October 28-30, 2003
Washington, D.C.
Meeting Minutes
The Subcommittee on Standards and Security of the NCVHS was convened on October 28, 2003 at the Silver Spring Hilton in Silver Spring, MD. The meeting was open to the public. Present:
Simon Cohn, M.D., Chair
Jeffrey S. Blair, M.B.A., Vice Chair
Stanley M. Huff, M.D.
Clement J. Mc Donald, M.D.
Kepa Zubeldia, M.D.
John P. Houston, J.D.
Vivian Auld, NLM
Suzie Burke-Bebee, ASPE
John Burke, OS
J. Michael Fitzmaurice, Ph.D., AHRQ
Maria Friedman, CMS
Marjorie S. Greenberg, NCHS, CDC
Gail Graham, VA
Betsy L. Humphreys, NLM
Debbie Jackson, NCHS
Donna Pickett, NCHS
Steve Steindel, Ph.D., CDC
Marietta Squire, NCHS
Karen Trudel, CMS
Gracie White. NCHS
Michelle Williamson, NCHS
Bill Alfano, BCBSA
George Arges, AHA
Richard Averill, 3M HIS
Carol J. Bickford, ANA
Meryl Bloomrosen, Altarum
Alicia Bradford, CHI
Pat Brooks, CMS
Steve Brown, VA
Jim Cimino, Columbia Univ.
Craig Cochran, CAP.
Michael J. DeCarlo, AAHP/HIAA
Theresa Doyle, BCBSA
Karen C. Eckert, Wolters Kluwer Health/Medi-Span
Jack Emery, AMA
Lizabeth Fisher, NCHS
Christine Frasier, Amer. Academy of Professional Coders
Jason Goldwater, CMS
Tom Grissom, CMM/CMS/HHS
Tom Gustafson, CMM/CMS
Margaret W. Haber, Natl. Cancer Inst.
Denene M. Harper, AHA
Linda Hawking, Magna Systems
Frank A. Kyle, Jr. ADA
Kathy Lesh, the KEVRIC Co.
Randy Levin, Ctr. for Drug Evaluation Research, FDA
Michael Lincoln, VA
Marilyn Figmund Luke, AAHP/HIAA
Michael Lundberg, NAHDO
Sheila Madhani, ALOG
Elizabeth McCullough, 3M
Carol Monaco, Amer. Osteopathic Assn.
Stuart Nelson, NLM
Karen Nostein, CHI
Frank Pokorny, ADA
Sue Prophet-Bowman, Amer. Health Information Mgt. Assn.
Ravi Raman, NIST
George Robinson, First DataBank
Dan Rode, Amer. Health Information Mgt. Assn.
Paul Seltman, AdvaMed
Louise Smith, McKesson Corp.
Merritt Smith, BCBS
Jim Sorace, CMS
Jeremy Spiegel, The Gray Sheet
Walter Sujansky, independent consultant to NCVHS
Maureen Thompson, ASHA
Shelley Walworth, Unicor Med.
Cynthia Wark, CMS
Cheryl Wilhide, Magna Systems
Tom Wilder, AAHP/HIAA
ACTIONS
1. The Subcommittee approved a letter to the Secretary concurring with three CHI recommendations as modified on medications, immunizations, laboratory interventions and procedures
2. The Subcommittee approved a letter recommending standards for PMRI message formats and technologies.
3. The Subcommittee accepted the RAND cost/benefit analysis of migration to ICD-10-CM and ICD-10-PCS as presented to the full Committee at the September meeting and approved
4. The Subcommittee approved a letter recommending that the Department initiate the regulatory process for the concurrent adoption of ICD-10-CM and –PCS.
CHI FINAL REPORTS
The interventions and procedures team evaluated LOINC, CPT, and SNOMED CT for laboratory test orders, choosing LOINC for vocabulary due to its penetration in the marketplace and government. Dr. Sorace cited the need for a hierarchy and structure that could hold test information as users “drilled down.”
The medications team picked FDA’s established name and UNII codes for active ingredients. The Semantic Clinical Drug portion of RxNorm was chosen provisionally for clinical drugs. For manufactured dosage forms, CHI selected FDA/CDER’s Data Standards Manual. FDA’s NDC for product names and codes was selected for drug products. Only one option was found for medication packages: FDA’s Data Standards Manual for Packaging. CHI focused on using HL7 vocabulary tables for race, ethnicity and gender and made a limited recommendation for hierarchies from NDF-RT for mechanism of action and physiologic effect.
Ms. Wark reported that the immunizations team recommended HL7 version 2.4 and higher for the messaging standard and CDC/HL7 CVX and MVX codes for the HL7. CHI will revisit adverse event reporting in concert with the medications workgroup in 12-to-18 months.
CHI PRELIMINARY REPORTS
Three categories were picked for Genes and Proteins: inherited genetic variation, passive inheritance, and infectious diseases. Dr. Sorace discussed human genetic disease databases, NCI’s project to populate chromosomal alterations associated with malignancy, and efforts to develop standards and a plan in NIAID for annotating genomes in terms of pathogenicity data.
Ms. Wark said the encounters team hoped to recommend a single terminology system for all data elements and sub-domains within encounter. They will suggest terminology code sets for encounter data elements and let business functions and processes shape determinations.
Ms. Wark said the billing/financial team understood that HIPAA transactions and code sets were minimum standards for the CHI billing administrative domain. Appeals and certificate of medical necessity also were considered sub-domains in which CHI might identify alternatives.
The text-based reports team concluded the domain and sub-domains in scope included text document structure and syntax, electronic signature, document section headings, clinical document types and titles. Alternatives are the clinical document architecture, ASN1 (abstract syntax), Notation one, HTML, XML, Rich Text Format, Portable Format, Clinical LOINC, CEN, 13606, Public Key Digital Signature and Electronic Signature.
Components in scope for history and physical and not addressed elsewhere are: history of present illness, review of systems, past medical/surgical history, family and social histories, non-medical allergies, physical exam observations and findings. Initial alternatives are SNOMED CT, ICD-10-CM, ICD-10-PCS, MEDCIN, DSM-VI, MEDRA, Clinical LOINC, and CPT-4 codes.
DISCUSSION OF CHI RECOMMENDATIONS – Steve Steindel, CDC
Members cautioned that alternatives were not mutually exclusive and noted the difficulty of evaluating if a list of domains was right when the structure was unsettled. Members considered asking CHI to look instead at how to structure information now, allowing areas to mature while clinical domains and terminology were defined. Ms. Wark said CHI watched the timeframe and resources allocated; the goal was to complete initial analysis and recommendations for the first 24 domains in 2003. Anything that could not be accomplished in that timeframe or had to wait for processes to evolve or work to be done would be brought to the Council. The Subcommittee concurred with the CHI recommendations as modified for clarity. (See Actions, above.)
HIPAA UPDATE - Karen Trudel. CMS
Ms. Trudel noted the absence of widespread cash flow/operations problems post the October 16 compliance date. Compliance increased (33 percent), people tested and were in production, and percentages of compliant claims increased. Publication of the provider ID final rule is tentatively set for December. The Department continues to work on an enforcement rule that is substantive and procedural, covering all aspects of HIPAA Administrative Simplification.
DISCUSSION OF HIPAA IMPLEMENTATION ISSUES - Simon Cohn, M.D.
Noting that HIPAA adoption of ICD-9-CM, Volume 3 for inpatient procedures, but not outpatient procedures in institutional settings, is an issue, participants cautioned that ability to assign a principal procedure code to a diagnosis could be lost. Members noted a lack of consistency among payers made coordination of benefits difficult. Dr. Zubeldia said companion guides conveyed 30-40 percent of the payers’ requirements, and transactions for requirements not in the guide were rejected. While noting they were “only ten days into this,” Dr. McDonald credited everyone who, so far, kept this working. Dr. Zubeldia said the impact on the bottom line for payers was becoming measurable.
