Washington, D.C.
The Subcommittee on Standards and Security of the National Committee on Vital and Health Statistics (NCVHS) held hearings on April 9 and 10, 2002, at the Hubert H. Humphrey Building in Washington, D.C.
April 9-10, 2002
On April 9-10 the Subcommittee on Standards and Security held the second of several hearings planned for 2002 on code set issues in an ongoing process focused on administrative simplification provisions of the Health Insurance Portability and Accountability Act (HIPAA). The focus was on the HIPAA medical data code sets and whether there was a need to expand the list of codes or replace some of the code sets. During the first day, the Subcommittee heard ten presentations and held discussions concerning the possible replacement of ICD-9-CM vol. 3 with ICD-10-PCS. The second day, eight presenters testified on previously identified gaps in code sets. Participants also discussed issues regarding the American Dental Association (ADA) raised at the February meeting.
Possible Replacement of ICD-9-CM vol. 3 (Procedures) with ICD-10-PCS
Mr. Gustafson said the World Health Organization (WHO) created the International Classification of Diseases, 9th Revision and the clinical modification (ICD-9-CM) was developed and implemented in the U.S. in 1979. Medicare introduced the inpatient prospective payment system in 1983, using ICD-9 codes to assign cases to diagnosis related groups (DRGs). A HIPAA final rule designated ICD-9-CM as the national standard for inpatient hospital reporting of diagnosis and procedures. Chosen largely because it was the system in use, ICD-9-CM was useful for research and payment, but structural limitations impeded its ability to keep up with the rapidly changing clinical environment. ICD-10-PCS was initiated as a possible successor and met NCVHS's criteria for development. A public hearing last May addressed this possibility. Today's hearing was the next step in consideration of ICD-10-PCS. The Committee would make a recommendation to the Secretary, who would decide whether to propose adopting ICD-10-PCS as the national standard.
Overview of ICD-10-PCS
Mr. Averill said ICD-10-PCS evolved in an open development process overseen by a technical advisory panel that represented American Health Information Management Association (AHIMA), American Hospital Association (AHA), American Medical Association (AMA), medical informatics, managed care, and other federal agencies. In November 1993, NCVHS proposed adoption of a single procedure classification system. The panel followed NCVHS's criteria for a procedure coding system (PCS) in developing ICD-10-PCS. Mr. Averill detailed how ICD-10-PCS's structure was based on a seven-character alphanumeric code. Only thirty terms (each with a specific definition) described all types of procedures done in what was traditionally thought of as surgery. The Clinical Data Abstraction Centers (CDAC) performed a testing evaluation and declared that ICD-10-PCS achieved its stated objectives and met the Coordination and Maintenance Committee's criteria. Based on testing and comments, Mr. Averill said ICD-10-PCS achieved its objectives of comprehensiveness, expandability and standardization of terminology and that its multi-axial structure allowed new procedures and technologies to be readily incorporated without overlap.
Summary of May 17-18, 2001 Meeting of the ICD-9-CM CMS Regarding ICD-10-PCS
Ms. Auld summarized the discussion about ICD-10-PCS at the ICD-9-CM Coordination and Maintenance Committee's May 17-18, 2001 meeting. Ten organizations (AMA declined) supported adoption of ICD-10-PCS as a national standard replacing ICD-9-CM vol. 3. Three (Federation of American Hospitals [FAH], AHIMA, and AHA) strongly recommended further testing, evaluation and planning prior to adoption. Participants agreed there were significant limitations in ICD-9-CM vol. 3. They also noted ICD-10-PCS wasn't perfect: its granularity fell short in a few areas and there were substantial departures from existing healthcare code sets. AdvaMed and PRG encouraged implementing immediately. McKesson HBOC and AHA proposed implementing three years after HIPAA implemented, in conjunction with migration to ICD-10-CM. Participants emphasized that implementation of ICD-10-PCS would require significant resources, and they noted the need for a well-defined maintenance and implementation process. AHA sought clear, unambiguous instructions and consistent official coding and reporting guidelines. AHIMA called for a process for developing rules and guidelines and implementing them on the same schedule as the code set. AHA asked that the cooperating parties (AHA, AHIMA, CMS and the National Committee for Health Statistics [NCHS]) continue developing guidelines and clarification. AHIMA and Ingenix argued in favor of a single procedure classification system. CMS stated in the meeting summary that there were no plans to implement ICD-10-PCS as a single PCS.
Panel 1: Possible Replacement of ICD-9-CM vol. 3 (Procedures) with ICD-10-PCs
Nelly Leon-Chisen, Director of Coding and Classification, AmericanHospital Association
Stephen Hull, Associate Vice President, Advanced Medical Technology Association
Alissa Fox, Executive Director for Policy, Blue Cross and Blue Shield Association
Tracy Gordy, M.D., Chair, American Medical Association CPT Editorial Panel
Sue Prophet, RHIA, CCS, Director, Coding Policy and Compliance, American Health Information Management Association
Lenore Whalen, Director of HIM, Providence Healthcare
Ms. Leon-Chisen said the more detailed code assignments made possible by ICD-10-PCS would greatly reduce hospitals' administrative burden. She cautioned that the timeline for implementation should be orchestrated to similarly minimize the burden to providers. AHA supported the hospital information system (HIS) industry's recommendation to implement ICD-10 three years after HIPAA. Noting implementation would be complex and costly, Ms. Leon-Chisen urged the Committee to encourage Medicare and other health plans to accordingly adjust payments. AHA recommended considering ICD-10-PCS, but emphasized it should be tested for all services in all settings before implementation. Ms. Leon-Chisen said no decisions should be made about a potential single PCS until all contending systems were tested for compatibility with existing payment systems. AHA asked the Subcommittee to consider nine implementation issues.
Advanced Medical Technology Association (AdvaMed) supported rapid adoption and implementation of ICD-10-PCS. Mr. Hull noted the system offered advantages in allowing for accurate billing, proper calibration of DRGs over time, and adequate payment for needed hospital services. He expressed concern about the availability and timing of code assignment for inpatient services involving new technologies and urged the Committee to encourage CMS to improve the process. AdvaMed urged CMS to consider assigning new codes for procedures related to technologies undergoing clinical trials. AdvaMed also encouraged CMS to expand the crosswalk table on CMS's Web site to reflect current ICD-9 codes, so everyone could understand how ICD-10-PCS would affect billing procedures. Mr. Hull also asked the panel to consider whether obliging hospital billing staff to learn two systems simultaneously might create training hurdles.
Blue Cross and Blue Shield Association (BCBSA) asked the Committee to postpone recommending conversion to ICD-10-PCS until the healthcare industry implemented HIPAA and completed a thorough analysis of the implications of such a change. Noting that converting to ICD-10 was a monumental undertaking and the Workgroup on Electronic Data Interchange's (WEDI's) white paper inventorying extensive issues involved, Ms. Fox urged the Committee to assemble a multidisciplinary team to identify all the issues. Because clinical codes underpinned virtually everything in the healthcare system, Ms. Fox said these changes not only impacted payment but also medical policies, quality improvement programs, benefit design, fraud and abuse prevention. Assembling a multidisciplinary team, the Committee could properly identify issues.
AMA stated the complexity of ICD-10-PCS coding structure increased the burden CMS/HHS already imposed and implementation would require significant resources (e.g., education of users, computer systems changes, testing of reporting functions, ongoing maintenance, and update process). Dr. Gordy expressed concern that inherent structural problems made commitment imprudent and questionable. Some descriptors were based on a reconfiguration of the standard anatomic organization, requiring a vocabulary “incompatible” with the language healthcare professionals used. User computer systems would have to be modified to accept the increased number and type of codes. ICD-10-PCS's structure and breadth called for more complexity and clinical decision making, and utilizing character positions with embedded meanings entailed substantial clinical input to ensure medical and anatomical accuracy. Dr. Gordy said further study was needed to determine any cost/benefit of ICD-10-PCS and if administrative simplification would be achieved. AMA believed ICD-10-PCS would perpetuate a dual reporting coding compilation and that the Current Procedural Terminology (CPT) code set had universal acceptance and should be used in all sites for reporting procedures.
Ms. Prophet said AHIMA believed the ICD-9-PCS for inpatient services was on the brink of collapse and imminent replacement was necessary. AHIMA believed ICD-10-PCS represented a significant improvement and met NCVHS's criteria for a PCS; designed to replace ICD-9, it was the only viable option to meet the critical, immediate need. AHIMA anticipated that experienced coding professionals would need fifteen content hours of instruction; users who didn't require hands-on coding skills would need four-to-eight-contact-hours. AHIMA called for a single ICD-10-PCS implementation date for all affected payers and providers. While noting that, ideally, ICD-10-CM and -PCS should be implemented simultaneously, she said it was possible to first replace ICD-9-CM PCS, so long as implementation was planned around conversion of both systems. AHIMA opposed stalling implementation of ICD-10-PCS to wait for leadership and funding for a study of a single PCS.
Ms. Whalen said FAH believed ICD-9-CM, vol. 3 wasn't adequate and that providers, payers and Medicare beneficiaries would be well served by conversion to ICD-10-PCS. Ms. Whalen presented FHA's considerations for implementation, noting health information management (HIM) professionals and coding staff required detailed education and training. Education had to be provided to administrative and billing personnel, researchers, and the physician community. Private payers also required training and education. And although others had roles in this process, CMS had to assume responsibility for educating physicians. Multi-hospital systems needed to provide in-house training for coding and administrative personnel and physicians. FHA believed that professional organizations would be instrumental in the successful delivery of education. Ms. Whalen encouraged the Committee to consider a future single classification system for procedural reporting, regardless of the treatment setting.
Subcommittee Discussion of ICD-10-PCS Testimony
The discussions focused on hospital inpatient procedures and replacing ICD-9, vol. 3. There was general agreement to take a wide view and move towards a singular PCS that handled all sites of service, including ambulatory procedures and hospital. Three options for replacing ICD-9 were identified: ICD-10-PCS, CPT, or postponing until they had more information. AMA said it wouldn't take much effort to use CPT for inpatient coding to determine DRGs and that it would redo a 1986 study showing CPT could be used for these determinations. Mr. Blair suggested that AMA present their proposed code system at the May meeting. Members noted that both ICD and CPT needed further evaluation, and that either could converge into a single coding system. Convergence was desirable, but not necessary. They'd heard that they shouldn't begin the change period while working on the existing set of changes, and that a reasonable timeline was two years, with coders and hospitals taking one-to-three years to transition. Noting payers had the heaviest load, Dr. Zubeldia questioned that they could transition in that time. One next step was asking about the implications of changing to ICD-10-PCS or CPT in October 2003. A separate issue was whether the change extended to other procedure codes.
Mr. Blair identified three key questions and a “boundary:” (1) what were the show stoppers with ICD-10-PCS (2) what were their alternatives, (3) what were the tiebreakers, and (4) setting a realistic time frame for the study and making the best possible decision. Dr. Zubeldia said six months was ample if the scope was ample for considering replacing ICD-9, vol. 3; evaluating a common PCS would take longer. Mr. Blair said he was inclined to limit the scope; which didn't preclude broadening it in the future. Dr. Fitzmaurice suggested setting aside time in October to hear about costs and benefits of using either system for inpatient coding and the cost of a single procedural code set. The Subcommittee could then synthesize and make their recommendation by February. Ms. Leon-Chisen clarified that the only section requiring further study was obstetrics, if only ICD-9-CM, vol. 3 procedures for inpatient was replaced. Noting essential codes were sprinkled throughout, Dr. Zubeldia said they couldn't “carve out” a chapter. Dr. Cohn noted the crosswalk needed to be updated, and the process had to be open for people to see and comment.
BCBSA, AMA AHA, and HIAA discussed investing and participating in a peer study; the Subcommittee would come up with a set of questions. Dr. Yasnoff suggested that the study the Committee commissioned examine the relative viability of 10-PCS vs. CPT based on the guiding principles the Subcommittee detailed in its 2000 report. Ms. Greenberg noted the Subcommittee's report on desirable criteria for a single procedure classification system and funds for an independent study were also available. The Subcommittee staff and Ms. Greenberg would identify possible resources and a design for the independent study. The Subcommittee would then begin to address the implementation problems Ms. Fox identified. Dr. Zubeldia said time could be allocated at the May meeting to decide whether to take the diagnosis and procedure codes together or separate. Dr. Cohn cited the need for a project plan and an overall scope.
Future Agendas - Issues Carried Over from February 6-7, 2002 NCVHS Meeting
Members reviewed topics stemming from the February hearings (some carryovers of comments made as part of the notice of proposed rule making [NPRM] and the final rule) suggested for the May meeting. Issues included: potential replacement of ICD-9-CM vol. 1-2 (diagnosis) with ICD-10-CM; open and public process for code sets; and how a criterion of peer review made it difficult, if not impossible, for new codes (e.g., CAM) to become part of the system. Other concerns included a process for updating the alphanumeric HCPCS and CPT-4 codes, identifying users of local codes to pay for home procedures who hadn't sought resolution, coding and reporting guidelines, frequency of the code set update (and the related administrative burden), problems with AMA's new Category 3 codes, gaps created because some code sets adopted as part of the initial code set weren't applicable to billing for supplies, and recognizing the NCPDP standard for billing professional pharmacy services. Dr. Cohn noted they'd look at the open issue related to cost in the second half of the year; compliance would also become an issue, and they'd need time for PMI next steps and thinking about the administrative and clinical procedure code sets.
Clinical Drugs - Project Overview
Ms. Humphreys reported for Dr. Stuart Nelson, National Library of Medicine (NLM), on the Clinical Drugs project, noting that in a previous Subcommittee discussion about NDC it became clear there was a missing code set of control vocabulary--the product level (e.g., a 50- or 100-count bottle). The provider prescribing often didn't have this information, which grossly complicated the notion of automated decision support reminders on drug-drug interaction checking. Ms. Humphreys illustrated the scale of missed synonymy: while FDA had approved 10,000 drugs, NLM's UMLS system had 81,000 individual clinical drug listings. HL7 was working on a standard form representing what was known when the drug was prescribed. The Food and Drug Administration (FDA), the Veterans Administration (VA), and NLM coalesced around developing a standard representation, the RX normal form, which included each ingredient, strength and dose form. This normalized form would be created for all approved drugs and the UMLS system would have separate concepts in it for ingredients, drug components (ingredient and strength), dose forms, drug formulations, and labeled relationships between concepts. The goal was to map every form of a clinical drug from the VA's National Drug Formulary, connecting them to this RX normal form.
Panel 2: Previously Identified Gaps in Code Sets
American Psychiatric Association (APA) requested that DSM-4 be designated as the code set for the diagnosis and reporting of mental disorders. DSM-4 was used nationally by virtually all providers of mental healthcare for communication, clinical management and record keeping. Over 650 federal and state statutes and regulations relied on or incorporated DSM diagnostic criteria. DSM diagnostic codes corresponded to more-or-less equivalent categories in ICD-9-CM. Many diagnostic terms were identical. The key difference was in the definitions of the disorders. DSM-4 provided operationalized diagnostic criteria using straightforward terminology easily assessed for each mental disorder. Dr. Mirin said ICD-9-CM had a brief (never updated) glossary of definitions, using vague, subjective, archaic terminology written in the 1970s. HIPAA regulations would require mental health practitioners, agencies and institutions to abandon DSM-4 diagnostic criteria in favor of definitions reflecting medical science in 1977. Dr. Mirin noted DSM-4 was widely used by providers and federal and state governments and satisfied the ten specified principles. Used for medical, reimbursement and forensic purposes, DSM would improve uniformity, meet healthcare provider standards, lower implementation costs, eliminate development costs, and have APA's ongoing support in maintaining the code set.
Dr. Kaplan explained that private payors had covered home infusion (the at-home-or-ambulatory management of specialty (often IV) medications and nutritional solutions for patients, by a specialized non-traditional licensed pharmacy) for nearly 20 years to reduce inpatient costs, yet there was no explicit Medicare coverage. Almost every payor had its own codes. National Home Infusion Association's (NHIA's) coding system followed per-diem methodology and was designed for use under the ASC X12 Professional Claims Standard. It wasn't approved under the HIPAA rules, but HHS recommended adoption of additional standards to fill existing coding gaps needed to meet home infusion's business needs. NHIA felt Home Infusion EDI Coalition (HIEC) fit appropriate criteria, however the Designated Standards Maintenance Organizations (DSMOs) rejected their request as they had not yet tried to meet their needs with the existing code set maintainer.. NHIA was directed to work with HCPCS and CPT maintainers to address the gaps and told to return to the DSMO process if that failed. Noting payors and providers were hesitant to convert to codes termed temporary, she recommended renaming the S codes to reflect permanent status. Dr. Kaplan said AMA developed home infusion coding without public input and published them over industry opposition. NHIA was distressed at how much “dogged persistence” it took to participate, however she did indicate that the vast majority of their coding needs have been met by HCPCS. Dr. Kaplan said they'd submitted considerable supportive documentation, but when they discussed this with the NUCC, members hadn't been prepared to discuss the request or why it was subsequently denied. Attempts to get transcripts or minutes explaining the decisions were unsuccessful. Dr. Kaplan said NHIA, which represented the home infusion industry, had no formal role in this process. Significant decisions were being made by entities lacking expertise in home infusion.
Ms. Giannini said ABCcodes were designed specifically for HIPAA compliance and addressing this rapidly expanding, critical gap in codes for alternative medicine, nursing and other non-physician services needed by more than 2.7 million licensed healthcare providers inadequately served by existing HIPAA code sets. The code set met the ten guiding principles outlined in the 1998 proposed rules. DSMOS hadn't recommended ABCcodes as a standard code set, but redirected Alternative Link to approach AMA and HCPCS to determine whether ABCcodes could be managed as part of HCPCS's Level 1 and 2 codes. She said Alternative Link complied, even though AMA had been consistently uncooperative and heavy-handed. Ms. Giannini concluded that ABCcodes filled a critical gap in coding, supported congressional intent and included intelligence that current HIPAA code sets and proposed ICD-10 codes weren't designed to provide. ABCcodes would provide critical information needed for benefit plan design, utilization management, claims management, and appropriate reimbursement as well as actuarial analysis and outcomes and health services research. She said Alternative Link would map ABCcodes and modifiers to a single national code set NCVHS selected.
Mr. Shugart testified on behalf of the State of Maryland Medicaid program and the National Medicaid EDI HIPAA (NMEH) workgroup on gaps in the current HIPAA medical code sets. In Maryland, any billing for alternative medicine services used local codes. Home infusion services were processed electronically through the pharmacy Point of Sales system. Claims were submitted in NCPDP 3.1 format using NDCs; plans called for processing through the HIPAA compliant NCPDP 5.1 transaction set. Other states used coding structures similar to the HIEC set within their local code schemes. In Maryland, mental health claims were submitted using ICD-9 diagnoses. Professional services were billed using CPT-4, HCPCS, or local codes; institutional services were billed using UB-92. At least one state used DSM diagnoses and would face gaps when it implemented ICD-9 diagnoses, because DSM codes couldn't be added to the standard transactions. Mr. Shugart said codes for acupuncture, massage therapy or naturopathic medicine would be a welcome addition, but he noted many codes in those code sets seemed complex and might not be useful. He said it would be easier to expeditiously add the codes for home infusion to the CPT or HCPCS code sets. Mr. Shugart said he'd received comments asking for DSM to be included in the ICD-10 structure after ICD-10 was adopted. He recommended implementing ICD-10 diagnoses two-to-five years after the HIPAA implementation date. Noting many states used ICD-9-PCS on outpatient claims, Mr. Shugart asked to use this code set on standard outpatient transactions. He also recommended that, if ICD-10 procedures was selected, implementation occur when ICD-10 diagnosis was implemented.
Ms. Leon-Chisen said hospitals and health systems were opposed to diverse reporting requirements for each of the clinical codes considered by the Subcommittee. AHA saw no gaps in the current HIPAA code sets that couldn't be addressed by ICD, CPT, HCPCS and CDT. AHA supported the current coordination and maintenance process for ICD-9-CM diagnostic and procedural codes and was encouraged by the AMA's CPT editorial panel's recent efforts to address AHA members' needs. But Ms. Leon-Chisen expressed concern over the lack of institutional provider input into the HCPCS Level 2 update process. AHA believed that ICD-9-CM met the needs of hospitals and health systems for reporting mental health diagnoses and conditions. AHA was sympathetic to needs of providers requiring a way to codify and classify alternative medicine services for other non-reimbursement related needs (e.g., research, benchmarking and public health tracking). She predicted that hospitals and other providers would report therapies considered complementary medicine as they became more widely adopted or covered by traditional insurance. AHA urged providers of CAM services to develop and present code proposals to the maintainers of HIPAA medical code set standards. Ms. Leon-Chisen said hospital and health systems that provided home infusion procedures through their home care services hadn't reported a gap in the current HIPAA code sets. AHA was encouraged that a substantial number of new home infusion codes were implemented in CPT-4 and HCPCS for 2002; if additional gaps in home infusion procedures were identified, Ms. Leon-Chisen said AHA would support adding those services into existing HIPAA medical data code sets.
Mr. Michael Beebe said AMA believed the current process for the selection of standards, transactions, and code set standards under HIPAA was successful. AMA agreed that the addition of new medical code sets would be disruptive and confusing to users, and that it was important that gaps in the existing data code sets first be remedied by the existing standards. He said AMA and the CPT editorial panel recognized a responsibility as a national medical data code set to work with other groups, and that they'd taken steps to expedite review and include new groups. CPT code change proposals, instructions for submission and a schedule with submission deadlines were available on the AMA CPT Web site. Mr. Beebe encouraged all current and prospective users to work with the panel to correct any shortcomings in CPT. He noted panel meetings were now open to all participating national medical specialty societies and the Healthcare Professionals Advisory Committee and that the panel was reviewing options to further open deliberations and advisory input, while maintaining the rigorous peer review process and confidentiality they believed essential in preserving the integrity of the editorial process and maintaining the industry-wide annual update schedule. The panel was also working with payors and the home health community to prepare the CPT code set for elimination of local codes.
Dr. McCabe expressed three professional organizations' belief that a substantial gap in the current code set recognized by HIPAA made it impossible for psychiatric nurses to document their care accurately or trace health outcomes and culminated in care disconnected from outcome measurement, making it difficult to control costs of care for patients. Models of healthcare based on advanced practice psychiatric nursing had been demonstrated to be efficacious, cost effective and well received by consumers, but were poorly reflected in the current HIPAA code sets. DSM-4 codes were modern, empirically tested, current, and gave detailed criteria for making a diagnosis not available with ICD-9 codes. Yet Dr. McCabe noted adoption of DSM-4 codes didn't resolve the issue of physician-oriented code sets, and still left psychiatric nurses with no way to code many health maintenance and promotion care practices. All three professional organizations believed: a significant gap existed in the currently recognized code sets, changes in those code sets were required, and psychiatric nurses needed to be included in the planning as new codes were defined and put in place.
Ms. Ward noted the DSMO Steering Committee included members of the National Uniform Billing Committee (NUBC), National Uniform Claim Committee (NUCC), ADA's Dental Content Committee (DeCC), ASC X12N, HL7, and NCPDP. For sixteen months, the steering committee has been addressing requests for changes to the transaction standards adopted by HHS for HIPAA. Several requests were for new code sets as standards under HIPAA. Ms. Ward said the steering committee recognized the benefit of a more structured process for considering additional code sets as HIPAA standards and believed it was being responsive to this need while adjudicating the monthly batches of requests. The DSMO was developing a guidelines outlining steps for submitters seeking to have new code sets adopted under HIPAA., The guidelines will include information about the process to follow, other organizations that may play a role in the process, , and preparatory questions to answer before submitting into the change request system (CRS). A liaison between the steering committee and each HIPAA code set maintainer would also be established. The DSMOs aimed to have this available to the public in May.
Subcommittee Discussion of Testimony on Gaps in Code Sets
Members explored the question of whether DSM could be a set of guidelines on how to code ICD, and decided to continue this discussion at the next hearing. They noted the need to determine whether resolution was underway to the problem of overlapping and conflicting codes in CPT and HCPCS because of a disconnect due to both systems being simultaneously changed. Noting NHIA, pharmacy, and other healthcare sectors lacked representation in the HCPCS and CPT process, members remarked on the need to ask code set developers how they proposed to establish full representation and interaction with domain expertise. Alternative Link was encouraged to work with the CPT editorial panel and the HCPCS code committee and bring a report to the May Subcommittee meeting. Dr. Zubeldia pointed out that Alternative Link could use the code set today for public health or actuarial services. Ms. Molina said most stakeholders looked to the Subcommittee and the DSMO to help them assess whether an investment in the capabilities of the code set was appropriate. Mr. Beebe said, hopefully, AMA would have demonstrated progress on developing code change proposals for distinct portions of the alternative medicine codes by the time he gave the May progress report.
