Hubert H. Humphrey Building
200 Independence Avenue
Washington, D.C.
Call to Order, Introductions - Dr. Detmer
Panel on Data Quality Standards
Reports on Subcommittee Work Plans and Recommendations
Panel on Standards for Computer-Based Patient Records
Future Meetings and Agenda Topics
Reports on Subcommittee Work Plans and Recommendations
DR. DETMER: What I would like to do first is, you have in front of you the drafted letter that we worked on last night. I would like to give you a couple minutes to read it and then we will see if we can't complete our wordsmithing on it and then move on to our subcommittee reports, before we get to our panel.
(Committee members peruse letter.)
MS. GREENBERG: The format, we didn't quite get it right, so don't worry about that.
DR. DETMER: Was this in here before? This wasn't in here, was it?
MS. GREENBERG: Yes, it was in, yes.
MR. SCANLON: Interesting, that is interesting apropos what we discussed this morning.
DR. DETMER: All right, are you ready to discuss it? I will wait for a minute. How much do we need to do on this, exactly?
MS. GREENBERG: I don't think much. One thing I did bring was the tentative planning for annual report.
DR. DETMER: It is called the biannual report, I guess.
MS. GREENBERG: Because it probably won't continue to be a biannual report.
DR. DETMER: Lynette, excuse me, how are you coming? How close are you to being finished?
MR. BLAIR: You don't need to wait.
DR. DETMER: I may go ahead, if it is okay, Jeff, and then you can let me know -- you can keep reading, if you wish, or do you want to hear our conversation.
MR. BLAIR: I will just listen. I have confidence.
DR. DETMER: Then, Jeff, if you have questions, we will pick them up and play them in.
Okay, I think what we have tried to do here -- and we may still have to reformat it slightly -- but what we really are wanting to say is, do we have sort of the substance that we were trying to capture with our discussion yesterday? Are there any concerns or comments?
What I have is that the last sentence before chapter one, the committee is concerned that some services may be devalued under this evidence base -- I would say is a narrowly defined evidence based standard.
Otherwise, I am at peace myself. Are there other questions or comments?
MR. SCANLON: A minor point. In the second paragraph it notes that 41 million people are outside the insurance system. They are also outside the reach of most data systems.
That is true for clinical and administrative data systems. It is not true for population-based surveys for data surveys for data systems. That is how we know how many there are. Could we say outside the reach of many, or make it more specific?
MS. GREENBERG: Or clinical and administrative.
MR. SCANLON: Or clinical and administrative. I mean, that is the numerator.
DR. DETMER: Clinical and administrative; okay.
MR. FITZMAURICE: I wonder also whether this should be at least a copy or maybe co-addressed to the co-chairs of the President's Commission?
MR. SCANLON: That is an excellent idea. That is a superb idea.
DR. DETMER: Yes, addressed to.
MS. GREENBERG: I cc'd the co-chairs of the data committee.
DR. DETMER: They were the chairs.
MR. FITZMAURICE: Yes, Secretary Shalala --
MR. SCANLON: That is an excellent idea.
MR. FITZMAURICE: The Secretary of Labor, the Secretary of -- the co-chairs of the commission.
DR. DETMER: Other ideas? If not, is there a motion to accept this?
DR. IEZZONI: I move that we accept the letter and send it forward.
(Motion seconded.)
DR. DETMER: Any further discussion?
All in favor say aye?
(A chorus of ayes heard.)
DR. DETMER: Opposed?
Abstentions?
(Voice heard in abstention.)
DR. DETMER: One abstention. Okay. Now, Paul Tang is not yet with us. So, let's go ahead and head on into our subcommittee reports. Bob?
Agenda Item: Reports on Subcommittee Work Plans and Recommendations. Subcommittee on Privacy and Confidentiality.
MR. GELLMAN: Okay, at the end of January, the privacy subcommittee held two days of workshops. One was on what is an identifiable record and one was on disease registries. I am going to offer some sort of conclusions from all of this.
I think the discussions were very interesting, especially the day on identifiability was quite eye opening.
Distinguishing between identifiable and non-identifiable records is really at the heart of every legislative proposal, including the Secretary's, with the idea being that identifiable records are regulated and non-identifiable records aren't.
That is very nice, but the distinction between identifiable and non-identifiable records has become much harder to make.
Whether something is non-identifiable depends on what else you know, and the amount of personally identifiable information that is available from public records or commercial sources has grown significantly in recent years.
Trafficking in personal information has exploded in the last decade, all fueled by cheaper computer power, the internet and the commercial sale of personal information.
We had a very stunning demonstration, or presentation, by a woman from the computer science lab at MIT, Latonya Sweeney.
In the materials, in the booklet that has been given out, her web site address is included there. I urge anyone who hasn't seen some of her work to go take a look.
Basically, when you listen to her stuff -- and this not theoretical activity; she has gone out and done this -- almost nothing is non-identifiable. You can link everything up.
She took the Cambridge, Massachusetts voter registration roles. She was able to, just by looking at non-unique identifiers, to discover that a very high percentage -- she was able to get over 90 percent of those records were uniquely identified based on non-unique characteristics, including birth date, sex, zip code and what have you. This is simply a problem for everybody.
I think it was clear -- we had participation at this workshop from health and statistical agencies, and it was clear from their activities -- Census Bureau -- that they have always been aware of this problem and they have always been careful to try and prevent any disclosures that they make from being identifiable back to specific individuals. This includes the disclosure of public use data tapes.
I think that the explosion of other information that is available means that the agencies have to constantly re-examine their policies to guard against inadvertent disclosures of identification of individuals.
I think this is a problem for anyone who deals with health data, to have to pay attention to the likelihood that their data, once even though stripped of overt identifiers, and perhaps even though co-mingled with other data, may still no longer be non-identifiable.
I think perhaps IRBs need to pay attention to this as well.
From a legislative perspective, relying on a simple word test or a simple phrase to try to distinguish between identifiable and non-identifiable data no longer seems to be adequate as a regulatory mechanism.
In the current environment we have to recognize that those who receive data without overt identifiers may still be able to recognize individuals and, as a result, some other substantive or procedural mechanism may be needed to protect individuals against identification in the research, public health, whatever context we are talking about, or just straight public disclosure.
I don't think this is a simple problem to solve. Mechanisms that may be appropriate include the use of contracts, so that you disclose data to an appropriate person under appropriate conditions, with a contractual restriction that says, the recipient agrees not to use this data to try to identify people. That is one way of trying to prevent this.
There may be other ways as well, but the goal is the same, and the idea is to support socially beneficial uses of data without impinging on the privacy interests of individuals, which is essentially where we have been all along with this. It is just that it is a little bit harder than it was before. Maybe it is going to be a lot harder, as more and more data is available and more and more data is used.
That is sort of the conclusion of the first day. We had hoped to have sort of a report basically saying what I just said, perhaps that we could send to the Secretary as a letter, and make it a public document.
The goal of the workshop, if we could have come up with a really snappy answer to this problem, that would have been nice.
We didn't, and it is a hard issue. The goal has already been accomplished, in part, which was to raise the visibility of this issue and get people talking about it.
I think we have been successful and I think that what we will do is try to take these conclusions, if they sort of sound roughly good to people at the moment, and come up with a letter and then let everyone have a shot at the written letter that we can move forward with. Does anybody want to talk about this at all?
MR. BLAIR: The only thought that I have, I don't know if you think this would be helpful or not, but I have been concerned about the fact that concern about the disclosure of genetic information has generated interest on Capitol Hill to generate legislation for that.
There are aspects of the patient's bill of rights, and other pieces. I am concerned about the fragmentation of legislation for confidentiality and privacy of the health record.
I would like to see something done to wind up encouraging, at a minimum, the coordination with work that you are doing, if not something stronger.
MR. GELLMAN: I agree with what you just said. I think that is a real problem. I am not sure it is directly on this point, but I think it is clearly something that is a problem.
I think actually there has been growing recognition of that, as people have looked at some of the sort of narrow sliced bills and said, well, wait a minute, how many of these narrow sliced bills are we going to have to implement.
There has been increasing recognition of that. I just don't know that it fits in directly with this, but maybe in the future.
MR. SCANLON: Bob, the other, I think, major point that many of -- the subcommittee meeting was hearing from the large statistics agencies, in terms of, when they do release a public use data file they have a whole set of policies and procedures.
These are probably three decades in the making, in terms of how they minimize the potential for disclosure in terms of -- there are various techniques.
Interestingly, as the technology issues challenges to those techniques, they seem to ratchet up the techniques themselves, so that you kind of get a balance here between data release and data confidentiality.
MR. GELLMAN: I think that is right. It seemed to me -- and I sort of tried to make this point -- it is not that I am being critical of anyone for not doing their job.
They are aware of it. I just think that perhaps the only thing I would sort of make a mild suggestion is they really have to pay more attention what is going out there in the world in terms of the availability of data, because it affects the decisions that they make.
It is really what they have been doing all along. I am not being critical of them at all. They are sensitive to it.
DR. DETMER: What else do you have?
MR. GELLMAN: All right, the second day, the second workshop was on data registries. It is sort of like, what is a data registry.
The answer is nobody knows. There are a lot of different registries for a lot of different purposes. There is a lot of diversity and structure and organization and operation and function.
There are some laws here and there at the federal and state level mandating different kinds of registries, but there is no comprehensive regulatory structure for registries.
In the context of health privacy legislation that will regulate the flow of health data here and there, registries will be affected, both in their ability to get data and in their ability to redisclose data.
This is clearly a concern. I don't think there is anyone on the committee, when we talked about this when we approved privacy recommendations, that really has a problem with the notion of registries that are carrying out legitimate research or public health function, or what have you.
The question is how do we know, how do we distinguish between what is a really good registry for a really good purpose and one that, at the end of the scale, is a registry of arthritis patients so they can do commercial marketing to them.
Why isn't that a registry? One answer is we know it when we see it, but that doesn't necessarily work for legislative purposes.
I think none of the bills, none of the legislation that has been proposed, really focused on the issue of registries. That is really the point, that this may need a little more attention.
I think that it is important that data be provided only to registries that conduct recognized and beneficial activities, whatever that happens to be, that legislation should not undermine the flow of data into registries that meet whatever standards we want to establish, and that registries have to be brought under a legislative framework when one is established, so that data subjects can have some assurance that data that flows into a registry is still protected.
That is not to say that anyone is doing anything wrong with data, but in the context of a comprehensive regulatory scheme, we need to pay attention to this little segment of the world and make sure that it is appropriately controlled.
So, basically the bottom line is that we will probably need some kind of clearer rules for registries in order to basically support activities that are going on today in a largely unregulated way.
This may assist in international flows of data for public health and research purposes because of all the international interest in privacy regulation.
The solutions here -- and I don't intend to propose anything specific -- but it may be that we need either a statutory definition of what a registry is, so that we can test whether any given registry meets the concept of sort of socially beneficial, or perhaps a process to do that.
The process could be something like the use of IRBs or approval by the Secretary or some other mechanism, so that people who want to disclose records to a registry can say, well, we know this is a recognized registry because it is on the approved list or it has been through an IRB, or some way that they can be assured.
One of the consequences of having more legislation -- if we ever get it -- is that all of a sudden disclosures that people have been routinely making, they are going to stop doing and say, hey, how do I know this is legal.
So, in order to support this function and to provide some kind of reassurance to patients, we need to be able to say this has to be addressed in some fashion.
Again, I don't propose to get into any really specific definition. That is something for somebody else. I think it is enough for us to flag the problem and say this needs some attention and it needs to be resolved.
Perhaps the meeting and perhaps whatever we have to recommend will get some of the registries thinking more about this and perhaps they will come forward with a solution on their own, or someone on the Hill will.
DR. DETMER: Okay, do you have other items?
MR. GELLMAN: The only other thing is that we have passed out some stories from The Washington Post that appeared a couple of weeks ago about information from the pharmacy systems at Giant Food and CVS being made available to a third company which sent letters to the patients that were funded by a pharmaceutical manufacturer.
This was a big headline in The Washington Post on Sunday, above what was the lead story, across the top of the paper. There was follow up.
Both Giant Food and CVS folded their tents within two days and said, we are not going to do this any more. There were lots of patient protests.
The issue of the use of pharmacy information and medical information more generally for marketing purposes is an issue that the subcommittee proposes to take up next and hold some hearings on.
It is a very complicated issue. It is not a one-dimensional thing where all these things are necessarily terrible.
There are lots of different factors that flow back and forth and perhaps we can do a little work and shed some light on this and find some more facts.
MS. FYFFE: Can you comment at all on the current federal proposals and how they might have prevented or not have prevented CVS and Giant?
MR. GELLMAN: Yes, it is very complicated, if you look at the bills. The question is, how do you characterize this disclosure, and I can characterize it three or four different ways.
They defended it by saying, well, this will improve the treatment of patients, and there is an argument that that is the case. So, we can say this is a treatment disclosure, so maybe it is okay.
I can say it was a marketing disclosure because they were being paid by someone to basically hawk their product. Maybe that isn't okay.
It is very complicated, and it is not all that clear. You have to go through a couple of stages in each bill, I think, in order to come to a conclusion.
I can say for the Condit bill, I can point you through the three steps that say, okay, you can't do this. But it is not overtly prohibited.
One of the things that happened is that Bennett and Jeffords have a bill that is in draft and in circulation, and it was going to be introduced last Wednesday.
When this story hit, they looked at their bill and I think that their conclusion -- and this was my conclusion -- was that their proposal would have authorized this.
MS. FYFFE: Not have prohibited it.
MR. GELLMAN: No, would have authorized it. That is my reading. There are other readings, of course. I don't want to explain it because it is too complicated to go into the details.
This would have authorized not only this kind of disclosure, but a lot more similar disclosures, and they said, uh-oh, and they went back to the drawing boards on this point.
It is not an easy thing to figure out, even if you look at this and say, this is terrible, and it shouldn't be allowed.
If that is your conclusion, it is really hard to figure out how to write a provision of law that says, this is prohibited.
DR. DETMER: I think what we may do, if you folks don't mind, is move to our panel, so those folks can get on with their day, unless they want to stay afterwards, because we are otherwise going to hear the rest of our reports.
If so, if actually Paul and Peter and Lew and Rob and Col. Ray, come on up to the tables, what we will do is, after we have had your presentations and our discussion, then we will go back into our committee reports.
While you are taking your seats, what is being handed out is essentially a letter we sent to the panelists on the kind of questions we would like to have them think about for their presentation today.
