July 9, 1997 Public Forum: Infrastructure/Cross-cutting Issues

07/09/1997

DR. MOORE: The next team to be presenting is the cross-cutting team. As I said, we wanted to move along. I will try to gather the others as the team starts, so I'll be taking off. Bob Mayes and Mike Fitzmaurice.

The cross-cutting team. As I said before, this was a team that was put together to look across all of the different standards to make sure that we were coordinating our efforts, that we were doing things in sync, that we didn't have one team off on a similar issue, going in a different direction than another. It was a team that was pulling together how we were going to write the regs, how the regs would be similar, so when these regulations come out to you all, they don't look so different that it takes a complete read to know what was happening. So we are trying to make that effort look like one.

Bob Mayes and Mike Fitzmaurice from the -- Bob is from the Health Care Financing Administration. He works in their health standards quality area. Mike Fitzmaurice is from the Agency for Health Care Policy and Research. Mike, I'm not sure what department you're in there. But they were the co-chairs of this particular team, and they are going to cover what the cross-cutting issues are, what is the status, et cetera.

Bob, do you want to lead off?

DR. MAYES: Well, good morning, everybody. I'm going to try to get us a little bit closer back on to our agenda. I think a lot of the questions and issues, even the cross-cutting types of issues, will come up again in the other implementation team's presentations, and there they will be discussing them as they apply to specific proposed standards. So I think many of you will have an opportunity to ask those questions later on during the day.

Also, I would like to point out that the rest of the presentations -- structurally, HIPAA requires both the adoption of standards for electronic transactions, as well as the privacy legislation and standards. The whole approach has been structured in such a way that all of us have been dealing on the technical standards side. The privacy issues have been dealt with by both the National Committee on Vital and Health Statistics and another group within HHS. So we will not be presenting any further presentations on privacy. We will be covering technical security issues with the security teams.

The recommendations from the Secretary to Congress are due August 21. At that point, there will be the official recommendations, as I said, on privacy from the Secretary. So when we talk about security issues here, it is really from a more technical perspective.

The overall purpose and charge from the implementation team that handled infrastructure and cross-cutting issues was to provide as Bob Moore had said, overall guidance and coordination to the standards activities, specifically to the other five teams, and to monitor the activity, to track it, to act as a focal point for discrimination of information, provide outreach, as well as some specific areas in terms of dealing with implementation issues.

A big one is to develop and maintain a master data dictionary, and look at data structures for all the standards, to develop time lines for the entire project, and as well to monitor progress of the various teams.

We have been working on developing and coordinating issues such as the regulation development. We also do report directly to the Data Council and the NCVHS.

We have had a number of activities that we have been involved in on the team. These five are ones that have taken a lot of our time. The first was developing guiding principles, that is, when we looked at all the standards, what kind of criteria were we putting to the various candidates to decide which one might be the most appropriate standard to adopt.

As I mentioned, we have been working on developing a master data dictionary that really is not only going to support this initial activity in the transaction standards, but that we hope will form a groundwork for the broader issues of some of the computer-based health record standards that will be coming out later.

We have been working on drafting boilerplate regulation language. There are quite a number of regulations that will come out of this legislation. For those of you familiar with that process, it in and of itself is a complex process, regardless of what it is that you are actually trying to regulate. So given the number that we have to put out, we have been trying to standardize the language to smooth the review process and to smooth the administrative path that is required for regulation.

There are a wide variety of cross-cutting implementation issues, and I'll mention just a few of those, and more will come out as we begin to talk about the specific standards themselves. Then as Dr. Detmer and Bob have both mentioned, outreach and educational activities have taken up a very large part of our time and effort of the team members.

Just to in a little more detail on those activities, the guiding principles. What sort of things did we look at when the various teams looked at the standards that were either out there already or under development, or if new standards had to be developed, what sort of criteria would be put to them.

Obviously, it needed to improve the efficiency and effectiveness of the system. It is possible to have standards that actually make things more complicated than they already are, and those were ones that we were not particularly interested in. Equally important, it must meet the needs of the users.

We did want to have a consistency to the extent possible across the various standards. We felt that that would also then help us lower the implementation costs, that that is an important criteria.

We did not want to see the government getting in the position of developing and maintaining national standards. We felt that that was much more appropriately handled by the range of organizations that are already involved in doing that.

