The Outcomes of an Inpatient Treatment Program for Geriatric Patients With Dementia and Dysfunctional Behaviors

Transcription

1 Copyright 1999 by The Cerontological Society of America The Cerontologist Vol. 39, No. 6, This study evaluated outcomes of an inpatient program designed to reduce severe agitated behavior in geriatric patients with dementia who could not be successfully treated on an outpatient basis. An individualized treatment plan was created for each patient (N = ) that involved pharmacological and nonpharmacological interventions with behavioral, environmental, and psychological components. Assessment of behavioral, cognitive, and functional status was conducted for each patient on admission to the program and at discharge. Significant improvements on these assessments were observed. We conclude that the longitudinal, multidisciplinary approach used in this study was effective in significantly reducing intrusive and dangerous behaviors while preserving or enhancing patients' cognitive and functional abilities. Key Words: Dementia, Geriatric pyschiatry, Disruptive behavior The Outcomes of an Inpatient Treatment Program for Geriatric Patients With Dementia and Dysfunctional Behaviors Alvin Holm, MD, FACP, 1 Matthew Michel, BS, 2 Greg A. Stern, BS, 3 Tsui-min Hung, MHS, 2 Theresa Klein, OTR, 4 Linda Flaherty, PhD, 4 Sara Michel, LSW, 4 and Gabe Maletta, MD, PhD 5 The dementias represent a group of largely progressive conditions that include some of the most devastating illnesses afflicting the elderly population. Dementia ranks as the fourth leading cause of death in adults behind heart disease, stroke, and cancer, and results in billions of dollars in health care expenditures annually (Dresse, Marechal, Scuvee-Moreau, & Seutin, 1994). Prevalence studies have indicated that as many as 5% of elderly individuals aged 65 and older suffer from dementia. Furthermore, the incidence of dementia increases with age; by the age of 85, as many as one in two elders may be afflicted (Evans et al., 1989; Livingston, 1994). This study was conducted at HealthEast Bethesda Rehabilitation Hospital's Geriatric Behavior Program in St. Paul, Minnesota, and was partially funded by a grant from the HealthEast Foundation. An oral presentation based on this study was presented at the Eighth Congress of the International Psychogeriatric Association in Jerusalem, Israel, in August This research was supported in part by a grant from the HealthEast Foundation. We thank and acknowledge HealthEast Bethesda Rehabilitation Hospital's Geriatric Behavioral Program nursing staff for their valuable assistance, Karen Brudvig for her library services, and Kari Nordin for data collection. 'HealthEast Bethesda Rehabilitation Hospital Geriatric Behavioral Program, St. Paul, MN, and University of Minnesota, Minneapolis, MN. 2 HealthEast Office of Research and Medical Education, St. Paul, MN. 'Address correspondence to Greg A. Stern, HealthEast Office of Research & Medical Education, 1700 University Avenue West, St. Paul, MN gstern1@ix.netcom.com "HealthEast Bethesda Rehabilitation Hospital Geriatric Behavioral Program, St. Paul, MN. 5 Geriatrics and Extended Care, Verterans Administration Medical Center, Minneapolis, MN. During the course of a dementing illness, impairments are realized in cognitive functions as well as noncognitive or behavioral functions. Cognitive difficulties include impaired memory, visuospatial, language, attention, and executive functions (McKhann, Drachman, Folstein, Katzman, & Stadlan, 1984). Behavioral disturbances include agitation, irritability, mood lability, verbally disruptive behavior, and physically aggressive behavior (Drevets & Rubin, 1989). Although cognitive losses are observed throughout the course of dementing illness, behavioral disturbances tend to occur more frequently in the moderate to moderately severe stages of the illness and can be so disruptive as to prompt institutionalization and increased restrictions on patients in institutionalized settings (Lacks, Becker, Siegal, Miller, & Tinetti, 1992; Reisberg, Ferris, Franssen, Jenkins, & Wisniewski, 1989). Dysfunctional behaviors are the most common reason that dementia patients are admitted to a skilled nursing facility (Streim, Rovner, & Katz, 1996). Pharmacological treatment options for the intellectual impairments seen in dementia include both agents that bolster cognition through enhanced cholinergic transmission and agents that slow the degenerative process by providing some neural protective effect (Schneider, 1996). Despite advances in the treatment of intellectual decline in Alzheimer's disease, untreated behavioral problems pose a considerable challenge. These so-called "excess disabilities" (Brody, Kleban, 668 The Gerontologist

