w
[This Trranscript is Unedited]
Hubert H. Humphrey Building
Room 705A
200 Independence Avenue, S.W.
Washington, D.C. 20201
Agenda Item: Call to Order, Welcome and Introductions, Review of Agenda - Dr. Lumpkin
DR. LUMPKIN: Good morning, welcome to the meeting of the National Committee on Vital and Health Statistics, we’re going to start off with introductions, we’ll go around the room and then the informational portion of the agenda. I would like to remind everyone that we are going live on the internet as we do for all of our presentations for people who can’t make it to the meeting and because of that we do ask that you speak into the microphone when you have a comment and for those who may speak from outside of the committee we would ask that you would identify yourselves again when you come to the microphone. My name is John Lumpkin and I’m senior vice president of the Robert Wood Johnson Foundation, and we’ll start off with Marjorie.
MS. GREENBERG: I’m Marjorie Greenberg from the National Center for Health Statistics, and executive secretary to the committee.
MR. ROTHSTEIN: I’m Mark Rothstein from the University of Louisville, a member of the committee.
MS. HANDRICH: I’m Peggy Handrich from Wisconsin Medicaid and Wisconsin Health Data.
MR. HOUSTON: I’m John Houston, I’m with the University of Pittsburgh Medical Center, I’m a member of the committee.
MR. LENGERICH: Gene Lengerich, Penn State University, member of the committee.
MS. SANCHEZ: I’m Linda Sanchez, I’m a staffer in the Office for Civil Rights in the Department of Health and Human Services.
MS. FRIEDMAN: Marie Friedman, CMS.
DR. STEINDEL: Steve Steindel, CDC, liaison to the committee.
MR. BLAIR: Jeff Blair, Medical Records Institute, member of the committee.
MS. BEREK: Judy Berek, CMS, liaison to the committee.
DR. ZUBELDIA: Kepa Zubeldia, Claredi Corporation, member of the committee.
DR. FITZMAURICE: Michael Fitzmaurice, Agency for Healthcare Research and Quality, member of the committee, staff to the Subcommittee on Standards and Security.
DR. SHORTLIFFE: Ted Shortliffe from Columbia University, Anatology(?) Physicians and Surgeons, member of the Committee.
MR. HUNGATE: Bob Hungate, principal of Physician Patient Partnerships for Health and member of the committee.
DR. MAYS: Vickie Mays, University of California, Los Angeles, member of the committee.
DR. COHN: Simon Cohn, Kaiser Permanente, member of the committee.
MR. SCANLON: Jim Scanlon, HHS, Office of the Assistance Secretary for Planning and Evaluation, executive staff director for the committee.
MR. BRUNELLI(?): Sam Brunelli, Team Builders International, Alternative Link.
MS. GIANNINI: Malina Giannini, Alternative Link.
MS. JONES: Katherine Jones, NCHS, staff to the executive committee.
MS. FISHER: Beth Fisher, National Center for Health Statistics.
MS. FYFFE: Kathleen Fyffe, ASPE, lead staff to the NCVHS Subcommittee on Privacy.
MR. SMITH: Merit Smith of the Robert E. Nolan Company.
MR. HAFFANA(?): Bill Haffana, Blue Cross/Blue Shield.
MS. FRASIER: Judy Frasier, American Optometric Association.
MS. TORAKOWSKI(?): Irene Torakowski, American College of Surgeons.
MR. MUSKO: John Musko with the Health Insurance Association of America.
MR. DESMARAIS: Henry Desmarais with the Health Insurance Association of America.
MR. KASTLE(?): Gene Kastle, American Psychiatric Association.
MR. REGERA(?): Daryl Regera, American Psychiatric Association.
MS. CANAAN: Susan Canaan, writer for the committee.
MS. JACKSON: Debbie Jackson, National Center for Health Statistics, staff to the committee.
MR. SPANGLER(?): Tom Spangler, American Dental Association.
MR. BURKE: John Burke, Office of the Secretary, staff to the Subcommittee for Standards and Security and NHII Workgroup.
MS. BEBEE: Suzie Bebee, Office of the Assistant Secretary for Planning and Evaluation, staff to the Subcommittee on Standards and Security.
MS. AULD: Vivian Auld, National Library of Medicine, staff to the Subcommittee on Standards and Security.
MS. MARSTELLER(?): Jill Marsteller, National Center for Health Statistics and an Academy Health Fellow.
MS. JOHN: Jody John, Quadrimed(?).
MS. WILLIAMSON: Michelle Williamson, National Center for Health Statistics, CDC.
DR. EDINGER: Stan Edinger, AHRQ.
PARTICIPANT: Tom (?), director of the Center for Medicare Management.
MR. RODY(?): Dan Rody, American Health Information Management Association.
MS. QUAS(?): Linda Quas, American Health Information Management Association.
MS. LEON-CHISEN: Nelly Leon-Chisen, American Hospital Association.
MS. PROPHET-BOWMAN: Sue Prophet-Bowman, American Health Information Management Association.
MR. GUSTAFSON: Tom Gustafson, Center for Medicare Management.
MR. STAEHLIN: Marty Staehlin, Price Waterhouse Coopers.
MR. BEBEE: Michael Bebee, American Medical Association.
MS. LUWELL(?): Mary Luwell, MedPac(?).
DR. LUMPKIN: Okay, the next item on our agenda are conflicts of interests and recusals, anyway, I can’t get that word out. We got a lot of lawyers in this room, how come none of you guys are helping me? Any conflicts of interests, recusals? Simon?
DR. COHN: Well in the event that we are discussing CPT everyone needs to know that I am on the CPT editorial panel and obviously would be recusing myself from discussion of CPT.
DR. HARDING: I am a past officer of the American Psychiatric Association, which has involvement with DSM and other things and would be recusing myself on a vote on that issue.
DR. LUMPKIN: Okay, actually we have two people on the phone, which we didn’t realize, I just heard Richard.
DR. HARDING: Yes, Richard Harding, member of the Committee, chair of the Department of Neuropsychiatry, University of South Carolina.
MS. GREENBERG: And new grandfather.
DR. LUMPKIN: And new grandfather, congratulations. Anyone else on the phone? And we have --
MS. GREENBERG: Did we have someone else on the phone?
MR. BLAIR: I have a disclosure to make.
DR. LUMPKIN: Yes, Jeff.
MR. BLAIR: I want to get in and not miss this opportunity, my son Brett and my daughter-in-law Vitka(?) are parents of my first grandchild.
DR. LUMPKIN: Congratulations. And Stan if you’ll introduce yourself?
DR. HUFF: Stan Huff with Intermountain Health Care and University of Utah in Salt Lake City.
DR. LUMPKIN: Member of the committee. Jennifer?
MS. MADDENS: Jennifer Madans, National Center for Health Statistics.
DR. LUMPKIN: Great, welcome. Okay, Jim, update from the Department.
Agenda Item: Update from the Department - Mr. Scanlon
MR. SCANLON: Thank you very much, John, welcome everyone. I’m going to try to make up some time on the agenda this morning but I wanted to report back to the committee since we met in late June I guess that progress continues on a number of developments relating to the NHII and health information technology and standards. And I think it’s a reflection of these kinds of issues becoming more systematically included in HHS planning and program operation, and even the Secretary himself on several occasions has emphasized the importance of the NHII and the importance of interoperable health information technology and standards both in our operations here in HHS as well as the health system generally.
Let me report on a couple things. The Secretary after our meeting on June 24th actually announced this at our NHII Conference in early July, as you know he’s emphasized repeatedly on the importance of applications of IT and their potential to make really major changes to the health system. And as you’ll recall in the draft strategic plan that you all had a chance to review and comment on for HHS, which we’re in the stages of finalizing by the way, we include several objectives related to the National Health Information Infrastructure and its importance in supporting the missions of the agencies in HHS and the health system generally. He is urging all of our agencies to try to look at opportunities to employ relevant information technology across all of our programs and to work to promote interoperable health information technology across the public sector and the private sector as well.
To assist in this goal he asks that we establish in HHS a Council on the Application of Health Information Technology, CAHIT. This is an internal group, it’s co-chaired by my office, ASPE, and the administrator of AHRQ. The goal is to coordinate HHS activities and investments in health information technology, so this is within and across HHS, but also to ensure a coordinated approach and more or less one voice in the HHS collaborative work with the industry and the public health and the research community as well. The Council has met about three times already, it’s on a pretty much regular basis, every other week or so or every third week, and Dr. Lumpkin participates in the meetings in his role as chair of the NCVHS. So that’s a very positive development.
Secondly, you may have already seen this in Science Policy newsletter, and it relates to some new proposed guidance from the Office of Management and Budget relating to peer review requirements for significant regulatory information that is disseminated by federal agencies. In the past I’ve briefed the committee on the Shelby I and Shelby II Federal Information Quality Guidelines, and the attempt there was really to make some of the research data more available to the constituents in the industries that use them, and to ensure that federal agencies have systematic procedures and policies in place to assure a quality review before they disseminate any substantive information, scientific or technical, epidemiological or so on.
And about three weeks ago OMB has added to that set of guidelines in a proposed bulletin on peer review and information quality and the intent here as stated on the bulletin, I’ll give you the website later because you can actually look at the bulletin and you can comment, provide comments to OMB, but the goal is to ensure that agencies conduct peer reviews of the most important scientific technical information, and technical information, relevant to regulatory policies that they disseminate to the public, and that the peer reviews are reliable, independent, and transparent. So the focus in general is on scientific regulatory information that an agency disseminates in support of regulatory actions. Now OMB definitions actually provide a bit of ambiguity about what the scope would actually be but the basic intent is to ensure that all agencies when they’re using scientific and substantive information and studies largely, that support a direction being proposed in a major regulation, that they have provide, that they have conducted or provided for a peer review. The peer review can be an internal kind of a peer review from other substantive experts within the agency but outside the program, or it can extend to the full external peer review much as you see in the scientific journal articles and the review articles as well.
The bullet defines regulatory information as any scientific or technical study that is relevant to regulatory policy and information may be relevant to regulatory policy if it might be used, this is the OMB definition again, by local, state, regional, federal, and international regulatory bodies. So it actually when you look at it more closely, it could extend by this definition of relevance to a larger scope then just purely regulatory actions, but I think we’re looking at it in terms of regulatory actions.
OMB is accepting public comments on the proposal through, well, they’ve just extended, it was October, it’s now December 15th, and the final bulletin would be taking effect early part of next fiscal year when it’s finalized. And they would, the bulletin would apply to all information that any federal agency disseminates that’s within the scope, well, originally it was planned for next January but it will probably be a bit later. So we in HHS are looking at this collectively, OMB has asked us to provide some comments and we’re in the process of pulling them together now. In the case of what OMB calls especially significant regulatory information OMB then, the bullet would then require an external peer review.
Let me go on just quickly to our conference follow-up activities. You’ll remember that in early July we convened a national meeting to try to develop a consensus agenda on action steps needed for the NHII, we had a very successful attendance and very successful participation. The secretary gave a keynote address, Dr. Lumpkin gave a keynote address as well. There were eight areas, breakout areas that you’ll recall, I won’t repeat them here, but there were more or less eight tracks ranging from research to public health, privacy and confidentiality, data architecture and so on, and we received a number of recommendations in all of those areas. We’re in the process now of pulling the recommendations together, there’s a plan to pull them together for publication and we also are beginning to plan some follow-up activities with some of the tracks and some of the groups of participants. I think one of the first groups is providers and another group is payers, and I think we’re planning to have specific follow-up meetings with those groups and we’ll do it for the others as well in the months ahead.
We’re also pulling together an inventory on our website, we now have an NHII website as you know as a follow-up to the meeting, and we’re pulling together an inventory of all of the HHS activities relating to the NHII and again, that’s on our HHS website.
