Outline
Graphics
Processes for validation and coding of the taxonomy for the Nursing Interventions Classification (NIC) are described.A sample of nurses expert in theory development rated the NIC taxonomy using five criteria. Following identification of a stable structure, the taxonomy was coded with each intervention receiving a unique number. A coded and valid taxonomic structure facilitates use of the classification in computerization and makes possible the collection of comparable data. A coded classification can also be used in reimbursement systems.
Efforts to construct a standardized language for nursing treatments have been underway since 1987 by a large research team at the University of Iowa. In May 1992, the Nursing Interventions Classification (NIC) was published by Mosby Year Book (Iowa Intervention Project, 1992) [11] and in fall of 1993 the NIC taxonomy structure was published in Image (Iowa Intervention Project, 1993). [12] This article describes the recent effort to validate the taxonomy and to code our NIC interventions.
Development of a valid and coded taxonomy of nursing interventions that is comprehensive to all of nursing practice is a unique occurrence. For the first time in the history of nursing, we now have a standardized language for nursing treatments that can be used in practice, education, and research. Use of this standardized language will facilitate the inclusion of nursing treatments within health care data sets and the computerized patient record and help to determine the effectiveness and costs of nursing care. The use of NIC makes nurses and nursing a more visible part of health care.
Overview of NIC^
Nursing Interventions Classification (NIC) is a comprehensive standardized language for nursing interventions organized in a 3-level taxonomy. NIC describes all interventions that nurses do on behalf of patients, including independent and collaborative interventions, and basic and complex interventions. The classification is useful to nurses in all specialties and in all settings. While an individual nurse may have expertise in only a limited number of interventions, the entire classification captures the collective expertise of all nurses.
NIC contains a standardized list of over 400 interventions performed by nurses.* Each intervention has a label name, a definition, a set of activities that a nurse does to carry out the intervention, and a short list of background readings. NIC interventions include the physiological (e.g., Acid-Base Management, Airway Suctioning, Pressure Ulcer Care) and the psychosocial (e.g., Anxiety Reduction, Preparatory Sensory Information, Home Maintenance Assistance); those for illness treatment (e.g., Hyperglycemia Management, Ostomy Care, Shock Management), illness prevention (e.g., Fall Prevention, Infection Protection, Immunization/Vaccination Administration), and health promotion (e.g., Exercise Promotion, Weight Management, Smoking Cessation Assistance). There are interventions for individuals and families (e.g., Family Integrity Promotion, Family Support). Recently, indirect care interventions (e.g., Emergency Cart Checking, Supply Management) have been added to the classification.
*The first edition of NIC published in 1992 had 336 interventions; the second edition of NIC which will be published by Mosby Year Book in 1996 will have over 400 interventions.
Ongoing research to develop, validate, and implement the NIC is being conducted by a large research team at the University of Iowa with funding from the National Institute of Nursing Research. Research methods to develop and validate the interventions included content analysis, expert survey, and focus-group review. Principles of label, definition, and activity construction were established for consistency across the classification. An ongoing review process is used to continually update and expand the classification. The placement of the interventions in a 3-level taxonomic structure facilitates selection of an intervention. The taxonomy was developed inductively by the team members using the methods of similarity analysis and hierarchical cluster analysis.
Validation of the Taxonomy^
The NIC taxonomy, before validation and as reported in Image (Iowa Intervention Project, 1993), [12] was a 3-tiered structure with 6 domains, 26 classes, and 357 interventions. At the highest, most abstract level are the domains. Each domain includes classes (groups) of related interventions, which represent the second level of the taxonomy. Each domain and class has a definition that helps to place and locate specific interventions. The third, most concrete, level of the taxonomy consists of the interventions that nurses can choose to use with particular patients. A few interventions are located in more than one class when they are integral to both classes. A rigorous research approach was used to develop the taxonomy. The interventions were first sorted into similar groups and then data from the group placements were entered into a computer program and hierarchical clustering was used to analyze the number of raters who put every two interventions in the same group. Small group review was used to examine the statistical output and determine the clinical usefulness of the results. An 8-step developmental process, which used powerful quantitative statistical analysis and expert review and opinion, is explained in detail in the 1993 Image article. While the construction of the NIC taxonomy was built logically from the interventions themselves, the work was done by the members of the research team.
