Assaultive behavior by psychogeriatric patients represents a substantial problem for caregivers. Ryden, Bossenmaier, and McLachlan (Chou, Kaas, & Richie, 1996) reported that 50% of older adults with cognitive impairment were physically aggressive within a 7-day period of observation. Research by Swearer, Drachman, O 'Donnei!, and Mitchell (1988) revealed that 21% of the patients at one dementia care facility were assaultive. Colenda and Hamer (1991) reported that approximately 20% of elderly individuals at a state psychogeriatric hospital were assaultive. In a study of assaults in one provincial psychiatric hospital, Cooper and Mendonca (1991) found that when assault rates according to diagnoses were considered in proportion to their occurrence, developmental delay and dementia were more often associated with assaultiveness than schizophrenia.
Responding to assaultive behavior is problematic for caregivers of psychogeriatric patients. One problem involves a misunderstanding in differentiating aggressive and resistive behavior (Gibson, 1997). Agitation has also been difficult to define, which makes responding to agitation, especially assaultiveness, a difficult enterprise (Roper, Shapira, & Chang, 1991). Yet the link between the two may be important - agitation has been hypothesized to be a harbinger of assault (Littrell & Littrell, 1998).
Different caregiver approaches to addressing patient behavior are cited as a factor in determining violent outcomes. Ray and Subich (1998) found that caregiver traits of external locus of control, anxiety, and authoritarianism were associated with increased frequency of assault. Cooper and Mendonca (1991) considered the possibility that some staff, particularly those with an approach that is "subtly adversarial," may become repeat victims of attack (p. 165). In a study involving registered nurses (RNs) and nursing aides (NAs), Haber, FaganPryor, and Allen (1997) observed that RNs and NAs did not agree on the use of physical intervention in response to aggressive behavior. This is important because it is showed that agreement among caregivers in aggression management leads to more effective patient care (Haber et al., 1997).
The purpose of this study was to examine the effect of implementation of a theory-based model, by educational in-service, on the occurrence of physical assault in a geriatric psychiatry setting. Whereas a number of models for managing aggressive behavior have been described, few have been implemented in a geriatric psychiatry setting and evaluated in terms of patient outcomes.
The study hypothesis was that introduction of a model might serve to decrease the occurrence of physical assault among psychogeriatric inpatients, and do so without having to resort to changes in the use of physical or chemical restraint. Implementation of the theory-based model for the management of assaultive behavior was the independent variable and the occurrence of physical assault was the dependent variable. A physical assault was defined as, "intentional personal physical violence" by a patient toward another person "for whatever purpose, irrespective of provocation" (Cooper & Mendonca, 1991, p. 163). A theory-based model was defined as a problem-solving approach consisting of established biological and psychosocial interventions intended to reduce agitation and prevent assault.
REVIEW OF LITERATURE
Roper et al. (1991) reported on a framework for agitation management in patients with dementia. These authors observed that agitation was related to combativeness. In particular, they recognized that kicking, grabbing, and hitting represented a threat to the well-being of others. The Psychiatric Self-Care Model, based on Orem's (1995) Self-Care Framework, was offered as a method to assess agitated behavior, assist with care planning, and direct caregivers to provide appropriate interventions.
One of the key principles of Orem's (1995) framework was that individuals possess the ability to learn to care for themselves. According to Orem, one role for nursing exists in the provision of care for those who have a care deficit (i.e., cannot care for themselves). Orem specifically argued that nurses "will be required to set limits for the behavior" of others (p. 335). Thus, as Roper et al. (1991) suggest, there is an appropriate fit between efforts to manage agitation among patients with dementia and Orem's concept of nursing practice. The authors successfully identified agitated behaviors in the context of Orem's self-care categories. Moreover, the authors recommended appropriate nursing interventions for managing disturbed behaviors according to self-care categories. Although the model proposed by Roper et al. (1991) was offered as a theoretical construct for managing disturbed behavior, its practical effectiveness was not examined.