In addition, given the very aggressive time lines that the legislation handed us, it was felt that the standards should be able to be adopted in a timely fashion. Also, we have to address with these standards current business process needs. In business today, as all of you are well aware, the business of delivery of health care services in this country is undergoing dramatic and extremely rapid change. One of the things that we were very concerned with and continue to be concerned with in the adoption process of the standards is the fact that these standards, while meeting today's needs, need to be as flexible as possible to meet both tomorrow's business needs and to take advantage of new technologies as they become available. So to the extent possible, we would hope that the standard would be independent of any particular specific technological platform.

We would like it to be simple, but at the same time, precise and unambiguous. We are mandated to keep data and paperwork burdens as low as possible. As I have mentioned, the standard needs to be flexible in order to accommodate what will definitely be changes in the health care system in the years to come.

The next activity that we have been involved in heavily is the building of a data dictionary. To date, we have put in 27 other data dictionaries into ours, including 15 X12 transaction standards for a total of about 4300 data elements.

Now, this is just a beginning process, as I said. The early needs for a data dictionary really revolve around element definition, normalization, identifying gaps and that sort of thing.

As we move to the broader world of clinical computer-based health information, it will be very useful to develop a more sophisticated model and structure within which to develop a more sophisticated model and structure within which to examine the kind of content that we hope to have in these kind of records. We hope that this activity will lay the groundwork for that.

In the regulation development process, we have been able to move forward and expedite several things about that process. One of the things -- and we will talk to the specific standard in just a moment, but we have now been able to reach a point where we think we will be able to cover all of the transactions themselves within a single regulation, versus having to have a separate regulation for every single transaction.

We will however probably have to have separate regulations for each of the various supporting standards, that is, a regulation for each of the identifiers separately, the medical code sets issues and the security issues. But being able to take it from the range of 15 or 16 regulations down to the range of six or seven should help us in keeping up with the mandated time lines for adoption of these regulations.

Now, there are a number of cross-cutting issues that have begun to come up. Initially, the focus of much of our activity was trying to identify appropriate candidates to be recommended for adoption. So a lot of our early work was just looking at the standards themselves and looking at them in conjunction with the criteria that was listed earlier.

Now that we have pretty much made up our minds as to which standards we feel are most appropriate to recommend, we are beginning to turn our attentions more and more to the not-trivial issues of, well, now that you've got the standard, how do you actually implement it and make it real.

As we look at that broad topic, quite a number of issues come up. I'm just going to throw a few up on the screen. Others will be discussed in more detail, and I'm sure will come up in questions as we get into the specific transaction standards and other standards themselves.

One issue which -- and by the way, we have not resolved any of these; these are still issues, and this is one of the reasons that we come before groups, is to try to get their opinions and ideas on how we might solve these. But certainly one issue is a conformance testing issue.

You've got a standard, you have put an implementation guide out. I say that my implementation of that standard is a legitimate one; who is to say if it is or isn't. So who does conformance testing, who is going to pay for it, who will monitor the testers. There is a variety of testers around that.

Data dictionary implementation guides. As I said, we have begun work on master data dictionaries. However, data dictionaries and guides are dynamic entities. So the question arises again, who maintains them, who pays for that maintenance and upkeep?

The legislation allows modification of standards to be made on an annual basis. The process for that needs to be spelled out. As I said earlier, while we want to have standards, we need to recognize that standards change over time in response to changes in both the business needs and the technology available to meet those needs.

So we need to develop explicitly a process which not only recognizes that there may be change, but in fact, encourages that change to take place as responsively as possible.

There are some specific issues. Some of these standards, it depends upon whether you are an employer, a plan or a provider. As many of you know, there are organizations that are all three of these at the same time. That is going to cause some issues. How do you actually draw the line? When is the organization an employer, when is it a provider, when is it a plan?

Finally, the issue of timing is one which comes up over and over again. Is two years a sufficient amount of time for organizations to actually implement the standard and make the necessary process and systems changes that will be required?

Also, one might question, is February the year 2000 the safest time to say that everyone has to comply, given the other issues around the year 2000 that are receiving a lot of popular press?

Finally, are the teeth big enough, if you will? In other words, a standard is really only useful if everyone uses the standard. If there are ways not to use the standard, then you really haven't accomplished much. Some have indicated, maybe it would be cheaper to pay the fines, at least in the short run, than it is to try and meet the time line.

Others question whether or not large programs such as Medicare or Medicaid, if they implement it, is that enough to tip the balance, so that others will follow suit and implement as well.

So these are just a few. I'm sure all of you are sitting out there thinking, what about this or what about that. This is the area that we are currently beginning to spend a lot of our time and effort.

