Journal of Gerontological Nursing

The articles prior to January 2011 are part of the back file collection and are not available with a current paid subscription. To access the article, you may purchase it or purchase the complete back file collection here

INTERVENTIONS FOR Decreasing Agitation Behaviors in Persons with Dementia

Mariah Snyder, PhD, RN, FAAN; Ellen C Egan, PHD, RN, FAAN; Kenneth R Burns, PHD, RN

Abstract

Agitation behaviors in persons with dementia are a major problem for caregivers. Hall and Buckwalter (1987) hypothesized that an increase in agitation behaviors occur when persons experience high levels of stress, The purpose of this pilot study was to explore the efficacy of hand massage, therapeutic touch, and presence (control group) in producing relaxation and decreasing agitation behaviors in persons with dementia. An experimental crossover design was used to study the effects of these interventions in 17 residents on one AlzheimeKs Care Unit. Hand massage and therapeutic touch were administered once a day in late afternoon for 10 days while presence was administered for 5 days. Five days of observation were done before and after the administration of each intervention. Significant differences were found in the level of relaxation from pre- to post- intervention with the use of hand massage and therapeutic touch. However, no decrease in agitation behaviors was observed.

Abstract

Agitation behaviors in persons with dementia are a major problem for caregivers. Hall and Buckwalter (1987) hypothesized that an increase in agitation behaviors occur when persons experience high levels of stress, The purpose of this pilot study was to explore the efficacy of hand massage, therapeutic touch, and presence (control group) in producing relaxation and decreasing agitation behaviors in persons with dementia. An experimental crossover design was used to study the effects of these interventions in 17 residents on one AlzheimeKs Care Unit. Hand massage and therapeutic touch were administered once a day in late afternoon for 10 days while presence was administered for 5 days. Five days of observation were done before and after the administration of each intervention. Significant differences were found in the level of relaxation from pre- to post- intervention with the use of hand massage and therapeutic touch. However, no decrease in agitation behaviors was observed.

The presence of agitation behaviors in a large percentage of persons with dementia has been identified as a major care problem (Argyle, Jestice, & Brook, 1985; Reisberg, Borenstein, Franssen, Shulman, Steinberg, & Ferris; 1986). The reported prevalence rates of agitation behaviors range from 24.4% (Jackson, Drugovich, Fretwell, Spector, Sternberg, & Rosenstein, 1989) to 93% (Cohen-Mansfield, Marx, & Rosenthal, 1989). The vast differences in the rates found in studies may be the result of settings used, methods employed in collecting data, or in the definitions used for these behaviors.

Various hypotheses for the occurrence of agitation behaviors have been proposed (Beck & Heacock, 1988; Birchmore & Clague, 1983). In the progressively lowered stress threshold model (PLST) advanced by Hall and Buckwalter (1987), increased levels of stress have been proposed as a cause of agitation behaviors. Anxious behaviors are manifested when persons experience increased stress. When the number of Stressors or the intensity of Stressors increases, dysfunctional (agitation) behaviors occur. Therefore, interventions or environments that would lessen stress should decrease the occurrence of agitation behaviors.

Numerous anecdotal reports in the literature suggest interventions that may decrease agitation behaviors in persons with dementia (Beck & Heacock, 1988; Schwab, Rader, & Doan, 1985; Robinson, Spencer, & White, 1989; Wondolowski & Chartock, 1989). Kaseman and Young (1988) emphasized the importance of assessing the effectiveness of interventions such as diversion activities, music, and walking, to reduce stress in persons with dementia.