2 Lawton, & Silverman, 1971) resulting from treatable psychiatric illnesses result in medical illnesses, hospital admissions, institutionalizations, and increased health care expenditures (Livingston, 1994). Unfortunately, many of the problem behaviors seen as a result of dementia are often viewed as expected and untreatable consequences of the degenerative condition itself. Roadblocks to successful treatment of dysfunctional behaviors in dementia include the sometimes confusing manner in which treatable conditions present in these patients and the difficulty of effective treatment plan execution. Patients with dementia and severe agitated behaviors are difficult to assess in an outpatient setting. Even when a patient is institutionalized, contact with the treating physician is limited to relatively brief encounters, and treatment decisions become heavily dependent on observations by staff with varying levels of training and experience. We established an approach to treat patients with dementia and severe agitated behaviors that could not be successfully treated on an outpatient basis. The approach allows for comprehensive multidisciplinary assessments in an ongoing fashion, thus enabling the treatment team to more accurately identify the etiology of problem behaviors. The approach is based on the assumption that a patient's behavioral, cognitive, and functional status not only forms the rationale for intervention, but also validates therapies used and forms the basis for alterations in therapy over time. The impact of this program on patients' behavioral, cognitive, and functional status was measured at admission and discharge. We hypothesized that dysfunctional behaviors seen in dementia patients are commonly the result of treatable psychiatric illnesses and that successful treatment of these illnesses will reduce these behaviors while at the same time preserving, if not enhancing, their cognitive and functional abilities. Methods The Impatient Unit The unit includes 16 beds, with audiovisual equipment in each room. A monitor located in the nurses' station can be switched from room to room, as needed, to observe patient behavior. There is a common dining area (also with audiovisual equipment) with a divider for patients with special behavioral needs. The staffing ratio is one nurse to every three patients. Program Admission Requirements Patients referred to the unit are screened by registered nurses before admission to determine if placement in the geriatric behavior program is appropriate. Patients are admitted if they meet the following admission criteria: (a) a diagnosis of a dementing illness; (b) unsuccessful prior attempts at nursing home and/ or outpatient behavior management; (c) behavioral problems such as agitation, poor attention, anxiety, aggressive behavior, delusional behavior, hallucinations, apathy, and regressed or bizarre behavior that result in the patient's inability or unwillingness to perform the activities of her or his expected daily routine; (d) behavior that threatens harm (to others or to property) and that requires close observation (e.g., 24-h supervision), a quiet room, and/or medication; (e) considered a high risk for self-destructive behavior; or (f) behavior that threatens displacement from the current living situation. The Program Model The evaluation and treatment of patients is multidisciplinary and longitudinal in nature. Treatment is provided in a traditional manner using the principles of geriatric psychiatry when diagnoses related to behavioral dysfunction can be made with an acceptable degree of certainty, such as in clear cases of depression, delirium, or psychosis (Jenike, 1989). In a situation in which diagnosis is unclear, treatment is more experimental and based on observation of the patient. Clinical decision making and treatment planning are guided by repeated assessments of broad-based behavioral, cognitive, and functional outcomes as various therapies are introduced and evaluated for effectiveness. The program's care team includes individuals from all of the disciplines necessary to provide comprehensive assessment and treatment of behavioral dysfunction in geriatric patients. The multidisciplinary care team consists of a geriatrician, a psychologist, a nurse manager, unit nurses, an occupational therapist, a recreational therapist, a social worker, a dietitian, a pharmacist, and a chaplain. A psychiatrist and a neurologist are available as needed for consultation. Initially, assessments are compiled and discussed with the care team members to determine the most appropriate therapeutic interventions. The care team creates an individualized treatment plan to address each patient's unique needs. Pharmacological therapies are used to treat diagnosable psychiatric conditions or to test appropriate hypotheses about the physiological basis of patients' behavioral dysfunction. At the same time, emphasis is placed on maximizing environmental and structural interventions to prevent overreliance on pharmacological treatment. Nondrug interventions often involve the creation of structure for patients who cannot provide this for themselves (Birchmore & Clague, 1983; Cornbleth, 1977; Knopman & Sawyer-DeMaris, 1990; Mintzer et al., 1994). In an attempt to improve structure, specific tasks and activities designed to decrease frustration, promote successful outcomes, and provide security are identified for each patient on the basis of objective assessments of function and intellectual capacity. For lower functioning patients, these tasks include gross motor activities. A higher cognitively functioning patient may be appropriately challenged with a higher level activity involving verbal skills. When appropriate, treatment also includes individual, one-on-one supportive sessions with the psychologist or occupational therapist, planned group activities, or both. Once again, the type of activity, as well as the length of session, is Vol. 39, No. 6,

