Hubert H. Humphrey Building
Washington, D.C.
The National Committee on Vital and Health Statistics was convened on September 23-24, 2003 at the Hubert H. Humphrey Building in Washington, D.C. The meeting was open to the public. Present:
ACTIONS
The Secretary is urging all agencies to find ways to implement information technology across their programs and to promote interoperable health information technology across the public and private sectors. HHS has established a Council on the Application of Health Information Technology (CAHIT) to facilitate coordination and collaboration. With respect to the Federal Information Quality Guidelines, HHS is preparing comments on proposed OMB guidelines calling for agency regulations to be based on peer reviews of the most important scientific technical information. The Department is compiling for publication the recommendations for a consensus agenda that were developed at a July 2003 meeting on action steps needed for the NHII, as well as planning follow-up activities.
The report on the National Academy of Sciences study on race and ethnicity will be released in December. A report on a special workshop for health plans will be published in September and is expected to dovetail with recommendations of the NCVHS Subcommittee on Populations. Mr. Scanlon also reported on budget planning and the HHS Gateway.
Ms. Friedman discussed the CMS guidance clarifying that covered entities that make a good faith effort to comply may implement contingencies to maintain operations and cash flow. In the course of this meeting, CMS announced that Medicare would invoke its contingency plan.
Ms. Sanchez reported on the activities of OCR in handling 1,937 Privacy Rule complaints. One third of the complaints have been resolved and closed, through voluntary compliance and technical assistance. No fines have been imposed, but a few cases have been referred to the Department of Justice
During the discussion period, NCVHS members raised questions about continued use of the DSM IV for mental disorders and about the HIPAA contingency plan.
OVERVIEW OF ICD-10-CM AND ICD-10-PCS
This series of presentations and discussions continued the Committee’s examination of whether there should be a transition from ICD-9-CM to ICD-10-CM and ICD-10-PCS, and if so, how. The goal, if possible, is for the Committee to make a decision at the November 2003 meeting.
Speaking on behalf of Dr. Sondik, Dr. Madans explained that the U.S. has used the clinical modification developed for ICD-9 morbidity coding since 1979. ICD-10 was adopted for mortality coding in 1999, representing a much-needed major revision that is needed at least as much for morbidity coding. ICD-9-CM is rapidly losing its capacity to incorporate specificity for newly identified disease entities and other advances, and it is becoming much harder to modify. Without the CM, ICD-10 does not provide the information needed for morbidity coding. The classification system is the key component underlying all the information systems that contribute to the creation, review and revision of health care policy. To provide the data needed to improve patient safety, health care research and health statistics and to preserve the status of the U.S. as an international leader, NCHS believes it is time to implement the revised morbidity classification.
Mr. Gustafson stated that ICD-9-CM offers inadequate descriptions of procedures, is out of date and difficult to update and expand, and is complicated. The precise specification of procedures is needed for accurate payment. Adequate procedure codes are also necessary for evaluating the efficacy of procedures, measuring outcomes, preventing fraud and abuse, and avoiding labeling people as “bad apples” when they are not. In addition, accurate procedure codes make it possible to understand technological diffusion, identify emerging technologies, and facilitate access to quality care. He described the attributes of ICD-10-PCS, noting that it was developed based on recommendations made by NCVHS in 1993.
In the discussion period, Ms. Handrich asked what discussion has taken place regarding Medicaid. Mr. Gustafson said the Committee would be provided with detail on those discussions.
Martin Libicki, Rand Corporation Science and Technology Policy Institute
The Rand study asked three questions, formulated by NCVHS:
The results are presented as ranges for both costs and benefits, reflecting the uncertainties of estimation. The two major cost categories are training and systems changes/upgrades. In sum, the study estimates the cost of the conversion for training and systems changes at between $425 million and $1.15 billion in one-time costs. In addition, it estimates a cost of between $5 and $40 million a year in lost productivity. The total of estimated benefits is between $700 million and $7.7 billion over ten years. In addition, potential benefits were identified in several categories for which no financial estimates were made. The additional estimated cost of switching sequentially (question 3 above) was estimated at an average of some $20 million per cost category (e.g., training physicians). The report states that the investigators see “no offsetting benefits from switching sequentially,” and no one who testified or was interviewed expressed a preference for sequential implementation.
In the discussion period, NCVHS members expressed interest in whether the benefits would accrue to those who will bear the major costs, the potential disruptions to payment systems, and the impact on states. Dr. Madans commented on the potential benefits to statistical systems.
Mr. Beebe introduced the presentation by Martin Staehlin of a cost/benefit study, commissioned by the AMA and conducted by Price Waterhouse Coopers (PWC). Among other topics, the report addresses the implementation costs of replacing ICD-9 CM Volume 3 with ICD-10-PCS. Mr. Beebe called the Committee’s attention to a letter signed by national medical specialty societies supporting the idea of replacing ICD-9 with ICD-10-PCS for hospital inpatient services only.
The PWC study calculated the one-time implementation costs for the aforementioned conversion at $178 million. Taking into account a multiplier effect related to further training and education, contract negotiation, and other factors led to an estimate of $1.1 billion for the total cost of the transition.
The Blue Cross/Blue Shield Association asked the Health Care Practice section of the Robert E. Nolan Company to look at the costs, benefits and operational implications of changing from ICD-9-CM to ICD-10-CM and ICD-10-PCS. The study found a low level of awareness among the people who would be involved in making the change, and a low level of organizational and technical readiness for it. It also concluded that the projected benefits represent “leaps of faith,” and it identified implementation issues and possible unanticipated consequences unless great care is taken in implementation. The study also provided its own cost estimates for the transition to ICD-10.
