A Compositional Approach to Nursing Terminology

TERMINOLOGY AND CLASSIFICATION

1

A Compositional Approach to Nursing Terminology

Hardiker N and Kirby J

Medical Informatics Group, Department of Computer Science, University of Manchester, ManchesterM13 9PL, UK

The development of standardised vocabularies within nursing has been an important research activity for anumber of years. Current representations generally take the form of taxonomic vocabularies. These are seen asimportant as they provide a structure for retrieving and analysing data from automated systems. However, thereis increasing evidence to show that traditional taxonomic vocabularies are unsuitable for capturing detailedclinical data. This paper describes how GRAIL (GALEN Representation and Integration Language) is beingused within the TELENURSE project to develop a representation of nursing terminology which is sufficientlyexpressive for documenting detailed clinical data while retaining the benefits of traditional taxonomicvocabularies.

IntroductionThe development of standardised vocabularies to represent nursing terminology and todescribe nursing practice has been an important research activity for many years. Thedevelopment and increasing use of computer-based nursing information systems have furtheremphasised the importance of this activity. The result has been the emergence of a number ofrepresentations.

Problems with traditional taxonomic vocabulariesThe majority of the commonly reported standardised nursing vocabularies take the form oftaxonomic vocabularies. Taxonomic vocabularies are terminological systems in whichconcepts are related by hierarchical relations i.e. generic ‘is-a’ relation and partitive ‘part-of’relation, and other associative and pragmatic relations1. Examples within nursing include theNorth American Nursing Diagnosis Association Taxonomy I (NANDA), the NursingInterventions Classification (NIC), the Home Health Care Classification (HHCC) and theOmaha Community System (Omaha).

These representations are seen as important because they provide a structure for retrieving andusing nursing data from automated systems2. Other reasons cited for organising nursingconcepts into taxonomies include: to formalise and expand knowledge about nursing practice,to assist in determining the cost of nursing services, to help to target resources moreeffectively and to make explicit the role played by nurses in health care 3.

Monohierarchical taxonomic vocabularies that are exhaustive and that guarantee disjunctionare seen as useful for statistical evaluation1. Thus it could be argued that taxonomicvocabularies have a useful role to play in activities such as data retrieval and data analysis.However there is increasing evidence to show that taxonomic nursing vocabularies are notable to represent the detailed clinical data within patient records4. As such they are poorlysuited for representing day-to-day nursing care.

One study was carried out to test the feasibility of using the third version of the SystematizedNomenclature of Human and Veterinary Medicine (SNOMED) to represent the terms used bynurses to describe patient problems5. It should be noted that SNOMED III contains all of thenursing diagnoses from NANDA. This study found that NANDA terms alone providedmatches for only 30% of the patient problems described by nurses in the study. It is clear fromthese results that NANDA alone does not provides the coverage necessary for nurses to record


2

patient problems (interestingly the inclusion of SNOMED III terms and combinations ofSNOMED III terms increased the proportion of matches to 69%).

Another study was carried out to compare the ability of terms from NIC and from themedically-oriented Current Procedural Terminology to represent the clinical terms used bynurses and patients to describe nursing interventions6. The results of matching NIC terms toclinical terms used by nurses and patients to describe nursing interventions are given as‘encouraging’. However the examples cited of selected clinical terms and their matching NICinterventions show that comprehensiveness of scope is at the expense of clinical detail. Forexample, the relatively abstract NIC term ‘Hypoglycaemia management’ is given as a matchfor the relatively detailed clinical term ‘Fingersticks for blood sugar’. NIC has been criticisedpreviously for being insufficiently fine-grained for capturing differences in practice4.

One reason concerns the fact that traditional taxonomic vocabularies are constructed byenumerating all of the possible terms that are to be represented; and organising these termswithin a hierarchy. In constructing any enumerative scheme, developers must limit thenumber of concepts to include as the total number of concepts would be unmanageable, bothin terms of development and in terms of practical application. As such, enumerativerepresentations tend to be tuned to a single purpose or to a group of closely-related purpose;re-use for other purposes proves very difficult. Indeed HHCC and Omaha have been criticisedfor lacking the specific vocabulary of acute care and NANDA Taxonomy I has been criticisedfor not covering all fields of specialty practice4.

Solutions to problems concerning expressivenessLinear listsAn alternative approach to the traditional taxonomic vocabulary is the linear list. A linear listis simply a collection of terms relevant to a domain1. One study claims that it may be possibleto develop a list of standard terms that is capable of representing the universe of termsactually used to record data elements in a patient record4. However, there are manyoutstanding issues arising from this study:

• The study was confined to two specialist fields, Orthopaedics and Thoracic/Cardiovascular surgery; there is no indication as to how the list of standard termsemployed within the study might scale up to include other areas of practice.

• 11 auditors were involved in the study, matching statements from patient records tostandard terms in a code book. While the reliability of the auditors, that is the accuracy ofterm matching, was a consideration within the study, the results given omit anydiscussion on the degree of detail captured by the standard terms and on the exactness ofterm matching.

• Within the study term matching was performed manually and it is not clear how thisprocess may be automated, nor how the standard terms might be re-used for otherpurposes.

Until these issues are resolved, the usefulness in practice of such an approach is questionable.

A compositional approachWithin the GALEN project (Generalised Architecture for Languages, Encyclopedias andNomenclatures in Medicine) a new approach to representing clinical terminology has beendeveloped in the form of the GALEN Representation And Integration Language (GRAIL) 7.


3

GRAIL is a terminological language that provides a means of capturing the knowledge thatunderpins clinical terminology in a formal compositional model from which all and onlysensible clinical concepts can be generated8.

By decomposing complex concepts into primitive concepts, other schemes such as SNOMEDattempt to address the problems associated with enumerative representations. However,although SNOMED provides a framework or meta-model for constructing complex clinicalconcepts, it is impaired by the lack of specific rules for determining which combinations areclinically sensible. Thus it cannot prevent the creation of clinically meaningless concepts; noris it able to control combinatorial explosion.

A GRAIL model consists of a taxonomy of concepts and a set of rules or ‘sanctions’ to dictatehow these concepts may sensibly be combined. For example, it might be sensible to combinethe elementary concepts Mobilising and Ability to create a composite concept which defines‘Mobility’:

(Ability whichrefersTo Mobilising).

As they are created, composite concepts are classified automatically in the taxonomy. Forexample, if the concept Mobilising subsumes Walking, then the composite concept whichdefines ImpairedWalkingAbility will be subsumed by ImpairedMobility.

Mobilising

Walking

Ability

Ability which < refersTo Mobilising hasState Impaired>

Ability which < refersTo Walking hasState Impaired>

The result is a multiple hierarchy of clinically sensible concepts which are defined to anarbitrary level of detail.

Practical application of GRAILThe TELENURSE projectThe TELENURSE project is an accompanying measure in the European TelematicsApplication Programme. Its primary aim is to promote consensus among nurses acrossEurope around the use of the International Classification of Nursing Practice (ICNP) which isbeing developed by the International Council of Nurses. Each concept within ICNP isexplicitly defined and classified in terms of the generic relation. The alpha version of ICNP


4

has two dimensions: nursing phenomena and nursing interventions. At the time of writing, theICNP classification of nursing interventions was undergoing change. The remainder of thisdiscussion is therefore restricted to the ICNP classification of nursing phenomena. As theICNP classification of nursing phenomena is monohierarchical, it may be well-suited forstatistical evaluation. However, as an example of a traditional taxonomic vocabulary it is notwell-suited to the task of recording day-to-day nursing care. In contrast, the development ofGRAIL has been driven in part by the data entry requirements of users of clinicalapplications.

The use of GRAIL within TELENURSEThe GALEN approach has been applied successfully within other areas9; nursing presentsnew challenges.There is within nursing a resistance to recording meaningful, concise information concerningthe nursing care of patients10. This is compounded by a general confusion about the nature ofnursing information. In the context of nursing interventions four main origins for thisconfusion have been identified3:

1. The contextual nature of nursing information leading to confusion between intervention(nurse behaviour) and assessment and evaluation (patient behaviour);

2. The use of synonyms e.g. action, activity, treatment, order;3. The lack of conceptualisation of how a number of actions might fit together, resulting in

long verbose care plans;4. Inadequate decision making among nurses in selecting and prioritising interventions.

Within TELENURSE the GALEN approach is being applied in an attempt to overcome thefirst three of these problems; the final problem requires a change in nursing practice.An existing GRAIL medical foundation model has been extended to incorporate nursingconcepts. This has involved the development of GRAIL definitions for ICNP concepts. Forexample, the ICNP concept ‘Nursing Phenomena’ has been explicitly defined in GRAIL as:

Phenomenon whichhasRelevantDomain NursingDomain.

Such definitions restrict the possibility of ambiguity and make explicit any contextualinfluences.

In GRAIL, any number of unique names for individual concepts is permitted, thus facilitatingthe controlled use of synonyms:

(Phenomenon whichhasRelevantDomain NursingDomain)

name NursingPhenomenon.

As specific detail is added, GRAIL concepts are classified automatically. The resultingsubsumption hierarchy provides a ‘bridge’ for different levels of abstraction and represents aconceptualisation of how concepts interrelate.

Transforming hope into working achievementA significant problem with enumerative representations is the fact that any reasoning behindthe decisions made during the construction of the scheme is locked inside terms or concept


5

definitions. For example, a nurse may have a clear understanding of the enumerated term‘Able to walk a short distance’. However a computer can have no such understanding andthus cannot utilise the underlying clinical concepts in managing the scheme. A majormotivation for the modelling activity within TELENURSE has been the need to ensure thatcomputers will be able to manage the ultimate structure and content of ICNP.

As part of TELENURSE two prototype nursing care management systems are beingdeveloped. GALEN technology will make ICNP, in the form of a GRAIL model, available tousers of these systems in order to transform the potential benefits of using standardisedvocabularies into working achievements.

SummaryA number of well-founded standardised nursing vocabularies have been developed overrecent years. The majority of these take the form of taxonomic vocabularies. While suchrepresentations may be appropriate for statistical analysis of relatively abstract data, they areunable to capture the detail of day-to-day nursing care. GRAIL provides a mechanism forrepresenting clinical data at any level of detail. Within the TELENURSE project a model builtin GRAIL will be used to make ICNP available to users of clinical applications withoutcompromising the operational needs of those users. The result will be a representation ofnursing terminology which is sufficiently expressive for documenting highly detailed clinicaldata; and one which retains the benefits of traditional taxonomic vocabularies.

