Thank you for inviting NANDA to give testimony at this hearing from the perspective of a classification developer. NANDA began its work 24 years ago when a group of nurses met to develop a vocabulary that captured nursing's contribution to patient care and was coded to enable computerization. A nursing diagnosis is a clinical judgment about individual, family, or community responses to actual and potential health problems and/or life processes. Nursing diagnoses provide the basis for selection of nursing interventions to achieve outcomes for which the nurse is accountable (NANDA, 1990). There have been twelve biannual conferences where the membership has approved additions to the NANDA Nursing Diagnosis Taxonomy (1996). Though a consensus process is used and the membership provides major input to the development and acceptance of a new nursing diagnosis, the criteria for inclusion are stringent. The new diagnosis must be research and evidenced based. NANDA developed a staging system, based on the DSM system, to facilitate the development and maintenance of the Taxonomy. NANDA was the first nursing classification to be recognized by the American Nurses Association.
1. What clinical codes do you use in administrative transactions? Strengths? Weaknesses?
Nursing has never had clinical codes used in administrative transactions. This has been a major problem in nursing. The charges for nursing care have always been embedded in room charges or some other charge, i.e. physician service. Therefore, a major resource in our health care system has never been captured so that impact to the system can be projected. Many policy analysts believe this contributes to the various rounds of nurse shortages and overages (the pendulum must swing widely since there is no statistical way to predict need). Recently, the advanced practice nurse who has a masters degree (nurse practitioner, clinical nurse specialist, nurse midwife, and nurse anaesthetist) have achieved the ability to charge for service. They use CPT 4. ANA has lobbied the AMA for many years and now has advanced practice nurses sitting on CPT panels. However, 80-90% of the nurses in this country have no administrative codes that cover their practice. The impact of nursing care on patient care is enormous, yet our administrative codes fail to capture this contribution to patient outcomes. Several major studies have demonstrated that they can predict allocation of resources and patient outcomes when nursing care is factored in (Halloran; Knaus [APACHE]).
2. What clinical codes and classifications do you recommend? Suggestions for implementation?
NANDA is a "niche" classification, designed to capture the conditions for which nursing is accountable. As such, we have been collaborating with other classifications to produce more comprehensive codes for health care. We have been working with the editors of SNOMED International and have placed nursing diagnoses in the Functional Axis. In two recent coding studies conducted by the CPRI (in which I participated) the addition of NANDA to SNOMED significantly strengthened its performance in capturing data from patient records. Those classification which did not have nursing content--ICD-9-CM and CPT 4--performed at a significantly less capability.
Since nursing does not participate in coding, we have no recommendations for implementation. However, NANDA is in every nursing textbook, taught in our schools, and is throughout our licensing exams, nurses know the terms and can select them rapidly from a code list.
3. Which version of ICD should be used?
NANDA has looked at ICD-10 (both ICD-10-CM and ICD-10-PCS) and is in favor of using it since it codes more human-level conditions, not just disease or organ-level content. Nursing diagnoses and treats human responses to health conditions, not the disease conditions. ICD-10 supports this view. When ICD-10 was being developed, NANDA submitted its work for inclusion and even developed a translation of our work into ICD-10 code format. This work was not accepted as only the International Council of Nursing could submit work to the WHO. Only the ANA could submit work to the ICN, and this work was done prior to ANA recognition of NANDA's work. Since then the two organizations have developed close working relationships. However, the ICN has decided to create its own classification the International Classification of Nursing Practice. The team producing this work is European, and while Americans participate, they are not leaders. NANDA has offered assistance to this group. As a result of our willingness to collaborate, we have become, along with the ANA, reviewers of this work. I will be meeting with this team in a few weeks to discuss issues of coding approaches.
While many are in favor of initially keeping ICD-9-CM and CPT 4 due to implementation readiness and the short time frame, that reluctance to change will always be there. If practicality forces that decision and it is known that the ICD-10 will replace the ICD-9- CM and the CPT 4, that announcement with a tentative time line for implementation must be release as soon as possible. The health care industry must know that the change is coming or it will always be put off and fought against. Worse yet, resources will be used to develop new classification instead of being used to collaboratively develop the ICD-10.
4. How would you reduce administrative burden, but obtain clinically meaningful information?
NANDA supports the migration to the CPR as rapidly as possible. In my own institution, I am co-chair of a team to develop and implement a computerized, multidisciplinary problem list. The clinician will select from a list of problems the appropriate problem to add to the patient's record. This list is already mapped to the ICD-9-CM, SNOMED, NANDA, and UMLS. Our billing office is also involved in this project to insure that administrative transactions are assured. In this way, clinical need for data are met as well as administrative needs.
5. How should ongoing maintenance of code sets and intellectual input and funding of maintenance be addressed? What are the arguments for having these systems in the public domain versus the private sector? Copyright?