The two panelists presented the findings of a field-testing project of ICD-10-CM conducted jointly by their organizations. The study assessed the functionality and utility of applying ICD-10-CM to medical records in a variety of health care settings. Regarding coding times, on average, the time required to code a record with ICD-10 was roughly double what it was for ICD-9 — 12.14 minutes compared to 6.37. There was no difference in the coding times for 58.6 percent of the record sample. Nearly 92 percent of the need for additional time was attributed to problems with the index file format, which are expected to be resolved before implementation.
The project identified several opportunities for system improvement, which Ms. Prophet-Bowman outlined. She also offered comments about the merits and potential benefits of ICD-10-CM, noting that with improved coding tools and training, ICD-10-CM might eventually require even less time to use than ICD-9.
The panelists also presented the findings of a follow-up survey on the level of education and training that professional credentialed coders would need to implement ICD-10-CM. More than 83 percent of the participants supported migration to ICD-10-CM, and favored doing so within three years.
The CMS study will focus on administrative costs, not benefit costs, because the Center does not expect substantial change in benefit costs as a result of this coding change. The study will encompass state as well as federal costs.
Mr. Gustafson focused on two issues for CMS. The first is the pressure to develop more precise payment cells for DRGs, in order to pay more sensibly for procedures and improve the Medicare program. The second issue is that Congress expects CMS to pay appropriately for the use of new technology, and this may not be possible in the future if ICD-9-CM is not replaced. Expandability is a critical problem, and one for which ICD-10 offers a substantial advantage.
Ms. Leon-Chisen’s comments reflect the advice of AHA’s member advisory panel on ICD-10 implementation. She stressed that migration to a newer version of clinical codesets is essential for hospitals. AHA agrees with the Rand study’s conclusion that the benefits associated with migration will exceed the costs, but it believes that the study overstates the costs of training and productivity losses for hospital coders, who already receive training on a regular basis. She referred to the AHA/AHIMA study described above. To the classes of benefits identified in the Rand study, AHA adds the ability to improve community health based on coding consistency between mortality data and morbidity data. Finally, Ms. Leon-Chisen offered AHA’s recommendations regarding transition dates and the timeline.
Ms. Kloss said that AHIMA supports all of AHA’s recommendations. She pointed out the cost of doing nothing, both in lost productivity and in missed opportunities for innovation, and she called attention to the opportunity to change from the status quo to a paradigm of information management. She urged the Committee to join this issue to its work on the NHII, the EHR and clinical vocabularies, and to move quickly to a decision.
Mr. Desmarais said HIAA looked forward, and would respond, to the forthcoming CMS report, especially with respect to DRG revisions. He suggested that the Rand study should “do the math completely” in comparing total costs and total benefits, following the 10-year trajectory in each case. HIAA’s overall reaction is that the Rand study underestimates costs. Regarding benefits, he expressed a desire for “real world examples” of the applications of the new codes. Above all, HIAA members stress the need for adequate time for everyone in the system to prepare and adapt, if a change is recommended
Dr. Cohn said the Subcommittee on Standards and Security would hold a hearing the last week of October on ICD-10, and it hopes to bring a letter on this topic to the full Committee’s November meeting.
(See Actions.)
On day one of this meeting, Dr. Mays presented two draft letters for review, one on the need for targeted surveys and the other concerning the collection of race and ethnicity data by health plans. Following extensive discussion and editing by Committee members and others, the Subcommittee revised the letters and the revised versions were approved on day two of this meeting.
Dr. Cohn presented a letter drafted by the Standards Subcommittee, commenting on a series of CHI Initiative recommendations. Following the Committee’s discussion, the Subcommittee revised the letter, and it was approved on day two of this meeting. The Committee also discussed the protocol for responding to future CHI recommendations, as the letter is expected to be the first of a series on that topic.
To aid the Workgroup in preparing its final report, Mr. Hungate invited discussion of the 21 provisional Workgroup recommendations, and in particular any that are problematic and need further discussion and refinement. He stressed the importance of having input and participation from all NCVHS subcommittees and workgroups because of the cross-cutting nature of quality measurement issues.
Members talked at some length, both about the content of the recommendations and about a process for moving ahead with the most difficult ones, especially those with implications for HIPAA transactions. The Executive Subcommittee was designated as the planner, coordinator and possible forum for this process. To initiate an effort to engage with DSMOs, an NCVHS representative may meet with X-12 to discuss the Workgroup’s findings on barriers to measuring quality and to seek that group’s views on possible solutions.
Members expressed interest in framing the forthcoming discussions in terms of paying for performance, quality and safety. It was suggested that the Committee consider what aspects of improved quality can be implemented within the current health system structure and, alternatively, to what extent the Committee needs to help embed quality in a new reimbursement structure or paradigm.
The Committee agreed to survey individual members on their views on each recommendation, with respect to 1) whether it should be recommended; 2) if so, whether the change is sought in the near term or long-term; 3) relative importance; 4) doability; and 5) suggestions on how to deal with the issue. Mr. Hungate and staff were asked to design the survey, in consultation with the Executive Subcommittee, with the goal of getting feedback prior to the November full Committee meeting.
(Please see the final few pages of the minutes for the brief summaries of these items. Actions are reported above.)