References1. Ingenerf J Taxonomic Vocabularies in Medicine: The Intention of Usage Determines Different Established

Structures. In: Greenes RA et al. eds. Proceedings of the Eighth World Congress on Medical Informatics.,Amsterdam North-Holland: 1995:136-139.

2. Grobe. Nursing Intervention lexicon and taxonomy study: Language and classification methods. Adv NursSci, 1990;13(2):22-33.

3. McCloskey JC, Bulechek GM, Cohen MZ, Craft MJ, Crossley JD, Denehy JA, Glick OJ, Kruckeberg T,Maas M, Prophet CM, Tripp-Reimer T. Classification of Nursing Interventions. J Prof Nurs 1990;6(3):pp151-157.

4. Ozbolt JG, Russo, M, Stultz MP. Validity and Reliability of Standard Terms and Codes for Patient CareData. In: Gardner RM (Ed.), Proceedings of the 19th Annual Symposium on Computer Applications inMedical Care. Hanley & Belfus Inc., Philadelphia, 1995:37-41.

5. Henry SB, Holzemer WL, Reilly CA, Campbell KE. Terms Used by Nurses to Describe Patient Problems:Can SNOMED III Represent Nursing Concepts in the Patient Record? J Am Med Informatics Assoc1994161-74.

6. Henry SB, Holzemer WL, Reilly CA, Miller TJ, Randall C. A Comparison of Nursing InterventionClassification and Current Procedural Terminology Codes for Representing Nursing Interventions in HIVDisease. In: Greenes RA et al. eds. Proceedings of the Eighth World Congress on Medical Informatics.Amsterdam:North-Holland, 1995:131-135.

7. GALEN Implementation and Modelling Teams. GALEN Documentation Volume H: A Guide to GALEN'sConcept Services. Available from: Project Coordinator, Medical Informatics Group, Department ofComputer Science, University of Manchester, M13 9PL, 1994.

8. Rector AL, Bechofer S, Goble CA, Horrocks I, Nowlan WA, Solomon WD (1995). Why GRAIL. Paper inpreparation. Medical Informatics Group, University of Manchester.

9. Kirby J et al. (1996), A System for Entering Detailed Clinical Data. To be presented at: HealthcareComputing, 18-20 March 1996, Harrogate, UK.

10. Howse E, Bailey J (1992). Resistance to documentation - a research issue. Int J Nurs Stud, 29 (4): pp 371-380.


6

Nursing Intervention Intensity and Focus:Indicators of Process for Outcomes Studies

Grobe SJa, Hughes LCb, Robinson Lc, Adler DCc, Nuamah Ic and McCorkle Rc

aSchool of Nursing, University of Texas at Austin, Austin, TX 78701-1499. bSchool of Nursing, Wichita StateUniversity, Wichita, KS, 67260-0041. cSchool of Nursing, University of Pennsylvania, Philadelphia, PA, 19104-6096.

Outcomes research has become increasingly important in the current health care environment and forinformatics research efforts. Recent efforts in automating clinical data for use in outcomes studies has focusedattention on the need to represent the processes of care in the classic structure-process-outcome models of care.This paper reports on use of the Nursing Intervention Lexicon and Taxonomy for classifying interventions tocharacterize two process of care variables: intervention intensity and intervention focus. Study resultsdemonstrate that these variables are descriptive and provide promise for describing processes of nursing carefor describing clinical care.

IntroductionOutcomes research and benchmarking studies are increasingly important in today's healthcarearena wherein cost and quality studies are being conducted to examine efficiency,effectiveness and quality of care. And, although much attention has been focused on outcomesof care, much less attention has been directed toward describing the processes of care. In thiscurrent environment of capitated and managed care it becomes increasingly critical to findways to describe and measure these processes of care. This paper describes a preliminary wayfor describing and measuring care process in a study of the home care of cancer patients.Models used to guide outcomes research are those of Donabedian and Holzemer (Donabedian,1982; Holzemer, 1995). In Donabedian's structure-process-outcome model, structure includesthe resources used in providing care, process includes the activities that comprise care, andoutcomes are the consequences of both of these on health. Holzemer's model extendsDonabedian's; it has a horizontal axis of inputs, processes and outcomes and a vertical axisthat includes the three constituents generally involved in health care encounters: the client, theprovider and the setting.

While much research in nursing and health care has been focused on organizational structure(i.e., Holzemer's inputs) and outcomes, little attention has been focused on describing careprocesses. This study reflects one of the dimensions of Holzemer's model: Process (from thehorizontal axis) as reflected by nurses' care interventions as documented during their homecare of breast and prostate cancer patients. Importantly, because the main study reflected acontrolled clinical trial, this nursing documentation was not encumbered by external Medicarereimbursement considerations. Although the study uses manual methods to test the feasibilityof using classified nursing interventions to describe the processes of care, it represents animportant preliminary informatics effort to demonstrate "proof of concept," prior toinvestigating automated methods for both extracting and classifying similar interventionsfrom narrative nursing recording.

Study purposeThe premise of this paper is that, it is possible to describe care processes, using two patternsof nursing care variables: nursing intervention intensity and intervention focus, asdemonstrated in a controlled clinical trial of home care of cancer patients (McCorkle, PI:NIH, NINR R01 NR03229, 1992-96). Intervention intensity is defined as the frequency ofinterventions; and, intervention focus is defined as the categories of interventions used most


7

frequently during care. These two variables, intensity and focus, once measured reliably andvalidly, can then be used for examining relationships among the care processes and specificcare outcomes.

This paper describes these two patterns of care variables for a subset of McCorkle'spostoperative cancer patients receiving the experimental home nursing care (the standardnursing intervention protocol (SNIP), specifically the breast (N=22) and prostate (N=33)cancer patients. McCorkle's standard intervention protocol consisted of 8 contacts including 3home visits and 5 telephone calls over 4 weeks by a clinical nurse specialist (masters preparednurse), begun within a week after discharge from the hospital. Clinical nurse specialistsdocumented their home care in paper records that served as the source of study data. Theirdocumentation was free of any constraints imposed by reliance on reimbursement. In fact,these nurses were encouraged to document completely all the care they provided for eachpatient and their caregiver(s) using a modified SOAP format.

Study proceduresOnce the cancer patients (breast and prostate) had received the standard nursing interventiveprotocol, interventions were manually extracted and transcribed, by a single member of theresearch team, from nurses' narrative SOAP recording (modified with an I category added toinclude interventions, making it a SOAIP format). Then, three trained individuals,independently classified each of the interventions using a 7 category scheme, i.e., the NursingIntervention Lexicon & Taxonomy (NILT) (Grobe, 1996). The category names and their shortdescriptive concepts include: Care Information Provision (CIP), teaching; Therapeutic CarePsychosocial (TCP), supporting; Care Vigilance (CV), monitoring status; Care NeedDetermination (CND), assessing need for care; Care Environment Management (CEM),obtaining resources for care; Therapeutic Care General (TCG), performing care procedures;and Therapeutic Care Cognitive Understanding and Control (TCCU&C), encouraging self-care.

A source definition, typical of those used by categorizers is provided for CIP along withprototypical intervention examples. Brief definitions for the remaining categories are providedin Figure 1. Care Information Provision (CIP): the deliberative, cognitive, physical or verbalactivities of informing or teaching that assist individuals (may be client, family, significantother(s) or caregivers(s), who are the focus of care to acquire or use care information intendedto maintain or improve the existing state or general condition and maximize the response totherapy. Prototypical examples include: Advise patient to seek information from ET nurse;Explain how to splint an incision; Review medications with patient; Teach patient to performdressing change; and Inform wife of possible side effects of medications.

Figure 1NILT Category Definitions and Examples

CND: Care Need Determination:•assessment of need for care including: past health, baseline health state, role management, healthbeliefs and values. Examples: Determined her perception of the cause of fatigue; Assessed patient’sknowledge of her disease; Explored patient’s expectations for recovery.

CV: Care Vigilance:•assessment of physical status, physiological, mental or emotional status, monitoring of status ordevices. Examples: Assessed pain; Monitored incontinence; Assessed mastectomy site.

CEM: Care Environment Management:


8

•evaluation of the environmental or familial context for care, referring or influencing the use ofresources. Examples: Referred patient to Reach for Recovery; Offered numerous resources fordepression; Suggested use of Wellness Center; Encourage to talk with someone in physician’s office.

TCG: Therapeutic Care-General:•performance of procedures, and physically based activities. Examples: Encouraged physical activity,i.e., short daily walks; Reapplied ace bandage to right chest area; Encouraged kegal exercises;Performed dressing change on incision line.

TCP: Therapeutic Care-Psychosocial:•performance of psychologically based therapies. Examples: Assured patient that feelings were normal;Reassured her that improvement is not unidirectional; Listened actively; Supported patient emotionally;Encouraged open dialogue with husband.

TCCU&C: Therapeutic Care Cognitive Understanding and Control•activities to enhance self care, control and independence. Examples: Supported patient’s need to decidewhen she is ready; Supported patient’s decision making ability; Provided positive feedback about herproblem solving skills; Supported patient in his self care efforts.

CIP: Care Information Provision:•informing about care and care procedures and therapies. Examples as above.

Once all the interventions were categorized, those interventions [for which there was not totalagreement] were then discussed by the 3 categorizers for placement into a single NILTcategory. This agreed-upon final category was the ‘category of record’ used for this paper. Allinterrater agreement scores (Cohen's kappa) were calculated prior to this discussion, whichwas simultaneously used to maintain coder training. Interrater agreements (Cohen's kappas)for the breast and prostate charts were .7492 and .7944 respectively.

Study resultsThe patterns of care variables include intervention intensity and focus. Each is described nextusing the intervention data from the records of 22 breast patients and 33 prostate cancerpatients.

Table 1Intervention intensity by contact* for breast and cancer patients

Contact Breast Prostate TotalN % N % N %

1 257 20.9 242 15.0 499 17.62 134 10.9 247 15.4 381 13.43 138 11.2 199 12.4 337 11.94 158 12.9 201 12.5 359 12.65 122 9.9 160 9.9 282 9.96 103 8.4 197 12.2 300 10.67 116 9.4 144 8.9 260 9.28 137 11.1 133 8.3 270 9.59 44 3.6 63 3.9 107 3.8

20 1.6 23 1.4 43 1.51229 1609 2838

N = 22 Breast and 33 Prostate Patients * Contact = either home visit or phone call

Intervention intensity for each patient contact is illustrated first in Table 1. Each contactrepresents either a home visit or a telephone call. For the breast cancer patients, the first


9

contact is more intensive with one-fifth of all interventions occurring during this first contact.By the fourth contact, both breast and prostate patients had received 55% of all interventionsfor their course of care. It is noteworthy that a few patients in both cancer groups receivedmore than eight contacts, with these additional contacts representing about 5% of allinterventions. Two patients received 10 contacts.