Just quickly let me bring you up on where we are with the budget and remember in any given year we’re dealing with three budget years. For the current year, which is really winding down, the current fiscal year, fiscal year ’03 will wind down at the end of this month, September 30th, the new fiscal year, fiscal year ’04, begins on October 1st. We and a very few of the, hardly any of the federal agencies have an appropriations yet for ’04 but it looks like the direction is heading in the manner we hoped relating to a number of investments that were included in the President’s budget for ’04 relating to NHII and data policy, and let me just remind you of what they are again, hopefully in the weeks ahead. Things look fairly optimistic actually that these will actually turn up in the budget, in the actual appropriation. And again these sort of reflect an increasing role for HHS in leadership, convening, and standards, and research and demonstrations for health information technology and some expansion in the public health area as well. At the overall departmental office in ASPE, my office, the budget continues to include a NHII initiative, about $3 million dollars, this would be for overall strategy planning and coordination, convening, and for some special projects that are more or less gap filling that wouldn’t be done otherwise. Some of the work has already begun this year and it would be extended in ’04 as well.
Another investment that’s proposed in the budget is a $10 million dollar investment, this is a departmental investment, it’s a partnership between ASPE and AHRQ, it’s in the AHRQ budget, it’s about $10 million dollars for standards, to accelerate the development and use of interoperable standards, so we’re looking forward to that as well. And in addition there is a $50 million dollar initiative in the AHRQ budget that would be available for research, particularly demonstrations, to improve patient safety through health information technology. I guess the focus would be largely on hospitals but it could extend to other health care organizations. And the Secretary there may give preference to small overall hospitals, but again, this was initiated about a year ago in the department’s planning and it hopefully will be coming to fruition. I think in the optimism and hope of having the money available fairly early we’ve already started to think about what the requests for applications would look like so we can get a fairly good start on that when Congress finishes action and again in the direction we hope.
Again, we’re working on the fiscal year 2005 budget, we’ve just begun work, we’re submitting our preliminary budget request to OMB. Our hope there as well is to continue some of these investments into the next fiscal year and to actually try to on the population statistics side to try to address some issues, a stable source of funding for some of the programs as well. Now again this is very preliminary, the budget has a long way to go for ’05, we don’t even have the ’04 budget yet, but this is at least the thinking and the direction of where we’d like to go.
And finally let me report on the National Academy of Sciences study on the adequacy of race and ethnicity date in HHS and other Health and Human Services systems. This began more than a year ago, Congress passed a law that directed HHS to fund the study at the National Academy of Sciences. The panel was established, it has met I understand three times, Dr. Lumpkin is on that expert panel and Ed Perren(?), Dr. Ed Perren I think many of you know, is the chairman. We asked them to include as part of their deliberations a workshop on particularly involving health plans and providers and to get a private sector focus on what the current policies and practices were in terms of collecting race and ethnicity data in regular and routine medical and administrative data. It was clear that there was a lot of misunderstandings prior to that and we were hoping to get some clarity about what the practices are and so on.
Well, the workshop was actually held, many of you were actually participants in the workshop, and I talked to the Academy folks yesterday and they expect to, the workshop report is expected to be released in a week or so, a week to ten days, so we’ll get that to the committee right away, and I think it will dovetail with some of the Population Subcommittee’s deliberations on improving race and ethnicity data in these very same systems, so I think they’re both heading in the same direction. The full report from what I understand is now in Academy review and I think they were hoping to have it around December, but again this is somewhat unpredictable, but later this year we probably would have the full, we look forward to the full report. I will keep you informed on that as well.
Two quick things from the Data Council, these are kind of follow-up to the vision for health statistics report that the committee completed. The gateway, our electronic web based gateway to Health and Human Services statistics that we talked about at a previous meeting, we’re now expanding and upgrading to include a number of additional sites, we now have about 2500 websites that provide reliable statistics in Health and Human Services. Again there’s some connection with HHS to be included and we have just, we’re just about to start a study to look at geo-coding issues, geo-coding data and standards, this is the HHS part of a larger government wide initiative to improve and standardize the geo-coding and geo-spatial information that we include in a lot of our data. In the health area I think the geo-spatial aspects are somewhat different then EPA and some of the other agencies have where they’re largely dealing with fixed assets that literally in a latitude and a longitude tells you where it is and that’s where it stays, dams and bridges and buildings and so on.
In the health area we have some of those but it’s a little more fluid, it’s more population based and government based and organization based but there are still some good practices and policies that we want to look at. As a start we’ll be looking at our meta-directory of the major HHS data systems, we’ll be looking at what geo-coding information is available now, what’s collected and what’s available, obviously there are confidentiality issues in what you can make available for small areas, and we’ll be looking at what kind of standards and best practices might best serve us all. Again, this is part of a broader government wide effort to kind of standardize and improve the geo-coding, geo-spatial data as well.
Let me stop there.
DR. LUMPKIN: Alright, any questions?
DR. MAYS: Jim, could you talk a little bit more about the geo-coding? When you said that you’re going to do a study, who’s the we that’s doing it?
MR. SCANLON: This is under the Data Council, we’re in the process of awarding a contract through one of our ask order contractors, and they’ll be working with, I can’t say who it is yet because we’re still negotiating, but it will be a contractor supported study, they’ll be working with a working group under the Data Council to kind of give them direction and to keep the project moving forward. But we’ll start with the meta-directory, which is pretty much our holdings, the HHS holdings and data collection systems and we’ll report back to the Populations Subcommittee, I think it’s particularly relevant there as well.
DR. LUMPKIN: Okay, thank you. We’ll go on with a report of HIPAA implementation, starting off with standards. Maria, welcome.
Agenda Item: HIPAA - Data Standards, Including Clinical Data Standards Adoption - Ms. Friedman
MS. FRIEDMAN: Thank you. A couple of significant activities have happened since the last time I reported to the full committee. In the end of July CMS issued guidance enforcement for the transactions and codeset provisions of HIPAA and in the, and a copy of this is in your book by the way, in the guidance we clarified that covered entities that made a good faith effort to comply with HIPAA transactions and codeset standards may implement contingencies to maintain operations and cash flow. And then as a follow-up to that not long ago we announced that Medicare has a contingency plan and a lot of people have asked us both about the guidance and contingency planning and the Medicare contingency plan, under that we would continue to accept and send transactions in legacy format in addition to the HIPAA compliance transactions obviously while trading partners work through issues on implementing the HIPAA standards. The contingency plan will apply to all Medicare’s fee for service contractors and in the next day or so we may be announcing whether we are going to invoke the contingency or not. It’s still up in the air, if that happens while we’re convened I will certainly let everybody know.
Aside from that we are working on contingency planning with our trading partners and continuing our outreach activities.
DR. LUMPKIN: Did you say something was happening next month?
MS. FRIEDMAN: No.
DR. LUMPKIN: Okay, any questions?
DR. HARDING: This is Richard Harding. I wasn’t on right at the start, they let us in when people were introducing themselves. Karen’s not there, it’s Maria instead?
MS. FRIEDMAN: Yes, Karen is up on the Hill at a hearing and I’m sure she’d much rather be here.
DR. HARDING: I imagine it’s tough to make those decisions. I had a question for Karen actually but maybe Maria could help, could I ask that now Jim?
DR. LUMPKIN: Sure.
DR. HARDING: Okay. I was kind of wondering about the issue of codes and what the status of one. Most of you know that about ten percent of electronic code transactions under HIPAA are in the area of mental disorders and I was wondering if Maria or Karen or whoever is there could update the committee now or later on how the department plans to kind of deal with the issue of mental disorder diagnosis codes under HIPAA and the electronic transaction regs. I guess more specifically will mental health professionals be allowed to continue to use the DSM evidence based criteria or has been suggested be required to kind of go back to ICD-9-CM descriptors to define mental disorders? If this isn’t the right time, I didn’t see that in the next number of things, but just wondered if you could comment on that issue.
MS. FRIEDMAN: I can’t comment right now, I know that it’s an issue that’s been under discussion with the committee and I believe we’re waiting for a report back from our Office of General Counsel on a couple of points, and certain when we have that information and are ready to make a report back we certainly will do that.
DR. HARDING: I guess the issue is October, and the Council has had that for some time now?
MS. FRIEDMAN: I think they have, I think they have been a long time is my understanding, so I think we’re due back for a report fairly soon.
DR. HARDING: It’s perplexing to many in the mental health professionals because the thought of going back to ICD-9, which of course was done in 1977, is a little bit mind boggling in today’s current health climate and the hope would be that the descriptors in DMS would continue to be allowed to be utilized. But you’re waiting on the Council to come back, is there going to be any FAQ’s or anything like that on that issue?
MS. FRIEDMAN: I believe that there will be FAQ’s on that issue and as I said I think we’ll be reporting on that fairly soon, just not today.
MS. GREENBERG: I was just going to say first of all Richard, that was a great segue into the next discussion on how old ICD-9-CM, so thank you. But I think, we recognize your frustration, it’s my understanding that there have been very recent meetings between the department and the American Psychiatric Association and that this is, so there has been very recent communication although as Maria said the actual official response has not yet come out, but I think there is an understanding that we’ll allow the profession to use the standard codes, which were mandated under HIPAA, which is the ICD-9-CM, but still not be required to stop using the diagnostic criteria that are currently applied. So I think there’s going to be a mutually agreeable solution, unfortunately we don’t have it all tied up with a bow by this meeting, but I know there have been a very recent meeting and I think a resolution is very close.
DR. HARDING: Okay, I appreciate that Margie.
DR. LUMPKIN: Thank you, any other questions? Kepa, is there any reason why you’re sitting in Clem’s seat? Okay, thank you.
DR. HARDING: I was going to ask the same question.
DR. ZUBELDIA: I have a question for Maria. There is another law that doesn’t effect October 16 and that’s the ASCA law by which all the Medicare providers will be required to file their claims electronically as of October 16. Has CMS looked at a contingency plan in case some providers cannot file their claims electronically as of October 16?
MS. FRIEDMAN: To be very honest we have issued a Medicare billing regulation and there is information on our website on that, I unfortunately am not up with all of the in’s and out’s that maybe our friends from another part of the house can handle that.
DR. LUMPKIN: We have a huddle on the 40 yard line.
DR. COHN: Maria, you’re saying you weren’t, I see the copy of the reg, you’re wondering whether there’s going to be a contingency plan in this one?
MS. FRIEDMAN: No, I think, my impression is and I don’t want to say anything correctly but my impression is that if we invoke the Medicare contingency plan, that’s the big question here, we don’t, as of right now we are not sure whether we’re going to invoke that contingency.
DR. ZUBELDIA: But that’s for HIPAA.
MS. FRIEDMAN: For HIPAA.
DR. ZUBELDIA: I’m asking about ASCA.
MS. BEREK: I just asked Tom Gustafson and he said we have the ability to waive for providers who can’t bill electronically, so that there isn’t a contingency plan but there is an ability to waive.
DR. LUMPKIN: I guess the follow-up question, assuming that you can waive, therefore your intermediaries would have to, because if there’s more then a few trying to do through Baltimore might be a little bit complicated. The issue of the contingency plan might be an important issue, how would someone, some guy who hasn’t been paying attention, be able to submit bills.
MR. BLAIR: Can I ask for clarification on this because this is on the internet, it does go out to the public, and if the word that we used in this particular case is inaccurate that could have some real consequences and Judy, can you verify that the word is waived? Because it was my understanding that the position of HHS was that it would be flexible in the enforcement but waiving is very different from being flexible in enforcement.
DR. LUMPKIN: Jeff, let me clarify, there are two different issues. One issue has to do with implementation of the transaction standards under HIPAA and the deadline that’s coming. The second is under ASCA, which mandates that anyone who, that CMS only pay electronic bills. Those are two different requirements and so there is flexible enforcement for the HIPAA and the question is the mechanism by which flexibility would be granted under ASCA would be the waiver.