Note: Please consult meeting transcript and supporting documents on NCVHS Web site for further details of all presentations — www.ncvhs.hhs.gov
—DAY ONE—
Dr. Lumpkin called the meeting to order and asked all present to introduce themselves.
Regarding conflicts of interest and recusals, Dr. Cohn stated that he is on the CPT editorial panel and would recuse himself from any discussion of CPT. Dr. Harding stated that as a past officer of the American Psychiatric Association, he would recuse himself from any vote involving DSM.
Data Council: Jim Scanlon, ASPE
Progress continues with regard to the NHII and health information technology and standards. They are systematically included in HHS planning and program operation with strong support from the Secretary, who is urging all agencies to find ways to implement information technology across their programs and to promote interoperable health information technology across the public and private sectors. To this end, HHS has established a Council on the Application of Health Information Technology (CAHIT), an internal group co-chaired by ASPE and the administrator of AHRQ that will facilitate coordination and collaboration in HHS and with industry and the public health and research communities. Dr. Lumpkin participates in the meetings on behalf of NCVHS.
Mr. Scanlon then reported on recent activities with respect to Shelby I and II Federal Information Quality Guidelines. OMB has released new proposed guidelines to ensure that agency regulations are based on peer reviews of the most important scientific technical information. The OMB definitions are somewhat ambiguous as to the scope, but HHS is interpreting the guidelines as applying just to regulatory actions. OMB is accepting public comments through December 15. HHS is looking at the guidelines and pulling together comments.
HHS held a successful and well attended national meeting in July to develop a consensus agenda on action steps needed for the NHII. It is now compiling for publication the recommendations developed in eight tracks (e.g., research, public health, privacy). The Department is also planning follow-up activities, including meetings with participant groups such as providers and payers. The NHII Web site maintained by the Department will have an inventory of all HHS activities relating to the NHII.
Regarding FY2004 budget and appropriations, Mr. Scanlon reported that things are headed in a positive direction. He described some proposed investments related to the NHII initiative. The Department has begun work on the FY2005 budget, which will continue the NHII investment.
The panel for the National Academy of Sciences study on race and ethnicity that began more than a year ago has met three times. Recently, at the Department’s request, the Academy held a workshop for health plans and providers to get a sense of private sector thinking on policies and practices for collecting race and ethnicity data. That report will be released around the end of September and is expected to dovetail with recommendations of the NCVHS Subcommittee on Populations. The full Academy report is expected in December.
The HHS gateway to the Web has been expanded and now links to 2500 sites. The Data Council is contracting for a study as part of an initiative to standardize geocoding and geospatial information in HHS data.
HIPAA: Data Standards: Maria Friedman, CMS
CMS issued guidance enforcement for HIPAA transactions and codeset provisions, clarifying that covered entities that make a good faith effort to comply may implement contingencies to maintain operations and cash flow. CMS announced in the course of this meeting that Medicare will invoke its contingency plan.
Discussion
Dr. Harding mentioned the ten percent of electronic code transactions under HIPAA in the area of mental disorders and asked if mental health professionals will be allowed to continue to use DSM criteria or will be forced to return to ICD-9-CM descriptors. Ms. Friedman said this question is under discussion and review by the Office of General Counsel. She predicted that there will be FAQs on the subject soon. Ms. Greenberg predicted a mutually agreeable solution.
Dr. Zubeldia asked about CMS’s plans regarding the ASCA law requiring all providers to file claims electronically as of October 16. This led to a discussion in which Ms. Berek informed the group that CMS can waive for providers unable to bill electronically. She also announced that Mr. Scully had announced that morning that CMS would be implementing its contingency plan under HIPAA. She promised to get back to the Committee on the application of the contingency to ASCA. Details will be on the CMS website.
Between April 14 and September 22, the HHS Office of Civil Rights received 1,937 Privacy Rule complaints. Regional offices investigate complaints and take the lead in handling them. One third of the complaints have been resolved and closed, through voluntary compliance and technical assistance. No fines have been imposed, but a few cases have been referred to the Department of Justice. Some complaints did not raise a privacy issue or concern a violated rule. Ms. Sanchez gave a breakdown on the kinds of issues arising in the complaints (see transcript).
OCR’s outreach activities include 100 presentations this year, telephone audio conferences, and the Privacy Rule hotline (which has received 14,000 calls since April 1, with the volume diminishing recently). The Web site’s FAQs, which are routinely expanded, have been accessed 1.2 million times. There are links to new NIH guides for researchers and the CDC guide on public health.
Mr. Houston suggested analyzing the complaints by type of covered entity and type of provider, to give insight into who is having problems with compliance. He praised the FAQs and suggested that the answers should point out that more stringent state laws may apply.
Dr. Lumpkin explained that the next series of sessions would continue the Committee’s examination of whether there should be a transition from ICD-9-CM to ICD-10-CM and ICD-10-PCS, and if so, how.
Dr. Cohn noted that the Committee had been studying code set and classification issues for many years. It decided not to recommend a transition to ICD-10 as part of its initial recommendations on administrative and financial transactions, recognizing the need for a thorough impact study before deciding what to recommend to the Department. The goal now, if possible, is to make that decision at the November 2003 meeting.
