Nurse aides employed in long-term care are responsible for 80% to 90% of all direct patient care. These employees, however, are often the least well trained caregivers in the facility.1 Consequently, nurse aides are often inadequately prepared to meet many of the complex needs of the residents they care for, especially the mental health needs of the elderly.2 This is a problem of some magnitude, particularly because nursing homes have become the major receiving site for mentally ill elderly.3 Previous research has suggested that behavior management may be one effective means of dealing with a range of problem behaviors among institutionalized elders,48 although it is not frequently used9 or taught to nurse aides.10 Although professional nursing journals regularly provide theoretical and practical information about mental health care for the elderly, nursing assistants themselves have little access to this information. It is not prepared directly for them as readers nor presented explicitly as guidelines for teaching and supervising aides.
We now see an increasing emphasis on preservice and inservice training to ensure the caregiving skills of nurse aides as mandated under the Omnibus Budget Reconciliation Act of 1987. Generally, it is not clear to what extent nurse aides have been involved in the very few studies of training that include paraprofessionals.11,12 One exception has been the success with nursing assistants reported by Chartock and associates.13 Thus, the means with which nursing assistants could be trained to use behavior management strategies with elderly nursing home residents deserve more careful attention.
This study is one part of a larger project that focused specifically on the work of nurse aides. The researchers developed and implemented a fivesession skills training program designed to increase nurse aides' knowledge and skill in the use of behavior management techniques with mental health problems presented by the institutionalized elderly. This training program is specifically aimed at three mental health problems that have previously been documented as the most problematic and frequent among nursing home elders: disorientation, depression, and agitation.3
Sites and Sample
Four nursing homes in western Pennsylvania were selected from a larger pool and matched on a set of organizational variables, including size (approximately 120 beds), non-unionized status, mixed levels of care, and comparable staff levels of job motivation and satisfaction.14,15 The analyses reported here focus on two sites where training was conducted. Each home employed 40 to 60 nursing assistants. The average nursing assistant in these analyses was a 35-year-old woman with a high-school diploma, who had been working at the facility for approximately 3 years.
All nursing assistants were encouraged to attend a series of five mental health training classes. Although attendance was considered mandatory in both sites, it was more closely supervised at Site 2. Of the nursing assistants at Site 1 , 58% (n = 29) attended at least four of the five sessions offered, whereas 67% (n = 48) of the nursing assistants at Site 2 attended at least four sessions.
Five 1 ½-hour sessions were provided at 1 -month intervals in each nursing home. Each session was presented four or five times over 2 days to accommodate staff on all shifts. In addition, the trainers reviewed the materials with supervisory staff each month to encourage administrative familiarity and support for the use of the skills taught. Two trainers, both registered nurses with advanced degrees and experience in small group training, conducted the classes. Instruction followed a standardized protocol supplemented by notebooks, overhead projections, case examples, and interactive discussions; all materials were designed by the trainers.
The topics of successive sessions were basic steps of behavior management and the application of steps to disoriented, depressed, and agitated behavior. The fifth session provided a review and application of principles in everyday life. Three consistent elements were included in each session: exploration of elementary behavioral concepts and steps, ie, the building blocks or ABC's of Behavior Management (Figure); demonstration of concrete ways to identify and change the triggers and reinforcers of problem behaviors displayed by residents; and practice with interactive exercises applying behavioral management steps to familiar cases.
THE ABC'S OF BEHAVIOR MANAGEMENT
Repetition of common strategies with slightly different emphasis was intended to encourage learning retention and generalization. Following each class, nursing assistants were encouraged to read the notebooks they had received and to do structured onthe-job homework exercises. The notebooks reviewed concept definitions and provided puzzles and case study exercises. Participation - both in class and by doing homework - was rewarded by the use of token incentives. Tokens were then entered into a raffle drawn at the end of the program.
At the beginning of the first and fifth sessions, participants completed the same 21 -item test of basic and applied knowledge. They completed a similar test after the first session and a detailed program evaluation form after the final review session. Thus, comparisons of performance on similar tests after one session and the same test after four related sessions were obtained.
The tests of basic and applied knowledge were designed for the present study. They contained items pertaining to behavior management as well as the specific problem behaviors being studied. Each test contained approximately equal numbers of true-false and multiple choice items. The true-false items focused more heavily on knowledge of normal behavior in aging, whereas the multiple choice items questioned the details and management strategies applicable to one or two briefly presented nursing home vignettes.