To ensure the taxonomy was valid, a review by others external to the project was desired. Since the interventions within the taxonomy had been developed with the assistance of expert practitioners and since two intervention validation surveys had been recently completed by practitioners and heads of specialty practice organizations (Barry-Walker, Bulechek, & McCloskey, 1994; Bulechek, McCloskey, Denehy, & Titler, 1994; Steelman, Bulechek, & McCloskey, 1994), [3,5,16] a sample of researchers and theorists was targeted for this study. It was judged that those who had some knowledge and experience with research methods and theory construction, as well as knowledge of the discipline of nursing, would make the best reviewers of an intervention taxonomy.
Thus, a questionnaire to assess the meaningfulness of the classes and domains was distributed in May 1993 to a sample of nurses who belonged to three interest groups in the Midwest Nursing Research Society (MNRS): 295 individuals from the interest groups of theory development, qualitative methods, and nursing diagnosis were polled for their willingness to participate in a validation survey of the NIC taxonomy. Those sent questionnaires were 161 MNRS members who responded that they were willing to assist with the research; 130 (81 percent) returned the questionnaires. For the analysis, 121 surveys were used (9 were returned late or did not include all necessary information). People in the sample were, on average, 47 years of age with a mean of 24 years of nursing experience. Twenty (16 percent) had a master's degree and the remaining 101 (83 percent) had a doctorate. Most of the respondents were employed as faculty in a school of nursing but 14 percent were employed by a hospital or other clinical practice agency. Respondents reported that they had expertise in the following areas: research methods (81 percent), theory construction (52 percent), clinical practice (84 percent), and administration (34 percent).
Participants were each supplied with a copy of the complete taxonomy and a questionnaire survey. In the questionnaire survey, each of the domains was color coded to match the color of the domain in the taxonomy booklet. This made it easier to find and rate the domains and the classes within each domain. Each participant was asked to rate each domain and each class according to how characteristic it is on five criteria:
1. Clarity: Class label and definition are stated in clear understandable terms.
2. Homogeneity: All interventions are variations of the same class.
3. Inclusiveness: Class includes every possible intervention.
4. Mutual Exclusiveness: Class excludes interventions which do not belong.
5. Theory Neutral: Class can be used by any institution, nursing specialty, or care delivery model regardless of philosophical orientation.
For each criteria, the characteristics of each domain and class were rated using a 5-point scale: 1 (Not at all characteristic), 2 (Slightly characteristic), 3 (Somewhat characteristic), 4 (Quite characteristic), 5 (Very characteristic). Data were entered into a mainframe computer and the Statistical Analysis System (SAS) was used to determine the number of respondents and frequencies for each criteria for each class and domain. Ratings four (Quite characteristic) and five (Very characteristic) were summed for each domain and class. In Table 1 (domains) and Table 2 (classes) the areas where there is less than 80 percent agreement are highlighted for ease of interpretation. (Eighty percent agreement was selected to be constant with the established convention in other diagnosis and intervention validation studies [Bulechek & McCloskey, 1992; Fehring, 1986].) [4,7]
1. All ratings were high: (a) 77 percent of the respondents rated the domains as either quite characteristic or very characteristic on all criteria, and (b) 88 percent of the respondents rated the classes as either quite characteristic or very characteristic on all criteria.
2. The criteria of theory neutral and mutual exclusiveness received the highest ratings.
3. The criterion of inclusiveness received the lowest ratings.
4. The "Physiological Complex" domain received the highest ratings and the "Health System" domain received the lowest ratings.
Respondents were also asked to make comments on each rating. We received numerous helpful observations. All comments were typed into a computer file and printed by domain and class. Next, six subgroups of the research team were established, one for each of the six domains. Each subgroup met several times to review and discuss the quantitative and qualitative comments for their domain. The chair of each subgroup wrote a summary of the group's findings and conclusions. These summaries were then reviewed by the leaders of the research team and the groups' suggestions for change were incorporated into the taxonomy. The revised taxonomy was then presented to the entire team and reviewed again. The principles and rationale used to determine what changes to make in the taxonomy were identified as follows:
1. Make changes that are supported by sound reason and compelling evidence. Do not make changes without a good rational. The taxonomy was constructed through a rigorous process of hierarchical clustering and clinical review. The numerical scores from the outside reviewers were excellent. Changes now should not invalidate all the previous work. In addition, nurses have begun to use the taxonomy and any change creates adjustment for those who are current users.