Intervention Intensity by type of contact (either home visit or phone call) illustrates that ingeneral, almost two-thirds of all interventions occurred during home visits (Table 2).Statistically significant differences exist between breast and prostate patients (p=0.046) onintervention intensity by type of contact, although this difference appears marginal.

Table 2Intervention intensity by type of contact* (home visit or phone call) for breast andprostate patients1

HomeInts

visits%

PhoneInts

calls%

Breast(n = 1229)

792 64.4 437 35.6

Prostate(n = 1609)

978 60.8 631 39.2

Total(n = 2838)

1770 62.4 1068 37.6

* p = 0.046 1 N = 22 Breast and 33 Prostate Patients

Intervention focus, i.e., what categories predominated during care, are illustrated in Table 3.Prostate patients received more teaching (CIP) (p<0.01), and slightly more psychologicalsupport (TCP) but this difference is not statistically significant. Breast patients received moreCND (assessing care needs)(p<0.01), CV (monitoring status)(p<0.05), and CEM (managingof care resources)(p<0.01). Breast and prostate patients were about equal with respect to TCG(performing procedures) and TCCU&C (encouraging self-care).

Table 3Intervention focus by breast and prostate patients for entire course of home care

Category Breast Prostate Total (focus of care) (n = 1229)

%(n = 1609)

%(n = 2838)

%CIP 35.5 45.8* 41.3TCP 16.2 19.5 18.0CV 12.6** 9.4 10.8CND 15.3* 9.0 11.7CEM 13.6* 8.9 11.0TCG 4.8 4.8 4.8TCCU&C 2.0 2.6 2.4

N = 22 Breast and 33 Prostate Patients * p < 0.01 ** p < 0.05

The focus of care variable illustrates a similarity between breast and prostate patients for thefirst and last contact with respect to TCCU&C (encouraging self-care) as illustrated in Table4. From the first to last contact, this self-care category increases from 1.5 to 6.9% for breast


10

and from .8 to 7.6% for prostate patients, reflecting a substantial increase in interventionsreflecting self-care encouragement toward the end of care. Teaching and informing (CIP) ispredominant for both breast and prostate patients for the first and last contacts, whileproviding emotional support (TCP) is the next most predominant category for prostatepatients for both the first and last contact, and for breast patients for the last contact. TCPincreases substantially from the first to the last contact for both breast and prostrate patients(8.6 to 20.6% (p<0.01) and 13.7 to 22.7% (p<0.05) respectively.

When combined, the monitoring (CV) and assessing (CND) categories for both types ofpatients are reduced by one half from the first to the last contact (from 37.15% to 19.4%(p<0.001) for breast and, from 25.6% to 13.5% (p<0.01) for prostate patients). For both typesof patients, even though there are increases in CEM (managing care resources and referring)from the first to the last contact (from 11.1 to 13.1%) for breast patients and from 8.0 to10.3% for prostate patients, these increases are not statistically significant.

Table 4Intervention intensity and focus of care for first and last contact for breast andprostate patients

Category(focus of care)

Breast(n = 160)

Prostate(n = 185)

Total(n = 345)

first last first last first & lastCIP 36.3 38.1 48.1 44.3 41.4TCP 8.6 20.6 13.7 22.7 21.7CV 18.0 8.8* 13.4 7.6** 8.1CND 19.1 10.6* 12.2 5.9** 8.1CEM 11.1 13.1 ns 8.0 10.3 ns 11.6TCG 4.9 1.9 3.8 1.6 1.7TCCU&C 1.5 6.9* .8 7.6* 7.2

N = 22 Breast and 33 Prostate Patients * p < 0.001 ** p < 0.01

DiscussionThese preliminary results demonstrate that it is indeed possible to characterize the processesof care using intervention intensity and focus. These preliminary results demonstrate that it isindeed possible to characterize the processes of care using intervention intensity and focus.Process of care variables measurement is important for two reasons: first, to characterize (orexplicitly describe) the nature of nursing care that has been delivered to a patient (i.e.,intervention focus) and second, to quantify (for comparative purposes), the amount of thatcare (intervention intensity). These process variables are then available for use in models suchas the structure- process- outcome models for studying the effects of nursing care.

Results should be interpreted cautiously because of the small number of patients (n=55) andinterventions (n=2838) included. Study results provide preliminary support that the patterns ofnursing care differ for breast and cancer patients with respect to the intensity and focus ofcare.

Intensity is consistently greater for home visits, and is greater for the first four contacts withboth breast and prostate patients, representing possibly the initial and critical phases ofpatients’ adjustments to care at home and their conditions. Fully 55% of all interventionsduring a course of care occurred during the first four contacts with both breast and prostatepatients. Decreasing intensity was observed for the remaining contacts (except for the 8th


11

contact for breast patients and the 6th contact for prostate patients), potentially attributable tointervention focus differences. However, increased substantiation of intervention intensitywith larger numbers of clients and interventions is warranted.

Differences in intervention intensity by type of contact for breast and prostate patients wasdemonstrated to be statistically significant. Although marginal, these findings demonstratethat it is possible to characterize process of care using intervention intensity.

Intervention focus represents a more descriptive way of characterizing processes of home carefor the breast and prostate patients. It is not unexpected that prostate patients received moreteaching (CIP: p>0.01) since the complexity of post surgical incontinence requires thelearning of many new ways of dealing with the operative sequelae. Both breast and prostatepatients received about equal TCP (emotional support). What is less able to be explained iswhy breast patients received more assessing of care needs (CND: p>0.01) and assessing andmonitoring of their status (CV:p>0.05), unless their needs were less obvious than those of themen. The breast patients’ higher CEM (p>0.01) can be attributed to the wider scope of careproviders and care resources available for women during this immediate course of home care.For example, women were referred to a variety of cancer support groups while the menreceived very few referrals.

Comparison of intervention focus from the first contact to the last are consistent with whatwould be expected. It is reasonable that the monitoring and assessing categories (CV andCND) would decrease toward the end of home care for both breast (p<0.001) and prostate(p<0.01) patients. It is also reasonable that TCCU&C would increase toward the end of homecare, since patients need to be moving toward enhanced self care. An unexpected finding isthat TCP increases substantially from the first to the last contact for both breast (p<0.01) andfor prostate (p<0.05) patients, an increase that might be explained by the nature of theinterpersonal aspects of the nursing care. In conclusion, this preliminary study demonstratesthat it is possible to use intervention intensity and focus to characterize the processes of care.Thus it represents a potentially useful way for examining the nature of nursing care acrossencounters and the continuum of care.

References1. Donabedian A. Explorations in quality assessment and monitoring: The criteria and standards of quality.

Ann Arbor, MI: Health Administration Press, 1982.2. Holzemer WL, Reilly CA. Variables, variability, and variations research: Implications for medical

informatics. J Am Med Informatics Assoc 1995; 2:183-90.


12

Evaluating Standardized Coding and Classification Systems for ClinicalPractice: A Critical Review of the Nursing Literature in the United States

Henry SBa and Mead CNa,b

aDepartment of Community Health Systems, School of Nursing, University of California, San Francisco, CA,bCarecCentric Solutions, Inc., Atlanta, GA

Developers of healthcare information systems are challenged by the difficulty of meeting the simultaneous goalsof 1) capturing and electronically representing the broad array of data related to healthcare with sufficientexpressibility to provide adequate documentation of the patient encounter, and 2) utilizing standardized codingand classification systems to facilitate linkages among computer-based systems. The evaluation studies to datehave primarily focused on matching actual clinical data with terms in the recognized classification systems.These studies have provided evidence that the classification systems are relatively domain complete forcategorizing patient problems and nursing interventions. Although several of the published criteria forevaluation relate to structure, it is noteworthy that this has not yet been a major focus of study in nursing. Thereare several areas of critical need that must be addressed. First, additional work is needed to develop and refinea standardized set of atomic-level terms relevant to nursing, including those for assessments, problems, andactivities. Second, knowledge representations must be developed to support the building of complex conceptsfrom atomic-level data.

BackgroundDevelopers of healthcare information systems are challenged by the difficulty of meeting thesimultaneous goals of 1) capturing and electronically representing the broad array of datarelated to healthcare with sufficient expressibility to provide adequate documentation of thepatient encounter, and 2) utilizing standardized coding and classification systems to facilitatelinkages to knowledge-based resources such as bibliographic databases, clinical practiceguidelines, therapeutic protocols, and decision support systems, as well as for abstraction toclinical data repositories. Significant pioneering research has focused on the development ofcoding and classification systems for nursing (e.g. Nursing Interventions Classification,1, 2

International Classification of Nursing Practice);3 the creation of architectures, e.g., theUnified Medical Language System (UMLS),4 to link standardized coding and classificationsystems; and the testing of standardized coding and classification systems with clinical data.5-9

The purposes of this paper are to review criteria for the evaluation of standardized coding andclassification systems, to critically examine the published evaluation studies related tostandardized coding and classifications applicable to nursing, and to suggest future directionsfor research and development.

Evaluation criteriaWhile a "gold standard" has not been identified, a number of authors have proposedevaluation criteria for a standardized coding and classification system designed to supportclinical practice. The authors and the resulting criteria represent a variety of perspectives.Cimino identified nine criteria for a multi-purpose controlled vocabulary for clinicalinformation systems.10 These criteria were aimed at increasing the sensitivity and specificityof information retrieval queries. Clark and Lang described criteria from the perspective of thedevelopment of the International Classification of Nursing Practice (ICNP).3 McCloskey andBulechek generated criteria specifically for the evaluation of the taxonomic structure of theNursing Intervention Classification (NIC).2 Several authors have focused on criteria related tothe clinical expressiveness of classification.6, 8 For recognition by the American NursesAssociation Committee on Databases to Support Nursing Practice, a system must meet


13

criteria for clinical usefulness, reliability and validity, and processes for revision andextension of the classification system.11

The evaluation criteria reported in the literature fall into five broad areas: 1) domaincompleteness, 2) conceptual clarity and coherence, and 3) data structures and relationshipsamong terms, 4) clinical concept capture, and 5) utility. The criteria are defined in thefollowing section.

