As the population ages, the age cohort of individuals age 85 and older will be increasing the fastest, with a projected increase of 300% to 400% between the years 1990 and 2010 (Mezey & Scanlon, 1988). Given that one in four people in this cohort have some need for nursing home care, the need for this care also will be increasing. In addition, patients who historically would have convalesced in the hospital currently are discharged to their homes, if possible, or to nursing homes for convalescence. Nationally, there are two times the number of beds in nursing homes than the number of acute care hospital beds (Ouslander, 1989). As the health care environment changes to accommodate this changing patient population, it is important to be able to articulate the nature of nursing work in long-term care to assure effective staffing and quality care. The purpose of this research was to evaluate the reliability and validity of the Leatt Measure of Nursing Technology (LMNT) for use in long-term care nursing environments.
The theory of comparative organizational analysis developed by Perro w (1967) allows the nature of work to be described and explained in terms of the work technology. Work technology is defined as the actions performed on an object or person, with or without tools, to make a change in that object or person (Perrow, 1967). Technology, as used by Perrow (1967), includes but is not limited to, mechanical technology. The technology of any work is measured by the amount of uncertainty, instability, and variability involved in making any change in an object or person, regardless of the amount of mechanical technology also in use. Work demands in varied situations can be compared using Perrow's (1967) theory, for example, comparing nursing to other kinds of work (e.g., police work, physicians' work) or comparing work demands across various types of nursing units.
In the 1970s and 1980s, Canadian researchers developed and refined an instrument based on Perrow's (1967) theory of work technology that measured the nature of nursing work in the hospital acute care setting (Leatt & Schneck, 1981, 1982a, 1982b, 1984; Overton, Schneck, & Hazlett, 1977). Leatt and Schneck (1981) operationalized the technology of the nursing profession as follows:
* Uncertainty is the percentage of patients with more than one diagnosis, with complex nursing problems, and the amount of nursing intuition or judgment is required in providing nursing care (i.e., predictability of nursing knowledge). Example: "What percentage of the nursing care given relies on nurses' intuition (judgement) rather than on set procedures or routines?".
* Instability is the percentage of patients requiring frequent observation and care, specialized monitoring, and potential emergency situations (i.e., predictability of patient condition). Example: "In your estimation, what percentage of the decisions made by the nursing staff are made independently from physicians?".
* Variability is the percentage of patients with similar health problems on the nursing unit and the variety of nursing techniques employed (i.e., predictability of work demands and patient needs). Example: "What percentage of the decisions made by nurses during their work are repetitive from one day to the next?".
The original Leatt study focused on the variety of nursing work environments in the hospital acute care setting. Of the 157 nursing units studied by Leatt and Schneck (1981), 10 were "auxiliary" units, comparable to current long-term care facilities according to Leatt (personal communication, fall, 1991). The Leatt instrument continues to be used for study of the nature of nursing work in the hospital acute care environment (Alexander & Bauerschmidt, 1987; Armstrong, Krahenbuhl, Muellenbach, & Savage, 1984; Mark & Hagenmueller, 1994; Mitchell, Armstrong, Forshee, & Lentz, 1989; Verran & Reid, 1987).
Much has changed in the organization, location, and delivery of services in health care. Many of the patients who would have been in the acute care general medical units identified in the Leatt studies currently are found in independent long-term care environments. In the original studies, the auxiliary units scored as "low" on all three measures of work technology (Lean's measures were either "high," "medium," or "low"). There is no information regarding how these auxiliary environments currently would score. The potential of the LMNT as a tool for studying longterm care nursing work environments has not been explored.
Ten long-term care (LTC) facilities in the Seattle metropolitan area were selected to participate in this study. This purposive sample was designed to maximize the variability of the LTC facilities and included for-profit and not-for-profit ownership, and sizes ranging from 70 to 250 beds. One facility withdrew during data collection, citing a management change as the reason, leaving a final sample of nine LTC facilities. Five facilities were not-forprofit, and four were for-profit ownership.
A total of 113 licensed nursing personnel from 35 nursing units returned usable questionnaires (45% response rate). The respondents were RNs (n - 69) and licensed practical nurses (LPNs) (n = 44). Eighty-four respondents worked full-time (»= 36 hours per 40-hour pay period), and 29 worked part-time (mean = 24 hours per 40hour pay period). Respondents had been in their current positions for a mean of 3.3 years (range = 6 months to 20 years). Mean years of experience in LTC was 8.4 years (range = 1 to 34 years). Seventy-six respondents (67%) also had previous experience working in acute care settings (mean = 6.8 years).