3 dependent on the patient's present level of cognitive functioning. Environmental management and adaptive equipment are also used to meet patients' individual needs and thus also to promote successful outcomes. Often, a decrease in trie level of visual and auditory stimulation decreases patient agitation and aggressive behavior. For patients who are higher functioning, treatment programs may include behavioral programming (e.g., placing a limit on the number of requests per hour a patient can make, with a care team member assigned to monitor requests). Often, higher functioning patients are given daily schedules with times for daily care, meals, activities, therapies, recreation, and daytime rest periods to control the amount of daytime sleep. The care plans and schedules are kept as brief and simple as possible to allow for greater consistency in implementation. As patients progress, their environmental and structural interventions are adjusted and modified in preparation for discharge. Care team members hold patient care conferences at least two times per week to discuss the results of patients' serial assessments and their responses to treatment. The patients' families are included as an essential part of the care team and are involved in the patients' treatment planning, assessment, and discharge preparations. Comprehensive discharge planning, coordinated by the team social worker, is an important part of the overall treatment plan. The goal at discharge is to place patients in the least restrictive environment that ensures behavioral control. Discussions with patients' responsible caregivers regarding discharge disposition begin at admission. Readiness for discharge is evaluated and is based on many considerations. These considerations include cognitive and functional status, medical needs, dependency for transfers, medication administration and monitoring requirements, and environmental needs for structure and supervision of daily schedules. Discharge options include long-term care skilled nursing facilities, board and care centers, assisted living apartments, or home with caregivers. For patients discharged to home, varied amounts of in-home care and community-based services may be recommended (e.g., adult day care, home health services, meals-on-wheels, chore services). In summary, the inpatient geriatric behavior program is essentially an assessment and treatment "laboratory" that allows members of the care team to make informed and rational choices in planning the most appropriate therapeutic interventions for each patient over time. The success of treatment is measured in relation to the patient's global function (i.e., behavioral, cognitive, and functional status) and not simply in relation to the effects of treatment only on specific problem behaviors. Evaluation Data were collected prospectively on patients admitted to the inpatient geriatric behavior unit between January 1, 1994, and December 31, Information was collected on patients' demographic characteristics including admission source, severity and types of problem behaviors, cognitive level, functional status, active medical comorbidities, psychiatric diagnoses, and pharmacological treatment. All patients were found to be testable despite varying degrees of behavioral dysfunction. Specific psychiatric diagnoses were assigned to patients by a geriatrician at the time of discharge according to Diagnostic and Statistical Manual of Mental Disorders (4th ed., or DSM-IV; American Psychiatric Association, 1994) criteria whenever these criteria clearly were met. Often, however, because of the patients' dementia and the complexity of their condition, psychiatric diagnoses could not be assigned on the basis of strict adherence to DSM-IV criteria. In these cases, the geriatrician's diagnoses relied on the interpretation of the patient's behavior in the context of her or his cognitive impairment as well as her or his response to treatment. Patients' behavioral, cognitive, and functional status were assessed at admission and discharge. Assessments of aggressive behavior were conducted by a trained clinical nurse manager, using the Rating Scale for Aggressive Behavior in the Elderly (RAGE; Patel & Hope, 1992). Cognitive performance was assessed by an occupational therapist certified in using the Allen Cognitive Level Test (ACL; Allen, 1982, 1985; Allen, Earnart, & Blue, 1992; Velligan, Bow-Thomas, Mahurin, Dassoric, & Erdely, 1998). Functional status was measured by a certified occupational therapist, using a modified version of the Functional Independence Measure (FIM), which included the self-cares, social skills, and leisure skills scales (Granger, 1990; Hamilton, Laughlin, Fiedler, & Granger, 1994; Pollak, Rheault, & Stoecker, 1996). The RAGE is a 21-item scale. RAGE scores range from 0 to 61, with higher scores indicating higher levels of agitated behavior. Nineteen of the items address specific kinds of aggressive behaviors; one item has to do with the patient's need for sedation or physical restraints to control her or his behavior, and the final item is a global rating of aggressive behavior. The item regarding the need for restraint to control aggressive behavior is rated on a 2-point scale where 0 = no and 1 = yes. All other items are rated on a 4-point frequency scale where 0 = not once in the past three days never; 1 = at least once in the past three days occasionally; 2 = at least once every day in the past three days often; and 3 = more than once every day in the past three days always. The ACL is based on a 6-point ordinal scale where Level 1 = severe cognitive impairment and Level 6 = normal cognitive functioning. The levels assess how patients receive and process information in order to perform activities of daily living (ADLs). The levels also allow for the assessment of attention, problem-solving skills, ability to follow directions, capacity for new learning, and capacity to perform ADLs. In addition, the levels can be used to monitor patients' status on a day-to-day basis, serve as a tool to educate caregivers, and suggest recommendations for discharge. The levels are determined by the patients' completion of an occupational therapy assessment. This assessment is a process whereby the patient completes a series of 670 The Gerontologist