I think we seem to know most of these people and appreciate their being here today.
Paul Tang is medical director of Medical Information Services at Northwestern, Associate Professor of medicine there, and also is co-chair of the public policy committee for AMIA.
He also has a chapter in the recent edition of the IOM computer based record report.
He will be here as well wearing a CPRI executive committee and board of directors hat. So, he has got a lot on his shoulders here today.
Peter Waegemann, I think everyone again knows. He chairs the ANSI health information systems board and is executive director of the Medical Records Institute.
Lew Lorton is executive director of HOST, Health Care Open Systems and Trials.
Rob Kolodner is the associate chief information officer for the Veterans Hospital Administration.
Col. Lynn Ray is in DOD in the health affairs office.
So, we are pleased to have you here. We also had a multi-member panel this morning and I sort of held all the discussion to the end, which really made it a little long.
What we may want to do is at least address maybe one question as we go along, just as a bit of a break. I know we will also want to have some just discourse with you as a group as well. Paul, if you want to kick it off for us, please?
Agenda Item: Panel on Standards for Computer- Based Patient Records - Paul Tang, Northwestern University.
DR. TANG: Thank you, Mr. Chairman and the committee. I am very pleased to be able to talk to you about a few things.
I have been given the charge to do four things, and those are, one, there was a seminal report issued by none other than Dr. Detmer in 1991, that called for the country to adopt computer based patient record systems.
The charge was to have this happen within a decade. Well, it has been a little over a decade, and Dr. Detmer thought it would be good to advise the committee about what has happened since that book.
The second point is, as you see in your memo from Don, is that we were asked to talk about who are the major users, the primary and secondary users, and what are the important functions of a computer-based patient record system, and provide some suggestions for things that the NCVHS could do.
Then, finally, I am supposed to represent CPRI because, unfortunately, Dr. Drazen, our current president, could not be here today.
Let me start out by trying to define CPR, and I will refer to the IOM definition. But first, let me present the definition not.
The definition not is that I don't think our goal, any of our goals, is to essentially electronify the paper record.
We do not want to make electronic all the tethers of paper, by just dumping this thing in here. So, let's turn to the IOM definition which said that it was an electronic patient record that provides a number of functions.
It is a complete and accurate repository of data. It provides alerts, reminders, clinical decision support, and links to other medical knowledge and other aids.
It will be helpful, probably, to review the seven recommendations from that IOM study, because actually they are quite interesting and we have made some progress on some of them, I am pleased to report.
The first and the premier recommendation was that this country, as the title of the book shows, computer based patient record systems, keeping this data in a computer is an essential technology for delivering health care anywhere.
We should, as a country, adopt CPR as our medical record.
The second was that -- this was at a time when the then-Senator Gore was talking about the information superhighway -- and at that time, health care actually wasn't even represented.
So, one of the concerns was that, even though in the committee all the stakeholders of health care were basically represented, even though we all felt this was important, there wasn't a central focus of activities to promote health care as a participant on the information superhighway.
That is one of the reasons why the IOM committee recommended that there was a new institute called the computer-based record institute, that should be publicly and privately funded and participated in, to focus our national agenda.
The third was to provide public and private support for CPR development and demonstration. The committee recognized that this was a major capital investment to do this, millions of dollars even for an individual institution, and we are talking about a billion dollars in the country. This needs to have both the public and the private support to do that.
So, it also gave some mandates to CPRI. It said that, when we create CPRI, CPRI should promote -- not necessarily develop, but certainly promote and promulgate national standards for data and for its security.
CPRI should also be reviewing the laws, the many state laws and federal regulations and ensure that they would facilitate and not impede the development and implementation of CPRs.
This is an interesting one, number six, which is that the CPR cost, because it is a rather high capital cost of both implementing and maintaining CPRs, should be shared among all users of the data, not just users of the system.
Then finally, we need to broadly educate people on CPRs, the concepts as well as the benefits.
Well, what has happened since? Two major things have happened. One is a change in the health care policies and the business landscape of health care.
We are now mandated, essentially, to cut costs, to capitate, to make the PCP the gatekeeper of care, to generate quality reports, regulate managed care, and audit the billing practices, as we heard yesterday.
Well, that is all well and good, but what if we don't have the data to do those things? One of the groups of people that doesn't have data to understand some of this is really the consumers.
When we told the consumers back then that we were all out to create integrated delivery systems, one of the surveys of consumers to say, well, what are integrated delivery systems to you, came up with some of these interesting answers.
One is where people of all races and creeds can go to the same hospital; one in which the doctors, hospital and all providers are in cahoots with one another; a system that is all messed up; and sort of a fancy HMO.
What if we tried to do the other things without data, like cutting costs? Without data, that can get us into some trouble, as we found out.
If we weren't accurately capturing our costs and being able to assess them, we can quickly run into trouble.
What if we are told to capitate and don't really have the data that we need to understand what is the cost of providing care and how to manage it.
Well, we can come up with arbitrary regulations like, well, let's kick people out with a condition that they don't know, like pregnancy, in 48 hours.
We can have just the same arbitrary response, which is to create laws that basically mandate certain amount of length of stay.
Well, neither of those -- they are more or less the administrative practice of medicine. In fact, I think they are practicing medicine, as we are finding out.
What about quality reports? If we don't have the primary data and we don't have ways of auditing that data, that can produce very untoward effects, such as, when the State of New York actually produced physician-specific bypass surgery mortality reports.
After the first report, there was a dramatic change in the population, or at least the health of the population of New York.
For example, there was a five-fold increase in renal failure and a two-and-a-half fold increase in congestive heart failure.
What happened is, that changed the prevalence of various serious comorbidities in that state. So, for example, you can see the range of prevalence in the various hospitals.
In fact, one hospital went from a prevalence of chronic obstructive pulmonary disease, a lung disease, of 1.4 percent to 60.6 percent.
Likewise, another hospital's unstable angina population used to be under one percent and now is 61.4 percent.
Of course the gaming going on here is to increase the severity of illness, the predictive mortality and reduce the risk adjusted mortality.
As a consequence, 46 percent of the surgeons switched haves in that report, in the subsequent report, that is.
Yesterday the committee heard about the fear and concerns raised with billing compliance audits. Again, we are dealing with some of the consequences of not having data, despite these mandates.
I wasn't able to be here this morning, but I am sure NCQA talked about their road map for information systems with you.
I think they had very good statements. They basically said, this is an extremely complex system we have, the United States health care systems, and clearly this is a place where we should have outstanding information systems; yet we don't.
More often than not, they say the information systems we do have actually get in the way. Oxford's predicament is one of those example.
So, their message, then, is that if you don't have an information system, you need to get one. It says that all of us participating in health care -- the plans, the providers, the payers -- should all have a sustained investment in this and that, in the future, these performance measures will assume that we have these things.
Like the 1992 campaign, I mean, where it is the economy, stupid, in a kinder, gentler way I would like to say that it is the data, sir, that is the focus, the underlying requisite to all of the things we are trying to do.
Finally, we may actually have, as we have been hoping for -- and maybe it is be afraid for what we wish for -- an alignment of the planets that says there is a lot of incentives for us to go gather this data. We have just got to gather the right data and we have to put it in a place where we can make use of it in the ways that we all want to do.
Imagine for a moment, if we did have data, and that we actually used it, these mandates, like cut costs, could be turned into managing costs. Capitating could be turned into manage the disease in patient care, and the gatekeeper function could instead be a management and a triage of demand.
Most importantly, the report of quality, which may be perfunctory now, could be in turn an opportunity to measure and improve our whole health care system and health care delivery.
The second major thing that has happened since the report is really some enabling technologies have come on with a vengeance. In particular, it has to do with the internet.
In a sense, we have created one world, at least interconnected. Nick Negroponti estimated that we would have a billion users of the internet by the year 2000.
That certainly brings up the question of the haves and have nots, and we address that in the chapter, but basically we are becoming one connected work.
We sort of have one-ish browser or view of way of viewing. At least we have one-ish standard of having the documentation.
One of the important things is that it solves some of our problems with legacy systems. We now can get data out of many of our legacy systems using one viewers or one way.
So, let me -- I put together, and there is already a big missing piece, sort of a graphical view. It is not comprehensive, and it is not necessarily even properly ordered, but it is sort of a way to describe the status of the development and implementation of computer-based patient record systems.
You may start out with -- and this was very popular about two or three years ago -- where everybody wanted to get a clinical data repository.
That was where you put everything either physically or virtually into a place where people could get data out.
The trouble with that is, if you don't have applications to get it out in the way we need it, then actually it may just be a repository, a place where data goes to die.
We need to then do something with it. So, one of the things -- and as I say, the order may change -- is to have physicians place their orders in this system.
At the time when you have physicians actually using the system and placing orders, that is when you can most influence the appropriateness of their decisions.
That is, the missing blob should be decision support and helping us both have the right information to make decisions as well as getting up-to-date information to make the most appropriate decisions.
Another big missing piece, and one of the reasons this is delayed, is capturing the progress notes, the actual documentation of our thought process, and the data that we use to make our decisions.
That is oftentimes delayed because that is awful tough, getting the physicians to enter that in, but that is coming along.
Once we have all this information, it still is not going to be of use if it is not ubiquitous. It is there every time we want to make decisions. That can include the physician on call being at home.
Then finally, we have to -- and have not met the challenge -- of taking care of people on a continuum; that is, continual care. That is yet unsolved.
Where are we on this continuum of development? We are probably somewhere in this vicinity. So, we are making progress, but we still have a ways to go.
Let me address some of Don's questions now about the primary and secondary users and the important functions in the CPR.
Let me list as primary users the clinicians -- the physicians and nurses -- but also let me stick in patients as a key primary beneficiary of CPR systems.
Some of these functions are access to the complete integrated data and the confidentiality and protection, ubiquitous access to decision support and patient information.
Rather than these words on acetate, I thought I would also bring some screen dumps on an example of a CPR system that we use.
For example, once we get information into the system, we need to be able to pull it up in ways that benefit us.
So, one of the ways is to pull up instant information about a summary, about the context, what is going on with this patient.
That important clinical summary includes the active problems, the active medications, the allergies, the immunizations, other kinds of details.
Then you would also want to have this continuum, this longitudinal record. Shown here for example is a sort of womb to tomb kind of thing, where you start out with the well-baby care.
You capture all the office visits, you capture the hospitalizations, and you capture the phone calls, which are part of that continuum of care. That is all there, and then you can probe deeper to get the details.
One of the interesting parts of computerizing the records is that you make it a lot more available. In fact, we tried to measure this.
So, for example, after you see a patient and we are trying to measure the patient from the time of your having an authenticated record of that visit in the document, well, in the computerized system most of the time -- about 90 percent of the time -- we have the thing signed off and available to anyone who wants to use it.
It turns out that on paper, the record, the chart itself, doesn't make it down to the medical records department until an average of 5.2 days.
That doesn't necessarily include the progress note from that visit which, on a single random sampling we found that it took 15 days.
The importance of this is, if you think of it, when people get sick, they are sick and they need to be seen frequently, which is the exact time when the paper record performs so poorly.
In contrast, for example, when we first came up live with our computerized system, I was on call and having to cover for a patient who had only been live for about two months.
I thought, well, gosh, I will check, but I don't know whether this patient is in there. In fact, she was in there.
So, it is interesting that sick people will get seen, which also means that they will have their information in the computerized record sooner.
So, it sort of works better for the electronic system and it actually works against you in the paper-based system.
The other important function area we talked about was decision support. As an example -- and these are things that Clem McDonald has done years ago, but we are now doing in a commercial system -- for example, if you see a patient and that patient is eligible for a flu vaccine, the system will not only remind you, but make it trivially easy to execute that decision. So, if you agree, you simply check the result.
The example was brought up yesterday about our missed opportunity in beta blockers after heart attacks; same kind of issue.
I also mentioned nurses as a beneficiary, as a primary user. So, we did sort of an interesting experiment.
Administratively, for some reason, they didn't want nurses to use our system, but nurses had been begging us to use the system.
We finally turned that on and they absolutely love it. We get about 1,000 calls in our clinic on Monday mornings, 40 percent of which are for medication refills, another big chunk for lab test results.
That goes through the system, the five people who have to see it. Typically you can't get the patient answered within a half day even.
With the system available to nurses, they can get that information right away. That is of primary benefit, not only to nurses, but also to the patients.
Another point I wanted to make was how it benefits -- how a computerized record system, how the primary user would be a patient.
One of the things we do, at the end of a visit, is to give them essentially a summary of this visit. It says, well, you saw somebody at this particular time. These are your vital signs. We hardly ever tell our hypertensive patients what their blood pressure is.
These are the allergies we have on record. And you can correct that, just as you might correct a credit report.
These are the orders and the medications that we prescribed today, and here are some patient instructions about what we have talked about today, all in writing, all for them to take away.
For diseases that don't hurt -- like blood pressure doesn't hurt, sugars don't hurt for diabetics -- we produce a graphical summary of what has been going on recently.
This has turned out to be -- patients really love this, and it turns out to be a wonderful motivator for them to stick to their regimen.
One of the things -- also, it is interesting. There was a Chicago Tribune article done about this system just a few weeks ago, and they interviewed one of my patients.
One of the things she liked about the system -- again, this is from a patient's point of view -- is that I could essentially make an electronic house call.
I called her in the evening to discuss things, and one of the reasons is because I have this at home.
To protect the confidentiality of the information, one of the things we insist, if you are off campus, is to have use go through the secure ID card. I think all of you are familiar with this?
MS. FYFFE: No.
DR. TANG: Okay, secure ID card is a smart card that has six digits that change every 60 seconds. So, that, in addition to my private PIN number, is the only way that I can get into the system.
So, this is what is called strong authentication. I am really quite sure that it is me on the other end of that telephone line.
In fact, if there are some of you who are not familiar with computerized patient records, I think you have a meeting in May -- I don't know the exact date -- in Chicago. I would certainly welcome any of you to conduct a site visit there, if you want to understand this further and see it in action.
The other pieces of use among the secondary users is really for the aggregate data, whether it is for clinical management, administrative management or clinical research.
We need the ability to,in a confidential way, aggregate data and analyze it.
I have an example there, too, from some of our use of aggregate data. If you look at some of the clinical guidelines -- you can even take a non-controversial guideline like the administration of flu vaccine.
For example, investigators have already shown that if you give flu vaccines, you will cut the admissions for pneumonia by half; very significant.