2. Keep the language simple and easy to understand. Do not make changes that are hard to understand. Terms should be understood by nurses in all specialties. The terms used in the domains and classes, as well as the terms for interventions, need to be clear as they may be used to help the public understand nursing.
3. Have the taxonomy structure represent what we have now, not what we might have in the future. The team's recently created "Interventions Under Consideration List" (a list of ideas for possible new interventions) is where proposed new interventions should be placed. Do not make new classes until there are sufficient interventions developed that require the class.
4. Try to place each intervention in only one class; never cross-reference an intervention to more than two classes. The cross referencing makes the taxonomy very long, presents complications for coding, and violates the taxonomy criterion of classes being mutually exclusive. Some cross referencing needs to be done to facilitate use but the temptation to overdue it should be resisted. No specialty group of nurses should expect that all of the interventions they use will be located in one class or domain; all nurses need to use all domains and classes as appropriate for their patients.
Based upon the results of both the quantitative and qualitative results, with the guidance of the principles, we made changes in the taxonomy. A revised version of the first two levels of the taxonomy is printed in Table 3. Changes occurred mostly in definitions. Three class names were modified and one new class; "Information Management," was created. In addition a few interventions were moved to different classes and some cross referencing was added or omitted. The review demonstrated the validity of the taxonomy; the changes that were made were done to enhance clarity. The "Health System" domain (whose classes were rated lowest on inclusiveness) was improved by the development and placement of several additional interventions. The order of the domains was revised slightly with the "Safety" domain being moved from sixth to fourth place. Reviewers though the safety domain should be a more central part of the taxonomy and that health system domain (now listed sixth) was best at the end.
One domain of the taxonomy with the coded interventions appears in Table 5 and a summary of our rules for coding is in Table 6. The entire validated and coded taxonomy will be in the second edition of the Nursing Interventions Classification (forthcoming in 1996). The codes for the 6 domains are 1-6; the codes for the 27 classes are A-Y plus a and b. The "Health System" domain appears last in the taxonomy with the last two of its three classes containing indirect care interventions coded with small letters (a and b). The use of small letters signals that these classes include interventions that, while important to patient progress and outcomes, may not be those that are billed for directly. (That is, these may be part of overhead rather than direct costs.) Each intervention has a unique number consisting of four spaces (e.g., 6140). If one wants to know the domain and class an intervention is in, one would use the domain and class designations with a hyphen between these and the intervention (e.g., 4U-6140). Activities are coded after the decimal using two digits (e.g., 4U-6140.01).
Each class has been allocated 300 numbers following the recommendations of our consultants because "the less ad hoc, the more uniform" and with classifications, uniformity is desirable. Three hundred numbers for each class allow for future additions as well as room for seven more classes should the taxonomy need to be expanded. The current numbering system which leaves spaces between numbers will help to minimize recoding in the future.
An early step in coding of the NIC taxonomy was to review other coding classifications (e.g., Current Procedural Terminology (CPT-4), International Classification of Diseases (ICD-10 and ICD-9-CM), Diagnostic and Statistical Manual of Mental Disorders (DSM-III-R), etc.). Table 7 compares NIC with six of the major classifications in medicine and with the version of the NANDA nursing diagnosis classification that was done for articulation with the International Classification of Diseases.
Conclusion^
The development of standardized language for nursing interventions and a coded taxonomic structure helps to advance nursing knowledge by facilitating the clinical testing of nursing interventions. It facilitates the study of linkages among nursing diagnoses, interventions, and outcomes; assists in the implementation of the Nursing Minimum Data Set and the building of large nursing data sets; and provides a language that can articulate with the classification systems of other health care providers. It provides an organized approach for curriculum development that should result in a more uniform method for socializing students to the profession. Nurses now have the opportunity to speak with a common voice when describing nursing treatments to clients, colleagues, and policy makers.