MS. BEREK: Right, right, and what Tom Grissam(?) just asked me to tell you, which is that in fact at 9:00 a.m. Tom Skully announced for HIPAA that we are in fact implementing our contingency plan, so since it is quarter to ten I am allowed, this is live, I am allowed to say it. At 9:00 this morning Mr. Skully announced to the press that we would in fact be implementing our contingency plan under HIPAA.
DR. LUMPKIN: You guys are such great troopers.
MS. BEREK: Maria is the person who gets the Academy Award.
MS. FRIEDMAN: I want to thank all the little people.
DR. LUMPKIN: Cause I’ve been there and I know how difficult it is, we appreciate it. Kepa?
DR. ZUBELDIA: So now that we know that, as a follow-up for those providers that need to comply with ASCA and file their claims electronically and they cannot do it in the HIPAA formats, would they be allowed to implement de novo the old legacy standard formats as a contingency under ASCA?
MS. BEREK: This is more detailed then any of us in the room know, we will get you an answer.
DR. LUMPKIN: Thank you.
DR. FITZMAURICE: Is there a copy of the contingency plan available?
MR. SCANLON: Well, there’s probably a press release, we’ll get it at the break.
MS. FRIEDMAN: We had issued a statement outlining it earlier but it’s very brief and I mean it had the details you might be looking for but I would assume that we’ll follow our usual protocol and make something available on the website.
DR. LUMPKIN: And Marjorie reminded me that that only applies to people who aren’t in this area who don’t have power, they can’t get access to the website anyway unless they have a wireless device with batteries. Simon, did you have, that’s kind of an East Coast thing, you guys out West can’t appreciate hurricanes. Earthquakes maybe.
DR. COHN: Actually I was remarking last night I hadn’t seen rain since about May, much less hurricane weather but thank you.
DR. LUMPKIN: Okay, any other questions? Maria, thank you. Then we have from OCR, Linda Sanchez.
Agenda Item: HIPAA - Privacy Rule Compliance Update - Ms. Sanchez
MS. SANCHEZ: Good morning. I just have a few things to say about enforcements, we’ve been enforcing now for five months, since the April 14th compliance date with the Privacy Rule. As of yesterday we had 1,937 complains submitted to us. We received 95 in the last full week we had data, which is the week ending September 13th. Last week as we all know we all got to go home early and batten down the hatches. As you all know also the OCR regional offices investigate complaints and they’ve been in the lead and actually handling everything that’s been coming in. Overall we’ve resolved and closed roughly one third of the complaints we’ve received so far for a variety of reasons. Some complaints submitted to us as a Privacy Rule violation did not actually raise a privacy issue, sometimes the action complained about didn’t actually violate the rule, or we’ve been able to resolve matters through voluntary compliance and technical assistance. We’ve actually been very gratified by the response we’ve had from covered entities when we contacted them, when we’ve received a complaint about a particular action they’ve all, in general they’ve been very responsive to our calls and letters and have taken appropriate action very swiftly.
I know the committee is interested in what kinds of issues that have been coming up in the complaints. Inappropriate uses of disclosures have come up, we did a rough count two weeks ago, these have come up in about 350 of those complaints. Possible problems with safeguards have come up in about 280 complaints. These include inappropriate waiting room encounters where things have been overheard or too much information has been left out where people could see, or other types of facility problems. Inability to exercise an individuals access right has come up in about 170 complaints. Notice, failure to provide notice or inadequate notice has been identified in about 50. Minimum necessary, often related to whether an incidental use of disclosure was actually committed or not has come up in about 50. We’ve also been seeing a number of cases involving inappropriate authorizations.
I’m going to talk a little bit about outreach, our headquarters and regional staff have made about 100 presentations this year to several thousand attendees at various conferences. We’ve also participated in many telephone audio conferences. Our Privacy Rule hotline, which is a joint OCR and CMS endeavor, continues to operate, calls about the Privacy Rule are generally taken by trained operators and issues that come up that they can’t handle are then referred to a Privacy Rule specialist in our office. Since April 1st we’ve fielded about 14,000 calls. The first week we had over 1300 calls, as of the week ending September 13 we had only 480 that week, so clearly volume has diminished and we believe this is because covered entities and consumers are becoming more aware of the rule, are less surprised and are more able to find answers to their questions on our website. We’ve also found that many of the questions that we’ve been getting can actually be resolved by referring people to information in our website.
Speaking of the website, since January our frequently asked questions have been accessed 1.2 million times, I think one of the most popular sites if not the most popular page on the HHS website altogether, so we’re very pleased about that personally. We’ve recently added several new questions regarding how to account for public health access to records, accounting for dates of access and creating notices that might apply in multiple states. And additional FAQ’s in a variety of topics are in development or in clearance right now.
As you know the website contains many other guidance documents, we’ve recently linked to two new NIH guides for researchers and the CDC guide on public health related matters, and additional targeted technical assistance materials for various parts of the industry are also under development through a contractor.
So I’m happy to take your questions.
DR. LUMPKIN: Any questions?
MR. HOUSTON: John Houston. Has there been any attempt to try to break down by type of covered entity as well as type of provider where the complaints are coming in or are they based around hospitals or physician practices or insurers? Have you done any statistics in that regard?
MS. SANCHEZ: No, we have not. Early on in the first couple of weeks we started looking at that, we can look at that again if you’d like.
MR. HOUSTON: I think that would be interesting to understand exactly who’s having the problems with complying or where the complaints received are coming from. And also I would like just to make a comment regarding FAQ’s, I think they’re very good, I know I reference them often and I only hope that you continue to add to them aggressively because I think that seems to be a good source of information. So I think one of the common misconceptions that is still out there, and I’m not sure how you’d remedy this, is that there’s a lot of people at least, I’m from Pennsylvania, who say well HIPAA says I can do this, unfortunately they naively simply discarded Pennsylvania law which in many regards is more stringent. And I guess there’s no way in a FAQ to put that necessarily in there that you still need to consult your state laws, but that has been a source of problems I know in what I deal with typically and just people seem to forget about the state laws.
DR. LUMPKIN: Let me see if I understand the report, nobody’s been thrown in jail?
MS. SANCHEZ: No.
DR. LUMPKIN: The system didn’t fall apart?
MS. SANCHEZ: Not that I’m aware of.
DR. LUMPKIN: Good, good, so I think that the volume actually of complaints is surprisingly low given as much attention was placed on that, and I know it’s still a challenge for your and OCR but we appreciate your hard work in there.
DR. FITZMAURICE: I want to follow-up on one of John’s questions. It’s been remarkable, complaints with it and resolution of conflicts in your report, has OCR seen fit to fine the first $100 dollars of fines to anybody who violated the Privacy Rule?
MS. SANCHEZ: We have not reached that point with anyone, no.
DR. FITZMAURICE: Remarkable.
MR. HOUSTON: Wasn’t there a number of cases that were referred from the Department of Justice?
MS. SANCHEZ: We have referred some cases to Department of Justice, yes.
DR. LUMPKIN: Well, thank you very much. We’ll move on to the next item on our agenda, you’re free to go. We’re now going to start a process of something that has culminated a fair bit of activity by this committee looking at the issue of the transition, whether there should be a transition and how from the ICD-9 to the ICD-10-CM and ICD-10-PCS. I’m going to start with an overview, we’re going to have a presentation of the impact study which was sent out with the meeting materials, which I found to be a very interesting study, and then we’ll have a reactor panel and some discussions which will take us pretty much through lunch.
Agenda Item: Overview of ICD-10-CM and ICD-10-PCS - Dr. Cohn
DR. COHN: Can I make an introductory comment?
DR. LUMPKIN: Please do, Simon.
DR. HARDING: Could everybody do their best to speak right into the microphone? I’d appreciate it.
DR. COHN: Can you hear me?
DR. HARDING: Yes, now.
DR. COHN: This is Simon Cohn. I think as you all know I’m chair of the Subcommittee on Standards and Security. I just wanted to give everybody sort of a brief, maybe much higher level overview then where Jennifer is going to talk about just because I wanted to reflect for a minute on the process and I think as you all know the Subcommittee and indeed the full committee has been involved with these issues for a number of years. We first discussed them in the late ‘90’s during the recommendations on the administrative and financial transactions, and I think at that point if any of you review the final rules for that which of course are now coming into play finally on October 16th, next month, it was felt to be too early to move to ICD-10. More recently the Subcommittee started reviewing these issues and holding hearings back in 2002, and after a number of days of testimony we realized that we really didn’t feel we had enough information upon which to reach a conclusion, which was part of the reason that we asked NCHS to help sponsor a Rand study on the cost on the cost of the impact, so I obviously want to thank both Marjorie Greenberg and Donna Pickett for helping to shepherd that effort through. We’ll obviously be hearing from that study later on today as well as reactors.
What I want the committee to realize is that in some ways we’re really, the focus of this work is really not to come to the perfect cost/impact study, though we will be hearing about the cost/impact, we will be hearing from reactors to it and hopefully we’ll be getting additional information from the industry, both now and in October when the Subcommittee holds hearings on this topic, to come back in November with hopefully the next steps for the full committee. Really the committee needs to remember that the purpose of these discussions is to try to come to a consensus and general agreement about what we ought to be recommending regarding both ICD-10-CM, for diagnoses, as well as uses for ICD-10-PCS. So I want everybody to sort of think about all these things as we listen to both the overview and the reactor panels, and the information on the cost/impact, the cost impact I think Martin Libicki will be first to comment on, is an advance draft at this point and can be further improved. But really we all need to be sort of familiarizing ourselves with what it is we’re talking about in costs and possible benefits and hopefully as I said as we move through this process we can come to a greater consensus about what we ought to be recommending.
Anyway that was just my initial comments. Jennifer did you want to go?
Agenda Item: Overview of ICD-10-CM and ICD-10-PCS - Ms. Madans
MS. MADANS: Thank you. For those of you who don’t me I’m Jennifer Madans, I’m the associate director for science at NCHS, and I’m representing Ed Sondik, the director of NCHS, and first let me convey to you his apologies for not being here today. He certainly was planning on it up until about 1:00 yesterday afternoon when a meeting that was carefully scheduled to be not when this committee was meeting was moved to exactly when it was meeting and unfortunately he had to be at that other meeting. It’s actually in this building so you might see him wandering around, but he did very much want to be here and I think is very much interested in the issue of ICD-CM and classification in general, both personally and I think also from the point of view of the Center. So perhaps the most important message I can convey to you today is that NCHS is extremely interested in classifications in general and in ICD and related classification in particular. It is a major part of what we do and how we view our responsibility to the statistical, health statistical field.
The mission of NCHS is to provide statistical information that guides actions and policies that will improve the health of the American people and in fulfilling that mission NCHS and other partners in the department, as well as throughout the statistical system, is kind of dedicated to improving and maintaining a very high level of data quality and there’s, I don’t know if it’s been spoken in this meeting but an OMB initiative on data quality and we’re very much involved in that, and we see classification systems is a particularly important part of that data quality initiative and really at the very basis of it. And this is I think particularly true in the area of health, and specifically in health care, where there are lots of changes going on. The field seems to be completely in flux and really do need a high quality classification system to keep track of it. The international classifications of disease, which is really a family of classifications, is a key component of NCHS’s statistical work.
I know that you’re aware that the ICD is developed and maintained collaboratively between the World Health Organization and the international collaborating centers. NCHS houses one of those collaborating centers and is responsible for implementing the family of international classifications in the United States, that North American collaborating center includes the U.S. and Canada. I think that it’s important that these are international systems. By joining with folks around the world we not only get the benefit of their knowledge and expertise in this area but we have the ability not only to look at the status of the situation in the U.S. but also across countries, that’s very valuable from a research perspective. I think also I have learned somewhat the hard way sometimes that you don’t really understand what’s not right in what you’re doing until you do do some international comparison, it seems to highlight the very basic flaws in the system very, very quickly. I think we gain a lot by having an international classification rather then just a U.S. based classification, and our collaboration with WHO in this area I think is extremely important.