Still, the report on the model offered by Roper et al. (1991) may be seen as important for a number of reasons. First, the authors recognized that disagreement existed among caregivers related to what constituted agitation. Second, the authors said there was a need for guidelines related to the assessment of agitation. Third, the authors proposed that understanding agitation was a "pivotal" feature in the effective overall management of patients with dementia (p. 17). This early analysis of the difficulties inherent in managing agitated and disturbed behavior among patients with dementia served as an important foundation for the development of strategies for managing agitation.
Concern for the development of guidelines for the assessment of agitated behavior occurred at the same time as research intended to produce just such guidelines. Cohen-Mansfield, Marx, and Rosenthal (1989) described the use of the Cohen-Mansfield Agitation Inventory (CMAI) to identify the nature and frequency of specific behaviors in 408 nursing home residents. An important finding of their study was that agitation did not represent unrelated behaviors, rather, it represented a cluster of behaviors that co-occurred within residents.
Use of the CMAI as an instrument to gauge agitation is increasingly widespread. Jackson, Templeton, and Whyte (1999) reported on the use of the CKlAI to assess aggressive behavior in a variety of long-term care settings. In the Canadian province of Ontario, the Ministry of Health and Long Term Care (1997) Psychogeriatric Training Initiative provided instruction on the use of the assessment to representatives of 498 long-term care facilities. Miller, Snowdon, and Vaughan (1995) reported on the assessment of 2,445 residents in 46 nursing homes in Australia, using the CMAI. This study demonstrated the reliability of the CMAI and showed that its ratings correlated well with other behavior rating instruments. According to Miller et al. (1995), an instrument for the accurate assessment of aggressive and disruptive behaviors, such as the CMAI, is a prerequisite to understanding and managing those same behaviors.
The assessment and treatment of aggressive and violent behaviors received some consideration from Littrell and Littrell (1998). These authors suggested that a lack of understanding existed among caregivers related to causes of aggression and optimal treatment interventions. Among psychiatric nurses in particular, Littrell and Littrell described the lack of understanding as "alarming" (p. 18). These authors recommended the use of a violence continuum model to assist with the conceptualization of aggression. According to the continuum, agitation is perceived as less threatening than aggression. However, agitation without intervention has the potential for evolving into more serious forms of aggression and violence.
Littrell and Littrell (1998) also recommended the use of an aggressive behavior model for assessing risk factors, developing and implementing interventions, and evaluating the outcomes of interventions. No indication was provided, however, that the model had been evaluated in terms of patient outcomes. As well, the aimpression is that the model was intended for general application in psychiatric facilities and not specifically intended for use in geriatric psychiatry programs.
Chou et al. (1996) described a model for guiding nursing interventions for assaultive behavior, specifically among geriatric patients. Their model proposed the importance of discerning a baseline stage, a preassaultive stage, and an assaultive stage. In the pre-assaultive stage, interventions for decreasing agitation were specifically recommended. Appropriate psychological, pharmacological, and physical interventions were also recommended for the deescalation of aggressive behavior. The authors' recognition of assault as a dynamic process, rather than an isolated act, was important. Moreover, the impression was that these authors agreed with Littrell and Littrell (1998) that assault developed along a behavioral continuum. Finally, Chou et al. (1996) suggested that by recognizing and responding appropriately to different stages of assault, nurses could not only decrease patients' agitation, but also prevent assault.
Gilley, Wilson, Beckett, and Evans (1997) reported on the relationship between symptoms of psychosis and frequency of aggressive behavior among individuals with Alzheimer's disease. The researchers found that the presence of delusions, but not hallucinations, were predictive of aggression. Thus, in the management of aggressive behavior, a role exists in the use of antipsychotic medication.
The findings of Gilley et al. (1997) emphasize the importance of the second tier, involving pharmacological intervention, in the threelevel model described by Chou et al. (1996). Indeed, efforts to appropriately address aggressive behaviors in geriatric psychiatry settings must be multifaceted.
Approval for the study was obtained from the Research Ethics Board (REB) of Queen's University. The REB reviews and regulates research prepared for university-affiliated hospitals.