Jennifer Madans, NCHS
Dr. Madans, the NCHS Associate Director for Science, spoke on behalf of Director Ed Sondik, who was unavoidably called to a meeting. She stressed the Center’s strong interest in classifications in general and in ICD in particular, calling it a key component of the Center’s mission of promoting data quality for effective health policy, health statistics and health care. NCHS houses one of the WHO collaborating centers that implement the family of international classifications that include the ICD. The classification’s international character contributes to research and to international comparisons.
Classification systems have to change as knowledge advances. The U.S. has used the clinical modification (CM) it developed for ICD-9 for morbidity coding since 1979. It adopted ICD-10 for mortality coding in 1999, representing a major revision that was sorely needed and that is equally needed for morbidity coding, if not more so. ICD-9-CM is rapidly losing its capacity to incorporate specificity for newly identified disease entities and other advances, and it is becoming much harder to modify the existing system. ICD-10 (unmodified) does not provide the information needed for morbidity coding.
The classification system is the key component underlying all the information systems that contribute to the creation, review and revision of health care policy based on accurate and timely health care data. To provide the data necessary for improving patient safety, health care research and health statistics and preserving the status of the U.S. as an international leader, NCHS believes it is time to implement the revised morbidity classification.
Tom Gustafson, CMS
Mr. Gustafson is Deputy Director of the CMS Center for Medicare Management, which runs the fee-for-service component of Medicare, with 85 percent of its caseload. Coding defines the products and services CMS pays for. His remarks focused on payment and business factors in a shift to ICD-10 classification systems.
ICD-9-CM has inadequate descriptions of procedures, is out of date and difficult to update and expand, and is complicated. The precise specification of procedures is needed for accurate payment. Adequate procedure codes are also necessary for evaluating the efficacy of procedures, measuring outcomes, preventing fraud and abuse, and avoiding labeling people as “bad apples” when they are not. In addition, accurate procedure codes make it possible to understand technological diffusion, identify emerging technologies, and facilitate access to quality care. Dr. Gustafson noted that the decision was made to develop a “new and refreshed coding system” that would provide a more precise linking of technology and procedure and allow improvement of underlying procedure codes.
Development of ICD-10-PCS, the procedure code piece of ICD-10 in the United States, was initiated by NCVHS in the early 1990s with an articulation of principles, funded by HCFA, and carried out by 3M. It has been tested and is ready for consideration as the replacement for ICD-9-CM Volume 3 as the standard codeset for inpatient use. Recommendations by NCVHS in 1993 called for a procedure codeset that was comprehensive, non-overlapping, expandable, hierarchical, multi-axial, used standardized terminology, and separated diagnostic information from the procedure descriptions. ICD-10-PCS has these attributes.
The codeset is based on a seven-character alphanumeric code, each of which functions in a specified and consistent way. There are 16 sections. Because of the hierarchical, axial structure, the codeset is simpler to use and understand than its predecessor. Trials indicate that after relatively minimal training, experienced coders can code material more quickly and accurately using ICD-10-PCS than with ICD-9-CM, Volume 3. The tools are already available to make it work, including a complete crosswalk between the two classifications, definitions manuals, user manuals, and training materials.
Discussion
Ms. Handrich asked what discussion has taken place regarding Medicaid. Mr. Gustafson said the Committee would be provided with detail on those discussions.
Martin Libicki, Rand Corporation Science and Technology Policy Institute
Dr. Libicki stressed that as the Rand report is still under review, the views expressed at this meeting are his alone. The study asked three questions, formulated by NCVHS:
It drew on three sources: prior testimony to NCVHS, literature review, and interviews with 90 people. The results are presented as ranges for both costs and benefits, reflecting the uncertainties of estimation. In his presentation, Dr. Libicki discussed the reasoning behind each conclusion outlined below. (See transcript and supporting materials for details.)
The two major cost categories are training and systems changes/upgrades. The groups to be trained include hospital coders, other coders, physicians, and code users. The study also estimated the cost of initial and long-term losses of productivity among coders and physicians. Systems change costs apply to providers, software vendors, payers, and CMS. Having no hard data for CMS, as the Center is still conducting its own analysis, Rand used a wide-ranging estimate as a placeholder for CMS costs. Dr. Libicki noted that neither the American Hospital Association nor the Federation of American Hospitals expressed concern about system change costs; they were more focused on training costs.
In sum, the study estimates the cost of the conversion for training and systems changes at between $425 million and $1.15 billion in one-time costs. In addition, it estimates a cost of between $5 and $40 million a year in lost productivity. The costs are distributed as follows: hospitals: 35%; outpatient facilities: 25%; payers, including CMS: 40%. The costs are roughly proportional to the number of patients or members.
Regarding the potential benefits of ICD-10, the study identified the following classes of benefits, created scenarios, and developed ten-year estimates for each:
The total of these estimated benefits is between $700 million and $7.7 billion over ten years. The following potential benefits of ICD-10 were also noted, but not quantified: better understanding of health care and its relation to outcomes, enhanced ability to evaluate providers, and more timely intervention for emerging diseases and bioterrorism.
Dr. Libicki explained that the benefits derive mostly from the additional detail available in ICD-10. The analysis assumes the codes will be used correctly, which is not always the case with ICD-9-CM. In conclusion, the study states that it is likely that switching to both ICD-10-CM and ICD-10-PCS has the potential to generate more benefits than costs.
To answer the third NCVHS question, the additional cost of a sequential change was estimated for each cost category. Depending on the cost category, the additional cost ranges from $0 to $50 million and averages about $20 million. The report states that the investigators see “no offsetting benefits from switching sequentially.” Dr. Libicki enumerated the reasons for switching both at once, which relate to training, testing, and DRGs. He added that no one who testified or was interviewed expressed a preference for sequential implementation.