The other thing that is easy to forget but very important to remember is that classification systems are not static. You don’t develop them and then leave them, they have to reflect the environment that they are devised to reflect. And as that environment changes they too have to change, as knowledge advances they have to reflect that knowledge and knowledge is advancing a lot faster then it used to it seems, or maybe I’m just getting a lot older, but it appears that we are updating and revising much quicker then we used to. And that’s a good thing, but it does make the classification business more of an ongoing kind of business then it had been in the past.
As you know we in the U.S. adopted ICD-10 for mortality coding in 1999 and the U.S. was, and NCHS I’m happy to say, was a leader in this implementation both in the U.S. of course and also worldwide. In the U.S. we have in the past developed a clinical modification of the ICD for use in morbidity coding, and this tack that we have taken is recognized in our agreements with WHO through that collaborating center. Since 1979 we have used ICD-9-CM for morbidity coding, and that classification was adapted from the ICD-9 through an extensive modification. What is important is that the two classifications, the ICD-9, the ICD-9-CM, can be collapsed and compared across the two, that you can look at morbidity and mortality statistics in some kind of coordinated way, and that classification has been of course adopted and is still in use.
The ICD-10 revision was a pretty major one, and also was sorely needed and does reflect the changes in the field and in the knowledge base, and those very same issues that required a change in ICD itself also required a change in the clinical modification and any coding system you’re going to use for morbidity. And I would venture to say that probably more so, it’s even more true in the case of morbidity coding then it is in mortality coding, that the changes in the field really require the revision of that coding system.
I think we’ve been pretty successful in trying to update 9-CM over time to reflect changes in the environment, but I know I have been convinced by my colleagues in NCHS and other places that 9-CM is rapidly losing its capacity to incorporate the additional specificity for newly identified disease entities and other advances in medicine. That it’s just becoming much, much harder to modify the existing system. It would be very hard to have that system grown even if we can do it in a short run. So if we’d come to an agreement that 9-CM is not going to be the answer for the future where do we look, I know there was some discussion of just using ICD-10 for morbidity coding and that also didn’t seem to be appropriate. That wouldn’t provide the necessarily information that we’d be needing to adequately describe the morbidity situation.
So let me just briefly summarize that we strongly believe that the creation, review, and revision of health care policy relies on the availability of accurate and timely health care data, that these data are generated by providers and payers and by statistical surveys and other research efforts, but that the key underlying all of these information systems is the classification system that’s used to interpret these encounters and to provide information on morbidity status of the population. The 9th revision of ICD CM has served I think role very well since 1975, 1979, sorry, but it is really time to move forward and to implement some kind of revised classification. I think it is imperative that we do this in order to improve the kind of health care research we do, to improve patient safety and other quality of care issues.
The U.S. has been a leader, an international leader in this field, I think we still can maintain that position, but probably more important we run great risk if we continue to collect data using an outdated classification system at a time when there’s a greater need for information about the health care system, and I think particularly in the area of quality of care and also in trying to reduce disparities in care. These are areas I know the department is very committed to and without having adequate data it’s very hard to move forward in these areas.
We are very happy to work with the committee as usual and with our other partners in developing a classification system and adopting a system that will meet the needs of all users. My particular interest, given my position, is how to maintain the quality of our statistical systems and how to use that information for policy development and it is really a pleasure to be able to be here with you today to represent Dr. Sondik and to be involved in these discussions. Thank you.
Agenda Item: Overview of ICD-10-CM and ICD-10-PCS - Mr. Gustafson
MR. GUSTAFSON: Thank you, Jen. I’m Tom Gustafson, I am the deputy director of the Center for Medicare Management within the Centers for Medicare and Medicaid Management, the agency formally known as HCVA(?), and my component runs the fee for service side of the Medicare program, which is still very much alive and kicking, we have over 85 percent of the Medicare caseload still in a fee for service environment, and that may change depending among other things on bills under consideration on Capitol Hill, but it is not likely to go away during any of our professional lifetimes.
We process about a billion claims a year, not all in the inpatient sector, which is what we’re talking about today, but we process a lot of claims, they’re done electronically, coding is profoundly important to us because it defines the products and services that we pay for, so we take a very keen interest in this entire area. I come to you as a health finance expert, not as a coding expert particularly, so I offer an apology if I get some of the terms and concepts just slightly askew here, it’s not because I intend to do that but I’m still being tutored as many of us are. And part of my role as I understood it this morning was to make sure that members of the committee were acquainted with exactly what we’re talking about here, and viewed from a payment standpoint and the business standpoint that goes along with that.
So why are we having this discussion in the first place? And the reason for that can be traced back to the existing coding system, the ICD-9-CM, volume 3, which is where the procedure codes lie, and that’s the stuff that the Medicare program uses for paying inpatient hospital services and in several other payment systems that have evolved more recently that use the same coding system. This codeset was first implemented in 1979, that’s 24 years ago by an easy count, for use with inpatient hospital procedures. It involves a four character hierarchical numeric code, so you have four digits and you can only use numbers. It has been criticized for providing inadequate descriptions of many of the procedures contained within it, it is exceedingly difficult to update and expand within its longevity, and it is complicated. It is reflecting a reality that is now out of date, technological change has moved very dramatically in the health care sector in the past few years as I think all of us are aware, and unfortunately we have a data system here, a coding system, that is worse just where it should be best in terms of reflecting the updating of the technology.
The precise specification of procedures is very important for us, it permits us to make accurate payment in prospective payment systems such as the inpatient prospective payment system, and these kinds of codes in the system that’s upon them is used widely, not only in Medicare but with other public and private payers here and abroad. It allows evaluation of the efficacy of procedures, because if you know what’s being done you can check out whether it’s working or not. It allows us to measure outcomes and to some extent measure quality. And it can be very helpful to us in limiting and preventing fraud and abuse, and one particular aspect I’d like to mention there is one of the great concerns of the medical profession at least at present is what I would describe as false positives, and to say as the federal government has become much more interested in seeking patterns, looking for patterns of fraud and abuse and going energetically after the bad apples, there’s a great deal of concern that folks would be labeled as one of the bad apples when they are not. Having a more precise coding system should help alleviate that concern.
It allows a better understanding of technological diffusion, where are new procedures being used, how are they being used, what kinds of cases are they being used for, and it allows us as a payer to identify new technologies emerging on the scene, to ensure that they are paid adequately. One of our major concerns is to facilitate access for our beneficiaries to quality care involving among other things lately arriving technologies, and if we’re not paying enough money for those hospitals may not offer them, and I can come back to that. So in general we think that a system that is simpler and more logical would be a step forward and obviously one that is more modern and up to date.
In considering how the world might proceed in this general area there are sort of two general approaches that could have been considered and I believe these were debated at some length early in the ‘90’s. One would have been to augment the existing ICD-9-CM system by expanding the number of characters or by making them alphanumeric. Now that would have involved systems changes throughout the health care system, but with comparatively little advantage in terms of the improvement in precision. Some, but not overwhelming. On the other hand we could move toward a new and refreshed coding system which would also involve systems changes, and there’s no hiding that, but would also provide a more precise linking of technology and procedure and allow improvement in the underlying procedure codes. And that is the development that has led to today’s discussion, the ICD-10-PCS is the procedure chunk of ICD-10, which was developed as a replacement to ICD-9. This was initiated by this committee, the activity here was initiated, discussed and initiated in this committee back in the early 90’s and was funded in its development by the agency then known as HCVA, now CMS, and most of the legwork on this was performed by 3M, contractor with significant history in this area. The codeset is now sufficiently developed and tested where it merits serious consideration as a replacement for ICD-9-CM as volume 3, as the standard codeset for inpatient use.
The National Committee in November of 1993 specified the characteristics that they wanted to see in a procedure codeset, and let me just tick those off for you very quickly. One is that it should be comprehensive, all procedures should be classified within the codeset. Second, that the codes should be non-overlapping so that substantially different procedures should be able to have unique codes. Third, it should have expandability so that new procedures and technologies can be incorporated as unique codes when the need arises. Fourth, that it should have a hierarchical structure so that the individual codes can be aggregated into larger categories for various kinds of analysis. Fifth, it needs to be multi-axial so that each code character has a consistent meaning, each of those seven digits does the same thing in different places. Sixth, standardized terminology so that all terms are precisely defined, which is not a characteristic of the existing system. And finally that diagnostic information is not included in the procedure descriptions so that you have a separation of what’s going on in the diagnostic code and in the procedural code, and these factors guided the development of the ICD-10-PCS. As I reflect on them they seem to me exactly the sorts of things that a committee like this would imagine might be useful functional characteristics of a code set. And I believe that the ICD-10-PCS does a pretty good job in meeting those kinds of needs.
The structure of this dataset, this codeset, it’s based on a seven character alphanumeric code, so you’re dealing, going from four characters to seven and going from straight numeric to alphanumeric. This obviously provides you a very substantially increased number of possible combinations. The codes are divided into 16 sections such as medical/surgical, obstetrics, imaging, and so forth, I won’t go into them all, but it allies out to the major activities that need to be covered. And each of the seven characters function in a specified and consistent way, and I’m sorry I don’t have a visual aid to show you here, but the hurricane interfered with my ability to produce one in a timely fashion. But if you look in the medical and surgical section, for instance, I would just review these very quickly, you’re first digit tells you what section you’re in, in this case I guess would be medical and surgical. Second, body system. Third digit, what is the root operation, what are we talking about doing here, are we sizing something for instance, would be an example. Fourth would be a body part, so you’re in the abdomen or you’re on a leg or something. Fifth would be the approach, are you doing this laparoscopically, is it open, things of that sort. Sixth would be whether a device is used and if so what kind of device, there’s room for a variety of digits in there obviously. And the seventh field is qualifier field so we can provide some additional information in that seventh field. The approach is slightly different in some of the other sections, imaging for instance, but the same general hierarchical structure, axial structure, persists. And as a result of this, even though it may sound daunting to imagine gee, there are lots and lots more codes, the picture we have is that it’s actually simpler to use and simpler to understand. This codeset has been run in a head to head trial against ICD-9, and it wasn’t an extensive trial but a limited and I think revealing trial, indicated that after a comparatively minimal amount of time of training that experienced coders could code material in ICD-10 more quickly and with more accuracy then in ICD-9.
So in general the approach allows greater precision, greater simplicity, and greater expandability. The current status as I’ve sort of already gone ahead of myself and summarized here, the codeset is developed, we have the tools necessary to make it work, there is a complete crosswalk between ICD-9 and ICD-10 so one can march from one to the other immediately. There are definitions manuals, user manuals, training material of various sorts are already available and have been disseminated on our web page so that members of the community could examine them, and I said earlier that we have tested this. So that is just a short characterization of what’s at stake here with this issue, I understand that we’re going to here the Rand presentation and they’ll be some opportunity for further reaction after that.
DR. LUMPKIN: Any questions on the two presenters before we move on to the Rand? Kepa?
DR. ZUBELDIA: Yes, I have a question of one of the last statements you made that there is a complete crosswalk between ICD-9 and ICD-10. Is there a crosswalk for each ICD-9 procedure code to one ICD-10 procedure code?
MS. GUSTAFSON: I don’t know if it was a one to one crosswalk, but Pat Brooks could perhaps help me here?
MS. BROOKS: It starts with a ten code, you look at the ICD-10-PCS code list on our web page and it gives you a nine code that it goes back to and many of those are one to one, sometimes they are not.