MEAN COHEN-MANSFIELD AG~TAT1ON INVENTORY (CMAI) SCORES (N=$)
The setting for this study consisted of one geriatric ward in a regional psychiatric hospital in eastern Canada. In such hospitals, geriatric psychiatry programs provide assessment, treatment, and continuing care to individuals experiencing mental health problems that are usually related to advanced age. The study ward contained 27 beds and provided continuing care for elderly, aggressive male in-patients. This was considered an appropriate ward for the study because of the chronically disturbed behavior of the patients and because of a trend of a rising number of assaults in the first half of 1999. The nursing complement on the ward consisted of 8 RNs and 12 registered practical nurses. Nursing staff members on the ward relied on individual nursing skills for both the assessment of aggression and the planning and implementation of interventions to address aggression.
The sample consisted of all of the patients on the selected ward. Repeated efforts were made to contact the substitute decision-makers for all of the 24 patients on the ward. Some substitute decision-makers declined the offer to participate; others did not return calls placed to them. Patients represented by the province's Office of the Public Guardian and Trustee were unable to participate because providing consent for research is not in the mandate of that agency.
Informed consent was eventually obtained from substitute decisionmakers for 10 patients to participate in the study. The mean age of the participants was 63.1 years (range 45 to 78). Eight of the participants had a diagnosis of dementia, one had a diagnosis of chronic schizophrenia, and one had a dual diagnosis of schizophrenia and developmental delay.
The intervention consisted of four major elements. First, participants were assessed with the CMAI (Cohen-Mansfield, 1991). Second, the perspective of the continuum of violence was applied to participants. That is, it was recognized that agitated behavior, if not addressed, could escalate into violent behavior. Third, there was frequent and regular discussion of each participant's behavior in relation to diagnosis and abilities according to the Global Deterioration Scale (GDS) (Reisberg, Ferris, DeLeon, & Crook, 1982) and Functional Assessment Staging (FAST) (Cohen-Mansfield, Werner, & Reisberg, 1995). Formally, these discussions occurred every time a participant was assessed with the CMAI. Informally, discussions occurred at any time that participants were observed to be exhibiting anxious and agitated behaviors. Fourth, evidence-based psychosocial interventions were implemented to attempt to allay agitated behaviors. All of the elements coalesced into a system, or model, for methodically planning and organizing individualized care for individuals with difficult and challenging behaviors. Because the system emphasized assessment of agitation, planning and implementation of appropriate interventions to prevent violence, and reassessment, it resembled the untried model described by Chou et al. (1996).
Initially, three formal presentations describing the study were provided for the ward staff. These were approximately 1 hour in length and were conducted in a small group format using visual aides, handouts, short presentations, and group discussion. Topics covered included:
* The importance of addressing assaultive behavior.
* The violence continuum described by Littrell and Littrell (1998).
* Assessment of agitation with the CMAI.
* The use of the model described by Chou et al. (1996) to guide interventions to prevent assault.
Both the initial educational sessions and ensuing participant CMAI assessment sessions were facilitated by an advance practice nurse of the hospital's geriatric psychiatry program.
Enrolled participants were assessed with the CMAI by nursing staff, with the assistance of the advance practice nurse. Assessment was conducted initially and then at 12-week intervals for 36 weeks. Possible psychosocial interventions accounting for the participant's individual diagnosis, level of developmental functioning according to the GDS and FAST (Groulx, 1998), and individual abilities were discussed in small group sessions. The individualized interventions were incorporated, as possible, by ward staff into each participant's care.
Figure 1. Frequency of reported assaults among participants (N= 8).
During the intervention period, any changes to the participants' regular medication profile were resisted as much as possible. This was for two reasons. First, this was in keeping with the philosophy of Chou et al. (1996). These authors emphasized that efforts should be invested in the provision of appropriate psychosocial interventions before considering chemical restraint. Second, changes in the participants' medication profile would act as a confounding variable. In other words, any changes observed in the participants' behavior could be more related to the effects of medication than to the effects of psychosocial interventions.
