The Omaha System: Bridging Hospital and Home Care
By Kathryn H. Bowles, Ph.D., R.N.
Although the aim of integrated health care networks is to provide seamless health care, the transfer of information from one health care agency to another has not been standardized. The current and evolving integrated health care system, in which information management, continuity, and communication are priorities, has important implications for the use of nursing classification systems.
Many studies have shown that the amount, type, and quality of information communicated across health care settings is inadequate, fragmented, or redundant (Anderson & Helms, 1993, 1994, 1995, 1998). As patients progress through the health care system, a standardized language to communicate patient problems, interventions, and outcomes is a valuable tool to improve efficiency and continuity. For nursing, many advocate the use of one of the American Nurse Association's (ANA) recognized nursing languages to improve communication, continuity, and information management (Head, Maas, & Johnson, 1997; Keenan & Aquilino, 1998; Lang, et al., 1995; Ozboldt, 1995). One of these languages, the Omaha System (Martin & Scheet, 1992a) may be suited for this challenge. The Omaha System was chosen for this study because it has been used for over 20 years in home care; it offers the ability to link patient problems, nursing interventions, and outcomes; it is of manageable size; and it performed well in pilot work with similar transitional care data sets (Brooten, Bowles, Kirby, & Miovech, 1997). Although beginning to be used in acute care (Doran, Sampson, Staus, Ahern, & Schiro, 1997; Sampson & Doran, 1998), the Omaha System has never been systematically evaluated for that expanded use. This study was the first attempt to evaluate the use of the Omaha System categories to code documentation from an acute care setting including the period from admission to two weeks after discharge to home. This article summarizes that study. A full report of the study is included in other publications.
Even though, the transition of patients from hospital to home is the most common opportunity for continuity of care, surprisingly, the impact of the amount, type, and quality of information exchanged during this transition on patients' post hospital outcomes has not been studied. In addition, no one has evaluated the Omaha System for this expanded use. These are important areas for future research to further evaluate the impact of the use of standardized languages across settings. This study was a much-needed first step in evaluating the Omaha System for such an application, especially as we move toward an electronic medical record.
The Omaha System is one of six ANA recognized nursing languages included in the Unified Medical and Nursing Language Systems (Humphreys & Lindberg, 1989; McCormick, et al. 1994; National Library of Medicine, 1994). It has been used successfully in home care and community health for over 20 years (Martin & Scheet, 1992a; Weidmann & North, 1987). As described in this issue by Martin, the patient problems are categorized into Environmental, Psychosocial, Physiological, and Health Related Behavior domains. The Intervention Scheme includes four intervention categories, Teaching, Guidance, and Counseling; Treatments and Procedures; Case Management; and Surveillance. Sixty-two targets provide terms to describe the object of the nursing intervention and patient specific information enables the nurse to write further description of the target, if needed.
The Problem Classification Scheme includes 40 patient problems each with defining signs and symptoms. Because of this, the patient assessment is very comprehensive, taking into account many of the problem areas experienced by patients who are transitioning from hospital to home. The application of a well established, valid, and reliable language from the home care setting is logical to consider as a possible instrument to bridge the gap from hospital to home. This study evaluated the adequacy of the Omaha System to capture the terms used by nurses as they recorded the patient problems and interventions used during hospital care and discharge planning. The study has implications for expanding the use of the Omaha System from community settings to acute care. A summary of the study is discussed below.
The Omaha System was applied to the in-hospital documentation from a prior clinical trial of "Comprehensive Discharge Planning for the Elderly" (Naylor, et al., 1994). The original clinical trial compared the effectiveness of a comprehensive discharge planning protocol implemented by advanced practice nurses (APNs) to the hospital's general discharge planning procedures. The patient problems and nursing interventions received by elderly patients with cardiovascular medical diagnoses were documented by APNs in hand written patient care logs and by staff nurses in the patient's hospital record. The content of the logs and hospital records were extracted using content analysis and the results were coded using the terms of the Omaha System. Simultaneously, the terms were rated for degree of match using concept match scores.