This is a measure of our consensus with flu vaccines. Basically, 100 percent think that we should be giving these things.
Actually, so we asked them, what is your perception of how often you do this, what you believe in. Well, two-thirds of us said that we did this 75 to 100 percent of the time.
Of course, with the computerized record, using the aggregate data, we can provide -- what we have done is provide them feedback on an individual anonymous basis what their individual rates are, and then we measured that effect.
So, for example, without this, for two years running we were pretty much stuck at about 50 percent. The good news is that that is no worse than the rest of the country -- about 50 percent. The bad news is it is only about 50 percent.
When we turned on the decision support, we improved 50 percent. So, that is an example of the secondary uses of data, going back and affecting clinical care.
Well, where are we about mid-decade from the IOM initial report? In these four goals, the first one was to adopt CPR.
I think that is still a goal. We are not there yet, but encouragingly, there are some early adopters who are implementing these systems.
Establish a public/private CPRI. CPRI was established in 1992. It is primarily funded privately. There are a couple of public agencies involved like the VA and the CDC.
What about public and private funding of research and development? Industry last week was HMSS, and there were 99 vendors who listed themselves as CPR vendors, plus or minus an order of magnitude, probably.
Public funds are coming in through the NLM, through grants and contracts of the NLM, AHCPR, NIST, DOD and VA.
Sharing the costs of CPR among users of the data, that is an area where I don't think we have been making much progress at all.
Right now -- and that is one of the impediments -- is that the costs are really borne by the users of the CPR, but not necessarily of the data.
Because the IOM report had recommendations for what CPRI should do, let me include some of what CPRI has been doing in the past six years.
One of the mandates was to promulgate national standards. Well, in 1996, CPRI organized a national summit on health information solutions.
We basically collaborated with major health information associations, like AHIMA, AMIA, CHIM -- the industry group -- CHIME -- the health care CIOs, HMSS and the Medical Library Association, to have this leadership summit where we had senior executives representing all of the health care stakeholders.
There were over 80 of them -- this was invitation only. What happened is we had consensus on three high level priorities in furthering the development and use of CPRs.
They were universal health identifier, confidentiality and security, regulation and laws, and terminology conferences.
The timing is interesting. This conclusion came in February of 1996, and the HIPAA came in August. So, I think we are all really on the same track.
Of interest, CPRI, about three years ago, proposed a standards acceleration project. The goal there was to grab some full-time effort and rotate those people in to a central place where we would literally try to bang out some of these standards and put some concerted effort.
That project was not funded at that time. The reason I bring this up is that maybe its time has come to do that kind of thing.
I am not at all going to talk about GCPR because Rob Kolodner is here, and I am sure he is going to go into that.
This is an example where we put some public monies into the standards development.
Another CPRI mandate from the IOM committee was to promote enabling policies and laws. Here, CPRI has done a lot in the confidentiality and security area. They have published seven documents that have been actually very well received.
In fact, the Massachusetts Health Data Consortium picked this up lock, stock and barrel in their regional work on confidentiality policies.
So, this has been very helpful. It includes everything from a job description for security officers, to sample confidentiality agreements.
The last recommendation that IOM made was to get the word out about CPR systems. CPRI, again, has been active in doing that.
CPRI is organizing an implementation conference this April on the HIPAA legislations, and how people should be working to implement those standards. Again, this is in conjunction with other groups -- HIMA, AMIA, the Joint Health Care Information Technology Alliance.
CPRI general meetings have been instructive as well, and very well received. For example, in July we are going to talk about evaluating CPRs.
That is in conjunction with an award program called the Davies award, after Nicholas Davies, one of the IOM committee members.
It is like the Baldridge, where it talks about best practices in everything from management to acquiring or developing the product and measuring the impact. That is a way of dissemination of information of best practices.
CPRI had a terminology conference last November and plans a second, follow-up terminology conference to bring the people developing terminologies together, later this year.
What are the remaining impediments? It is not that they have gone away. One is just in the area of the CPR themselves.
We need to continuously and aggressively evolve these systems, so that they serve all the needs of the multiple stakeholders.
We need to more quantifiably understand the benefits and the costs of both developing and using these things.
We need to communicate the results of the experience that the early adopters have had. We can do this through -- one is the Davies award symposium, and there are other conferences and media that are being used to communicate this information.
The infrastructure still remains something. We may be getting a little tired, but we do have a sense of urgency.
Standards. Clearly, we are accomplishing a lot through working on the HIPAA legislation. The universal individual health identifier is still something of major concern.
One of the interesting aspects about using a CPR before, I didn't care what anybody's number was because I didn't have the record anyway.
Now that I have access to the record, I actually need the number to get access to the record. It turns out that most people know their social security number, for example, and that is a quick way I get in.
Clinical terminology was the number one impediment mentioned just at last week's HMSS leadership survey. That is just a top area of concern in order for us to both use the record as well as gather information from the aggregate data.
Confidentiality and security, I certainly don't need to talk about here.
The cost sharing models are something that still need a lot of work. We need to find ways of getting people over that initial capital cost barrier.
There is a leadership education and strategic planning in executing these plans.
So, opportunities for NCVHS, which I hope are not insurmountable opportunities, are in the standards area primarily.
Certainly, I know that you are planning to conduct hearings. I hope that you will be issuing a recommendation on the UHI, because that is something that is sorely needed.
Clinical terminology and the data model that serves as a framework for that terminology is important and, I think, a top concern of the industry.
If this committee could help couch the definition of that problem, and recommend public funding -- because it is going to be a big problem to accelerate the development, say, by 2001 -- that would be of great service.
Policies, confidentiality is out there. As Bob Gellman talked about, we have legislation, bills, being introduced, and we need to keep this on track. All of us do.
This whole notion of data and cost sharing and the policies -- maybe there are recommendations. I don't have any specific ideas, but maybe there are recommendations where HHS can play a role in setting the model for cost sharing.
There are a lot of beneficiaries of the data and we have imposed a cost on the users. The docs are the users, but they don't have the capital. Somehow we have got to merge the capital with the users. One way is to look at the beneficiaries of the data.
I want to part with a little sense of urgency. I know we are all on the right track, but as Mark Twain said, even if you are on the right track, you will get run over if you just sit there.
DR. DETMER: Thank you, Paul. Are there one or two questions or comments we would like to make at this point?
All right, we will proceed. Then we will, in fact, I am sure, be getting back to you, Paul. Thanks. That was excellent. Peter?
Agenda Item: Panel on Standards for Computer-Based Patient Records - Peter Waegemann, ANSI.
MR. WAEGEMANN: Good afternoon, everyone. Thank you for allowing me input into the committee. I have passed out some handouts, and you have seven pages of handouts.
I am not going to read them. I am going to give you an animated version of it in free style, and I hope we can highlight the points I think which are important.
I started out describing that I am speaking for both the ANSI HIS-B and the MRI, the Medical Records Institute.
What both of them are doing is described in the paper, so I don't have to go into that. I also want to shorten all of my various description of what is in there.
I was hoping there would be time to give you some entertaining description of the history of medical records, because it is worthwhile to understand how did we come here, and what has been developed and, for instance, what has not been developed and what was left out.
That would take probably another 10 or 15 minutes. So, I will skip that and try to get back to this as we are going along.
I want to go right to the questions and then I will get to the various issues.
When you look at what -- the key of what we are talking about is what is the electronic patient record. One really needs to go back and say, originally, if I have looked at early medical records -- St. Bart's Hospital in London has medical records from 1684, you have many organizations in Europe that they have medical records from the Eighth Century onwards -- it was to remind the caregiver, to remind the physician what he had done with patients, or what was the process.
Some of them were very interested in that. It was always focused to remind that one person, that one caregiver, of what has been going on in that process.
Even if you look back into the Fourteenth Century where the first public actions were in Venice and so on, you will find that the individual records were just being kept for one caregiver.
Now, this has changed, and I don't have time to go into the whole evolution of it. We had three surveys in the last 10 years where hospitals have been saying, we have up to 122 people, caregivers, who need to get access to that medical record, because it has become so complex, we have so many people who need to get to it.
As a result of the evolution, what we have now is the documentation of the complete care process of a patient, plus the decision-making processes, the documentation of that, really, of the electronic patient record.
The moment that we look at it that way, we can see that there is no division between management data, between financial data.
We need to know why a surgery was delayed, for financial reasons and so on. All of this is part of a future electronic patient record, and therefore, we have to see, when we look at the major uses, of course the caregivers.
I have just listed here a few, from physicians, nurses, therapists, technicians and so on. Then, of course, the persons who are involved in the handling process, they have information -- management professionals, transcriptionists.
Do you know, for every two physicians we have one transcriptionist in this country, 250,000? Then we go on to the systems and computer professionals.
If you add the handling persons together, you have at least the same number, if not a higher number, than caregivers. They are all involved in the process.
Let's just go quickly over the four major functions of the medical record, if I may put this down. The first part is share-ability.
When we are saying, what really -- if so many people need to get to that medical record, what is the key function.
This is really sharing that record. I always use in my speeches the example of the east coast teaching hospital where the patient had emergency surgery. They called the west coast HMO hospital.
They got the fax running after nine minutes, except it was 162 pages. So, here they shared the medical records between two facilities.
It is not the time it took until the record came through on the fax machine, but it took an additional three hours to figure out what is what and how is the record structured and what does the terminology mean.
When we are talking about sharing information, we have to realize that we have a tremendous task in front of us. So, share-ability is a major issue.
The second one is comprehensiveness. We have in the average hospital up to 17 places where medical records are being stored.
You go to any hospital and you have an inpatient record and an outpatient record, an ER record, and X-rays are stored somewhere else, and feeder strips are stored someone else, and you can go on and on and on.
The first part to create an electronic record within a hospital would be to have one data base. So, comprehensiveness is a major issue, even on an enterprise basis.
I will be talking about that when we go to the IOM study and we are talking about having for the patient a comprehensive data base from the psychiatrist and possibly the dentist, and that is much more complex.
The electronic patient record is the nucleus of the health system. We need to realize that management systems, financial systems, public health, research systems, telemedicine, all of these are dependent on the nucleus of the electronic medical record or electronic patient record.
The final part is, where the paper medical record was a passive medium, we are moving into an interactive medium. Paul talked about that with physician support system. This is really where the future is.
This is where the real benefit of the electronic patient records are coming from.
Now, what we have to realize when we talk about the requirements for the record, the key now are the documentation requirements.
We need uniform information, and that means information and presentation, coding systems and so on. We need to have uniform content. Remember that example I just mentioned. We need a uniform structure, logical index.
We need to have integration of information from various specialties. All of this needs standards, and that is where we need to talk about things.
Most important of all of those are the documentation rules. It has been overlooked by all of those, but I will be mentioning over and over, when people are talking now about CPRs, implementing them and there are, as Paul mentioned, there are 99 companies at HMSS who say, we are selling right now a CPR system.
At the TPR conference -- Toward an Electronic Patient Record -- there are over 160 companies saying, I can sell right now such a system.
The point I am trying to make is that none of them really have what I call, or what the vision was of an electronic patient record.
All of them need to have paper back-up records. All what we have at this point is a parallel computer system to the paper record, which is the legal and archival, and main documentation at this point.
So, what we have to realize, the key elements to get to electronic patient records are really to have a signature, and to have it in a way that accountability is there, so that any provider can go 20 years later to court and say, Your Honor, I am certain this was created by Dr. Tang and no one else.
So, this is an important part that is missing by about 90 to 95 percent of all vendors at this point.
The second point is data integrity. Once a signature has been attached, it should not be changed. It is the same as what we have in any commercial art. It is the same as what you have for your checks and so on.
The minute it has documented, has been signed, there should be a ceiling process and such information cannot be changed.
The standard from ASTM E31 goes even further and says, any changes can only be made as amendments and you cannot pull up the record without the amendments. They must be provided at the same time.
All of these are requirements we must have in order to have a legal record.
We need to have information available all the time. This means, really, when there are breakdowns, when there are maintenance of patients, we must have the information available at that point.
We need to be able to audit and we need to be able to say, at any time, for confidentiality reasons, when has it been printed out, for whom, when has it been sent, to whom, and what has been happening in the whole process.
On top of that, we need uniform access and handling rules. We could just spend two hours discussing what is meant here.
When you look at what is happening in various countries in regard to electronically patient records, then you find that they are very different worlds.
In France, a physician is not allowed to give a patient record to anyone but maybe his head nurse. No colleague can look at the medical record.
We need to look at what has to be put into the record, what kind of minimum requirements for documentation out there. I don't have the time to describe those in detail.
What are the secondary uses? I will just try to skip this. It is management, financial, research, public health, and many other people who have non-identifiable information.
As we have heard early on and I agree fully, one particularly can buy commercial patient identification systems, where you can take patient records without the name and information, you just take my social security number and where I live and the age and whatever, and you will have within 30 seconds my identification.
What we really have to look here is what is an electronic medical record, or patient record, and what is not. Paul talked about that.
Any physician who puts information on a computer, on his pentium, and has it on an axis data base, it is an electronic medical record.
No, it is not. It will not stand up in court. It has no signature. You can go and change it at any time because it is just on the hard disk. There is no indication that it has been changed. It will not be legal. It will not be archival. It is not acceptable, in most cases.
What we have to realize is that only information which is compliant with the principles of documentation can be accepted.
That is really, when you talk and you go to Sweden or you go to The Netherlands, and you find that 80 to 90 percent of the GPs have patient information there. They do not have a CPR because they do not share it with anyone.
They write a specific letter if they have to, but it would never stand up in court anyway, because it is just information put on a hard disk and anyone can change it and there are no specific rules for that.
The Medical Records Institute has been for years concerned about the fact that people are saying we are implementing CPRs right now.
What we have been saying, how can one get somehow a handle on it. We have been promoting for five stages, which have been quite acceptable, and actually are now part of the legislation in Australia, New Zealand and some additional countries.
What we are saying, any hospital today buying computer systems is doing this parallel to the paper records. They all keep their paper records.
Actually, as a matter of fact, all have shown that 60 percent of the paper in an average hospital now is computer print-outs which go into the chart.
So, the paper record gets bigger as we do not have these requirements that I have been talking about. What these people are doing is automating.
In most cases, when we are talking about implementing electronic paper records, then what we have to realize that in hospitals it is still decentralized.
They computerize what I call empires -- the radiology department, outpatient departments. There is no company at this point which has even a consensus on a comprehensive health care information system.