Finally, Dr. Libicki observed that nearly every industry has found that a program of continuous improvement begins with measuring input and output. In the health care sector, which has been “laggard in this approach,” the use of codes is central to the measurement process. Better codes could make more accurate measurement possible and thereby lead to better outcomes.
Discussion
Dr. Zubeldia expressed concern that the study does not look very far into the costs for payers. He also asked how and when CMS costs and benefits would be assessed, and he raised questions about potential disruptions to the payment system.
Dr. Libicki said the CMS cost numbers were expected soon. The Rand study did not identify any benefits for CMS. He observed, however, that on a philosophical level, getting better information can lead to better decisions. If payers get better data from providers, they will have less work to do to adjudicate claims, which will save them and the entire system money. In response to Dr. Zubeldia’s final question, he pointed out that negotiations take place in any case over a multi-year period, with or without a change in ICD.
In response to another question, he said he had not parsed out the distribution of benefits as was done with respect to costs, adding that it would be statistically impossible to find the effect of several hundred million dollars worth of costs in a $1.4 trillion industry in which there is considerable variation from year to year. He speculated that making the medical reimbursement system more rational could move it toward optimality. Dr. Cohn said it would be useful to know whether the benefits would accrue to the same groups that would bear the costs.
Ms. Handrich stressed that the impact of the proposed change on Medicaid programs, and in particular on states, must be thoroughly investigated. She expressed concern that the coding refinements would be inflationary for states. Ms. Brooks said CMS is currently doing a study of the potential impact, looking at all parts of CMS including Medicaid. The report is expected soon.
Dr. Madans observed that while it is difficult to estimate a benefit for the statistical component, just as it is for the research component, it can be assumed that there would be benefits in this area from the proposed shift.
Michael Beebe, American Medical Association
Mr. Beebe introduced the presentation by Martin Staehlin of a cost/benefit study commissioned by the AMA and conducted by Price Waterhouse Coopers (PWC). The study presents three cost scenarios: costs involved with CPT as the single procedural codeset, ICD-10-PCS as the single procedural codeset, and replacement of ICD-9 CM Volume 3 with ICD-10-PCS. Today’s presentation focuses only on the third scenario. The AMA monitors and has presented testimony about ICD-10-PCS to the effect that it has serious technical limitations, which Mr. Beebe detailed. The current study, which looks only at implementation costs and not at technical limitations, builds on a 1989 study of the costs of a single procedure codeset and on other studies, placing them in the current regulatory context.
Mr. Beebe called the Committee’s attention to a letter signed by national medical specialty societies supporting the idea of replacing ICD-9-CM, Volume 3 with ICD-10-PCS for hospital inpatient services only.
Martin Staehlin, Price Waterhouse Coopers
The PWC study was conducted through literature search and interviews. The study predicted that a shift to ICD-10-PCS for hospital inpatient coding would have advantages in the areas of flexibility and patient profiling. It did not look at the benefits identified in the Rand study. It foresees disadvantages from the shift to ICD-10-PCS in the areas of training, continuity of data, systems costs, and systems testing.
Mr. Staehlin noted that the PWC study integrates and follows from an August 2002 GAO study, which talks about the practical constraints to a single procedure codeset. The PWC study calculated the one-time implementation costs for conversion to ICD-10-PCS for inpatient hospital coding at $178 million.
In trying to develop a full cost impact, PWC found that people in the industry associate the ICD conversion with the costs of other transitions, including Y2K and HIPAA. One section of its report addresses the multiplier effect related to further training and education, contract negotiation, and other factors. Taking these factors into account led to an estimate of $1.1 billion for the cost of the transition to ICD-10-PCS, representing the high end of the range of estimated costs.
Merit Smith, Robert E. Nolan Company
The Blue Cross/Blue Shield Association asked the Health Care Practice section of The Robert E. Nolan Company to look at the costs, benefits and operational implications of changing from ICD-9-CM to ICD-10-CM and ICD-10-PCS. Mr. Smith described their methodology, which was similar to that of the other studies and also looked at experience in other countries.
The study found a low level of awareness among the people who would be involved in making the change, and a low level of organizational and technical readiness for it. At present, “they’re consumed with the transactions and codeset change and getting through privacy.” Mr. Smith supported Mr. Staehlin’s statement that people associate the ICD change with Y2K and HIPAA.
The study also found that the putative benefits of ICD-10 were “quite speculative,” and it was difficult to see how making the changes would lead to the anticipated benefits. The Nolan Company study came to the conclusion that the projected benefits represent “leaps of faith.” It also identified implementation issues and possible unanticipated consequences unless great care is taken in implementation. These include disruptions to provider reimbursement, clinical data, existing clinical knowledge, and (in the short term) fraud detection. The study also provided detailed cost estimates for a transition to ICD-10.
Nelly Leon-Chisen, American Hospital Association and Sue Prophet-Bowman, American Health Information Management Association
The two panelists presented the findings of a field-testing project of ICD-10-CM conducted jointly by their organizations. The study assessed the functionality and utility of applying ICD-10-CM to medical records in a variety of health care settings. The study used a descriptive survey research model, involving 169 participants from all regions of the U.S. who coded a total of 6,177 records. It also assessed the level of education and training that professional credentialed coders would need to implement ICD-10-CM, using a follow-up survey. Ms. Leon-Chisen noted that the field test was conducted without benefit of an electronic alphabetic index and other standard resources; these will be available to coders if and when ICD-10-CM is implemented. She described the methodology of the study, which included validation of the accuracy of the coding. Information was collected about the number of years of experience of the coders, and findings were stratified with respect to experience. The health informatics staff at Ohio State University collected, tabulated and reported the results of the field test.