Issues Regarding ADA Raised at February 6-7 2002 NCVHS Meeting
Frank Pokorny, Manager Code and Third Party Issues, American Dental Association, updated members on ADA issues raised in February. Ms. Trudel recounted that the previous testimony raised what Mr. Pokorny categorized as four key issues: participants, meeting protocol, process for future code review and revision, and access costs for CDT. Mr. Ward noted the dispute over what happened at the meeting and the procedures showed the lack of a necessary structure: they needed to determine the procedures and make sure they worked together. Mr. Phil O'Brien, General Counsel for DDPA, said they were open to changing the fifth payor “rep” on the committee and that they wouldn't oppose NADP as representative. Jay Grant, Washington counsel for NADP, confirmed that NADP was willing to participate. The Subcommittee noted they'd like to hear when this was resolved. Mr. Pokorny said he would report on progress. Noting the issues seemed to mostly involve governance, Mr. Blair suggested considering identifying an independent secretariat that all sides felt comfortable with. Dr. Zubeldia said he hoped that suggestion wouldn't be necessary; with a little collaboration, those directly involved might be able to work out the friction.
Possible Replacement of ICD-9-CM vol. 3 (Procedures) with ICD-10-PCS
Mr. Gustafson said the Purchasing Policy Group handled payment policy for the acute care portion of the Medicare program; ICD-9-CM coding efforts of HCFA; and CPT coding efforts of HCFA. WHO created ICD-9 and the clinical modification developed in the U.S. added specificity to the diagnosis codes and a procedure coding section implemented in 1979. Medicare introduced the inpatient prospective payment system in 1983, using ICD-9 codes to assign cases to DRGs used in determining payments. HHS created the ICD-9-CM Coordination and Maintenance Committee in 1985 to maintain the currency of the code set as the healthcare environment changes. The committee meets biannually to discuss proposed revisions to the codes in an open, public process. Mr. Gustafson noted the process was viewed as a successful model. HIPAA had an administrative simplification title that created a national process for establishing standards for coding systems and designating standard code sets. The goal was to improve efficiency of the entire system by enabling electronic exchange of information in standard formats. A final rule published in August 2000, designated ICD-9-CM as the national standard for inpatient hospital reporting of diagnosis and procedures. Chosen largely because it was the system in use, ICD-9-CM was useful for research and payment, but structural limitations impeded its ability to keep up with the rapidly changing clinical environment.
ICD procedure codes utilized four numeric digits. The first digit designated categories in a hierarchical structure based on organ systems, allowing for automatic grouping useful for research and analysis of payments. Trailing digits specified the procedure. Ten codes generally were available per category. One had to jump elsewhere in the numeric sequence when more codes were needed, disrupting the groupings and confusing coders and analysts. Mr. Gustafson compared ICD-9 to a library with full shelves that stymied its collection's growth. The basic categorization was 23 years old and the substantial number of procedures invented since the categories were set were now utilized extensively in the healthcare community. Updating this inelastic system had become difficult and required compromises that degraded its efficiency for research and payment purposes.
ICD-10-PCS was initiated as a possible successor to ICD-9-CM. NCVHS established criteria for development, stating the system should be: complete, expandable to readily accommodate new procedures and technologies, employ standardized terminology, multi-axial so each code character had a common meaning, and the procedure description should be limited and not rely on diagnostic information. ICD-10-PCS was developed by 3M Health Information Systems (3M HIS) under contract with the Centers for Medicaid and Medicare Services (CMMS). The development process was open and included a technical advisory panel available throughout the process. The Coordination and Maintenance Committee provided frequent updates. ICD-10-PCS met the criteria, using seven alphanumeric characters that provide greater capacity to give an order of magnitude. The current draft of ICD-10-PCS had nearly 200,000 procedures (ICD-9-CM currently designated less than 4,000)) and there was ample room for expansion. The final draft released in 2000 incorporated results based on Center for Medicare Services (CMS) and 3M HIS's multiple field tests. A training manual and ancillary products were available. A public hearing held last May addressed the possibility of replacing ICD-9-CM with ICD-10-PCS. Under HIPAA, the next step in consideration of this new system was today's hearing. Mr. Gustafson said they would consider the costs and benefits of staying with the current system vs. those of moving to a new system, particularly ICD-10-PCS. The Committee would make a recommendation to the Secretary, who would decide whether to propose adopting ICD-10-PCS as the national standard.
Discussion
Dr. Fitzmaurice asked if the desirable criteria Mr. Gustafson outlined for a PCS applied to ambulatory procedures. Mr. Gustafson said, in the abstract, they were desirable features of any coding system. He reported he didn't know any other PCS that could be used for institutional medical procedures. Ms. Greenberg recalled that those criteria, laid out in a report developed by the Committee on a single procedure classification system, weren't specific to institutional. Mr. Gustafson observed that the Purchasing Policy Group was reminded daily of the desirability of having more codes available. The agency was under substantial pressure from those paid and the drug and device industries for a differentiated payment system that allowed them to identify and track when particular items (e.g., a particular pacemaker or drug) in the inpatient setting were used, in order to facilitate differentiated payments.
Dr. Cohn noted the Subcommittee was looking for someone with a broad view within CMS who could help with broader issues related to gaps in the procedure code sets. The next day, they'd hear about proposed code sets (e.g., alternative services) and they sought guidance from CMS about issues related to needs for codes for identification. Dr. Cohn asked Mr. Gustafson if they should follow up with him on issues he'd identified around supply and product codes. Mr. Gustafson noted Ms. Trudel was a lead in many areas. He said he was a resource for acute care payment policy. His colleague, Tom Hoyer dealt with chronic care policy and issues related to home health, skilled nursing facilities, and ESRD. Noting Dr. Zubeldia expressed a desire for members to better understand payment policy issues, Dr. Cohn said a session with both would be welcomed. Mr. Gustafson said he wasn't aware of a specific source within Medicaid, but noted a third portion of the coding world, related to Healthcare Common Procedure Coding System (HCPCS) Level 3 codes and durable medical equipment (DME) reported to Mr. Hoyer.
Mr. Gustafson explained that a portion of the code set was maintained jointly by several major payor organizations, including CMS. A committee, which included representatives of state Medicaid agencies, advised them. There wasn't a separate Medicaid presence, but CMS (which actually was CMMS) included both Medicare and Medicaid interests. Insofar as they dealt with CPT or ICD-9, they primarily served as a voice for Medicaid.
Dr. Fitzmaurice remarked on how the states worked together to winnow down 30,000 local codes to under a 1,000, reducing duplication and linking with CPT and other codes so they meant the same thing across states. They'd heard testimony that CMS was active in the application of this consolidation. Dr. Fitzmaurice credited the Purchasing Policy Group for what he considered a grand result for the medical informatics community to witness.
Overview of ICD-10-PCS
Mr. Averill shared his perspective as the principal investigator on the CMS contract to develop ICD-10-PCS, developed as a replacement for ICD-9-CM procedure codes used for inpatient reporting. A contract for preliminary system design was signed in 1991. CMS let a contract for the complete system in 1995. The first draft was completed in 1998 and multiple field tests were conducted from 1998 to 2000. In November 2000, CMS released the updated version incorporating field-test results. A minor update, based on comments from the field, was done in November 2001. The open development process was overseen by a technical advisory panel, which provided review and comment throughout development. Several members of the committee were on the technical advisory panel. Members represented AHIMA, AHA, AMA, medical informatics, managed care, and other federal agencies. The panel provided frequent updates to the ICD-9-CM Coordination and Maintenance Committee, sought suggestions from both the committee and audience, and strove to incorporate comments in the development process. The entire system was available on the CMS Web site. Many individuals downloaded the system and a number of foreign governments (e.g., Germany) were evaluating moving to ICD-10-PCS.
In November 1993, NCVHS proposed adoption of a single procedure classification system. Mr. Averill said the panel considered characteristics of a PCS recommended by NCVHS, found those criteria thoughtful and insightful, and followed them in developing ICD-10-PCS. Criteria included: (1) comprehensiveness--all procedures classified; (2) non-overlapping--all substantially different procedures had a unique code; (3) expandability--new procedures and technologies could be incorporated as unique codes; (4) hierarchical structure--individual codes could be aggregated into larger categories; (5) multi-axial--each code character had a consistent meaning; (6) standardized terminology--all terminology precisely defined; (7) diagnostic information--not included in the procedure description.
Mr. Averill detailed how the structure of ICD-10-PCS was based on a seven-character alphanumeric code using the digits 0-9 and the letters A-H, J-N, and P-Z. ICD-10-PCS was divided into 16 sections. The medical and surgical section (largest and most familiar) contained traditional surgical procedures. Separate sections covered: obstetrics, placement, administration, measurement and monitoring, imaging, nuclear medicine, radiation oncology, osteopathic, rehabilitation and diagnostic audiology, extracorporeal assistance and performance, extracorporeal therapies, laboratory, mental health, chiropractic, and miscellaneous.
The first character of the alphanumeric code indicated the section--e.g., medical/surgical; second, body system--e.g., heart and great vessels; third, root operation--underlying objective of the procedure; fourth, body part/region; fifth, approach--e.g., how the site was reached (percutaneously, endoscopically, open procedure); sixth, device--e.g., a pacemaker; seventh, a qualifier for additional specificity to any code. Only thirty terms (each with a specific definition) described all types of procedures done in what was traditionally thought of as surgery. Mr. Averill noted the tabular in ICD-10-PCS was set up quite differently than that used in ICD-9-CM. ICD-9 tended to be a list of code numbers; ICD-10 was set up in a true tabular form with the first three characters of the code at the top of each page. If someone had a PTCA on three coronary vessels and a stent put in, ICD-9-CM would utilize two code numbers: 3605 for multiple vessel PTCA or coronary atherectomy and 3606 for insertion of coronary stent(s). ICD-10-PCS would be 02725DZ--dilation three coronary arteries, percutaneous intraluminal with intraluminal device. Code was build up from tabular representation.
A definitions manual was available with computer-generated tabular and index. Mr. Averill explained the table-driven system: one turned on a new cell to add a new code, and the computer automatically built the English description, regenerated tabular, and updated the index. Mr. Averill said generating everything by the same program ensured a comprehensive, consistent index. A user manual, presentation materials, summary articles, and complete mapping back to ICD-9-CM procedures were available on the CMS Web site. Everything was in the public domain.
CDAC performed a testing evaluation. Both centers abstracted and coded 5,000 records in ICD-10-PCS. A subset of the 5,000 records was coded both in ICD-9-CM and ICD-10-PCS. CDAC said ICD-10-PCS achieved its stated objectives and that the Coordination and Maintenance Committee's criteria were met. CDAC reported: (1) coders quickly became proficient in ICD-10-PCS; (2) time to code was the same or shorter and non-specific codes were easier to code in ICD-10-PCS; (3) the precise specificity of ICD-10-PCS made coding faster and provided a more precise, complete clinical description of the procedure; (4) ICD-10-PCS with its definitions and straightforward nature would be at least as, and probably more, accurate than current ICD-9-CM coding. Based on testing and comments, Mr. Averill said ICD-10-PCS achieved its objectives of comprehensiveness, expandability and standardization of terminology and that its multi-axial structure allowed new procedures and technologies to be readily incorporated without overlap.
Discussion
Asked if the panel had tested the feasibility of arranging the ICD-10-PCS codes in clusters similar to useful DRGs, Mr. Averill said 3M HIS was the contractor that updated the DRGs and that much of the initial motivation for developing ICD-10-PCS came from frustration with assigning codes to DRGs. He said mapping to DRGs would be straightforward and satisfying; many problems with DRGs that originated from lack of specificity in the codes could be corrected. With so many codes, the challenge would be keeping the definitions book to a reasonable size.
Noting ICD-10-PCS covered areas beyond the medical/surgical, Dr. Cohn asked how the laboratory portion of ICD-10-PCS related to LOINC, which appeared to be a developing national standard lab terminology. Dr. Goldfield said 3M HIS collaborated with LOINC and did a mapping that had to be updated.
Observing that revenue policy drove a lot of decision making, Dr. Cohn noted some segments of industry hoped any change would bring hospitals additional revenue. He asked if the intent in developing new sets of DRGs was to generate additional revenue or be revenue neutral. Mr. Gustafson said Congress constrained how they set system updates; any change had to be revenue neutral. What could be expected in a new coding system was potential for paying in a more differentiated fashion; under some circumstances, an expansion of the DRGs might enable more accurate payment. But he added there wouldn't be intent to accompany that with any substantial increase in the amount of payments Medicare made in the aggregate. Recalling that Mr. Gustafson had mentioned the desirability of a cost benefit analysis, Dr. Cohn asked if CMS had conducted one. Mr. Gustafson said all parties involved in these deliberations needed to be aware that there were costs and benefits to retaining the current system. They'd heard about the current system's substantial limitations and that moving to a new system required costs on somebody's part. Mr. Averill had pointed out the cost of training for the testing exercise and cost associated with training inpatient coders.
He also remarked that, beyond the reimbursement side, there was the efficacy of procedures and outcomes (e.g., for all practical purposes, the current system provided no information about different therapies applied in radiation oncology). Mr. Averill observed that, while they'd had trouble constraining radiation oncology to roughly 1,200 codes, they could get precise information and track it over time. Currently, they had no information in terms of the national coded database. Beyond payment issues, asking efficacy types of questions and understanding what was done and the trends were crucial for moving forward. Mr. Blair said that Mr. Averill's response to the question on cost benefit seemed focused on the cost to HHS for education and maintaining and updating ICD-10-PCS compared to ICD-9-CM. Mr. Blair said his perception was that much of the cost savings would be for users within the provider and payer communities and clinical research. He urged 3M HIS and others looking at cost benefits to consider the savings to all groups. Mr. Gustafson encouraged the Committee to make a national decision, taking into account the costs, benefits and downsides for everyone. He noted a need to acknowledge costs associated with whatever future course they chose. Staying with the current system, they'd forego opportunities to learn about what happened or gain a more differentiated payment system. Moving to the new payment system, there would be the cost of the virtues Mr. Averill mentioned.
Dr. Zubeldia remarked that with this flexible, comprehensive code system it seemed there would be a complete overlap of the laboratory section of the CPT book. He asked about overlap of the miscellaneous procedure codes covering acupuncture and yoga therapy with other alternative codes for alternative medicine. Mr. Averill said 3M HIS did a fair amount of international work and had scoured all the PCSs to make sure that coded procedures considered important were included. Dr. Goldfield added that 3M HIS worked closely with LOINC and collaborated with American College of Pathology, American Society of Microbiology, and the other colleges and rarified groups of laboratory professionals to gain input on specific disciplines and codes. That development process created de facto overlap with the other systems. Mr. Averill said contractually 3M HIS had been required to develop all these sections; ultimately it would be up to CMS and NCVHS whether LOINC or the laboratory portion of ICD-10-PCS was used. Mr. Averill clarified that, excluding things specifically related to office visits that weren't applicable to an inpatient setting, everything codeable in ICD-9-CM, CPT and most other developed PCSs should be codeable in ICD-10-PCS. Mr. Blair asked how much overlap that caused. Dr. Cohn said historically there'd been a large overlap. Withe ICD-10 vol. 3, there was ability for an in-hospital procedure to be coded by the hospital at their level of specificity and the physician would also use CPT to code the procedure, usually at a slightly more specific level. Mr. Averill observed that the objective of any PCS was to allow the coding of any procedure done. They were trying to create a PCS that allowed any hospital procedure to have a unique code. To the extent that any other coding systems had the same objective, there would be overlap. Dr. Cohn noted Dr. Gordy would talk that afternoon about AMA's view of the overlap.
Returning to the question of ICD-10 as a possible successor for inpatient services, Mr. Gustafson noted laboratories and outpatient services were currently paid using the CPT code. Laboratory payments were split roughly in thirds between hospital-based laboratories, independent laboratories, and physician office labs. Noting the need, in contemplating appropriate code sets or another code for laboratories, to consider the various parties using these code sets, Mr. Gustafson pointed out that ICD-9 might be a simple, obvious answer in a hospital context, but quite different in a practicing physician's office.
Mr. Averill mentioned that individual codes could be aggregated into larger categories, Dr. Fitzmaurice asked about characteristics and hierarchical structure. He said he interpreted that as enumerating fine enough entities for the aggregation and classifying them meaningfully. He observed that if hierarchical structure meant people had to enumerate finely enough for intended purposes, then different levels of enumeration might have to be classified as well. He asked how one made enumeration of individual elements (e.g., individual procedures) and classification serve multiple purposes, or if one had to fall back on different classifications for different purposes. Mr. Averill suggested going back to another criteria that didn't overlap, noting it was this understanding that drove them to make each procedure a unique code. The question was how to put this together structurally to meet an entity's hierarchical and aggregation purposes. Depending on purpose, what was wanted, and granularity, one could aggregate multiple code characters in a hierarchical way to get the desired subset of procedures and level of specificity. One could look at: the second character and get all procedures done on the heart and great vessels, the fourth character and get procedures done in the coronary arteries, or go to the fifth character and focus on all procedures on the coronary arteries, heart and great vessels done endoscopically. Dr. Fitzmaurice reiterated: it wasn't only making the enumeration fine enough, but capturing enough characteristics of that individual element to be useful.
Mr. Blair built on Dr. Fitzmaurice's question: was the focus specifically on reimbursement and statistical purposes or did individuals need to go to a certain level of granularity to improve patient care or clinical research? Mr. Averill said, obviously, payment wasn't the only objective. One would be unlikely to make payments based on many levels of distinction, and they could get by with far fewer codes if their only purpose was payment. But a great deal of input said those levels of specificity were desired for monitoring clinical care for quality evaluation.
Summary of May 17-18, 2001 Meeting of the ICD-9-CM CMS Regarding ICD-10-PCS
Ms. Auld summarized the May 17-18, 2001 meeting of the ICD-9-CM Coordination and Maintenance Committee's discussion about whether or not ICD-10-PCS should be named as a national standard, replacing the current ICD-9-CM vol. 3 (procedure codes) for inpatient hospital use. As both Mr. Gustafson and Mr. Averill cited, significant thought, planning and testing had gone into the development of PCS and the meeting gave the industry a chance to react to 3M HIS' and CMS's work. Eleven organizations testified: Advanced Medical Technology Association (AdvaMed), AHIMA, AHA, AMA, the American Speech-Language Hearing Association (ASHA), DRG Review, FAH, Ingenix Syndicated Content Group, Medical Technology Partners (MTP), McKesson HBOC, and Princeton Reimbursement Group (PRG). Five were testifying today.
Ten organizations (AMA declined) supported adoption of ICD-10-PCS. All agreed implementation would require significant resources. Three (AHIMA, AHA and FAH) strongly recommended further testing, evaluation and planning prior to adoption. Participants agreed with Mr. Gustafson and Mr. Averill that there were significant limitations in ICD-9-CM vol. 3 and discussed how ICD-10-PCS addressed those limitations. People noted that ICD-10-PCS wasn't perfect: its granularity fell short in a few areas and, in some cases, it was a step back from ICD-9-CM. There were substantial departures from existing healthcare code sets. Because it required significant resources, some found it less than perfect.
The time frame for implementing ICD-10-PCS was the most contentious point. Two options were discussed. AdvaMed and PRG encouraged implementing immediately. Citing indications that moving only ICD-10-CM or PCS forward first could cause problems in overlaps and maintaining too many code sets, McKesson HBOC and AHA proposed implementing three years after HIPPA implemented, in conjunction with migration to ICD-10-CM. Issues affecting the time frame were coordination between parties (considered imperative) and additional testing needed prior to implementation.
Participants emphasized that implementation of ICD-10-PCS would require significant resources. Areas requiring resources included: software changes (e.g., redevelopment of related systems including groupers, payment policy and performance measurement systems), crosswalking among coding systems, extensive retraining due to ICD-10-PCS' significant code set differences, and the cost factor Mr. Gustafson detailed.
Participants noted the need for a well-defined maintenance and implementation process. AHA felt the current ICD-9-CM C&M process worked well and should be adopted for PCS. FAH suggested PCS should be updated annually in conjunction with the federal government's fiscal year. AdvaMed suggested quarterly updates; others said more than once a year. Another request was for a minimum time between when a code was proposed and implemented. AHA sought clear, unambiguous instructions and consistent official coding and reporting guidelines. AHIMA called for establishing a process for developing rules and guidelines and implementing both on the same schedule as the code set. AHA asked that the cooperating parties (AHA, AHIMA, CMS, and NCHS continue their role in developing guidelines and clarification. Additional suggestions included: identifying one body within CMS to promptly respond to questions (especially from software developers) regarding implementation, making all materials associated with code sets easily accessible in print and on the Internet, and developing a Web site as a clinical coding resource. Ms. Auld said people could pose questions on the Web site and get responses from CMS and other users. Q&As could be archived, so information built and was available over time.
Noting overlaps in services forced some facilities to use and maintain two PCSs, AHIMA and Ingenix argued in favor of a single procedure classification system. CMS stated in the meeting summary that there were no plans to implement ICD-10-PCS as a single PCS. Testimony and minutes were available on the CMS Web site.
Discussion
Recalling Ms. Auld had reported a discussion about implementing both ICD-10-CM and PCS simultaneously, Dr. Zubeldia asked about the correlation between them. Mr. Averill said there wasn't an intrinsic relationship, but an implementation convenience. If the industry revamped how diagnoses and procedures were coded, it preferred undergoing both changes simultaneously to being disrupted twice. Dr. Fitzmaurice observed there was a national shortage of qualified coders; he asked about studies or pilots that indicated the easiest ways to train a new coder in a particular system. Ms. Pat Brooks, CMS, said it took weeks in school for people to learn procedure aspects: she'd been pleasantly surprised when both sets of CDACs trainees used the system in two-and-a-half days. Dr. Cohn pointed out that these weren't novices but trained coders learning another system. Dr. Fitzmaurice noted that training was another cost differential to consider.
Panel 1: Possible Replacement of ICD-9-CM vol. 3 (Procedures) with ICD-10-PCs
Ms. Leon-Chisen explained that the American Hospital Association's (AHA's) Central Office on ICD-9-CM served as the U.S. clearinghouse for issues related to ICD-9-CM. The central office stemmed from the 1963 memorandum of understanding (MOU) between AHA and HHS. It also published Coding Clinic for ICD-9-CM, the official publication for ICD-9-CM coding guidelines and advice, as designated by the Cooperating Parties (CMS, NCVHS, NCHS, AHA, and AHIMA). She noted that accurate, precise reporting of clinical codes was vital because these codes were key to benchmarking, quality assessment, research, public health reporting, and strategic planning, in addition to accurate reimbursement. ICD-9-CM codes allowed hospitals to develop critical pathways for high-volume, high-risk or high-cost diagnoses and procedures and those with similar courses of treatment. Hospitals could identify high-volume procedures by examining clinical abstracted data and selecting ICD-9-CM codes appearing most frequently.
Ms. Leon-Chisen said ICD-9-CM didn't adequately capture many new procedures and innovations in medical practice. And physical numbering constraints limited an ability to expand enumeration for any procedure category. Consequently, some categories provided vague, imprecise procedure codes. She reported that chapters 00 and 17 were recently opened for the creation of new codes. But, at the current rate (i.e., 24 new codes created this last year) the two chapters wouldn't meet the need for new codes. Because of numbering constraints, ICD-9-CM grouped together under one code a number of distinct procedures performed on different parts of the body with widely incongruent resource utilization (e.g., code 99.29, “Injection or infusion of other therapeutic or prophylactic substance,” encompassed such diverse procedures as: an injection of epinephrine to cauterize a rectal ulcer, infusion of a narcotic into a pump for pain relief, insertion of an eye implant for slow release of an antiviral drug, and injection into the uterine artery to treat a fibroid). Medical records must be pulled and examined to identify patterns or treatment procedures: a researcher studying outcomes of antiviral drugs released via implantation in the eye would first need to examine all 99.29 medical records to identify the subset of patients that received the antiviral drug, a very labor intensive task. Ms. Leon-Chisen said the more detailed code assignments made possible by ICD-10-PCS would greatly reduce hospitals' administrative burden. More detailed codes would ease requirements for submission of additional documentation to support claims, allow capture of accurate data on new medical advances, and provide data to support performance measurement, monitoring resource utilization, outcomes and cost analyses.
Ms. Leon-Chisen cautioned that the timeline for implementing the new system should be orchestrated to minimize providers' administrative burden. Hospitals already faced numerous regulatory changes that, over the next several years, significantly stretched their burden and costs. The ability of hospitals to absorb more regulatory changes had to be considered. The vast majority of hospitals were dependent upon HIS vendors for programming changes. AHA supported their request that the ICD-10-PCS implementation for procedure coding be conducted in tandem with migration through the ICD-10-CM diagnosis codes. The AHA also supported the HIS industry's recommendation to implement ICD-10 three years after HIPAA implementation.