Dr. Cohn noted in January they would talk with DSMOs and others about obsolete implementation guides. In 2004, the Subcommittee will look at successes, failures and lessons learned from HIPAA and work with DSMOs to identify changes and extensions to the standards and prepare recommendations on where to go next. Ms. Greenberg noted the need to gather systematic information and report the status of the standards adoption for the annual report to Congress. Mr. Blair noted that if they knew the transition cost for the covered entities as well as perceived or measured savings, they could show any return on the investment for HIPAA in their report. Dr. Cohn said he hoped CMS would take care of most of the emergent issues, but the Subcommittee would stay close and, if necessary, provide time.
OVERVIEW OF SUBCOMMITTEE HEARINGS ON ICD-10 AND SYNTHESIS OF TESTIMONY – Simon Cohn, M.D.; Donna Pickett, NCHS
Ms. Pickett reviewed the timeline and industry comments on ICD-10 and summarized implementation issues identified in the testimony.
SUBCOMMITTEE DISCUSSION - Simon Cohn, M.D.
Members discussed whether to determine the feasibility of a procedure classification suitable for all settings. While acknowledging the history of testimony, diversity heard, and that there was no single view, they noted NCVHS had to “move something forward so industry could comment.”
ICD-10: NEXT STEPS: DISCUSSION PANEL
Mr. Grissom outlined issues with ICD-9-CM, compared it to ICD-10, and clarified the Center for Medicare and Medicaid Services (CMS) concerns. He said ICD-10 clinical modification (CM) provided proper treatment protocols, opportunities to understand what worked - how someone or an illness, disease, or its control progressed. ICD-10-CM’s potential benefits outweighed its risks. CMM will present their perspective at the November full Committee meeting.
Mr. Smith said delays and backlogs in coding could impede the flow of dollars and data. Extending current coding systems could make time to prioritize and build a framework for health care data for 20 years. He urged NCVHS to sort priorities and sequencing between alternatives.
Ms. Doyle stressed the cost burden of migration. She noted Nolan indicated cross walks could not address all comparability issues or solve the problem of data continuity. Noting concern about ICD-10’s impact on research and data trends, she called for a high-level stakeholder commission to forge a consensus on goals and objectives of NHII and a strategy for adoption and implementation of voluntary standards.
ICD-10 NEXT STEPS: TESTIMONY AND PUBLIC COMMENT - Michael Lundberg, Vice Chair, Board of Directors, National Association of Health Data Organizations
NAHDO advised an initial dual approach to procedure coding with ICD-10-PCS for hospital inpatient services and CPT-4 for outpatient, while an index and cross reference system was developed for patient services. Mr. Lundberg called for a two-year voluntary transition process and a defined date for adoption of both ICD-10-CM and ICD-10-PCS as replacements for ICD-9-CM. He suggested phasing in ICD-10-PCS for all services.
DISCUSSION OF ISSUES AND ICD-10 RECOMMENDATION LETTER – Subcommittee
Members discussed the findings in the RAND report presented to the full Committee at the September meeting and the ICD-10 recommendation letter. (See Actions, above.)
PATIENT MEDICAL RECORD INFORMATION STANDARDS: FINAL RECOMMENDATIONS ON CORE SETS OF TERMINOLOGIES - Walter Sujansky, M.D.
Members reviewed and updated as necessary the PMRI terminology recommendation letter. (See Actions, above.)
PLANNING FOR UPCOMING MEETINGS - Simon Cohn, M.D.
First priorities for the November meeting deal with any full Committee input on the letters and additional CHI final recommendations. December 9-10 hearings focus on CHI; claims attachments, CDA and how they work together; and pieces related to a document issue. The January 27-28 Subcommittee meeting agenda includes the security final rule, the DSMO’s yearly report, dental content relating to PMRI, and initial discussions about lessons learned from HIPAA. Hearings are also scheduled for March 30-31 and May 25-26.
The details of all presentations and letters drafted can be found in the meeting transcript posted on the NCVHS Web site, www.ncvhs.hhs.gov
- DAY ONE -
The Subcommittee on Standards and Security of the NCVHS held three days of hearings on October 28-30, 2003. The Committee is the main public advisory committee to HHS on national health information policy. The first day members heard three proposed final and five preliminary Consolidated Health Informatics Initiative (CHI) reports and provided input into the CHI process. An update was presented on the HIPAA regulations citing progress on further final and proposed rules and implementation of administrative and financial transactions. The next days the Subcommittee heard testimony on ICD-10-CM and-PCS and discussed next steps and recommendations to the full Committee. Members also reviewed and updated recommendations on Patient Medical Record Information (PMRI) core terminologies. Recommendation letters were drafted on the CHI final reports, ICD-10, and the PMRI core terminologies.
CHI AGENDA REVIEW, Cynthia Wark, CMS
Ms. Wark reported that CHI made every earnest effort to keep on schedule and complete work on all 24 domains in the portfolio by the end of 2003. NCHVS had heard presentations on lab result contents, demographics and units domains at the Subcommittee’s August and September meetings. Subsequent presentations were held with the CHI Council and participating agencies. Recommendations are in clearance. The next round will be interventions and procedures non-laboratory and laboratory test orders, diagnosis and problem list, anatomy/physiology, nursing, medical devices and supplies. Teams are deployed for encounters, disability, population health, text based reports, history and physical, multi-media, genes and proteins, billing/financial, and chemicals. They intend to have remaining preliminary reports available for the next Subcommittee meeting.
Ms. Wark noted the CHI process has four stages: (1) deployment of subject-matter-expert teams to define scope and identify candidate terminologies; (2) analysis and feedback to evaluate terminologies and assess deployment followed by a preliminary report to the Subcommittee; (3) technical presentation to the CHI Council along with department/agency review and feedback; and (4) Council consensus followed by a final report and (if everyone is in-sync) standards adoption with government wide policy rollout and agency/department specific implementation.
CHI FINAL REPORTS - INTERVENTIONS AND PROCEDURES – Laboratory Test Orders - Jim Sorace, CMS
In considering the standard for laboratory test orders, Dr. Sorace said the team (representatives of VA, DOD, NIH, and Indian Health Service (IHS)) took “a machine-oriented view” of how computers might represent laboratory test names, working backwards to end-users to determine if there was a vocabulary for test names and panels doing clinical and anatomical pathology. The workgroup evaluated LOINC, CPT, and SNOMED CT, recommending LOINC for its well-focused vocabulary and considerable penetration in the marketplace and government.
Noting an interface problem, Dr. Sorace cited the need for a hierarchy and structure that coding-naïve users could use to gain general test information as they “drilled down.” LOINC named and standardized over 50 panels and worked with the laboratory community to add flexibility, addressing panels designed based on rapidly updatable epidemiological data.
Discussion
Asked about CPT/LOINC mapping for billing, Dr. Sorace said the workgroup recommended this and LOINC was constructing a map. Participants mentioned copyright issues, but noted LOINC was loaded into the UMLS and already “knew” much of the mapping. Dr. Huff said orderable codes and panel definitions were always the intent and within the work’s scope and that LOINC spent about seven years defining batteries.
Noting CDC considered panels a fundamental feature of LOINC important for grouping tests for reporting and ordering purposes, Dr. Steindel said CDC always considered this from both sides. Until recently, CDC had a prescriptive approach (a limited group of tests were defined as a LOINC battery) and another with a set of tests defined as “part of the battery” along with optional tests fleshing it out (i.e., how LOINC commonly defines a CBC). The epidemiological/ surveillance community asked LOINC to add an open panel and they are working on that. He noted the CHI recommendations asked for and LOINC also was creating more generic test codes that might be designated the preferred ordering form. Any LOINC code could be orderable.