DR. ZUBELDIA: So from ICD-10 to ICD-9 --
MS. BROOKS: Yes.
DR. ZUBELDIA: And it’s many to one.
MS. BROOKS: Yes, there can be many 10’s that have come from 9, because ICD-9 procedure codes have a lot of things involved, they’re procedures of every artery in the body in some codes, so obviously there would be a lot of 10 codes list, go back to one 9 code.
DR. LUMPKIN: Okay, any other questions? Simon.
DR. COHN: I think this is just one for you, Tom, I actually don’t have a specific question to ask you about, I just wanted to make sure that you were going to be around as we hear from reactors because there are a number of questions I have but some of it relates to information we’re going to be receiving later on. So you’ll be with us for the remainder of the morning and the afternoon?
MR. GUSTAFSON: I’ll be here, I think you’ll discovered me listed among the reactors. I’m hoping, I was intended to be instructive at this point and more of an advocate later.
DR. LUMPKIN: Well we look forward to your advocacy role. Peggy?
MS. HANDRICH: Hi Tom. I’m wondering whether in your overview of the coding you represent the CMS view from the perspective of Medicaid, has there been discussion on the Medicaid side about this issue?
MS. GUSTAFSON: Yes I’m told. We are the Centers for Medicare and Medicaid Services, but I don’t know, I believe that there have been some discussions with representatives of the Medicaid community and we could perhaps give you some more detail on that.
DR. LUMPKIN: Thank you, we’ll move on to the presentation of Dr. Libicki.
Agenda Item: Presentation on ICD-10 Impact Study - Dr. Libicki
DR. LIBICKI: Good morning Mr. Chairman and members of the committee. My name is Martin Libicki, I am a senior policy researcher at the Rand Corporation, a non-profit research organization providing analysis and solutions for public and private sectors. My presentation today will cover the work of the Science and Technology Policy Institute of Rand did on the cost and the benefits of switching from ICD-9-CM to ICD-10-CM and ICD-10-PCS. This work it should be noted has not been completely through the Rand quality assurance process although a draft of the report is available for public comments. The views expressed are mine alone and do not represent Rand or the sponsors of our research.
Rand was asked three questions. First, what are the costs and the benefits of switching from ICD-9-CM’s diagnostic codes to ICD-10-CM? Second, what are the costs and the benefits of switching from ICD-9-CM’s procedure codes to ICD-10-PCS? Third, if it is advisable to switch to both ICD-10-CM and ICD-10-PCS should they be done sequentially or simultaneously? In answering these questions we draw on three sources. Prior testimony before the Subcommittee on Standards and Security of the NCVHS, literature review, and interviews mostly by phone of experts and stakeholders throughout the health care sector. In the middle of April until the beginning of August we conducted 80 interviews in which we talked to about 90 people.
Our presentation today will focus first on costs and then on benefits. We will not repeat the introductory and methodological material from last May, nor go into detail about the codes themselves, except to note briefly that their most salient feature from our perspective remains the greater specificity that they offer, a mostly two fold increase in codes for ICD-10-CM and a 50 fold increase for ICD-10-PCS. Although neither ICD-10-CM nor ICD-10-PCS were designed as a strict one to many expansions of their predecessors, ICD-10-PCS by virtue of the 50 fold increase in codes is a de facto one to many expansion while ICD-10-CM is only a one to many expansion for about two thirds of its codes. We intend to present both costs and benefits as ranges in order to reflect the uncertainties of any such estimation. In so doing we will start with costs.
We estimate the costs of the transition to be $425 million to $1.15 billion for the first year, and $5 to $40 million dollars a year thereafter. These numbers assume that the switch to ICD-10-CM and ICD-10-PCS occur simultaneously. If considered separately one can add up to another $20 to $170 million dollars to that number.
The costs basically can be divided largely into two costs, those for training and those for systems changes. The costs of training people to use ICD-10-CM and 10-PCS can in turn be subdivided into the costs of training hospital coders, non-hospital coders, code users and others. The amount of training required for coders has been studied and results from other countries serve as benchmarks. We start off with the roughly 50,000 hospital coders, which can be understood more or less as tantamount to the number of full time coders even though the overlap is not 100 percent. The time to get such coders to proficiency in the new codes, mostly ICD-10-PCS, has been estimated at several days to a week by AHIMA among others. If such training costs $2500 dollars per person, split between expenses and lost work time, then the training bill is roughly $125 million dollars, reflecting uncertainties that gives us a range of about $100 to $150 million dollars. One caveat, however, merits note. Gail Graham from the Veterans Administration informally tested ICD-10-PCS on coders and concluded that while good coders essentially had no problems with the switch mediocre ones did because their knowledge of anatomy and physiology was incomplete. If they have to bolster their knowledge of such fundamental subjects this would endeavor training of several months with a cost of perhaps $10,000 dollars each or more. When multiplied by 10,000, if one talks about the least proficient 20 percent of coders, this would add another $100 million dollars to the costs. That noted there is no evidence that such broad training was required in Canada, where an analogous transition was made in coding procedures except their code expanded for procedures by a factor of six rather then 50. In Canada it took two days of intensive training plus additional follow-on sessions of a few hours each to get coders up to speed. There were initial losses in productivity, they were modest and they disappeared largely or totally within six months. Australia’s training experience was similar.
What of those who code only ICD-10-CM? That is those who deal with outpatient records? With 200,000 doctors offices one can think in terms of an equivalent of 200,000 coders. Yet almost all the diagnoses in any one office usually involves a small number of diagnostic codes, some offices preprint the codes either in hard copy or on disk. Many physicians officers do not record diagnoses with the highest degree of specificity in any case. It is not unknown for physicians, presented with a choice of 40 different diabetic codes in ICD-9-CM to use the most general code in billing, thus little or no training will be asked for in many offices. Some inpatient coders will simply train themselves as they go and there are potential efficiencies in training that may be available by using the web as an instructional vehicle. More broadly AHIMA has estimated the retraining time for part time coders at roughly eight hours, yielding a cost figure of $100 million dollars. We will use $50 to $150 million dollars, however, as a range of costs to train coders outside of hospitals.
Incidentally, those present at the August meeting of the subcommittee heard an estimate from the AHIMA AHA study that indicated that coders for ICD-10-CM wanted an average of two and a half days of training. This figure does not invalidate the eight hours that we used, first their testers were largely full time coders and even teachers, not part time coders. Secondly, working training is not the same thing as getting paid for training. As to the training, retraining of physicians, one responder indicated that the requirement for training would depend on how different ICD-10 was from ICD-9. At the high end another estimated that each physician would need four hours of education on the new diagnostic codes, plus two further two hour sessions in the first year for a total of eight hours. If multiplied by the half million physicians in clinical practice this sums to roughly 2,000 man years of training at a cost of several hundred million dollars. Nevertheless, no physician we talked to said that they themselves would have to be retrained. Professional coders for their part are extremely leery of having doctors do their own coding.
Finally as noted, many physician offices, for better or worse, rely on super bills to capture their diagnoses for the record. If ten percent of all physicians take eight hours of training the overall costs would be roughly $50 million dollars as a point estimate, and perhaps $25 to $100 million dollars as a range reflecting these uncertainties. Code users may also need some time to become familiar with the new codes, perhaps half of the quarter million employees of payer organizations work directly with codes. It is unclear how much training they will get. Formal training in ICD-9-CM remains the exception today.
To some extent employers may count on people seeing codes and descriptions on claims forms often enough together to link them. No payer has mentioned any formal program, formal training program, although to be fair few have yet to focus on the issue. If the average employee needs the aforementioned eight hours of training then roughly $50 million dollars worth of training man years would be required. If only half are trained the cost would be $25 million and these two numbers form our range.
There may also be some loss in productivity. CMS formally tested ICD-10-PCS in 1998 by using two clinical data abstraction centers. One of them, Dinecheypro(?) report indicated that no significant time difference in coding was found between ICD-9-CM and ICD-10-PCS. The other one, F-MAST(?), reported that ICD-10-PCS generated more codes then the ICD-9-CM, and that each record on average took 1.7 minutes longer to code, or 50,000 extra coding hours equal to $3 million dollars when multiplied by the 1.8 million discharges a month from hospitals that have procedures associated with them. If one assumes a six month learning curve, at the end of which only a third of the difference persists, in the long run productivity costs are roughly $12 million dollars a year. But these data must be qualified by the observation that 200 records is a very small sample upon which to base conclusions and most hospitals use automatic encoders these days, which were not part of the clinical data abstraction center studies. So we use a range of zero to $15 million dollars a year to reflect lost productivity among coders.
Another possible source of productivity could stem from the fact that these codes, requiring as they do more input from the patient record, will require in some cases that more information be added to the patient record in order to fill out the codes. If focusing on ICD-10-PCS and assuming one, that 80 percent of the patient records either describe routine procedures or are well documented, two, coders will get back to physicians only half the time they need information, and three, physicians and coders need to spend five minutes for each such patient then multiplying by the aforementioned 1.8 million inpatient discharges gives you at least initially 15,000 physician more work hours, and the same number of additional coder work hours, or $3 million dollars worth in the first month. Again, this figure will decline as physicians get used to what they have to note in the record whenever certain types of procedures are done. So we figure a range of $5 to $25 million dollars in lost productivity over the long run.
Systems upgrades account for the other half of the cost of transition. Providers will have to alter their billing and administrative systems to handle new codes, which in ICD-10-CM and PCS are longer and contain alphanumeric characters rather than just numbers. Most providers, however, buy their systems from vendors and most of the provider/vendor relationships have a software maintenance component which pays for periodic upgrades, especially in response to government regulation. Nevertheless, Bob Hardesty(?) from Cerner did testify before the subcommittee that “it would be expected that there would be professional service fees for planning, testing, training, conversion, and implementation. Providers themselves may have to do some of their own planning, engineering, and user services such as report generation, testing, ensuring backwards compatibility with earlier data and systems integration.”
Overall the level of concern we found with such codes was modest, with neither the AHA nor the FAH raising such issues before the subcommittee, both of which were more concerned about covering training costs. Major hospital chains similarly were not terribly concerned about the change, believing that they would need little additional consulting and for them the major costs would be testing and integration. One hospital chain representing .5 percent of the total inpatient market estimated the cost of training and integration at a maximum of a man year. Another said it would be close to a trivial event. The Veterans Administration, which accounts for roughly three percent of the total potential patients does not plan to spend more then a million dollars. But there were outliers, “easily a million dollars worth of work among hospitals,” which may account for a disproportional share of the costs and makes estimates difficult to generate on the basis of small sample sizes. We therefore used a wide range, $50 to $200 million dollars, to represent the cost of updating provider services.
Estimates were also made for the costs of updating software. At the high end two software vendors expressed a need for as much as three years to make the change. The man year estimates were in both cases roughly one man year for every ten employees they currently have on board. A third major vendor cited a rough estimate of 50 to 100 man years for total adaptation to ICD-10. Others such as Quadrimed, Cerner, Solution, Core Solution, Symetry Ingenics(?) talks of a few man years of effort, so rounding this all up we used a figure of $50 to $125 million dollars for this sector.
Fixing payer systems is also a large cost estimate and payers vary widely. Some, generally the larger ones, develop and maintain their own systems. Their costs could be large and seem to fall either into a lower rate, of 40 cents per member, or a higher rate of $1 dollar a member. Others, perhaps buffeted by earlier changes required by Y2K or HIPAA, have come to rely on vendors and face fewer direct costs and perhaps fewer costs overall. Based on the former, that is those who do not use vendors, we estimated the cost of adapting software to accommodate ICD-10 between $100 to $250 million dollars if extrapolated to the entire payer base.