Regarding coding times, on average, the time required to code a record with ICD-10-CM was roughly double what it was for ICD-9-CM — 12.14 minutes compared to 6.37. There was no difference in the coding times for 58.6 percent of the record sample. Nearly 92 percent of the need for additional time was attributed to problems with the index file format. Regarding validation, roughly 80 percent of the participants’ and validators’ code assignments matched, with the other 20 percent attributed variously to index problems, unfamiliarity with ICD-10, or erroneous assumptions. None of the participants coded enough records (55 maximum) to become proficient in using ICD-10-CM.
The project identified several opportunities for system improvement, which Ms. Prophet-Bowman outlined. She also offered comments about the merits and potential benefits of ICD-10-CM, noting that with improved coding tools and training, ICD-10-CM might eventually require even less time to use than ICD-9-CM. She speculated that because of its design, this project is likely to have identified most of the common problems coding professionals would face with ICD-10-CM.
The follow-up survey asked coders to estimate the amount of training they would need to shift to ICD-10-CM. 60 percent said they would need 16 hours or less. Nearly 72 percent said the clinical descriptions in ICD-10-CM were better than in ICD-9, and more than 83 percent of the participants supported migration to ICD-10-CM, and favored doing so within three years.
Tom Gustafson, CMS
Mr. Gustafson reiterated that CMS is conducting a detailed study of the costs of the proposed change, including the state Medicaid dimension, and will make its findings available as soon as possible. He noted that the study will cover the impact on coverage policies, adding that coverage policy is constantly under review. The CMS study will focus on administrative costs, not benefit costs, because the Center does not expect any substantial change in benefit costs as a result of this coding change. He reserved the possibility of reporting to the Subcommittee on Standards and Security on the business case for a change to ICD-10 in October.
After commenting on the potential benefits of the change in the areas of disease management and clinical research, he called attention to two significant issues. The first is the need for greater precision in the payment rates for diagnostic reimbursement groups (DRGs), which CMS is “under regular pressure to improve.” Severity-adjusted DRGs, or more precise payment cells, would contribute to the ability to pay more sensibly for procedures and would improve the Medicare program. Second, on the diffusion of new technology, Mr. Gustafson said that while coding limitations have not prevented an appropriate payment policy with respect to new technology to date, that is a possibility in the future if ICD-9-CM is not replaced. The rapid rate of technological change in the medical sector necessitates rapid adaptations of codes in order to pay for new procedures. Congress expects this to happen, so “we have to do something.” Whether or not there is a shift to ICD-10, there will be costs in the future.
In summary, he said that expandability is a critical problem from the perspective of CMS and one for which ICD-10 offers a substantial advantage.
Ms. Leon-Chisen, American Hospital Association
Ms. Leon-Chisen said her comments would reflect the advice of AHA’s member advisory panel on ICD-10 implementation, which is composed of hospital health information management professionals, corporate compliance officers, clinical and financial data leaders, and others. She stressed that migration to a newer version of clinical codesets is essential for hospitals. AHA agrees with the Rand study’s conclusion that the benefits associated with migration will exceed the costs, but it believes that the study overstates the costs of training and productivity losses for hospital coders, who already receive training on a regular basis. Hospitals have already initiated training to bolster coders’ knowledge of anatomy and physiology, in anticipation of the shift to ICD-10. Regarding possible productivity losses, the AHA/AHIMA field testing found no increase in coding time between the two versions of ICD for more than half the records. Its findings actually suggest that the coding productivity under ICD-10-CM may prove to be greater than under ICD-9-CM.
The AHA agrees with the classes of benefits identified in the Rand study, and adds the ability to improve community health based on coding consistency between mortality data and morbidity data. Furthermore, comparisons based on international health data will be more meaningful.
Ms. Leon-Chisen offered AHA’s recommendations regarding transition dates and the timeline. It supports a transition to ICD-10-CM and ICD-10-PCS for hospital inpatient reporting at the same time, on a date certain, preferably October 1; two years from the publication of a final rule; preceded by ample testing; and facilitated by a crosswalk.
Linda Kloss, AHIMA
Ms. Kloss said that AHIMA supports all of AHA’s recommendations. She stressed that there is a real cost of doing nothing, because the productivity of those managing the data is deteriorating and that deterioration will compound itself. “We are paying the price for having delayed too long.” The price is both in the loss of productivity and in the loss of innovation.
She called attention to the opportunity the proposed shift affords to change to a paradigm of information management rather than continue the status quo, and she urged the Committee to join this issue to its work on the NHII, the EHR and clinical vocabularies, where extraordinary progress has been made. It would be a mistake not to view the codeset decision in that context. Work such as mapping SNOMED to ICD-10 is on hold because of the delays in a decision. Finally, she congratulated the Subcommittee for getting the impact study done and urged the Committee to move quickly to a decision.
Henry Desmarais, Health Insurance Association of America
Mr. Demarais referred the Committee to HIAA’s written statement. He said his organization was disappointed not to see more in the Rand report on the Medicare and Medicaid implications, especially related to revising the DRGs. It plans to respond to the forthcoming CMS report on the subject.