NANDA supports a modification of public domain system. Our system is copyrighted. Royalties are collected from publishers and other vendors who use our work to develop a product that is sold for profit. This income is used for fund the ongoing development and maintenance of NANDA since there is no other current source of money for this work. Clinicians may used the work by requesting permission for use, but there is no fee. We want the work used. The copyright allows us quality and version control. We and other groups publish textbooks and manuals for using the NANDA Taxonomy.
We have problems with proprietary systems that are a major source of income for an organization and that are required for use by the federal government and other payers. The classification should be copyrighted but be free to users. Money to support development should result from sales of tools and educational products concerning the use of the classification or should come from a government/private sector collaboration.
6. What would be the resource implications of changing from the coding and classification systems that you currently using to other systems?
As a developer, NANDA would need to re-evaluate its strategic plan and budget. We have several collaborative projects underway with the ICN, SNOMED, the ANA, the UMLS, and three research teams at the University of Iowa (nursing diagnosis extension project, the Nursing Intervention Classification, and the Nursing Outcome Classification) that would need to be reconsidered. However, we are deeply committed to serving the data and information needs of the nursing profession and will work collaboratively to achieve that goal. We believe that as these systems capture more clinical variance to facilitate control over the process and outcomes of care, health care costs will decrease and this will offset any start-up costs or costs of changing systems.
7. Concerns about the process being undertaken by DHHS to carry the law in regard to coding and classification issues?
At this time, we have no concerns about the process. The Department seems to being listening to all the stakeholders for which we are very pleased. We plan on monitoring the process and are willing to provide whatever assistance we can to this process. We see this activity as a major way to standardize health care data, both administrative and clinical, which will improve the quality of care and hopefully reduce its costs.
We do caution the Committee to be aware of the implications of word use. Please be concerned about the health care community, not just the medical community and about health care in this country, not just medical care. We have become a country that is concerned about wellness and health promotion, not just disease and disease prevention. Let us strive to develop codes and classification that reflect the needs of the country.
Additional questions:
1. Can one system serve most purposes? If not, which systems can serve most functions?
No one current system can serve all needs. NANDA supports mapping efforts between systems if the owners of the systems allow that to occur. We are currently working with UMLS, SNOMED, and the ICN to develop mappings between the systems. These organizations are willing to collaborate. We do not have access to work with WHO on any ICD project, but would if it were possible.
2. Is it simpler to use same disease classification for administrative and statistical reporting?
Not for nursing data, since we do not treat disease but the human response to it. Statistical reporting, if I understand what that means, must include clinical data; administrative data is skewed to the billing process and does not accurately reflect clinical data. Nursing is not represented in any administrative system. We have worked with Werley as she developed the Nursing Minimum Data Set (NMDS), of which nursing diagnosis is one data element and NANDA is a set of appropriate values, and we have worked with the American Organization of Nurse Executives to create the Nursing Management Minimum Data Set which captures workload and work force data to be used with the NMDS. Again, whatever classification system(s) is used must include nursing diagnoses and nursing interventions.
3. To what extent do you feel that your discipline and practice setting are well represented by the current system?
Nursing is invisible in the current systems, yet significantly influences patient care and outcomes. The only way to capture this influence is through research which is very expensive to conduct. The National Institute of Nursing Research in NIH no longer funds this type of research, nor do they fund classification development research.
4. What issues do you encounter linking data coded with different classification systems and trying to crosswalk between them?
NANDA has participated in several projects. A previous president assisted the NLM in adding and cross walking NANDA into the UMLS which required mapping to the Home Health Care Classification and the Omaha System (two other ANA recognized nursing classifications) and I have been working with one of the editors of SNOMED to do the same thing (including mapping the fourth ANA recognized classification, Nursing Intervention Classification). In the first project a set of filters was developed that facilitated an accurate mapping. These were used to facilitate mapping in SNOMED but using their rules of classification, not UMLS's rules. We are still working on developing the filter or other approach to cross walking the interventions.
The major issue in cross walking between a medical and a nursing classifications is the issue of disease versus human response. The nursing diagnosis of Pain is classified as a symptom in SNOMED; however, the scope of treatments between the clinicians is different. Mapping between different scopes of practice is like mapping behaviors between different cultures--it can lead to great misunderstandings. You need experts from each discipline that are open to understanding and learning about the classifications in the other disciplines.
5. What are the impact on and implications for current clinical classifications as we migrate towards a CPR?
It is very difficult to change codes and the database structure for the codes. We must designs structure that facilitate ease of changing codes (because that will be a fact of life as health care science and knowledge evolves). However, as migration occurs to a CPR, the computer can provide the code (unseen to the user) when the clinician directly enters the patient's diagnosis and interventions/orders. As a result of the two CPRI studies, a recommendation was made the SNOMED was an excellent clinical classification and could be used to provide decision support to the clinician. UMLS can also be mapped to the clinical data and thus provide bibliographic retrieval support for the clinician by organizing literature searches around the actual patient data in question. These and other advantages have all be noted in the Institute of Medicine's report on the computer-based patient record. NANDA fully endorses this report and vision of the CPR. The CPR can support nursing classifications and make visible nursing's contribution to patient care and outcomes.