Gwyther (1986) proposed using touch to reduce stress in persons with cognitive impairments. Massage and therapeutic touch are two forms of touch. Most of the studies about the effects of massage used a 3-minute slow stroke massage of the back. In a shady on the use of back massage with institutionalized elders, Fraser and Kerr (1993) found a significant difference between the anxiety scores of subjects receiving massage with conversation and subjects receiving no treatment. Fakouri and Jones (1987) reported significant changes in heart rate, blood pressure, and temperature following the administration of slow stroke back massage to elderly nursing home residents. Therapeutic touch is a process by which energy is transmitted from one person to another for the purpose of potentiating the healing process of one who is ill or anxious (Egan, 1992). In two studies conducted to determine the effect of therapeutic touch on reducing anxiety in adults, subjects receiving therapeutic touch had a significantly greater reduction in anxiety scores than did persons in the control group (Heidt, 1981; Quinn, 1984). No studies were found in which the efficacy of massage or therapeutic touch in reducing anxiety in persons with dementia had been examined.

The regulations contained in the Omnibus Budget Reconciliation Act (OBRA) of 1987 make it imperative that health care workers explore the use of alternatives to physical and pharmacological restraints for the management of agitation behaviors. The purpose of this pilot study was to explore the efficacy of hand massage and therapeutic touch in producing relaxation and decreasing agitation behaviors in persons with dementia; presence was used as a control condition. The specific hypotheses studied were:

1. Subjects will have a greater level relaxation and fewer anxious befollowing the administration hand massage and therapeutic than prior to administration of the interventions.

2. Subjects will have a decrease in agitation behaviors followthe administration of these interventions.

3. Greater changes in the level of will occur with the use of massage or therapeutic touch with the use of presence.

METHODS

An experimental crossover design used with each subject serving his or her own control. Six groups three subjects were formed with being collected on one group at a time. The order of administration of the three interventions was alternated so that Group 1 subjects received hand massage as the first intervention. Group 2 subjects received therapeutic touch first, while subjects in Group 3 received presence first. The Figure depicts the time frame and order of observations and interventions for subjects in Group 1.

Sample

Subjects were recruited on a 31bed Alzheimer Care Unit. A letter describing the study and a consent form were sent under the auspices of the agency to the families of all residents on the unit; all families agreed to have their family member participate. After consultation with the nursing staff, the 18 residents who manifested the highest incidence of agitation behaviors were selected for the study. One subject moved to another setting before all three interventions had been administered and was dropped from the study. Five subjects were male and 12 were female. Subjects ranged in age from 66 to 90 years of age with a mean age of 77.7 years. All but four of the subjects were on psychotropic medications. However, no differences in level of relaxation, anxious behaviors, or agitation behaviors were found between subjects on psychotropic medications and those not receiving them (Wu, 1993).

Table

Figure. Time frame and order of observations and interventions.

Figure. Time frame and order of observations and interventions.

Table

TABLE 1Protocol for Hand Massage

TABLE 1

Protocol for Hand Massage

Table

TABLE 2Protocol for Therapeutic Touch

TABLE 2

Protocol for Therapeutic Touch

Instrumentation

The subject's level of dementia was determined from scores on the Haycox Rating Scale (Haycox, 1984). This scale consists of eight behaviors (language-conversation, social interaction, attention-awareness, spatial orientation, motor coordination, bowel-bladder, and dressing-grooming) that are rated on a six point scale. The possible range of scores is from O to 48 with higher scores indicating more advanced dementia. Haycox reported high interrater (r = .90) and intrarater (r = .97) reliability for the scale. A correlation of r = .61 with the Blessed Cognitive Scale was reported. Haycox scores in this study ranged from 12 to 43 with a mean of 23.1.

Two indices were used for measuring the level of relaxation: behaviors and pulse rate. The Relaxation Checklist developed by Luiselli and colleagues (1982) lists nine areas to observe: forehead, eyes, neck, head, arms, hands, legs, feet, and breathing. Each behavior is rated on a scale of 1 to 5 with 1 representing the greatest level of relaxation. Luiselli and colleagues reported an interrater reliability of r = .82 for the Checklist. The pulse rate was obtained via manual palpation of the radial artery. Relaxation indices were measured by the researcher immediately before and after administering the intervention.