There are several indications that NIC is becoming a standard. NIC has been recognized by the American Nurses Association as one of the nursing classifications that should be part of a unified nursing language and in January 1993 it was one of the first two nursing languages to be included in the National Library of Medicine's Metathesaurus for the Unified Medical Language System. Both the Cumulative Index to Nursing Literature (CINAHL) and Silver Platter have added NIC to their nursing indexes. NIC is included in the Joint Commission on Accreditation for Health Care Organization's (JCAHO) scoring guidelines for the 1994 new section, "Management of Information" (JCAHO, 1994) [13] as one nursing classification system that can be used to meet the standard on uniform data. The National League for Nursing has produced a 30-minute video about NIC. There are numerous requests for presentations and information about NIC. Many practitioners are beginning to adopt NIC for use in standards, care plans, nursing information systems; nursing educators are beginning to use NIC; authors of major texts are beginning to incorporate NIC to discuss nursing treatments; and researchers are using NIC to study the effectiveness of nursing care. Interest in NIC has been demonstrated in several other countries, notably, Canada, Denmark, the Netherlands, Iceland, and England. NIC is a central part of the International Council of Nurses' (ICN) International Classification of Nursing Practice. The future of NIC as a mechanism for reimbursement to nurses and as an integral part of federally mandated health care data sets has yet to be determined. The accomplishment of a valid and coded taxonomic structure, however, facilitates achieving these goals.
REFERENCES^
1. American Medical Association. (1993). CPT: Physicians' current procedural terminology (4th ed.). Chicago: Author.
2. American Psychiatric Association. (1992). Diagnostic and statistical manual of mental disorders (3rd ed.). Washington, DC: Author.
3. Barry-Walker, J., Bulechek, G.M., & McCloskey, J.C. (1994). A description of medical-surgical nursing. MEDSURG Journal, 3(4), 261-268. [Context Link]
4. Bulechek, G.M., & McCloskey, J.C. (1992). Defining and validating nursing interventions. Nursing Clinics of North America, 27(2), 289-299. [Medline Link] [Context Link]
5. Bulechek, G.M., McCloskey, J.C., Denehy, J.A., & Titler, M. (1994). Report on the NIC project: Interventions used in practice. American Journal of Nursing, 94(10), 59-66. [Medline Link] [Context Link]
6. Cote, R.A., Rothwell, D.J., Palotax, J.L., & Beckett, R.S. (1993). SNOMED International. Northfield, IL: College of American Pathologists.
7. Fehring, R.J. (1986). Validating diagnostic labels: Standardized methodology. In M.E. Hurley (Ed.), Classification of nursing diagnoses: Proceedings of the 6th conference (183-190). St. Louis: Mosby. [Context Link]
8. Fitzpatrick, J.J., Kerr, M.E., Saba, V.K., Hoskins, L.M., Hurley, M.E., Mills, W.C., Rottkamp, B.C., Warren, J.J., & Carpenito, L.J. (1989). Nursing diagnosis: Translating nursing diagnosis into ICD Code. American Journal of Nursing, 493-95.
9. Health Care Financing Administration. (1992). HCFA common procedure coding system (HCPCS). USDHHS: Washington, DC: U.S. Government Printing Office. [Context Link]
10. Iowa International Project. J. McCloskey & G. Bulechek (Eds.). (1992). Nursing Interventions Classification (NIC). St. Louis: Mosby Year Book.
11. Iowa Intervention Project. (1992). Nursing Interventions Classification (NIC): Taxonomy of nursing interventions. Iowa City, IA: University of Iowa College of Nursing. [Context Link]
12. Iowa Intervention Project. (1993). The NIC taxonomy structure, Image: Journal of Nursing Scholarship, 25, 187-92. [Context Link]
13. Joint Commission on the Accreditation of Health Care Organizations. (1994). Section 2: Management of information, JCAHO Accreditation Manual for Hospitals (p. 10). Chicago, IL: Author. [Context Link]
14. Practice Management Information Corporation. (1992). HCPCS: National level II Medicare codes. HCFA common procedure coding system. Los Angeles: Author.
15. Practice Management Information Corporation. (1989). ICD-9-CM international classification of diseases (9th rev., 3rd ed. clinical modification). Los Angeles: Author.
16. Steelman, V., Bulechek, G.M., & McCloskey, J.C. (1994). Toward a standardized language to describe perioperative nursing. AORN Journal, 60(5), 786-795. [Medline Link] [Context Link]
17. World Health Organization. (1992). International statistical classification of diseases and related health problems (10th rev.). Geneva, Switzerland: Author.