Noting implementation of ICD-10 would be complex and costly, Ms. Leon-Chisen urged the Committee to encourage Medicare and other health plans to be sensitive to increased regulatory costs and accordingly adjust payments. AHA believed the costs of implementing significant new regulations should be worked into the Medicare prospective payment rate updates. AHA also believed Congress should establish grants to help hospitals convert to a new coding system and meet other enormous costs of complying with HIPAA rules.
Ms. Leon-Chisen noted the bulk of the costs associated with the adoption of a new procedure classification system for hospitals involved training personnel. Hospital support staff (coders and billers) had to attend training seminars on new coding guidelines, rules and definitions. Hospitals and their medical staffs had to ensure that appropriate documentation was available to support the new system. ICD-10-PCS code selection required more specific and detailed physician documentation in the medical record, and this greater level of specificity might require that coders and billers expand their knowledge of medical terminology, anatomy, physiology and disease process. Changes to the coding system also required extensive, costly modifications to information systems. Software applications requiring modification encompassed: code assignment, medical records abstraction, aggregate data reporting, utilization management, clinical systems, billing, claim submission, groupers, and other financial functions. In essence, every electronic transaction requiring an ICD-9-CM procedure code had to be changed. Changes included: software interface, field-length formats on screens, table structures holding codes, expansion of flat files, coding edits, significant logic changes, report formats and layouts. Hospitals bore the financial burden associated with software changes and possible hardware upgrades.
Ms. Leon-Chisen emphasized that they would occur these changes moving to any procedure classification system. And during the transition, information systems had to support both PCS and CM coding systems, requiring additional data storage space. Ms. Leon-Chisen noted small, rural healthcare providers in particular (many of whom already faced serious financial challenges with less sophisticated information systems that further handicapped their ability to accommodate change) might require additional resources and support to acquire information and coding system support programs.
AHA worked closely with institutional members in initial informal field testing of ICD-10-PCS. Ms. Leon-Chisen reported that ICD-10-PCS held promise and AHA recommended considering it for future use. However, AHA also emphasized that the new system should be tested for all services in all settings before implementation. To date, testing was limited to primarily medicine and surgery sections in the inpatient hospital setting. Sections to be tested included: obstetrics, measurement and monitoring, imaging, nuclear medicine, radiation oncology, osteopathic, rehabilitation, audiology, therapies and mental health. Further formal Beta testing should be undertaken with coders using real records. Additionally, testing should evaluate the compatibility of the new system with existing payment systems, such as DRGs, ambulatory payment classification (APCs), or simple fee schedules. AHA believed that use of ICD-10-PCS wouldn't lower providers' reimbursement levels.
Ms. Leon-Chisen emphasized that no decisions should be made about a potential single procedure classification system until all contending systems had been tested for compatibility with existing payment systems. Testing should be undertaken by an objective organization, preferably under NCVHS' direction.
Ms. Leon-Chisen asked the Subcommittee to consider nine implementation issues: (1) support migration to ICD-10-PCS after testing and establishing funding; (2) Medicare and other payers should be sensitive to increased regulatory cost linked to migration and adjust payment accordingly; (3) a well-defined, broad-based, predictable implementation and maintenance process that considered all users' needs, their capabilities to adapt the coding changes, establishment of routinely scheduled meetings to review coding changes, a date certain for using approved coding changes (AHA supported the current ICD-9-CM coordination and maintenance process and would support the same process for ICD-10-PCS); (4) clear, unambiguous instructions and consistent official coding and reporting guidelines readily available and accepted by all payers, preferably as part of the HIPAA standard code set; (5) continue the role of the cooperating parties in the development of guidelines and clarification on the application of ICD-10-CM and ICD-10-PCS; (6) continue AHA's long-standing MOU with HHS to provide ICD coding advice and training; (7) a detailed implementation timeline along with milestones; (8) lessons learned from other international implementations of ICD-10-CM should be applied; (9) backwards and forwards electronic crosswalks between ICD-10-PCS and ICD-9-CM vol. 3 codes available free of charge or at reasonable cost.
Panel 1: Possible Replacement of ICD-9-CM vol. 3 (Procedures) with ICD-10-PCs
Mr. Hull said that AdvaMed supported rapid adoption and implementation of ICD-10-PCS. AdvaMed was concerned about difficulties experienced because additional codes had to be assigned according to the rational construction of ICD-9 and often there wasn't room for a new procedure code in the appropriate section for new services in technologies. Mr. Hull emphasized that an optimal inpatient billing system must be capable of describing multiple aspects of a medical procedure including the body system, organ, surgical approach, specific procedure, and specific technology used during the procedure. AdvaMed believed the ICD-10-PCS system offered advantages in allowing for accurate billing, proper calibration of DRGs over time, and adequate payment for needed hospital services.
Mr. Hull urged the Committee to encourage CMS to implement important changes to the process for assigning new codes. He expressed AdvaMed's concern about the availability and timing of code assignment for inpatient services involving new technologies. The current ICD-9 system rarely provided for codes until after regulatory approval. Because of the annual coding cycle, new services might wait up to 18 months before a new code was even assigned. Medicare's inpatient prospective payment system might lead to additional delays of two years or more before an item was assigned to an appropriate payment group. Mr. Hull said this approach was neither timely nor resulted in a sufficient number of new codes to describe procedures relating to technologies. AdvaMed also urged CMS to consider the assignment of new codes for procedures related to technologies still undergoing clinical trials. Mr. Hull said earlier assignment of codes would allow for tracking and analysis of clinical and economic benefits and better outcomes research for new types of medical therapy.
Mr. Hull said AdvaMed was grateful for the crosswalk table for ICD-9 and proposed ICD-10-PCS codes recently posted on CMS's Web site. He encouraged them to create a version reflecting all current ICD-9 codes so everyone could better understand how the new ICD-10-PCS system would affect billing procedures. Mr. Hull also urged CMS to publish in similar format and provide an opportunity for comments on any revisions to the DRG grouper that might be driven by the adoption of the ICD-10-PCS system. Noting it was suggested earlier that ICD-10-PCS be considered for roll out in conjunction with the introduction of the ICD-10 diagnosis system, Mr. Hull asked the panel to consider whether obliging hospital billing staff to learn two systems simultaneously might pose training hurdles.
Panel 1: Possible Replacement of ICD-9-CM vol. 3 (Procedures) with ICD-10-PCs
Ms. Fox said BCBSA asked the Committee to postpone recommending conversion to ICD-10-PCS until two things occurred: the healthcare industry successfully implemented the initial HIPAA administrative standards specified in the law, and a thorough analysis of the implications of such a change was completed. She noted the healthcare industry had a massive job ahead and providers and health plans already faced significant costs and the complexity of complying with multiple federal rules simultaneously. In addition to changes otherwise made at the state level and by private accreditors, these rules required significant and costly reengineering of systems, business process changes, constant retraining of staff, and diversion of patient care from providers and reduced product innovation from health plans. Ms. Fox cautioned against overwhelming the system at this critical junction with yet another massive change.
BCBSA urged the Committee to assemble a multidisciplinary team to identify all the issues involved in converting to a new system. Ms. Fox said BCBSA plans learned a lot about the impact of system changes on their business as they worked toward compliance with the initial HIPAA standards. They'd approached the transaction and code sets as an information technology [IT] issue, but soon realized it was bigger than initially imagined and affected virtually every aspect of their business. In order to identify all changes required and develop effective solutions, BCBSA health plans assembled interdisciplinary teams (claims adjudication, medical management, actuarial, legal, and provider contracting) to determine the impact of these new rules on all operating areas. Ms. Fox said this process made it clear that converting to ICD-10 was more than a coding issue and a monumental undertaking. WEDI had drafted a white paper that inventoried the extensive issues involved. She noted they'd heard some issues that morning. These astronomical changes ranged from systems overalls to staff training, recontracting, and other business process changes. Because clinical codes were the underpinning of virtually everything in the healthcare system, these changes not only impacted payment but also medical policies, quality improvement programs, benefit design, and fraud and abuse prevention. Ms. Fox said that by assembling a multidisciplinary team, the Committee could build on WEDI's work, assuring that all these issues were properly identified.
Discussion
Ms. Leon-Chisen said they didn't have a choice about moving to ICD-10-PCS. ICD-9-CM vol. 3 wasn't fixable. Even if CMS funding wasn't available and payments continued to be revenue neutral, AHA would still support the three-year transition. Mr. Hull said AdvaMed members believed something more expedited ought to be considered than three years after the financial administrative transactions compliance date was in place. He said he would send a letter to the Committee identifying areas that could be accelerated. Ms. Fox explained that BCBSA was asking NCVHS to postpone their recommendations until a multidisciplinary team could be assembled to look at all the implications and cost everything out. She said they didn't know enough about whether implementation should be done separately or together, how much it would cost to go from ICD-10 to PCS, issues around doing just PCS and not CM, implications to the Medicare system (or benefit payments to hospitals), how having more codes would change the DRGs, cost differentials, and if there were winners and losers. Ms. Leon-Chisen said the diagnosis codes had more room left than the procedure side, but they, too, were more than 20 years old and “patched up”; AHA recommended migration to ICD-10-CM because 10-CM diagnosis would fix many needs not addressed under ICD-9. Ms. Leon-Chisen clarified that AHA recommended waiting until three years after the Claims Attachment Rule was finalized (2008-2009) because hospitals were “at the mercy of” HIS vendors; if HIS people could be quicker, AHA would reconsider the timeline.
Ms. Fox said the several years AHA and BCBSA agreed on gave people time critically needed to inventory all the issues and together come up with answers. Ms. Leon-Chisen suggested it would be helpful to know, up front, what direction the Committee intended to take as hospitals considered purchasing information system upgrades and budgeted for the future. The more time people had to plan, the better off they'd be. Ms. Trudel asked about the risk involved in continuing to operate 9-CM when it appeared to be losing its ability to be responsive. Ms. Leon-Chisen acknowledged there would be a time when it actually broke, but she said she was more optimistic they'd have a few years because the two chapters Ms. Brooks and her staff found brought some leeway. She clarified that there wouldn't be a point where one couldn't report a code for a procedure. One could limp along. The reimbursement and the data wouldn't be accurate, but you would still be able to assign a code to every procedure. Mr. Hull expressed concern over how the system and its specificity of payment eroded. He noted that other countries had 50 percent more DRGs than the Medicare program, essentially providing more specific payment for different cases. Ms. Fox clarified that BCBSA believed they needed to make sure that the coding systems were appropriately precise and accurate. At the same time, they felt they had to clearly state the advantages of a new healthcare system and the costs and implications associated with changing. Healthcare inflation was up 14-15 percent this year, hospitals felt they needed more money to ensure appropriate levels of service, and the states were cutting back on children's health programs. As a society, they needed to invest in revamping the systems. Ms. Fox emphasized she wasn't saying immediate difficult problems shouldn't be addressed, but BCBSA wasn't sure that going to ICD-10 right away was the best solution. They wanted to examine other short-term solutions, make sure things were coded appropriately, that they had the complete picture, and knew all the costs and implications.
Ms. Greenberg said she was struck by Ms. Fox's statement that clinical codes were the underpinning of virtually everything in the healthcare system. She said she assumed others on the panel agreed and noted this supported Ms. Fox's proposal to convene in an expedited fashion a multidisciplinary group to address these issues. She observed that finding an objective solution would be a challenge. Nonetheless, they'd heard that these clinical codes that underpinned everything were 20-plus years old and had serious deficiencies. Yet AHA recommended going the entire decade with them. She asked if that would work for hospitals or if they needed another way to run an effective, efficient healthcare system. Ms. Leon-Chisen explained that one reason AHA advised waiting three years was that they had to rely on HIS vendors for changes. She suggested there might be a way to work with them, have them inventory their systems, and expedite the timeline. Recalling Ms. Leon-Chisen had said the main cost would be training coders, Ms. Greenberg asked if this issue with vendors might be equally large. Ms. Leon-Chisen said it might be larger. AHIMA's members, the coders, stood ready to learn the new system. It would take time and development of training materials, but it was doable. AHIMA didn't have as much influence over HIS, but was willing to be a leader in that area.
Asked how long they could wait before doing something, Ms. Fox replied she needed more information before she could know. Ms. Leon-Chisen didn't know either, but she reiterated that AHA was willing to reconsider the timeline if vendors could help. Ms. Fox said six months would be a reasonable time for convening a multidisciplinary group, working with a consultant to come up with cost estimates, and mentoring these issues. It took BCBSA three-to-five months to look at standard transaction in code sets and do an analysis on their own. She said the industry should fund the study and that BCBSA would contribute. They needed representation from the states and all the other stakeholders at the table. She emphasized that the industry needed to realize the importance and commit to it. She said she didn't know if BCBSA could take the lead, but said they definitely wanted to participate.
Ms. Humphreys commented that, in order for the vendor community to refocus and speed up activities that addressed serious deficiencies in the ICD-9-CM code sets within a decade, the Committee had to come up with an early recommendation. Dr. Zubeldia remarked that it was difficult enough to get things done with a hard mandate, such as HIPAA. Noting hospitals held the purse strings that drove vendors, he asked if it was reasonable to think that a recommendation from NCVHS to move to ICD-10-PCS by a certain date would start vendors migrating before the deadline, or whether it took a hard mandate. Ms. Humphreys pointed out that, in order to be paid, hospitals had to run the old system until the day declared for the new system. Dr. Zubeldia replied that they'd heard today there was a relatively easy transition and CMS had already mapped between ICD-10 to ICD-9. Vendors might be able to implement ahead of schedule and still produce ICD-9 in out-going administrative transactions by a simple mapping. He asked if that was feasible, or if it should be a “one-time jump.” Noting that 3M HIS did the bulk of hospital medical records and was probably the HIS vendor most affected, Mr. Averill observed that the conversion to OPPS, which had a bigger impact on the vendor community in terms of creating new processes and wholly new software, had been done in less than a year. Here they were only modifying systems. If the Committee clearly set a direction, 3M HIS would work on it well in advance and be prepared for whatever date was chosen. He emphasized that they should choose a date. Mr. Augustine contended that this would be a major and not easy change. He said the AHA report summed it up when it detailed changing software interfaces and file and report formats. These codes were ubiquitous in everything a hospital did. Every screen, and everything one did with it, had to change. And many systems were proprietary. Finding programmers for archaic languages hospitals still used would be difficult and costly.
Dr. Cohn had a question for Ms. Brooks, who he credited with developing and maintaining vol. 3. Mr. Hull had commented that vol. 3 was already broken, but they'd also heard about opening up a new chapter. Dr. Cohn asked about the validity of concerns that there weren't enough DRGs to assure appropriate payment and about how long they had until things broke to the point they couldn't fix them. Ms. Brooks replied that, if they'd had a date that day, they still wouldn't have enough code space to make it. All they'd opened up with two chapters were 200 codes. Some proposals they'd discuss in May could use up 20 codes in a single proposal. She said they'd gotten stingy; a good proposal that needed five codes might get one. If they had a date, she said they'd limp along and stick things in very odd places.
Recognizing they dealt with a world lacking perfection, Dr. Cohn asked how many other cubbyholes they had to get them through. Ms. Brooks said they could pigeonhole a lot of body parts under eye, but nobody had better do anything in the cardiovascular area in the foreseeable future. She remarked that sticking codes in another body system might not be much of a problem for coders, who could track through the index. But for people doing data analysis it was more complex (e.g., a researcher looking at outcomes for all the spinal procedures who wasn't aware they were scattered throughout the book). Mr. Blair said they might not have the luxury of continuing on for nine or even six years. Noting Ms. Fox raised valid concerns, he asked if she'd consider slightly modifying her recommendation, putting it in the context of what needed to be done to enable a successful transition to ICD-10-PCS. Ms. Fox reiterated that they needed more information before making a recommendation. But if there was no alternative, she agreed they should look at how to move to ICD-10-PCS, realizing it was critical that they understood all the implications and costs when they made their recommendation. Dr. Cohn said he'd heard the Subcommittee's first step was to look at all the options and make sure they really understand and were convinced they'd picked right before mobilizing. Ms. Fox said that was exactly BCBSA's feeling.
Ms. Gilfoy supported looking into the process Ms. Fox proposed. Noting points raised around the issue of HIS vendor involvement, she reported looking at a recent gap analysis of HIS vendor-supplied packages. Ten out of 16 vendors could support a seven-digit procedure code or seven-character position code. She said Mr. Augustine's comment about IT implications was well said, but if vendors already supported seven positions, they had to assume they were supported throughout their database in all screens and reports. She said she was encouraged that at least 10 out of 16 vendors were able to move forward and support the PCS structure, but she emphasized that didn't take away the issue everybody mentioned about business implications.
Panel 1 (contd.): Possible Replacement of ICD-9-CM vol. 3 (Procedures) with ICD-10-PCs
Tracy Gordy, M.D., Chair, American Medical Association CPT Editorial Panel
AMA had followed the development of ICD-10-PCS closely and Dr. Gordy said AMA appreciated participating in the Technical Advisory Panel established by CMMS. AMA believed the project had made valuable contributions to many issues related to coding and terminology, but AMA and its medical specialties had identified problems with the code set that posed limits on its acceptability for practicing physicians, including the introduction of up to 197,000 new codes. AMA believed the complexity of ICD-10-PCS coding structure added to the extraordinary burden CMS/HHS rules and regulations already imposed on a regulated community of hospitals, physicians and other healthcare providers. She said it was ironic that a more complex coding system was being contemplated as the Secretary's Advisory Committee on Regulatory Reform urgently sought to remove much of the regulatory burden in federal programs.
Dr. Gordy said implementing ICD-10-PCS's structure would require significant resources in terms of education of users, computer systems changes, testing of reporting functions and ongoing maintenance and update process. AMA was concerned that inherent structural problems made that commitment imprudent and questionable.
ICD-10-PCS was a substantial departure from ICD-9 and other healthcare codes sets. Instead of pre-composed descriptors, codes and descriptors were built from tables that assigned values and terms by the placement of characters in the code string. Dr. Gordy cautioned that constructing codes and descriptors could cause significant problems for some specialties--e.g., in trauma care, the requirement to code every subset of a multi-system problem with a separate seven-character code string would be complex, time consuming, and exacerbating for procedures involving six-or-more types of body systems such as a Whipple procedure that was a proximal subtotal pancreatectomy associated with a total duodenectomy, partial gastrectomy, a choledochoenterostomy and a gastrojejunostomy. All six had to be coded with a separate seven-code string; a problem compounded for the coder because all the eponyms had been removed from ICD-10-PCS.
AMA believed this complexity and excessive formalism would cause problems for users and require significant education of physicians, coders and others billing or paying for services. In creating the descriptors, ICD-10-PCS relied on character positions with embedded meaning. In some instances, descriptors weren't based on the natural language of healthcare professionals but on a reconfiguration of the standard anatomic organization. The 31 body-system characters didn't conform to traditionally named body systems (e.g., physicians didn't differentiate between upper and lower body arteries and veins—how would they code an umbilical artery transfusion?). Dr. Gordy noted they'd heard that morning that it took significant education to use ICD-10-PCS. Informed individuals couldn't begin coding ICD-10-PCS using an index, intuitive reasoning or clinical knowledge. Physicians, other practitioners and coders would need to learn a vocabulary incompatible with the language they'd always used to document services and procedures. A new cadre of ICD-10-PCS coders would have to be developed at great expense.
AMA also believed the seven-character alphanumeric structure limited expansion by assigning meaning for each character slot. Thirty-four characters were available for each chapter/section; thirty root operations already had been defined in character position three, leaving little room for new root operations. Dr. Gordy noted that for this reason other code sets were moving away from this type of intelligent numbering.
Dr. Gordy noted implementation issues. ICD-9-CM, vol. 3 was a four-character numeric code set. All user computer systems would need to be modified to accept the increased number and type of codes. Existing software would have to be changed and/or new software acquired. And new hardware might be required to accommodate software changes. Dr. Gordy noted that systems changes necessitated by ICD-10-PCS were in addition to extensive changes already taking place, making the overall bill even larger and the project more complex. He said any changes to accommodate replacement should wait until HIPAA changes were accomplished.
Dr. Gordy noted that in order for ICD-10-PCS to be implemented successfully, DRG assignment must not only be accomplished but the assignment shouldn't differ from that achieved under ICD-9-CM. He said they'd heard that morning that consistent, reproducible payments based on a sample of past years had occurred, but they didn't know it had been tested. Flawed mapping could result in delayed payments and inconsistent DRG assignments could result in incorrect ones, causing problems for the administration of the Medicare Prospective Pay System. Dr. Gordy said problems associated with longitudinal data quality and comparability were also an issue; as coders gained experienced, coding interpretations and mappings might change, affecting coding comparability and retrospective analysis.
AMA advised that a code set with the complexity and built-in rigidity of ICD-10-PCS would require establishment of an editorial board with sufficient resources and independence from CMS and HHS to adequately maintain and update it. Dr. Gordy cautioned that ICD-10-PCS's structure and breadth involved a greater degree of complexity and clinical decision making and use of character position with embedded meaning suggested the need for substantial clinical input to ensure character placement was medically and anatomically accurate. Because ICD-10-PCS was limited to thirty-four characters in each of seven positions, considerable thought had to be given to how the code set grew. Factors for future growth included changes in therapeutic and diagnostic medical devices, clinical specialties and subspecialties that experienced a proliferation of new services, and clinical services on the horizon such as gene therapy. The detailed specificity and large number of codes required that ICD-10-PCS be maintained as a relational database, allowing the editorial board to build and maintain relationships between the codes, preserve consistent terminology, and eliminate duplicate codes or multiple ways of coding the same service. Development and support for such a tool was expensive, but necessary for timely, accurate maintenance of ICD-10-PCS.
Dr. Gordy reiterated that the expenses of educational needs and system changes plus the possibility of reporting errors could result in serious system-wide disruptions and financial disorder. The AMA believed that considerable further study was necessary regarding the cost/benefit of implementing ICD-10-PCS and whether administrative simplification would be achieved by its use.
AMA believed that, ideally, one procedure code set should be used uniformly by health professionals in all sites of service, allowing for true administrative efficiencies, reducing burdens on those currently mandated to use multiple code sets, and facilitating creation of databases to effectively analyze differences in treatment patterns and outcomes. Dr. Gordy said ICD-10-PCS would only perpetuate a dual reporting coding compilation that medical professionals found awkward and wasteful. CPT was used in the outpatient services for physicians and other healthcare providers that had the ability to use it under CMS and for the 60 percent of the population not governed by CMS. CPT also handled outpatient surgical procedures. All that ICD-10-PCS would replace was the inpatient services, which was dwindling in importance as more services normally done in the hospital setting were developed in an outpatient setting. AMA believed the CPT code set enjoyed universal acceptance and should be used in all sites for reporting procedures.
Panel 1 (contd.): Possible Replacement of ICD-9-CM vol. 3 (Procedures) with ICD-10-PCs
Ms. Prophet said AHIMA's members (who manage health records and databases in provider, health plan, government, and private organizations) believed the ICD-9-PCS for inpatient services was on the brink of collapse and imminent replacement was necessary. Ms. Prophet said the issue was how many different procedures could they classify to the same code before acknowledging the system was broken. Unless the situation was addressed quickly, in concert with other crucial coding system decisions, she cautioned the industry would face serious consequences.
AHIMA believed ICD-10-PCS represented a significant improvement over ICD-9-CM and met the characteristics outlined by NCVHS for a PCS. ICD-10-PCS provided more complete, accurate descriptions of procedures performed and its detail and completeness were essential. Its level of specificity would provide payers, policy makers and providers with more detailed information for establishing appropriate reimbursement rates, evaluating and improving the quality of patient care, improving efficiencies in healthcare delivery, reducing healthcare costs, and effectively monitoring resource and service utilization. Based on AHIMA's involvement with the ICD-9-CM, the Coordination and Maintenance Committee's struggles to maintain the existing system, and its own testing, AHIMA believed ICD-10-PCS was an acceptable replacement within the current use designated under HIPAA.
Ms. Prophet noted that because payment systems were now based on coded data, there were many stakeholders and the size of the work force directly engaged in coding had grown along with the range of coding skills. Thousands of mastery-level coders held CCS and CCS-P credentials. Ms. Prophet said AHIMA remained committed to training the health information industry, as it had for ICD-9-CM, through Internet-based communities of practice and networks of state associations and coding professionals. She said interactive Internet-based training was an excellent methodology for relatively quickly and cheaply training coding professionals and users of coded data. Audiotapes could also be economically delivered to large audiences. And face-to-face seminars, widely used to train for ICD-9-CM, remained an effective vehicle. AHIMA anticipated that experienced coding professionals would require about fifteen content hours of instruction in ICD-10-PCS, assuming the system was only used for the services that the ICD-9-CM procedure codes used (for the hospital inpatient setting, primarily the medical-surgical section). Instruction would include structure, principles, rules and guidelines of ICD-10-PCS and hands-on practical application using clinical data. Ms. Prophet said four-to-eight-contact-hours of training would be needed for users who didn't need hands-on coding skills.