Three anxious behaviors that were specific for each subject were identified. Tapping fingers, rocking back and forth, flicking a hand through hair, and swinging one's legs are examples of anxious behaviors. The frequency of these behaviors was rated on a five point scale with 5 indicating that the behavior occurred throughout the intervention while 1 indicated that the behavior did not occur. Ratings were made by the researcher administering the intervention immediately before and immediately after intervening.

Agency staff, over a period of time, had identified agitation behaviors that were specific for each subject; these were termed targeted agitation behaviors. Examples of agitation behaviors included hitting, striking out, grabbing, pacing, verbally abusing others, and crying. Behaviors specific for each subject were listed on the observation form. The frequency of occurrence of these behaviors were recorded by the agency staff in 3-hour increments from 6 AM to 9 PM during the 45 days of data collection. Instructional sessions about use of the form, parameters of specific behaviors, and an overview of the study were provided for the agency staff.

Interventions

Hand massage and therapeutic touch were deemed to be appropriate stress reduction interventions to use with this population as both were of short duration, noninvasive, and had produced a relaxation response in other populations. The protocols for these interventions are found in Tables 1 and 2. Presence was included as a control intervention so that the effects of "just being with the person" could be differentiated from the effects obtained from the use of the other two interventions. Members of the research team, all of whom were registered nurses, implemented the intervention protocols for 10 minutes in the late afternoon for 10 days. Presence was administered for 10 minutes in the late afternoon for 5 days. During presence, the researcher attempted to maintain the subject's attention by speaking to or touching. Only brief touching with no stroking was done. (Because of the extended length of time for data collection, a decision was made to only administer presence for 5 days.)

Table

TABLE 3Comparison of Chongos fn Relaxation Behavior Scores Before and After Administration of Interventions

TABLE 3

Comparison of Chongos fn Relaxation Behavior Scores Before and After Administration of Interventions

Prior to data collection, sessions to establish interrater reliability for administration of the instruments and protocols were held for the research team. Discussions during the weekly staff meetings were also used to clarify ratings and adaptations in protocols for specific subjects.

Analysis

Each subject's pre- and postintervention scores for level of relaxation, frequency of anxious behaviors, pulse rates, and agitation behaviors were averaged for a 5-day period rather than analyzing data for each day of the study. Repeated measures analysis of variance (ANOVA) and post hoc t-tests were used to analyze changes in the outcome measures.

RESULTS

The repeated measures ANOVA results revealed no significant differences in the pre-intervention relaxation behavior scores, pulse rates, or anxious behaviors. The repeated measures ANOVA in which the amount of change from pre- to post-intervention that occurred in relaxation behaviors (F[5, 23] =5.23, p = .002), pulse rate (F[4, 40] = 3.75, p = .001), and anxious behaviors (F[4, 24] = 4.95, p = .005) indicated an intervention effect. Therefore, post hoc comparisons were made. Because of missing data due to subjects' illness or reluctance to participate in sessions, the degrees of freedom vary.

Table

TABLE 4Comparison of Changes in Pulse Rates Before and After Administration of Interventions

TABLE 4

Comparison of Changes in Pulse Rates Before and After Administration of Interventions

Following the administration of hand massage and therapeutic touch, the indices of relaxation (relaxation behaviors and pulse) reflected an increased level of relaxation as compared to pre-intervention. These changes were statistically significant for both the first and second 5 days of these interventions. No significant changes in the pre- and post-intervention relaxation behaviors or pulse rates were found with presence. These findings are shown in Tables 3 and 4.

Significant differences in the preto post-intervention scores for anxious behaviors were found during both 5-day periods of hand massage (f[13] =4.16, p = .001) (?[12] = 6.32, ? = .001) and during the second 5 days of therapeutic touch (f [15] =2.45, p = .027). No significant differences were found in the anxious behavior change scores with the use of presence. Based on the findings for changes in the level of relaxation (relaxation behaviors and pulse rate) and anxious behaviors, support exists for the first research hypothesis.