RELIABILITY COEFFICIENTS FOR LEATT MEASURE OF NURSING TECHNOLOGY (LMNT) SUBSCALES
The original LMNT, introduced in 1981, contains 20 items in three subs cales (Uncertainty =10 items, Instability = 7 items, and Variability = 3 items). The instrument later was adapted using the scoring mechanism introduced by Armstrong et al. (1984), changing the 5-point Likert scale to a 10-point percentage scale based on feedback from RN survey respondents. For this study, the instrument was adapted further by adding an introductory paragraph explaining the study and guaranteeing anonymity of responses. A section of demographic questions also was added. At the end of the instrument, respondents were asked to identify up to three nurses who they believed could "best discuss the work of long-term care nursing in this facility" who would take part in a focus group interview. They were told they could list themselves.
The research plan and instrument were approved by the Institutional Review Board of the University of Washington to assure protection of human subjects. The Director of Nursing in each of the facilities was contacted for permission for research participation and one of the longterm care faculties required an additional full research review process.
Questionnaires were distributed to all RN and LPN personnel in each of the nine facilities via inhouse mail and message systems. A large collection envelope was located in each facility (e.g., at the main reception desk, in the staff development office or library) for staff to deposit responses in their individual, sealed envelopes. The names of licensed staff identified as potential key informants were collected from the questionnaires. The most frequently named staff representatives from all areas of the facility (5 to 7 names for each facility) were contacted to be participants in a focus group interview. The focus group interviews were conducted in the facility at a time convenient to the majority of the participants, and each facility provided a conference room for the interview. The group interviews were semistructured around the themes of the LMNT subscales. Participants were asked about the work of long-term care nursing in general and then were given copies of the questionnaires to review. Interviews ranged from 40 to 90 minutes and were audiotape recorded and transcribed verbatim. Comparative analysis (Knodel, 1993) was used to identify major and minor themes in the interviews with the different focus groups. This involved reviewing themes that were consistent with the LMNT subscales as well as themes that were not covered by the LMNT but were identified by participants from across focus groups.
An initial concern was whether questionnaire responses were systematically different between RNs and LPNs or between full-time or part-time workers. Therefore, the questionnaires were evaluated using a t test to determine if there were differences in the three subscale scores between RN and LPN respondents, or differences between respondents working full-time and part-time. No statistically significant difference was found between the groups of scores at the p =.05 level, and all questionnaires were used as a single group in this analysis.
Total scale reliability was evaluated by the item to total correlations. These ranged from .75 to .78, with a standardized item alpha of .77. The internal consistency of the subscales was evaluated using Cronbach's alpha. The subscale alpha coefficients for the original instrument, two subsequent studies, and this LTC pilot study are presented in Table 1. Armstrong et al.'s (1984) scoring change from a 5-point Likert scale to a 10-point percentage scale, also used by Mitchell et al. (1989), increased the variability of the responses and subsequently lowered the alpha coefficients. While the alpha coefficients are not strong, the internal consistency of the instrument is supported moderately.
The LMNT items were analyzed using a principle factor extraction factor analysis with varimax rotation to evaluate the construct structures. The number of subjects is lower than the generally identified rule-of-thumb 10 per variable. However, the analysis provides a rough check on the construct structure. The resulting factor structure was consistent with the factor structure in the original LMNT with the exception of two items (Items 1 1 and 19). Table 2 lists the subscale items and factor loadings. The two items that loaded very low (Items 11 and 19) on the identified LMNT subscales were addressed specificaiIy in the focus group interviews and are discussed further below.
FACTOR LOADINGS FOR LEATT MEASURE OF NURSING TECHNOLOGY SUBSCALES
Content validity was evaluated by an analysis of questionnaire themes and individual questionnaire items in the focus group interviews. Respondents independently identified the nursing tasks and responsibilities in long-term care settings and articulated these themes as consistent with uncertainty, instability and variability, supporting the general content (factors) of the original LNMT subscales. Items 11 and 19 were a specific focus of a portion of the interviews.