Thirty records were randomly chosen from the 138 records generated in the experimental group of the prior clinical trial. The hospital records contained the staff nurse's notes, nursing care plans, and teaching plans. The patient care logs contained free form, hand-written APN notes of the patient problems and nursing interventions recorded during the patient's hospital stay and discharge to home.
Each patient record was coded using content analysis. Signs and symptoms, problems, interventions, and the targets of those interventions were matched to Omaha System terms. The degree of match between the Omaha System term and the unit of analysis was rated according to concept match scores. The scores ranged on a five-point scale from no match to perfect match. The coded terms and match scores were entered into a relational database that enabled data retrieval and descriptive analysis. Evaluation criteria were appliedduring and after the coding.
Validity and Reliability
A validity check was performed to assure that the unit of analysis did not lose its meaning once removed from the context of the patient record. Reliability was tested by an independent coder, a doctorally-prepared nurse experienced with the Omaha System, who re-coded a randomly selected 100 units of analysis and a percent agreement and Kappa statistic were calculated. Stability was determined when the investigator coded the same 100 units of analysis twice and the results were compared for percent agreement and Kappa. In addition, prior to coding the study charts, the investigator coded three case studies and sent them to content expert Karen Martin to evaluate accuracy. Ms. Martin discussed the coding approach with the investigator and provided guidance regarding interpretation of definitions and the development of decision rules.
The sample of 30 elderly, cardiac patient records yielded a variety of signs and symptoms or problem terms and corresponding nursing interventions. The ability of the Omaha System to capture the terms used by nurses in acute care were evaluated in relation to: 1) used and unused categories, 2) terms coded as "other", 3) degree of match between the Omaha System terms and the hospital terms, and 4) reliability scores.
Categories Used and Unused
All four domains were used and as expected for an in-hospital sample, nearly half of the problems were in the Physiological domain. Review of the unused categories revealed that they are appropriate for other types of patients or care settings and do not require modification.
All four intervention categories (Teaching, Guidance, and Counseling; Treatments and Procedures; Case Management; and Surveillance) and 86% of the targets were used. Unused target categories such as bonding, cast care, discipline, and family planning are appropriate for other types of patients.
Items Coded as "Other"
The Omaha System provides one "other" category within each problem domain to code terms that are not found in the Problem Classification Scheme. In this study, the Omaha System was adequate to code almost all of the problems in the hospital records. Of the signs and symptoms or patient problems documented, only three were coded as "other". The "other" category was used when specific nursing interventions occurred for a problem that had no Omaha category label.
In the Omaha System all interventions have targets defined as the terms used to describe the object of the intervention (Martin & Scheet, 1992a). In this study, very few targets required coding into the "other" category.
Degree of Match
The concept match scoring quantified the degree of match between the hospital terms and the Omaha System. The scores ranged from no match to perfect match. A small percentage of problem terms were coded as no match.
The match scores for the targets were also favorable with a small percentage scored as no match. The majority matched in the higher levels of the concept match rating scale.
Reproducibility (intercoder reliability) scores demonstrated the ability of two independent coders to produce consistent results when coding hospital records using the Omaha System. Stability (intracoder reliability) scores represented the consistency or stability of the same coder when using the Omaha System. Kappa scores for this study were all >0.79, indicating substantial to almost perfect agreement.
Strengths and Limitations of the Omaha System
During this evaluation, strengths and limitations of the Omaha System in acute care became apparent. The first strength was the ability to code the majority of the problems, interventions, and targets from the hospital record. The Omaha System provided four domains to holistically document patient problems in Environmental, Psychosocial, Physiological, and Health Related Behavior domains. In this study, the patient's problems were coded into the Problem Classification Scheme using all four domains. Analysis of unused categories revealed terms appropriate for other populations. The unused terms would not be expected to occur in this elderly sample with cardiovascular conditions (i.e.: post partum).