As I mentioned as well, one of our major activities has been in outreach and education. I would like Mike Fitzmaurice to come up and talk a little bit about the kind of issues or questions that have come up, as well as some of the barriers to the successful implementation of this legislation that we seem to have perceived. Mike?

DR. FITZMAURICE: Thank you, Bob. I think the most important questions you're going to have are going to be of the specific implementation teams working with the standards.

I will say that it is important for us to understand the business process of these standards, and how they are going to be used. There has been a lot of discussion and a lot of outreach on that.

We have built a network for dissemination, feedback and rapid responses. But actually, we are part of your network for getting this information and disseminating it to your constituents. It is easy for us to forget in the midst of all this hard work that there is a good sense of accomplishment, but you always think of what is the bad news around the corner, rather than what is the good news of the day.

It kind of reminds me of the lawyer who come in to see his client. He says, I have just come from the district attorney's office, and I've got good news and bad news. What is the bad news, says the client. From the blood at the crime scene, they have definitely matched your DNA. They are 100 percent sure. Oh my gosh, what is the good news? Well, your cholesterol count is down to 140.

Frequently asked questions. Is Big Brother forcing this on the industry? The answer is, we sure hope not. It is not intended to be that way, and it didn't start out that way. The industry asked Congress to provide uniformity of health care transaction standards. The Secretary must develop standards developed by NCA accredited standards setting organizations, and the Secretary must consult with the industry with NCVHS, with standard developing organizations. So it has been very directed to the health. There is not a lot of leeway on how to do this.

Will only providers benefit from the HIPAA standards? No, providers, clearinghouses and payers will benefit. Providers will benefit from reduced claims hassle, dealing with formats of different insurance companies. They will benefit from mandated acceptability of electronic transactions by payers and clearinghouses, and by greater opportunity for direct electronic transactions with payers or their designated clearinghouse representatives.

Clearinghouses will benefit from a simplification of a reduced set of acceptable formats for electronic transactions, leading to less frequent format changes by payers. At some point, you can say, will this drive the clearinghouses out of business? No, because they are still needed to consolidate the claims, provide a batching service and provide efficient transportation of the claim from the provider to the payer.

Will the payers benefit? Yes, from reduced clerical work and expertise needed to check on the format of the submitted claim. This work should be much easier, and they will either have to do less of it themselves or pay a smaller amount to the clearinghouse. But that is going to be left for the market to work out. This is like paving the roads, and we will see what the size of the trucks are that drive down it and how fast they go and how efficient the delivery of the information is.

Is DHHS doing this alone? Nope, other federal agencies are participating. You heard Bob Moore earlier talk about participation through the DHHS Data Council on the implementation teams as well. The Department of Defense, Veteran Affairs, the Office of Personnel Management, Department of Commerce's National Institute for Standards and Technology are all participating as members of the implementation teams. NCHS advised the Data Council and provides input to the implementation teams. So there is broad input from the public through these sources.

We have had public hearings. Don Detmer talked about six days of hearings on privacy and confidentiality, six days on health care data standards, including clinical codes, and two days on both of them out in San Francisco. DHHS staff participate broadly in outreach to get people to know what is going on at the time it is going on.

Will DHHS merely adopt Medicare standards? The fear has been that the Medicare program is the largest payer, and therefore there is the incentive for the government to take the Medicare standards, therefore Medicare won't have to change, and everyone else will have t conform to the Medicare standards.

Not so. Standards from accredited SDOs, standard developing organizations, must be considered first, says the law. HIPAA strongly encourages the use of existing standards approved by SDOs, such as the X12N standards. No adoption decision has been made by the health yet. You are seeing the works in progress of our implementation teams. The Medicare program will be bound by HIPAA just as the other payers are bound by HIPAA.

Will all the standards be adopted in 18 months? Nobody wants to face that question, because it is hard to give a definite answer for all the standards taken together. The intention of course is for DHHS to meet the Congressional target. The industry acceptance and the lack of controversy over many of the standards will propel this adoption. But Congress appropriated no additional resources for this mandate, so we are all taking from existing resources and away from other projects. The public notice and proper administrative procedures will take time.

So where does that leave you? It leaves you with a sense from today's meeting that you should know when you walk out of the door which standards to expect to be published in October and adopted in February, which standards may be problematic. I'll have a little bit more to say on that in the next set of frequently-asked questions.