Item analyses using data from two pilot sites were used to identify unclear items; the final iteration of the tests included 42 items, some original, some reworded, and some new items. No prior psychometric data were therefore available for the use of these tests. Cronbach's alphas for Test 1 were 0.44 at first administration and 0.54 on second administration 4 months later. The alpha for Test 2 was 0.63. Cronbach's alpha expresses the mean correlation of each item with the test total and is, therefore, an estimate of moderate internal consistency of these measures.
The program evaluation form contained three sections: items assessing the appeal, helpfulness, and interest of the modules; a self-report of their completion of optional between-class reading and writing assignments; and an eight-item, Likert-type self-report scale, asking the nurse aides to rate the frequency with which they applied behavior management principles to residents before and after the training program. Specific dimensions assessed in this section included their tendency to look for environmental triggers of problem behaviors, interpret problem behaviors as learned, prompt or teach new behaviors to residents, and reinforce desirable resident behaviors. Each performance dimension ranged from 1 (never) to 10 (always). Scores across the four dimensions were summed, and thus the past performance scores were compared with the summed current performance scores.
Overall Knowledge Improvement
Overall knowledge improvement was assessed by comparing individuals' performance on the pretest with their performance after one session and after the four core sessions. There was significant improvement within the first session with a mean change of 12.52% (SD= 12.68%, r(91) = 9.47, P<.0001). After three additional sessions of supportive material, the mean knowledge increase remained 10.32% (SD= 13.16%, /(67) = 6.47, P<.0001).
Training Gains Versus Self-Reported Caregiving Behavior
The mean past caregiving performance score was 21.5 (SD =7. 27), reflecting infrequent to occasional use of the skills, whereas the mean current caregiving performance score was 29.7 (SD = 5.72), reflecting more than occasional or moderate use of the skills. This is a significant increase (r(73) = 8.66, P<. 0001).
The correlations of these perceived caregiving performance scores with actual test scores were examined. The Table displays these bivariate correlation coefficients. The small but generally significant correlations indicate that strong test performance consistently is associated with recall of poor behavioral skills use prior to training and more frequent skills use following training. These correlations occurred in the absence of any feedback to the participants about their test performance.
The results of this investigation indicate a significant improvement in caregiver knowledge as a function of the training program implemented. This improvement suggests that nurse aides can acquire new knowledge relevant to their care of elderly residents.
There was significant improvement within the first session that was sustained across later sessions. Although gains in the first session are modestly larger than the final gains, they were measured on similar but not standardized test forms. Thus, the clearest indicator of knowledge gain may be found in the significant overall training gain when the same test was given at the beginning of the program and at the completion of the core program. This comparison controls for differences in test difficulty and reliability.
The initial psychometric evaluation of the measures used suggested both strengths and weaknesses to the measurement approach. Although the internal consistency of the measures can be described as only moderate, the absence of high internal consistencies is less disturbing if one hypothesizes that behavioral management knowledge is indeed multidimensional and composed of several diverse cognitive steps and activities.
It is interesting to note that the internal consistency of this measure increased from 0.44 to 0.54, which suggests that the training may have improved the interconnection of the relevant concepts in the minds of the nurse aides. The higher internal consistency of the measure at its second administration, therefore, may suggest that the training had taught the subjects a more coherent pattern of thinking (and perhaps of responding) to resident behavior problems. Of course, some of the performance gain may be attributable to retest effects (performance improvement as a function of practice) and decreased situational anxiety (after repeated positive interactions with the trainers). Thus, this test performance improvement may be a beneficial consequence of just being involved in relevant training sessions. In this era of competency testing and mandatory training for nursing assistants, training that helps to reduce employee resistance to ongoing education and performance testing may have its own value.
CORRELATION OF SELF-RATED BEHAVIORAL CAREGIVING PERFORMANCE WITH TRAINING TEST PERFORMANCES*
Perhaps the most promising result obtained was the finding of a positive relationship between test performance and self-assessed current caregiving performance. For this study, the only indicator of caregiving performance used was a short self-rating inventory, administered retrospectively. Thus, important cautions must be exercised in interpreting the significant gain (eg, social desirability response set, high motivation levels associated with an enjoyment of the training program). Nevertheless, to the extent that these performance ratings were modestly yet significantly correlated with test scores, these results serve as a preliminary indication that nursing assistants have altered their behavioral performance as a function of training.
The direction of the correlations obtained suggests an additional trend. All of the correlations between test performance and ratings of past performance were negative, whereas the correlations between test performance and ratings of current performance were positive. This suggests that what may have been affected most by the training was caregiver's awareness of behavioral opportunities and their own actions. Those who scored highest on the knowledge measures (ie, by extrapolation, those who had learned the most about behavior management) were most aware of the degree to which their pretraining caregiving performance was lacking in behavior management strategies, and were most likely to rate their pretraining behavior management performance as low. Similarly, those who scored highest on the knowledge measures also appeared to be most likely to integrate behavior management strategies into their current caregiving.