Ms. Prophet said AHIMA was also prepared to address appropriate coding documentation management and data analysis issues related to the system change. Noting AHIMA's February testimony regarding medical codes and standards in HIPAA indicated the need for such codes and standards to include rules and guidelines for proper use to ensure consistent application of reliable data, Ms. Prophet emphasized that these rules and guidelines need to be developed well in advance of ICD-10-PCS implementation. She pointed out that many procedures included in ICD-10-PCS weren't typically reported with ICD-9-CM procedure codes (e.g., laboratory codes); the standard had to clearly identify sections of 10-PCS that needed to be included as a required data element of electronic transactions.
Ms. Prophet said a modified maintenance process was needed to ensure representation among stakeholders and public process in streamlined system revision. She noted six principles: (1) a committee representing all stakeholders, (2) open, public meetings, (3) streamlined procedures, (4) minimum lag time for process, (5) established process for developing system rules and guidelines, and (6) publicly available documentation. And she said ICD-10-PCS met these requirements. Ms. Prophet expressed AHIMA's belief that responsibility for maintaining this coding system and development of associated rules and guidelines should be the domain of a single agency, so decisions balanced needs of users, payers, and providers and ensured data integrity wasn't compromised. Ms. Prophet reiterated her earlier testimony, recommending NCVHS as the logical choice for this central authority. She suggested an advisory group representing all stakeholders be established to provide input into the maintenance and guideline development processes. AHIMA also recommended continuing the cooperating party structure.
AHIMA called for a standard effective ICD-10-PCS implementation date for all affected payers and providers. While noting that, ideally, ICD-10-CM and ICD-10-PCS should be implemented simultaneously (limiting disruption of data quality, coding accuracy and productivity, and ensuring proper emergence of claims) Ms. Prophet acknowledged that the systems were distinctly different, it was possible to implement separately, and implementing ICD-10-PCS first would allow for easier, quicker implementation and training. She said the question of simultaneous or separate implementation must be addressed in the context of the healthcare environment and accommodating the entire industry's needs. Ms. Prophet expressed AHIMA's belief that the question of separate implementations in the area of information systems (IS) and databases should be explored and she suggested there might be a way for IS professionals and vendors to make a one-time change in their information systems and databases, provided they were confident that conversion to ICD-10-CM was a certainty, had sufficient details as to forthcoming system and database needs, and agreed it was most economical. Ms. Prophet noted such an approach would permit the flexibility HIPAA called for in coding conversion and implementation and was consistent with comments heard that morning about a date certain. Ms. Prophet emphasized that, after evaluating all these considerations, AHIMA believed it was possible to replace the ICD-9-PCS before full implementation of ICD-10-CM, so long as this decision was made and planned around the conversion of both systems.
Recalling their previous testimony supporting adoption of a single PCS, Ms. Prophet said AHIMA opposed holding up implementation of ICD-10-PCS to wait for leadership and funding for a study. AHIMA believed that the crisis surrounding limitations of the current inpatient PCS and its need for replacement and granularity couldn't wait. And implementation of ICD-10-PCS in the inpatient acute care setting would both relieve the problems inherent in ICD-9-CM and provide better experience for a uniform PCS study. Ms. Prophet said AHIMA recognized other PCSs, but ICD-10-PCS was designed to replace the ICD-9 and was the only viable option for meeting the critical, immediate need for a replacement system. While significant cost implications were associated with adoption of ICD-10-PCS as a standard code set, Ms. Prophet remarked on significant costs associated with maintaining the status quo. Everyone paid a high price for delaying so long. They were already a decade behind in implementing new ICD modifications, and catching up could be more costly than staying current. Further delay wouldn't reduce direct costs, and indirect costs couldn't be ignored any longer. Any more fixes of ICD-9-CM weren't possible without severely disrupting the system's structure. ICD-10-PCS had the capacity to grow along with medical science and could serve healthcare procedural data needs for many years. AHIMA believed the benefits of the improved data resulting from ICD-10-PCS was worth the cost and difficulty encountered during the transition period.
Ms. Prophet emphasized that AHIMA and its national network of coding professionals were uniquely capable of assisting in the research of the right system, implementation strategy, and design of a new maintenance process, as well as taking the lead in retraining and user education for new procedural and diagnosis coding systems.
Panel 1 (contd.): Possible Replacement of ICD-9-CM vol. 3 (Procedures) with ICD-10-PCs
Ms. Whalen said FAH's members (1600 privately owned or managed community hospitals and health systems ranging from small rural hospitals to large urban medical centers offering acute hospital care, outpatient services, skilled nursing care, rehabilitation, and psychiatric care) believed ICD-9-CM, vol. 3 wasn't adequate and that providers, payers and Medicare beneficiaries would be well served by conversion to ICD-10-PCS. FHA recommended ICD-10-PCS as an appropriate replacement and the system of choice based open the Federation's essential criteria and elements: (1) in the public domain (facilitating training and usage of codes in official communications); (2) familiar to hospital inpatient coders; (3) use of standardized terminology that allowed for consistency of terms throughout the system; accumulation of statistics for financial, quality and outcomes measurement; payment; and information retrieval; (4) updating of the code set at the beginning of the federal fiscal year, in line with usage in the payment methodology; (5) usable on the current UB-92 form locators 80 and 81; (6) compatible with the Uniform Hospital Discharge data set; (6) a unique code assignment for each substantially different procedure; (7) expandability; (8) supports reporting of all procedures performed; (9) minimum need to change/update the system; (10) structure and terms in the code set facilitate easy data retrieval for one main code set; (11) codeable manually from paper records; and (12) absence of diagnostic information in the procedure code.
Ms. Whalen presented FHA's considerations for implementation of ICD-10-PCS, noting health HIM professionals and coding staff required detailed education and training. Education had to be provided to administrative and billing personnel, researchers, and the physician community. Training also needed to be available to fiscal intermediaries and private payers. And time had to be allocated to test and finalize educational and training tool development and prepare trainers.
FHA stated that CMS must continue to assist with training development and disseminating information to medical publications. Web-based training, the Payment Error Prevention Payment Program, and a hot line capability for trainers and coders were essential. Ms. Whalen said CMS must assume primary responsibility for educating and ensuring consistent application and understanding by the fiscal intermediaries. She noted private payers also required training and education. And, although others had roles in this process, CMS must also assume responsibility for educating physicians.
Ms. Whalen said multi-hospital systems needed to provide in-house training for coding and administrative personnel and physicians. Smaller facilities might have to rely on professional organizations and private consultants. CMS must ensure that every facility had fair and equal accessibility to quality training at a reasonable price. FHA believed that professional organizations would be instrumental in the successful delivery of education. These organizations would need time to plan their educational activities.
FHA believed it was critical that all payers converted at the same time using the same reporting requirements. Ms. Whalen said a high priority should be placed on national standards for all payers with the same reporting requirements. All discussions regarding development and maintenance of the system needed to be in a public forum with an opportunity for users to provide comments and have questions and concerns addressed. FHA recommended that the maintenance process be conducted in the public domain and allow for comments and questions in a cost-free public forum (i.e., the Federal Register). Development and implementation of guidelines, rules and interpretations should be conducted using this same open process. All guidelines and rules should be available in the public domain. Providers shouldn't have to purchase software, a printed code set, or other publications explaining official use of the code set. Copies of software, code sets, guidelines and rules should be available at cost through the National Technical Information Service, the Federal Register and/or other public documents. All publications should also be in the public domain. Committees responsible for maintenance and guideline development should meet publicly. If possible, a listen-only capability for meetings (real time, post meeting) should be provided on the Internet. Another administrative area of HHS (e.g., NCVHS) could take responsibility for the PCS, if there was any concern about an arm of HHS that administered Medicare payment being in charge.
Ms. Whalen noted that time had to be allocated for the DRG conversion methodology to be developed and tested for fairness. Adequate hardware and software must be tested and certified prior to training and implementation. She said costs for necessary hardware enhancement should be reasonable, not contribute to the increased healthcare costs, and affordable for small hospitals. Ms. Whalen emphasized that implementation was urgent and that FDA recommended that it occur quickly. She said the provider community could respond to the challenge of a new procedural classification system, if adequate attention was provided to a concise, reasonable and timely implementation plan.
FHA encouraged the Committee to consider a future single classification system for procedural reporting, regardless of the treatment setting. Ms. Whalen said it was in the best interests of patients, providers and regulators that a nationally recognized standard for reporting procedures be adopted. She noted it would be extremely difficult for healthcare providers to be sure that they complied with multiple standards, rules and regulations. Trying to provide consistent education and training as well as development and implementation of policies and procedures with the current approach was operationally inefficient and complicated efforts to promote, ensure and validate compliance activities. Although no system was ready to fulfill this need today, Ms. Whalen said FHA urged that ICD-10-PCS be used as the system's foundation. She said implementing the new inpatient procedural classification soon would position everyone for the future of a single procedural classification system.
Discussion
Dr. Zubeldia asked Ms. Whalen if she was referring to inpatient coding when she said there had to be one PCS. Ms. Whalen replied she meant all areas: ICD-10 lent itself easily to the outpatient setting and, from the perspective of other systems, it could easily become the one PCS. Ms. Prophet clarified that she'd excluded ICD-10 when she'd said she didn't see any other replacement for ICD-9-CM. All others would require significant, radical changes in processes, perhaps even structure.
Dr. Fitzmaurice reflected on the question of CPT-4 vs. ICD-10-PCS. He noted parameters of training, existing use, and possibly separate domains. They'd been reminded that procedures were being moved out of the hospital into ambulatory surgery centers. He asked if they knew enough about whether either could replace the other or ICD-9-CM in the hospital or ambulatory setting. He hadn't heard about overlaps between the two systems or any suitability for domains they didn't employ. One argument was that CPT-4 could do the job for hospitals. Other people preferred ICD-10-PCS. Remarking that he didn't want to trigger a raft of studies, Dr. Zubeldia noted they'd heard much testimony calling for one system for coding procedures. He said he was confused about criteria. Ms. Prophet told them there wasn't any basis at this time for judging between them. As AHIMA recommended in February, a feasibility study was needed to see which would be best and if either could work across all settings. Dr. Gordy reiterated that CPT was used on a prospective basis in outpatient surgical centers and a study never fully developed on CPT in the hospital could be repeated to see if it could be crosswalked to the DRG system. He remarked that if CMS went to a prospective paying system it might be more like the outpatient setting in hospitals that already had such a system then like the DRG. CPT currently covered everything but inpatient.
Ms. Whalen said it was difficult for a coder to go between ICD-9 and CPT. The two didn't talk or look the same. One had to switch hats between inpatient and outpatient coding. Ideally, there would be one PCS. CPT was familiar to many coders, but wasn't in the public domain, which would make it difficult to provide training materials. Ms. Whalen emphasized the Federation's belief that the procedure system and code set needed to be in the public domain. Asked about non-procedures (e.g., evaluation of management services, supplies coded with HCPCS), Ms. Whalen said they wouldn't fit unless ICD-10-PCS accommodated them. She said they could develop something on the evaluation and management side. Ms. Prophet reiterated that AHIMA didn't see HCPCS as a duplicative coding system and preferred to see the issue of naming supplies and devices separate from the performance of a procedure system. Ms. Whalen pointed out that typically those weren't coded by HIM staff or coders, but through a charge master.
Noting Dr. Gordy had recommended considering CPT codes and that the only area they didn't cover at the moment was the inpatient medical/surgical, Mr. Blair asked how long it would take for AMA to develop and test CPT codes for that area. Dr. Gordy said she wasn't prepared to answer that but she said crosswalking a Whipple procedure to the hospital setting was basically a matter of understanding the morbidity of the issue; the physician coded the procedure the hospital did under Part B for CMS. She doubted that crosswalking to the DRGs could be done in the six months mentioned that morning, but she said it might be done within a year. Mr. Blair asked about the body of codes that weren't in CPT and needed to be developed and tested. Dr. Gordy said CPT had almost 10,000 codes and the capability of 99,999. Ms. Prophet said that while there were inpatient medical surgical procedures in CPT, the system was designed primarily for physician reporting and so there were many situations where distinct codes had been created on a different premise than granular reporting of the procedure. A feasibility study could determine the number of procedures that had to be split because different data would be needed from the facility perspective. Ms. Prophet said another concern, beyond the structure of the codes, was the six principles of code set maintenance AHIMA set forth in its written testimony as ideal for any single PCS.
Dr. Zubeldia asked if bundling and unbundling of procedures was feasible under ICD-10-PCS. It seemed that if Whipple was dealt with as independent procedures, they'd get bundled under one procedure code, which wasn't possible with PCS. He questioned whether that could work for reimbursement. Ms. Prophet said PCS was granular and richer. One thing AHIMA had proposed was a single procedural system granular enough that one wouldn't have to change the reimbursement side to match the coding system. If six codes were reported the payer could decide to bundle into one payment or pay based on three codes; either would be fine and wouldn't change the code reporting structure. The codes would be granular enough that they could bundle and unbundle them and do whatever they wanted to design the payment system around them. AHIMA believed 10-PCS was granular enough to do that. Ms. Prophet said a reason AHIMA was glad not to see eponyms in 10-PCS was because those terms (e.g., the Whipple procedure) weren't consistently used. She said that was why 10-PCS boiled down to “call it what you would … but these are the actual components you did.” Ms. Whalen emphasized this granularity was necessary to use 10-PCS in the research field or any other statistical data set.
Dr. Zubeldia recalled that they'd looked at the HCPCS J codes, found they weren't granular enough, and decided to go to NDC codes. He expressed concern about running into a similar situation. Ms. Prophet acknowledged that, to an extent, more detailed medical record documentation was required, which concerned many members until they applied the system and found, most the time, that the level of documentation supported that level of detail. Dr. Cohn said he was concerned about how this would work in the future. If a physician described a local procedure as a Whipple procedure and affixed the CPT code, could the UB92 Professional reverse engineer that report to produce those six codes? Ms. Prophet explained that the coder in the HIM department didn't just code the name of the operation based on what the physician said. Coders were instructed to read the entire operative report and pick up all appropriate ICD-9-CM, vol. 3 codes to describe the report's contents. Ms. Prophet agreed with Dr. Cohn that vol. 3 was a relatively non-granular procedural code set, but she pointed out there could be more than one ICD-9, vol. 3 procedure code in an operative episode. Ms. Humphreys said vol. 3 had been tested and presumably met the medical-surgical part. They had the comparative data and what it was originally coded from; it could be retested.
Ms. Whalen remarked that from the standpoint of inpatient coding they weren't using CPT, but ICD-9. They read the operative report and coded all components of the operative procedure. Whipple might be three or four codes in ICD-9. Mr. Averill said the technical advisory panel had discussed this issue and there was strong consensus not to allow evidence in the system. Guidance was put in the index. A reference (e.g., appendectomy) routed to the right place in the tabular. Terms like Whipple that lacked clear consensus about their definition and constituent parts weren't included.
Asked about the feasibility of using ICD-10-PCS for non-procedures like services, an ambulance run or supplies, Mr. Averill said clearly the system wasn't designed for that, but the structure could accommodate it (it was no different than any of the other sections for which codes were developed) if one decided to take it on as a separate task.
Dr. Cohn recalled that one of the early principles the technical advisory group discussed concerned meaningless identifiers that weren't accepted by ICD-10-PCS. Some observed that caused interesting structural issues in ICD-10-PCS (e.g., differentiation between upper and lower veins and arteries). He asked if, in retrospect, that was something they'd do differently. Mr. Averill said the advantages of clearly identifying in a hierarchical sense what the character did in setting standard values far outweighed any disadvantages.
Asked what would happen with ICD-10-PCS' architecture when they ran out of character pieces, Mr. Averill replied that unless they changed Gray's Anatomy the anatomical sites were quite stable. They'd struggled to find more group procedures (originally they'd had two characters) and wound up with 30 after going through medical-surgical and no one could think of possible additions. He expressed confidence in the structure for root procedures and body sites. The one candidate for expansion, the device column, could be made specific to the procedure or body site, adding almost a limitless amount of space.
Dr. Fitzmaurice said he bought into Dr. Cohn's concept of codes assigned to elements that could lead the coding system into a classification system, and so he was puzzled. With each procedure classified and numbered, he thought there was still an advantage to having meaningless, unique numbers attached to each procedure in case something had to be shoehorned in. He said he could see genomic procedures expanding and meaningless numbers would allow a computer to read through them and perhaps map more quickly into another coding system. Mr. Averill said the technical advisory panel discussed that at length and, while they decided to go with the seven-character meaningful representation, they'd also observed that CMS or any collective body who decided how to implement the system could take their seven-character alphanumeric and make that part of the English description and sequentially number all the codes for reporting purposes. Both ends could be served implementing that way. Dr. Fitzmaurice agreed. The seven characters were unique themselves; another unique number wasn't needed.
Dr. Cohn asked Ms. Prophet how consistent Ms. Fox's suggestions about a feasibility study were to what she'd proposed in her previous testimony. Ms. Prophet said she thought Ms. Fox was referring to a study of the implications of implementing 10-PCS for acute care inpatient services vs. maintaining 9-CM. AHIMA's feasibility study posed that 9-CM couldn't last much longer and they'd like a single system across all sites of service, and so looked at the different systems and how they would work as a single system in all these settings. Dr. Cohn said he thought Ms. Fox also addressed identifying what ought to happen about 9-CM being broken and that Ms. Prophet had added the piece about simultaneously looking at exclusions that might move them towards a single PCS. Ms. Prophet agreed. One could look at it as two problems or a variable of the same problem. One element was 9-CM was used in the inpatient hospital setting and had to be replaced. They could focus solely on that or say, because they had to do something anyway, that they'd look at the full context of the duplicative code set issue for a way to concurrently resolve that larger problem within a reasonable period of time.
Dr. Zubeldia asked how long it would take for the system to transition from 9- to 10-PCS if they started tomorrow and had no other interferences or HIPAA transactions on Security. Ms. Prophet noted they weren't even close to a proposed rule. From a training perspective, she thought they could easily do it within the time allotted. Even once they had a final rule, there would still be a two-year implementation timeline. It would be close to 2005 by the time it was ready to go, even if they got going tomorrow.
Noting the IT problem was small compared to the system business problem, Dr. Zubeldia asked how long the healthcare system would take to transition. Ms. Prophet said she couldn't address the full healthcare industry, but she thought two years would be reasonable. A year would be reasonable to develop products and conduct training. Dr. Gordy said two years was optimistic. He noted they'd heard that morning that 9- and 10-PCS had to run for a while concurrently. And CPT had experience with people still using codes that were no longer in the book. Some systems would be slow to develop. Ms. Whalen said it depended upon whether they were talking about the whole industry. She said her experience with physicians was that it wasn't easy to teach them something new, especially when it came to coding. She said, at the most, it would take three years. She noted AHIMA was good at training. With the media available today, a year would be sufficient for training.
Mr. Blair observed that clearly they had to change, because ICD-9-CM, vol. 3 wouldn't last much longer. Whichever they chose would involve the costs of change. They needed to address the question of whether ICD-10-PCS was a better answer than CPT, based on the functional value of each system, not whether it cost more or was disruptive. They had to understand the cost of change and carefully work with whatever choice they made. But in discerning between CPT and ICD-10-PCS they needed to get and consider information comparing functional value, openness, and growth opportunities.
Dr. Zubeldia asked how CPT-5 fit in the picture. Dr. Gordy explained CPT-5 was already being incorporated into CPT and they'd find it in the book. She described a seamless integration. Components of CPT-5 were being implemented in CPT-4. Currently, diagnoses were being removed from the codes. Dr. Zubeldia asked if that included the codes necessary for inpatient procedures. Dr. Gordy said it did, in the sense that they'd increased granularity development in the CPT process. Medicare gave a DRG. A benefit company gave a fixed rate in the hospital for days of non-procedural kinds of care; and pneumonia or depression in the psychiatric industry got a fixed rate for so many days. Terms for the care were already there when a procedure was done. What they didn't have was the morbidity portion, which could be adopted from the DRG standards. They didn't have a study to show they could map what they had in CPT to the DRG system, but Dr. Gordy said that increase in granularity in CPT should carry over into the hospital. Asked how the hospitals billed for non-procedural stays like depression, Ms. Prophet explained vol. 3 was only procedures. A hospitalization for an appendectomy was billed with an ICD-9-CM procedure code. Hospitalization for depression would be billed with ICD-9-CM diagnosis code, without a procedure code.
Mr. Blair asked how ICD-10-PCS and CPT-5 were prepared to accommodate the growing area of procedures for complementary and alternative medicine (CAM) and areas involving human genomics. Dr. Gordy explained CPT was based on peer-reviewed literature. When the codes for acupuncture were presented the two requests that provided peer-reviewed literature to establish efficacy were accepted. The Relative Value Update Committee attached a value to each code relative to the system involved. Dr. Gordy noted that a set of category 3 codes was for new technology that hadn't yet established efficacy and clinical areas not yet established. Category 3 codes weren't valued by the committee, but were carrier priced based on the submitter. Dr. Gordy said CPT had some genetic codes, but that process was difficult even for the panel to understand.
Dr. Delbert Goldfield, Medical Director, CMHAS, said a variety of ICD-10-PCS coding systems specifically looked at alternative and complementary medicine. Acupuncture codes were already present and CMHAS most recently worked collaboratively with a number of societies on a set of codes for substance abuse. He said CMHAS worked collaboratively to come up with codes that reflected what was in practice, believing the best coding system was not only used for reimbursement but had clinical meaningfulness and could be examined for purposes of research on outcomes. Noting the codes Dr. Goldfield mentioned were institutional codes, Dr. Zubeldia asked if CMHAS contemplated alternative medicine services for an outpatient setting. Noting there was partial hospitalization these days and the term "hospital" was rapidly losing its meaning, Dr. Goldfield said CMHAS tried to cover services provided by individuals in an inpatient setting (e.g., members of the American Physical Therapy Association).
Dr. Yasnoff rephrased Dr. Zubeldia's earlier question, asking about the absolute minimum transition time for any change. Ms. Prophet and Ms. Whalen said a year was the absolute minimum. Asked if all CPT codes could be represented in ICD-10-PCS, Mr. Averill said there hadn't been a formal study looking at potential mapping, the objective had been to create a coding system that represented all inpatient procedures. He said there might be some dimensions CPT felt were important from a physician's work perspective (e.g., how many centimeters a laceration might be) that weren't particularly relevant in terms of an inpatient procedure, and in all probability there would be overlap. He said “the length of the repair of the laceration” could easily be put in the qualifier column. Structurally it would be simple to add those distinctions, and it would probably be possible to represent 99.9 percent of current CPT codes in the structure of ICD-10-PCS. They'd chosen not to create a section of I-10-PCS to deal with the ENM codes, but a structure was there. Dr. Gordy said AMA hadn't tried to map back to ICD-9 or ICD-10, but she thought Mr. Averill was correct. ICD-10 addressed procedures and wasn't designed at this point to address cognitive care. Ms. Prophet said an issue AHIMA had between 9-CM and CPT that she presumed they'd also have with 10-PCS was that one might be able to represent the procedure or service per se, but how it was represented and the pieces of information considered important could be quite different because of the focus on physician work on one side and the inpatient-facility perspective on the other. It wasn't a one-to-one match and might not mean exactly the same thing.
Dr. Zubeldia asked how much time they wanted for transitioning to ICD-10-PCS, considering that ICD-9 was broken. Ms. Prophet said AHIMA would prefer not to go beyond 2005 with ICD-9; by then, only one code would be left. "Other." Ms. Whalen agreed: two years after HIPAA. Dr. Fitzmaurice commented that a payment rate schedule might require more information than found in a diagnosis or procedure code. Payment might be based on qualifications of the medical professional or whether the procedure was done in the office, emergency room, or ambulatory clinic. Coding systems might not be able to handle all the needs for a payment schedule or research, but Dr. Fitzmaurice said describing accurately what was wrong with the patient and what a medical professional had done was the right start.
Subcommittee Discussion of ICD-10-PCS Testimony
Ms. Trudel recapped. They'd begun that morning discussing a primary issue with ICD-9, vol. 3 for inpatient procedures and what they could do: vol. 3 appeared to be breaking. As the discussions went on through the day, they'd talked about how this code set and procedures related diagnosis codes to other codes for services and procedures, outpatient services and procedures, physicians' non-institutional services and procedures, devices and supplies. And they discussed the notion of one big code set that would handle all or parts of this. Ms. Trudel said she wanted to put this primary issue and downstream issues on the table, so the Subcommittee could decide which dimensions they'd use to look at this problem. The first point of discussion was whether to look at the issue of inpatient procedure codes and the need to do something about vol. 3, or if they immediately needed to look at this in a broader sense.