Will the private sector standards be adapted with no change? Changes may be made during the adoption process; there is no prohibition against that. But any changes should improve the uniformity, cost effectiveness and acceptability of the standards. Any changes will be publicly announced and public comments solicited.

Which standards will be adopted first? Well, some standards will be completed first and published in 1997. These are likely to be the national identifier codes for providers and payers, likely to be the code sets and the claims standards.

What is the other side of that? The unique personal identifier will most certainly take a little bit longer, because it raises questions about the privacy of information that may be linked to it.

What about testing of HIPAA standards? We all agree that standards should be tested in use. It can be done in either the public or the private sector. Testing would require resources however that currently aren't available in the public sector.

There will be two years from the announcement of the Secretary's adoption of the standards before they are mandated. So that is going to give two years of time to do testing, to do practice runs and to decide how best to implement them. I think two years is a long time, but others have said, could we have another year or two. Those decisions and your input on those decisions are most welcome.

Some of the barriers we see to adapting national standards is that, right now we have conflicting standards. We have ANTSE standards, we have industry de facto standards, and we have government standards. We have conflicting implementations, ways of using the standards and using the data within the standards. There is a proprietary collection of unique or differently defined data. And we have incomplete standards. Some of them have no implementation guides, but for the X12N standards, that is being worked on very rapidly.

There are proprietary code sets. Professional associations and others make dollars selling code sets. Will it interrupt their flow of normal commerce?

There is a cost of change, even of changing the cost of an idea. If we move to a unique health identifier or we modify the social security number, it is very likely to mean more digits. Yet many programs are written with just a specific number of digits in mind.

Bob raised the question of, is the year 2000 the best time to do all this stuff. It may be a very good time, because as you are going through looking for how do we change the date, you might also look through the lines of code for how do we make the size of the variables variable. How do we expend them, so that we can be prepared for all of these identifier changes.

Another barrier and perhaps the biggest one is privacy. It faces us when we look at the use of the social security number and when we look at security as well. The question that gets raised is, how can you do security standards if you don't know what privacy you're trying to safeguard. The answer is that you can look at the components of security, but you may not be able to know where to set the bar, high, low, low cost, high cost. A lot of barriers ever getting out, very few barriers against the data ever getting out.

The privacy legislation is expected to tell us what is wrongful disclosure, what information is covered perhaps beyond the transactions data, and what the penalties are if you violate the privacy laws. The security standards are to let you know, are the data being held securely, do you know who is getting at the data both internally and externally, do you know to whom it is being disclosed and/or redisclosed, and for what purpose.

I want to open it up now with Bob for questions from the audience on the infrastructure and cross-cutting team. if you have questions, I would ask you to go to the microphones.

PARTICIPANT: I have a question in general about the bill and implementation of it. Won't the federal government now by virtue of this bill have electronic medical dossiers on each citizen in this country?

DR. FITZMAURICE: The bill doesn't require that information to go to the federal government. So the answer is no.

PARTICIPANT: Well, it does state in the conference committee notes, and I read the bill quite thoroughly, apparently all health care providers will be required to use these standard code sets. And basically, what the bill does is lay the groundwork for a cooperative federal, state, local statistics system. That is in the bill itself.

The question that I have is, since this information will obviously be linked together, the government, by virtue of these different governmental agencies, will be regulating how this information is going to be used. From my understanding and from what I have read, health insurance companies in addition to your own insurance company, law enforcement potentially will have access to it, other physicians, so it will be very difficult to get an objective second opinion.

Let's suppose you have a very bad terminal situation, and you need to go and get an objective second opinion. If there is a national patient database, how will you be able to get that objective second opinion when doctor number two can go and see what doctor number one had to say? You are taking away the right of privacy from the patient, and the autonomy from the patient. I don't have anyone that has been able to answer that question for me, as a practical matter.

DR. FITZMAURICE: I don't agree with your assumptions, and I don't rush to the same conclusion that you do.

PARTICIPANT: I'm certainly not rushing to it. I have been looking into it for a number of months now.

DR. FITZMAURICE: But when you talk about a national federal database that has the intimate details of people's medical encounters, I find that nowhere in the bill.

PARTICIPANT: Well, it seems to me that when you talk about equivalent encounter information and claims attachments, you're talking about diagnosis, are you not?

DR. MOORE: I would like to point out that actually, the legislation merely discusses the standardization of information that is currently already been transmitted. Claims are being transmitted every day. Enrollment and eligibility information is transmitted every day. The legislation calls for no new information to be standardized. It also doesn't call for the information to be transmitted to any particular point. It simply calls for a standard transmission format and content between current business partners. So there is really not -- it is not building a national database.