A highly individualized medication regime for participants, unrelated to the purposes of the study, preexisted the introduction of the study intervention. Psychotropic medication profiles varied widely for each participant. Of the eight participants enrolled for the entire study:
* Five received typical antipsychotics (perphenazine, methotrimeprazine, and haloperidol).
* Five received an antidepressant (trazodone).
* Four received an atypical antipsychotic (risperidone).
* Four received anxiolytics (lorazepam and oxazepam).
* Three received a sedative/hypnotic (chloral hydrate)
* Three received an antiparkinsonian agent (benztropine mesylate).
* One received a mood stabilizer (carbamazepine).
Two instruments were involved in the study. The first was the CMAI. This was used to gather data for baseline and post-intervention comparison of agitation. Cohen-Mansfield (1991) reported inter-rater reliability averages between .88 and .92 for the CMAI. Miller et al. (1995) reported on the concurrent validity of the CMAI with the Behavioral and Emotional Activities Manifested in Dementia (BEAM-D). Using Pearson's correlation, the researchers found a high association between existed CMAI scores and BEAM-D scores for the day, evening, and night shifts (r =.91, .79, and .92; ? < .01).
The second instrument was the standard incident report form for psychiatric hospitals. In regional psychiatric hospitals, incident report forms are completed for unusual occurrences, such as falls and assaults. The original copy of the form is filed in the affected person's casebook. A duplicate copy is forwarded through administrative channels and, eventually, entered into a database. The incident form was used to compare numbers of reported assaults during the 36 weeks prior to the intervention with numbers of reported assaults for 36 weeks after the onset of the intervention. Pearson, Wilmot, and Padi (1986) also used incident forms to analyze the occurrence of violent behavior in a psychiatric hospital.
One participant died at the onset of the intervention. Another participant was transferred from the study ward to a regional long-term care facility approximately halfway through the intervention period. Data for both participants were discarded.
The CMAI data were analyzed for the remaining eight participants. Following the work of other researchers, the CMAI behavior items were divided into three main factors (Cohen-Mansfield et al., 1989; Jackson et al., 1999; Miller et al., 1995). These included 21 items relating to aggressive behavior, physically non-aggressive behavior, and verbally agitated behavior. Following the work of Mintzer et al. (1997), mean total CMAI scores and mean CMAI factor scores were compared. In this case, the mean scores were derived from baseline and final CMAI assessments. As with Mintzer et al. (1997), a t test was used to analyze the difference between scores (Table). The differences between pre-intervention and post-intervention total CMAI scores and CMAI factor scores were not significant.
Frequency of assaults was compared under the conditions of standard interventions and experimental interventions (Figure 1). In view of the ordinal nature of the data, the Wïlcoxon signed ranks test (Munro, 2001) was used for analysis. The distribution of assaults differed significantly under the conditions of the two interventions (p < .05). Significantly fewer assaults occurred under the experimental intervention than under the standard intervention.
The results of this study suggest that a comprehensive approach to preventing aggressive behavior may be beneficial in decreasing assault occurrence among male patients in a geriatric psychiatry setting. Agitation assessment and re-assessment, recognition of the role of the agitation in the development of aggression, and the use of appropriate individualized interventions are critical elements of a systematic approach to reducing aggression.
The CMAI assessments were useful for stimulating discussion with nursing staff about the definition of agitation. This was especially helpful for arriving at some consensus about whether participants were or were not agitated according to criteria described by Cohen-Mansfield (1991). In one case, a participant who was routinely considered not agitated according to the patient assessment flow chart (a standard ward document) was considered agitated upon assessment with the CMAI. This participant was also found to have assaulted staff four times during the pre-intervention period of data collection. Realizing the severity of this participant's agitation and his propensity for assault served to emphasize the utility of the CMAI for formally defining and gauging agitation.