No one should have to go out and buy today an order entry system. This should be a byproduct of the health care information system.
So, what can we do today? What are these people doing? They automate departments, they do certain sections. We do not have the time to analyze the whole health care field.
The main problem in all of these discussions is that we take a medical record as a medical record. What a non-accredited counselor in a mental health clinic is writing in a medical record and what you have in an ICU unit are so different in content, in structure, in many of the additional features, that it is very hard to call them both medical records. We need to look at all specialties.
Again, what we can do right now is to automate sections, and that is what our hospitals and providers are doing at this point.
If you want to get rid of your paper, you can do it by creating document imaging. About 22 percent have been doing this.
We have been doing surveys at the Medical Records Institute, so it is about 1,200 hospitals have document imaging and have what we call a computerized medical record.
It is an interim step to get to the full benefits of an electronic medical record. What we believe the current task is, is to create within an enterprise an electronic medical record.
This means that there is one data base. This means that there is one agreement on how to use clinical information and all information within one provider enterprise.
This means integrating legacy systems. It means integrating many of the policies, identifying the patient once. Remember, in 95 percent of all hospitals, the patient is still being identified between six and 14 times because in different departments they have to give their name and number again and so on.
So, the third section is really to create within an enterprise an electronic medical record. Now we are moving on to the patient record.
We are far from creating it within an enterprise. What we are talking about is calling to fruition the Institute of Medicine patient information system, with multiple disciplines, multiple providers who have different incompatible systems, and therefore, this is a long-term situation out there.
The final one, and again, I don't have time, is really where the patient is involved. Now we are talking about information going away from the patient to whom care is given.
We are talking about a health record which is being kept in combination of the patient and the provider, or care giver. I wish I had more time on that.
What of the benefits? What has been happening? Right now we are moving along on the journey toward an electronic patient record.
What we need to understand here is that there is no quick solution. I get, once a week at least, a call from a hospital saying, Peter, we just got $22 million approved. Where can we buy a computer based patient record in the next two weeks? We have to spend it by April 1st. We want to get rid of all of our paper records.
I have to tell them over and over, we cannot do it at this point. It is a journey. It will take some time. We need to understand that providers are moving along -- I don't want to get to payers, vendors and so on.
What are the main hurdles? If the ideal system would be available at HMSS or TAPR, these systems would fail tomorrow.
What we need to understand first are the professional changes, behavioral changes, and I described those for you in the handout.
The systems changes, technologies are moving along and the industry is moving along. I want to focus on the two ones which really should concern this committee.
The first one is standards. What we need to understand here is that there is a tremendous complexity. If you take the electronic patient record as the nucleus of the total health care information system, then you can understand why standards are so complex and what is happening in that field.
We need to understand that there are, in the United States, more than 5,000 people working on health care application standards, all related to that whole field.
We need to understand that this is an area that really needs to move into a new era, because we do not have, at this point, interoperability at a plug and play level.
We are at a stage -- and I have studied this -- if you would have had -- I should say the first electricity works was really created in Brooklyn, New York in 1876.
If you would have, in 1898, to plug into their electricity system, you would have found that two or three systems of office buildings would have blown the bulbs because each was on different wattage and different strengths.
We are at that point right now. What we have to realize is that it will take substantial time to get to precise detailed plug and play standards.
It is very complex. We are looking for some leadership. We have the support of AHCPR and ANSI and HISP, and we need to keep it and we need to get it in the future.
Technologies are changing. Again, I don't have the time to go into that. What you see now will be outdated in a few years. We are moving toward SGML and XML and different ones.
What I think this committee should be really concerned with is kind of recommending legislation. What I believe is needed most is some support of empowerment, which means that any provider can create electronic medical records, if they comply to a minimum of requirements, such as signature, data integrity and ceiling process such as availability, and an audit process.
This piece, which could empower and which could move the whole process of creating electronic patient records along could really be most valuable in that field.
What I would really say is, my recommendations would be to have rules for uniform documentation, on signatures, on data integrity, on availability, on audits, uniform rules for information handling and, at the same time, to look, if you really want to go further, you can go to uniform access and handling rules to uniform data sets and information recommendations on code sets and information structure and so on.
What really should be -- I would like to see a closer cooperation between ANSI and HISP and ANSI VHS. We need to recognize that the main deliverable of HISP is the current clinical inventory of clinical standards under the leadership of Jeff Blair, who is in front of me.
We need to go the next step. As this is happening, we need to look at the analysis stage. We need to know, once we have our standards, what really is missing, where are inconsistencies, where is duplication and so on.
We need to look, at the same time, at the coordination process in the whole standards field, and we need to realize for this we now have a United States -- as we now have an ISO technical committee 215 for health informatics, that there will be substantial activity in the United States tag for it, and I hope that there will be close cooperation.
Thank you very much. I apologize for having so -- just to indicate a few things, but maybe we can afterwards go into more detail.
DR. DETMER: That is another excellent presentation. I thank you for the stage setting for the committee's work on this, and I also thank you for the precision of some of your recommendations. One or two questions before we move on to Lew Lorton?
MR. FITZMAURICE: I wonder if we could get a copy of those recommendations?
DR. DETMER: You have it. I think they are in here.
Agenda Item: Panel on Standards for Computer-Based Patient Records - Lew Lorton, HOST.
MR. LORTON: It is quite a disadvantage to have no overheads and no accent, but I am going to try to establish some different parameters to talk about. It is impossible to talk behind Paul Tang and Peter and tell you anything that -- I don't know anything that they don't know. So, I would like to take a slightly different tack.
I am the director of HOST consortium. We are a non-profit association that started in mid to late 1994, not to serve as a forum, but to serve as a neutral workplace for the industry.
The industry that we have is health care providers, health care organizations and vendors of different types.
We do a lot of projects in order to find what we believe are exemplars of good practice.
I have always been intrigued by the health care information systems, because they are so large -- they are the largest elephant around. Everyone has a different slice at it.
I just read this book, or part of this book, last night and this morning called, demanding medical excellence in the information age by someone named Michael Millenson. It was recommended in Robert Samuelson's column yesterday.
On the way in, I didn't know where I was, but I remembered I was reading about computers. So, I looked up information systems.
There was one entry. It says, information systems, see capital allocations.
So, there are so many different ways of looking at information systems.
I would like to look at where we were and why we got to where we are. When our telecommunications infrastructure was small and our health care organization was different, it was very easy to exist on a little island.
We could make little information systems and people were quite happy with the idea of using whatever standards they had and then adding to them.
So, they added on and they had local systems, local language, local information things and everybody was fairly comfortable with that.
They were even comfortable with the fact that they might have two or three different systems running on the same network that didn't share data.
As long as the hospital was stable and we could satisfy all the health care providers without there being an open revolt, they went along that way.
Well, all of a sudden, a few years ago, we had a huge -- we had two revolutions. There was a health care revolution which changed organizational structure completely and made us realize that health care is incredibly flexible, and there is the information revolution, in which the change in technology made everything able to happen.
So, there aren't too many technology challenges that can't be dealt with, but there are a lot more organizational challenges which then illuminate some of the problems we have with non-standard systems.
All along there were visionaries who worked on standards, people who knew -- either because they were engineers in some previous life, or because they understand the issue that we need to share data, but they knew that we had to develop these standards.
Unfortunately, they were always under-resourced, always under-resourced. If you look at standards organizations, they don't have paid employees to work on health care standards, many of them.
Companies do this because they have to or they want to, and certainly not all companies are fully represented.
If there were 90 companies generating CPRs or HMSSs, not all those companies have somebody working actively on standards.
So, that is our basic problem, is that the development of information technology is always outrunning the ability of standards organizations to keep up with it.
Most of the people who have been working on this are very bright, and they have been working very hard, and yet technology changes are outrunning this.
So, I look at this and I say, well, why are we under-resourced. The reality that has always been with me is that when you have to convince the person that something is good, up until the time that they have to sign the check, they are very happy to go along with you if the amount of resources that they have to commit are small.
You can see that in the development of the CHIMS up until a few years ago. A lot of CHIMS would start to develop and then, at the last moment, just before they implemented the major equipment, people would back out.
They would say, what is in it for me. What am I going to get out of this X million dollars.
So, the issue for us, the challenge for us, is not necessarily in technology development because that is essentially, as you can tell, invisible to the outside.
It is the challenge of showing the world, showing the users, showing the providers, showing the people with the checks that what information technology can do for them will pay them back in a term that is acceptable to them -- in the short term, the long term, whatever that term is.
The other challenge is to look at how we develop things like this, and I would like to give three examples of how things are developed, two that we were involved with, that I was quite pleased with.
They are perhaps not the best example of development, but they were done quickly and were successful.
One is, the current network which is done by all the electronic health care transaction companies, that transfers billions of transactions every year successfully, confidentially, privately, why did they do that.
They do that because they made money out of it. There wasn't a four-year cycle to figure out what the standards were to get it going. They pushed to make it happen, and it happened and it works, and it is invisible, because there was a benefit to them.
The second example is, about a year ago, a little bit more than a year ago, Margaret Mariaku(?), the former executive director of CPRI and I, and somebody from Los Alamos National Laboratory were sitting at breakfast at a HOST meeting.
They said, we would like to have a meeting about the master patient index. Over the next six months, we sponsored two meetings, one at Los Alamos in Sante Fe. We expected 30 people and got 105.
One was here in D.C. We expected 50 people because we figured the first one was boring as hell and most of those people wouldn't come back, and we got 140.
What happened from that is that the project transitioned into HL7. The project also was taken over by -- was actively worked on by Corbamed. Now Corbamed has a draft reference standard and at least eight companies are producing active -- how many -- well, I must admit, I am not much on the intricacies of Corbamed. I get a translation, a simultaneous translation when I go to the meetings.
In a little over a year, people who had a commercial investment in making this happen, made it happen.
A third example is a project that I am very pleased that we are involved with. Through the NIST, ATP and HOST sponsored a project down at the University of South Carolina in gastroenterology.
Over a period of two or three years, the gastroenterologists have developed an extension to their language, have established all the additions to dicom that they need to make this work, and now they have a computerized patient record for their work, which works for them.
This is another example of a situation where you have the stakeholders very actively involved, and they can see the feedback.
So, without any criticism about the way things have been done up to the future, the reality is that it is a different world now and we have to satisfy different masters in health care.
It is no longer deciding that this is just good for the patient. We have to say that this is good for the industry in some way.
I would suggest that some of the challenges are that we have to define exactly what we want to do and why it is good to do it in real terms.
Somebody from SCMS, a very large medical records firm, speaking off the record for his identity, says that the industry has not done enough to establish the cost benefits of computerized patient records and they need to do more.
So, it is that kind of establishment of real life cost benefits, a prioritization of what we need to understand about being implemented, a prioritization of what we see as transferring between different sites to establish the prioritization for the needs for standards.
I don't have any tasking for the NCVHS. I do think that because -- this is a unique industry. If we were interested in the automobile industry, there are only four large manufacturers that control the whole industry.
If they wanted to establish an infrastructure that makes things work, they meet together in pre-competitive meetings and do this.
That is why you can use any brand of gas, you can use any brand of spark plug, there are a whole bunch of things which are interoperable between violent competitors.
In the health care industry there are 90 people who say that they make computerized patient records, or at least at HMSS, and there are 700 companies listed in the director of information systems manufacturers.
We have an incredibly fragmented industry. I think that the only organization that has the power to spearhead this movement is the federal government.
I am not much for a growing federal government. I don't believe that the federal government should either dictate or fund all of these activities.
I think that the only activity large enough, the only organization large enough with enough presence to do this, with enough reason to do this -- because they pay for so much of the health care -- is the federal government.
I think the challenge for the future is how to make the federal government serve as the leader in changing health care. Thank you.
DR. DETMER: Thank you. Questions or comments before we move on? Another excellent talk. Rob, you are on.
Agenda Item: Panel on Standards for Computer-Based Patient Records - VA/DOD/IHS/LSU Partnership.
DR. KOLODNER: Actually, Col. Ray will start. We are going to split our presentation.
COL. RAY: We are going to kind of do a tag team here. We are going to split this in half. What I am going to do in the first half here is talk a little bit about the effort that we have embarked upon as a partnership.
That partnership is the Indian Health Service, the Veterans Administration, the Department of Defense and LSU Medical Center. Rob and I will be representing that partnership, not just us two. Also Richard Ferrans from LSU is also in the audience here.
We want to tackle this complex problem. We know it is going to be difficult. I think that is why we want to do this as a partnership, that we take our combined knowledge of where we have been and try to move forward.
The goals, we got together in January and sat down and tried to come up with a vision and a set of goals and objectives, where we want to go with this partnership.
Everybody talks about the 90-or-so vendors that were at HMSS. We had an industry day on the 29th and announced this partnership. There were over 140 vendors there looking for work.
Our goals from this GCPR is that we will create a collaborative partnership and there is a lot of energy here. We are excited about doing this.
Our goal is that we can appropriately share, I think, that the right data gets to the right person for the right reason.
We understand privacy and confidentiality and we will certainly know that security is going to be a high priority.
We know that standards are a key for our success. Without this committee's support and help and letting us work with this committee, we cannot be successful.
Where there are not standards, we want to work with you -- the partnership and with you -- to try to create those standards.
So, we are willing to just step out and try to create those, if you look at the size of the footprint of this partnership.
This is kind of a conceptual representation of what we have agreed to do. This slide, if you look at it, looks kind of like a doughnut. If you look at the outer ring of the doughnut, it really is where we are today.
We all have a set of applications. In the Department of Defense we use the composite health care system. In the VA, of course, they use the DHCP, and then the LSU has a set of applications software and they are largely into telemedicine as they look through all the public hospitals throughout the state of Louisiana.
So, what we said is, because we are on such different journeys in this outer ring, what we want to do is focus on the center of this doughnut, or the whole of the doughnut.
That is what we call the GCPR, or government computer based patient record.
What we see are key problems that we have to solve, as we have to come up with a standard lexicon. We need a health data dictionary that we can map our current applications to so that, when we share information, we understand what the data are that we are reading, as we move that data across. In order to do that, we need a common data model.
We know that we have to come up with a set of business rules and event triggers that says, who needs to read this data, what data needs to be pushed to this CPR.
Communications protocols, such as HL7, we need to say this is the way we are going to share information and pass that data back and forth.
For the utmost, is security and patient confidentiality and privacy, trying to understand what that is and how we are going to solve that problem as we look at who needs to see what data, when.