Participants noted LOINC worked directly with UMLS to map LOINC codes within UMLS. Dr. Cohn said the draft PMRI core terminology letter cited the importance of maintaining availability of mappings. Dr. McDonald suggested that they could readily commit to mapping to UMLS; there was already agreement on both sides and it was a “three-open” standard. Dr. Steindel noted the letter also stated that NLM would maintain these maps.
CHI PRELIMINARY REPORTS - GENES AND PROTEINS - Jim Sorace, CMS
Dr. Sorace said the workgroup strove to ensure that, ultimately, a health care information infrastructure was served and took a deliberately top-down clinical approach, adopting the view of a health care worker in the future who has a clinical problem and must research what is known about a disease state in terms of genes and proteins. The workgroup established three categories: inherited genetic variation (i.e., applications of genomics to disease), passive inheritance--acquired genetic mutations, and infectious diseases (genomes important for health care, including medically important human pathogens).
He cited big human genetic disease databases: National Center for Bioinformatics (NCHI) Online Mendalian Inhertance in Man (OMIM, http://www.ncbi.nlm.nih.gov/omim/), GeneTest, (http://www.geneclinics.org/), CDC Human Genome Epidemiology Network (HuGENet, http://www.cdc.gov/genomics/hugenet/default.htm). He said National Cancer Institute’s (NCI) Genome Cancer Micro Anatomy Project to populate chromosomal alterations associated with malignancy hopes to be ready for review in six months. He noted efforts to develop standards in the National Institute of Allergy and Infectious Diseases (NIAID) with a pathogen anatomy branch and a plan for annotating genomes in terms of pathogenicity data, in order to build a cohesive set of databases regarding infectious organisms.
Discussion
Dr. Sorace noted that it was difficult to find “a keeper” for the protein nomenclature. Dr. Steindel suggested the workgroup consider the FDA’s investigations as they “tightened down” on ASR reagents used by laboratories for present testing for genes and also proposed Clinical Laboratory Improvement Act (CLIA regulations for genetics subtests as a special category of laboratory testing defining a genetic test. Dr. Sorace clarified that environmental exposures fit under all three categories. He added that a group at NIH, predominantly NCBI, would keep the genome, be responsible for nomenclature issues, and maintain the sequence. Noting GeneTest was adopting primitive but useful approaches (e.g., XML tagging gene reports and developing a set of documents that could be marked up and include UMLS vocabulary), he said no standard in these fields was a “slam dunk” but that, hopefully, they could provide input on how to formulate these standards.
He clarified that, while protein and gene nomenclature were not in scope for this domain’s short- term work, Human Genome Organization (HUGO, http://www.gene.ucl.ac.uk/hugo/, which he considered would be a de facto standard) and other gene nomenclature organizations tracked DNA-level work. The workgroup decided no one could legitimately claim themselves a protein nomenclature or address that within the CHI timeframe.
Dr. Sorace said the workgroup would flesh out plans for: annotating and keeping microbiological genomes; better understanding relationships between OMIM, GeneTest and other contributors; and how to mark up and store the work. He noted tradeoffs: it had to be well structured, keeping up with the literature was a challenge, and there were issues with structured vocabularies.
Asked about cross pollination, sharing best practices, and building on lessons learned, he said there was a plan to bring resources together. Immediate focus was on answering those questions.
CHI FINAL REPORTS - MEDICATIONS - Steve Brown, M.D., VA
Dr. Brown reported that since their preliminary report in May the workgroup (representatives from CDC, DOD, FDA, IHS, and NLM) completed the CHI process and reached consensus in the Council. The group “cast a wide net” asking which areas of medication-related terminologies were important. Drugs were broadly defined as products intended for the diagnosis, cure, mitigation, treatment or prevention of disease, manifestations or symptoms of disease, or altered structure or function of the body. Areas initially considered in scope included active ingredient, clinical drug, manufactured dosage forms and drug products including finished dosage form, packaging, labeling, section headers, populations, and drug classifications. Analysis for adverse events included gene administration, indications, contraindications, and aspects of pharmacokinetics and pharmcodynamics.
Options considered for active ingredients (substances responsible for effects of a medication) included CAS numbers, MolFile structures, FDA’s, United States Adopted Names’ (USAN’s), The International Union of Pure and Applied Chemistry’s (IUPAC’s), United States pharmacopeia (USP)-NF’s and international established names. FDA’s established name was picked along with Unique Ingredient Identifier (UNII) codes. Dr. Brown cited FDA’s regulatory authority, the established name’s wide use internally within FDA, and UNII codes becoming availability free through NLM.
Alternatives considered for clinical drug (a pharmaceutical preparation with components, defined as active ingredients and strengths, and the dose form as administered) included RxNorm’s Semantic Clinical Drug section as distributed in the UMLS, certain core clinical drug portions of SNOMED CT, and representations at the clinical-drug level by First Databank, Multum, Micromedex, and MediSpan. The Semantic Clinical Drug portion of RxNorm (a public domain system developed with NLM in conjunction with FDA, VA, and others in consultation with HL7 which has mappings freely available to the VA’s NDF, FDB, Micromedex, MediSpan and Multum) is still under development and recommended on a provisional basis.
For manufactured dosage forms, the group gauged approved drug products from the FDA/CDER Data Standards Manual and dose forms in HL7 and selected FDA/CDER which Dr. Brown noted was FDA-regulated, authoritative, widely used, appropriately granular and included definitions.
Drug products (one or more finished dosage forms, each containing one or more ingredients) considered included FDA’s National Drug Code (NDC) for product names and codes, FDB’s Drug Facts and Comparisons, and Micromedex. NDC, a HIPAA approved code set broadly used in transactions domestically and internationally was selected. Dr. Brown noted opportunities for improvement and encouraged FDA to further revise NDCs and address well-known issues in their co-generation processes.
Only one option was found for medication packages: FDA’s Data Standards Manual for Packaging. Noting FDA’s and LOINC’s work in moving towards electronic representation of labels and HL7’s November balloting, Dr. Brown reported putting forward consideration of label section headers. Noting the near limitless number of ways to “slice and dice” recommendations about special populations and the CDC/ Census/HL7 collaboration, he said the group decided on using HL7 vocabulary tables for race, ethnicity and gender.
Concluding that probably they could never come up with a single classification useful in all instances and that drug classifications had to be use-case based, the group made a limited recommendation for hierarchies from NDF-RT for mechanism of action and physiologic effect. Dr. Brown said the intent was to spur sharing for decision support rules in patient safety issues.
Discussion
Dr. Brown said initial publication of UNII codes was in scope, but he predicted a comprehensive list of UNII codes was unlikely in 2003. Ms. Wark clarified that the CHI recommendation was contingent upon FDA determining what had to be done, the resources needed, and making the codes freely available. Dr. Levin noted FDA had a system in place that generated codes for ingredients and sought to make it publicly available. FDA was improving their processes, had already generated 2,000 codes, and was working legacy data into their processes so codes would be assigned to investigational drugs and the system kept up to date and comprehensive. He said generating codes for most ingredients would not require regulatory change; there were no restrictions for investigational products (soley a resource issue). FDA aimed to have this process in place within 2004. He added, however, that ingredients of products already in the market were captured through FDA’s listing process; improving that will require regulatory change.