Finally we have no hard data for CMS, which at the time of our research had only begun to estimate the impact of ICD-10-CM and ICD-10-PCS on their costs. Based on the size of CMS, but tempered by the fact that CMS does not do everything that an insurer does, we used a figure of $25 million to $125 million dollars as a placeholder for their costs.
Last is a word about how costs are distributed. As a rough estimate, hospitals are likely to bear 35 percent of the costs, outpatient facilities 25 percent of the costs, and payers, including CMS, about 40 percent of the costs. Within these categories we found that the likely cost to accommodate a code change are roughly proportional to the number of patients the providers see or the number of members that a payer has. We were not able to detect or predict any systematic shift in payments as a result of the change in coding between ICD-9-CM and ICD-10-CM and PCS. Current plans suggest that DRG’s will not be redefined until sometime after ICD-10-CM and PCS goes into effect. The fact that CMS pays for nearly half of all hospital costs mitigates against any large immediate shift in reimbursement rates elsewhere within the medical sector.
Now for the benefits side of the ledger, where we note that the discussion will be more qualitative in nature and that the ranges assigned to benefits will be accordingly much wider. Assumptions play a large role in these calculations.
We identified five categories of benefits where there could be a reasonable basis for enumeration. The total benefits over ten years we estimate to be between $700 million dollars and $7.7 billion dollars, depending on estimates and assumptions. Following discussion of these five categories of benefits will be mentions of four categories where benefits are possible but which could not be enumerated for reasons detailed below. The five categories are one, more accurate payment for new procedures. Two, fewer rejected reimbursement claims. Three, fewer questionable reimbursement claims. Four, better understanding of the value of new procedures. And five, improved disease management.
We start with category one, more accurate payment for new procedures. As one researcher has argued the predictive power and clinical utility of a DRG classification system is limited or empowered by the precision of the available diagnostic and procedure codes. This dictum applies through particular force to ICD-10-PCS. With more codes available it is now possible to differentiate simple and complex inpatient procedures, there they’re two lumped together. Different procedures can be reimbursed differently. This in theory is a net plus if procedures that would have been optimal for patient care were not performed because providers feared they would lose money doing so. In practice, however, this possibility was not raised by any provider we talked to and certainly not by any payer.
In recent years CMS has started a program to reimburse elected new procedures at rates higher then what is indicated by the DRG’s that says procedures are associated with. The limitations on expanding ICD-9-CM volume 3 are a major reason that very few such procedures have been approved. With ICD-10-PCS the additional approval rate for additional payment, with ICD-10-PCS the approval rate for additional payments for new procedures could rise to several dozen a year. Estimating this benefit required making assumptions about the number of new procedures introduced annually, the number that would not get their own code in ICD-10, excuse me, the number that would not get their own code where ICD-9-CM volume 3 in use, the percent of them whose benefits would exceed their costs but whose costs would exceed the reimbursement allowed for their DRG, and the percentage of Medicare/Medicaid patients who would not get procedures as a result. In this scenario therefore we estimated benefits to range somewhere between $100 million and $1.2 billion dollars over ten years.
Category two, fewer rejected reimbursement claims. As a workgroup for electronic data interchange observed ICD-10-CM and ICD-10-PCS incorporate much greater specificity and clinical detail, which may help reduce the number of cases where copies of medical records need to be submitted for clarification for claims adjudication. By one estimate only one in five hospital claims make it through with complete payment and no additional questions asked. Nevertheless, almost everyone who commented on this issue felt rejection rates would rise initially as people unfamiliar with the codes put down the wrong codes, or as payers who also have to learn the new codes err and reject correct codes. We estimate that within a year or two the rejection rate will be below what it had been otherwise, and otherwise you’ll initially see a rise and then a fall. And now within five years a cumulative number of rejections will fall below what they would have otherwise have been. In other words it takes about five years for the up part of the curve and the down part of the curve to balance off, leaving the down part of the curve in year six through ten. Our overall estimate is based on assumptions about the share of health care financing costs arising from having to adjudicate claims, the number of cycles the average claims makes, and the percent of claims that can eliminate one cycle thanks to more detailed discharge records. We estimated a range of $200 million to $2.5 billion in benefits for this category over ten years.
Category three, fewer improper reimbursement claims. As for fraud, the WEDI report goes on to argue, if a new coding scheme were implemented there would be increased opportunity for fraud in the beginning when people are less familiar with the new codes. It might be more difficult to detect potential duplicates, unbundled services, or up coding of procedures during the transition when two versions of codesets would be in effect. In the longer term it is possible that fraud could be reduced since ICD-10-CM and ICD-10-PCS are more specific and there are fewer gray areas in coding. Although the General Accounting Office has used the figure of $13 billion dollars to represent the sum total of all questionable reimbursement claims, the vast majority are either differences of opinion or providers seeking the benefit of the doubt on close calls. Only a minority of them are improper per se, our estimate was based on the assumptions related to what percentage of the $13 billion are clearly improper, of these how many exploited the ambiguities of ICD-9-CM, of those how many would recognize a greater clarify of ICD-10-CM and PCS and abandon that tack and not exploit other weakness. So in other words we took a large number and we subsequently divided it by a number of percentages. We concluded after that by assigning a benefits range of $100 million dollars to $1 billion dollars for this category.
Category four, better understanding of the efficacy of new procedures. In estimating the value of research on procedures several assumptions could be kept in mind. First, only a certain percentage of procedures, that is those that did not get their own ICD-9-CM volume 3 code are likely to win new separate identifications in ICD-10-PCS and their identification assumes correct coding. Second, insofar as a point of research is to effect decision one is looking at the decisions of dual procedure more often, say for a broader selection of patients or less often, that is to say consider alternative procedures in some cases or not do the procedure altogether. Such research will say little about how to do the procedure, except when two different ways of doing the procedure are identified as different procedures in ICD-10-PCS and are therefore found to yield different outcomes. Perhaps needless to add practice is unlikely to follow research results immediately or totally. Third, many of these procedures are so well accepted that no feasible amount of research is likely to effect a decision to undertake those procedures. Fourth, research using ICD-10-PCS hospital discharge data is only one of many sources of information and procedures. Payers, for instance, already have access to physician based claims and coded NCPT, and there are also in many cases controlled studies that rely on information and patient records, or conveyed directly by procedures. Last but perhaps most important in judging the efficacy of new procedures are stories passed around among specialists of failures and success.
Our estimate picks up many of the assumptions of the first benefit, notably the number of new procedures done every year that do not get their own code, but requires a separate estimate on what percentage of such procedures are significantly informed by the analysis of discharge statistics on who underwent them. We then need to estimate how often based on such information procedures are introduced to new populations or contraindicated for populations who would have otherwise gotten these procedures. Calculating through these assumptions generates a range of $100 million to $1.5 billion dollars to represent this benefit.
Category five, improve disease management. Increasingly care management organizations and providers are working with patients, especially those with chronic conditions such as diabetes, asthma, or heart condition, to ensure that they receive the correct quality of care, reminders for treatments, and referrals to classes and other self management techniques. The point here is to postpone or eliminate the onset of the more serious conditions that send people to hospitals. True, many of these services could provide, could be performed by primary care physicians, these physicians in turn would rely on personal knowledge of the patient or detailed notes on the patients rather then simply on discharge codes. Yet many patients lack primary care physicians or have no single such physician who could manage the patient. Since much of the work of disease management, such as finding eligible patients, sending them reminders, and giving them information can be automated organizations can run through patient records en mass and pick out patients who need disease management and deliver such management automatically.
Although disease management programs are proving popular a tempered optimism is called for in evaluating how better codes may attribute to its efficacy. As it is many primary care physicians do not use the specificity already present in ICD-9-CM. It is unclear whether they will react to a more specific set of codes by using distinctions that previously lay unused. Others produce diagnosis that indicate not the underlying cause of the encounter, such as diabetes, but its proximate cause, such as kidney dysfunction. One respondent has suggested that some differentiation in diabetes, such as neuropathy, exceed what are typically recorded even in doctor’s notes. As a result many care management organizations use other indicators to differentiate diabetic patients from one another. Some use HBA 1C tests and diagnostic codes for comorbidities for diabetes. Perhaps half of all payers classify diagnoses by broad episode treatment groups of which the most popular set has 500 to 1,000 and thus are more aggregated then ICD-9, let alone ICD-10. The more sophisticated algorithms use a mixture of diagnosis, procedures, laboratory tests and medications to determine what the patient may be suffering from and thus whether disease management is warranted. The finer details available from ICD-10 cannot hurt the implementation of this algorithm but responses to date do not suggest it will contribute that much more accuracy to ETG determinations either. One provider is moving away from codes altogether and going over to SNOMED for such purposes.
We chose to focus on diabetes partially because it is such a large disease management challenge and partially because of the anticipated five fold expansion in number of codes available as one goes from ICD-9-CM to ICD-10-CM, which are available for this disease. Two category of benefits were estimated, those from getting more of the right people into disease management, and those that come from being able to refine disease management for the type of condition the patient has among those who are already in disease management. Our calculations first required an estimate of the number of diabetes patients in programs that offered disease management, excuse me, first required an estimate of the number of diabetic patients in programs that, in disease management programs and that use ICD-9-CM as a primary source of information on payments. From that basis we made assumptions about the percentage of diabetics who enter the program as a result of more detailed information available from ICD-10-CM who therefore reduced their primary indicators as a result of being in this program, and who therefore added quality months or years to their life. Under the scenario we used the benefits estimate, the points estimate was $500 million dollars.
The second half of the calculation required estimating the number of patients in the program, the percent of those whose program is fine-tuned by ICD-10-CM information, and the percentage of those who thereby reduced their primary indicators and add quality months or years to their life. Under the scenario we used the point benefits estimate was $300 million dollars.
Our last assumption was that two thirds of all the benefits of disease management that come specifically from using ICD-10-CM accrued to diabetic patients. By contrast the code expansion from coronary disease was modest and the same applies to asthma. Our final estimate range therefore was $200 million dollars to $1.5 billion dollars. As noted there are an additional four categories of potential benefits that we could not estimate very well, or in the latter cases even determine whether they were at all significant, but for fairness they should be presented.
Therefore, category of six, better understanding of health care. The tens of millions of annual hospital discharge records should hold enormous potential for assessing the relationship between health care and outcomes. One prominent researcher observed that ICD-10-CM had the potential to say a good deal about the quality of care and its codes so that data could be used in a more meaningful way, enabling better understanding of complications, better design of clinical robust algorithms, and better tracking of the outcome of care. Nevertheless, several factors should be included when thinking about this benefits. Studies that infer conclusions from hospital discharge statistics are limited in their ability to control population samples or data quality. Furthermore, some of the gains from having finer distinctions are vitiated by the smaller population sizes within each category, which in turn erodes statistical significant. Indeed, many researchers agglomerate information to super categories, not only to create statistically significant sample sizes but to avoid the misleading effect of minor coding errors or coders unwilling or unable to make the fine distinctions that ICD-9 already allows.
Last, hospital discharge datasets alone do not say what happened to the patients after the discharge, how long they lived, how they fared, or whether they reentered the hospital even if it’s the same hospital. There are two ways around the problem. One is that AHRQ maintains a database with ID numbers and if one can get permission to access such data, which a few but perhaps too few researchers have, more follow-up information would be available. Two is that payers and data consolidators that work with payers have access not only to names but to inpatient records as well. Mining such data has already generated useful conclusions about procedures such as coronary artery bypass grafts. On that greater specificity has to be understood to be beneficial but exactly what we will learn from greater specificity is extremely hard to forecast with any confidence. For instance, it is altogether possible that ten years later we will look and say this fact we learned because of greater specificity, and that fact may in fact account for more then half of all the benefits from medical research that have been aided by greater specificity. But that’s an extremely difficult statement to make looking forward and therefore extremely difficult to put a number on that kind of conclusion. Such uncertainty typifies all serious research and all serious attempts to measure research.