Dr. Cohn questioned whether they were putting devices and supplies on the table or if that was another added code. He suggested there was a continuum of inpatient, hospital outpatient, and ambulatory procedures and that the point was to ensure they had a fix in the future for hospital inpatient procedures. But he said he'd heard general agreement that whatever was done ought to take the wider view and make sure they didn't miss an opportunity to move towards a singular PCS that might handle all three sites of service. The main focus was on hospital inpatient, but as part of that they needed to look at the feasibility of moving towards fewer PCSs. Dr. Zubeldia said, theoretically, he liked the goal of fewer procedure systems. It simplified things when everybody coded the same thing the same way. But he said he still had a bitter taste of NDC and J codes and was concerned that the institutional view of services rendered wasn't the same as the physician's who rendered those services. He doubted that they could have one encoding system if the physician counted the number of stitches he put in an incision and the hospital couldn't care less. Dr. Zubeldia disagreed with the movement that morning towards replacing the diagnosis codes, just because they were also in ICD-9. He questioned change unless it was well defined and justified and was concerned about a coding system that would be applied for all healthcare services, whether rendered in an institution or by a physician. He agreed that the main focus was on hospital inpatient and replacing ICD-9, vol. 3 if it was broken. He wasn't sure anything else was needed.
Ms. Humphreys commented that ICD-10-PCS hadn't been developed to handle everything that took place in an ambulatory care setting and CPT hadn't been developed to handle inpatient. Expansion for a purpose never intended, full testing to see whether it worked, and all that remained to be done clearly added to the equation. Unless they decided they couldn't possibly get a replacement for CM vol. 3 until 2010, that didn't seem workable. Mr. Augustine disagreed. He suggested there was a short- and long-term solution. The ideal situation would be one code set for procedures. But now they did all that had to be done with ICD-10-PCS. That was the most important thing. Recalling Ms. Fox's testimony, Ms. Humphreys suggested that they might not fully understand the full system in practice or ICD-10-PCS's impact on the healthcare system. Dr. Zubeldia noted that they'd heard it would take the recorders and hospitals one-to-three years to transition. He said the load on the payers to transition was much heavier. The PCS had to be fit into groups of about 1,500-1,700 payers who had to identify how much they'd pay for each of almost 200,000 procedures in a hopefully neutral fashion. He wasn't sure the payers could do that in three years. Ms. Humphreys pointed out that not all 197,000 procedures were done; payers didn't have to deal with those that weren't conforming. Dr. Zubeldia said the problem was that, if the payer had never seen a procedure code, he or she had to determine whether any factor affected payment. Mr. Augustine considered that, from the patient care perspective, PCS added with each character the ability to track, trend and monitor patient quality care, adding possibilities unavailable in existing systems.
Dr. Yasnoff identified three options for replacing ICD-9: ICD-10-PCS, CPT, or postponing until they had more information. Mr. Blair said he'd heard they had the choice of examining whether CPT could be an option in the future: it wasn't even developed yet for the inpatient area. Dr. Yasnoff and other members concurred; both systems would need further evaluation. Dr. Yasnoff also noted they'd heard that they shouldn't begin the change period while still working on the existing set of changes, and that a reasonable timeline was two years. They were looking at 2003 to 2005. Dr. Yasnoff said he didn't have enough information to make a decision, but he thought a next step was to ask those involved about the implications of changing to ICD-10-PCS or CPT in October 2003. Dr. Zubeldia noted a separate issue was whether the change extended to other procedure codes.
Dr. Cohn said Dr. Yasnoff put it together nicely. He noted a feeling from the industry that they needed to think hard about what made the most sense. Dr. Cohn presumed they'd come back, after thinking, and help explore options and timelines. Ms. Humphreys suggested that, in asking people to respond, they needed to be clearer about what they were reacting to. Her understanding, based on Ms. Prophet and other's testimony, was that a piece of PCS was very comparable to ICD-9-CM, vol. 3, which was broken. The question was what were the costs and issues around implementing ICD-10-PCS.
Mr. Blair cautioned against putting themselves in a position where the perfect was the enemy of the good. Whatever solution they came to would be imperfect, and if they kept pressing and delaying, they'd have a growing problem and a system crumbling at their feet. He said there were only three key questions and a “boundary.” One question was the focal point of the study Ms. Fox brought forward: What were the show stoppers with ICD-10-PCS? Another question was what were their alternatives. He suggested consulting an outside entity about CPT 5 (Did it have potential as a better solution, how long would it take to be implementable, how did its maintenance costs compare?) The third major question, assuming viable alternatives, was what were the tiebreakers? Mr. Blair said he'd heard several: maintenance, ability to accommodate future growth, and flexibility to grow beyond the defined scope (e.g., in ambulatory or acute care). And he said the fourth point or boundary was setting a realistic time frame for the study and making the best decision they could at that point. He said he'd hoped to wrap this up in June, but reluctantly was suggesting a hearing in the fall. They needed a certain level of confidence as they went forward. He asked if the members felt comfortable with a six-month window.
Dr. Zubeldia said six months was ample if the scope was to replace the brokenness of ICD-9, vol. 3. But it wasn't enough time to extend to a common PCS for everything. Mr. Blair said his inclination was to keep the scope limited; that didn't preclude broader scope in the future. Dr. Yasnoff reminded them that Ms. Fox said BCBSA would put resources into a study. Mr. Augustine suggested that, since they couldn't move on this for a few years, they use the time to thoroughly look at it from all angles, rather than just fix the short-term solution. Ms. Humphreys pointed out that October 2003 wasn't far away in terms of making a recommendation and publishing an NPRM so it could be implemented two years later. Dr. Fitzmaurice suggested they might set a date in October to hear about costs, benefits and models to help formulate a decision, then synthesize, and make their recommendation by February.
Mr. Beebe said in 1989 AMA hired a Cooper's librarian to study the costs of going to a single PCS; things obviously changed since then and he'd been working with PricewaterhouseCoopers on the costs of using CPT or ICD-10-PCS for inpatient coding and the cost of a single procedural code set. AMA had already put time in this and, hearing what the Committee said, they might invest in a study. Ms. Greenberg pointed out that the Committee had resources available if a study could be designed to address some of these issues. They'd heard that stakeholders would put their own resources into whatever they needed to do, but there was an advantage in not relying on a stakeholder for the study. Ms. Leon-Chisen said AHA would be interested in a study, and she agreed that it should be conducted by an independent, objective organization. She said AHA would be willing to help develop a model, but a study conducted by the Committee would have more credibility.
Dr. Zubeldia asked if was possible, as Ms. Humphreys suggested, to replace ICD-9, vol. 3 with sections of ICD-10. Pat Brooks, CMS, responded that today hospitals mostly code and report because it affects DRG and most those codes were medical-surgical codes in ICD-10-PCS. Coders developed additional types of codes people code internally, and a lot of codes were developed for other purposes. People might never use them, but they were there. The medical-surgical chapter contained most of what was coming in now. Ms. Brooks said the obstetrics measurement, monitoring, imaging, nuclear medicine, radiation, oncology, osteopathic, rehabilitation, audiology therapies and mental health mentioned in AHA's recommendations as needing additional testing and evaluation didn't have any impact on the payment system. She didn't know of anyone who had a DRG based on any of the imaging codes.
Ms. Humphreys said that in computing replacement costs they didn't want to count every bell and whistle of another system that didn't have to be phased in or used in order to solve core problems. She cautioned that they needed to be clear in asking about transition costs because they could get very different answers. Ms. Leon-Chisen clarified that only the obstetrics section would need further study if ICD-9-CM, vol. 3 procedures for inpatient was all they replaced. The other chapters would be useful if they looked at a single procedure system, because many hospitals tracked internally. She reiterated that AHA wanted vol. 3 studied for how it played into the current payment systems. Mr. Averill had told them that 3M HIS would be positioned to do that mapping and study, but apparently it hadn't been done yet. Dr. Zubeldia said the Subcommittee would recommend adopting all of the chapters because essential codes were sprinkled throughout; they couldn't carve out only a chapter.
Ms. Brooks clarified that there were two issues and people would comment on them separately. CMS would do a notice to implement ICD-10-PCS, get comments, and then do a proposed notice on how the DRGs would be changed. The year it was implemented, they'd have another public notice showing the mapping. Dr. Zubeldia acknowledged the difference, but noted the need for the flexibility to choose codes from all sections in order for the DRG grouper notice to indicate how they would do the grouping. Ms. Brooks explained that many codes in ICD-9 didn't affect the DRG assignment, but some were flagged as affecting DRGs. She said the same thing would happen with ICD-10-PCS. She clarified that they weren't talking about implementing part of a system, but that an activity wouldn't be mandated just because a designated code set could cover it. Dr. Zubeldia noted that concept was in line with the induction of the code set only for inpatient procedures.
Dr. Cohn remarked that they seemed to be reflecting on Mr. Gustafson's recommendation for a number of implementation activities. He said clearly the crosswalk needed to be updated; that was probably a show-stopper. And the process needed to be open for people to see and comment on. He noted mapping ICD-PCS into the DRG system would be CMS's responsibility. Another DMA would hopefully come in six months with whatever DRG mappings it had, and they'd have a sense of how well it would work using CPT. Dr. Cohn said the question then was did they do a study themselves or an evaluation.
Observing that they seemed to be driving towards getting the information together and making a decision in six months, Mr. Blair suggested that at the May meeting AMA might present their proposed code system for inpatient, discuss how it compared in terms of function and value, and how long it would take to develop. Dr. Zubeldia suggested they divide this problem into what they had to have and what would be nice to have. Something had to replace ICD-9, vol. 3 before 2005. It would be nice to have a single PCS for all healthcare and additional studies. It was clear that CPT didn't have anything to replace ICD-9 with today, and that they could converge into a single coding system with either CPT or ICD-9. Convergence would be nice, but wasn't necessary. Ms. Humphreys noted some people felt expanding CPT as a replacement within the time frame was an option; she said Mr. Blair was suggesting AMA could tell them whether that was feasible.
Mr. Beebe said AMA believed it wouldn't take much effort to use CPT for inpatient coding to determine DRGs. AMA wanted to redo a study done in 1986 that showed CPT could be used for these determinations. He said AMA didn't think 200,000 codes were needed; the nine thousand in CPT would be more than sufficient. Ms. Humphreys said that reflected a view that the only reason ICD-9-CM was used in inpatient settings was to drive the DRG. Her view was that, at the minimum, whoever used ICD-9-CM had to be given something that supported the functions it currently fulfilled in the inpatient setting.
A participant remarked that whenever his organization did ICD-9 codes, they were careful because one couldn't have anything overlapping. But because physicians bundled, concepts of one CPT overlapped with another CPT code, making DRG structures difficult. Everyone would have to analyze all CPT codes with overlap and pull them apart. They couldn't get it done in six months. CPT was the work of a physician and the concepts were so different it was hard to think that the bundles could work well.
Dr. Patricia Arzuaga said BCBSA hadn't begun studying what it would take to migrate to ICD-10 or the possible impacts on their business systems. She said if the Subcommittee told them what they wanted to know, BCBSA would talk with PricewaterhouseCoopers and deliver something in that six-month time frame Ms. Fox mentioned. Noting it was important to have peer reviews, Dr. Cohn said the Subcommittee needed to come up with a set of questions.
Mr. Tom Musco with Health Insurance Association of America (HIAA) noted there were reimbursement systems that weren't solely based on DRGs, but had payments made on a per-diem basis or through contractual arrangements on discounted charges. Anything that replaced the codes currently in that ICD would have to address more than just that DRG bundled payment. He agreed that replacing ICD-9, vol. 3 wasn't a simple substitution. Because of the various systems involved, patient history, utilization review, and disease management systems all used these codes. It wasn't just one set or group that had to be replaced. As Ms. Fox mentioned, there was a ripple effect, not just with reimbursement, but also throughout these systems that payers must update periodically. He said HIAA might participate in that study.
Dr. Yasnoff recalled that the Subcommittee detailed desirable characteristics of some guiding principles for selecting standards in its 2000 report. He suggested that the study the Committee commissioned could examine what choice reflected those guiding principles best, given this set of circumstances and a broken coding system. Ms. Greenberg said she saw the value in narrowing the scope for practical reasons, but she noted the reports the Subcommittee did in the nineties on desirable criteria for a single procedure classification system were available so members could know the Committee's rationale at that time. Dr. Cohn expressed gratitude for the resources, but he said his view was that they'd delegate as much as possible, recognizing that, though a certain part of what they did had to be under their control, there was value in having others in the industry participate, develop their own views, and provide widespread input. Ms. Greenberg agreed, but she said she wanted the Subcommittee to realize there were funds and asked them to think about how they might use them to pull this together. When people talked about six months, they meant a fairly concerted effort and the staffing provided to the Subcommittee had broader job responsibilities.
Dr. Cohn said they seemed to jump from being focused on the near-term fix to ICD-9, vol. 3 to anything beyond that was valuable, which he thought Mr. Blair proposed in his structure and Ms. Greenberg seemed to address. Ms. Greenberg said that, while the Committee didn't want to lose two months, they also didn't want to look at this completely isolated from replacement to ICD-9-CM, vol. 1-2. She suggested considering the pros and cons of replacing the whole ICD-9 suite at the same time or separately. Reviewing her notes, Ms. Trudel said the first thing to be looked into was the relative viability of 10-PCS vs. CPT. The Subcommittee staff and Ms. Greenberg would identify possible resources and a design for taking an independent look at what was the best code set based on the guiding principles, as Dr. Yasnoff proposed. The next sequential decision for the Subcommittee could be to consider the implementation problems Ms. Fox brought up.
Dr. Zubeldia said time could be allocated at the May 29-30 meeting to decide whether to take the diagnosis and procedure codes together or separate. Ms. Greenberg advised that anything done under a task order could be under way and they would have something preliminary; the key was to identify a source that could do this in a scientific, objective way. Dr. Cohn noted the need for a project plan and an overall scope, recognizing some of it would be under their control, and some of it they'd ask others to do. Dr. Zubeldia said the scope decision needed to be made relatively quickly, but they might have to get the project started first. The scope had to go through a single code set and would continue with the ICD-9. Ms. Greenberg noted the question Ms. Trudel posed was ICD-10-PCS vs. the CPT for a replacement for ICD-9-CM, vol. 3.
Mr. Averill suggested the Subcommittee begin with the criteria for how to make the decision, debate them and see if they felt criteria in the Committee's 1993 report were still applicable and if there were new criteria. Starting off, independent of either system, defining the criteria or properties a system should have could help the whole industry understand any recommendation ultimately made. If everyone went through all this time, effort, and expense to gain a new system, they'd want to achieve certain benefits in return; the Subcommittee could articulate those properties the system should provide and make a decision based on them.
Ms. Greenberg supported Dr. Yasnoff's recommendation. The 1993 report was primarily related to procedure classification, while the 2000 criteria were for standards more broadly under HIPAA; focusing on both sets together and the overlap between these different views would be helpful. Dr. Cohn noted that there had been other work since 1993. Ms. Humphreys pointed out that they'd selected the HIPAA code set and so the criteria for selection and HIPAA standards trumped anything else. Mr. Blair observed that they'd threaded the CPRI and ANSI-HISP studies on criteria for selecting good clinical terminologies that flowed into an ASTM standard and incorporated them in the PMRI report and their guiding principles for selecting standards.
Future Agendas - Issues Carried Over from February 6-7, 2002 NCVHS Meeting
Dr. Cohn said he'd ask members the next day to indicate their availability for hearings during three two-week periods in the fall. Dr. Zubeldia noted that the Subcommittee had an action plan with Committee deliverables to bring to the May 29-30 meeting. Ms. Pickett reviewed a list of topics stemming from the February hearings suggested for the May hearings. The outline, comprised of comments made by two or more testifiers, represented a broad-based view for the future agenda. Heading the list was the potential replacement of ICD-9-CM vol. 1-2 (diagnosis) with ICD-10-CM. Other issues related to the process for updating the alphanumeric HCPCS codes and a presentation from CMS staff addressing issues raised in previous testimonies. Similar issues raised about AMA's process for updating CPT-4 were included.
Ms. Humphreys noted another process issue heard a lot in February was whether more frequent updates of HCPCS and CPT meant the payers and hospitals would get uncoordinated updates every month. The question had been raised with CMS and she said a similar set of questions would be sent to AMA and CPT. Noting concern about whether there were gaps in HIPAA code sets with nursing, she said a conversation with Carol Bickford at ANA suggested they might identify people using local codes to pay for home procedures who were among those Dr. Cohn noted in February hadn't come forward seeking resolution. She suggested the Subcommittee might follow that thread further, based on what was heard the next day.
Ms. Humphreys noted a number of issues discussed in February were carryovers of comments made as part of NPRM and the final rule and were ongoing issues needing attention. Many in the healthcare industry were concerned that, among the HIPAA code sets, only the coding and reporting guidelines for ICD-9-CM were adopted as part of the code set. Historically, that had created problems because payers that created their own rules used these code sets differently. Ms. Pickett said CMS would address the issue of guidelines for HCPCS.
Stakeholders were going in opposite directions regarding the frequency of the code set update. Many on the technology side, where advances occurred at a rapid rate, thought it wasn't sufficient to update annually. Ms. Humphreys said having additional information from AHA, AHIMA and others about how the frequency of updates might increase administrative burden was salient for discussion. She also noted comments about problems with how AMA's new Category 3 codes were implemented and that some payers couldn't accept them because the structure of the codes varied from what normally went into the system, causing disruptions in process. And, as they'd heard that day, an open and public process was needed for these code sets; there was concern that users might not have the ability to actually submit code requests, be involved in discussions about updates, or be able to provide comment. National Council for Prescription Drug Programs (NCPDP) had provided information regarding gaps created with the initial standards because some code sets adopted as part of the initial code set weren't applicable to billing for supplies. NCPDP had cited NDC, UPC and HRI as other codes that should be considered under the HIPAA code sets. And there was the issue about recognizing the NCPDP standard for billing professional pharmacy services in addition to the X12N 837.
Dr. Cohn asked how many issues raised in the letter to CMS still required answering. Ms. Humphreys said CMS had everything and was expecting to address all of them when they came. Ms. Pickett noted that some were quite specific to CMS and the alphanumeric HCPCS process, but others on the timing of updates more frequent than once a year were generic across all code sets. Stakeholders could provide additional information on many issues. Mr. Beebe reported that the CPT editorial panel, BCBSA, HIAA and CMS agreed upon an update system releasing Category 3 codes in January and July with six-month phase-ins for implementation.
Dr. Cohn reflected on when or how they might deal with the issue of open and public processes. He said he'd been reminded, listening to the initial discussions about this issue, that nobody seemed to know exactly what “open process” meant. He suggested that later in the year they should get a better idea of what open process was considered to be. Dr. Cohn said he expected some code set developers might be willing to move towards something open if the Subcommittee could define it. Ms. Humphreys said they'd heard testimony to that at the last meeting. Ms. Greenberg recalled that the coding and classification team had developed a piece on what they felt was an open process. Ms. Humphreys noted language and guidelines in the preamble. She suggested there was a lot of discussion around this issue, and that the Committee had to do what it thought was right, based on what eventuated. The Committee had felt that there wasn't a single process that met the definition of open, and hadn't wanted to give twelve steps indicating there was no other way. She said they needed to figure out whether there was too much leeway, because some stakeholders felt excluded.
Ms. Pickett said this was a recurring theme that came up during the comments to the NPRM and at every hearing on code sets. Stakeholders clearly considered it an issue, but the Subcommittee hadn't received additional input about what stakeholders thought might happen to code sets considered more closed or recommendations for changes. Dr. Cohn questioned whether openness had to do with everybody being heard or with access to decision making; there seemed to be two very different sorts of openness. Ms. Humphreys said a third piece was everyone being able to listen to what went on. Dr. Zubeldia added another piece associated with openness that came up at every meeting: the cost of the codes. Some people believed an open process led to free codes. Ms. Pickett pointed out that the guiding principle said, "reasonable cost."
Mr. Blair remarked that talking about an open process led him to think that the criterion of peer review made it extremely difficult, if not impossible, for new codes (e.g., CAM) to have a chance at becoming part of the system. Dr. Zubeldia said he thought the open process was for obtaining existing code sets or, in the future, obtaining code sets in general. Ms. Humphreys remarked that open could be defined any way. From the Committee's point of view, a legitimate need in the industry either had to go into an existing code set or another code set had to be approved for it. She noted the next day they'd hear more testimony about how the current systems dealt with perceived gaps. Dr. Cohn noted tomorrow they could reflect on that and a letter asking them to consider how to deal with an issue raised in previous testimony about which transaction and code set to use billing supplies for NCPDP. They'd probably look at the open issue related to the cost issue during the second half of the year. Dr. Cohn said, obviously, they'd defer to those capable of handling some of these additional issues. The Subcommittee had a fairly active set of code issues and other things they were also supposed to be doing. Compliance would become an issue, and they needed time later in the year to deal with PMI next steps and begin thinking about the administrative and clinical procedure code sets.
Clinical Drugs - Project Overview
Ms. Humphreys reported for Dr. Stuart Nelson, NLM, on the Clinical Drugs project, noting in a previous Subcommittee discussion about NDC it became clear there was a missing code set of control vocabulary--the product level (e.g., a 50- or 100-count bottle). The provider prescribing often didn't have this information, which grossly complicated the notion of automated decision support reminders on drug-drug interaction checking as hundreds of NDC codes might represent an aspirin of a comparable dose. Ms. Humphreys illustrated the scale of missed synonymy by noting that, while FDA had approved 10,000 drugs, NLM's UMLS system has 81,000 individual clinical drug listings. Health Level 7 (HL7) had already identified this problem and was working on a standard form representing what was known when the drug was prescribed. The Food and Drug Administration (FDA), the Veterans Administration (VA), and NLM coalesced around developing a standard representation, the RX normal form, which included each ingredient, strength and dose form. The plan called for creating this normalized form for all approved drugs and having separate concepts in the UMLS system for ingredients, drug components (ingredient and strength), dose forms, drug formulations, and labeled relationships between concepts. The goal was to map every form of a clinical drug from the VA's National Drug Formulary (from Multum, Micromedex, First Databank, and Medispan), connecting them to this RX normal form.
Ms. Humphreys explained that VA was interested in a full drug reference terminology that had additional properties. The project dealt with standard nomenclature for representation of what was known about the drug when it was prescribed, plus additional relationships and attributes. Everyone would be given a standard way of representing this and, if everyone mapped to this, they'd be able to exchange information, coalesce, and inter-operate in ways that they couldn't today.
RX normal forms for a big chunk of the drugs in the VA's national drug formulary were released in the first 2002 version of The Medical Source in January 2002. The next step was to look at the scaleableness of this approach, create the normal forms, and map them for other drug services. The dose forms and model were refined as they moved through more drugs. Ms. Humphreys said they expected to complete the bulk of this work for the second 2002 release of The Medical Source available in May. She said they would also make sure that the hundred-or-two-hundred most commonly prescribed drugs were fully substantiated for all of these relationships within that smaller subset. They hoped to complete the mapping for the January 2003 release and have more enhancements available in July. The normal form created by NLM in conjunction with the other organizations would be available free and feedback was encouraged. Hopefully, additional forms would be available when FDA approved drugs in 2004.
Mr. Blair asked whether First Databank and other drug information system vendors felt this had potential for them to adopt, rather than retain their proprietary systems. Ms. Humphreys said they were relatively enthused. They felt this lifted some work from them and enabled them to spend their resources on value added drug-drug interactions and other information. Mr. Blair asked if FDA might adopt RX normal form to the extent that there was consistency with MEDRA. Ms. Humphreys explained that MEDRA dealt with adverse reactions in testing and post market release of a drug. Noting FDA was interested in ratcheting up its participation in standards-related activities and coordinating across the agency, she suggested the Subcommittee might want to hear from Dr. Janet Showalter in the Administrator's Office about what FDA did in this area. She said she would extend Mr. Blair's thanks to Dr. Nelson. Dr. Zubeldia asked if NCPDP had been involved in this for their electronic prescription transaction. Ms. Humphreys said her understanding was that, as they'd heard, NDC code was applicable in this arena because they were selling a product. Dr. Zubeldia noted the prescription transaction didn't have an NDC code to use at the time of a prescription. Ms. Humphreys said she would follow up with Dr. Nelson to see how this might be workable.