Interview participants reinforced the adequacy of the content of the Uncertainty subscale (see list of items in Table 2), with the exception of Item 11. This item asks, "What percentage of patients have individualized nursing care plans?," as part of the measure of Uncertainty, and it loaded independently at .80 in the factor analysis. Focus group respondents verified that LTC residents are required by state and federal regulations to have individualized care plans on file. Item 1 1 does not contribute to the content on uncertainty and will be dropped in future studies.
Content validity of the Instability subscale was evaluated through a general discussion, with specific focus on Item 19, which asks for the percentage of nursing decisions that are made independently of physicians. This item also loaded independently in the factor analysis at .81. Focus group respondents discussed the autonomous role of LTC nurses, commenting on the inf requency of physician visits and the frequent inability to contact physicians before nursing actions were required. Working independently from physicians was not associated with work instability. This item must be evaluated further as a single-item factor.
While respondents basically agreed with the content of the Variability scale, they did expand on the content of the three-item subscale. Specifically, the focus group participants articulated the variety of goals of patient care and treatment in long-term care settings. Some LTC patients have care and treatment goals similar to acute care patients (i.e., rehabilitation and discharge home). However, other LTC patient goals were identified (e.g., maintenance of optimum function with expected decline) that may be either rapid or slow. Terminal care also is a patient goal and may be either rapid or slow. These varied care goals were articulated by one nurse who said:
Coming out of working in CCU [coronary care unit], I wanted to save everybody and I had a hard time with that - it's really hard because you're dealing with quality of life and the dignity of dying.
Another respondent described patients who were in
the dwindles - it's a gradual decline [and] we know it's happening. We're trying to have them eat or whatever - and truly fighting upstream on that. It's a gradual thing - in decline.
This variability in care goals and patient outcomes is not captured in the original LMNT questionnaire, and additional items should be added and tested.
In the focus group interviews, participants were asked if they found any other questionnaire items problematic for any reason. They identified Item 8 as confusing. This item measures instability and asks what percentage of time is there a greater pressure to give care quickly because of residents' critical physical conditions. The participants discussed how they felt pressure to give care quickly because of the volume of work to be completed, citing an average patient load of 12 to 30 residents per nurse. They further discussed the concept of "critical condition," explaining that the concept of "fragile condition" also is appropriate to long-term care environments. Item 8 should be clarified in terms of the pressure to give care quickly because of resident condition as well as because of volume of work, and could be expanded into two separate items.
While LTC residents may be relatively stable, their conditions are fragile, or as one respondent explained, they have "no ability to adapt to a physical or psychological insult." Therefore, if a resident sustains a minor injury, it could signal a major change in condition and increased need for physical and psychological care. The Instability subscale needs to be expanded to include fragility as an element of predictability of patient condition. Respondents articulated this as requiring keen observation skills and knowledge of the residents to be able to identify subtle changes. One example given was knowing a resident well enough to recognize a subtle behavior change as an early indication of a urinary tract infection, indicating a need to intervene early. Another example was the need for rapid response to a noncritical insult such as a skin tear to minimize the effect on the frail physical condition. On a broader level, focus group participants commented on the ways in which a change in one resident's condition affects the other residents in the close community of the LTC environment - an element of their work that was not captured in the LMNT.
This study supports the use of the LMNT in long-term care nursing environments. The general nature of nursing work measured as uncertainty, instability, and variability is consistent with the Leatt interpretation of Perrow's (1967) theory of organizational comparison. Comparisons can be made using the original instrument. However, some adaptations are suggested for future development in LTC, as well as in acute care environments.
Analysis of the focus group responses suggests the need for changes to the instrument that would be appropriate for all nursing work environments. One example of this is the elimination of Item 1 1, the percentage of patients with complete nursing care plans. This item is obsolete in acute care settings as well as in LTC settings because the Joint Commission for Accreditation of Healthcare Organizations (JCAHO) standards require care plans for all acute care patients.
The focus group participants described the importance of knowing the patients in LTC nursing work. Swanson (1993) defined the theoretical construct of knowing as "informed understanding of the clinical condition (in general) and the situation and client (in specific)" (p. 355). The idea of individualized patient care and knowing the patients as individuals has been the focus of research on nursing clinical decision-making (Evans, 1996; Tanner, Benner, Chesla, & Gordon, 1993). This research has been conducted primarily in acute care settings and in intensive care units in particular (Radwin, 1995, 1996). Long-term care nurses in this study articulated "knowing the patients" and knowing their patients intimately, within a very different environment, over a longer period of time, and with a broader perspective. These nurses spoke of knowing the patients as individuals, as members of families, and as members of a care community. Research that measures this aspect of nursing work is needed. Adapting the LMNT to include a measure of this concept would expand the understanding of the nature of nursing work in all nursing environments.