Category seven, fewer erroneous claims. Now I had talked about fewer claims that were sent back and fewer claims that were questionable. These are simply miscoded claims that have no other implications other then random ones. The cost of reducing random error is difficult to put a number on without knowing the consequence of such errors. If errors that are not caught result in erroneous payments and such uncaught errors are randomly distributed, in other words people are as likely to over bill as under bill, then the effects washout entirely. If errors are non-random, but not large enough to bias a provision of medical care, the payers’ gains will be the providers’ losses, or if it turns out to be the other way around, vice versa. If the errors are caught the result is likely to be a returned claim, the benefits of whose reduction have already been estimated. Therefore we did not put a number on fewer erroneous benefits, but one has to believe that fewer errors is a good thing.
Category eight, enhance the ability to evaluate providers. One could analyze hospital discharge statistics to gage the ability of various providers to undertake specific procedures or respond to diagnoses that were previously lumped in with others. But one is doing is taking existing data and parsing it finer, bringing statistical significance from such additional information is very difficult when no one provider accounts for more then three percent of the U.S. patient population. In other words to evaluate providers you have to basically say using this procedure how well did they do. You can evaluate it against a total corpus of procedures if you had the data but any one provider only provides on average at most three percent of these procedures, so you have a real problem with statistical significance when you evaluate providers that you don’t have when you take a look at the nationwide database in a particular procedure.
There is an alternative, one might choose to aggregate all new procedures performed by a given provider for analysis. That would generate perhaps statistically significant data but its unclear whether this would generate meaningful data. It is also possible that finders diagnostic and procedural codes could be used to better norm the patients that providers treat so that one understands that hospitals and other providers that have on average, that have sicker patients on average, would have that fact recorded when it comes times to evaluate quality of care. Unfortunately ICD-10-CM’s contribution to indicating severity is slight. We therefore conclude that the measurable benefits from greater specificity when it comes time to evaluate providers is likely to be quite modest.
Category nine, more timely intervention for emerging diseases. It is also theoretically possible that more precise categorization of diseases could help the public health community detect emergent diseases, including bioterrorism, faster. If better diagnostic codes could improve the odds of even by a few percentage points of heading off the next SARS like disease the switch could pay for itself many times over through that benefit alone. SARS is said to have cost the Asian economy over $11 billion dollars and SARS had not run its cost at the time that estimate was provided. Similar benefits may come from being able to spot clusters of diseases that might be traced to environmental or occupational conditions. But that noted, ICD-10-CM was not designed for syndromic surveillance and access to hospital discharge records would have to be accelerated to provide for timely intervention in case of a fast moving epidemic.
Finally, if switching to both ICD-10-CM and ICD-10-PCS is warranted, is it better to do both at once or one after the other? Many witnesses to the NCVHS and its subcommittees and those we interviewed shuddered at the cost of trying to implement new codes before they had finished complying with HIPAA requirements. But no one who testified or who interviewed suggested it was better to implement the code switch sequentially. Canada, which also implemented changes in both diagnoses and procedures, did them together. There are several reasons that switching both at once makes sense. It is easier to get people such as coders mentally prepared for a major change once rather then twice. It is also more efficient. Training only has to be done once. It is easier to review a complex billing software system where ICD-9-CM is used and has to be replaced then it is to do it twice, first for volumes 1 and 2 and then for volume 3. Simultaneous change means testing only once. Dealing with transition issues such as getting through the errors of miscoding or misinterpretation again is best done all at once rather then in two periods separated by months or years. People can focus attention and resources on the process of making a smooth changeover and move on. Finally, the long awaited DRG’s, dependent as they are on both diagnosis and procedures, would have to be further postponed if it had to wait for first diagnosis and then procedures to be revamped.
We have endeavored to estimate the cost and characterize the benefits of switching from ICD-9-CM to ICD-10-CM and PCS but in concluding we thought it would be helpful to introduce a broader perspective of the discussion. At almost a trillion and a half dollars a year the U.S. health care industry is enormous, it would not take a very large percentage shift in its efficacy or efficiency to make the kind of economic difference that would easily swamp cost estimates that are made in the hundreds of millions of dollars. Nearly every other industry has found that if it wants to get on a program of a continuous process of improvement it has to start by measuring input and particularly output. Only after such measurements can alternatives be systematically compared. As an industry it’s fair to say that the health care sector has been laggard in this approach, leaving room for improvement. As such the use of codes, or at least the information contained in these codes, is central to the measurement process and hence central to the goal of continuous process improvement. Anything that improves such codes provided that they contribute to good measurement cannot help but have positive long term ramifications. But all this is predicated on the idea that we know what we want to measure to improve outcomes and how we want to measure them.
Many people we have talked to had done quite a bit of thinking on this score. Patient safety advocates laud the present on arrival adjunct to diagnosis as for from hospitals in California and New York. In other words when you discharge somebody from a hospital and you go through the various diagnoses you ask yourself was this diagnosis present when the person entered the hospital or did this condition arise after somebody entered the hospital.
Those who measure patient quality of care are interested in whether providers comply with HEDIS standards, in other words are ace inhibitors or beta blockers given to heart attack patients within the first 30 minutes in an emergency room is one such measure. Others had noted that they would find indicators that differentiate diagnoses by severity to be quite useful. In many cases severity can be directly indicated through laboratory test results, such as HBA 1C for diagnosis or creatinine for kidney conditions.
Overall there are good grounds for believing that there will be benefits, which we estimate to be $700 million to $7.7 billion dollars over ten years, and that these benefits are likely to exceed the initial break in costs of $425 million to $1.15 billion dollars within a few years of when the code is adopted. What ICD-10-CM and PCS permit is a considerable expansion in what we can know about the health of a population and the care it receives. The extent to which this information will be properly and intelligently exploited is hard to gage at this point, but the potential gain is large if the opportunities are taken.
DR. LUMPKIN: Questions? Kepa.
DR. ZUBELDIA: Thank you for the report, I think it’s a very interesting report and you’ve done a lot of work, I want to thank the Rand Corporation for doing this work. I have a question about, perhaps it’s related, but the lack of CMS information and how could that be assessed rather then just estimate the cost is going to be such, there should be also some benefit for CMS, and there should also be benefits for payers. The concern that I have is that it seems that the study is looking at the benefits from the statistical side, the top three benefits are related to claims payment, but it doesn’t seem to look very far into the costs for payers. And let me give you some scenarios, the payers are going to have to redo their coverage plans and renegotiate the coverage plans for the employers as to what is covered and what is not. They’re going to be have to redo the DRG pricing, every single reimbursement schedule is going to have to be redone. And the potential there is that perhaps it can be done as a probably neutral change, but that’s going to be a substantial disruption to the payment system and I don’t see in the study the costs related to these changes in the payment system. Is that something that you’re planning to do as you continue and finish the study?
DR. LIBICKI: Well, let me kind of address your questions in order. In terms of CMS, when we talked to CMS, I think in the May/June time period, they were about to start studying what their costs were going department by department, so I have to believe that numbers from CMS will be available if they’re not already available from CMS directly itself. In terms of the benefits to CMS from this we did not identify any as such, mostly because we were concentrated on what decisions would change as a result of better information. Not to wax terribly philosophical about this but what you end up with when you change information, when you get better information, is you end up changing decision. If you haven’t changed decisions you really haven’t gotten a whole lot of benefit from changing information, so the difficult part was to try to say okay, which decisions actually get changed and that was the basis of our estimates, is to try to identify okay, we do procedures differently, we do procedures, when we don’t do procedures, we are able to expend less manpower in evaluating, in going through claims, etc., which by the way gets to the second question, which is to say the benefits to the payers. If the payers start getting better data from providers presumably they will not have to do as much work in order to adjudicate claims and that will save them costs, I mean it will save the entire system costs but they will be the first people to get the costs.
Now the last question you have raised is a rather, is a trickier question entirely and there are two aspects of the question. One is how much work is it going to take in negotiation/renegotiations and the second is what’s going to be the impact on the payment schedules. The quality of negotiation/renegotiations actually almost entirely falls on ICD-10-CM, not PCS, and the reason is because if you wish to for your own purposes say okay, here’s a condition in ICD-9-CM volume 3, and I don’t want to change the rules, it is not in my benefit, it is in my benefit to make the least possible changes, then you have 10-PCS things you can crosswalk into easily, because there’s just so many 10-PCS codes and you do a one for one substitution. Where you cannot do a one to one substitution very easily is in ICD-10-CM because you don’t have a one to one relationship and then you’re basically looking at where you have the mismatch, because you are going to have some one to ones and not one to ones. Okay, so you have to take that into consideration.
The second thing one has to take into consideration, particularly when it comes to DRG’s, is we’re talking about a process that takes place over a multi-year period under which you’re going to have to have renegotiations anyway. I mean I think it’s unrealistic to say on the first of October the year 2006 everybody goes into renegotiations, there are various cycles, people are going to anticipate them, some people have annual renegotiations, some people do things on two or three year periods, and some of that can be folded in. Now, if it turns out that a payer says you know, now that we have all this greater specificity I don’t want to pay for this or I’m willing to pay for that, that’s a different issue but then I have to assume that the payer benefits directly from that negotiated change and therefore on net has to be a beneficiary on that basis.
When we talked about methodology back in May that was one of the basically the assertions I made in terms of how you do a cost/benefit analysis. In other words, and this is going to sound silly but it really works out this way, if as a result of new information you do something smart, that’s a benefit. But if as a result of a new information you do something which ends up being dumb, that’s not a cost. It sounds strange but that’s the way we deal with the economics of information, that’s kind of how it comes out.
DR. LUMPKIN: Simon.
DR. COHN: I’ll have to think about that last comment, I’m not an economist and I don’t play one on TV either. It’s almost a follow-on with Kepa but first of all I want to thank you, I think this report is getting much better and really appreciate the way you’re beginning to put sort of bounds around the issues where I did notice in your presentation for the first time began to posit at percentages of costs by payer, provider, hospital. I didn’t hear the same thing, and maybe I just wasn’t listening closely enough, around the benefits. Did you parse out the benefits similarly?
DR. LIBICKI: The benefits, can I parse out the benefits, short answer, no I didn’t parse out the benefits. And the second reason I didn’t parse out the benefits is something called flow through. In other words if as a result, if payers can more efficiently process claims does that become a profit to the payer or a lower cost to the, if XYZ payer now saves $10 million dollars because they can process claims more efficiently, does that mean that XYZ company makes $10 million dollars more worth of profit? Or does that mean on the next cycle they can hold down inflation and reduce costs by $10 million dollars? Unless you know those subsequent decisions you don’t know who the ultimate beneficiary is and therefore it’s safer to basically say there was just a total benefit without having to make that overall statement. On costs at least it’s possible to be more specific on the first order of how costs are allocated because you see direct payers on that. Let me give you another example. I am a patient that gets a $10,000 dollars procedure that I wouldn’t have gotten because now we understand the benefits of the procedure. We will make for assumption sake that I get $15,000 dollars worth of benefits from that. On the other hand, so I’m $15,000 dollars better off but then somebody has to pay for the $10,000 dollars, it’s either my employer or whatever, actually in this case I suppose it would be CMS, so CMS is $10,000 dollars worse off. And it goes on and on and on, what economists call the incidence issues are subject to a large number of further assumptions but they don’t enter into the total if you just have to make an overall guess of cost and benefits. In other words without worrying about who pays for it you at least can conclude that somebody pays for it and it turns out to be a cost.