Panel 2: Previously Identified Gaps in Code Sets
Dr. Mirin said that, in the case of mental disorders, the proposed adoption of ICD-9-CM and the codes and accompanying definitions would impede rather than facilitate efforts to reduce fraud and abuse, and inhibit rather than promote administrative simplification. ICD-9-CM diagnostic codes were already required for Medicare claim submissions by CMS and private insurance carriers, and so making it the only HIPAA-approved diagnostic code set wouldn't create difficulties for the healthcare community. But for over twenty years, mental health practitioners had used an alternative set of descriptors: APA's Diagnostic and Statistical Manual of Mental Disorders, now in its fourth edition (DSM-4). Dr. Mirin and APA requested that DSM-4 be designated as the code set for the diagnosis and reporting of mental disorders. DSM-4 was used nationally by virtually all providers of mental healthcare, including psychiatrists, psychologists, social workers, counselors, mental health administrators and policy planners for communication, clinical management and record keeping. Epidemiological surveys and studies of mental health practice patterns use DSM-4 definitions for ascertainment of caseness. Practice guidelines for improving and standardizing patient care were keyed to DSM definitions. Virtually all research studies on mental disorders defined study populations in terms of DSM categories. Students of medicine, law, psychiatry, psychology, social work and all other mental health professions relied on textbooks that described mental disorders based on DSM definitions.
DSM was the de facto official code set for various funding agencies and virtually all states. Over 650 federal and state statutes and regulations relied on or directly incorporated DSM diagnostic criteria. The Department of Veterans Affairs' disability program used the diagnostic criteria in DSM-4 to assess whether an applicant qualified for a disability on the basis of a mental disorder. CHAMPAS required that a mental disorder be a condition listed in DSM-3. Medicaid beneficiaries who applied for admission to nursing facilities due to a mental disorder had to meet DSM's diagnostic criteria. In California, Medicaid reimbursement to hospitals was keyed to DSM-4. Mental health qualifications to serve as a police officer in Tennessee and driver's license provisions in Pennsylvania incorporated by statute reference to the DSM. Both federal and state courts relied heavily on the DSM as the common diagnostic language in discussing the possible role of mental disorders in serious crime.
In recognition of the importance of keeping the diagnostic nomenclature in mental health linked with the nomenclature used by the larger medical community and to meet regulatory requirements of both the federal government and private insurance carriers, the DSM diagnostic codes corresponded to more-or-less equivalent categories in ICD-9-CM. Many diagnostic terms were identical, reflecting adoption of modifications requested by APA over the years to facilitate efforts by clinical coding specialists to use DSM terms to locate ICD-9-CM codes in the alphabetical index. DSM-4 was fully compatible with ICD-10.
The key difference between the systems was in the definitions of the disorders. Unlike DSM-4, which provided operationalized diagnostic criteria using straightforward terminology easily assessed for each mental disorder, Dr. Mirin said ICD-9-CM had a brief (never updated) glossary of definitions, using vague, subjective, archaic terminology written in the mid-1970s and unfamiliar to most practicing mental health professionals. For example, he noted both systems classified major depression under code 296.2. In ICD-9, the discussion of this highly prevalent illness consisted of three sentences with terms like “wretchedness,” “mood of gloom” and other archaic descriptors. DSM-4's description occupied a dozen pages, was specific, and the product of an evolutionary process overseen by the APA that involved numerous organizations over twenty-five years. DSM diagnostic criteria were updated every seven-to-ten years. Two major revisions were made since 1980. Updates were based on comprehensive, methodologically sound review of scientific data and extensive input from and collaboration with federal government agencies, including NIMH, NCHS, WHO, and clinicians, researchers and policy makers. A collaborative process had begun to update DSM by the end of the decade, incorporating advances in scientific research and classification of mental disorders.
Dr. Mirin said implementation of HIPAA regulations would thus require mental health practitioners, agencies and institutions to abandon their use of the DSM-4 diagnostic criteria in favor of the ICD-9-CM definitions reflecting medical science as it was in 1977--impeding efforts of public and private payors to promote administrative simplification and enforce fraud and abuse laws and efforts. For example, Medicare auditors seeking to determine whether a psychiatric hospitalization was medically necessary reviewed a patient's chart for signs and symptoms consistent with DSM-4 diagnostic criteria and its severity. Mandated use of ICD-9-CM would undermine efforts to fight fraud and abuse as well as all the federal and state laws incorporating DSM, needlessly complicating use of medical information and the practice of medicine.
Dr. Mirin noted DSM-4 was widely used by providers and federal and state governments and satisfied the ten specified principles. He requested it be adopted as a designated code set and that APA be included as a DSMO. Used for medical, reimbursement and forensic purposes, if adopted as the mental health diagnostic code set, DSM would improve uniformity, meet healthcare provider standards, lower implementation costs, eliminate development costs, and have APA's ongoing support in maintaining the code set. DSM was precise, unambiguous and as simple as possible, would minimize data collection and paperwork requirements, as it already was the mental health standard used by providers, private payors, federal and state agencies. It was adaptable, regularly updated, and technologically independent of computer platforms and transmission protocols.
Dr. Mirin said APA didn't expect significant problems with this addition. Only the definitions, diagnostic terms and coding conventions differed. The actual numerical codes were identical to ICD-9-CM. He noted precedent for proprietary code sets to be adopted under HIPAA (e.g., CPT-4 and Code on Dental Procedures and Nomenclature). APA had already established an open and inclusive process for periodic revision of the DSM. And no burden would be imposed, because use of DSM-4 was already the status quo.
Panel 2: Previously Identified Gaps in Code Sets
Lorrie Kline Kaplan, Executive Director, National Home Infusion Association
Dr. Kaplan explained that NHIA was a 501c(6) non-profit organization. NHIA's HIEC coordinated its coding and electronic claiming activities. Home infusion was the at-home-or-ambulatory management of specialty (often IV) medications and nutritional solutions for patients, by a specialized non-traditional licensed pharmacy. This market represented four-to-five billion dollars annually in healthcare expenditures for patients who otherwise would have their care managed in inpatient facilities. Private payors covered home infusion for nearly 20 years to reduce inpatient costs, yet there was no explicit Medicare coverage. Almost every payor had its own codes, although most shared common features and used a per-diem approach to cover home infusion. Similar to the method adopted for Medicare home health and outpatient services, a bundled daily rate was prospectively determined based on the therapy provided. Payors knew in advance the cost of care and providers knew the reimbursement.
Dr. Kaplan explained that NHIA's coding system followed per-diem methodology and was designed for use under the ASC X12 Professional Claims Standard. It was not approved under the HIPAA rule, but HHS recommended that additional standards be adopted to fill existing coding gaps needed to meet home infusion's business needs. These gaps weren't minor and at that time no codes available in the HCPCS or CPT systems for home infusion were even similar to the HIEC or other private payor codes. In a December 2000 policy change request, NHIA applied to the DSMOS for approval of HIEC as a HIPAA medical code set. NHIA felt HIEC fit the criteria it believed appropriate for evaluating this request: (1) Was this an established healthcare sector? (i.e., Was there a business need for these codes?), (2) Were there significant coding gaps in the approved systems? (i.e., a lack of duplication with current systems), (3) Was this a sizeable body of codes? ( i.e., more than 50—HIEC has over 200), (4) Did the system methodology reflect current coverage models in the marketplace?, (5) Did it meet provider and payor' needs, (6) Was specialized expertise required to properly design and maintain the system?, and (7) Did the code set maintain or have necessary systems in place to professionally maintain these codes to HIPAA requirements? Dr. Kaplan reported that NHIA's NIEC system met these criteria, however the DSMOS rejected NHIA's request. She said it was still unclear what criteria the DSMOs used. NHIA was directed to work with HCPCS and CPT maintainers to address the gaps. Lacking other reasonable options, NHIA pursued that directive.
Dr. Kaplan reported that many of the home infusion coding gaps had been addressed through the 2002 HCPCS. Much of the HIEC coding system (some 80 new per diem codes, themselves comprising a fairly comprehensive system) was freely adapted for inclusion in the 2002 HCPCS. While NHIA still believed approval of the HIEC codes would have been best, they commended the panel for the steps it took to address these issues. Dr. Kaplan said NHIA worked hard to facilitate adoption of the HCPCS, and found that most providers and payors were pleased with them.
NHIA urged NCVHS to consider whether additional changes to the HCPCS infrastructure might be needed to accommodate private payor coding and other new demands placed on this system. NHIA recommended renaming the S codes non-Medicare codes or another name reflecting permanent status. NHIA was told that these non-Medicare S codes were permanent, but had no formal assurances, and payors and providers were hesitant to convert their systems and contracts to codes termed temporary. Dr. Kaplan said HCPCS S codes that cover home infusion to reduce inpatient expenditures should be usable by Medicare Plus Choice plans. And she said the HCPCS application should be updated: HCPCS increasingly addressed services, but the application was designed for products and manufacturers--not providers or payor groups who identified coding gaps under HIPAA. Noting providers had no formal opportunity for review and comment on coding requests and issues, Dr. Kaplan said healthcare providers should have input on HCPCS. Dr. Kaplan requested that NHIA have a formal advisory role in the HCPCS panel's considerations related to home infusion coding. NHIA already maintained systematic published processes for reviewing and recommending coding requests, providing a clearinghouse and review function before sending requests to the panel, and strove to see these functions formally incorporated into the review process for home infusion coding and other healthcare sectors.
Dr. Kaplan said NHIA appreciated recent inclusion in key meetings with AMA and hoped to continue working closely, but noted obstacles. She said AMA developed home infusion coding without public input and published them over industry opposition. The CPT codes didn't fit the home infusion clinical or business model and were at odds with HCPCS. NHIA considered the codes the major obstacle to coding standardization in home infusion and continued to ask that they be withdrawn. She said they were an example of what happened when one sector of healthcare developed a coding framework for a sector in which it had no appropriate experience. Dr. Kaplan noted the lack of adequate coding for home infusion nursing visits was another obstacle. NHIA was pleased to participate in an AMA workgroup recommending CPT coding approaches, but was troubled that changes wouldn't appear until at least 2004. Another obstacle was codes that didn't fit the needs of the marketplace, triggering coding misuse. Payors already talked of changing the descriptors. Dr. Kaplan noted education was needed to gain standardization in coding use and was vitally important for the success of HIPAA. One comprehensive resource, National Coding Standard for Home Infusion Claims Under HIPAA, was available free at NHIA's Web site, www.nhianet.org/hiec_issues.htm.
Noting that Ms. Ward's written testimony showed many issues were already being addressed, Dr. Kaplan offered observations about NHIA's experience as an early requester in the DSMO process. NHIA had been distressed at how much “dogged persistence” it took to participate and provide input. They'd submitted considerable supportive documentation for a NUBC and NUCC meeting, but members hadn't received the materials and weren't prepared to discuss the request. Dr. Kaplan said they were surprised to find their request turned down after that. After they appealed the original rejection and were again denied, none of the other DSMOS offered to discuss their request. Attempts to get transcripts or minutes explaining the decisions were unsuccessful and Dr. Kaplan said she read the February Subcommittee testimony with great interest, as NHIA had heard none of the reasons for denial. She said many reasons cited appeared to be based on misunderstandings that could have been cleared up. In Dr. Kaplan's view, some were addressed prior to the denial.
Dr. Kaplan said all DSMOS and code set maintainers should be expected to meet standards required of ANSI accredited organizations: fairness, openness, equal representation and public accountability. NHIA felt fortunate that many needs had been addressed for now, but Dr. Kaplan said luck had been on their side. If they hadn't been a squeaky wheel, the outcome would have been far different--and still could be. The home infusion industry had a significant volume of coding needs, but Dr. Kaplan said the organization representing the sector, NHIA, had no formal role in this process. Significant decisions were being made that would control their lifeblood by entities lacking significant expertise in home infusion.
Panel 2: Previously Identified Gaps in Code Sets
Ms. Giannini addressed the critical gap in codes for alternative medicine, nursing and other non-physician services needed by more than 2.7 million licensed healthcare providers inadequately served by existing HIPAA code sets. ABCcodes had five characters followed by a two-character modifier. The same length as HCPCS's Level 1 and 2 codes, ABCcodes fit into existing software applications and health information systems as well as on UB-92 and HCFA CMS-1500 forms. ABCcodes supported non-physician practitioners licensed as healthcare providers in one or more states who had completed accredited training programs and met national and state certification requirements.
Ms. Giannini said the code set was designed specifically for HIPAA compliance and addressing this rapidly expanding gap identified by the NLM and the White House Commission on CAM Policy. She said ABCcodes delivered infrastructure needed to determine which non-physician interventions could help manage access, improve healthcare quality, and alleviate skyrocketing healthcare costs. In developing the codes, Ms. Giannini said Alternative Link explicitly followed guidelines created under HIPAA for code set development and approval, especially those pertaining to administrative simplification and avoidance of billing fraud and abuse. The code set met the ten guiding principles outlined in the 1998 proposed rules. Market-ready information, products and consulting services were available to support HIPAA-covered entities and business associates in implementing the code set.
Alternative Link submitted extensive testimony pertaining to the development and acceptance of ABCcodes and followed specific instructions from HHS, DSMOs, ASC, X12N, HL7 and EDI experts. They'd submitted letters of recognition and support from thought- and practice-leading healthcare organizations demonstrating that ABCcodes addressed public health and business needs not met by code sets adopted under HIPAA. The code set supported benefit plan design, utilization and claims management, and appropriate reimbursement. It facilitated development of evidence-based practice guidelines through clinical trials, health services research and actuarial analysis. ABCcodes offered functionality, fulfilling a unique and essential need of non-physician practitioners that couldn't be achieved by other code sets. The code set design enabled users to link services and supplies to state scope-of-practice laws for multiple practitioner types.
Ms. Giannini said the DSMOS hadn't recommended ABCcodes as a standard code set, but redirected Alternative Link to approach AMA and HCPCS to determine whether ABCcodes could be managed as part of HCPCS's Level 1 and 2 codes. She said Alternative Link complied, even though AMA had been consistently uncooperative and heavy-handed. Recently, Alternative Link was encouraged by the tone of AMA's response to a request to collaborate. Still, Ms. Giannini expressed concern that collaboration could continue to be difficult and time consuming. Within the last month, she said AMA encouraged Alternative Link to apply for CPT codes through the existing code application process without informing them that this would transfer intellectual property to AMA without compensation. When Alternative Link asked for more specific guidelines, AMA didn't respond until the day the CPT application was due. Had Alternative Link not anticipated these responses, Ms. Giannini said they could have lost their ability to represent non-physician practitioners and efforts could have been stalled for three more months.
Alternative Link's adapted CPT application points out the unique and essential functionality of ABCcodes and the extent of code set development (about 4,000 service and supply descriptors and corresponding legal scope of practice information and relative value units). It asked the AMA to either acknowledge the differences in code set functionality and recommend ABCcodes be named a separate standard, or propose a way to accommodate the codes and serve the needs of non-physician practitioners.
Ms. Giannini said Kaye Riley, Director of the HCPCS panel, supported national code set designation for ABCcodes and if delays were expected, then Kaye was going to ask CMS if the ABCcodes could be added to HCPCS Level 2, and that adoption by payors was forthcoming, once ABCcodes was named as a code set standard. She noted different laws in each state governed alternative medicine, nursing and other non-physician providers and that the AMA CPT editorial panel didn't have expertise in non-physician interventions. Only a handful of non-specific codes had been developed, and claims for non-physician interventions often had to be manually processed or supplemented by extra paperwork to document treatment rationale. In relation to non-physician interventions, CPT codes didn't meet congressional intent vis-à-vis HIPAA. Ms. Giannini said Alternative Link was the only organization that had built this type of information into coding functionality on a code-by-code, state-by-state, and provider-by-provider basis and that the healthcare industry needed ABCcodes.
Ms. Giannini said Alternative Link had analyzed overlaps with CPT and found less than ten percent crosswalk. Anyone could submit applications for code set maintenance. Code development approval and obsolescence were overseen by independent coding administrators, based on written procedures and formal standards with systematic reviews by training, licensing and other authorities. Procedures and standards were described in coding publications and on Alternative Link's Web site. The code set was updated continually; new releases were issued annually. Coding manuals included general guidelines for coding and expanded definitions for each procedure. Expanded definitions were available in electronic and hard copy formats. Current users adhered to guidelines and hadn't needed to develop their own definitions. Alternative Link codes were available as an ASCII flat file ($49.95) and as a PDF coding manual ($59.95). State code reports for each provider in each state were also available ($19.95). Updates were disseminated by Alternative Link and its distribution partners (Thompson Learning and Relative Value Studies, Incorporated). The licensing fee structure was comparable to fee structures established for existing code sets.
Discussion
Ms. Giannini clarified that one HMO already used ABCcodes. A PTO network that represented 23 million covered lives was ready to adopt the code set once it became a HIPAA standard. Some large payors (e.g., BCBSA) were interested, but first wanted to see it a HIPAA standard. Ms. Giannini explained that alternative services were billed as a carve-out with the insurance company paying caps on dollars or numbers of visits; little data was captured or used. Often, insurance companies pulled in claims and counted dollars or numbers of visits. Acupuncture and electrical acupuncture both had a code. Chiropractic services had four codes; practitioners also used some physical modality codes in CPT. BCBSA had S codes specific to some services. There was no data to prove efficacy based on current coding. Dr. Kaplan noted the confusion over these permanent/temporary codes. Payors resisted adopting them because they were temporary. She said the nomenclature could easily be resolved.
Dr. Fitzmaurice asked whether potential users didn't find the coding system useful or didn't want to pay for services. Ms. Giannini said users would like an actuarial analysis first, because they didn't know their exposure. But without codes, they couldn't do the analysis. Dr. Fitzmaurice suggested they could hire a researcher to categorize what was being done, then look at evidence and decide what to pay for. He noted those steps hadn't been taken and said he wondered about coding something people didn't want. Ms. Synthia Molina, CEO, Alternative Link said they'd encountered many self-insured payors and third-party administrators who'd experimented with benefits in this area. The challenge they faced was the lack of a mechanism to control utilization management. They were flooded with demand for services when they opened up benefits, without evidence of differential outcomes. Many had to withdraw those benefits, because they lacked any way to access quality and cost of interventions. She said there was tremendous benefit in the actuarial potential alone. Ms. Giannini added there was also tremendous benefit in a provider being able to manage his own business. Providers were coding things as hours and not putting specific detail into their own billing systems.
Dr. David Riley, an internist in Santa Fe, New Mexico told how he'd administered an alternative medicine IPA of providers, doing peer review of quality assurance. The state insurance commission required that all insurance plans sold in New Mexico provide alternative medicine benefits, yet the insurance industry didn't have an effective mechanism and the lack of codes was a particularly onerous issue. He said it was impossible to develop actuarial data to look at what interventions might be most effective for certain conditions, because there was no coding information on which to base a rational decision.
Noting Dr. Mirin referred to comparisons of ICD-9-CM, Ms. Beebe asked if there were any comparisons to ICD-10-CM. Dr. Mirin said APA worked closely with WHO, so current DSM-4 codes would be identical to codes used in ICD-10. The advantage of ICD-10 was the descriptors would be much closer to what was understood about mental illness in 2002, rather than 1977. Dr. Mirin said APA would be pleased if the rest of the world adopted ICD-10 and APA would do what it had done in the past: the mental health portion of ICD-10 would be DSM-4 or its successors.
Dr. Cohn said, the way DSM was set up, APA seemed to have effectively expanded the ICD codes, giving them greater depth and understanding and developing a knowledge base around what the codes were and the requirements. Dr. Mirin concurred. To avoid confusion, APA used the same codes for the particular diagnosis, although sometimes in the crosswalk between DSM-4 and ICD-9-CM, it could be difficult for a practitioner to be certain that, apart from the name of the illness, they were actually talking about the same illness, because descriptions varied significantly from what was understood today. Dr. Cohn asked if it made any difference whether one used DSM or ICD in sending a bill to a managed care environment or a health plan. Dr. Mirin said practitioners uniformly referred to and used the DSM-4 code. In some instances, even Medicare used DSM-4. The numbers were identical.
Remarking that they tended to think of codes and guidelines being separate, Dr. Cohn contemplated whether DSM-4 might really be a large set of guidelines around the application of those ICD codes. Instead of asking for another set of codes, he suggested considering identifying DSM-4 as guidelines for the use of the mental health codes. Dr. Robert Charo explained that, while ICD-9-CM was the official code set for Medicare reimbursement purposes, DSM-4 was the official code set for audits and the Office of the Inspector General's purposes. Inspection forms in the CFR and inspection criteria for hospitals listed DSM-4 criteria and the inspector checked medical records against those criteria. DSM-4 spelled out the definitions of each code. Dr. Cohn said he thought Dr. Charo agreed with him. Dr. Charo said he did and he didn't: it was difficult having two words spelled the same with different definitions. Dr. Cohn asked if DSM had codes beyond what was in ICD. Dr. Mirin explained that the original set of codes matched ICD-9. As DSM evolved, medical advances resulted in more precise definitions of mental disorders and additional codes. Those codes and descriptions of those disorders were included in a separate appendix to ICD-9-CM. He said he'd check and respond back, but he believed those additional codes were included in ICD-10.
Dr. Zubeldia asked again about this issue. He said it seemed to him that DSM-4 and ICD-9 were the same codes with the same sort of definitions. He asked if the same code was used for conflicting definitions. Dr. Mirin said that was his point. They'd endeavored to make the coding numbers the same, in order to enable physicians across all disciplines to use the ICD system in coding diagnoses. But it was precisely in the area of definition that they were vastly different. Saying he understood they were different, Dr. Zubeldia asked if they were in conflict. Was the code for depression used for another disorder in the other code set? Dr. Mirin said the code number and illness remained the same, but it was like saying, “Malaria is a disease characterized by high fever, rash and death vs. this is a mosquito-borne ….”
Responding to the question about harmonization between ICD-10-CM and DSM, Ms. Pickett clarified that NCHS had worked closely with APA in the expansion and development of Chapter 5 and ICD-10-CM, so there was complete harmonization. NCHS worked well and collaboratively with APA so it would be almost seamless in ICD-10-CM. They couldn't accomplish that in ICD-9-CM. Expanding on the discussion about definitions, she noted diagnostic criteria weren't included within ICD. Chapter 5 contained a glossary and some definitions WHO placed there. But the clinical modification had deferred to the medical specialty groups that maintain diagnostic criteria independent of the classification and code number assignments. Diagnostic criteria for any medical condition (not just psychiatric conditions) resided outside the classification.
Dr. Humphreys asked if identifying clinical guidelines for the various parts of the code set (APA for this set, others for other sets) would cause any one a problem. Ms. Pickett said clinical guidelines existed for a number of medical conditions (one could find criteria for diagnosing various conditions on medical Web sites). Medical providers referred to them independently in diagnosing patients. Hopefully, when codes were assigned, that information was translated into the appropriate assignment of an ICD code. The criteria had been independently developed and resided outside the classification, but were routinely used by professionals diagnosing patients.
Dr. Kaplan clarified that NHIA's first choice was approval of the code set. However, they'd gone through the DSMO process and were rejected. Continuing to appeal wasn't an option. She said they felt they had a solution and were optimistic, despite language like “temporary coding.” They believed they had the expertise and input to make this process and everyone's life easier and help guide things forward, but they didn't have a role and victory was still tenuous and temporary. Dr. Fitzmaurice noted he'd heard NHIA was looking for greater appropriate use of these codes for home infusion and would like greater representation on coding bodies that consider codes for medical services. He also noted Dr. Kaplan hadn't mentioned the reasons for the DSMO denial. Dr. Kaplan referred back to reasons provided in testimony at the February hearings, reiterating that reading that testimony was the first NHIA learned of them. Dr. Fitzmaurice recalled from Dr. Kaplan's testimony that there was no explicit Medicare coverage, yet private payors covered home infusion for nearly twenty years. He asked if they'd used HIEC codes. Dr. Kaplan explained they'd used per-diem and payor-specific proprietary codes. Almost every one had their own and, even before HIPAA, NHIA tried to create a standardized code set that reflected what was actually happening in the industry and could be put forward as a national code set. But the administrative efficiencies that were the carrot accrued to providers, not payors, and they'd had no success in gaining standardization. NHIA saw HIPAA as a major opportunity to finally address the inefficiency of having so many coding systems for so many payors. HIEC codes hadn't gotten far in the marketplace; however that code set was similar to what was taking place there. Dr. Kaplan noted NHIA did a crosswalk of its codes vs. HCPCS and CPT's and the efficiency concerns she'd raised were reflected in mapping done against code sets used by many large payors
Dr. Mirin confirmed the assumption was that when anyone read ICD-9 or DSM-4 codes, they used DSM-4 definitions. He said the only cost to the provider was the price of the DSM-4 book (about $50); a volume with the codes and brief definitions, but no accompanying text, was significantly less. He believed the code definitions were available online at no charge. And he clarified that, while it was true that many medical specialty societies developed clinical criteria for particular diagnoses, the DSM-4 code set and accompanying criteria were the only ones used uniformly in the promulgation of both state and federal law and regulations. DSM-4 was referenced in 650 state regulations; OIG used DSM-4.