The results of this study also indicate that LTC nursing is not a less intense or scaled-down version of acute care nursing. There are distinct differences in patient outcomes that require further study (Ouslander, 1989). Specifically, research is needed to clarify the care and treatment outcomes associated with nursing actions in both acute care and LTC settings. The enormous data set provided by the heavy regulation and reporting mechanisms of the nursing home Minimum Data Set (MDS) may provide an important method for connecting outcomes to nursing actions. Further study is needed to expand the knowledge of the nature of nursing work, taking into account the variety of care goals in LTC as identified.
The LMNT can be used in longterm care nursing environments and can provide a reasonable basis for comparing the nature of nursing work with other nursing environments. However, it also can be improved based on the results of this study to capture more successfully some aspects of nursing work that currently are not a part of the measurement.
- Alexander, J.W., & Bauerschmidt, A.D., (1987). Implications for nursing administration of the relationship of technology and structure to quality of care. Nursing Administration Quarterly, 11(4), 1-10.
- Armstrong, S., Krahenbuhl, C, Muellenbach, R., & Savage, J. (1984). Effects of hierarchical position on perceptions of nursing technology. Unpublished master's thesis, University of Wisconsin-Madison.
- Evans, L. K. (1996). Knowing the patient: The route to individualized care. Journal of Gerontological Nursing, 22(3), 15-19, 47-53.
- Knodel, J. (1993). The design and analysis of focus group studies. In D. L. Morgan (Ed.), Successful focus groups. Newbury Park, CA: Sage.
- Leatt, P., & Schneck, R. (1981). Nursing subunit technology: A replication. Administrative Science Quarterly, 26, 225-236.
- Leatt, P., & Schneck, R. (1982a). Technology, size, environment, and structure in nursing subunits. Organizational Studies, 3(3), 221-242.
- Leatt, P., & Schneck, R. (1982b). Work environments of different types of nursing subunits. Journal of Advanced Nursing, 7, 581-594.
- Leatt, P., & Schneck, R. (1984). Criteria for grouping nursing subunits in hospitals. Academy of Management Journal, 27(1), 150-165.
- Mark, B.A., & Hagenmueller, A.C. (1994). Technological and environmental characteristics of intensive care units. Journal of Nursing Administration, 24(4 Suppl.), 6571.
- Mezey, M.D., & Scanlon, W. (1988). Registered nurses in nursing homes. In Secretary's commission on nursing: Support studies and background information: Vol. 2. Washington, DC: United States Department of Health and Human Services.
- Mitchell, P., Armstrong, S., Forshee, T, & Lentz, M. (1989). AACN demonstration project: Profile of excellence in critical care nursing. Heart & Lung, 18, 219-237.
- Ouslander, J.G. (1989). Medical care in the nursing home. Journal of the American Medical Association, 262, 2582-2590.
- Overton, P., Schneck, R., & Hazlett, C. (1977). An empirical study of technology of nursing subunits. Administration Science Quarterly, 22, 203-219.
- Perro w, C. (1967). A framework for comparative analysis of organizations. American Sociological Review, 32, 194-208.
- Radwin, L.E. (1995). Knowing the patient: A process model for individualized interventions. Nursing Research, 44, 364-370.
- Radwin, L.E. (1996). "Knowing the patient:" A review of research on an emerging concept. Journal of Advanced Nursing, 23, 1142-1146.
- Swanson, K.M. (1993). Nursing as informed caring for the well-being of others. Image, 25, 352-357.
- Tanner, CA., Benner, P., Chesla, C, & Gordon, D. (1993). The phenomenology of knowing the patient. Image, 25, 273280.
- Verran, J.A., & Reíd, PJ. (1987). Replicated testing of the nursing technology model. Nursing Research, 36, 190-194.
RELIABILITY COEFFICIENTS FOR LEATT MEASURE OF NURSING TECHNOLOGY (LMNT) SUBSCALES
FACTOR LOADINGS FOR LEATT MEASURE OF NURSING TECHNOLOGY SUBSCALES