Domain completeness. The classification system must include all the terms necessary todescribe the domain.2, 10 In addition, from the perspective of the ICNP, domain completenessmeans that the classification system is broad enough to serve the multiple purposes requiredby different countries. McCloskey and Bulechek apply the notion of completeness at the classlevel in the NIC taxonomy, that is, the intervention class will include all the interventionsbelonging to that class.2

Conceptual clarity and coherence. The classification system should be consistent with aclearly defined conceptual framework, but not dependent upon a particular theory or model.2, 3

Clark and Lang propose that the conceptual framework should be reflective of the commonvalue system of nursing across the world as expressed in the International Council (ICN)Code for Nurses.3 Other criteria related to conceptual clarity and coherence include: 1) clear,understandable definitions;11 2) only one way to express each concept (non-redundancy);10 3)terms should refer to only one concept (unambiguous);10 and 4) all terms within a category aremembers of the same class (homogeneity).2

Data structures and relationships among terms. The relationships among terms should beexplicit.10 For instance, in the Nursing Interventions Classification,(1) Bowel IncontinenceCare IS-A Elimination Management intervention. IS-A is a statement of explicit relationship;other types of explicit relationships among terms include EQUIVALENT-TO, PART-OF, andASSOCIATED-WITH. Another useful perspective on structure is that of Ingenerf who hasexplicated four types of taxonomic vocabularies or standardized coding and classificationsystems for health care based on the underlying structure and related knowledgerepresentation formalism.12 Thesauri are defined as lexical vocabularies containing definitionsand cross references (e.g. UMLS Metathesaurus). Classification systems are vocabularies thatcan be represented as hierarchies or decision trees, and that have as a main emphasis thedisjunctive and exhaustive classification of terms. Nomenclatures are combinatorialtaxonomic vocabularies containing more complex polyhierarchies or axes. Terms within anomenclature may be combined into complex concepts using semantic grammars; however,explicit rules for canonical (disambiguated) representation of terms is lacking. Formalterminologies, such as the GRAIL representation language developed in conjunction with theGALEN project, are systems that are based on concepts, rather than on terms and that includeexplicit rules for sensible composition of primitive concepts into complex concepts.13 Theconcepts are represented using knowledge formalisms such as description logic or conceptualgraphs.

Clinical concept capture. Classification systems for clinical practice should be clinicallyexpressive, that is, include the types of natural language terms used to describe patientproblems and health care interventions in the medical record.6, 8, 14 To do this, theclassification system should include modifiers such as those related to time and severity.


14

Utility. Three criteria related to utility of a classification system have been described in thecontext of the ICNP.3 First, the system is "...simple enough to be seen by the ordinarypractitioner of nursing as a meaningful description of practice and a useful means ofstructuring practice" (p. 111). Second, the classification system is complementary with thefamily of disease and health-related classification systems. Third, the classification system isbased on a central core which can be updated through a continual process of development andrefinement. Others have also noted the significance of the last criteria.11 Ozbolt recentlyemphasized the importance of having a standardized set of terms that can capture the variedand evolving clinical practice in addition to formal vocabularies as exemplified in nursing byNIC.14

Evaluation strategies in nursing studies in the USA variety of strategies have been utilized to evaluate the standardized coding andclassification systems for use by nursing in the US. Excluded from the studies in this revieware those done by system developers for the purposes of creating, validating, and refining thesystems and studies aimed a validating a single entity within a system, e.g., validating thedefining characteristics for a particular nursing diagnosis. The studies are listed individuallyin Table 1. n the following section the studies are discussed from a chronological perspectivewhich relates to the type of evaluation strategies utilized. In the early 1990’s, Griffith andRobinson conducted two provider surveys focused on the degree to which Physician's CurrentProcedural Terminology (CPT) coded services were provided by nurses in a variety of nursingspecialties.15, 16 These studies provided evidence that nurses do perform a limited number ofinterventions that can be represented using the CPT codes, however, the determination ofwhether or not the CPT codes can represent the scope of nursing was not an intent of thestudy. While Griffith and Robinson identified the potential overlapping functions ofphysicians and nurses in some areas (as identified by CPT-coded procedures), Zielstorff etal.’s study highlighted the differences among systems in the UMLS and the nursingclassification systems that were not at the time included in the UMLS.22 Subsequently, thenursing classification systems that have been recognized by the ANA Steering Committee onDatabases to Support Clinical Practice have been added to the UMLS.

The majority of recent evaluation studies have tested existing classification systems withclinical data to examine the extent to which the systems capture clinical concepts and aredomain complete.5, 17, 18, 20, 21 The systems examined were Systematized Nomenclature ofMedicine (SNOMED), North American Nursing Diagnosis Association (NANDA) Taxonomy1, NIC, CPT, and Home Health Care Classification (HHCC). The earliest study used a semi-automated lexical matching approach to determine if exact matches could be found betweenterms used by nurses to describe patient problems in the patient record and standardized termsin SNOMED which includes NANDA terms in its functional axis. The study found that 69%of the terms could be matched. Of particular note was the fact that terms in SNOMED otherthan NANDA accounted for 35% of the matches; that is, nurses frequently used symptoms,signs, and medical diagnoses to describe patient problems in their documentation.

Recognizing that the standardized systems developed for nursing to date were aimed atclassification or aggregation of atomic-level data into categories, four studies focused on thedomain completeness of existing nursing classification systems. Henry et al. compared theability of NIC and CPT codes for categorizing nursing activity terms from three acute carehospitals and reported the superiority of NIC to CPT in representing the domain of nursingactivities.17 Holzemer and associates used a related data set to examine the utility of theHHCC for categorizing patient problems and nursing interventions in the hosptial


15

environment.18 The findings were that the HHCC was useful beyond the home health caresetting for which it was designed. Ozbolt noted that the care components of the HHCC were auseful organizing framework, but that a standardized set of more atomic-level terms wasneeded.20 Parlocha tested the utility of the HHCC to abstract chart data related to psychiatrichome care with the intent of developing a critical path for Major Depressive Disorder.21 Whilethe HHCC problem scheme worked well for this data set, subcategories for psychiatricnursing interventions were added to adequately capture that area of nursing activities.

Congruent with the shift of research focus in the larger health care vocabulary arena to anexamination of data structures and representations in addition to content coverage, Henry andMead recently critically analyzed three nursing intervention schema related to two sets ofcriteria described earlier in this paper; the typology of taxonomic vocabularies published byIngenerf12 and the Cimino10 criteria for a multi-purpose controlled vocabulary.23 Theiranalysis demonstrated that the recognized systems in the US have classification as theirprimary purpose and that there are no nomenclatures or formal terminologies for nursing thatmeet the definitions proposed by Ingenerf.

The evaluation studies to date have primarily focused on matching actual clinical data withterms in the recognized classification systems. These studies have validated that theclassification systems are relatively domain complete for categorizing patient problems andnursing interventions. Only one study had a nomenclature, rather than a classification system,as its focus and had exact lexical matches rather than categorization as an aim.5 The studyfinding that SNOMED terms other than NANDA diagnoses were exact matches for termsused by nurses to describe patient problems in the patient record, suggests that nomenclatures,as well as classification systems, have a role in representing terms for computer-basedsystems. Studies are needed that focus upon the utility of the existing standardized coding andclassification systems for purposes other than classification, e.g., how useful are the systemsfor representing nursing terms in a multi-purpose clinical information system and what are theimplementation barriers related to the existing standardized coding and classificationsystems? Additionally, although several of the criteria for evaluation relate to structure, it isnoteworthy that this has not yet been a major focus of study in nursing as compared tomedicine.8, 24

Directions for future development and studyIn addition to the ongoing significant work related to the recognized nursing classificationsystems in the US and elsewhere, several research teams are working on the gaps identified inthis review of evaluation studies. For example, Ozbolt and associates have developed andcontinue to refine a more atomic-level set of standardized problem and activity terms for theacute care environment.14, 19, 20 Grobe and associates are utilizing complex natural languageprocessing techniques to examine both content and structure of nursing documentation as anextension of the work on the Nursing Intervention Lexicon and Taxonomy (NILT).25, 26 In aneffort to compare the relative domain completeness of five existing classification systems(NANDA, NIC, HHCC, Omaha, and SNOMED) across three care settings (acute care, skillednursing facility, and home care), Henry and associates are testing each system with the samedata sets. Another aspect of the same research project is aimed at comparing the frequenciesof matches along a 9-point scale6 to essentially provide a fingerprint of each system related toits proportion of abstract as compared to atomic-level terms.

The large scale efforts of the National Library of Medicine related to identifying the meta-setof terms needed for multiple uses in health care and supporting the linkage among terms


16

through the structure of the UMLS has the potential to significantly accelerate thedevelopments in coding and classification systems.27 However, there are several areas ofcritical need that must also be addressed. First, additional work is needed to identify andrefine a standardized set of atomic-level terms relevant to nursing, including those forassessments, problems, and activities. The ongoing work related to the ICNP andTELENURSE has provided an excellent foundation to be built upon. Second, knowledgeformalisms must be developed to build more complex concepts from primitive concepts(atomic-level data). Several investigators have reported the applicability of conceptual graphsas a type of knowledge representation for medicine,8, 28 but little is known about similaritiesor differences in structure between nursing and medical knowledge or whether the samestrategies for representing knowledge are appropriate or feasible across disciplines. Thesignificant work in Europe within the GALEN project on this topic29 must be criticallyexamined to determine the scope of its applicability. During the last decade excellent progresshas been made in the development and validation of classification systems for nursing. Therapidly evolving nature of computer-based system implementation in health care hashighlighted the need for nomenclatures and formal terminologies in addition to classificationsystems to support nursing practice with a variety of systems including those for decisionsupport. The profession of nursing must address these areas of critical need while continuingthe refinement of the existing classifications.

References

1. Iowa Intervention Project. The NIC taxonomy structure. Image 1993;25:187-192.2. McCloskey JC, Bulechek GM. Nursing Interventions Classification. (2nd ed.) St. Louis: C. V. Mosby,

1996.3. Clark J, Lang NM. Nursing's next advance: An international classification for nursing practice. Int Nurs

Rev 1992;39:109-112.4. Humphreys B, Lindberg DAB. Building the unified medical language system. In: Kingsland L, ed.

Symposium on Computer Applications in Medical Care: IEEE Computer Society Press, 1989:475-480.5. Henry SB, Holzemer WL, Reilly CA, Campbell KE. Terms used by nurses to describe patient problems:

Can SNOMED III represent nursing concepts in the patient record? J Am Med Informatics Assoc1994;1:61-74.