No significant differences in the occurrence of targeted agitation behaviors were found from before, during, or after the administration of any of the interventions. Differences in behaviors during various timeframes, particularly late afternoon, were examined. No significant differences among the frequencies of targeted agitation behaviors across the time periods were found. Great variation in the occurrence of agitation behaviors existed among subjects and across time in the same subject. The findings in this study do not provide support for the second hypothesis.

As shown in Tables 3 and 4, hand massage and therapeutic touch produced significant changes in the indices used to determine the level of relaxation. No significant changes were found when presence was used. Thus, support is provided for the third hypothesis.

DISCUSSION

Hand massage produced a greater level of relaxation than did therapeutic touch. Both were more effective than presence. The use of presence to calm non-cognitively impaired individuals had been suggested in the literature (Moch & Schaefer, 1992). Trying to gain and maintain the attention of persons with dementia may be appraised by the subject as being an irritant and thus interfere with relaxation occurring.

One plausible reason for decreases in the frequency of targeted agitation behaviors not being found is that the relaxation response obtained may have been of short duration. Observations of the level of relaxation were only made immediately following administration of the interventions. The precise length of time that the relaxation response persisted is unknown. Subsequent studies that employ measures to determine the length of time the relaxation response persists need to be undertaken.

Producing a relaxation response of short duration does not, however, negate the possible utility of use of hand massage and therapeutic touch. Administration of these interventions, particularly hand massage, by nursing staff prior to initiating cares that tend to precipitate agitation behaviors may produce a relaxation effect of sufficient length to diminish the frequency or intensity of agitation behaviors during a care activity.

The degree to which the researchers were able to implement the intervention protocol and the unit environment were two factors that the investigators noted during the study. The investigators believed that these factors may have had an impact on the efficacy of the interventions. Incidents, such as subjects withdrawing their hands before the end of the ??-minute massage or a subject suddenly arising to walk, were encountered. Environmental factors included someone walking into the subject's room during administration of the intervention, other residents screaming, or high levels of noise on the unit. Scales for protocol implementation and environment were developed (Egan, Snyder, & Burns, 1992). Because of the unpredictability of subjects in this population, the investigators believe that gathering data on the extent to which a protocol can be implemented and the presence of adverse environmental variables are important in assessing the efficacy of interventions in persons with dementia.

The researcher administering the interventions made the observations on relaxation and anxious behaviors pre- and post-intervention. It is recognized that this procedure may have resulted in researcher bias. Various other methods for obtaining ratings were explored, but characteristics of this population presented problems for introducing a second person. Having another person enter the room prior to intervening and immediately after the intervention is completed to make observations may increase the agitation of the subject. Videotaping of pre- and post-intervention behaviors was considered. However, trying to get a subject to a particular room for videotaping was problematic as this could increase the subject's agitation. The use of mechanical devices to obtain objective data is also difficult. Attaching a subject to a machine may increase the level of agitation. Also, a subject may suddenly decide to walk, thus causing detachment from the machine. Identifying methods that will provide valid and reliable indices of the relaxation response is indeed a challenge in this population.

This pilot study is only a beginning effort in establishing a scientific basis for nursing interventions that can be used in caring for persons with dementia who manifest agitation behavior. Studies using larger sample sizes are needed, especially to enable the researchers to compare the impact that confounding variables may have on the outcomes. Also, studies exploring the efficacy of other interventions need to be conducted.