He called attention to the ”startling” fact that Canada and Australia ended up with much fewer procedure codes than in ICD-10-PCS, and invited people to reflect on that. He suggested that the Rand study should “do the math completely” in comparing total costs and total benefits, following the 10-year trajectory in each case. HIAA’s overall reaction is that the cost estimates are understated. He cited the costs of software adaptation for payers and of payment policy adjustments. Regarding benefits, he expressed a desire for “real world examples” of the applications of the new codes —for example, how the new diabetes codes would affect clinical decision making. Finally, Mr. Demarais said HIAA members are stressing above all the need for adequate time for everyone in the system to prepare and adapt, if a change is recommended. In response to a follow-up question, he said HIAA has not taken a formal position on adopting CM and PCS; for now, it wants to emphasize the importance of timing.
Dr. Zubeldia asked for comments on the notion of a staged implementation. Mr. Gustafson, Ms. Leon-Chisen, and Ms. Clausen called the idea impractical. Mr. Desmarais agreed and added that a key concern for payers is how long they need to maintain dual systems. He pointed out that clearinghouses are another stakeholder. Mr. Gustafson stressed that sufficient lead time and testing are essential. Dr. Zubeldia proposed “permissiveness” during the transition, to allow for readjustments. Mr. Desmarais pointed out that while payers have to receive the new information, they do not have to immediately adjust all their payment and coverage policies.
Ms. Handrich urged that CMS seek the input of states in its analysis of the impact of the proposed change. Mr. Gustafson assured her that this would be the case.
Summarizing the Committee’s next steps, Dr. Cohn said the Subcommittee on Standards and Security would hold a hearing the last week of October on ICD-10, and it hopes to bring a letter to the full Committee’s November meeting. Dr. Lumpkin stressed the importance of coming to a decision then, if possible.
Dr. Cohn made a presentation to Jared Adair, who is retiring. He acknowledged her exceptional contributions with respect to the Consolidated Healthcare Informatics (CHI) Initiative, and thanked her on behalf of the Committee for her devotion and leadership. A motion was unanimously passed that a letter of appreciation be written, reviewed and approved by the Executive Subcommittee, and sent to Ms. Adair.
Dr. Lumpkin observed that the well-documented health disparities in the U.S. are receiving attention from both the Workgroup on Quality and the Subcommittee on Populations, reflecting the fact that health disparities are both a population health issue and a quality of care issue.
As background, Dr. Mays noted that recent and forthcoming reports, including recommendations being developed by the NCVHS Quality Workgroup, highlight the need to collect additional race and ethnicity data to provide adequate information on the quality of care in the U.S. She presented for the Committee’s review two letters addressing this concern.
A letter on the need for targeted surveys, is the result of a consistent finding in Subcommittee hearings about the inadequacy of survey data, especially on small groups or ones in remote areas, but also sometimes even on African Americans or Latinos. The other letter, concerning the collection of race and ethnicity data by health plans, is a response to findings in hearings held by the Workgroup on Quality. Dr. Mays read the two letters and received comments from Committee members and others. The letters are to be revised by the Subcommittee and presented for final review and approval on day two of this meeting.
Dr. Cohn presented a letter drafted by the Standards Subcommittee, commenting on a series of CHI Initiative recommendations. He read the letter and accepted comments from Committee members. The group also discussed the protocol for responding to future CHI recommendations. The revised version of the letter will be presented for approval on day two of this meeting.
Mr. Hungate noted that the NCVHS reports on the NHII and 21st century health statistics take a much longer view than the evolving QWG report. While the latter addresses a wide range of issues and barriers to measuring quality, most of the proposed solutions are framed in terms of today’s systems. To aid the Workgroup in preparing its final report, he asked for comments on the 21 provisional Workgroup recommendations, which along with the full draft report were distributed to members prior to the meeting. In particular, he invited comments on any recommendations that are problematic and need further discussion and refinement.
The recommendations were developed by the QWG in response to the findings from 16 panel presentations to the full Committee over a five year period, on the barriers to quality measurement. They concern data content, data standardization, functionality requirements for electronic health records, and data linkage.
A major theme of the ensuing discussion concerned the best process for moving ahead with recommendations that are problematic and/or that cut across subcommittees. This attention to process was necessitated not only by the large number of recommendations and the shortness of meeting time, but also by the fact that quality cuts across all subcommittee and workgroup domains and broad participation is needed to shape the report.
It was agreed that uncontroversial recommendations would be referred to the subcommittee with the best expertise to fine-tune them. For issues that require debate and discussion, Committee time will be scheduled. (A prime example of the latter was what mechanism to recommend for collecting and reporting race and ethnicity data, a question that received considerable discussion at the meeting.) The Executive Subcommittee is to serve as a forum for planning the process and possibly conducting necessary discussions.
Some members expressed a desire to further investigate recommendations related to HIPAA transactions, seeking a way to achieve the desired ends in a manner compatible with the established process for modifying HIPAA transactions.
Regarding the scope and intent of the report, Ms. Coltin (former QWG chair and now a consultant to the Workgroup) referred to the initial QWG charge. With many other organizations working on how to fix the problems in health care and measure quality, the Committee saw its unique contribution as addressing problems and gaps in the underlying data infrastructure, to provide the information others rely on to assess and improve quality.