Dr. McDonald suggested that the Subcommittee identify weaknesses in NDC in order to make those changes easier. Dr. Cohn said the Subcommittee would consider whether to hold hearings about NDC code features in 2004 as post-HIPAA implementation
Dr. Levin clarified that FDA was working on: the structured product label balloted in HL7, better ways to communicate drug information in package inserts, and how to structure the label in HL7.
Members recommended CHI investigate dosage and administration sub-domains as a next phase activity. Ms. Wark described a phase-two work plan for out-of-scope areas that could benefit from further work. Noting some recommendations about NDC code changes required regulatory change, Dr. Steindel suggested advising the Secretary to be expeditious in starting this process.
Dr. Steindel said NCVHS could be proactive and hold hearings to advise the Secretary about regulatory changes that might be needed or wait until the Department published an NPRM with changes FDA wanted to see and react to it. Dr. Levin noted regulatory changes might not change code characteristics, but much else about how the code is obtained and assigned might have to change to improve how the code is handled. Noting their intent was to be expeditious, Mr. Blair cautioned about being so expansive that they moved into any unnecessary regulatory process.
CHI FINAL REPORTS - IMMUNIZATIONS - Cynthia Wark, CMS
The immunizations workgroup (representatives from CDC, USAID, FDA, DOD, IHS, and CMS) reassessed the messaging standard and looked at vocabulary standards for immunizations, defining that domain as implementation of a data standard for immunizations providing an organized, streamlined means of communicating between federal partners by offering a real time means of transferring information regarding immunization encounters, vaccine events, patient records, and other immunization related information. Immunizations scope was defined broadly and encompassed patient safety, including standardization to and from providers and users of vaccine information (e.g., primary care physicians, schools) and providing an up-to-date, standardized method of communication to keep vaccination records current and complete. Alternatives considered were HL7 version 2.4 and SNOMED CT.
Final recommendations were HL7 version 2.4 and higher for messaging and CVX and MVX codes for HL7 immunization terminology. Conditions and gaps were identified. Noting that many sub-domains for medications (e.g., dosage, administration) were identical for immunizations, the group chose to revisit adverse event reporting in concert with the medications workgroup within 12-to-18 months. Members concurred after modifying these recommendations for clarity. Ms. Wark will send the Subcommittee an edited version to forward to the Secretary.
CHI PRELIMINARY REPORTS - ENCOUNTERS - Cynthia Wark, CMS
Realizing “clinical encounters” encompassed other domains that subsequent teams worked on, Ms. Wark said the group was put on hold until this fall. They defined clinical encounter as (1) interaction regardless of setting between a patient and a practitioner vested with primary responsibility for diagnosing, evaluating or treating the condition or providing social services, and (2) contact between a patient and a practitioner with primary responsibility for assessment and treatment at a given contact exercising independent judgment. Noting encounters link clinical, administrative and financial data, the group considered several sub-domains in scope.
The workgroup began with HL7, basic encounter message, noting HL7 messaging was already adopted as a standard and served as a model for viewing clinical encounters. The HL7 standard consists of 25 message segments comprising 612 data fields. Some 517 data fields falling within other domains’ scope were eliminated. Another 57 of the remaining data fields were determined to not need standards adopted, leaving 38 candidates. Alternatives identified after analysis were: ASTM standards, HL7 version 2.4 admission/discharge transfer messages, X-12N 837 health care claim message, SNOMED CT, and HL7 version 3.
Data elements only supporting internal operations within a facility were considered out of scope. Noting HL7 encounter messages could report information before, during, and after encounters, they chose not to limit scope based on life-cycle phase. The group will attempt to recommend a single terminology system for all data elements and sub-domains within encounter. They decided to suggest terminology code sets to use for encounter data elements (rather than propose mandatory elements) and let business functions and processes make those determinations. The group chose to clarify the analysis by identifying sub-domains within encounter (e.g., provider, admission, discharge, accident information, death and autopsy information) and sought the Subcommittee’s advice.
Noting NCHS collects information on inpatient discharges and ambulatory care that overlap the encounter domain, the group’s preliminary stance was not to try harmonizing clinical encounter recommendations with other government reports.
Discussion
Asked about how encounters related to the national provider identifier (NPI), Ms. Wark recalled that in discussing demographics for a patient they looked at variables that could identify medication was administered to the right person. Encounters “didn’t encompass a provider identifier per se,” other elements described provider characteristics. Ms. Trudel said CMS would look at the National Provider File, which might have elements such as a specialty taxonomy code, and ensure harmonization.
Mr. Blair suggested the value in what the workgroup was doing was that there would be a declaration that, within their facilities, people would move towards compliance with ADT standards and strive to eliminate optionality. Ms. Wark said the feedback was valuable and reflected the workgroup’s belief in the value of standardizing as much as possible, even internally.
Members suggested provider-to-provider discussions as consultations and outpatient and less-direct encounters (e.g., diagnostic labs, e-encounters, and telemedicine) as sub-domains.
Noting CMS answered questions and identified issues around encounters for years, Dr. Cohn cited CMS regulations and advisories as a source. Noting encounters was a complicated domain interfacing with many others, Ms. Wark said CHI chose to make it manageable by using HL7 as a model and looking at data elements in that. Provider information was one of those areas.
CHI PRELIMINARY REPORTS - BILLING/FINANCIAL - Cynthia Wark, CMS
Ms. Wark said the billing and financial domain called for validation rather than full analysis of alternatives because of existing HIPAA standards. Scope was defined as the standards used to implement electronic exchange of health-related information needed to perform administrative functions in the federal health care enterprise. Alternatives identified were code sets adopted under HIPAA plus ICD-10-CM.
The workgroup began with sub-domains identified under HIPAA: claim submission for reimbursement, health care claim payment/advice, eligibility determination, prior authorization referral, enrollment/disenrollment, coordination of benefits, and claims status inquiry. No standards had been adopted in two other sub-domains (claims attachments and report of injury) considered out of scope. The group identified additional areas in which to consider alternatives for standardization within the federal government: appeals and certificate of medical necessity.
Noting health plans and providers transmitting designated HIPAA transactions electronically within the federal government will be HIPAA-covered entities required to be compliant with the HIPAA transactions and code sets, the group’s initial thought is that HIPAA transactions and code sets are assumed to be minimum standards for CHI’s billing administrative domain.
Ms. Wark said claims attachments are seen out of scope due to scheduled publication of the attachment NPRM in 2004. HL7 attachment special interest group and CHI staff are mapping and aligning CHI clinical standards with the proposed HL7 claims attachment standards.
Discussion
Ms. Bickford suggested that CHI consider the ABC codes that are complementary to existing CPT-4 codes for procedures and activities and the financial component attached for the billing pieces. While ABC codes were not a complete, all-inclusive data set, she noted that others lacked their important components for complementary and alternative therapies.
CHI PRELIMINARY REPORTS – TEXT-BASED REPORTS - Alicia Bradford, CHI
Ms. Bradford said the text-based reports and history and physical groups considered merging reports but in defining their scope realized that would be too difficult. The text-based reports group’s scope was defined as a domain determined to include standards and terminologies used to define messaging architecture and syntax of clinical text documents. All clinical document types were considered possible sub-domains. Additional sub-domains were delineated from analysis of content of clinical document types including section headings and data types. The group decided that including those sub-domains resulted in scope too broad for the time frame and resources allocated. Document components and data domains contained in text-based documents overlap broadly with other CHI domain teams. The group concluded that the domain and sub-domains in scope include text document structure and syntax, electronic signature, document section headings, clinical document types and titles. Alternatives are the clinical document architecture, ASN1 (abstract syntax), Notation one, HTML, XML, Rich Text Format, Portable Format, Clinical LOINC, CEN, 13606, Public Key Digital Signature and Electronic Signature.