Consideration of individual patient diagnosis was equally important. In one case, it was considered that the nature of one participant's disorder, described in current literature as related to a high level of anxiety, really required staff to approach the participant with a inordinately high amount of caution. This participant was also considered to have provided behavioral signs of inability to cope with certain situations, such as being bathed in a whirlpool tub. The observation by one nurse that the participant responded much more appropriately when provided breakfast in bed and then a bed bath, served to change the strategy of care for that participant for the post-intervention phase of the study. Moreover, there is support for exactly this kind of approach in the literature. As Radar, Lavelle, Hoeffer, and McKenzie (1996) observed, where "maintaining cleanliness" is concerned, "individualization is the key" (p. 33).
Discussion related to participants' individual presentation, stage of dementia, and corresponding abilities was also worthwhile. The impression was that, for most staff, Groulx's (1998) description of Reisberg's (cited in Groulx) comparison of the abilities of someone in the seventh stage of dementia with the abilities of a 1- to 15-month-old infant was a revelation. The comparison of the abilities of someone in the sixth stage of dementia with the abilities of a 2- to 5-year-old also made an effect because this stage categorized the presentation of most of the patients on the ward.
Figure 2. Frequency of reported assaults among non-participants (N= 12).
Reflecting on the lost abilities of late-stage dementia also served as a catalyst for reconsidering expectations for functional competence of the patient at times of care and that patient's competence in the context of the general ward environment. This is in keeping with the "enablement perspective" advanced by Dawson, Wells, and Kline (1993, p. 1). According to that perspective, one therapeutic role for nurses exists in encouraging individuals with dementia to continue to use their remaining abilities to prevent "disuse" and "excess disability" (p. 9). Thus, an individual with relatively high competence may benefit by an environment that offers opportunities to use those abilities.
On the ward, this prompted interest in stimulating capable participants to contribute more to their own care. Promoting self-care activities, included, for example, encouraging and supporting some patients to attempt to shave, dress, and feed themselves. This also included attempts to engage participants in reading or using a newly acquired billiard table.
The enablement perspective also cautions that where the stimulation in the environment exceeds the competence of the individual with dementia, agitation may ensue. Because an attempt was made in this study to decrease assault by decreasing agitation, this aspect of the enablement perspective was of particular interest. Nurses were sensitive to the possibility that some participants were unable to cope with the level of stimulation in the environment and attempted to provide, as possible, quieter and less crowded settings for some participants.
One particularly novel approach served to not only increase the opportunity for one participant to engage in activities and use their remaining abilities, but also to temporarily decrease that patient's contribution to any excess stimulation on the ward. This was with a participant who was often found pacing, being repetitive, and teasing other patients. One nurse adopted the strategy of accompanying the participant on a walk to the hospital foyer where, once there, he would sit and converse, quite appropriately, with familiar staff and patients passing by. This simple, but thoughtful intervention actually served to engage the participant in an activity that was mentally stimulating, social, physical, and emotionally supportive. These are four of the seven types of activities described by Helm and Wekstein (Radar, 1995).
The concept of promoting abilities and compensating for lost abilities also had application for other participants. One participant had lost the ability to locate and use the toilet. The outcome of this was a tendency to void and defecate anywhere on the ward. The same participant also had a tendency to remove his clothing. These were issues for the study because "removing clothing" is one behavior assessed by the CMAI, and "other inappropriate behavior" another. An examination of the gerontological nursing literature revealed the perspective that a patient who removes his clothing may either be hot or in need of the use of a toilet (Roper et al., 1991). This was shared with the nursing staff and an effort was made to assist the participant to use the toilet on a more frequent basis. Although the strategy did not serve to decrease the tendency of the participant to undress, it did serve to substantially decrease the number of times the participant would void and defecate about the ward. This decreased in frequency from 5 (once or twice a day) on the baseline CMAI to 1 (never) on the final CMAI.