REFERENCES

  • Argyle, M., Jestice, S., & Brook, C.P.B. (1985). Psychogeriatric patients: Their supporters' problems. Age and Aging, 14, 355-360.
  • Beck, C. & Heacock, P. (1988). Nursing interventions for patients with Alzheimer's disease. Nursing Clinics of North America, 23, 95-124.
  • Birchmore, T. & Clague, S. (1983). A behavioral approach to reduce shouting. Nursing Times, 79,(16), 37-39.
  • Cohen-Mansfield, J., Marx, M., & Rosenthal, A.S. (1989). A description of agitation in a nursing home. Journal of Gerontology: Medical Sciences, 44,(3), M77-84.
  • Egan, E.G. (1992). Therapeutic touch. In M. Snyder (Ed.), Independent nursing interventions (pp. 173-183). Albany, NY: Nyelmar Publishers, Inc.
  • Egan, E.C., Snyder, M., & Bums, K.R. (1992). Intervention studies in nursing: Is the effect due to the independent variable? Nursing Outlook, 40(4), 187-190.
  • Fakouri, C. & Jones, P (1987). Relaxation Rx: Slow stroke back rub. Journal of Gerontological Nursing, 13(2), 32-35.
  • Fraser, J. & Kerr, J.R. (1993). Psychophysiological effects of back massage on elderly institutionalized patients, Journai of Advanced Nursing, 18, 238-245.

Gwyther, L.P. (1986). Treating behaviors as a symptom of illness. Provider, 12(5), 18, 20-21.

  • Hall, R.R. & Buckwalter, K.C. (1987). Progressively lowered stress threshold: A conceptual model for care of adults with AIzheimer's disease. Archives of Psychiatric Nursing, 1 (6), 399-406.
  • Haycox, J.A. (1984). A simple, reliable clinical behavior scale for assessing dementia patients. Journal of Clinical Psychiatry, 45, 23-24.
  • Heidt, P. (1981). Effect of therapeutic touch on anxiety level of hospitalized patients. Nursing Research, 30, 32-37.
  • Jackson, M.E., Drugovich, M.L., Fretwell, M.D., Spector, W.D., Sternberg, J-, & Rosenstein, R.B. (1989). Prevalence and correlates of disruptive behaviors in the nursing home. Journal of Aging and Health, 1, 349-369.
  • Kaseman, D.F. & Young, S.H. (1988). Stress: An added incapadtator. Geriatric Nursing, 9, 274-277.
  • Luiselli, J.K., Steinman, D.L., Marholin, D., & Steinman, W.M. (1982). Evaluation of progressive muscle relaxation with conductproblem learning-disabled children. CAiW Behavior Therapy, 3(2/3), 41-55.
  • Moch, S.D. & Schaefer, C.C. (1992). Presence. In M. Snyder (Ed.) Independent nursing interventions (pp. 238-243). Albany, NY: Delmar Publishers, Inc.
  • Quinn, J-F. (1984). Therapeutic touch as energy exchange: Testing the theory. Advances in Nursing Science, 6(2), 42-49.
  • Reisberg, B., Borenstein, J., Franssen, E., Shulman, E., Steinberg, G-, & Ferris, S.H. (1986). Remediable behavioral symptomatology in Alzheimer's disease. Hospital and Community Psychiatry, 37, 1199-1201.
  • Robinson, A., Spencer, B., & White, L. (1989). Understanding difficult behaviors. Ypsilanti, Ml: Eastern Michigan University, Geriatric Education Center.
  • Schwab, M., Rader, J., & Doan, J. (1985). Relieving the anxiety and fear in dementia. Journal of Gerontological Nursing, 11(5), 8-11, 14-15.
  • Wondolowski, C., & Chartock, P. (1989). Guidelines for nursing management in Alzheimer's care. Journal of Advanced Medical Surgical Nursing, 1, 76-87.
  • Wu, EY. (1993). Comparison of intervention outcomes in subjects receiving psychotropic medications and subjects not receiving these medications. Unpublished master's plan B project, University of Minnesota, Minneapolis, MN.

Figure. Time frame and order of observations and interventions.

TABLE 1

Protocol for Hand Massage

TABLE 2

Protocol for Therapeutic Touch

TABLE 3

Comparison of Chongos fn Relaxation Behavior Scores Before and After Administration of Interventions

TABLE 4

Comparison of Changes in Pulse Rates Before and After Administration of Interventions

10.3928/0098-9134-19950701-11

Sign up to receive

Journal E-contents