In addition, the high concept match scores are a reflection of the abstract nature of Omaha categories. The terms are broad which enables the coding of many hospital terms that are more descriptive, yet, fit into the Omaha categories. The system is also comprehensive having four domains available to code a variety of patient's problem areas. The signs and symptoms listed under the problem categories provided many of the perfect matches. This is consistent with recent findings by Zeilstorff, Tronni, Basque, Griffin, and Welebob (1998). When mapping the Omaha System problems with the North American Nursing Diagnosis categories and the Home Health Classification, they found the Omaha System signs and symptoms contributed to the consistency and reliability of usage.
Another strength was the ability to achieve high Kappa reliability scores. This represents the ability to remain stable when coding terms from the hospital setting. This may be a reflection of the manageable size of the Omaha System. The consistency in coding may have been achievable because the Omaha System has only 40 problem categories, four intervention categories, and 62 targets. There is also a pocket guide (Martin & Scheet, 1992b) that was invaluable as a reference for definitions and examples. In addition, several decision rules were generated to guide the coding.
The availability of the Omaha System problem categories for common transitional care problems such as Nutrition, Income, Residence, Communication with Community Resources, Caretaking, Cognition, Pain, Neuro-muscular-skeletal Function, Physical Activity, Health Care Supervision, Prescribed Medication Regime, and Technical Procedure is a strength. The Omaha System was derived from home care records so it guides the assessment of areas of importance during discharge planning and after hospital discharge (Bowles & Naylor, 1996; Martin & Scheet, 1992a). This strength suggests it may be suitable for documenting patient needs and nursing interventions in transitional care from hospital to home (Bowles, 1997a).
The availability of the "other" categories was another strength. The Omaha System provided a method to code terms from the hospital record not found in the present Omaha System. This open architecture accepts the addition of new terms.
Although the availability of the "other" category can be seen as strength, the need to use that category does indicate a limitation of the Omaha System. It may indicate a need to add additional categories to the existing classification system. Further evaluation of the Omaha System is needed to determine the need for new problem categories.
Finally, a limitation of the nursing classification systems identified by Henry and Mead (1997) was also found in this study. Henry and Mead reported that data transformation with the Nursing Intervention Classification (Iowa Intervention Project, 1993), the Omaha System (Martin & Scheet, 1992a), and the Home Health Care Classification (Saba, 1992) all resulted in data loss. Similarly, to avoid data loss in this study, the Omaha System "patient specific information" category was used to describe the targets of interventions in more detail. Patient specific information is a category within each target that allows a narrative description of the target. For example, patient record data that reads, "taught signs and symptoms of wound infection", if coded as a teaching intervention to target #50 physical signs and symptoms, the abstract target #50 would not indicate what specific signs and symptoms were taught. This loss of detail is an important limitation not exclusive to the Omaha System. At least the Omaha System provides a means to capture that description, although it is in narrative form.
The broad, abstract, and holistic categories in the Problem Classification Scheme of the Omaha System provided adequate coverage for terms described within hospital records. This study suggests the potential for using the Omaha System in the acute care setting, especially as a tool to facilitate communication and continuity of care from hospital to home care. This is critical within integrated health care systems especially if the home care agency already uses the Omaha System. The potential to communicate the signs and symptoms, problems, nursing interventions and the status of outcomes across settings is a valuable clinical and research tool. It is also an important feature of evolving nursing information systems (Bowles, 1997b). In this study, and in over 20 years of clinical use and testing, the Omaha System has demonstrated many strengths (Bowles, 1996; Cell, Peters, & Gordon, 1984; Hays, 1992; Helberg, 1993; Marek, 1996; Martin, Scheet, & Stegman, 1993).
Presently there is no perfect or ideal nursing classification system. Users must evaluate research findings and examine their needs before choosing a system. Although these results are encouraging, this study has several limitations. The analysis was limited to the hospital records of elders with five different cardiac diagnoses, in one hospital. Therefore, further research, testing, and development of the Omaha System in acute care is encouraged using diverse sites, patient ages and diagnoses, and methodologies.