I think I have hit most of the highlights on here. It is a collaboration between VA, DOD, Indian Health Service and LSU.
Appropriate sharing I think I have already touched on, so I won't talk about that.
We do have a set of shared populations. If you look at -- a lot of this was kicked off on the 8th of November, the Presidential statement.
President Clinton, based on a report from the Presidential Advisory Committee on the Gulf War Syndrome says that we will build this CPR, because as we have looked at Persian Gulf Illness, one, we didn't collect the data very well in the Gulf, number two, we can't find it when we come back home because it is paper based.
You can't take paper and try to look across and say what caused symptoms. Certainly these now are veterans that both we and the VA have to share.
A lot of those are reservists. Some of those are National Guardsmen. We need to be involved with the states. We can't, you know -- when we discharge these people and they go back and do their jobs in the reserves, we have to be able to share that information with the states.
If you look at the Indian Health Service, a lot of work that we do in the State of Alaska, between the VA, the DOD and the Indian Health Service, all three of us share a patient population up there.
So, we know that we do have a requirement that we need to do this immediately, and we do have a sense of urgency based on things coming up.
Another Gulf War is possible. We have talked about anthrax. Are we tracking immunizations? Are we tracking side effects of immunizations? There is really a need that we need to share this information.
If you look at the purpose of the partnership, we know that we have to sit down. We have done some of this already at a high level, that we have to come up with the technical and functional requirements for the development of a worldwide CPR.
I won't go through each one of the functional requirements and the technical requirements. We will just talk about those.
Our overall goal, why we need to do this, is to improve the public and individual health status by sharing this information.
I think everyone down through here has said the same thing. This is something that we have to do if we are going to improve health care.
I won't go through these. You can just throw these up. I think everyone has a set of handouts. There is a set of business characteristics and technical characteristics that we came up with in our two-day meeting before we did this industry day announcement.
Why do we want to do this? Why does it make good business sense technically? What is it that we need to actually do.
What I will do right now is kind of turn this back over to Rob. He is going to answer some of those questions specifically that you asked us to address.
DR. DETMER: Okay, thank you. Rob?
DR. KOLODNER: One of the things to recognize is that the partnership is somewhere between eight and 10 weeks old.
We have had one face-to-face meeting. We have another one scheduled next week. So, we are very early in our process, and the agreement is really at a high level.
I think it is that commitment to work together, to take on some of those things that have been barriers for us and for others, and for us to have enough of a presence, and enough resources to bring to the table, as a number of previous speakers have talked about, that we need to do it.
We are not doing it in the abstract. We are doing it, as Lynn pointed out, because of real business needs. So, it is our leaders who are behind it saying, we must get this going.
If there are standards that don't exist, we are going to have to move forward rather than waiting for those standards.
Whereas, up to now, on an internal basis, we could kind of do our own thing and get by, once we start crossing the boundaries we have to advance something that at least looks like standards, even if we recognize that we are not a standards organization, but we intend to work with them.
The next slide shows you the standard organizations that one or more of our partners participates in. So, we have been active out there.
We are not intending to go around but work in conjunction with those organizations. But we recognize that at some times our time table will be different. So, we will press on and try to do it in the smartest way possible, that might eventually come close to what that standard will evolve to.
In some of the questions that I will be answering next, some of the questions we didn't even touch, because as a partnership we really couldn't provide a single answer, but at least on a few of these we will move forward on them.
One of them is the major users and functions of the CPR. The primary and secondary users, I think, are similar to what a number of speakers -- Paul and Peter -- have talked about previously. So, I don't need to spend more time on that.
I think one of the areas to spend a little bit of time is on this distinction between the EMR and other related topics.
I really get the sense that each of us who are presenting today are answering this from our own perspective, which is understandable, our perspective being that each of the partners is coming in with our own systems, our own applications.
When we first looked at the kind of high level charge from our leaders which was, well, get a common medical records system or a common data system, we weren't quite sure what they had in mind.
But we sat together and came up with something that we thought made sense, which was not to try to buy the same system, the same vendors, the same set of applications.
We are each at different points. We have different pathways that we need to proceed. We have different priorities from within our organization as to what functions and what things we need next, and where we are starting from.
So, that is how we came up with that common focus on what would allow us to exchange the information about the patient.
That then guided the three level distinction that we have laid out here, between the electronic medical record, the term that is used in your letter to us, the CPR which really came from both the IOM and the CPIR, and what we are calling the CPRS or the CPR system.
What I want to point out is that, however we define this -- and it is interesting to see all the different ways that we can define it -- that there is the aggregate of just the data, Paul's first slide, of just pouring the paper in and the information that is on there.
By itself, it is a start -- I think it corresponds to one of Peter's levels of being able to scan in the data, and it is a level above not having anything, but not a whole lot above.
The CPR, to us, is this core that is vendor independent. It is the conceptual, it is the standards level, that allows that exchange of data, and that is really what we are focusing on.
Then there can be a multitude of solutions, of individual solutions of the systems, that then each of us can run, that ties into that and allows us to exchange the information, the patient specific and actually population specific data.
As you notice in our goal, it is both individual and public health advancement.
So, you see there, at least, where we came from and why we draw a distinction between the CPR and the CPRS. I think the importance in that is that -- I haven't run this by all the partners, so I can't say we have a total endorsement, but my sense is that where your focus is, is that middle layer, is that standards.
It is not in a system, but it is in those things that facilitate the exchange of that information, so that it is readily available any time that it is needed, whether that be for patient care -- which should certainly be the core reason for anything we are doing -- or for doing other analyses, either population or reimbursement kinds of activities that are necessary as part of our society and as part of our health care system.
In terms of the capabilities that are successfully being provided, again, we have got some examples of emphasis from providers on clinical decision support and places such as Intermountain Health Care and specifically a starting place in LDS hospitals, and registries that are notable examples of some really advanced work that has been done.
Again, it is ones that have been specific to a site or small set of sites that haven't been able to spread as a system beyond those particular, relatively small sites.
DR. STARFIELD: Can I just ask you, are they both in hospital, or do they go beyond the hospital?
DR. KOLODNER: Regenstrief is outpatient, at least a lot of its work began in outpatient, whereas LDS was inpatient.
I am not sure of the current status. I am sorry Clem isn't here today, because he would be able to tell us just how extensive it is.
DR. TANG: They also have inpatient. It is a bit of a mix. It is sort of in, out, in, kind of.
DR. KOLODNER: Then in terms of vendors, they have a slightly different focus in terms of information, and then the standards organizations -- certainly the HL7 -- as the vehicle for moving the information, but still needing the standards of the coding and the vocabulary to go inside that vehicle is something that I think you are all familiar with.
In terms of impediments for the EMR or the CPR, which is really where our focus is, there are a number, certainly starting, as we saw today, with the differing definitions.
I think it is useful for us to just kind of say, here are the levels, or here are the different terms that we will use, so that we can all be using those.
We really aren't wedded to one or another, but I think it helps to clarify what we are talking about, and whether distinctions, for example, should be made between the CPR and the CPRS, so that we can separate out some of the functions and help clarify what we may or may not be focusing on.
Certainly as we heard, the issue of the complexity, cost, et cetera of the implementations, the lack of data models -- and as you look down this list, you will see many things, again, that have been mentioned before, as I would certainly expect.
There are also some things like the limited advances in voice and handwriting recognition. Although they are improving, as Paul alluded to, they still don't fit the practice of care. They still get in the way. They still take longer.
There are occasional places where you will have the champions and the motivated clinicians to use it, but it isn't something where you see clinicians flocking to it in droves and saying, wow, this is really the thing that we need, and oftentimes you need to do trade offs.
We still have some technology barriers. I think what that also means is that we need to really focus on what is that core data set we need to capture, and focus on getting that, not necessarily capturing everything that we always did, in fact, put in paper records, and at least getting started and chunking out the advancement, rather than trying to do it all in one fell swoop.
There is also another thing that was, again, mentioned, particularly in Peter's talk about the level five, and that is the data ownership and when do you move that data, and who owns it, and does the patient have access, that I think really need to be clarified and I guess are part of the President's patient's bill of rights.
I think that is an important area that needs to advance that may actually help us, rather than hurt us, if we can help to particularly empower the patient to have more say in all the things that are going on.
Then, as we have heard a number of times, the issue of confidentiality and security. We actually have an advantage in our partnership, because at least three of us are under the Privacy Act, which is more stringent and more well defined than what is out there in the private sector.
Actually, having that structure eases our burden, rather than making it more difficult.
In terms of recommendations, as far as addressing the impediments, I think they really follow some of the key impediments such as defining what it is that we are focusing on and what it is that you will do.
Part of that will then be to see what the overlap will be with the GCPR partnership because many of us are, in fact, participants and representatives in a lot of the HIPAA activities.
In fact, we expect that the partnership is, in a sense, an open one. If you notice in the diagram that Lynn Ray pointed out, the doughnut is open. It is not closed.
That was done on purpose, because we expect to bring in and to partner with a variety of other people, and to bring in basically the best minds and the best thoughts and things to address these problems.
So, we have talked in our work groups a little bit -- although they haven't formed yet -- that we expect there will be representatives not only from within the Federal Government -- NLM and HIPAA activities -- but also we expect that we will be reaching out to those who are active.
Erica Drazen has had some conversations with us, and I expect that a number of people who are doing premier research and cutting edge kinds of activities, that we will want to bring them in to participate, to let us know at least what not to do based on lessons learned, as well as some suggestions as to what we should do moving forward.
As you see, a number of other things. The unique patient identifier, there what we are saying is we are going to be using something to link those people together.
Whether we need a unique identifier, or whether we have some sort of a linkage amongst our master patient indices, we will be using something to move forward.
In fact, each of the partners is establishing a master patient index, so we will probably use that in the meantime.
Whether or not there should be a unique patient identifier, we are not going to weigh in on that, but we know that is on your plate, and there is at least some way of moving forward.
As Bob Gellman said, you don't need a unique identifier to uniquely link the patients. Just a few elements will do.
Finally, in terms of the recommendations, they really follow the previous slides. In particular, as with a number of the previous presenters where we would like to be working together, we think that we have an opportunity to serve as a test bed, maybe to in fact take on some of the things that we find are our first business needs.
That may be Persian Gulf data, that may be some other kinds of things, where we will be moving forward and learning some things and actually testing it out in real settings.
There are others, I know, such as Simon and Chris Shute and others, who have been doing that, that may be drawing in some of the data sets that are more well defined or are beginning to take hold.
Again, addressing some of those content standards and coding standards in chunks rather than trying to do the whole thing may be a strategy for you to consider.
DR. DETMER: Great. I really hope that this partnership continues with the level of interest and enthusiasm that you have gotten underway. I think this is fantastic.
Okay, what we wanted to do today was actually re-acquaint ourselves with where things seemed to be at the moment, as we move forward working on the next order of business of the HIPAA mandate, which has to do with computer based patient records.
I think this is a really just outstanding first step for us. With that, let me open it up for questions and comments.
MR. SCANLON: Focusing on the hospital and care area for a minute, the current array of hospital administration systems that you are familiar with now, how do they relate to what you are envisioning as the computer-based patient record?
Are they something that can be transitioned into? Are they systems that can be interrelated with the computerized patient record or are they really business oriented, or administration oriented, and really are a separate and parallel sort of system?
DR. TANG: That is a good question. Unfortunately, I think the current hospital systems are mostly business oriented, as you suggested. Most of us are actually having to replace those to a more clinically oriented system.
I am also hoping that, as we move into the ambulatory care side, as we heard yesterday about data collection, that we look toward clinical data collection, as well as the administrative data collection.
I understand the sense of urgency and the need to do with what you have, but I would hate to force the ambulatory care side into being another billing system and overlook the clinical side of it.
DR. DETMER: The comment relating to the discussion on E&M codes and the recent visit to HMSS, where a number of software companies are targeting to meeting HCFA kinds of requirements rather than essentially working toward a --
DR. TANG: As well as the BBA stuff that we heard, in terms of getting a number for people based on essentially billing, administrative, data, we would like to certainly like to get closer to clinical data.
DR. DETMER: Others have any comment?
COL. RAY: I think DoD is kind of the opposite because we are way behind in billing systems. We have not billed in the past. We have kind of become this HMO that goes to war.
With our Tricare system we kind of fell into this, that we are trying to do it. So, most of these systems that we have out there are really clinically based.
Our composite health care system, which is by far our largest, has a central patient core and it has pharmacy, lab, radiology, appointing and scheduling, and some administration on top of that. That is deployed to 450 hospitals worldwide.
That will be the centerpiece, and the feeder of the data that we need to share with the VA and Indian Health and LSU.
DR. KOLODNER: Just to expand briefly on what Lynn just said, actually the three partners -- VA, DoD and Indian Health Service -- have cousin systems. We have a common base.
Indian Health Service has probably done the most in terms of taking the system and moving it into ambulatory care, long before the rest of it did. VA and DoD stayed hospital focused and they actually started moving data, not just reports, amongst each other.
I think Lynn's comment there is we probably have more of a clinical system than a business system. So, we will be feeding some of that in, but there certainly are elements that I think many of the private sector has, that pharmacy, radiology kinds of things, that can be fed into the electronic medical record, or whatever term we want to use.
MR. LORTON: Several years ago I was a CIO of a medical foundation and we were introducing inpatient medical records and we had direct board entry.
I was talking to one of the senior physicians, who was pretty irascible. I said, look, you just sit down here, log in and type your order, it goes to everywhere, and it works. He said, I write it down, I give it to the nurse. That works.
So, if you see that one of the major problems is a cultural problem of trying to persuade people to be something in addition to what they originally signed on to be, I think that there is a technology challenge here to bridge that gap and say, you can put in data however you want to put it in.
We are going to make certain that your data gets put in correctly in whatever way we have to do it. But there is a technology challenge there, to do handwriting input, to do voice input.
We shouldn't downplay that. If we can do that, if we can meet that technology challenge, then we can counteract that sort of cultural ennui we have to implementing these things. That is why a lot of people are putting a lot of effort into it.
DR. COHN: I just wanted to express my appreciation of the panelists. I think you did a great job. As one who sometimes has to present on the other side, I was just sort of sitting and listening and was impressed with the amount of knowledge and expertise at this table.
I was sort of reminded that I think some of the issues, I mean, are really going to be very daunting over the next year and we have got a lot of work that needs to be done, as well as focusing on where we can really make a difference.