DR. COHN: Can I follow up on that, certainly if you have thoughts about how we can begin to look at that because certainly there are things we’ll have to think about, is that we now know how the costs break down, the question is is do the benefits accrue to the same groups or is there a disparity there, so any thoughts you have on that will be helpful. I read through your document, obviously a number of them as you’ve draft them, and you made a comment just now about this issue about how costs get transferred and what gets reflected in there. One question I have had and I don’t know how this reflects in your transfer payment model, and maybe it’s the same question that Kepa was asking, but on a more global view there’s, given a new codeset coming in or new codesets coming in that somehow out of that implementation, and lets just sort of posit that on one hand global health care inflation for whatever reason might jump at some point, on the other hand it might increase at some point, and that we were somehow knowing that there are 12 different reasons why that might happen but say we were able to posit some relationship to the introduction of the new codesets, how would you, I mean is one considered and the other one not or are these things that are costs in the system in your model?
DR. LIBICKI: If you were to go back statistically and look at sort of the prices of services and you say ah ha, this changed because of ICD-9/10, only under the case where you could point to a change in the DRG and say ah ha, they now cover this and I can trace this through is it possible to actually make a statement like that. But if you don’t have changes in the DRG’s or you have many changes in the DRG’s, some of which cancel each other out, when we talk about a $1.4 trillion dollar industry to find the effect of several hundred million dollars worth of costs is statistically impossible.
DR. COHN: So it would have to be a major differences --
DR. LIBICKI: Yeah, that’s right, because you have so much variation from one year to the next coming from so many different issues. If I may go, I think the Latin term is ex cathedra, on a further point, a lot of cost/benefit analysis starts with the assumption that you already have an optimized system, and then it makes a change to that system and say okay, this system is already optimal and then how are the costs and what are the benefits relative to that optimum. But there are a lot of people, and I don’t want to get into this issue except to mention it, who suspect that the medical reimbursement system is not optimal today and therefore changes to that system may in fact move you farther or closer to this optimality. But to do that I would have to go into are patients getting the right care, are we having too much care, too little care, etc., and we would have ended up with a much larger study.
DR. LUMPKIN: Michael.
DR. FITZMAURICE: I want to follow on what Simon said about the distribution of costs and the distribution of benefits. I find that the same incident assumptions hold for benefits as for costs, for example, if the health plan incurs $1 million dollars in costs from implementing ICD-10 does this reduce that health plans profits by $1 million, or does it result in higher premiums by a million to its payers of the health plan premiums, or somewhere in between, with any kind of incident, instead of zero to 100 percent, it’s somewhere in between. Likewise the same thing with benefits so I think you could have looked at the initial incidence of costs as you did and also the initial incidence of benefits but realize that over the long run markets will play out and we don’t really know how it’s going to play out.
DR. LIBICKI: That’s a good point.
DR. LUMPKIN: Okay, we have time for one more question and then we’re going to take a break. Peggy?
MS. HANDRICH: I’d just like to put on the table that at some point, either in this study or as analysis of this issue moves forward, that the estimate of the impact on CMS must include Medicaid. And I think when you talk about CMS you’re thinking Medicare and states contribute to the cost of Medicaid programs, they run their own fiscal intermediaries and the cost to states to make such a change, I just think needs to be more thoroughly investigated. Likewise I would encourage us at some point to give some consideration to states in general that operate data repositories that have public health data functions and that use health data for the purpose of state health policy development. And I just don’t see that being addressed in any way.
Kind of moving from the philosophical plane that you were just one to a more practical level, I think the bottom line issue from state Medicaid programs perspective will be with more opportunities for refined coding, how can it possibly be anything but inflationary. States have their own coverage policy, they have their own reimbursement policy, they have their own service limitations, but they do in many cases build on Medicare policy and I think that will be from their perspective their biggest worry. And in spite of all that you’ve said I don’t know that you’ve made a compelling case that that kind of inflationary spiral that would be invited by having more codes to bill could somehow be controlled.
DR. LIBICKI: People we’ve talked to expressed opinions on both sides of that one.
MS. BROOKS: I can address your question about the impact, that CMS is doing a study right now, we’re right in the middle of it, and all parts of CMS are being asked to evaluate their costs so that includes the Medicaid part as well as Medicare, so the entire agency is in the middle of this right now. I can’t give you a time when it will be completed but a lot of the work has been done and we hope to have a report out very soon.
DR. LUMPKIN: Let me see if I understand because Medicaid actually is a shared program and the highest level of match, which my former state was the unfortunately recipient of, is 50 percent so actually the bulk of the costs are borne by CMS for Medicaid expenses, and MMMS changes are also reimbursed at an accelerated rate. So I think it’s very appropriate that we need to look at the balance of the costs --
MS. HANDRICH: If I may just react to that, that may be the case but there is state match and in the current circumstances states capacity to heighten state funds available for systems changes or for reimbursement changes is limited and it might even be fair to say it rates from awful to really terrible.
DR. LUMPKIN: I’m not going to ask you which one you’re in. Jennifer.
MS. MADANS: A quick comment. I’m certainly sympathetic to the difficulty of trying to estimate a benefit, for example, for the research component. I think the same is true for the statistical component of this which I don’t think was really dealt with on the benefit side and maybe can’t be dealt with in some kind of quantification but I think these, the records that these codes are imbedded in are a big part of the statistical system that we use to monitor what is going on out there and even if it’s just in verbiage I think that has to be in there even if it’s an unmeasured benefit.
DR. LUMPKIN: Thank you. We’re going to take a 15 minute break and then we’ll resume.
[Brief break.]
DR. LUMPKIN: Let’s get started. Before we get started with the panel Simon Cohn has asked for a point of special privilege to make a motion.
DR. COHN: Yes, well actually, well first of all we actually have Jared Adair present and I just want to take a moment, I think as you all know that she’s been really the shepherd not only of HIPAA over recent years, CMS, but at least from our usually important and maybe more important Consolidated Healthcare Informatics Initiative, she’s going to be retiring at the end of this month, I personally am very sad, it’s been wonderful working with her and I think I speak for certainly the Subcommittee on Standards and Security, I’m sure for the full committee, in terms of her devotion and work to things that we’ve all been trying to make happen for many years. So I want to make sure that everyone has a chance to acknowledge her and thank her and obviously I’m going to be making a motion for us to write a letter to her expressing our gratitude and thanks for her exceptional contribution.
DR. LUMPKIN: It’s been moved that a letter be written and the procedure that I think Simon made in his motion that it would then be reviewed by the Executive Committee and sent as soon as that’s available.
DR. COHN: Hopefully before she retired, which is very quickly.
DR. LUMPKIN: Is there a second?
MR. BLAIR: I will second the motion.
DR. LUMPKIN: Been moved and seconded, is there a discussion on the motion to send a letter and the procedure for the adoption of the language of that letter? All those in favor signify by saying aye.
COMMITTEE: Aye.
DR. LUMPKIN: Those opposed? Great, thank you.
[Applause.]
MS. ADAIR: It is probably inappropriate for me to say, but appropriate has never kind of stopped, I would thank the NCVHS for their support of our initiatives and the tremendous work that they’re doing to help us move these things forward. CHI is, the success of CHI is due in large part to your assistance and thank you, thank each and every one of you personally, also for your support as we moved on those activities. It has been my privilege to work with you and I thank you for your support. Thank you.
DR. LUMPKIN: We’re now going to move on to the panel of reactors and lets see, you’re not sitting in the order in which I have on my list, so if you would please get up. We’ll see if you can introduce yourself and who you represent, we’ll go down the list and then we’ll start off, was there an order that, maybe we’ll just go in the order that it’s printed on my list even though you’re not sitting in that order.
MS. LEON-CHISEN: Nelly Leon-Chisen, American Hospital Association.
MS. PROPHET-BOWMAN: Sue Prophet-Bowman, American Health Information Management Association.
MR. SMITH: Merit Smith from the Robert E. Nolan Company, talking for the Blue Cross Association.
MR. GUSTAFSON: I’m Tom Gustafson talking for the Centers for Medicare and Medicaid Services.
MR. BEBEE: Michael Bebee with the American Medical Association.
MR. STAEHLIN: Marty Staehlin, Price Waterhouse Coopers as a consultant to the American Medical Association.
DR. LUMPKIN: Great, why don’t we start off with Michael, he’s first on my list.
Agenda Item: Reactor Panels to Impact Study and other Studies - Mr. Beebe
MR. BEEBE: That’s the benefit of having the B I suppose. I’m glad to see it spelled correctly, I guess if you appear before the committee enough you get your name spelled correctly.
Thank you for having me back again, I appreciate it. What we’d like to do today, I’d just like to give a brief introduction to Marty Staehlin and he can get through the details of the cost/benefit study that we, rather the cost study that we had commissioned. The study actually, the formal study presents three cost scenarios, cost involved with CPT as the single procedural codeset, ICD-10-PCS as the single procedural codeset, and then also basically a replacement of the status quo scenarios which is replacing ICD-9-CM volume 3 with ICD-10-PCS. This is the study we commissioned. Today we will only be discussing the third scenario, which are costs associated with replacing ICD-9-CM volume 3.
Briefly, a historical perspective on this, the AMA has been involved with ICD-10-PCS for quite a while, it seems a bit of my adult life at this point. The AMA as the developer and maintainer of CPT has been involved in monitoring ICD-10-PCS as part of the technical advisory panel in 1996. We also contracted with Aspen Systems Corporation in 1999 to evaluate the technical attributes of ICD-10-PCS. We presented testimony at the ICD-9 Coordination and Maintenance Committee, and at this committee, the NCVHS Subcommittee on Standards and Security, regarding technical limitations and costs associated with implementation of 10-PCS.
Our testimony was generated from extensive experience as a codeset developer, but also from the work that Aspen did. And the basic kernel of our arguments in past testimony has been that there are serious technical limitations. That there was significant physician and other health care professional services that are not included in 10-PCS. That the terminology does not conform to traditionally named body systems or procedure descriptions that are commonly used by physicians that they’re familiar with. That there are multiple codes that are needed to fully describe commonly performed procedures and this has the potential to be confusing. And that the procedures cannot be fully understood without reference to diagnosis. We also went through implementation issues such as education issues, system issues, reporting and payment issues, and maintenance issues, all of which relate to cost.
As I said, the current Price Waterhouse Coopers study in ICD-10-PCS was intended to build on our previous study, 1989, when I was in college, 1989 study, which looked at the single procedure costs of theoretical new single procedure codesets. It also is building on other previous studies in this area, such as the 2000 WEDI white paper and the 2002 GAO report. We also wanted Price Waterhouse Coopers to consider the current regulatory context, i.e., HIPAA, the existing codesets in this area, ICD-9 and CPT, and also the health care payment systems in which codesets are used in their analysis of the three scenarios.
The Price Waterhouse Coopers study only looks at implementation costs, it doesn’t look at technical limitations, certainly these limitations effect educational and systems costs. Although the costs of a single procedure codeset were studied by PWC the AMA believes that the committee correctly limited the Rand study to examine costs associated just with the replacement of ICD-9 and as such that’s all we will be discussing here today.
Before turning this over to PWC I just want to direct the committee’s attention to a letter that’s been signed by many national medical specialty societies, and this letter supports the idea of replacing ICD-9 with ICD-10-PCS only, following along the rationale that ICD-10-PCS was developed as an inpatient hospital coding system only and is not suitable for physician services.
So with that, I will turn it over to Marty Staehlin.
Agenda Item: Reactor Panels to Impact Study and Other Studies - Mr. Staehlin
MR. STAEHLIN: Thank you, Michael. Somebody earlier said that information is expanding exponentially, unfortunately my ability to retain information is not expanding at the same rate, so this is a cooperative study, it was done by a number of people and I am here representing the study to give an overview of it and to answer questions as might arise. I have seen copies around so I think most of you if you haven’t read it you at least have a copy of the study and so the alternative that I’m going to be addressing today is basically in section four and it starts on page 38 and talks about the alternative to a single procedure coding system with option of ICD-10-PC