These interpretations provide the first review of this training effort. Future analyses using additional data, including control groups, will examine the relationship between other indicators of knowledge gain and of job performance (eg, supervisor ratings, experimentally blind behavioral observations). Although these additional studies are warranted to clarify the full efficacy of this training program, it is reassuring to find such significant improvement among nursing assistants in the domain of mental health skills.
In summary, inservice education for nursing assistants, beyond basic safe practice requirements, has not typically been a high priority in nursing home settings;11 it is costly to make an extensive commitment to it. The current mandate for certified paraprofessionals, however, will be followed by demands for more informed, more "responsive' '6 care by these bedside attendants. Mental health skills involving careful observations, judgments, and communications will be valuable at every level of the caregiving staff, but particularly by the aide on the spot.
It appears that any practically useful training program for nursing home aides must be both brief, relevant, and in manageable units. The behavioral approaches provided here in brief interactive sessions at monthly or more frequent intervals are immediately applicable, measurably mastered, and viewed as useful by these caregivers. Although additional follow-up to support and supervise the practice of these skills must be done by the licensed staff, particularly the staff development personnel, such efforts require further demonstration and study.
The success of this training suggests the need for the professionally prepared registered nurses and geriatrie nurse practitioners in these settings to consider the practical usefulness and availability of behavioral strategies in reducing disruptive behavior. It is the professional staff on site who can now offer routine training in and support for behavioral options to the legions of aides who most immediately can make use of such ABC's.
- 1. Heine CA. Burnout among nursing home personnel. Journal of Gerontological Nursing. 1986: 12(31:14-18.
- 2. Canar MJ. Johnson JC. An employee learning needs assessment concerning mental health needs of residents in a long-term care setting. Journal of Continuing Education in Nursing. 1986; 17(1 ):5-11.
- 3. Smyer MA, Conn MD. Brannon D. Mental Health Consultation in Nursing Homes. New York: New York University Press; 1988.
- 4. Patterson RL, Jackson GM. Behavior modification with the elderly. In: Hersen M, Eisler RM, Miller PM, eds. Progress in Behavior Modification, Vol 9. New York: Academic Press; 1980:205-239.
- 5. Burgio LD, Burgio KL. Institutional staff training and management: A review of the literature and a model for geriatric long-term care facilities. Int J Aging Hum Dev. 1990; 30:1287-1302.
- 6. Hussian RA, Davis RL. Responsive Care: Behavioral Interventions with Elderly Persons. Champaign, 111: Research Press; 1985.
- 7. Thompson LW, Gallagher D, Breckenridge JS. Comparative effectiveness of psychotherapies for depressed elders. J Consult Clin Psychol. 1987,55:385-390.
- 8. Baltes M, Reisenzein, R. The social world in long-term care institutions: Psychosocial control towards dependency? In: Baltes MM, Baltes PB, eds. The Psychology of Control and Aging. Hillsdale, NJ: Lawrence Erlbaum Associates; 1986.
- 9. Guy DW, Morice HO. A comparative analysis of behavior management in the nursing home. Clinical Gerontologist. 1985; 4(2):11-17.
- 10. Wisocki PA. Behavioral approaches to gerontology. In: Hersen M, Eisler RM, Miller PM, eds. Progress in Behavior Modification, Vol 6. New York: Academic Press; 1984:121-157.
- 11. Priddy JM, Teitelman JL, Kivlighan DM, Fuhrmann BS. Overcoming learned helplessness in elderly clients: Skills training for service providers. Educational Gerontology. 1982;8(5):507-518.
- 12. Thompson LW, Gallagher D, Nies G, Epstein D. Evaluation of the effectiveness of professionals and non-professionals as instructors of "Coping with Depression" classes for elders. Gerontologist. 1983; 23:390-3%.
- 13. Charlock P, Nevins A, Rzetelny H, Gilberto P. A mental health training program in nursing homes. Gerontologist. 1988; 28:503507.
- 14. Brannon D. Nursing Home Caregiver Intervention Project: Preliminary findings. Presented at the 41st Annual Meetings of the Gerontological Society of America; November 1988; San Francisco.
- 15. Brannon D, Smyer MA, Conn MD, Borchardt L, Landry JA, Jay GM, et al. A job diagnostic survey of nursing home caregivers: Implications for job redesign. Gerontologist. 1988;28:246-252.
CORRELATION OF SELF-RATED BEHAVIORAL CAREGIVING PERFORMANCE WITH TRAINING TEST PERFORMANCES*