I certainly am also reminded, I think the CPRI work on the terminology conference, which of course, NCVHS participated in and was on the planning committee as well as the NCHS, the National Library of Medicine, HCFA and others, turns out probably to be a very strategic move and one that we need to look at, how we want to move forward.
With that preamble, I actually had a specific question for Rob and Lynn. That actually has to do with -- as I look at what you are doing and I look at your overall business model, as last I heard it, which is that you are going to be integrating with the community, and that, I think, is the future of how a lot of VA and DoD are going to be working, how does some of that relate to the overall parts of your problem and really making sure not only that you can communicate data together, but with everyone else in the world.
DR. KOLODNER: I think that was one of the opportunities we had with LSU Medical Center joining us. It turns out, just to give context, that even before we change more, $1 billion of the VA $17 billion goes to health care provided outside VAs.
We get precious little information back about the care that is given, and that is something that we need to do more of.
LSU Medical Center, which has a budget of over $1 billion, which was rather surprising, is one of those places where we send patients, and DoD mentioned that reservists get their care in the state.
So, it broke the mold of it just being a federal CPR. So, we moved beyond that to a government. But the intention is, if we can figure out how to exchange that information amongst the four of us, we will have a target that will facilitate us being able to get to many of our partners and many of our care providers outside the group.
Our intent, as you see, is really to advance the standards in any way that we can. So, by working with LSU, we start with a non-federal provider, but expect and hope that it will go well beyond that.
DR. DETMER: Does our colleague from LSU want to weigh in with a comment? If so, please go up to the mike, if you would.
DR. FERRANS: This could be characterized as a cameo appearance. When I was a house officer, I remember -- house officer here in Washington, D.C. -- I had a medical student once who needed to get a record when we were at the VA.
He actually drove to one of the local hospitals to get the paper record because it was much faster to do it that way, and the patient was sick and he had to do it. He was an excellent medical student.
Some of the things that we would like to do, and getting into the business model, one of the keys of the Department of Defense's future health information systems is a personal health information carrier, which would be the smart cards that the soldiers would carry.
Since there are a number of reservists that are seen in civilian health systems, our ability to see reservists and to somehow capture that data in a standard format, and be able to electronically transfer that information to DoD data bases so, at the time when our forces are mobilized and the Louisiana National Guard and some of our other reserve elements are sometimes the first to go, that that information can be quickly brought down to their personal information carriers, so that those can be kept updated.
I know from the military standpoint that deals with issues of readiness, and those are important. From our standpoint, we would also like the ability, as Col. Ray had mentioned, for us to be able to survey those people if there are any long-term effects from low level exposures to different agents.
I did want to go back and address one of the questions about business systems before, and say that, you know, we have had problems in the past, and I think the LSU Medical Center certainly reflects other academic and non-academic health centers, in looking at the value that we get out of administrative information systems.
We see that often we base clinical decisions upon data that is collected through those systems, and there is a question as to the validity of that data, number one.
Number two, I think there has been a lack of advancement in some of the technology. Some of the technology that we have deployed, frankly, is 15 years old.
Part of that has to do with -- and this is in deference to Lew Lorton -- with the lack of open standards being used.
We feel that open standards and vendor independent solutions allow for innovations. That is one of the things that we certainly hope can be promoted, to move things forward.
So, we are certainly looking in this partnership to try to advance the standards process. Given the size of our organization that we serve -- close to two million people -- we are typical in the private world in terms of having a lot of complex relationships.
We can simply not impose the same clinical information system on all of our partners. If we can't do that, we can't communicate with them.
So, we have a real business need for interoperability of computerized patient record systems, as Dr. Kolodner had alluded to about the systems.
The only way for that is for the core elements to come together. We think that is something that the GCPR with its footprint can help facilitate, with your help.
DR. DETMER: Thank you. You know, it is interesting, in the preface for this latest edition, I was also reflecting on what has happened since 1989, when we first started getting together to talk about this whole issue, this iteration, at least.
I ultimately see three computer-based health records, the CPR that is computer-based personal record, the CPR that is the patient record as it relates to health delivery systems, and the computer based population, whether you are worrying about what happened to a cohort of soldiers somewhere or populations for other purposes.
They do need to have the right kind of interface, and it really is quite a daunting challenge. I think that is really interesting because all of that sort of emerged, in some respects, since we first started sitting around a desk or a table at IOM and working on this.
I have Jeff, Peter wanted to make a comment. I want Betsy Humphries, if she would, please, to comment on the study you were starting at the NRC, if you would just make a comment about it. It could be interesting for the group to hear about that. Then Gellman, also, and then we will probably have to wrap it up. And Kathleen, sorry.
MR. BLAIR: I thought the presentations were just excellent. As many times as we have gone through this, there were just a lot of different new perspectives that I found very helpful.
I would like to ask each of you to just pick one example of the many that you indicated in terms of impediments that still remain, and I would like you to refine it in terms of which one of the impediments you feel would benefit most by government facilitation or assistance.
This is in the context that the role of the NCVHS right now is to make recommendations to Congress on electronic health records for legislation about two years from now.
We need to focus on addressing impediments. The piece is that some of it could be legislative, like privacy and confidentiality. Some of it could be facilitating standards activities like interoperability and clinical specificities. Some of it could just be as a forum.
It is a broad piece, but I would just like to hear what your number one area is, where you feel that we could make a recommendation that would assist.
COL. RAY: I guess I get to go first. I am closest to him. I think if I looked today, what we have been trying to work on for the past year is really focused on the clinical lexicon.
We feel that if we could come up with a standard health data dictionary -- you know, 3M has been working on one with DISC for the last 20 years -- if we could just say this is our selection for our standards, what that helps me to do is to take -- right now under the CDCS-2 program office I own 49 clinical systems, of which none of them talk to each other.
If I could have a standard health data dictionary which I could map those two, then from mapping it I could store those into a new clinical data repository and, from there, share that information.
Today, with all these systems having non-standard data in them, if I had a standard data dictionary and a clinical lexicon, I really believe I could map those elements to it.
DR. KOLODNER: I agree, that the biggest barrier that I see is the clinical vocabulary, clinical lexicon. Again, that is an area where even in the partnership we haven't defined what goes into which term, but that area of the terms and the descriptions.
In thinking about it, that is something that you need to advance, but I would expect that you are probably going to have to depend on lots of others to do that.
The role there is more of a coordination and maybe when there are two competing ones, just choosing one, as I know we have been trying to do with HIPAA and some of the areas of overlapping standards.
I think one of the other areas that may be appropriate -- and part of it may be not understanding the full boundaries of the committee -- but the issue of the data ownership, I think, is one that I think probably needs to get clarified, and would be one that would be doable that would then lay the foundation for a lot of other things, including privacy and confidentiality and other kinds of things.
I think that may be an opportune one that isn't often focused on, but may be an important thread.
MR. LORTON: There are so many it is like being at Baskin-Robbins. I agree with Lynn Ray about the data lexicon. I would just embellish that a little bit as to, rather than look at something as profusing a static system, but develop also the functional thing where it can change rapidly.
I think that is the most frustrating thing. A very quick, when I did this work at the Medical Foundation, we used SNOMED, which is great as far as it went.
Then I would have a clinician come to me with, here we have these 100 terms, and I had no way to get this implemented because he is standing there with the data in his hand that he wants entered.
So, that is the kind of -- that is the way research is done, that is the way health care is done. We need health care to be responsive to the needs of the clinical environment.
MR. WAEGEMANN: As I have said earlier, I think that documentation rules are the most important requirement.
I think to get providers to a point where they can create legal and archival electronic medical records, which do not require that paper back-up records have to be kept, would be the biggest step forward.
My second option would, of course, be interoperability. I think we would need not only an afternoon. We would probably need a day or two just to explore interoperability.
We have identified 19 items. I just believe that a lexicon isn't doing it. An identifier isn't doing it. An operating system isn't doing it.
We need to look at all of these if you want to look at interoperability. We have right now 56 companies developing medical records systems, on the Microsoft platform.
We have at this point 47 companies probably by June, we have 70 companies working, then, on the OMG carbol platform.
These are just two platforms, somewhat compatible. This is just one level out of the 19, and we could look at all of them, and it is much more complex.
But I think by focusing on documentational rules, one could really make a difference, because any provider could really save tremendous money. It could really enable and empower vendors as well as providers.
I also would like to add here that we have to realize that we haven't talked in this whole session about the various niches of medical records, which have been paperless, which have been quite successful: home health care, physicians offices, some of the nursing systems out there.
So, therefore, it is so easy to just focus on the hospital. I warn that this might be wrong, because there are additional areas that one needs to address here.
When we just look at the hospital, what we need to realize is that what we have right now is a stage of homegrown system.
We only get somewhere in the hospital field when we can say, here is one hospital which is implementing it in June 1998, and within six months an additional 500 hospitals can implement the same.
This is the big vision that we have to keep in mind, that we have to get from homegrown systems to commercial systems which can be implemented on a large scale and which can bring the benefits that we need.
DR. TANG: I think I would have to support the terminology as being the top priority. As the advisor to HHS, I think it is particularly relevant for NCVHS.
I think there is no one better than the government to increase the attention and priority of an issue. I think there is a lot of money that needs to be -- there is a lot of funding that is required, and this is a public good and I think it is an appropriate role for the Federal Government.
The Federal Government is certainly one of the beneficiaries of the use of aggregate data. So, I think that is what puts it at the top.
One last thing, the individual identifier, if I see a patient and they have a lab and an X-ray, that is three different numbers. That is one of my biggest impediments, already having a record, of getting the patient data.
DR. DETMER: Betsy, would you share with us? Come up to the mike and just tell briefly. I think some people know that, but I don't know if everyone does, and I think it will be of interest to the committee as it relates to this issue.
MS. HUMPHRIES: Yes, the National Library of Medicine is funding another study by the Computer Science and Telecommunications Board, having been overjoyed with the results of the last one, which you are all familiar with, in terms of the privacy and security recommendations.
This new study is going to focus on the specific health requirements for the next generation internet. That is, if the next generation internet, the public network that it becomes, is to be useful for routine health applications in health care and public health, so that people will not have to build private networks in order to get either the guaranteed quality of service, response times, security, confidentiality features, then what does that mean.
Also, to have this prestigious group which will include a lot -- as you would imagine -- of strong network technical expertise as well as a broad view of informatics and health care and what have you, public health.
They will also be focusing on are there, in effect -- are the requirements for health in conflict with any of the other large scale routine uses that society would like to see in the next generation internet.
I don't know that anyone expects the answer to that to be yes, but the issue of, okay, what is the most important thing in terms of guaranteed level of service for health applications looking for all kinds of things.
I mean, there is obviously the real high end application, sending the tremendous series of mammograms over the line.
Then there is the other issue, if everyone in the DoD and the VA and the Indian Health Service and all of their providers were, in fact, making routine use just to get the labs back and forth, we could scale this up and then you could have a traffic issue where certainly the routing and so forth has to handle a lot of this.
Then there is the issue of maybe I don't need instantaneously. Can you guarantee me that it will arrive within N amount of time and that there will not be a missing bit here or there that might affect the diagnosis or the decision.
DR. DETMER: I appreciate your sharing that with us. I also want to compliment, I think, you and Don Lindburgh.
I think part of the reason you have gotten excellent responses from those studies is you have sent them very focused questions. I think this is another excellent example of that.
MS. HUMPHRIES: I think that is fair, and they will also be discussing this whole problem directly with people who have had to build their own networks or who have not been able to use public communication networks, and exactly what were the issues behind those decisions.
MR. GELLMAN: I have got a question for the government CPR guys. You have to operate uniquely, I believe, in this country at least for anyone doing this, in that you have to operate under the Privacy Act, which is probably the only full-fledged code of fair information practices applicable to health records.
I wonder if you could comment on what are the consequences, implications, problems, whatever that you see as a result of that, if any.
DR. KOLODNER: At this point, I think those of us who are involved at this level of the partnership are not knowledgeable in the depth that we would need.
One of the reasons, if you look again at that graphics slide, the security and actually security and privacy is something that needs to be built in, in a way that supports all of our legal requirements.
I would be turning to our medical records, our HIM types of people and say, okay, in order to do this, how do we validate that there was a signed release, and where do we document that, and how do we get that incorporated into the system, or document that there are emergencies.
DR. DETMER: The documentation challenge.
DR. KOLODNER: The documentation challenge. Obviously, what we have, at least in VA and I believe that is true in the other agencies, is we have some rulings by our general counsel, that in the way that we have put our system together and the encryption that we have, that we have a way of being able to demonstrate that it has not changed, and how we make changes, additions and things.
So, we would be looking to see, how do we do it, in what is in essence a unique way right now but hopefully would evolve as documentation standards go, to fit into more general standards as they fit.
We would, in essence, be turning to our work group to say, what does that really mean and what do we need to do to support the requirements from legislation.
MR. GELLMAN: Let me just make two points. One is, I am actually asking a question and I didn't expect to have a really good answer right off the top of your head. I am actually less interested in security than all the other requirements. Security is technical and boring; not unimportant. It is all the other requirements.
Just keep the question in mind as you keep going and you may have something to come back and tell us one day.
DR. DETMER: I am sure you will, as a matter of fact, because you are already living under a mandate that we will eventually be spreading in some form to the rest of the country. Kathleen?
DR. FYFFE: To get back to Jim Scanlon's question about the state of readiness of the hospitals in the United States in terms of being ready for computerized or electronic patient record, would it be possible -- and I am making this as a suggestion -- to talk with the three or four or five major hospital information system vendors that probably have installations in half the hospitals in the United States.
Have we thought about doing that at all? Actually, I am sort of going --
MR. SCANLON: We have heard from a number of them.
MR. LORTON: Of course, they would be happy to talk to them. I was talking to somebody from SMS last night, who volunteered to drive down and sit in the audience in case there were any questions.
DR. STARFIELD: Could we do the same for the major outpatient vendors, if there are any?
DR. DETMER: Anybody; we can mix and match at will.
MR. SCANLON: What are the questions that you would focus on if they were here again?
MS. FYFFE: Well, with all due respect to the panel, I think we have to talk with the people who are out there actually designing and doing it.
MR. LORTON: Was that, with all due respect, you don't know anything?
MS. FYFFE: Yes, well, I mean, I would like to hear from some of the major vendors.
DR. DETMER: That is no problem. This is the first of a series. That is no problem. It is the start of this thing, not the end of it.