Dr. Mays urged the Committee not to let lengthy deliberations result in missing the wave of interest and attention around the release of the first AHRQ National Quality Report. If necessary, she proposed breaking up the issues and releasing a series of NCVHS reports. Related to this, Dr. Steinwachs asked the Committee to prioritize the various issues and proposals addressed in the report.
Ms. Greenberg stressed the importance of having a single report summarizing the findings from the five years of hearings and panel discussions on this topic. She noted that recommendations in more sensitive or difficult areas could be made more general or open-ended, if necessary. She expressed hope that a report could be produced and approved by March 2004.
Dr. Lumpkin encouraged the Committee to think outside the box, especially with regard to more controversial issues. Dr. Zubeldia suggested engaging the DSMOs in a discussion of these issues, and the group agreed to ask for time at an impending X-12 meeting to begin that process. The NCVHS representative was asked to present the Workgroup’s findings on the barriers to measuring quality and to talk with X-12 about how those barriers might be overcome and what role X-12 could play in doing so.
Ms. Greenberg observed that variants of some QWG recommendations had already been brought to the NUBC and NUCC by the Consumer/Purchaser Disclosure Project. Ms. Coltin commented on the trend toward pay-for-performance that is being pushed by large self-insured purchasers. Mr. Blair observed that the QWG report provides valuable information for the dialogue related to paying for performance, quality and safety. He suggested expanding the Committee’s discussion to consider what aspects of improved quality can be implemented within the current health system structure and, alternatively, to what extent the Committee needs to help embed quality in a new reimbursement structure or paradigm. The Executive Subcommittee was asked to frame future discussions.
After determining that the remaining meeting time did not permit a consensus process on the specific barriers to quality measurement, the Committee agreed to survey individual members on their views on each recommendation, with respect to 1) whether it should be recommended; 2) if so, whether the change is sought in the near term or long-term; 3) relative importance; 4) doability; and 5) suggestions on how to deal with the issue. Ms. Coltin urged members to think in terms of whether the information supports a pay-for-performance strategy. Mr. Hungate and staff were asked to design the survey, in consultation with the Executive Subcommittee, with the goal of getting feedback prior to the November full Committee meeting.
Dr. Lumpkin pointed out that for issues with a high value rating and a low doability rating, the Committee will need to come up with an outside-the-box solution. Dr. Zubeldia noted that X-12 requires a business case for recommended changes.
—DAY TWO—
Dr. Cohn, who chaired the second day of the meeting in Dr. Lumpkin’s absence, asked everyone present to introduce themselves.
Executive Subcommittee: Dr. Cohn
The Subcommittee will hold a retreat on November 21 in Princeton. Ms. Greenberg asked other Committee members to submit their agenda suggestions.
Subcommittee on Standards and Security: Dr. Cohn
The Committee reviewed a revised version of the letter to the Secretary reviewed the previous day. Dr. Cohn read it aloud. A motion was passed approving the letter, as revised.
The next Subcommittee hearings are scheduled for October 28-30, covering CHI domain recommendations, ICD-10, and PMRI recommendations, all for possible action in November. Future hearings are scheduled for January 27-28, March 30-31, and May 25-26, 2004.
Subcommittee on Privacy and Confidentiality: Mr. Rothstein
The Subcommittee will hold hearings related to the privacy rule on November 19-20; the hearings will focus on public health, research, and a discussion with invited stakeholders about issues related to HIPAA implementation. A February 3-4 hearing will address schools, law enforcement, and information-sharing problems related to health care reimbursement.
Subcommittee on Populations: Dr. Mays
The Committee was asked to review revised versions of the letters discussed the previous day. Both letters were read aloud and motions were passed approving them.
Dr. Mays reported that the Subcommittee would like to provide input on a proposed national childhood longitudinal survey to be conducted by NICHD. It proposes a presentation to the full Committee. In addition, the Subcommittee proposes a presentation to the full Committee by the CDC Director on the Centers’ strategic planning process.
The Subcommittee plans to take up the issue of mental health statistics, a neglected area of its charge. This is prompted by awareness that consideration is being given to dropping the mental health questions from the next round of NHANES. Finally, the Subcommittee has asked the Subcommittee on Privacy and Confidentiality to cosponsor a hearing on how to increase access to small-area data without compromising privacy and confidentiality. The Subcommittee will also hold a hearing in San Francisco, where the APHA is meeting in mid-November, on the collection of data on Asian, Native Hawaiians and other Pacific Islander populations.
Quality Workgroup: Mr. Hungate
The survey discussed above will be circulated to Committee members in mid-October. Mr. Hungate thanked Dr. Zubeldia for attempting to arrange time on the X-12 agenda to discuss the QWG project.
NHII Workgroup Letter to Department
Dr. Cohn presented a letter to the Secretary, drafted by the NHII Workgroup, encouraging the Department to increase its participation in federal interagency research and development initiatives. After discussion and editorial suggestions, the Committee approved the essence and intent of the letter, with the understanding that a clarified version will be forwarded to the Executive Subcommittee for approval.
Dr. Deering announced that the Markle Foundation’s Connecting for Health Initiative is turning its attention to some of the issues that have been on the Workgroup’s agenda.
The following topics were mentioned for the November meeting:
For the March meeting:
Dr. Cohn then adjourned the meeting.
I hereby certify that, to the best of my knowledge, the foregoing summary of minutes is accurate and complete.
/s/ John R. Lumpkin 03/29/04
_____________________________________________________________________
Chair Date