6. Chute CG, Atken GE, Ihrke DM. An empirical evaluation of concept capture by clinical classifications.. In:Lun KC, Degoulet P, Piemme TE, Riehoff O, eds. MedInfo92. Geneva, Switzerland: North-Holland, 1992

7. Chute CG, Cohn SP, E. CK, Oliver DE, Campbell JR. The content coverage of clinical classifications. J AmMed Informatics Assoc 1996;3(3):224-233.

8. Campbell KE, Musen M. Representation of clinical data using SNOMED III and conceptual graphs. In:Frisse ME, ed. Symposium on Computer Applications in Medical Care. Baltimore, MD: McGraw-Hill, Inc.,1992.

9. Campbell JR, Kallhenberg GA, Sherrick RC. The clinical utility of META: An analysis for hypertension..In: Frisse M, ed. Proceedings of the 16th Annual Symposium of Computer Applications in Medical Care.New York: McGraw-Hill, 1993

10. Cimino JJ, Hripsak G, Johnson SB, Clayton PD. Designing an introspective, multi-purpose, controlledmedical vocabulary. In: Kingsland III LC, ed. Symposium on Computer Applications in Medical Care.Washington, D.C.: IEEE Computer Society Press, 1989.

11. McCormick K, Lang N, Zielstorff R, Milholland DK, Saba V, Jacox A. Toward standard classificationschemes for nursing language: Recommendations of the American Nurses Association Steering Committeeon Databases to Support Nursing Practice. J Am Med Informatics Assoc 1994;1:421-427.

12. Ingenerf J. Taxonomic vocabularies in medicine: the intention of usage determines different establishedstructures. In: Greenes RA, Peterson HE, Protti DJ, eds. MedInfo95. Vancouver, British Columbia:HealthCare Computing & Communications, Canada, Inc., 1995.

13. Rector AL, Nowlon WA, Kay S, Horan B, Wilson A. Foundations of an electronic medical record.Methods Information Med 1991;30:179-186.

14. Ozbolt JG. From minimum data to maximum impact: using clinical data to strengthen patient care. AdvPrac Nurs Q 1996;1(4):62-69.


17

15. Griffith HM, Robinson KR. Survey of the degree to which critical care nurses are performing CurrentProcedural Terminology-coded services. Am J Crit Care 1992;1:91-98.

16. Griffith HM, Robinson KR. Current Procedural Terminology (CPT) coded services provided by nursespecialists. Image 1993;25:178-186.

17. Henry SB, Holzemer WL, Randell C, Hsieh S-F, Miller TJ. Comparison of Nursing InterventionsClassification and Current Procedural Terminology codes for categorizing nursing activities. Image, Inpress.

18. Holzemer WL, Henry SB, Dawson C, Sousa K, Bain C, Hsieh S-F. An evaluation of the utility of the HomeHealth Care Classification for categorizing patient problems and nursing interventions from the hospitalsetting. NI97. Stockholm, Sweden, 1997.

19. Ozbolt J, Graves J. Clinical nursing informatics. Developing tools for knowledge workers. Nurs Clin N Am1993;28(2):407-425.

20. Ozbolt J, Fruchtnicht JN, Hayden JR. Toward data standards for clinical nursing information. J Am MedInformatics Assoc 1994;1:175-185.

21. Parlocha PK. Examination of a Critical Path for Psychiatric Home Care Patients with a Diagnosis of MajorDepressive Disorder [PhD]. San Francisco, CA: University of California, San Francisco, 1995.

22. Zielstorff RD, Cimino C, Barnett GO, Hassan L, Blewett DR. Representation of nursing terminology in theUMLS Metathesaurus: a pilot study. In: Frisse M, ed. Proceedings of the 16th Annual Symposium onComputer Applications in Medical Care. New York: McGraw-Hill, 1993:392-396.

23. Henry SB, Mead CN. Standardized nursing classification systems: necessary, but not sufficient forrepresenting what nurses do. In: Cimino J, ed. Fall Symposium of the American Medical InformaticsAssociation. Washington, DC: Hanley & Belfus, 1996.

24. Evans D, Chute C, Cimino JJ, et al. CANON: Toward a medical concept representation language.Proceedings of the American Medical Informatics Association Spring Congress. Bethesda, MD: AmericanMedical Informatics Association, 1993.

25. Grobe SJ. Nursing intervention lexicon and taxonomy: Methodological aspects. In: Hovenga EJS, HannahKJ, McCormick KA, Ronald JS, eds. Nursing Informatics 91: Proceedings of the Fourth InternationalConference on Nursing Use of Computers and Information Science. Berlin: Springer-Verlag, 1991

26. Grobe SJ. Personal communication, June 1996.27. Humphreys B. Vocabularies for computer-based patient records: identifying candidates for large scale

testing. Invitational Meeting at the National Library of Medicine. December, 1994:28. Moorman PW, van Ginneken AM, van der Lei J, van Bemmel JH. A model for structured data entry based

on explicit descriptional knowledge. Methods Information Med 1994;33:454-463.29. Hardiker NR, Kirby J. Overcoming terminological barriers in nursing. In Brendies J et al, eds. Proceedings

of the 13th International Congress of the European Federation for Medical Informatics, Copenhagen,1996:227-231.

AcknowledgmentsThe preparation of this paper was supported by NIH-NR03874, Representing Nursing Concepts for Computer-Based Systems, S B Henry, Principal Investigator. The first author thanks the participants in the 1994 EuropeanSummer School of Nursing Informatics for their thoughtful discussions regarding evaluation criteria for nursingcoding and classification systems during the Summer School.


18

An Evaluation of the Utility of the Home Health Care Classification forCategorizing Patient Problems and Nursing Interventions from

the Hospital Setting

Holzemer WL, Henry SB, Dawson C, Sousa K, Bain C and Hsieh SF

Department of Community Health Systems, School of Nursing, University of California, San Francisco, CA94143-0608

The purpose of this study is to evaluate the utility of the Home Health Care Classification for categorizingpatient problems and nursing interventions from the hospital setting. The data set comprised 5,844 problemterms and 20,055 interventions terms. All terms could be categorized using the Nursing Components and MajorCategories for Nursing Diagnoses and Interventions. A total of 1,767 (30.2%) patient problem terms could beplaced into Major Nursing Diagnosis categories, but not subcategories even though there were subcategoriesrelated to the major category. All intervention terms whether they were classified at the Intervention Categoryor Subcategory could be coded according to Type of Nursing Action. These findings demonstrate that the HomeHealth Care Classification, at the level of Nursing Components and Major Categories, was domain complete forthe data set. The fact that not all terms could be classified according to the existing subcategories suggestssome areas for future development, but is also a reflection of the level of detail expressed in the data set itself.The results suggest that the Home Care Classification will be adequate and appropriate for categorizingproblems and interventions across settings for the next phases of the research project.

IntroductionStandardized coding and classification systems are important building blocks for all types ofcomputer-based systems.1-5 Standardized coding and classification systems vary along manydimensions including the domain covered, the degree of abstract versus atomic-level data, andthe structure, thus, the selection of a standardized coding and classification system mustmatch its intended purpose.6-8 Ingenerf has defined four types of taxonomic vocabularies forhealth care based on the underlying structure and related knowledge representationformalism.9 Thesauri are defined as lexical vocabularies containing definitions and cross-references. Classification systems such as the Omaha System,10 the Nursing InterventionsClassification,11-13 the International Classification of Nursing Practice,1 and the Home HealthCare Classification,14-15 have as a main emphasis, the disjunctive and exhaustiveclassification of terms. More structurally complex are nomenclatures (e.g., SNOMED)16 andformal terminologies (e.g., GRAIL representation language)17 which are necessary torepresent primitive concepts using knowledge formalisms such as description logic orconceptual graphs.

A classification system is best suited to meet the need of this particular research project tocollapse terms used for patient problems and nursing interventions into a manageable numberof categories in order to examine linkages among problems, interventions, and patientoutcomes. Three nursing classification systems that have undergone extensive developmentand testing were considered: the Omaha System,10 the Nursing Interventions Classification,11-

13 and the Home Health Care Classification.14-15 Henry et al. previously reported on theutility of the Nursing Interventions Classification for categorizing nursing intervention termsfrom the hospital setting.18 The evaluation of the Home Health Care Classification (HHCC)was undertaken for several reasons. First, the parent research project (Quality of NursingCare of People with AIDS, NR02215) was expanded to include data collection in non-hospitalsettings including home care and skilled nursing facilities, and the research team desired touse a single system across settings if possible. The report by Ozbolt and associates on theutility of HHCC for the hospital setting suggested that this might be feasible.7 Second, theability to code both problems and interventions using a single system and to aggregate to


19

common Nursing Components will potentially provide useful comparisons for subsequentdata analyses. Third, it was thought that the discrimination among Types of Nursing Actionswould be useful when linking interventions with problems and outcomes (including thoserelated to resource utilization).

Therefore, this study was undertaken to evaluate the utility of the HHCC for categorizingproblems and nursing interventions in the hospital setting. Three questions were of particularinterest in the evaluation: 1) Is the HHCC domain complete for the hospital setting? 2) Towhat level of the HHCC can the problem data be classified (Nursing Component, MajorCategory, Subcategory)? and 3) To what level of the HHCC can the intervention data beclassified (Nursing Component, Major Category, Subcategory, Type of Nursing Action)?

MethodsThe data for this analysis represents more than 600 patient encounters for 201 patients livingwith AIDS who were hospitalized for Pneumocystis carinii pneumonia. The data werecollected as part of a larger study aimed at examining the linkages among patient problems,nursing interventions, and patient outcomes (NR02215). The data were collected from threehospital settings which had three different types of care planning systems and three types ofnurses’ notes. In the first institution, the care plans were computer-based and the nurses’notes were written in narrative style. In the second institution, the care plans werehandwritten and the nurses’ notes were written on a flowsheet using charting by exception. Astandardized printed care plan with a flowsheet and once-daily narrative note were used in thethird institution. Data were collected near hospital admission, approximately midpoint inhospitalization, and near discharge. Patients who had shortened lengths of stay had fewerthan three hospital data collection points. Patient interview data were also collected at threeand six months after hospitalization for those patients receiving follow-up care.