Dr. Cohn reiterated his own view that DSM-4 could serve as a guideline. He thanked Dr. Kaplan for providing a case study that created a solution, and said the Subcommittee would confirm how the final rules described and characterized guidelines. He said the biggest concern he'd heard that the Subcommittee needed to deal with was how groups with particular expertise could provide input into these codes. Dr. Kaplan said NHIA's most important issue was their temporary status. It damaged adoption of the codes--in three months, they might not have any codes at all. Noting it was difficult for a small group nobody understood to get its needs met within this process, Dr. Kaplan said she'd tried to focus on recommendations to make this process go better. NHIA was pleased to have achieved what it had with the help of the HCPCS panel, but sought a formal process for representation on the panel. Dr. Cohn asked Dr. Kaplan to cc the Subcommittee the letter NHIA sent the DSMOs regarding professional pharmacy services.
Panel 2 (contd.): Previously Identified Gaps in Code Sets
Mr. Shugart testified on behalf of the State of Maryland Medicaid program and NMEH's workgroup on gaps in the current HIPAA medical code sets: alternative medicine, home infusion procedures, and DSM. Mr. Shugart was a state representative for the National Association of State Medicaid Directors (NASMD) to both the Center for Medicaid and Medicare Services' (CMS') Alpha Numeric HCPCS Workgroup and Technical Advisory Group (S-TAG).
Mr. Shugart said he'd received comments about the need for alternative medicine codes. He said if those services were billed in Maryland, they were billed with local codes. In Maryland, home infusion services were processed electronically through the pharmacy Point of Sales system. Claims were submitted in NCPDP 3.1 format using NDCs. Plans called for processing these claims through the HIPAA compliant NCPDP 5.1 transaction set. Other states reported using coding structures for home infusion programs similar to the HIEC set within their local code schemes. HIEC methodology was considered good and the structure easy to understand. In Maryland, mental health claims were submitted using ICD-9 diagnoses. Professional services were billed using CPT-4, HCPCS, or local codes. Institutional services were billed using UB-92. At least one state used DSM diagnoses and would face gaps when ICD-10 diagnoses was implemented, because DSM codes couldn't be added to the standard transactions.
Mr. Shugart received comments welcoming the codes for acupuncture, massage therapy or naturopathic medicine, but noted that many codes in those code sets related to small increments of nursing time that seemed complex and might not be useful. He cautioned that adding another code set would increase complexity. He suggested that it would be easiest for the states and providers to have alternative medicine codes and home infusion expeditiously added to CPT-4, CDT or HCPCS. Another option was for home infusion services to be billed through NCPDP in the pharmacy system.
Mr. Shugart said he'd received comments asking for DSM to be included in the ICD-10 structure after ICD-10 was adopted. He noted that he'd recommended at the February 2 hearings that ICD-10 diagnoses not replace ICD-9 until two to five years after the HIPAA implementation date, even though this would delay implementation of DSM. He also noted collaboration between state Medicaid, mental health and state substance abuse programs, CMS, HIAA, and BCBSA concerning creation of national procedure codes for mental health and substance abuse services.
He remarked that, in retrospect, anticipating the number of change requests HIPAA required, the code maintenance committees should have developed an expedited process to address gaps in code sets and elimination of all local codes. The process had been labor intensive, slower than anticipated, and might result in extensive workarounds.
Mr. Shugart reported requests to use universal product numbers for medical equipment and supplies, which represented the largest number of local codes Medicaid programs addressed and its major gap. Significant collaboration and work between NEMI, CMS and the medical supply and equipment industry went into addressing development of national codes, but no one was confident this would be accomplished by the deadline. Considerable manual processing might be required to properly pay these claims.
Noting Medicaid home and community-based services waiver programs included both health and non-health services, Mr. Shugart said non-health services (e.g., home and vehicle modifications and homemaker services) needed to be incorporated into the standard code sets in order to insure uniformity of processing and reporting. He noted current code sets should be adequate to meet these programs' needs. Pointing out that many states used ICD-9 procedures on outpatient claims, Mr. Shugart asked the Subcommittee to consider allowing this code set to be used on standard outpatient transactions. And he reiterated that, if it was decided to replace the ICD-9 procedures with ICD-10 PCS, implementation should occur when ICD-10 diagnosis was implemented: two to five years following the October 16, 2003 deadline.
Panel 2 (contd.): Previously Identified Gaps in Code Sets
Ms. Leon-Chisen reported that AHA worked with HHS, CMMS and NCHS to ensure appropriate interpretation and application of ICD-9-CM codes. AHA also worked with HHS and AMA to insure that the same level of integrity and quality was achieved in the hospital application of CPT and HCPCS. Standardization of clinical codes was extremely important to AHA members, particularly as they related to HIPAA transaction standards. Hospitals and health systems were opposed to diverse reporting requirements for each of the clinical codes being considered. Dealing with multiple code sets would be costly and burdensome. And wide variations in use of the clinical code sets compromised the primary purpose of clinical code reporting: providing an accurate record of the patient encounter.
Ms. Leon-Chisen said AHA saw no gaps in the current HIPAA code sets that couldn't be addressed by approaching those responsible for maintaining the designated clinical code sets: ICD, CPT, HCPCS and CDT. AHA supported the current coordination and maintenance process for ICD-9-CM diagnostic and procedural codes. ICD-9-CM had a well-defined and effective maintenance and implementation process that was open, broad-based, and strove to consider needs of all users. Committee meetings were routinely scheduled and open to the public, proposed changes well publicized, and an annual date certain for adopting approved coding changes.
Ms. Leon-Chisen said AHA was encouraged by recent efforts of the AMA's CPT editorial panel to address the needs of AHA members and consider interests of non-physician stakeholders. But she expressed AHA's concern over the lack of institutional provider input into the HCPCS Level 2 update process. Public meetings had been held to discuss proposed DME HCPCS codes, but there wasn't a comparable process for the bulk of the HCPCS codes. She emphasized that the code development process for all codes should be: sensitive to needs of all users, capable of recognizing shifts in organized delivery systems, structured with easily accessible instructions on proper code usage understandable for a variety of provider settings, and maintained through a clearly defined organizational structure capable of achieving high-level coding integrity. The maintenance and implementation process should be predictable and account for the capabilities of all users in adapting to coding changes. Providers should be able to rely on routinely scheduled meetings in order to review coding changes and identify implementation dates. Ideally, coding changes should occur no more than once a year. Providers should be able to utilize the same code set version over this period. A uniform implementation date should apply across all users, including payors. Code request procedures should be available to the public. Code maintainers should be acutely aware of the need to also have a broad-based process receptive to the needs of all users. Long considered by many to be designed for physician use only, CPT needed to account for the needs of non-physician providers, especially when making decisions regarding new code proposals. CPT editorial panel decisions regarding codes shouldn't be based solely on how much physician work was involved in a service, but on whether there were sound business or other data collection and reporting requirement needs.
AHA members used ICD-9-CM to report mental health diagnoses and conditions. Psychiatrists primarily used DSM as a diagnostic tool for assessments. In recent years, several proposals approved by the ICD-9-CM Coordination and Maintenance Committee made ICD-9-CM more compatible with DSM. A crosswalk (and guidelines) was available between ICD-9-CM and DSM-4. Noting that hospital coders didn't diagnose patients but based their coding on documentation provided by the physician, Ms. Leon-Chisen said AHA believed that ICD-9-CM currently met the needs of hospitals and health systems for reporting mental health diagnoses and conditions. As Dr. Mirin had noted, ICD-9-CM and DSM-4's code numbers were identical.
Congress established the National Center for Complementary and Alternative Medicine (NCCAM) as part of the National Institutes of Health in 1998. NCCAM defined CAM as “those treatments and healthcare practices not taught widely in medical schools, not generally used in hospitals, and not usually reimbursed by medical insurance companies.” About 15 percent of community hospitals provided CAM services and the consumer generally paid. HIPAA electronic transaction standards didn't apply, because no electronic claim was submitted to a health plan. However, Ms. Leon-Chisen said AHA was sympathetic to needs of providers requiring a way to codify and classify alternative medicine services for other non-reimbursement related needs (e.g., research, benchmarking and public health tracking). She predicted that hospitals and other providers would report therapies considered complementary medicine as they were widely adopted or covered by traditional insurance. AHA strongly urged providers of CAM services to develop and present code proposals to the maintainers of HIPAA medical code set standards. AHA opposed proprietary or copyrighted alternative coding systems that added to the already significant expenses incurred by providers submitting claims. Many of these codes represented bundled or packaged sets of services and didn't follow traditional (mainstream) coding conventions. These codes might represent the provider or insurance coverage as well as the service provided.
Ms. Leon-Chisen said hospital and health systems that provided home infusion procedures through their home care services hadn't reported a gap in the current HIPAA code sets. AHA was encouraged that a substantial number of new home infusion codes were implemented in CPT-4 and HCPCS for 2002. If additional gaps in home infusion procedures were identified, she said AHA would support adding those services into existing HIPAA medical data code sets.
Panel 2 (contd.): Previously Identified Gaps in Code Sets
Michael Beebe, Director of CPT, American Medical Association
Dr. Beebe expanded upon his comments at the February hearings. AMA believed the current process for the selection of standards, transactions, and code set standards under HIPAA was successful. AMA agreed that the addition of new medical code sets would be disruptive and confusing to users, and that it was important that gaps in the existing data code sets first be remedied by the existing standards. He said AMA and the CPT editorial panel recognized a responsibility to work with other groups and had taken steps to expedite review and include new groups. CPT code change proposals, instructions for submission and a schedule with submission deadlines were available on the AMA CPT Web site. Code requests could be submitted via e-mail. AMA was developing and updating the Web site: the public would be able to view the panel meeting agenda and electronically comment on issues. CPT advisors could review and comment on code change proposals electronically and panel members could download agenda books. Any one could submit a CPT code change proposal. Mr. Beebe encouraged current and prospective users to work with the panel to correct any shortcomings.
In response to the CPT-5 project, Mr. Beebe reported that panel meetings were now open to all participating national medical specialty societies and the Healthcare Professionals Advisory Committee, which included the American Nurses Association; Occupational Therapy Association; Physical Therapy Association; Podiatric Medical Association; Chiropractic Association; Speech, Language and Hearing Association; physician assistants and dieticians. Mr. Beebe said the panel was reviewing options to further open deliberations and advisory input, while maintaining the rigorous peer review process and confidentiality they believed essential in preserving the integrity of the editorial process and maintaining the industry-wide annual update schedule.
Mr. Beebe said another issue from the CPT-5 project was the attempt to prepare the CPT code set for elimination of local codes. Private payors were surveyed to determine why local codes were developed and to capture a set of local codes for possible inclusion in CPT. Based on this survey, the CPT editorial panel, working with payors and the home health community, developed an initial group of home health codes. Further refinement of these home health codes occurred with the assistance of BCBSA.
The panel worked with HIEC to further refine the codes, approving those recommendations in time for CPT 2003. HIEC submitted subsequent changes to the panel in February, in time for CPT 2003. The panel considered new changes significant and re-established the Home Health and Home Infusion Workgroup to examine these coding issues. Mr. Beebe noted they'd also begun working with Alternative Link. He considered both experiences a good learning process for the panel. The HIEC codes initially were quite different from CPT or HCPCS codes, but he said both parties realized, after working together, that their needs could be met with a combination of HCPCS and CPT codes to cover services, supplies and equipment. The panel was still uncertain how Alternative Link codes were currently billed, but looked forward to collaborating and learning more about the needs of alternative medicine providers.
Panel 2 (contd.): Previously Identified Gaps in Code Sets
Dr. McCabe conveyed the American Psychiatric Nurses Association's, the International Nurses Society on Addictions' and the International Society of Psychiatric Mental Health Nurses' belief that care provided by advanced practice psychiatric nurses was distinct and different from care provided by physicians. She said psychiatric nurses' advanced practice was based on a distinct body of scientific knowledge different from medicine, built on theory, and tested through research. Psychiatric nursing was holistic and considered health needs and strengths of individuals, families, groups and communities, rather than orienting exclusively to illness.
She expressed their shared belief that a substantial gap in the current code set recognized by HIPAA made it impossible for psychiatric nurses to document care accurately or trace health outcomes and culminated in care disconnected from outcome measurement, making it difficult to control costs of care. Psychiatric nurses had little input into the development of these code systems, and no role in authoring, maintaining or authorizing these sets, yet were required by HIPAA to use them to code their clinical practice. Models of healthcare based on advanced practice psychiatric nursing had been demonstrated to be efficacious, cost effective and well received by consumers, but were poorly reflected in current HIPAA code sets.
Dr. McCabe said CPT and ICD codes didn't adequately capture psychiatric nursing interventions such as health promotions and health maintenance (e.g., general health assessment, assertiveness training, stress or anger management, parenting classes); promotion of self-care interventions (weight management programs, promoting recreational activities, facilitating development of the practical skills of community life--shopping, managing a checkbook and using public transportation); psychobiological interventions (relaxation techniques, sleep/rest scheduling, and other semantic complementary therapies); and psychiatric rehabilitation interventions (development of life skills, increasing use of environmental supports, and the design and implementation of mental illness prevention interventions such as screening and at-risk evaluations). ICD-9-CM code didn't capture functional disabilities that were often drivers of high healthcare costs (e.g., patients who lacked the ability to care for daily needs such as feeding themselves, obtaining and maintaining adequate housing, or self-administering medications).
Psychiatric nurses were unable to code for such common activities as nutritional education and counseling for a patient with anorexia nervosa or bulimia, relapse prevention and the recovering alcoholic adolescent, development and implementation of a weight management program for a schizophrenic patient at high risk for developing diabetes as a consequence of medication treatment, or preventive family education that wasn't therapy without using a collateral therapy code. Dr. McCabe said the lack of coding led to lost detail about patient need, lost opportunities to fully manage needed healthcare services, and an inability to fully track care rendered and relate it meaningfully to health outcomes.
ICD-9-CM codes were antiquated and less useful for psychiatric nurses than the DSM-4 codes. DSM-4 codes were modern, empirically tested, current, and gave detailed criteria for making a diagnosis not available with ICD-9 codes. Yet Dr. McCabe noted adoption of DSM-4 codes didn't resolve the issue of physician-oriented code sets, and still left psychiatric nurses with no way to code many health maintenance and promotion care practices. She reiterated that the lack of appropriate code sets forced psychiatric nurses to attempt to squeeze their care practices into existing code sets that didn't fully reflect their practice and, in that process, make arbitrary and imprecise decisions in order to match the care actually provided with non-representative codes.
Because the code sets poorly reflected the state of nursing science and had no mechanism to accurately code the totality of care delivered by psychiatric nurses, their practice was constrained to what was coded, limiting reimbursement and ability to measure outcomes of care. Without specificity, important details for analyzing delivery of mental health services were lost while others for costing healthcare couldn't be maintained. Dr. McCabe reiterated that using current HIPAA recognized code sets was time consuming and burdensome, didn't improve advanced practice psychiatric nurses' efficiency, and constrained their practice. Psychiatric nurses needed a code set that embraced care practices distinct and unique to nursing and had been demonstrated to be efficacious, cost effective, and produce positive health outcomes. All three professional organizations believed a significant gap existed in the currently recognized code sets, changes in those code sets were required, and psychiatric nurses needed to be included in the planning as new codes were defined and put in place.
Panel 2 (contd.): Previously Identified Gaps in Code Sets
Ms. Ward noted the DSMO Steering Committee included members of ASC X12N, HL7, NCPDP, NUBC, NUCC, and ADA's Dental Content Committee (DeCC). Referring to an earlier comment, she clarified that AMA, AHA and ADA weren't members of the steering committee, though the NUBC, NUCC and DeCC were DSMO members and each of these committees had many members, including the AMA, AHA and ADA. . For sixteen months, the steering committee has been addressing requests for changes to the transaction standards adopted by HHS for HIPAA. While most pertained to data elements used in X12 and NCPDP implementation specifications, several requests were for new code sets as standards under HIPAA. Ms. Ward said the steering committee recognized the benefit of a more structured process for considering additional code sets as HIPAA standards and believed it was being responsive to this need while adjudicating the monthly batches of requests. The DSMO was developing a document outlining steps submitters should follow, information about organizations available to them, and preparatory questions to consider before entering a submission into the change request system (CRS). Ms. Ward said a copy would be sent to a WEDI SNIP team developing a white paper on code set maintenance. The CRS guidance documentation would be available on the DSMO Web site (www.hipaa-dsmo.org). The DSMOs aimed to have this available to the public in May.
The Web site would provide an introduction to the issue, including an executive summary for familiarizing requesters with how code set requests were handled differently from requests for changes to implementation specifications. Another document would outline the framework for how the steering committee worked with the national code set maintainers. The DSMO chair would send a letter to all code set maintainers named in HIPAA explaining the DSMO's role and charge for maintaining the standards under the MOU and how the DSMO advised requesters to proceed. A liaison between the steering committee and each named HIPAA code set maintainer would also be established (e.g., someone from DeCC as liaison to CDT), enabling bridging potential gaps between what a requester expressed to the DSMO and what they needed to present to the code set maintainer. When appropriate, liaisons would work directly with the code set maintainers. They would also report to the steering committee on a regular basis.
The process for requesting that a code set be recognized as a HIPAA standard would be posted on the Web site. This process would emphasize that requests must first be addressed with all appropriate, existing code set maintainers. As with all DSMO recommendations, NCVHS would have an opportunity to review the decision. If, after the requester worked with the existing code set maintainers, it could be demonstrated that the business need hadn't been met, the requester would be advised to submit the request through the DSMO process. Any request most document code set maintainers approached, outcomes, and reasons the solution wasn't acceptable or the request was denied. The process would identify the next steps for entering a request, which began with selecting the enter-a-request option on the Web site and indicating a national code set from a dropdown box. (the steering committee was considering other options in addition to national code sets.) Selection of a code set triggered a questionnaire the requester completed to make the request. An error message indicated all required questions must be answered before a change request ID number could be assigned. Requests were considered by the DSMO in the next monthly batch.
The DSMOs were currently developing a questionnaire that would be added to the CRS. Questions and issues to be addressed included: (1) have you addressed the business need for adding this proposed code set with the appropriate code set maintainers named in HIPAA (a list and contact information was included)--those answering no would be guided to material regarding code sets and instructed to contact the appropriate code set maintainer (they could also contact a DSMO directly for guidance); if yes was selected, requesters were asked to two questions, (2) one dealt with demographic information (e.g., requesting organization, contact information), (3) the other asked to state a business case for adding a new code set (sub-questions dealt with entities affected by the request, percentage of overall business affected, documentation regarding the code set maintainers approached, outcomes, reasons why the solution was unacceptable or the request denied, and if an appeal has been filed with the code set maintainer--and the outcome or why that path wasn't pursued).
A U.S. postal mailing address would be posted on the Web site for people who want to send in hard copy information.
Discussion
Dr. Cohn thanked Ms. Ward for putting so much thought into how to deal with code sets and requests. He asked how this process would handle a request that initially should have triggered an analysis of the portion of the business need that had to be part of a medical code set and the part that could be handled by some other part of the implementaiton guide, for example, taxonomy codes. Ms. Ward said the inclination was to understand why folks came forward with new code set requests before addressing their need with existing code set maintainers. Noting they'd heard testimony for and against adding individual new code sets or incorporating current needs into existing medical code sets, she said DSMOs had to know that had been considered and the dialogue around the appropriateness of accommodating those code set needs in existing medical code sets.
Dr. Cohn clarified that he was asking how to handle a code set that addressed one field and required other fields to span needs. Ms. Ward said she heard him suggesting that the DSMO committee had to thoroughly review upon receipt any request for an additional code set and documentation, vs. first asking whether the requester had talked and worked with CPT or HCPCS. Dr. Zubeldia suggested the issue was similar to the situation with vision services: there was need for both a code set and a process to structure billing different components (e.g., lenses, frames). Ms. Ward remarked that a lot of work had to be done before they knew whether a code set was the answer to the vision problem. Dr. Zubeldia asked how the DMSOs addressed issues where the code set wasn't enough, and there was another component. Ms. Ward replied that many of the morning's comments (e.g., the HIEC testimony about only NUBC and NUCC being actively engaged) illustrated the need to educate people about how this process worked and what it was meant to do. When DSMO members felt a request affected their domain and constituents and had something to contribute, they participated in that discussion through the DSMO process. NUBC and NUCC were engaged because they were affected and opted in.
Noting a business need was a criterion for a new code set that comes to AMA or, with their appeal, to the DSMO process, Mr. Blair asked how each organization assessed whether there was a valid business need. Ms. Ward said the DSMO's were working on their guidance and criteria. Mr. Beebe said a business need translated in CPT-speak as a service performed by a large number of healthcare professionals throughout the United States. Mr. Blair asked if, by definition, that didn't preclude a code set addressing CAM. Mr. Beebe said CPT had chiropractic, osteopathic, and acupuncture codes. Anyone who could demonstrate a service was being performed by a healthcare professional throughout the country would be seriously considered for a CPT code. Asked if practitioners of Oriental medicine were considered healthcare professionals, Mr. Beebe replied they would be for the CPT process and the editorial panel, but not with AMA.
Dr. Humphreys pointed out that by 2003 there would only be the HIPAA code sets. If transactions were going back and forth in five states for procedures which hadn't yet been accepted, and so didn't have national adoption, those procedures had to be covered in some code set or they would have to send a form when everything else was electronic. Dr. Humphreys emphasized that there had to be a sense of what was a business need, and if insurers and providers sent electronic transactions for procedures that were legal where they were performed, a coast-to-coast requirement would bring the system to a halt. Mr. Beebe clarified that procedures that were legal according to a state law met CPT's requirements.
Mr. Blair said he agreed that it made sense when a code set applied for alternative medicine to discourage a proliferation of new code sets and first see if it could be accommodated within the existing code set. But he added that directing a new code set developer to a code set organization with practitioners and criteria that didn't represent that new healthcare domain appeared to be a way to delay and discourage adoption of that new code set and meeting that new business need. Noting the person that represented NUCC also represented AMA, Mr. Blair said it did seem as if it wasn't a level playing field. Ms. Ward replied that NUCC's participation in the DSMO didn't represent an AMA perspective. She said she sat (representing HL7) and had a vote on NUCC, as did Dr. Cohn and BCBSA, and many others. Results and decisions were representative of the committee, not AMA. If she'd thought otherwise, Ms. Ward said the Subcommittee would have heard. Mr. Blair said he completely accepted that.
Following up on Dr. Zubeldia's initial comments, Dr. Humphreys said helping to educate people and meet legitimate business needs through existing code sets made sense. But she saw a problem when people had a business need, but not sufficient sophistication to know whether they should ask for an adjustment, an additional data element, another feature within a transaction set, or apply to a code set maintainer. She suggested people needed something to help them work through that, rather than filling out lengthy forms only to be told by AMA that this was a transaction issue. Dr. Humphreys expressed confidence that the DSMOs tried to address that in their educational materials.
Trying to put all this in context, Ms. Ward remarked that this process couldn't be newer. Starting out, they tried to make decisions that worked for the entire industry. She expressed confidence that the DSMO Steering Committee had the expertise to hold and consider those conversations and seek a parallel track. She noted they didn't want a questionnaire so lengthy and time consuming that people didn't want to do it. Feedback suggested they might guide a requester through steps that resulted in more specific requests. Ms. Ward noted she also co-chaired the Claims Attachment Committee in HL7. She made reference to the DSMO website for requests and that X12 only posted a text box on their Web site, and the Steering Committee often had difficulty deciphering those requests. Requesters using the CRS told Ms. Ward the form / questionnaire for HL7 requests forced them to do their homework; when those requests came through, they were clear. She noted the Subcommittee brought up a good point that hadn't been discussed much in the context of medical code sets: what to do when there were issues about the transaction, as well as the code set.
Noting he'd expressed anxiety and concern about a level playing field for complementary medicine, Mr. Blair invited Mr. Beebe to correct his impressions by explaining how AMA considered accommodating CAM code sets. Mr. Beebe reflected that the editorial panel's experience working with HIEC provided a good example of the kind of dialogue needed. They'd looked at each other's code set and come to understand it. The panel saw how they could help HIEC in terms of coding and home infusion services in certain areas, and knew they couldn't help in others. HIEC had an amalgam identifying the actual healthcare professional providing the service, supplies, drugs, and the service itself. AMA realized it could only achieve one portion of that amalgam and that the provider taxonomy would achieve the rest. Mr. Beebe anticipated something similar when AMA sat down with alternative medicine. AMA was more than willing to look at the actual services provided. If there was a need, they'd expand current services to address it, but AMA might not be the complete solution. AMA wouldn't address which provider did the service; from their point of view, that was the provider taxonomy code's job.