Discussion
Mr. Blair suggested adding the continuity of care record: an initiative from ASTM, HIMSS and the Massachusetts Medical Society related to documentation records for radiology, discharge summaries and claim attachments that is complementary to HL7 and can be used independently. Ms. Bickford noted that ASTM’s Committee E31 groups worked on format and structure for clinical records and standardization of headings. Noting reports coming from radiology on imaging studies, she emphasized procedure notes and operative reports as well as laboratory findings. Ms. Bradford explained that laboratory findings were determined out of scope and that the group chose to standardize the format of the document rather than numerous types of text documents, so it could be exchanged between different platforms. A multi-media workgroup will look at radiology. Dr. Zubeldia suggested adding as a sub-domain text encoding and alphabets.
Noting the Subcommittee held hearings on electronic signature, Dr. Cohn recommended contacting HL7, ASTM and others. Dr. Zubeldia suggested a federal eGov project on electronic signatures and e-authentification might provide an idea bridge between eGov and HIPAA electronic signatures through CHI. He recommended looking at electronic signature for transactions as well as human-readable text.
CHI PRELIMINARY REPORTS - HISTORY AND PHYSICAL - Alicia Bradford, CHI
The history and physical group defined the terminology as used to identify, classify and name components incorporated into a patent’s medical history and the physical exam process performed by a physician. Noting the process is not standardized and dependent upon clinical judgment, Ms. Bradford said they took a broad-brush approach, keeping in mind JCAHO accreditation requirements.
The group developed a list of major components or sections that might be included in these documents and compared each to domains addressed by other groups. Components not addressed elsewhere included: history of present illness, review of systems, past medical/surgical history, family and social histories, non-medical allergies, physical exam observations, and physical exam findings. Initial alternatives were SNOMED CT, ICD-10-CM, ICD-10-PCS, MEDCIN, DSM-VI, MEDRA, Clinical LOINC, and CPT-4 codes.
DISCUSSION OF CHI RECOMMENDATIONS, Steve Steindel, CDC
Dr. McDonald cautioned that alternatives were not mutually exclusive: LOINC was headers, sections or questions; SNOMED was answers, findings or diagnoses; MEDCIN included a logic to follow within a structure. Dr. Cohn remarked on the difficulty of evaluating whether a domain and sub-domain list was right or comprehensive when the structure was still unsettled. Ms. Bickford noted the language of the definition focused exclusively on pathology and did not accommodate wellness and health promotion initiatives. She point out that the clinical record component of the ASTM standard supported the group’s activities.
Emphasizing that they did not want work to stop on text-based standards but much work was in process, and recalling comments about the difficulty of defining a history and physical, members asked CHI to consider looking instead at how to structure information now and allow some things to mature until clinical domains and terminology were better defined. Ms. Wark clarified that the schedule was to complete initial analysis and recommendations on the first 24 domains by the end of 2003. Everyone looked at the timeframe, resources allocated, and left out of scope what couldn’t be adequately covered. What could not be accomplished in this timeframe or had to wait for processes to evolve or work to be done would be brought to the Council. The Subcommittee concurred with the CHI recommendations as modified for clarity and voted unanimously to accept the recommendation letter.
HIPAA UPDATE - Karen Trudel, CMS
Ms. Trudel reported no indications of widespread cash flow or operations problems following the October 16 compliance date. Compliance was increasing, people were successfully testing and going into production, and there was a sharper increase in the percentages of compliant claims. Medicare’s October 10 statistics indicated 28 percent compliance. She estimated up to a third of the claims in production were fully compliant. Medicaid also reported that roughly 33 percent of the non-pharmacy claims were compliant, in production and “without problems.” Ms. Trudel noted similar reports from BCBSA and other plans.
She reported that the provider ID final rule moved to OMB for review; publication is tentatively set for December. Ms. Trudel said the Department continued to work hard on an enforcement rule that was both substantive and procedural, covering all aspects of HIPAA Administrative Simplification. The proposed rule would augment the procedural rule published in the spring. She anticipated considerable comment.
Discussion
Ms. Trudel clarified that the NPI database would not be available when the final rule is published (until the Secretary’s and OMB’s final decisions on the data elements and identifier they “couldn’t go past a certain point”). The final rule will identify what the identifier will be, how enumeration takes place, if providers had to pay for an identifier, and ability to access and use the data. The ultimate compliance date will designate that enumeration begins two years after the database is available. Ms. Trudel noted a system was in place to accept complaints; two on transactions and code sets had been received.
She provided an update on a request for an OGC opinion about aspects of the DSM-4/ ICD-9 issue worked out at the Department level. Frequently asked questions (FAQ) addressing issues raised by the American Psychiatric Association (APA) and posted on the CMS HIPAA Web site (www.cms.hhs.gov/hipaa/hipaa2) clarify that no diagnostic criteria were adopted under HIPAA and that they are outside the scope of HIPAA. FAQ specify that DSM-4 (or any other diagnostic criteria) can be used and that, although ICD-9 codes need to go on HIPAA transactions, it is acceptable to crosswalk between short descriptors in ICD-9 and DSM-4. APA will have a period of time to request changes to short descriptors in ICD-9 that seem especially outdated or inappropriate. Ms. Trudel explained that once the NPI is fully implemented UPINS and other previous identifiers for providers will no longer be used on the transactions. But for some time, perhaps permanently, there will be behind-the-scenes cross mapping to existing identifiers.
Ms. Trudel doubted that the law permitted assigning different compliance dates to various kinds of covered entities, but suggested this could be done on a voluntary industry basis. Noting that people paid attention to the WEDI-SNIP sequencing paper with its testing levels, she suggested that WEDI-SNIP could get enough consensus that things ran fairly smoothly.
Members considered whether the industry could move to the NPI earlier in the two-year cycle, noting the implementation guide required sending a tax ID along with the primary NPI and that sending two accommodated both kinds of payers, providing an easier, more flexible transition.
Asked about potential for annual modification to the transaction rules, Ms. Trudel said the first was underway but they did not yet have a sense of the content. She noted that the appropriation continues to contain language prohibiting allocating funds for an individual identifier.
DISCUSSION OF HIPAA IMPLEMENTATION ISSUES - Simon Cohn, M.D.
Participants said an issue for the industry was that HIPAA adopted the ICD-9 for inpatient procedures but not for outpatient procedures in institutional settings. Some carriers required on the claim to manually crosswalk CPT procedure codes to ICD-9- Volume 3 procedure codes. Noting many hospital providers used the ICD-9 procedure code internally to compare inpatient and outpatient environments, they cautioned that ability to assign a principal procedure code to a diagnosis could be lost. Members also mentioned a lack of consistency among payers (some insisting on coding anesthesia with minutes, others units or both) made coordination of benefits difficult, if not impossible, in some cases. Dr. Zubeldia said so far Claredi corrected over 300 companion guides and implemented edits for more than 160. Companion guides only conveyed 30-40 percent of the payers’ requirements and payers rejected transactions for requirements not in the guide. He noted there already had been several iterations of some guides. Dr. Zubeldia said the advantage of going to HIPAA transactions was reduced because new transactions carried the same friction. While noting that they were “only ten days into this” and that the rule might not have hit places with 60-day delays in payments, Dr. McDonald credited everyone who, so far, kept this working. Dr. Cohn agreed that there would be more to discover. In January members will talk with DSMOs and others about obsolete implementation guides.
Reporting he heard from a large payer that their auto adjudication rate on claims went from the low 60’s before HIPAA to the high 80’s with HIPAA transactions, Dr. Zubeldia said the impact on the bottom line was becoming measurable. Dr. Cohn noted that in 2004 the Subcommittee will look at successes, failures and lessons learned from HIPAA. Members also will work with DSMOs to identify changes and extensions to the standards and prepare recommendations on where to go next. Noting that, while the focus was mostly on claims, many providers looked to “get the bang for the buck” on eligibility, claims status, inquiry and response, remittance, and other transactions, Mr. Argus expressed interest in industry’s extent of utilization.