There are two unit of analysis for this study, terms used to describe patient problems andterms related to nursing interventions. These data were collected from multiple data sources:1) patient interview, 2) nurse interview, 3) chart audit of nurses notes, flowsheet, and careplan, and 4) intershift report. The resultant sample size for this analysis is 5,844 patientproblem terms and 20,055 nursing intervention terms.

Data from the multiple data sources was entered verbatim into a relational database foranalysis. The patient problem terms and nursing intervention terms were placed intocategories of the HHCC by Master’s prepared nurse research assistants after they were trainedin the use of the system. Discrepancies in classification among raters was resolved throughconsensus of the research team.

ResultsAll terms in the data set could be classified into the HHCC nursing components (Table 1).The frequencies of the problems classified by nursing components ranged from less than 1%(Fluid Volume and Tissue Perfusion) to 16% (Respiratory). The nursing components leastfrequently used to classify interventions were Coping, Metabolic, Role Relationship, andTissue Perfusion. Nursing components used to classify at least 10% of the terms forinterventions were Fluid Volume, Medication, Physical Regulation, and Respiratory.

Table 1Frequencies of patient problems and nursing interventions in the hospital data set categorizedby HHCC nursing components


20

____________________________________________________________________

Nursing Component Problems Interventionsn (%) n (%)

Activity 548 (9.4) 1282 (6.4)Bowel Elimination 525 (9.0) 502 (2.5)Cardiac 102 (1.7) 213 (1.1)Cognitive 347 (5.9) 358 (1.8)Coping 284 (4.9) 176 (<1)Fluid Volume 46 (<1) 2073 (10.3)Health Behaviour 60 (1.0) 1686 (8.4)Medication 111 (1.9) 2153 (10.7)Metabolic 313 (5.4) 52 (<1)Nutritional 412 (7.0) 641 (3.2)Physical Regulation 689 (11.8) 3136 (15.6)Respiratory 933 (16.0) 2886 (14.4)Role Relationship 134 (2.3) 98 (<1)Safety 62 (1.0) 862 (4.3)Self-Care 116 (2.0) 845 (4.2)Self-Concept 428 (7.3) 788 (3.9)Sensory 522 (8.9) 536 (2.7)Skin Integrity 145 (2.5) 1464 (7.3)Tissue Perfusion 9 (<1) 33 (<1)Urinary Elimination 58 (1.0) 271 (1.4)Total 5844 (100) 20055 (100)___________________________________________________________________________

Problem data were classified into major nursing diagnosis categories and into subcategorieswhen they existed and fit the data. As shown in Table 2, the six nursing diagnosissubcategories for Activity Alteration were aggregated to obtain the frequency of the majornursing diagnoses, Activity Alteration, in addition to the frequencies of the subcategories.Further aggregation to the Nursing Component level would also include combining thefrequency of the other major diagnosis, musculoskeletal Alteration (n= 1), with that ofActivity Alteration (n = 547) resulting in a total of 548 for the Nursing Component of Activity(Table 1). In other instances, e.g., Comfort Alteration, not all of the terms could be classifiedinto a subcategory, so the total frequency for Comfort Alteration included the frequencies forthe subcategories of Acute Pain, Chronic Pain, and Unspecified Pain, as well as an additional226 not otherwise specified Comfort Alterations. A total of 1,767 (30.2%) patient problemterms could be placed into Major Nursing Diagnosis categories, but not subcategories eventhough there were subcategories related to the major category.

All the hospital intervention terms could be classified at least to the level of InterventionCategory. The frequencies of terms classified at the Category versus Subcategory level of thetaxonomy varied greatly by intervention. In the instance of Activity Care, 65.9% of the termswere classified at the Intervention Category as opposed to the Subcategory level. In contrast,for Nutrition Care, 87.1% of the terms could be subcategorized. As shown in Table 3, allterms whether they were classified at the Intervention Category or Subcategory could becoded according to Type of Nursing Action. The majority of the terms were categorized asAssess (53.3%).


21

Table 2Example of frequencies of nursing diagnosis termsclassified by subcategory and major category_________________________________________

Diagnosis subcategories

A01.1 Activityintolerance

25

A01.2 Activityintolerance risk

93

A01.3Diversional

activity deficit

0

A01.4 Fatigue 305A01.5 Physical

mobilityimpairment

72

A01.6 Sleeppattern

disturbance

52

___

Diagnosis major category

A01 Activityalteration

547

__________________________________________

Table 3Frequencies of hospital intervention terms categorizedby type of nursing action___________________________________________

Type ofnursing action

n (%)

Assess 10683 (53.3)Care 7316 (36.4)Teach 1432 ( 7.1)Manage 624 ( 3.1)___________________________________________

DiscussionThe findings of these analyses demonstrate that the HHCC, at the level of NursingComponents and Major Categories, was domain complete for the data set. Not all problemand intervention terms could be subcategorized even when Subcategories existed for MajorDiagnosis and Intervention Categories. This finding suggests some areas for subcategory


22

development, but also is a reflection of the level of detail in the data set so it is not inherentlya weakness of the HHCC. The ability to categorize interventions according to Type ofNursing Action had been identified as a potentially useful attribute by the research team forfuture comparisons among care settings. The study findings indicate that the interventionterms in the data set could be classified into the four categories.

No single existing standardized coding and classification system can meet all needs. Theselection of a system must match the purpose for which it is to be used. The results presentedhere demonstrate the utility of the HHCC to classify large numbers of atomic-level data into alimited number of abstract categories for subsequent analyses thus it is well-suited to meet theneeds of the current research project.

Future studies are needed not only to evaluate the manner in which these abstractions areuseful, but also to examine the impact of the data abstractions in terms of loss of potentiallyuseful information. What level of granularity of the data needs to be preserved in order tolink problems, interventions, and outcomes? Is it possible that abstractions will result in theinability to detect potential process of care differences that could affect outcomes? What isthe role of nomenclatures and formal terminologies as complements to the current existingnursing classification systems? These types of questions can best be answered by a broad-based approach to the development, validation, and implementation of standardized codingand classification systems for nursing.

References1. Clark J, Lang N. Nursing's next advance: an international classification of nursing practice. Int Nurs Rev

1992;39:109-112.2. Dick RS, Steen EB. The computer-based patient record: an essential technology for health care.

Washington, DC: National Academy Press, 1991.3. Simpson RL, Waite R. NCNIP's system of the future: a call for accountability, revenue control, and

national data sets. Nurs Adm Q 1989;14:72-77.4. Lange L, Jacox A. Using large data bases in nursing and health policy research. J Prof Nurs 1993;9:204-

211.5. Henry SB, Holzemer WL, Reilly CA, Campbell KE. Terms used by nurses to describe patient problems:

can SNOMED III represent nursing concepts in the patient record? J Am Med Informatics Assoc1994;1(1):61-74.

6. McCormick KA, Lang N, Zielstorff R, Milholland DK, Saba V, Jacox A. Toward standard classificationschemes for nursing language: recommendations of the American Nurses Association SteeringCommittee on Databases to Support Clinical Nursing Practice. J Am Med Informatics Assoc 1994;1:421-427.

7. Ozbolt J, Fruchnicht JN, Hayden JR. Toward data standards for clinical nursing information. J Am MedInformatics Assoc 1994;1:175-185.

8. Zielstorff RD, Hudgings CI, Grobe SJ et al. Next-generation nursing information systems: essentialcharacteristics for professional practice. Washington, DC: American Nurses Publishing, 1993:

9. Ingenerf J. Taxonomic vocabularies in medicine: the intention of usage determines different establishedstructures. In: Greenes R, Peterson H, Protti D, eds. MedInfo95. Vancouver, British Columbia: Health CareComputing & Communications, Canada, Inc., 1995.

10. Martin KS, Scheet NJ. The Omaha System: applications for community health nursing. Philadelphia: W BSaunders, 1992.

11. McCloskey JC, Bulechek GM. Nursing Interventions Classification. St. Louis: C V Mosby, 1992.12. Iowa Intervention Project. The NIC taxonomy structure. Image 1993;25:187-192.13. Iowa Intervention Project. Validation and coding of the NIC taxonomy structure. Image 1995;27:43-

49.14. Saba VK. The classification of home health care nursing: diagnoses and interventions. Caring

1992;11(3):50-56.15. Saba VK. Home Health Care Classification. Caring 1992;11(4):58-60.16. Côte´ RA, Rothwell DJ, Palotay JL, Beckett RS. SNOMED International. Northfield, IL: College of

Pathologists, 1993.


23

17. Rector AK, Nowlon WA, Kay S, Horan B, Wilson A. Foundations of an electronic medical record.Methods Inf Med 1991;30:179-186.

18. Henry SB, Holzemer WL, Randell C, Hsieh S-F, Miller TJ. Comparison of Nursing InterventionsClassification and Current Procedural Terminology codes for categorizing nursing activities. Image inpress.

AcknowledgmentsThis study was supported by NIH-NR02215, Quality of Nursing Care of People with AIDS (W.L. Holzemer,Principal Investigator) and by NIH-NR03874, Representing Nursing Concepts for Computer-based Systems(S.B. Henry, Principal Investigator).


24

Strategies and Tools for Creating a Common NursingTerminology within a Large Health Maintenance Organization

Zingo CA

Clinical Systems Development, Kaiser Permanente, Southern California Region

A common nursing terminology is essential for outcomes research, data comparability and clinical documentation in anelectronic health record. Kaiser Permanente has recognized the need to develop a common nursing and medical terminologyacross the program. The Interregional Nursing Nomenclature Committee has developed a model for developing a commonnursing terminology integrated with other healthcare terminologies.

BackgroundA common nursing nomenclature has been identified as necessary both within the UnitedStates as well as on an international level.1,2This common nursing terminology is necessaryfor the nursing profession to achieve multiple objectives. These objectives include:- to improve communication within nursing as well as between nursing and others,- to describe nursing care across the continuum of care,- to ensure comparability of nursing data,- to assist with the allocation of resources to patients according to patient needs,- to stimulate nursing research through data available in nursing and healthcare information

systems,- to provide data regarding nursing practice to influence health policy decisions, and- to populate electronic clinical documentation systems (electronic health records) to improve

both communication as well as data retrieval for both outcomes studies as well as decisionsupport for the practising nurse.3

Formation of Interregional Nursing Nomenclature CommitteeIn May, 1995, a teleconference was held within the Kaiser Permanente Medical Care Programto provide an overview of national activities in the area of nomenclature development. KaiserPermanente is the largest Health Maintenance Organization within the United States with oversix million members nation-wide. After the conference, nursing representatives from three ofthe Kaiser regions, Northern California, Southern California, and Northwest, met to discussalternatives on how to achieve a standard nursing terminology within the Kaiser Permanenteorganization.