PARTICIPANT: In the bill itself, it refers to a federal, local and state cooperative statistics system. If you want me to, I can point out the exact clause in the bill for you. No one has been able to answer for me what that is referring to, and what is the reason for having these data sets if not to link up and have health research, have access to it.

DR. MOORE: These aren't specific data sets. These are simply to say, if you send a claim, the claim should contain this information.

PARTICIPANT: But you are all going to be speaking the same computer language. So inevitably it will all be connected up.

DR. MAYES: Let me say that a lot of the concerns you raise are concerns that we also face, and we face them in privacy and security, on how to implement whatever the privacy law will tell us. Bob, would you like to make a comment that addresses your comment, or do you have a different comment?

PARTICIPANT: It is on the same issue.

PARTICIPANT: Basically, what I feel is that the right of privacy is actually being eroded, because the patient will now have law enforcement, health researchers, statisticians, health insurance companies, and according to the chairman of the subcommittee on privacy, most probably your employer will have access to your medical records without your consent.

In a recent House subcommittee on technology meeting, computer experts from all over the U.S. testified how easy it is to break into computers. What I find illogical is, it is illogical to require patient medical records on a national database. It would be as if every citizen in the country had their money in a central bank. Where do you think hackers will go for this medical information?

As it stands now, doctors and hospitals have paper files, they have internal computers. It seems to me, if you want to insure better privacy, the way to do it is not make a national database, but rather make more stringent requirements of these separate institutions. I just don't understand how a broader level of dissemination is going to -- by opening up a national database, is going to insure privacy. To me, that is completely illogical.

DR. FITZMAURICE: I think your comments are better suited for a privacy law than for HIPAA, because this does not create a national database on people.

PARTICIPANT: Then let me just get the cite from the Kennedy-Kassebaum bill, the health insurance bill itself, and perhaps you can tell me what it is referring to.

DR. MOORE: I would just like to address that last comment. As Mike said, this in no way establishes a national database for all people in the country's health records. Health plans, health payers are still going to be in charge of those records, as they are today. Medicare is in charge of Medicare beneficiary records. State agencies are in charge of Medicaid beneficiaries' records. Blue Cross, Travelers, other plans, are in charge of those records. In today's world, the last statistic I saw was something approaching 70 percent of the private citizens, outside of Medicare and Medicaid, are in managed care plans, and they sure as hell move information from one provider to another in order to treat that patient appropriately for whatever disease or illness they have.

DR. FITZMAURICE: Are there other questions?

PARTICIPANT: In referring to the bill itself, 104 PL 191, Section 263, and I am quoting, the committee shall assist and advise the Secretary to determine, approve and advise the terms, definitions, classifications guidelines for assessing health status and health services.

Now I'm going to go down, by all programs administered and funded by the Secretary, including the federal, state, local, cooperative health statistics system referred to in Subsection E and 3.

Now, I have asked numerous people, where is Subsection E and 3. Where is that section in the bill itself?

DR. FITZMAURICE: I'm afraid without taking some time to look through the bill and look through any supporting law, I can't answer that question.

But what you are referring to there is the charge given to the National Committee on Vital and Health Statistics, is that right?

PARTICIPANT: Yes.

DR. FITZMAURICE: And the kind of assistance they are to give to the Secretary? I don't see that it directs the Secretary to do anything. It tells what kind of advice the national committee should be giving to the Secretary.

PARTICIPANT: Well, I think that you are evading what I am raising to you, which is, it is not something that you can deny, that once the computers are speaking the same computer language, and you are saying that health research will have access to our medical records and we are all going to get unique health identifiers which are patient I.D. numbers, to me you are not being responsive. How will health research access our medical information? How will it have the access to that?

DR. FITZMAURICE: But if you look through the bill, that has to be discussed broadly in Congress, and they plan to come up with the law in the next three years. We are not directed to do that. Congress is saying that it wants to take on that for itself, and then only if they can't reach a conclusion at the end of three years and pass a law do they give any power to the Secretary.

PARTICIPANT: But you are directed to look into the privacy and confidentiality aspects, and that is what I am speaking toward.

The other thing is that in this bill itself, it also talks about the fact that past, present and future medical information will be required to be reported in these code sets.

DR. FITZMAURICE: The Secretary is directed to report to Congress on her recommendations on privacy. That report is due August 21st.