Members noted much information passed around was anecdotal and that they had to find a way to get universal data. Ms. Greenberg said it was time to begin reporting what happened with the standards adoption. She recalled discussions about using groups who routinely collect this information, but agreed that they needed to think about a way to gather systematic information. Mr. Blair observed that if they could find out the transition cost for the covered entities as well as perceived or measured savings, they would be able to show a return on the investment for HIPAA in their reports to Congress. Dr. Cohn said his hope was that CMS took care of most emergent issues, but the Subcommittee would stay close and, if necessary, provide time.
OVERVIEW OF SUBCOMMITTEE HEARINGS ON ICD-10 AND SYNTHESIS OF TESTIMONY - Simon Cohn, M.D.; Donna Pickett, NCHS
Ms. Pickett built on comments heard over the last year at hearings and meetings as well as earlier deliberations about ICD-10-PCS and ICD-10-CM. She provided information in segments and timelines, highlighting NCVHS meetings, hearings and letters received as well as development of ICD-10-CM and -PCS. She outlined development of both, adding input gained from AHA’s and AHIMA’s joint presentation on pilot testing. Her history and overview of ICD-9-CM and the coordination/maintenance update process that strives to keep it clinically relevant to the current health care environment noted that limited space blocked updating parts of the classification.
She also outlined NCVHS’s timeline, chronicling its history in relation to the classification systems. She detailed highlights and major milestones of NCVHS’s and its Subcommittees’ proceedings related to ongoing discussions to move forward with implementation of ICD-10-CM for morbidity diagnosis classification and ICD-10-PCS for inpatient procedure classification. The timeline included reports, letters and other external actions concerning the need to implement these classifications. Ms. Pickett noted the document will be updated to include letters from AAHP/HIAA and the American Academy of Professional Coders distributed at the meeting.
An appendix synthesizes in table format the timeline and perspective of 83 organizations’ testimony. A summary of implementation issues raised in testimonies and letters to the full Committee or Subcommittees noted disadvantages of remaining with ICD-9-CM: limited ability to expand classifications to incorporate new diseases and technology, lack of comparability with U.S. mortality data and international data, limited ability to adjust payment methodologies due to a lack of specificity and detail, and limited ability to evaluate health care outcomes. Potential benefits included: better understanding of health conditions, health care outcomes, and the value of new procedures; improved disease management; potential for fewer miscoded, rejected and improper reimbursement claims; and more accurate payments for new procedures.
Noting many issues related to timing, Ms. Pickett highlighted the sense that: no new code sets should be implemented until after initial HIPAA standards were implemented; lessons should be learned from HIPAA implementation; the implementation and maintenance process should be well defined; the timeframe for implementation should be two-to-three years or no later then 2005; crosswalks should be available prior to implementation; and quality controls had to be in place to monitor systems during transition. The issue of resources surfaced in numerous testimonies along with the suggestion to adjust payment methodologies to be sensitive to increased regulatory costs of implementing new code sets.
Issues related to training included ample time for education and training; broad-based training and retraining; a train-the-trainer approach; identified tools for training; and coordinating training across sites, regions and stakeholders.
Issues related to systems changes included: product redesign, testing, systems integration between applications and foreign systems, upgrading or total revamping of hardware/software, software conversion, optical scanning systems, decision support systems, field lengths, coding edits, order entry, lab, radiology, pharmacy, medical record, patient accounting systems, revisions of product and user manuals, and the need for coordination between vendors. Also cited were the need to run parallel systems, increase storage capacity, changes to state reporting systems, cost containment and discharge reporting systems, and quality reporting systems.
Other changes were related to policy and fee schedules, fee schedule calculations and structures, authorization streams, DRG assignment and pricing interfaces, changes to groupers and end coders, and the need to renegotiate contracts. Also noted were changes to databases used for modeling, adjustments for care management policies, redesigned paper forms, data abstracts, report layout and format, and the need for new claim and other treatment forms.
SUBCOMMITTEE DISCUSSION – Simon Cohn, M.D.
Members discussed whether, rather than continually revisiting the issue of dual coding systems, it might be better to determine the feasibility of a procedure classification suitable for all settings - an option NCVHS wrestled over for decades and most countries adopted that probably was not currently feasible for more than political and institutional issues. CMS and the hospital industry said CPT, as currently constituted, did not meet the needs for hospital statistics and reporting.
Mr. Argus emphasized the benefit in moving forward with ICD-10-CM and ICD-10-PCS for the inpatient institutional setting. Ms. Bickford stressed the context: new technologies and an evolving health care system, with most care conducted in outpatient settings, often by other practitioners than physicians, with multiple technologies unrepresented in current coding systems. Dr. Cohn proposed a letter acknowledging the long history of testimony and the point where, as an advisory body, NCVHS had to move something forward so the industry could comment, while recognizing the diversity heard and that there was no single view on how or when to move forward.
- DAY TWO -
ICD-10: NEXT STEPS: DISCUSSION PANEL
Mr. Grissom outlined issues with ICD-9-CM, compared ICD-9 and -10, and clarified CMM’s concerns. He noted: ICD-9-CM had been seen for at least a decade as “fast becoming outmoded;” examples of where ICD-9-CM had reached its limit; and ways ICD-10 would lead to more accurate, appropriate payments and improve the quality of health care. He said ICD-10-CM’s increased specificity and breadth of provisions provided the ability to develop metrics and measure and follow patients to do disease management and develop protocols in ways not possible with ICD-9-CM. ICD-10-CM provided an opportunity to understand what proper treatment protocols should be and how someone or an illness, disease or its control progressed over time.
Mr. Grissom said crosswalking DRGs and moving diagnostic related groupings would involve “a fair amount of work,” but he predicted the number of claims processed and accepted electronically the first time would increase (98 percent of CMM’s claims on the inpatient side pass through electronically the first time; most rejections are not denials but suspensions due to insufficient documentation). Considering the dollar value of the benefit and returned/reworked claims, he expressed confidence that benefits would outweigh costs. Medicare and Medicaid’s estimate will be presented at the November full Committee meeting.
Discussion
Mr. Grissom advised developing best practices and treatment protocols that focused on outcomes as well as the number of providers, dollars paid, and services delivered.
Asked if ICD-9-CM, Volume 3’s coding hierarchy could be converted into just an identifier for the procedures, Mr. Grissom cautioned that, rather than simplify, temporary codes “that never died” created entropy and made systems more complicated.
ICD-10: NEXT STEPS: DISCUSSION PANEL
BCBSA commissioned Nolan to analyze the costs and benefits of migration to ICD-10-CM and ICD-10-PCS. Nolan conducted secondary research and primary interviews, developed cost models, and built-out and compared a range of costs against experiences in other countries. Nolan estimated a two-to-three-year implementation cost for hospitals, physician groups, and health plans of 3-to-14 billion dollars.
Mr. Smith noted that implementation was a “blank suit” to many interviewees and entire categories of provider organizations and payer types were excluded from the estimate. He cautioned that, without extraordinary care in implementing proposed changes, delays and backlogs in coding would impede the flow of data and dollars. He noted experiences in Australia and Canada and AHIMA and AHA’s experiment at Ohio State University indicating substantial reduction in the productivity of coding. He noted that much of the expense of implementation would come from testing, confirming and “becoming comfortable” with newly displayed and different data. Mr. Smith predicted that it would be some time before data users returned to normal conditions of use because of their conservative nature, the importance of the data for clinical and financial uses, and new federal regulations requiring very detailed assurance of the quality of data used in creating financial statements. He suggested that other alternatives might give more benefit than having facilities, physician organizations and HMO’s make the conversion. Mr. Smith urged NCVHS to sort priorities and sequencing between alternatives.