One note of caution that NANDA would give to other developers, since we have committed the error ourselves, concerns the development of code structures for the classification systems. There are two issues about coding: 1) ease of the human user, 2) ease of the machine user. Coding is the abstract representation of a concept. Classification is commonly used to reveal the relationships between concepts. When a human user is considered, classification trees reveal these relationships in a pictorial way that can be readily grasped and understood. A code can be developed that represents the location of each concept within the classification. The problem begins when the basic understanding of the classification system change or evolves to include new or the creation of different relationships. The classification trees must change to help the human understand the new relationships. Therefore, the code structure must also change as it encodes information about the concepts. For humans we can provide new understandings and map the changes from the old to the new even though the mapping is not a precise translation of the new location/relationship of the concept. A machine user cannot "understand" this new relationship unless it is a one-for-one change (which rarely occurs in the world). Changes are usually "kaleidoscopic" in nature--same elements, very different patterns and relationships. Therefore, codes for the machine user need to be without meaning or random. Only the concept is coded not the relationships or locations in the classification systems. The humans may change the classification system (relationships with other concepts) without changing the identification of the concept. Classification trees assist human understanding and confound machine handling of the concept. As developers of classification systems, we should code concepts for the ease of the user. Today the user is a machine. The quantity of information has grown too large for a human to remember and/or manipulate. A machine does this very well. Humans manipulate large quantities of information with metaphors or models--machines cannot. Therefore codes should have no information about the classification model encoded in them. NANDA, after consultation with the NLM, recommends a multi-digit (alphanumeric) code (for machine use only) as the true code for the concept and a conceptual model to communicate the concept relationships to the human user. To rephrase, as long as concepts keeps same code and codes are not reused when concepts are deleted, how we display NANDA to nurses/humans may change over time without having to completely re-code and handle mapping of same concepts from one version to another. Secondarily, codes without meaning will facilitate different cultural views or relationships of the same concepts so that all nurses can use NANDA and share data about the concepts (which are language neutral, much like the GALEN project) while exploring how relationships are the same and/or different.
A final note of concern, as more of health care migrates to the home and families become the major caregivers, they will need a language to document the observations and care they deliver. As our health care system evolves, we must make sure that all care providers, whether professional or family, contribute to the documentation of clinical care and the resources given to support that care. Without a doubt a traditional clinical classification filled with professional jargon will not serve this group of care givers. The CPR has a goal of becoming a life long record across many sites. Let us not forget to consider these sites that are not hospital or clinic abased.
6. What is the market acceptance of the recommended clinical classification? What are mechanisms for low cost distribution?
NANDA is widely used in schools of nursing, by all major textbook publishers, and numerous hospitals and other sites of practice. We actively encourage all nurses to submit candidate nursing diagnoses for the ongoing development of the NANDA Taxonomy. The ANA has recognized NANDA as the nursing organization responsible for developing and maintaining a nursing diagnosis classification. We are a member of the Nursing Organizations Liaison Forum which brings together all the nursing speciality organizations to address common concerns. A major concern is the classification of nursing phenomena and the creation of clinical guidelines. As a result, the specificity of some of our diagnoses is developing and the flow of new diagnostic submissions is growing.
NANDA does not charge a fee for using the Taxonomy in schools of nursing or facilities that provide nursing care. We require permission for use so that we can ensure quality and version control. Only organizations that use the Taxonomy in the development of product that generates income/profit for that organization must pay royalty fees.
7. How might NCVHS work with DHHS to assure that the USA coordinate development of an international medical dictionary, classification?
Any international effort should be coordinated through WHO. It is arrogant to think that the United States should coordinate an international medical system. This approach did not work for nursing. NANDA strongly supports that all discussion concerning coding and classification be focused on clinical systems, not just medical. There is discussion of the formation of a new ISO committee. NCVHS and the DHHS might join the ANSI-HISB effort to become sponsors of this ISO committee and that way provide leadership in this effort--this will require money and resources.
8. Is it practical to move to a single procedure classification?
It is practical, only if the procedures of all disciplines are included, not just the ones that currently receive reimbursement.
9. Within a selected classification, should providers be able to use all codes or should requiring the information be allowed to restrict reporting of certain codes?
NANDA supports the clinician being able to accurately describe and code the care given. They should not be forced to use Procrustean measures that distort the work in order to be reimbursed for their work. The implications of this question are enormous. The ANA published a monograph (Zielstorff, R. D., Hudgings, C. I., & Grobe, S. J. (1993). Next- generation nursing information systems: Essential characteristics for professional practice.) that described a data pyramid. At the bottom was the accurate clinical data. At the top the data had been abstracted and summarized for use by policy makers. This data no longer represented the actual clinical conditions but only those that administrative transactions allowed. Yet we set public policy and laws based on that flawed data as if it accurately represented the health of our country.