DR. MAYES: Let me just clarify, as I did in the beginning of the talk. Whether or not in fact the law states this, we are not the appropriate group to be answering that. In the implementation of this legislation, as I said before, it was structured two separate groups or entities. We were concerned with developing -- well, identifying and adopting technical standards for the ten transactions, for the four unique identifiers, for code sets used within the ten transactions, not any broader than that, with security for the transmission of that data.

There is a separate group under John Fanning, whose name was mentioned earlier, in the department who is dealing specifically with the privacy issues. They are right now finishing up the draft recommendations for the Secretary, and those will be available the 21st.

Now, obviously there is some linkage. It is difficult as Mike pointed out to discuss some of the details of security without understanding where whatever privacy legislation comes up with where it sets.

So I don't want you to get the impression that we are not concerned with these issues and that we are not trying to answer it. It is just, that really wasn't the people that are going to be talking today's specific charge. We are much more concerned with just the technical issues of adopting the standards.

So just to clarify, if it seems like we don't have the answers and know the sections, it is because we haven't had any mandate to do so under our group.

DR. FRAWLEY: The Public Law 104-191 not only designates responsibilities to the Department of Health and Human Services, but also the National Committee on Vital and Health Statistics. Section 263 is clearly the charge to the National Committee on Vital and Health Statistics, in terms of the cooperative health statistics system.

As Dr. Detmer pointed out this morning, that committee has been in existence for 46 years, and has been responsible for evaluating the public health infrastructure in this country, particularly a lot of the reporting that is taking place at local, county and state level, and also with federal, in terms of a lot of the surveys that go on, that many of you are familiar with.

Some of those are done with non-identifiable data, some with identifiable. I think the important point for everyone in the audience is that most of the transactions that we will be hearing about today are ongoing transactions. There are no privacy protections as they exist today. I think the law does at least start the ball rolling, in terms of saying to Congress, you wanted to enact privacy legislation, you gave yourselves the opportunity for three years, and if you don't do it, the Secretary has the final authority.

So I think it is very important, when you read the final legislation, to make sure you understand what the responsibilities of HHS are, what the responsibilities of the national committee are, and what the responsibilities of the private sector are.

DR. FITZMAURICE: Thank you, Kathleen, for clearing that up. I have time for one more question. Let's take the question, and then we'll bring up the claims and encounter implementation team.

DR. KORNETSKY: I just wonder if you would briefly comment on the type of feedback you received in some of the public hearing stop date on the time frames. I know for -- I speak here primarily for multi-employer plans. I know there is a great deal of concern about the aggressiveness of the time frame. A number of these plans are still heavily manually handled in their administration, and they deal particularly with very complex eligibility and enrollment issues. I know they are terribly concerned that the costs will be absolutely prohibitive and the time frame, impossible for them to come up to speed. It might really have a serious impact on how that industry works, and whether they will be forced into using national clearinghouses and so forth.

So I wonder if you would comment just generally on the kind of concerns that have been raised by other industry groups or payers about whether it is really feasible for them to implement these standards within two years of their adoption.

DR. FITZMAURICE: Let me just give you a general answer, and say that that question is really best directed toward the specific implementation team that has to deal with those particular standards.

My general response is that while we think that two years is a long time to do the implementation of the standards, we are aware that other standards are firmly embedded in different organizations. We have been listening to different comments about transitions about how long it would take. Even the National Committee on Vital and Health Statistics has addressed that issue.

There is some feeling that we should take a look at a longer time period. Whether we will go with that feeling, we certainly have to listen to the industry that says, we can do it, we can't do it. We have to listen to other parts of the industry that say, we can help them do it. It is not a matter of forcing anybody into an organization that can help them do it faster. It is a matter of having market opportunities.

That is a very open issue with us. Bob?

DR. MAYES: Let me just add one other -- two other points. One is that, interestingly enough, we have had mixed reaction. We have certain sectors of the industry that say, no, don't lengthen -- whatever you do, don't lengthen it, because people will just put it off. If you give them four years, they will put off for four years doing it.

The other issue, remember, it is a two-year implementation time period. If in fact for particular sectors or overall, it becomes clear that that is not going to be an adequate time to effectively and efficiency do this, there is that time period to make adjustments. It's not like, -- February we have to adopt them, but it is not like you have to be using them by March. So there will be opportunity over the next two years to make the case whether from a particular sector or overall, about the adequacy of the time available.

We will be here all day, so if you have any further questions to either Mike or myself, feel free to grab us.