These nursing representatives drafted a proposal for an interregional project for thestandardization of nomenclature in the Divisions of Nursing. This proposal was presented tothe Interregional Nursing Committee in July, 1995. The Interregional Nursing Committee(INC) is composed of identified nursing leaders from all eleven regions within the KaiserPermanente organization. The proposal was reviewed by the committee and recommendationsfor clarity of the proposal were forwarded to the group. Agreement was reached by the INCthat the concept of common nomenclature for interregional use was needed. It was alsoessential the nomenclature developed needed to be multidisciplinary, that is, understood bythe entire health care team and not only the nurse.

The group reconvened based on the feedback from the INC and established a succinctdocument which outlined the mission statement and goals and objectives of the InterregionalNursing Nomenclature Committee (INNC).


25

Our mission statement currently reads, the INNC, in support of program strategic objectives,is to develop and implement a convergent nursing terminology for the Kaiser PermanenteMedical Care Program.

The goals agreed upon by the Committee are:- to co-ordinate the process within and among Kaiser Permanente regions, by which a

convergent nursing terminology is developed,- to assure the convergent nursing terminology is of sufficient specificity and robustness to be

useful for outcomes research, clinical care, identification of best practices, implementationof guidelines and the management of care,

- to assure the convergent nursing terminology will focus on human responses to actual orpotential illness and provide shared meaning for specific nursing diagnoses, interventions,outcomes and patient intensity across all settings where nursing is practised.

Consensus was reached by the group on the objectives to achieve our goals. (see Table 1)

Table 1Objectives of the Interregional Nursing Nomenclature CommitteeTo receive direction from and provide feedback to the Interregional Nursing Committee andto the Convergent Medical Terminology (CMT) GroupTo define standards for the use of convergent nursing terminology and/or linkage of localterminology to the convergent terminologyTo provide consultation and educational opportunitiesTo develop model educational programsTo recommend successful processes for implementation of convergent nursing terminologiesby local groupsTo act as a clearinghouse for terminology activities throughout the regionsTo consult with projects introducing convergent nursing terminologies into developing paperbased protocols and standardsTo maintain a library of resources, sample documents, and expertise for use within and acrossthe regionsTo refer locally identified terminology linkage issues to the KRep/ CMT project for resolutionTo maintain currency with state and national efforts in nursing terminology including thosedefining common outcome measures.

Linkages with the Controlled Medical Terminology ProjectAt the same time that the nursing group was pulling together resources and discussing ways toachieve a common terminology, the medical groups within Kaiser Permanente were alsolaunching a program wide effort to achieve a common medical terminology. This effort ispartially funded by the National Library of Medicine and is being done in conjunction withMayo Clinic. This project is the Convergent Medical Terminology (CMT) project and isutilizing a developmental tool from IBM called KRep. KRep is a knowledge representationtool that allows for graphical representation of modelling languages as well as subsumption ofterms. This subsumption of terms allows for easier retrieval of data for outcomes research.The CMT has populated the KRep tool with the Systematized Nomenclature of Human andVeterinary Medicine (SNOMED).4


26

The INNC reached agreement that the Common Nursing Terminology needs to be integratedwithin the CMT. The CMT group was approached by a representative from INNC and theINNC work was adopted as part of the CMT project.

Selection of Nursing TerminologiesINNC reviewed the terminologies and classifications currently recognized within nursing inthe United States. North American Nursing Diagnosis Association (NANDA), NursingInterventions Classification (NIC), and terms for patient findings from SNOMED wereselected as a base to begin our work. NANDA is currently contained within SNOMED andnegotiations are underway to include other nursing classification systems to SNOMED.

Subsequently, the INNC will follow the model of distributed development currently beingutilized by the CMT group and will use KRep for modeling of the common nursingterminology.5

Modeling of Common Nursing TerminologyThe modeling, strategies and tools being utilized in the CMT project were reviewed. to assurethat the nursing modeling would integrate well. Currently, the CEN standard for operativeprocedures out of Europe has been adapted,6 as well as Laboratory Observation IdentifierNames and Codes (LOINC) roles and facets for laboratory procedures.7 Roles are utilized inconcept definitions to describe relationships with other concepts. While facets attach toconcepts or roles information which cannot be expressed with roles and which have no impacton the classification.8

Table 2 provides a preliminary listing of roles for modelling NIC.

Table 2Preliminary Listing of RolesIs-ARelated-ToHas-EquipmentHas-Body-SiteHas-ApproachSpecimen-Role

Table 3 demonstrates some of the proposed facets for modelling NIC.

Table 3Proposed FacetsNIC CodeSNOMED CodeCPT-4 CodeICD-9 CodeComplete NamePreferred NameSynonyms

A proposed model for NIC was developed by the INNC. Figure 1 demonstrates the beginningof that model.


27

Figure 1Proposed Kaiser Permanente Model for NIC

Facet:NIC C ode:Dom ain 3

Facet:NIC C ode:

Class S

Facets:NIC C ode: ?

SN OM ED C ode:PA-60140N

Facet:N IC C ode:

5620

Grasp:Planned Teaching

Van Slyck:Short Term Teaching/C ounseling

Synonyms:

Facet:N IC CO de5620.01

AC TIVITYD em onstrate the Skill for the Patient

IN TER VEN TIO NTeaching Psychomotor Skill

R ole:R elated To

Facet:SN O MED C ode

P3-72949

Glucose M easurem ent by Monitoring D evice

Blood Glucose Testing

Blood Glucose q 2-4 hours (GR ASP)

Blood Glucose (FS) M onitoring (GR ASP)

(C apillary) FSBG FSBS

Glucose Monitor ing(Van Slyck)

Synonym s

B lood Glucose MonitoringR ole:

H as-Equipment

SU B C LASSD iabetic Pt. Education

R ole:Is-A

C LASSPatient Education

R ole:Is-A

D OMAINBehaviora l

R ole:Is-A

The model as shown in Figure 1 looks at the structure of NIC, allows for activities to bemapped to interventions, interventions mapped to either a NIC class or subclass, and then theNIC class to be mapped to the NIC domain. Activities that are found in multiple NICinterventions will be mapped to both with one code applied. A subclass was added to allowfor protocols or guidelines to be mapped to a specific area. The NIC class by itself did notallow, for example, a group of activities such as diabetic patient education to be easilyaggregated and retrieved.

Lessons LearnedThe INNC has learned some valuable lessons in the evolution of this project. One of the firstgroups we engaged was nursing leadership within the organization on an interregional level.Without their support, the project would not have proceeded this far. The INNC needed tocreate a business case for continuing work with this project. It was also important to link withthe CMT project so that an integrated approach versus a fragmented, discipline specificapproach could be achieved. In order for the INNC to reach our goals, we needed to utilize thestrategies and tools that were already in place within the CMT project.

The INNC needed to tie our project with the quality initiatives within the organization as wellas those external to the organization. It is important that the terminology be modeled so thatdata can be easily retrieved and be comparable across the Kaiser Permanente program.

Next StepsThe INNC is now seeking the resources to perform the modelling work to be done. Initiallythe INNC expects registered nurses within several of our regions to model the terminology.


28

The INNC is also exploring as a next step the linkage of our work with an outcomestaskforce project, Health Status Outcomes Dimensions (HSOD).Kaiser Permanente is seeking to enhance our clinical information systems as well as ouroutcome reporting and data comparability. This is a movement forward for all healthcareorganizations within the United States. The INNC within Kaiser Permanente has built thebeginnings of a foundation to create a common integrated healthcare terminology that isrobust enough for clinical documentation, but is modeled so that retreivability andcomparability of data is possible across the organization.

References1. International Council of Nurses, Nursing’s Next Advance: An International Classification for Nursing

Practice (ICNP), Geneva, Switzerland, 8 October 1993.2. Zielstorff R, Hudgings C, Grobe S, Next-Generation Nursing Information Systems, Washington (DC),

1993; 5-11.3 International Council of Nurses, Nursing’s Next Advance: An International Classification for Nursing

Practice (ICNP), Geneva, Switzerland, 8 October 1993.4 Cote R, Rothwell D, Palotay J, Beckett R, Brochu L, The Systematized Nomenclature of Human and

Veterinary Medicine, Northfield, IL, College of American Pathologists, 1993.5 Campbell K, Distributed Development of a Logic-Based Controlled Medical Terminology, Dissertation

Proposal February, 1996, Stanford University.6 CEN, European Committee for Standardization, European Prestandard, Medical Informatics, Structure for

Classification and Coding of Surgical Procedures, November, 1994.7 Regenstrief Institute, Laboratory Observation Identifier Names and Codes (LOINC), Release 1.0f, 12/21/958 IBM, KRep User’s Guide and Reference- Draft, Yorktown Heights, NY, 1995.

AcknowledgmentsThe author wishes to acknowledge the work of the INNC members: Linda Dietrich, Northwest, Mary Lush,Northern California, Barbara Carroll, Southern California, and Carly Kirby, Colorado. The author would alsolike to thank the members of INC and the CMT Project, as well as Simon Cohn, MD, Keith Campbell, MD, andJohn Mattison, MD for their continued support of the project.


29

The Danish National Health Classification System - Nursing InterventionsClassification - a Part of a Common Danish Health Care Classification

Madsen Ia,c and Jytte Burgaard SA b,c

aDepartment of Health Care Informatics, Skejby Sygehus, Århus University Hospital, Brendstrupgaardsvej 100,8200 Århus N, Denmark, bDepartment of Nursing Management, Hvidovre University Hospital, Kettegaardsalle30, 2650 Hvidovre, Denmark, cNursing Classification working group, The Danish National Board of Health.

The Danish Nursing Intervention Classification is the first national nursing intervention classification. Becauseof the official status of the project it is to be a national standard. It will thus be possible for the system to be ingeneral use nation-wide, which is a necessary condition for an efficient communication in the health care sectorin Denmark.

Background.The Nursing Intervention Classification is a part of a project to develop a common health careclassification system in Denmark. This project was started in 1992 by the Danish NationalBoard of Health. This common Danish classification system will unify all existing officialclassification systems in a joint hierarchical structure which will form the basis of allclassifications in a common system. The Nursing Intervention Classification is as previouslymentioned a part of a common classification and it is thus possible to use codes from the otherclassifications e.g. ICD 10, and drugs classification. Supplementary classification will also beuseful in connection with the specification of a level of detail of for example, right or leftsided, or personnel involved.