MR. LORTON: Kathleen, we were just talking. We have a meeting in July here in town, and we were tossing around the idea of trying to get the major system vendors, each one of them contribute a high ranking person to talk on a panel about how they would transition to this new stage.
MS. FYFFE: Yes, that is exactly what I am driving at.
MR. LORTON: We have already started creating the lists.
DR. DETMER: That is no problem.
MR. MAYES: Just to clarify, at least three -- I am not sure about Lew and Peter -- but at least three of the people on the panel have actually operationalized and built very large systems. It is not like they are not familiar with it.
MS. FYFFE: That is fine, but I think in terms of numbers of hospitals; okay.
MR. MAYES: You are talking about hundreds, though.
MS. FYFFE: Okay, and if there are approximately 4,000 to 5,000 hospitals in this country -- and of course, the number keeps going down -- what I would like to hear is three or four of the five major vendor systems who cover maybe 2,000 of the hospitals and have a discussion.
MR. BLAIR: One of the things that we have to balance that with -- I think it is very appropriate to have them in. I absolutely agree, and I think we already -- we didn't communicate well enough, that a number of us had already started to this about that.
But I think we need a diversity. In the past few years a lot of the momentum in moving toward computer based patient record systems has switched to the ambulatory side as outpatient side as managed care and integrated health delivery systems have come into place.
So, you have a lot of technical innovation that is taking place over many, many years in an acute care setting. However, the last, I would say, five or six years a lot of the activity is in the ambulatory.
So, we need to have at least those two areas and then, of course, this post-acute care area, I think we should hear from some vendors.
DR. DETMER: My goal, actually, is to have us not only look at the past as it was and legacy and big systems, but also the future as it is moving.
So, I mean, I think what we will want to do is tap people's thoughts in putting together the right mix of folks to let us know what their challenges and successes are.
MS. FYFFE: Sure. Just as a perspective -- and you have probably figured out I am a bean counter -- when WEETI did an informal or semi-formal survey a few years ago, they found that there were about 200 different major hospital information systems in the United States, and about 2,000 to 3,000 practice management systems.
I would be willing to bet that if you took a small percentage of those, it is the old 20/80 rule. You know, 20 percent of those might be covering 80 percent of the market. That is really what I am driving at.
COL. RAY: I would just caution, one, if you just go after -- remember the doughnut that I talked about. Those major vendors are in that outer ring.
What you also have to find now is in this complex information society is you need true integraters that can help us build those pieces and parts to help integrate all of those major vendors on the outside.
Many of these vendors sell applications, but they don't do a very good job of integration.
DR. DETMER: I was wanting to describe that as your leaky O ring.
MS. FYFFE: The word that used to strike fear in my heart when I was a systems designer was the word interface. Okay, that got a rise out of some of you, but it is a problem. It is really a problem.
Hopefully, if we get some folks in here from the major companies, they will talk to the people who are in the back rooms having to do these interfaces among the other systems.
DR. DETMER: I am going to have to cut this off, because I know people are going to be leaving and we have some other business that we need to finish.
This was a terrific introduction, though, to this iteration of our task and I can't thank you enough. So, thank you very, very much, and good luck to you, as we will be staying in touch.
MS. GREENBERG: While Rob is here, I would just like to thank the VA for supporting our broadcast of these meetings.
DR. DETMER: I think it has been very useful. Thank you very much.
MS. GREENBERG: I have several staff who are listening as we speak, I think.
DR. DETMER: I am willing to see us take a break or we can press on with the other two reports. What is your pleasure?
Okay, we will take a literally quick break and then we will come back. I want to finish before people leave.
(Brief recess.)
DR. DETMER: We will go next with a discussion on future meetings and agenda topics and then we will come back to our two committee groups.
Agenda Item: Future Meetings and Agenda Topics.
MS. GREENBERG: Okay, my guide here is tab K.
DR. DETMER: She has assured me this is a brief item.
MS. GREENBERG: Yes, tab K, pages 12 and 13, and in particular page 13, but that is the June agenda items. I think if you look just to your left here on page 12, you will see that we did have all of these presentations, and maybe more.
They feed into some of the ones for the June, and also may need to be revisited in June. We don't have that listed under June.
We had our initial discussion on data quality and that was a decision to look at really the quality of the data for looking at quality assessment before we had a panel on quality more focused on the quality of care, although I think as we saw this morning, it is impossible to separate them.
DR. DETMER: What I think we were wanting here, I think, if I remember our executive committee issue, was whether John Lumpkin's group might have a shot at a draft for us to look at, relating to vision or something like that, as well.
DR. STARFIELD: With regard to claim attachments.
DR. DETMER: Excuse me?
DR. STARFIELD: With regard to claim attachments or no?
DR. DETMER: No, we are charged with trying to come up with data sets that we think are relevant for the future.
One of the questions is, if you are going to do that, what kind of a system are you even thinking about that you want to drive to, both as a system for quality and management?
DR. STARFIELD: I misunderstood. I was thinking of the work group, the K-2 work group.
DR. DETMER: No, I was thinking about Lumpkin was going to head up a group. That is not what you --
MS. GREENBERG: You were talking about the -- it is here in the work plan under H. Let's see, there is a quality kind of initiative that is actually on page 5 that John was asked to head up.
DR. DETMER: Right, we will see if he is ready. He may not be by then.
MS. GREENBERG: I think we had also talked -- maybe Lisa can help us here -- about having a panel of some of the really leading people in addition to those around the table and those that we heard from today, who have really done an enormous amount of thinking, such as Kathy Lohr and some others.
DR. DETMER: Lisa is nodding. I think I recall that conversation, too. Is that what your recollection is?
MS. IEZZONI: Yes.
MS. GREENBERG: So, I think at a minimum we could put that together for the June meeting.
DR. DETMER: All right, we will see if that is a possibility.
MS. GREENBERG: Maybe work with John on how that fits into what he might want to do. It might jump start to his efforts.
DR. DETMER: Right.
MS. GREENBERG: We will have had a session in Chicago, hopefully May 4 and 5, on unique identifier. We are back to page 13 here of tab K.
The day before the full committee meeting, we are planning to have one here in Washington. The full committee, I believe, is the 16th and 17th, and this would be the 15th; is that right, Lynette?
MS. ARAKI: Yes.
MS. GREENBERG: So, we wouldn't be ready, probably, to develop recommendations but we would certainly be ready to at least have some dialogue or discussion about what came out of these two meetings.
DR. DETMER: That is fair. I think what we were aiming for was probably some recommendation by September, if I was recollecting, to summarize that.
MS. GREENBERG: We need to come to closure on the third hearing, if we are going to have it, the southern hearing that somehow ended up in Denver.
DR. DETMER: Well, it is south of Alberta.
MS. GREENBERG: I don't know whether you would want to revisit that or not, but I think that we are trying to zero in on a date for that, either the third or the fourth week of July.
MS. ARAKI: Actually, I was told to hold on because we were going to discuss that.
DR. STARFIELD: It has been a long time since we were polled and we need to get these things on the calendar.
DR. DETMER: I can't hear you, Barbara.
DR. STARFIELD: It has been a long time since we were polled.
MS. GREENBERG: Yes, so things may have changed, too. I think in all of these things, I think we get a less than good response rate often when we poll people, lower than 70 percent, often.
DR. DETMER: What do you mean?
MS. GREENBERG: We don't hear back from people often, when we pool, to some degree.
MS. ARAKI: Yes, people don't respond when we are polling the dates and locations and to check your calendars. That makes it difficult to select the meeting date and time.
DR. DETMER: Understandably.
DR. STARFIELD: Why don't you give priority to the people who respond quickly?
DR. DETMER: Like we just ought to have a 48-hour turn around or something like that, or something to get back, and if not call them, and then if you can't get them, let me know.
MS. FYFFE: Some very nice people have called me and I answered their questions. What is the problem?
MS. GREENBERG: I don't know. We seem to have a hard time coming to closure on this. We will redouble our efforts, because we have to come to closure on that hearing.
At one point we actually thought that the NCVHS CINSTAT meeting on health statistics might even take place in the spring. That is clearly not going to happen, but I think we have had some productive discussions just during these two days.
Dan tells me that he thinks we are lurching forward. So, we will have something to report on that.
DR. DETMER: I thought you said surge.
MS. GREENBERG: Surging forward. But we won't have certainly a report of the workshop because it won't have taken place yet.
I think actually a number of activities, hopefully, will have taken place by then, preliminary types of activities. So, we will probably have something on that.
There was a request for survey integration status report, and we did, I think, get really -- launch the population-based data work group now of the subcommittee on health data needs, standards and security.
That subcommittee did have a report on survey integration probably over a year ago now -- I think it was a year ago February or something. I think we have to talk within that work group --
DR. STARFIELD: Yes, we should have a copy of that to review.
MS. GREENBERG: -- as to whether that should come to the full committee or not, but that is something that Paul has raised also.
We have talked about having the presentations on the status of, and moving toward ICD-X-CM and ICD-X-PCS. There was original interest in having it at the March meeting, this meeting.
Because both of those have been out for a public comment period, and I know from the ICD-X-CM point of view, that it is sort of -- we had quite a few comments but this last week, these last few days since last Friday which was the unofficial closing date, although we are still taking some comments, they have really been pouring in.
That is why we delayed it until June, but we should have the results of the X-PCS field test, all the comments we have gotten on both, and hopefully have a good discussion on that.
As Jim has mentioned, the X-CM is actually going to be on the data council agenda next week, and I think your mentioning this as being on the radar screen of the committee raises its importance to the data council also.
I have something for the executive subcommittee members or anyone else who is interested as kind of a work plan that our reporter, Susan Kinnen, has developed for our -- I refer to it as a two-year annual report and Don has suggested maybe we should call it a biennial report.
It may resume to being an annual report, but this is the report of the overall committee, which builds on -- certainly will incorporate information from the HIPAA report, but also cover the full range of activities of the committee, as our annual report used to, but will be differently organized, thematically rather than by subcommittee.
A major piece of it is to be this vision statement that Susan is working on with the executive subcommittee and others.
DR. DETMER: So there is not misunderstanding on that, what we were thinking about doing was the executive committee drafting this, but then obviously circulating to the full committee for input and then sign-off.
MS. GREENBERG: That definitely would happen at the June meeting.
DR. DETMER: Yes. I just want to make sure people understand that this isn't something that is going to be not weighed in on.
MS. GREENBERG: Definitely. I think at the June meeting we will want a substantive discussion on that -- we will, that is our plan, on that document, as well as we will have a detailed proposed outline for the annual report. So, we will need some discussion on that.
DR. DETMER: What I would actually like to do, if we could, is have a draft such that we really are discussing the draft at the meeting.
MS. GREENBERG: Oh, definitely. Oh, not the draft annual report.
DR. DETMER: That is what I am saying. I would like to see us being able to discuss the draft at the meeting.
MS. GREENBERG: We were actually thinking in terms of having a draft of the full annual report at the November meeting, because it is a 1997-1998 report.
DR. DETMER: Okay.
MS. GREENBERG: The vision statement, we will discuss the draft at the June meeting.
DR. DETMER: All right.
MS. GREENBERG: Yesterday -- I am not quite sure where we are on claim attachments and whether we will be prepared to come forward with --
DR. STARFIELD: Keep going, Marjorie, you are giving my report.
MS. GREENBERG: What?
DR. STARFIELD: Keep going. You are giving my report. That is fine.
MS. GREENBERG: Barbara will tell us what we will be able to talk about in June, whether the subcommittee is likely to have any recommendations.
DR. COHN: I just had a question. Are you giving us a report or are you asking us for input on what we would like in the June agenda?
MS. GREENBERG: What I am telling you is all the things that at least at the executive subcommittee meeting, and polling from the full committee meeting that have been put on the agenda.
DR. DETMER: Tentatively. The point is, we will open this up and say, do you want those on, do you want them off, do you have other things that you want on.
DR. COHN: So, we should e mail you if we have other things.
DR. DETMER: I think if you have other things, too. She will be talking about claim attachments.
MS. GREENBERG: I won't mention any more about claim attachments, but we thought it was something that would be coming up at the June meeting, possibly recommendations.
Jeff has asked to have some time at the June meeting to present the clinical inventory which ANSI HISB is preparing, the inventory of clinical information standards. Did I get the name right, Jeff? It has just the information.
The templates were due about a week ago and they are putting that all together and they will have something to present -- final -- to the committee and department at the June meeting.
MR. BLAIR: We have had 77 submissions about an average of 10 pages apiece. So, it is exceeding our expectations in terms of the response. So, yes, I think I will try to have hard copies, if possible, for the committee members in June.
Actually, in the May time frame, if I can give you an update and give you any information before then, that might be helpful, on what we want to do to analyze that information.
In June, it will be the compilation of all the information. We are going to need to go further and analyze that information.
Then the other thing that I also would like to talk about maybe in that May session, if possible, if we could pull together some kind of a work plan, because even analyzing the standards, probably isn't all that we are really going to need to do to prepare for our recommendations on electronic patient records.
There are a number of other topics and areas, some of them, you know, we hit today. So, I was mentioning to Don that I will just send an e mail around to, I guess, everybody in Barbara Starfield's committee.
DR. DETMER: Right.
MR. BLAIR: With just some straw person suggestions as to things that we may want to include in a work plan to have it a little more focused and structured.
MS. GREENBERG: Okay, you are talking about the May 4, 5 meeting in Chicago that we talked yesterday about carving out a few hours on that?
MR. BLAIR: Yes, if that is possible.
MS. GREENBERG: On some of these subcommittee activities.
DR. COHN: I was actually just going to add two comments. One is that I think the overall agenda sounds fine. I was hoping that we could have a report on the terminology conference that was held last November to the committee, which I think, based on our panel presentation here, it might be very pertinent.
DR. DETMER: Right. We will have to refocus this as we get closer, but that is what we want to do right now.
MR. COHN: I had just one other comment, and I don't know if this needs to be further structured, but I think it is time for the committee to revisit, or further visit, the whole issue of data.
About a year ago we sent the Secretary a letter saying that we were --
MS. GREENBERG: Do you mean content?
DR. COHN: Yes, content, in relationship to the standards that we have produced, in relationship to work that Bob Mayes has done around registries and approaches to how we are going to deal with all of this.
It has not been an area that we have made further progress on. So, I don't know what it needs to look like, but I do think it is time for us to go back and revisit some of the overarching issues around this.
DR. DETMER: Okay, other items that we put on here?
MS. GREENBERG: I was just going to mention that we may want to revisit the E&M issue.