Not every effort to decrease agitation was successful. One disappointing aspect of the study was die lack of decrease in the frequency of agitated behaviors despite the implementation of nursing interventions in place for this purpose. In fact, the mean score for CMAI Factor 1 (physically aggressive) behaviors was slightly, but not significantly higher in the postintervention phase than in the preintervention phase. This could be explained, perhaps, as a consequence of the progression of the participants' dementia over the long course of the study. The effect of maturation, of the dementia in this case, would also serve to explain the increase in assaults noted among non-participants present on the ward over the same two time periods as for participants of the study (Figure 2). This conjecture does not, however, explain the simultaneous decrease in assault occurrence among participants over the study time periods. The relationship of agitation to assault occurrence among older adults with dementia is one area requiring considerably more research.
An examination of the relationship between frequency of committing assault and length of hospitalization, for continuing care patients with dementia, may be one particularly interesting subject for future research. Such research may be useful for determining whether propensity for committing assault increases or decreases over the course of hospitalization, along with progression of the disorder. An increase in assaultiveness as, perhaps, a consequence of increasing cognitive impairment may be one possible explanation for the increased number of assaults noted among nonparticipants in this study.
Some rival explanations to the experimental intervention could be offered to attempt to explain the decrease in assault occurrence among participants. These may include, for example, the influences of physical restraint use and administration of psychotropic medications. Unfortunately, program evaluation for the selected ward was only expanded to include frequency and length of physical restraint use at the onset of the post-intervention time period. Similarly, tracking the frequency with which psychotropic medication was administered on an as-necessary basis started only during the first month of the post-intervention time period.
Restraint use for all participants decreased from 1,138 hours during the first month of the post-intervention time period to 724 hours by the end of the second month. Restraint use for all participants never exceeded 512 hours per month for the remainder of the post-intervention time period. One of the recommendations of the guide to interventions for assaultive behavior, emphasized as a strategy in dus study, was for the use of physical restraint as a last resort only after psychosocial and pharmacological interventions had failed (Chou et al., 1996).
Use of psychotropic medications on an as-necessary basis (i.e., pro re nata [prn]), initially decreased from six times for all participants during the first mondi of the post-intervention time period to two times by the end of the second month. Administration of prn medication was variable over the remaining months of the post-intervention time period. The mean number of times prn medications were administered, for all participants, was 10.5 for the remaining 7 months of the study. Thus, an increase in prn psychotropic medication to participants over the course of the post-intervention time period may have also influenced the frequency of assaults and amount of time that physical restraints were used. In other words, both the decrease in assault and physical restraint use among participants may have been related to not only the effect of the psychosocial interventions, but also the effect of increased use of prn medications.
There are some limitations to this study that may limit the generalizability of its findings. The sample size was exceedingly small. The study was only conducted on a single ward of a single site. The nature of the nursing staff was not consistent for the duration of the study. There was a gradual turnover among half of the 20 staff members during the entire course of the study. However, new staff assigned to the ward were oriented to the project, the individualized approaches for participants, and invited to participate.
During the course of this study, an attempt was made to decrease assault occurrence by emphasizing psychosocial approaches to decrease agitation, as opposed to physical or chemical restraint. There are other approaches worthy of investigation. The effort to investigate such approaches is certainly worthwhile because strategies to decrease assault may not only provide a more physically safe environment for those afflicted with dementia, but may also provide a more satisfactory psychosocial environment.
- Chou, K., Kaas, M., & Richie, M. (19%). Assaultive behavior in geriatric patients. Journal of Gerontological Nursing, 22(11), 30-38.
- Cohen-Mansfield, J. (1991). Instruction manual for the Cohen-Mansfield agitation inventory (CMAI). Rockville, MD: Research Institute of the Hebrew Home of Greater Washington.
- Cohen-Mansfield, J., Marx, M., & Rosenthal, A. (1989). A description of agitation in a nursing home. Journal of Gerontology, 44(3), 77-84.
- Cohen-Mansfield, J., Werner, P., & Reisberg, B. (1995). Temporal order of cognitive and functional impairment in a nursing home population. Journal of the American Geriatria Society, 43(9), 974-978.
