel like that to you guys? It's been a long day.
Kathy, would it be okay if we adjourned at this point? Do you feel that you got what you need for Thursday?
PARTICIPANT: I'm not going to here for Thursday, because we've got RWJ coming in for a site visit. It may be that once the work of the work group is underway, and a framework for quality is developed, and how we want to address the issue, it may be easier to deal with certain pieces of it that are related to vulnerable populations, and other activities as we try to address the whole quality pie.
So I just want to hold that out that maybe we want to look at an evolving process, where it may go through the work group, and then back to the full committee for certain pieces of it that are more consistent with the full committee's charge. Then I'm ready to adjourn.
DR. IEZZONI: The committee is officially adjourned. Thank you everybody.
[Whereupon, the subcommittee was adjourned at 5:15 p.m.]