Today a number of classification systems are used in the Danish Health Service, but there is atpresent no official classification in nursing. As a consequence of this, a number of localclassifications have emerged, preventing communication across units, hospitals and sectors. Inaddition, a comparable registration of all essential interventions will in future contribute toensuring the quality of patient treatment in the Danish National Health Service. In 1992, theDanish National Board of Health started a project with the purpose of establishing a commonclassification system in Denmark. This system will form the basis of an up-to-date use ofinformation systems in all relevant connections in the Danish Health Care Service. The nameof the project and the system is: The Danish Health Classification System (DHCS) whichtranslates into Danish : "Sundhedsvæsenets Klassifikations System. (SKS)". 1

Joint hierarchical structure.The SKS system combines all the existing official classifications in a joint hierarchicalstructure, which will form the basis of the preparations of classifications to come. Commonrules for the structure of a classification system and for the adoption of codes have been made.This will ensure that new classifications in all areas of the health service will be elaboratedand maintained according to the same principles and with clear limits. The official status ofthe system will ensure its general use nation-wide, which is a necessary condition for anefficient communication system in an integrated health service

The Danish National Board of Health have stated the following requirements for theclassification:

1. One coherent system with clear interfaces between the sub classifications. It would be anadvantage if all existing concepts can be coded in one coding system. This would facilitate thesearch for the right code, the coding process and the use of data. In addition, it would make it


30

easier to understand the structure of the classification, and the coding of one concept withseveral codes according to different classifications is prevented.

2. One concept - one code. For any relevant concept within the health care service there mustbe one code, and only one code.

3. Consistent hierarchical structure. A hierarchical structure ensures that the code for aconcept indicates the logical position of the concept within the classification structure. Codingcan be made at several levels without losing the possibility for comparison. In addition, avarying degree of detail is essential for order entry and finally the hierarchical structureconsiderably simplifies retrieval and analysis of data .

4. Unambiguous and homogeneous codes. The individual codes must be unambiguous, welldefined and cover a relatively uniform "piece of work" or concept. This is a prerequisite forusing point systems for estimating the production of the units and calculating the price of theindividual services. The use of codes such as "others" should therefore be limited to aminimum through the creation of a sufficient number of specific codes.

5. Clear and Pure. The division of the classes of the classifications must as far as possible bebased on one logical criterion. Any sub-division aimed at compensating for deficiencies inother parts of the classification must be avoided. If additional information is required, it mustbe added by registering more (possibly combined) codes rather than linking several non-related concepts in one code.

6. Danish texts. The code text must correspond to Danish terminology. If good, unambiguousDanish words exist they should be used, as the classification system will often be used bypersonnel groups who are not proficient in Latin or English. The code texts have to be able tostand alone and be immediately understandable to the target group.

7. Fast and consistent updating. If a classification is to be used in a rapidly developing healthservice, it must be currently updated. In particular when making forward registrations, forexample order entry and booking, updates must be made swiftly and preferably months beforethe codes come into force. It is essential that changes and modifications are strictly controlledby a few competent persons to avoid the classification becoming polluted with codes whichdo not comply with the rules that have been adopted.

The first level structure in the Danish Health Care Classification is:

A AdministrationB Treatment (= non- surgical therapeutic procedures) NURSING INTERVENTIONS.CD Diagnoses ( WHO's classification of diseases)E Special types of diagnoses (e.g. pathology diagnoses)FGH Medical technology, aids, etc.IJK Surgical procedures


31

L Living conditionsM Drugs (the ATC classification)N (Reserved for NURSING DIAGNOSES.)O (not used)P Classification of primary careQRS Medical history / symptoms/ clinical findings / effect of treatmentT Topography / anatomyU Examinations ( = non surgical diagnostic procedures ) NURSING INTERVENTIONS.VX (Reserved for local use)YZ Other classifications.

The practical approach.A classification system of this type and size cannot be made by just a few people, whoproduce the complete code from A to Z, as once it was finished it would already be obsolete.It is necessary to build up the system in modules, using an approach where the individualparts of the classification are made layer by layer.

This means starting by making an overall structure where the present official classificationcan be inserted as modules and where a first crude classification can be made in areas wherethere is no official classification today. 1

The working process.The Danish nursing working group, appointed by the Danish National Board of Health, isresponsible for the development of a nursing classification system concerning nursingdiagnosis, nursing intervention and nursing assessments, and started working on the nursingclassification in 1993.

All but two of the members are nurses. One member is a representative from the nursingauxiliary group and the last member is a representative from the steering group.

The work will gradually be carried out in greater detail within the areas and in the order thatmay be deemed appropriate. Older modules can also be replaced by new ones, as thisbecomes necessary. Another advantage of this procedure is that many persons can workindependently of each other on each part of the classification. As the hierarchical codes arenamed at all levels, the classification can be implemented step by step top-down without therisk that later extensions will cause problems.

It is intended that the work will be performed by a number of working groups with membersfrom the various scientific societies and health care personnel associations. In order to set upworkings groups that can work with fairly well-defined parts of the total classification system,the planned structure of the classification system is in several respects based on the medicaldivision into specialities which are found within the health service.

The Danish Nursing Intervention Classification contains 12 domains. These domains containa total of 523 interventions collected from nurses all over Denmark. The interventions have


32

been classified according to the hierarchical structure framed by The Danish National Boardof Health and each intervention will have been coded - also according to the rules of theDanish National Board of Health.

Bottom-up approach.The working group chose in co-operation with the Danish Institute for Health and NursingResearch, which was developing a classification of nursing diagnoses, to arrange theme daysall over Denmark with the purpose of informing and motivating with regard to the work withfuture nursing classifications.

We chose a participant perspective because we wanted as many nurses as possible to feelownership of the classification systems from the start. It is our hope that having contributed tothe development of the classification system, nurses will use the system in practice.

On the theme days and through articles in the nursing magazine "SYGEPLEJERSKEN" 2,3

nurses were requested to submit concepts for nursing diagnoses, nursing interventions andnursing assessments from their daily practice.

In parallel with these initiatives, the working group collected and analysed existing Danishnursing classifications 5 and studied other international nursing classifications 6,7,8,9,10. Thegroup also studied recommendations for international classifications 11,12,13,14,15,16.

We chose to study the structure of the American nursing classification system (NIC),developed by Bulecheck and McCloskey 7. The structure of this classification was interestingwhen compared with the structure of the Danish Classification System.

We received 750 interventions, which were sorted and grouped into a new Danish nursingintervention classification system. Danish nurses have shown themselves to be very interestedin this classification work and the members of the working group have given many lectureson nursing classification all over the country.

For almost 3 years, we worked intensively with the assessment and interventionsclassifications, including nursing data collected from all over Denmark. This work was donein parallel with our employment in different places in Denmark.

ResultsAt present, we have finished the intervention classification. This classification was releasedfor use spring 1996 after the system had been submitted to the profession and to an officialhearing. The interventions are coded and are thus ready for use in computer-baseddocumentation in nursing information systems. At present the interventions classificationcontains 12 domains. These domains contain in all 569 interventions. The interventions havebeen classified after the previously mentioned hierarchical structure framed by the DanishNational Board of Health. In order to give a more explicit formulation of the nursinginterventions for each patient, the structure of the system makes it possible to usesupplementary codes.


33

The future.The Nursing Intervention Classifications are suitable for use in an electronic patient record,national databases, statistical specification control, quality improvement, minimum nursingdata set, research and in the curriculum for education of nurses and students.

The nursing intervention classification will be used on a trial basis for two years after it hasbeen coded and will then be ready for integration into future international classifications 11,17.The work with the nursing diagnoses and assessment classification is still going on. TheNational Board of Health have not yet determined the evaluation criteria for the trial period.

Nursing practice can not be described by a nursing classification alone. Nurses will only beable to document nursing practice if they use the whole Danish Health Care Classification.

References :1. Sundhedsvæsenets Klassifikations-System (SKS). The Danish Health Classification System (DHCS)

Danish National Board of Health, Copenhagen.Sundhedsstyrelsens medicinal statistisk afd., Denmark,1992.

2. Mortensen, Randi : Øjeblikkelig indsat ønskes. Sygeplejersken 1993;(2).3. Madsen, Inge & Burgaard, Jytte : Til vurdering i praksis . Sygeplejersken 1994;(39)4. Madsen, Inge : Fremtidens sygepleje dokumentation. . Sygeplejersken 1996;(19).5. Salling, Anne Lise. Stimulation af patienters aktivitet og udvikling. Dansk Sygeplejeråd, Denmark, 1991.6. Saba V.Georgetown University, School of Nursing.Washington DC, USA.: A Home Health Classification

Method.Informatics '94. Proceeding of the Fifth IMIA International Conference on Nursing Use ofComputers and Informations Science, Texas. Amsterdam:Elsevier 1994:697.

7. Buleheck G.M, Mc Closkey JC. Nursing Intervention Classification (NIC). Taxonomy af nursingintervention. St. Louis Mosby year book,1992.

8. North American Nursing Association. Taxonomy I : Revised. St. Louis, MO: NANDA,1990.9. Grobe S, Pluyter-Wenting ESP. Eds. Classifications. Nursing Informatics '94. Proceeding of the Fifth IMIA

International Conference on Nursing Use of Computers and Informations Science, Texas. Amsterdam:Elsevier, 1994: 687-701.

10. Saba V et al: A Nursing Intervention Taxonomy for Home Health Care. .Nursing & Healthcare 1991;12(6).11. Clark, J, Lang N.: Nursing's Next Advance : An International Classification for Nursing Practice."

International Nursing Review 1992;39(4).12. Iversen I et al.: Minimun data-sett for sykepleie. Sykepleien 1993;(2).13. Nielsen GH, Mortensen R. Nursing Terminology as a Means of Identifying Nursing Diagnoses Basic Level

Diagnostic Categories in Nursing. Danish Institute for Health and Nursing Research, Copenhagen.14. Ball M, Douglas J Healthcare Informatics. Healthcare Informatics Magazine, May 1990.15. Roger F. The Road to Standards. Healths Informatics Europa. 1993;12.16. Standardization Framework in Europa, Annex 2 to SESAME, part II-I. : Classification and definitions

vocabulary.17. Wake M et al. Toward an International Classification for Nursing Practice. International Nursing Review

1993;40(3).

A Compositional Approach to Nursing Terminology

Documents

Transcript of A Compositional Approach to Nursing Terminology