- Colenda, C, & Hamer, R. (1991). Antecedents and interventions for aggressive behavior of patients at a geropsychiatric state hospital. Hospital and Community Psychiatry, 42(3), 287-292.
- Cooper, A., & Mendonca, J. (1991). A prospective study of patient assaults on nurses in a provincial psychiatric hospital in Canada. Acta Psychiatrica Scandinavica, 84, 163-166.
- Dawson, P., Wells, D., & Kline, K. (1993). Enhancing the abilities of persons with Alzheimer's and related dementias: A nursing perspective. New York: Springer.
- Gibson, M. (1997). Differentiating aggressive and resistive behaviors in long-term care. Journal of Gerontological Nursing, 23(4), 21-28.
- Gilley, D., Wilson, R., Beckett, L., & Evans, D. (1997). Psychotic symptoms and physically aggressive behavior in Alzheimer's disease. Journal of the American Geriatria Society, 45(9), 1074-1079.
- Groulx, B. (1998). Nonpharmacologic treatment of behavioral disorders in dementia. The Canadian Alzheimer Disease Review, 2(1), 6-9.
- Haber, L., Fagan-Pryor, E., & Allen, M. (1997). Comparison of registered nurses' and nursing assistants' choices of interventions for aggressive behaviors. Issues in Mental Health Nursing, 18, 113-124.
- Jackson, G., Templeton, G., & Whyte, J. (1999). An overview of behaviour difficulties found in long-term elderly care settings. International Journal of Geriatric Psychiatry, 14, 426-430.
- Littrell, K., & Littrell, S. (1998). Current understanding of violence and aggression: Assessment and treatment. Journal of Psychosocial Nursing, 36(12), 18-24.
- Miller, R., Snowdon, J., & Vaughan, R. (1995). The use of the Cohen-Mansfield agitation inventory in the assessment of behavioral disorders in nursing homes. Journal of the American Geriatria Society, 43(5), 546-549.
- Ministry of Health and Long Term Care. (1997). Putting the PIECES together: A psychogeriatric guide and training program for professionak in long term care facilities at Ontario. Toronto, Canada: Queen's Printer.
- Mintzer, J., Colenda, C, Waid, L., Lewis, L., Meeks, ?., Stuckey, M., Bachman, D., Saladin, M., & Sampson, R. (1997). Effectiveness of a continuum of care using brief and partial hospitalization for agitated dementia patients. Psychiatric Services, 48(11), 1435-1439.
- Munro, B. (2001). Statistical methods for health care research. Philadelphia: J.B. Lippincott.
- Orem, D. (1995). Nursing: Concepts of practice. St. Louis, MO: Mosby.
- Pearson, M., Wilmot, E., & Padi, M. (1986). A study of violent behavior among in-patients in a psychiatric hospital. British Journal of Psychiatry, 149, 232-235.
- Radar, J. (1995). Individualized dementia care: Creative, compassionate approaches. New York: Springer.
- Radar, J., Lavelle, M., Hoeffer, B., & McKenzie, D. (19%). Maintaining cleanliness: Individualized approaches. Journal of Gerontological Nursing, 22(3), 32-38.
- Ray, C, & Subich, L. (1998). Staff assaults and injuries in a psychiatric hospital as a function of three attitudinal variables. Issues in Mental Health Nursing, 19, 277-289.
- Reisberg, B., Ferris, S., DeLeon, M., ßc Crook, T. (1982). The Global Deterioration Scale for assessment of primary degenerative dementia. American Journal of Psychiatry, 139(9), 1136-1140.
- Roper, J., Shapira, J., & Chang, B. (1991). Agitation in the demented patient: A framework for management. Journal of Gerontological Nursing, 17(3), 17-21.
- Swearer, J., Drachmen, D., O'Donnell, B., & Mitchell L. (1988). Troublesome and disruptive behaviors in dementia: Relationships to diagnoses and disease severity. Journal of the American Geriatria Society, 36(9), 784-790.
MEAN COHEN-MANSFIELD AG~TAT1ON INVENTORY (CMAI) SCORES (N=$)