Ms. Trudel said NCVHS' own characterization of a PCS was an additional tool that might be of use to DSMOs and maintainers as they looked at requests for code sets that didn't fit into existing code sets. Among the requirements were that codes had to be setting and provider neutral, limited to classifications of procedures, and free of diagnostic information or other data elements.
Ms. Giannini said she'd been on the X12 Committee and helped develop provider taxonomy. She said it was an extensive list of practitioner licensure types, but had no relationship to legislative issues according to state statutes, administrative “regs,” and case law. Ms. Giannini emphasized that if they were going to have 663 regulations governing one service, they needed an electronic way to determine if fraud and abuse regulations allowed service for a particular provider and state. She agreed that the provider taxonomy issue was separate from the procedure coding. However, using the modifier along with the code to get into information that existed in state laws, one could electronically adjudicate a claim that would otherwise cost people hours, time, energy and put the provider and the payor at high-risk of processing a fraudulent claim. Dr. Zubeldia encouraged her to work with the Provider Taxonomy Workgroup to ensure that the provider taxonomy covered their providers' needs. Ms. Giannini said the intent of HIPAA would be met, because the provider taxonomy could reduce paperwork and fraud and add specificity. Mr. Beebe said Ms. Giannini had summarized his view of the critical importance of Alternative Link: linking procedures, providers of procedures, and state licensing issues alternative medicine had to deal with across the country. He said doing all that on the back end to help process claims was a tremendous value.
Dr. Humphreys remarked on a distinction between the standard code set or provider taxonomy and a set of additional services rules, algorithms and data put together and highly useful to the community, that determined whether a particular claim could be submitted in a particular state for a particular service. She said only part of that fell within the definition of a HIPAA code set or provider taxonomy. Dr. Zubeldia agreed; noting they'd run into that problem with vision coding. He suggested there were at least three ways: an attachment, coded like compound drugs, or an implementation guide. He said they also needed code sets, and assistance in doing the right thing. He asked Dr. McCabe if a different coding structure was an alternative. Dr. McCabe clarified that the problem for the psychiatric nursing associations was representation and access to current code sets. The concept of the level playing field summarized it best. Many panels were physician controlled and access and representation was limited. Psychiatric nurses sought inclusion and the voice to make clear what they did and needed.
Mr. Shugart remarked that the mental health associations had submitted to the HCPCS panel specific mental health codes to describe services that might meet some of the psychiatric nurse's needs. Responding to another question, Dr. McCabe said some other classifications came closer to addressing needs vis-à-vis functional status. The difficulty was a philosophical perspective. Nurses perhaps were more health and preventative oriented than some other healthcare providers and much preventative work occurred before a disease was diagnosed. One could argue that preventative work clearly was cost effective, but she said it was difficult, within the current structure, to code for a parenting class before an at-risk child had a medical diagnosis. She said any strategy that looked at health and coding for health would enhance their work. Dr. Humphreys said Mr. Shugart touched upon issues CMMS would address in the next hearing. Dr. Mirin said a number of coding and definitional issues had to be addressed in an inclusive process. He noted that this was a key issue that needed to be addressed in the development of DSM-5. APA supported that, although they believed a single code set would be less confusing than multiple sets used to diagnose and ultimately treat people with mental illness.
Dr. Cohn expressed surprise that Mr. Shugart considered the lack of national codes for medical supplies and equipment a major problem. Mr. Shugart recounted that when the Medicaid program started there were about 25,000 DSM and DME codes established throughout the states. Some 5,000-7,000 codes explained supplies and equipment. There were differences, in part because of reimbursement methodologies. Many services done at the generic level were broad based, which he said was fine if that was how one was reimbursed. But if one was on cost basis or reasonable charges, going from ten codes to explain a service and proper reimbursement became complex. Supplies and equipment were more like national drug codes: thousands had to be adopted down to a smaller level. And some original codes were done from a Medicare perspective, whereas many Medicaid services were child oriented and distinct. Mr. Shugart said NMEH was working with the national associations to solve this problem, but people were nervous and wanted an answer today.
Dr. Cohn said Mr. Shugart's written testimony astonished him when it asked the Committee to consider ICD-9 procedure codes for outpatient claims. Mr. Shugart (speaking on behalf of Maryland, which had a Medicaid waiver--hospital reimbursement was based on a commission rate; they didn't price at the procedure level) explained that Medicare and others used ICD-9 procedure codes for paying hospital outpatients. He said they would follow the HIPAA rule, but questioned that would gain anything in processing claims. He noted the other state he'd heard from that didn't have a waiver also processed ICD-9 procedure codes and diagnoses.
Ms. Trudel noted some functional status was built into MDS and Oasis assessment tools used for home health and long-term care. Ms. Greenberg noted NCHS was working on the first of two issues with CMMS: a healthcare financing review collecting functional status for payment and quality scheduled for spring 2003. As they'd heard from Ms. Humphreys, the International Classification of Functioning, Disability and Health recently approved by WHO was a comprehensive classification for collecting functional status information that the full Committee singled out as meriting attention and further study. Ms. Greenberg said it would be premature to bring it forward now, but she predicted they'd hear more about it as development occurred.
Ms. Molina encouraged the Subcommittee to consider a couple themes they'd heard. There had been some 629 million visits to CAM practitioners in 1997, compared to 427 million for primary care physicians. Noting four years later there were still no codes available, Ms. Molina said this indicated a problem with existing code sets and implementing codes in response to public health needs. She asked members to look at attachments to Alternative Link's written testimony suggesting that, at least in the past, code set authorities hadn't been responsive to public health and business needs and, in fact, attempted to block access to administrative simplification for a whole category of practitioners. Ms. Molina said she was encouraged by the tone of recent interactions with AMA, but the testimony noted recent actions that reinforced the perception that there wasn't a level playing field. Alternative Link urged the members would to look at issues of a level playing field, public health, and business needs and make a rational decision in the interest of congressional intent and public health.
Subcommittee Discussion of Testimony on Gaps in Code Sets
Noting they'd heard about the different code sets from the developers' and industry's perspective, Dr. Zubeldia asked if it was possible for the Secretary to adopt the DSM coding guidelines while still using the ICD-9 coding structure. Dr. Cohn said the question of whether DSM could be a set of guidelines on how to code ICD or was incompatible was a valuable question. He'd considered them guidelines, until he'd heard about a crosswalk. Ms. Pickett said the staff would flesh that out. She noted that guidelines generally were to help someone apply codes appropriately and consistently, which was a different issue from what had been described about diagnostic criteria that morning. Diagnostic criteria for developing a definitive mental health diagnosis were a separate issue beyond coding guidelines. Dr. Cohn contended this was a slippery slope with shades of gray. Ms. Pickett recalled AHA saying the diagnostic criteria were what a physician used to develop a definitive diagnosis. Coders used guidelines to code from what the physician definitively diagnosed, without applying diagnostic criteria reserved for definitive diagnosis. Dr. Humphreys said the slippery slope would be having clinical criteria as part of a federal regulation. She asked if they'd go through a review procedure when DSM-5 came out. The guidelines for ICD-9 said, “Given the doctor reported X, code it this way.” That wasn't the same as: “This was how the doctor determined it was X.” DSM addressed that and it was “off the page” of an administrative regulation.
Dr. Zubeldia noted APA recommended the psychiatrists use DSM-4. He suggested they could support the use of DSM diagnostic criteria to support the ICD-9 for coding the diagnosis. Both had the same code for diagnosis. While this might be an appropriate position for NCVHS, Dr. Humphreys questioned that they wanted a federal regulation specifying a particular method. Regulations didn't get revised and published as rapidly as needed, and there could be other problems if the “reg” wasn't in synch with the revision. Dr. Zubeldia clarified that he wasn't saying support DSM codes but support the DSM diagnostic criteria. Based on the diagnostic criteria, they could come up with the ICD-9 code that represented a diagnosis. Ms. Trudel expressed concern over where this led. Ms. Pickett had mentioned this went well beyond a traditional coding mind set, and they'd heard that APA wasn't the only medical organization that developed and proposed its own clinical diagnostic guidelines. What if they heard from organizations with conflicting guidelines? Ms. Greenberg noted they'd also heard about conflicting language in ICD-9-CM and DSM; hopefully this was a short-term problem that adopting ICD-10-CM would resolve. Noting this linked into discussion of the ICD-10-CM update, Ms. Trudel proposed tabling the discussion until the next hearing. Dr. Cohn concurred. Meanwhile, he suggested they talk with colleagues and resources about crosswalk identification and how this might fit together. He noted they needed to reflect on whether, in this area, coding guidelines weren't closer than they'd thought.
Dr. Zubeldia remarked that the National Home Infusion Codes had an unanticipated solution: a combination of CPT and HCPCS codes seemed to work. Ms. Trudel said she would resolve with CMMS the issue of S codes being defined as temporary codes.
Dr. Humphreys noted another question was whether resolution was underway to the problem of overlapping and conflicting codes in CPT and HCPCS because of a disconnect with changes made simultaneously to both systems. Noting Mr. Beebe explained the CPT editorial panel's expectations in assigning those codes, Dr. Zubeldia said it might be a matter of education and documenting the connection between provider taxonomy, service with CPT, supplies with HCPCS, and how to use them in combination to produce the desired result. He suggested that the Home Infusion EDI Coalition put together an informational paper to educate the industry, especially payors. Dr. Kaplan reported that NHIA's national standard coding document had been downloaded free by about 600 organizations, including 100 payors. But AMA hadn't been able to rescind those codes for 2002: conflicting and duplicative codes wouldn't be resolved until at least the beginning of 2003 and potentially later. Dr. Kaplan said AMA was working towards an effective solution for the nursing visit component for home infusion within the CPT codes and NHIA was pleased with everything else surrounding home infusion that payors bundled in HCPCS. NHIA was optimistic that, eventually, HCPCS and CPT would fit together. But today it was still a major problem. Dr. Zubeldia concurred. NHIA could educate members and payors on the solution for next year. Dr. Humphreys said the Subcommittee planned to following up with AMA and seek clarification at the May hearings.
Recognizing confusion and overlap, Mr. Beebe said the panel was considering earlier release of the home infusion codes through the Internet, notifying appropriate payors this was being done to help them code these services and comply with HIPAA. Dr. Kaplan asked if the per diem home infusion codes that conflicted with HCPCS would be in the next CPT book. Mr. Beebe said they were working on that book now and needed to sit down with NHIA as part of the re-established Home Health Workgroup and figure out exactly what was wanted. Dr. Zubeldia noted the Subcommittee would act on the S codes Medicare or CMMS needed to tend. He congratulated everyone for his or her impressive work.
Noting NHIA, pharmacy, and other healthcare sectors lacked representation in the HCPCS and CPT process, Dr. Kaplan asked about future consideration of an advisory role. Temporary code designation would be one aspect of making this a more permanent situation, but there was no representation for pharmacy or home infusion within the Healthcare Providers Advisory Committee of the CPT editorial panel or on the HCPCS committee. Dr. Humphreys said that issue had already been raised and CMMS and AMA would address it at the May hearings. Dr. Yasnoff asked for an update in May, noting they should know of any problems.
Dr. Cohn observed that business needs didn't go away overnight; when a code set developer incorporated domains into its code sets it took on a responsibility for ongoing dialogue. He noted they needed to ask code set developers how they proposed to do that. Dr. Kaplan clarified that NHIA's most pressing concern was being housed in HCPCS, which led to concerns on the CPT side about potential ongoing issues within the Healthcare Professionals Advisory Group. NHIA sought recognition of its domain expertise. Dr. Cohn noted this was another question for HCPCS. He said full representation was needed on the HCPCS national panel, along with a structure for all the code sets incorporating interaction with domain expertise. Mr. Beebe concurred, noting there also was overlap between S and CPT codes. The editorial panel had concerns about how to eliminate that overlap, so that its participation in the HCPCS decision making process didn't result in duplicate codes.
Dr. Zubeldia said it was encouraging to hear that Alternative Link was working with AMA, and might bring their domain expertise to the CPT editorial panel to come up with new codes in CPT and perhaps in HCPCS that represented their needs. Ms. Giannini asked that when the members deliberated on including ABCcodes in HCPCS, they remember an X12 indicator (added to X12 4022, with implementation guidelines developed in the 4050) was also needed to realize the depth of information available from the code set. Dr. Zubeldia suggested incorporating at least some of Alternative Link's codes into CPT and HCPCS so they could be used immediately. In that case, the codes wouldn't need a separate qualifier. Internal medicine codes incorporated in the S section of HCPCS would become HCPCS codes. Ms. Giannini foresaw a design issue, but said Alternative Link was willing to see if that was a solution. Dr. Zubeldia encouraged them to work with the CPT editorial panel and the HCPCS code committee and bring a report to the May meeting. The agenda was already full, but if Ms. Pickett brought the report they could set some time aside for a discussion with Alternative Link, APA, and Mr. Beebe representing the CPT editorial panel.
Ms. Beebe noted that this also was an issue at X12. She drew an analogy of an issue she anticipated. A public health reporting guide had been accepted and X12 was building it, based on the 837 standard. Going to CPT as Dr. Zubeldia recommended, the guidance was that anything in the reporting guide didn't need to be in the institutional. She asked how the process took out of the HCPCS what had been put in while awaiting acceptance. Dr. Zubeldia explained that his understanding was that Alternative Link worked with the CPT editorial panel to build certain codes in the next CPT release. Because a CPT already referred to that service, the ABCcodes that came out as a standard probably wouldn't have those same codes, although there might be duplication. It would be up to the code set maintainers to tell their practitioners what to use. Ms. Pickett cautioned that this could create confusion and problems. As AHA testified that morning, it wasn't just the lack of a new code set, but how that rippled through all the vendor systems and the multiplicity of codes a provider then had to use to report the same procedure or service, depending on what payors accepted. Dr. Cohn reflected on the coordination needed between code set developers, saying he didn't know where these codes would find a home, but it once again spoke to what they'd already heard about HCPCS and CPT talking together better, so there was overlap. This percolated through the industry, and they had to be aware of it as they moved forward. Dr. Humphreys agreed; that was why they kept hearing it would be nice to have a similar procedure code set.
Members noted that even if the ABCcodes were recognized as an external code set in the 4050, the code set wouldn't necessarily be adopted in the HIPAA implementation guide. If adopted today, the code set couldn't go in the transactions because there wasn't a qualifier for them. The best short- and long-term solution was a CPT-HCPCS combination and provider taxonomy. Dr. Zubeldia suggested there might be a business need for time billing. Ms. Molina asked what time frame and objective criteria the Subcommittee used to assess whether the relationship between AMA and Alternative Link worked effectively. Dr. Zubeldia said the criterion was that both Alternative Link and AMA would report back at the May meeting on their perceptions of progress. Ms. Molina reminded members that Alternative Link existed because of philanthropic donations of a benefactor and time was of the essence. She expressed concern for the CAM and nursing communities if this took too long. Asked if a month-and-a-half was a reasonable time frame, Ms. Molina said any distance between them and the implemented code set was a challenge. They'd been trying to implement for six years and had gone through many rounds of contributions from their donor, tapping into resources heavily. She expressed concern about what this meant for public health.
Dr. Cohn advised Ms. Molina to set the expectations of Alternative Link's sponsor; a long HHS process had to occur before there was a final rule. Ms. Molina said she understood that, but lack of direction kept Alternative Link from being self-sustaining. Stakeholders indicated they'd mobilize behind the codes, once a firm direction was identified. Dr. Zubeldia noted version 4050 implementation guides weren't even in draft form. They had to go through draft comment for the industry, an NPRM process, and an implementation period. They were looking at 2004 or 2005. Dr. Zubeldia pointed out that Alternative Link didn't have to wait for the Secretary to adopt a code set before using it. Most sections weren't billed for insurance; they could use the code set today for public health or actuarial services. Ms. Molina noted most stakeholders looked to the Subcommittee and the DSMO to help them assess whether an investment in the code set was appropriate. Stakeholders already had significant financial burden in implementing HIPAA, and wanted to be sure they made good business decisions. Direction was needed quickly to assure that some aspect of this code set was going to make it.
Mr. Beebe advised the Subcommittee that AMA would still be working on CPT code change proposals for codes for alternative medicine when he gave the progress report in May. By then, hopefully AMA would have demonstrated progress on developing code change proposals for distinct portions of the alternative medicine codes.
Issues Regarding ADA Raised at February 6-7 2002 NCVHS Meeting
Frank Pokorny, Manager of Code and Third Party Issues, American Dental Association, updated members on ADA issues raised in February. He said he'd responded to issues raised by representatives from the National Association of Dental Plans (NADP), HIAA, and Delta Dental Plans Association (DDPA) in a March letter to the Subcommittee. He described the tone of the letter as appreciative of the issues raised, tempered by disappointment that members of the Code Revision Committee hadn't raised them during the January 28-29 meeting (planned by representatives of NADP and ADA to discuss NADP having a seat on the committee). Mr. Pokorny said it would be premature to address the outcome of those discussions, because they were only one step in the modification of the settlement agreement which defined the procedure and participants gathered to revise the Code of Dental Procedures and Nomenclature.
Ms. Trudel recounted that the previous testimony raised what Mr. Pokorny categorized as four key issues: participants, meeting protocol, process for future code review and revision, and access costs for CDT. She noted Mr. Pokorny responded to all four issues and read that morning. additional comments from Delta and NADP about his responses. Participants discussed the meeting protocol. Ms. Trudel noted there'd been a discussion about protocols for conducting the sessions. ADA recalled a vote to not use Roberts's Rules of Order, but operate more on a consensus basis. Ms. Oshensky, Director of Provider and Plan Relations, DDPA, said they didn't go over all the protocols and she didn't recall a unanimous vote on that procedure. Mr. Pokorny said he'd compared notes with his colleagues who noted a unanimous vote. He said they'd discussed an evolutionary process and “a spirit of working together in the collegial environment” mentioned several times during their discussion of operating procedure. He pointed out that he'd just learned from Ms. Oshensky about a board meeting of Delta USA and a need to review and approve the code changes. Noting he hadn't heard about a need for review and approval by Delta or other parties, he said they needed more feedback to ensure they worked together.
Recalling Ms. Ward's comment about the code sets and the DSMO process, Ms. Trudel observed that the coding committee's process also couldn't be newer. She said the question was how to work this out. Members noted no one wanted this to be a recurring item on the Subcommittee's agenda; the coding committee needed to address and take care of this at their next meeting. Mr. Pokorny agreed. He said one of the obligations ADA took on, because the committee couldn't begin the conversation on January 29, was an overall scheme for moving ahead: a timetable, activities, who did what, public notice, a process for solicitation of changes for the next cycle. He said he had an obligation to craft that, but first he had to be sure they had agreement on what they did January 28 and 29. He said a report of the Code Revision Committee's actions circulated to all participants. It recapped introductory comments, initial agreements about the operating protocol, and was “a recitation in detail of every motion, vote, tally, and resultant action” (deletion, change or addition) in the next version of the code. Mr. Pokorny said members' comments would be reviewed for agreement on April 18. A final version of the code set was intended to be available for use by the payor sector of the dental community on or about July 1, 2002.
Mr. Musco expressed appreciation for Mr. Pokorny's comments about putting together components of this process. He pointed out that another meeting of the Code Revision Committee wasn't scheduled and there wasn't a time frame for submitting code change requests or thoughts about the eventual composition of the committee. Much of the process still needed to be put into place. Mr. Pokorny said his first concern had been meeting information dissemination needs, so everyone knew what was coming and prepared for it. Ms. Trudel asked how other participants in this process could take on some of the load of drafting these procedures. Mr. Pokorny said, as secretariat, he'd assumed responsibility for drafting a discussion topic on how to make it work. He said he'd had several conversations with representatives and the feedback was valuable. Mr. Musco and he agreed that the next step was to put that into something people could look at and respond to. Mr. Musco said HIAA had conversations with ADA about a continuing process that kept code revisions at the forefront so, as needs arose, code revisions could be suggested to ADA. He also noted a need to keep in close contact with the Code Revision Committee. Mr. Pokorny said that was the social glue that made an organization work over time. He expressed confidence they'd find ways to talk more frequently. Ms. Oshensky said DDPA would be more than willing to participate in developing a timeline; she said she thought HIAA would also help. Mr. Ward said NADP would volunteer, soon as they become part of the committee.
Mr. Ward noted the dispute over what happened at the meeting and the procedures showed a lack of necessary structure: they needed to determine the procedures and make sure they worked together. Mr. Pokorny said he'd talked with Evelyn Ireland who volunteered to keep NADP's members informed about upcoming changes and the process. He noted ADA was also dealing with harmonizing the data contents of the electronic and paper claim form, something the others agreed was important to do together. He said he fully expected this to carry forward into the code revision process. He agreed that Ms. Ward said it well: this was the start of a new process. Circumstances were different now, and they could build.
Mr. Musco said representation at the initial meeting wasn't representative of payors and providers. HIAA welcomed additional input on the Code Revision Committee from NADP. The representative from CMMS was supposed to represent a payor perspective, but HIAA considered the representative was from the Medicare side of CMMS, which didn't pay for dental benefits. He suggested that someone from the Medicaid perspective who looked at and reimbursed for dental services might have more of a payor perspective. Ms. Trudel noted Mr. Musco's concern.
Mr. Phil O'Brien, General Counsel for DDPA, said the association strongly supported the settlement agreement and recognized the process wouldn't be perfect, but that they had to work together. Once they had a process and rules of protocol to govern meetings, a lot of issues would be minimized. He said DDPA was open to changing the fifth payor “rep” on the committee, because the national purchaser representative called for in the settlement agreement hadn't proven workable. He said they wouldn't oppose NADP as representative. Jay Grant, Washington counsel for NADP, confirmed that NADP was willing to participate as the fifth representative. Dr. Zubeldia said the Subcommittee wanted to hear when this was resolved. Mr. Pokorny said he had the feeling it might be beneficial to report on progress to the Subcommittee periodically. Dr. Zubeldia suggested reporting by letter or e-mail, at least until issues were settled and the group worked in harmony.
Mr. Blair remarked that, in listening to the testimony, the issues seemed to mostly involve governance. He suggested considering a way to identify an independent secretariat that all sides felt comfortable with. Mr. Pokorny asked if that would embrace an independent chair. Mr. Blair said they were talking about a mechanism to achieve an additional degree of harmony and confidence in the working relationship. He would leave those decisions to them. Mr. Pokorny recalled a statement in the settlement agreement that the ADA would be secretariat. Mr. O'Brien said DDPA was willing to discuss an independent chair; the settlement agreement could always be modified. Dr. Zubeldia said it was a suggestion he hoped wouldn't be necessary; with a little collaboration, they might be able to work out the friction. Doing this on their own could be easier.
Mr. Musco asked about an issue raised at the Subcommittee's last meeting and in today's testimony: accessibility and availability of the code set at reasonable cost. Dr. Zubeldia noted they'd heard DSM sold their book for $50; CPT's was between $35-$45. Some developers incurred a cost, and there was intellectual property in the codes. He believed the CDT code set was in that same range or lower. Mr. Musco pointed out there also was a licensing arrangement. Mr. Pokorny explained that a licensee would normally be a third party payor organization; the fee for a license to use the code was $1000. A vendor of a practice manager software system would pay about $10 per seat. Mr. Ward said NADP had some very small plans working in very small areas that felt a thousand dollars was unreasonable. Mr. Blair asked if ADA could accommodate a separate fee schedule for payors smaller than a designated size. Mr. Pokorny said he would bring these comments and discussions to those within ADA who could make that decision.
Dr. Zubeldia remarked that whether there was any consideration for small payors was up to ADA. Personally he thought a payor who couldn't afford $1,000 for the code system on which payments were made probably needed to consider another business. Mr. Musco pointed out that one issue raised was the equitable nature of the cost differential between payors and providers and requiring licensure just to be able to use the codes to process claims. Ms. Greenberg commented that code set developers charging a license fee was part of a broader issue, raised with several of the code sets as well as the NPRM process, and part of the Committee's ongoing review. Noting Mr. Blair added cost to their list yesterday, Dr. Cohn noted they needed to have a panel devoted to feelings about the costs and any solutions. Mr. Grant said he would ask NADP's executive director to respond with a letter.
Ms. Ward offered to be available for the next meeting. Dr. Cohn said that beyond a bit of agenda planning (Ms. Ward had proposed having the annual DSMO review at the October meeting and Ms. Trudel said she'd provide guidance in terms of timing) there wouldn't be a need for a DSMO representative. Drs. Zubeldia and Cohn adjourned the meeting at 12:30 p.m.
I hereby certify that, to the best of my knowledge, the foregoing summary of minutes is accurate and complete.
/s/ 1/22/03
_________________________________________________
Chair Date