Contending that it was hard to argue that “we’re running out of codes” when less then 15 percent of the code space is used, Mr. Smith suggested instead that the codes were “sitting in the wrong place.” He said extending the life of current coding systems could provide time to safely prioritize and establish the framework for health care data for 10-20 years.
ICD-10: NEXT STEPS: DISCUSSION PANEL
Ms. Doyle explained that an industry advisory panel composed of provider organizations, national accreditation organizations, and representatives from the National Association of Medicaid Directors assisted Nolan in developing the cost/benefit analysis’s initial scope and provided feedback during initial research.
She stressed the industry cost burden of migration to ICD-10-CM and -PCS on the heels of Y2K and HIPAA, clarifying it was actually 5.5 to 14 billion dollars, and questioned the wisdom of wholesale change to the codes at this time. Nolan projected the cost for a large hospital between 1.5-and-5 million dollars and for a large multi-regional health plan between 10-to-20 million dollars. Ms. Doyle emphasized that the numbers were conservative.
Ms. Doyle noted the report concluded that benefits were speculative. She expressed BCBSA’s concern about the potential impact of ICD-10 on medical research and data trends, citing the report indicated that existing medical knowledge would be degraded significantly over a three-to-five year period. Noting the report said cross walks could not address all comparability issues and would not solve the problem of data continuity, she urged the Committee to ensure that these issues were addressed before making a recommendation.
She noted the analysis questioned the sequence of standards adoption, arguing that in order for any benefits from ICD-10-PCS to be achieved - particularly improvement in outcomes measures - standardization of a clinical vocabulary was a prerequisite, Ms. Doyle stressed an urgent need for a national information technology strategy that established an industry roadmap to guide policy makers, standards development, and technology investments.
Ms. Doyle said ICD-10-CM and -PCS could not be considered in isolation of other health care technology initiatives and that the industry needed a blueprint, prioritization, and a timeline to provide order, predictability and cost effective, efficient implementation. She called for creation of a high-level stakeholder commission to create a consensus on goals and objectives of the National Health Information Infrastructure (NHII) and develop a comprehensive strategy for adoption and implementation of voluntary standards.
Ms. Doyle predicted that the initial standards of HIPAA would “take place “over the next five years, considering to the time to get the rule promulgated and the 24-month implementation period and compliance. She observed that the Patient Safety Bill had been approved by the House and was pending full Senate consideration, the Medicare Reform Bill was in conference, and both were viewed in piecemeal isolation. Contending that, if the bills passed along with ICD-10 recommendations, the demands on the industry and HHS to develop standards would not be feasible, she called for shared understanding of where they wanted to be in 15 years, their limited resources, and how to get there.
Discussion
Dr. Huff said he saw more differences than similarities in the HIPAA, Y2K and code set processes. The flow of information was different with HIPAA and the process did not change. With Y2K, every application used dates, but in his institution the code set change was reflected in only several systems. Contending that changes in multiple systems would be pervasive and that the diagnosis codes underpinned the health care system, Ms. Doyle wondered about all the systems it would touch and the number of reports that would have to be reconsidered.
Noting that differences between ICD-9, -9-CM and -10-CM were greater than when they switched to ICD-9 in 1978 (even then there were enough code changes that the entire pricing study had to be repeated), Ms. Doyle cautioned that building DRG values would be necessary and that, in order to ensure fairness to providers and payers and solvency of the health care system, rebuilding had to be done off-line in a parallel system and would be a tremendous effort.
Mr. Blair concurred with the need for a national cohesive strategy for NHII, noting NCVHS put forth a proposal. Leadership, coordination, and architecture were needed. He cited an item in the RAND report indicating benefits might be substantial. While it was difficult to quantify and put a cost on improving quality and patient safety, he said that, as both improved, there could be significant cost savings. Noting that costs were based, so far, on very short periods of estimating, he suggested that, as people accommodated, costs might change dramatically. Mr. Blair noted the need to do a better job of quantifying benefits and that no one had attempted to identify or quantify the cost of not acting.
Noting a lack of consensus on whether or how to go forward, members observed that the NPRM process would provide additional learning and the possibility of a mandate would be four-to-six years away. The cost of not beginning an NPRM now could be high; the cost of a later transition higher. Ms. Doyle noted serious reservations about the appropriate sequencing of moving forward, whether benefits of ICD-10 were valid or alternatives fully explored, and the lack of consensus and commitment among the industry about how this fit an overall strategy. Noting the lack of sufficient crosswalks and doubting that an NPRM process would provide much information needed, Ms. Doyle said going forward was premature. BCBSA supported creating a high-level commission to agree upon 10-to-15 year strategy on moving forward on information technology standards that would report back within 18 months. She suggested continuing studies on the appropriateness of ICD-10 and whether it solved issues, emphasizing it should be part of an overall strategy.
Dr. Fitzmaurice said that with NHII the Committee took on a good part of the responsibility Ms. Doyle called for. Following that vision, the Subcommittee encouraged development of tools providing those benefits. He noted hearing the need for more precision in measuring health care inputs and what those inputs did, patient conditions and disease categories, addressing problems to improve public health, examining DRG’s efficiency for payment based upon more precise diagnostic and cost data, the quality of care and patient safety, resource use and incentives for the wise purchasing use of health care.
Noting the difference between what RAND and Nolan projected as benefits and savings, Ms. Doyle called for a critical analysis of the benefits and further discussion and examination of whether patient safety came because of coding, another behavioral change within the physician community, or information. Remarking that implementation would take considerable financial resources and that the only way to get there was for high-level decision makers to share that vision, she emphasized that Capitol Hill needed to share in the consensus.
Noting literature from the UK did not cite improvements in patient safety linked to the quality of data and a decade of writing on patient safety said little about the data’s importance, Mr. Smith suggested experimenting in a specialty within a defined area to determine the benefits of finely grained data. He said the main way to minimize disruption and cost of adoption of ICD-10 today would be to reduce uncertainty about the implementation schedule. He noted the study indicated that sequences of task and events a facility had to do for HIPAA and for ICD-10 implementation were similar. Anything that increased confidence in the fidelity of the crosswalk might encourage organizations to use them rather than expensive, time consuming testing and retesting.
Ms. Doyle emphasized learning lessons from the initial HIPAA before moving forward on implementation of additional standards. BCBSA advocated a credible cost benefit analysis, pilot test programs, ironing out issues before implementation, and a large-scale educational process. Ms. Frasier conveyed the American Academy of Professional Coders belief in the need for greater specificity in order to better understand conditions, symptoms, and diagnosis. Mr. Smith clarified that, if the missing entities that hadn’t been accessible were included, the estimated cost would be closer to $14 billion dollars. Mr. Rode advised beginning now and implementing over three years, realizing that today’s costs will be lower then when a much larger electronic health system is in place. Mr. Argus urged NCVHS to move forward with the NPRM.
Tom Wilder cautioned that ICD-10’s additional codes and complexity would cause problems the Subcommittee had to address. Noting that HIPAA taught that it took longer and cost more than estimated, he conveyed AAHP/HIAA members’ caution. He emphasized that payers would not focus on implementation until a final rule came out. Under HIPAA, they would then have two years to “get their ducks in a row.” Members said HIPAA had shown that was not enough time.