DR. DETMER: We may want to follow up on E&M, too.
MS. GREENBERG: I don't know whether there will be anything new on the Bubba.
DR. DETMER: What I think we may need to do on some of these issues, where we can, is just write up a summary and just have it as an information item, and then if people have questions, we can talk about it, but otherwise probably try to handle some of these as just information items.
MR. SCANLON: We will have the quality commission report, of course, but maybe the quality care panel can submit recommendations. I don't know what the follow up will be.
DR. DETMER: Okay, any others? Okay, that is great.
MS. GREENBERG: This will keep us busy.
DR. DETMER: Oh, yes, that is never a problem. Now, let's go back to --
DR. STARFIELD: Do we need dates?
DR. DETMER: They are set. The June meeting is set.
MS. GREENBERG: It is the 16th and 17th.
DR. STARFIELD: We are set only through September, though, aren't we? Do we have November?
MS. GREENBERG: We have November also, I think, 11-12 or 12-13. I have it here somewhere.
DR. DETMER: Yes, so it is just a matter of reminding you of the dates, but the also respond on the Denver dates or whatever other dates. Okay, thank you.
All right, who is next?
MS. GREENBERG: Lisa has some.
DR. DETMER: All right, Lisa
DR. IEZZONI: Yes, I would really like to. I am going to roll out of here as soon as I finish talking.
Agenda Item: Reports on Subcommittee Work Plans and Recommendations - Subcommittee on Population-Specific Issues.
DR. IEZZONI: The subcommittee on population-specific issues, basically what has happened since I reported to you since last November is that we have held two series of hearings, one in January in Washington, D.C., and one in Phoenix in the first week in February, where we basically gathered a lot of information.
Fortunately, because of diligent staff work, even at the Washington meeting, from people from states, which has been really wonderful, to have people coming in from states to talk to us, about Medicaid managed care.
Again, what we are doing is asking, what are the major questions that people have about Medicaid managed care, what are the data that they need to feel that those questions have been answered, and what are the gaps in the data, and what would we recommend to the Secretary around filling these gaps, recalling, again, that we have a contract with Sarah Rosenbaum from George Washington University.
Also, we are to review current contractual language around data reporting for Medicaid managed care, and also propose prototype data reporting contract language for the future.
That is where that effort currently stands. We will be having another hearing in Massachusetts, in Boston, in April.
We are proceeding on the meeting on the islands and territories. It will be held in mid-July. We are very grateful to the Office of Minority Health. Olivia Carter-Pokras' office has agreed to pay for people to come in and talk to us from the Pacific Islands.
DR. DETMER: Gee, I thought the committee was traveling to Bali or something.
DR. IEZZONI: No, we are not traveling 4,000 miles. We are hoping that we will also be able to have people come in from Puerto Rico and the Virgin Islands.
It actually is shaping up to be a very exciting and informative meeting. We are actually totally thrilled about it.
Jim, we have to thank you and a variety of other staff -- Del and Lynette. The people on your staff have really pitched in and this is now looking as if it is going to be a very, very exciting meeting.
What we learned yesterday from somebody from the Department of the Interior who came to speak to us in our breakout session was that we could really make a difference by conveying some of the needs of the people from the islands and territories around very fundamental things to improve their information infrastructures.
So, we are hopeful that even in these two days we will be able to potentially make some very useful recommendations.
DR. DETMER: Lisa, if I might give an example, some of the infrastructure is very basic, like back-up generators.
DR. IEZZONI: Yes, back-up generator for the computer systems, when they die -- which they do do, apparently.
So, the third and final thing that I wanted to talk about was the hearing that we held on March 2, which was Monday, in Baltimore. I want to thank Bob Mayes, Bob Moore and Carolyn Rimes -- who is not here -- and the other people from HCFA for hosting us at what was actually a superb day, although a sobering day, I think.
It was one which I would like to just summarize by giving you a little tiny bit of information -- this will take a minute -- from the Balanced Budget Act.
The Balanced Budget Act requires that a prospective payment system be implemented for skilled nursing facilities by July of 1998, that a prospective payment system be implemented for home health agencies by October of 1999, that a prospective payment system be implemented for rehabilitation hospitals by October of the year 2000, and that a prospective payment system be implemented for long-term care hospitals by October of 1999.
For three of these -- for the home health agencies, the rehab hospitals and the long-term care hospitals -- the prospective payment systems have not yet been created.
In fact, for two of them -- the rehab hospitals and the long-term care hospitals -- which will apparently have their prospective payment systems be based on the minimum data set post-acute care method, the minimum data set post-acute method hasn't been finished yet.
They are still drafting it. We saw draft number six on Monday when we met. This is going to require enormous new data gathering procedures.
I think there was something like -- Bob, correct me, I am probably going to be off by an order of magnitude -- 50,000 or something like that, home health agencies.
MR. MAYES: Yes, there are.
DR. IEZZONI: Fifty thousand home health agencies. They are talking about implementing potentially a 149-data element data collection procedure for supporting the prospective payment system for home health agencies.
So, this just gives you a sense of the magnitude. Trust me, if HCFA were talking about implementing a 149-data element data collection procedure for the 15 percent of Medicare beneficiaries and Medicare capitated products, all hell would be breaking loose right now.
That is why I asked the question that I did yesterday. Is this going on behind the scenes, that all these new data gathering mandates are coming forth.
They are not, but it is obvious that people haven't really understood the impact and the implications of some of this.
We heard on Monday from basically the developer and vendor of the home health system, who I do not think answered the question in a realistic way when they said that this data collection mandate will have no impact on the home health agencies.
The woman -- the nurse -- who was from the developer said that this will not make a difference in terms of the time that is required for documentation. That simply is difficulty to give credulity to.
DR. DETMER: Bob, I just want to say, are all the details on all of these things on your web site at HCFA or not?
DR. IEZZONI: No.
MR. MAYES: No, I don't believe so. On the web site, we really don't publish any of the draft materials.
DR. DETMER: It would be useful to at least have some summary.
MR. MAYES: The minimum data set, the long-term care minimum data set.
DR. IEZZONI: The long-term care minimum data set, it is established, it is in place. They can create the skilled nursing facility. What we are talking about is the other things.
DR. DETMER: That is what I am saying, these new things.
DR. IEZZONI: This needs to be on the agenda of the committee, because these are attachment issues. This is on their data gathering mandates that need to be, I think, on our radar screen.
We are thinking about having the population-specific subcommittee have as its next big project kind of thinking about an overview approach toward gathering data for this area.
We are not going to call it post-acute care. We have to come up with a better name for it.
It is obvious that Congress has mandated these requirements that, suffice it to say, are going to be challenging to me and are going to have enormous documentation burden, not only, let me say, on providers but potentially on patients.
The only way that you can validate the credibility of this home health agency data is by having patients be examined by two different people.
So, you are going to be having patients be visited twice. So, you know, there are a lot of issues around this. I think our committee needs to be a lot more on to of this than we have been.
I think a lot of this, again, is Congressionally mandated. So, HCFA has to respond. I think that our committee has the potential to help HCFA think through how to respond to some things that, in fact, are going to be a bit difficult to fulfill.
MR. MAYES: If I can just add one thing, I mean, one other -- the minimum data set in long-term care is, I think you will discover, becoming the de facto center of long-term care information systems, when we talk about CPR.
DR. IEZZONI: Right; exactly.
MR. MAYES: This is such detailed reporting that is coming out of this, that it will have a profound impact on the overall structure of computer-based health records, especially in these niche markets.
DR. STARFIELD: We can review that again.
DR. IEZZONI: The resident assessment document for the minimum data set is this thick.
MR. MAYES: It is 700 questions.
DR. IEZZONI: It is this thick. It is the computer based medical record for these places. I think our committee has to begin to think about this. We should probably figure out exactly how this should be done. It may be too big an issue just for --
DR. DETMER: It doesn't sound like a briefing. It sounds like a lengthening.
DR. IEZZONI: I am sorry, but all of this is a really big deal and we need to be aware of it and on top of it.
MR. MAYES: I would just like to say there are quite a number of people from HCFA staff who don't normally come to these meetings, that I heard a lot of very positive feedback from the staff, who thought that the hearings were extremely helpful and of great interest to them, since they don't often get a chance to hear the outside world quite so clearly.
DR. DETMER: I appreciate that. I understand you love the cafeteria.
DR. IEZZONI: I was going to say, the cafeteria is worth the periodic meetings in Baltimore.
DR. DETMER: Are there other questions or comments for Lisa? All right, Barbara?
Agenda Item: Reports on Subcommittee Work Plans and Recommendations - Subcommittee on Health Data Needs, Standards and Security.
DR. STARFIELD: We were the health data needs, standards and security subcommittee. We were pleased to co-sponsor that meeting. We were also very pleased that you want to take this on in your committee.
DR. IEZZONI: No, I think it is still an issue about whether there needs to be some larger involvement, frankly.
DR. STARFIELD: It is a really big issue.
DR. IEZZONI: It is a really big issue.
DR. STARFIELD: Okay, my brief report is going to be even briefer because Marjorie did a lot of it. It turns out that the June meeting was projected to include things that are mostly in the jurisdiction of the health data needs, standards and security -- can we get a shorter name for this -- subcommittee.
There are three things that I just want to mention briefly. Marjorie has already mentioned all of them.
First of all, we now have a work group on population-based issues and hopefully we may have a work plan for you to approve in June. I would think we could do that.
We talked about this conference -- are we calling it a conference, symposium or what -- that is going to be jointly held with the National Academy, Institute of Medicine, that Ed Sondik has been suggesting that we undertake with this.
I think Dan has been working on this. Within a month he has really come very far and I think we are well on our way to having something that is going to be successful.
It is going to deal with, what are health statistics and what is a health statistics system, and how can we do better and know we are doing better. That is, how can we evaluate what we are doing. Dan may want to add to that in a minute.
The second thing is the national health surveys, not only the integration but what else needs to be done in terms of looking at the surveys, doing away with some, integrating them, adding some things.
Paul Newacheck, virtually overnight, helped us to develop a plan on that. So, we will have more on that afterwards.
The last thing we did was deal with claim attachments. That is proving to be a nightmare because we can't even define what they are. Clem defines them as anything that can be stapled.
DR. DETMER: Or photographed.
DR. STARFIELD: He gave us a very nice typology of a five-category kind of thing, going all the way from things that are in forms, to things that are completely amorphous, which I think was actually very helpful.
In terms of developing standards for that, I think probably what we are going to do is come up with a pathway to standards, rather than standards themselves.
HCFA is doing a lot of work on this proof of concept plan. Clem is working very actively on developing some standards for some levels of these attachments.
We certainly will not be able to, even if we wanted to -- and we probably don't -- want to develop standards for attachments as they are today, because they will be replaced eventually when we get the computer-based record.
We will be meeting as a subcommittee, and perhaps with input from the whole committee in May in Chicago. I think a lot of people on the other subcommittees will be interested, in the unique identifier.
DR. DETMER: Right. Dan, do you want to add anything?
DR. FRIEDMAN: We had a very productive meeting this morning with Paul and Barbara and George and Ed Sondik and Marjorie and Jerry Henderschott, talking about what Marjorie calls the vision quest.
MS. GREENBERG: Actually, Ed Hunter coined that. I don't want to take responsibility -- I don't want to take credit.
DR. STARFIELD: Not to be confused with the vision statement.
DR. FRIEDMAN: So, you know, I think that we are reaching some explicit agreement on goals and, during the process of reaching some agreement on some nice iterative steps toward reaching those goals.
Hopefully within the next few weeks we will have developed essentially a work plan.
DR. DETMER: Good. Other questions or comments for Barbara? I certainly wouldn't change the name of the committee. If you want to just call it the health data committee as a way of shortening it, I think that would be fine.
MS. GREENBERG: Data needs, I call it.
DR. DETMER: We have a fairly thin group in the audience at this point, but we usually, before we adjourn, give anybody a chance to say something if they would like to the committee. If you have the spirit moving you, now is the time to move.
I am seeing some movement but not necessarily toward the Mike.
DR. HARDING: The executive committee thing, there was a thing on attendance policy and you were going to bring this up and discuss it with us during this meeting.
DR. DETMER: Yes. What the executive committee talked about was that we obviously would be urging people to come to the meeting and if, in fact, they can't, to let me know, if they cannot.
Then, also, our minutes will reflect who wasn't there as well as who was. I think those were the two major points.
I think actually people have been working pretty diligently to, in fact, attend, and also our subcommittee and work groups. So, I appreciate that.
On the other hand, it is also very apparent that with the work we have got, people need to attend. So, that is really the spirit of it. Did you have a comment or a question you wanted to make?
DR. HARDING: No, I think the idea of 20 days is a reasonable one. I think last year we went way over that. It gets to be difficult, I think, for everybody here to make it one day every two weeks loss of your time at home. But I think 20 days is a reasonable time.
More than that, that is pretty tough for somebody working. That is a reasonable thing to shoot at, I think, 20 days.
DR. DETMER: Other comments from others of you on that point?
MS. FYFFE: I agree. Actually, 20 days is a month. That is a whole work month.
MS. GREENBERG: Twenty days is what we actually tell people when we recruit them. Right there, they see it as a fair commitment. But you are saying that you really went over 20 days.
DR. DETMER: I think the K-2 really did give us more.
MS. GREENBERG: Depending on how many subcommittees a person participated in.
DR. HARDING: There have been seven days so far since January 1.
DR. DETMER: Yes, we are off to a roaring start this year as well. I think that is the point. We also need to keep an eye clearly on that.
MS. GREENBERG: It impacts on our budget also, particularly as we see the need to meet in other areas for good reasons.
That is why we need to think probably more about piggy-backing meetings together.
DR. HARDING: The July Denver thing, that is two days plus a day of travel. That is three days. Is that necessary or is that something that could be condensed and so forth?
I certainly want to participate and be fully involved, but any time that can be pulled together, that is very helpful.
DR. DETMER: I hear you. I think that is very well taken. I also particularly want to thank Kathleen, Dan and Paul for their willingness to be here. Obviously we have already benefitted from your counsel. That is wonderful.
I want to thank the rest of the committee and also the staff. I thought we had a very busy but very effective agenda, and that is great.
Are there any other new items? If not, I would take a motion to adjourn.
(Motion made and seconded.)
DR. DETMER: All right. Adjourned. Thank you.
(Whereupon, at 4:10 p.m., the meeting was adjourned.)