ICD-10 NEXT STEPS: TESTIMONY AND PUBLIC COMMENT - Michael Lundberg, Vice Chair, Board of Directors, National Association of Health Data Organizations
Mr. Lundberg said the National Association of Health Data Organizations (NAHDO) - a non-profit membership and educational organization promoting public availability of health care data and improved state-wide health care surveillance systems - recommended an initial dual approach to procedure coding with ICD-10-PCS for hospital inpatient services and CPT -4 for outpatient while an index and cross reference system is developed for patient services. He said together they would capture greater specificity and clinical detail and new technological procedures. NAHDO urged NCVHS to announce a two-year voluntary transition process from ICD-9-CM to ICD-10-CM and ICD-10-PCS and a defined date for adoption.
Data collection costs include a one-time cost of changes to data systems, groupers and software as well as information technology and vendor costs. The move to ICD-10 system would impact data retrieval, necessitating crosswalks, and increase information system personnel’s workload. Vendors and IT staff had to modify existing audit and edit systems. There were also analytic costs. People costs include the need to prepare, train and educate coding personnel across the health care industry. Radiologists, pathologists, and other physicians and clinicians needed to know documentation requirements for accuracy and completeness about the patients’ health care.
Mr. Lundberg said benefits of a transition to ICD-10 include codes that lend more specific data for outcomes studies on acute myocardial infarctions and trending conditions in cerebral vascular disease. ICD-10-PCS would provide a marriage of dual procedure systems, with structure to capture new technology, greater specificity and detail and new medical services and technologies. The ICD classification system would be more consistent and uniform with HIPAA standards and provide comparability with international data and state/national mortality data.
Mr. Lundberg noted implementation’s financial impact on the outpatient health industry. ICD-10-PCS will impose new medical terminology, reassignment of names and conventions to procedures, and an index/cross reference system. He said ICD-10-PCS could be phased in over time for all services, eliminating duplicative efforts of dual procedural classifications.
Noting that the National Uniform Billing Committee was in the process of approving the next version of the paper billing form, UB02, he said NAHDO believed new forms should support ICD-10-CM and -PCS coding.
Discussion
Asked the estimated cost of converting state hospital discharge data systems from ICD-9-CM to ICD-10, Mr. Lundberg said the cost of converting to a new grouper and additional fields would be minimal. He noted there were trending issues every time CMS made significant changes to the DRG system; transition to ICD-10-CM and the resulting grouper would be more significant than annual changes, but he expressed confidence that it could be overcome. He noted that Virginia uses two grouper systems, including a proprietary system that changes every three-to-five years for severity: adding it was part of doing business and the net gain was much greater for outcomes measurement.
Mr. Lundberg said NAHDO supported a mandatory move. Mandatory systems gave providers a breather: they knew they had to do, and with everyone following the same rules there was comparability and, ultimately, compliance. He added that having a reasonable transition period and listening to industry about the issues and addressing them were critical.
Asked what might ease transition, Mr. Lundberg said issues had to be addressed proactively. He urged the Subcommittee to do a post mortem to understand why software wasn’t there when payers might have been ready to implement changes associated with HIPAA. Emphasizing that a vendor developing a grouper system or way to handle edits associated with codes could not move without direction from the federal government or those adopting ICD-10, he said identifying and minimizing such situations would reduce costs and smooth the process.
Mr. Lundberg reflected that the health system’s reliance on multiple things (hospitals had multiple layers of vendors associated with administration of data collection, submission and retrieval of payment; payers dealt with hundreds of different providers) made transition expensive and a landmine. Asked if there would be efficiencies for hospitals in consolidating data with a compliance data vendor that supplied requesters with the data in appropriate formats, Mr. Lundberg said this might be a challenge because the data was collected for different needs. He said he believed vendors would “step up to the plate” and, because ICD-10 had been out internationally for years, applications probably could be found fairly quickly.
Asked why NAHDO supported 10-PCS as well as CPT-4 for the procedure component, Mr. Lundberg said NAHDO believed that requiring adoption of ICD-10-PCS for physician components where payments were separate and different was a burden at this point, given the lack of information and huge conversion cost to many small providers.
Mr. Rode suggested that they were down to four decisions. Whether to recommend and adopt ICD-10-CM, considering they heard it was time to look at how they would handle all the uses of data (for public health, research, clinical and managed care as well as reimbursement). Considering ICD-10- PCS, he said the ability of tertiary care in educational organizations needed the detail PCS provided for external purposes (e.g., reporting to Medicare, merging a ICD-10-CM/ PCS system that enables adjusting severity outcomes) as well as internal. Regarding the question of doing ICD-10-PCS and -CM together, he noted no one suggested doing them apart. As to the question of costs, he said, one way or another, there would be further costs – and they would trickle down. AHIMA counseled the Committee to recommend looking at the purpose for the cost and how to invest in the future. Emphasizing that for now and years to come there were no alternatives for the aggregate terminology system ICD-10-CM and -PCS presented, Mr. Rode said ability to move forward with PMRI and build an electronic health record system depended on how this issue was addressed.
DISCUSSION OF ISSUES AND DRAFT ICD-10 RECOMMENDATION LETTER – Subcommittee
Noting that the RAND report was a “remarkably candid and useful document” that portrayed the issues the industry faced and could be used in the impact analysis if they went to an NPRM, members voted unanimously to accept it as presented to the full Committee at the September meeting and forward the report and their letter with modifications for clarity and grammar to the full Committee for the November 5-6 meeting. Dr. Huff abstained due to a past consulting relationship with 3M. Noting that the Subcommittee had done what they could do to elicit further input and clarity from the industry and still there was no consensus,. Dr. Cohn thanked everyone for moving the issue forward.
- DAY THREE -
PATIENT MEDICAL RECORD INFORMATION STANDARDS: FINAL RECOMMENDATIONS ON CORE SET OF TERMINOLOGIES - Walter Sujansky, M.D.
Mr. Blair thanked Dr. Sujansky and Dr. Steindel for pulling together the PMRI terminology recommendation letter and explained that last August the Subcommittee reviewed the beginning of the second draft, considering recommendations on SNOMED and LOINC, but found issues with federal drug terminologies that they “couldn’t resolve at that time” and hoped to come to consensus on that day. Members voted unanimously to move the letter with minor word and grammatical modifications to the November 5-6 full Committee meeting. Dr. McDonald and Dr. Huff abstained because of LOINC.
PLANNING FOR UPCOMING MEETINGS - Simon Cohn, M.D.
Remarking that there would be wordsmithing on the ICD, PMRI, and CHI letters, Dr. Cohn said, hopefully, by the end of the three-day meeting the letters would be sent to the rest of the full Committee for review before the November meeting. Noting Ms. Pickett had additional work to do on appendices for the CHI letter, he said they would be sent to the meeting as attachments. The first priorities for the Subcommittee’s November meeting were to deal with any full Committee input on the letters and additional CHI final recommendations, which he noted could trigger more modification to the CHI letter or a sequel. Members noted a process issue. Preliminary recommendations had to be considered first; finals could “drop back” to December, but preliminaries had to be presented before they could become final.
Noting a proposed rule on claims attachments might come out early in 2004, Dr. Cohn said the December 9-10 hearings would focus on CHI; claims attachments, CDA and how both work together; and pieces related to a document issue. The agenda for the January 27-28 Subcommittee meeting includes the security final rule, the DSMO’s yearly report, dental content relating to the PMRI, and initial discussions about lessons learned from HIPAA. Hearings are scheduled for March 30-31 and May 25-26.
Thanking everyone for three days of hard work and three major deliverables, Dr. Cohn adjourned the meeting.
I hereby certify that, to the best of my knowledge, the foregoing
summary of minutes is accurate and complete.
/s/ June 10, 2004
____________________________________________________
Chair Date