4 three leather lacing stitches that progressively increase in difficulty. Task analysis (breaking down a task into steps) is used to assist the patient in completing each task. The amount of steps, cueing, and expectation of performance vary with level of impairment. The three scales of the FIM address basic ADLs (eating, dressing upper body, dressing lower body, grooming, and toileting), leisure, and social skills. Each item was rated on a 7-point ordinal scale of disability where 1 = total assistance (less than 25% independence), 2 = maximal assistance (at least 25% independence), 3 = moderate assistance (at least 50% independence), 4 = minimal assistance (at least 75% independence), 5 = supervision, 6 = modified independence, and 7 = complete independence. FIM scores ranged from 7, representing need for total assistance in all areas, to 49, representing complete independence in all areas. Results Patient Characteristics Two hundred fifty patients were included in this study. The mean length of stay was 25.6 days {SD 19.1) and ranged from 7 to 152 days. The patients' average age was 81 years {SD = 8). Twenty-three patients (9%) were admitted from home and 164 patients (67%) were admitted from a skilled nursing facility. Patient demographics are summarized in Table 1. Table 1. Summary Characteristics of Patients Admitted to an Inpatient Geriatric Behavior Program Behavioral Diagnoses Patients' behavior problems were classified into five categories (see Table 2). The three most common problem behaviors identified at admission were physical aggression (68%), agitation (62%), and verbal disruption (54%). Table 3 lists the definition of the presenting psychiatric illnesses used by the clinical team. The three most common psychiatric illnesses were major or minor depressive disorders without psychosis (30%), major or minor depressive disorders with psychosis (25%), and bipolar disorder or bipolar symptoms (23%). Patients' common admission behaviors, presentation of psychiatric illnesses, and common active medical comorbidities were identified by a geriatrician (see Table 4). An association between psychiatric illness and admission behaviors was detected. We found that a higher proportion of patients with bipolar illness were physically aggressive than patients with depression (79% vs 64%, p =.04). In addition, we found that a higher proportion of patients with bipolar illness were both physically aggressive and verbally disruptive as compared with patients with depression (58% vs 41%, p =.03). The patients were medically complex, with an average of 2.3 active medical comorbid conditions on admission, including atherosclerotic cardiovascular disease (37%), hypertension (30%), and cerebrovascular disease (20%). Pharmacological Treatment A patient's pharmacological treatment regimen at discharge represents the medication management strategy that was most effective for achieving the goal of decreasing agitated behaviors and maximizing function. Of the 229 patients (92%) who were discharged Demographic Characteristics N a % Gender Male Female Age < Marital status Single o Married Widowed Divorced Race Caucasian African American Other Admission living status Home Skilled nursing facility Acute or transitional care Other assisted living" a Numbers of patients do not always total because of missing data. Percentages are based on valid, nonmissing data. b "Other assisted living" includes assisted living facilities, board and care facilities, and group homes. Table 2. Behavior Name Agitation 3 Verbally disruptive Physically aggressive Resistive behavior InaDDroDriate sexual behavior a Cohen-Mansfield (1995). Definitions of Problem Behaviors Description Inappropriate verbal, vocal, or motor activity not explained by apparent needs or confusion, per se A form of agitated behavior whereby one's utterances, intentionally or unintentionally, result in a prolonged disruption of the environment A form of agitated behavior whereby physical actions potentially resulting in injury (e.g., hitting, spitting, scratching, kicking) are directed toward self or others Behavior, either verbal or physical, indicating an unwillingness to comply with requests or directives Verbal or Dhvsical behavior indicating or inappropriately intimating a desire for sexual contact Vol.39. No fi71

5 Table 3. Definitions of Presenting Psychiatric Illnesses Table 4. Clinical Characteristics Presentation of Psychiatric Illness Major depression without psychosis Minor depression without psychosis Major depression with psychosis Minor depression with psychosis Bipolar disorder Definition DSM-IV criteria 3 Features of depressive illness but does not meet full DSM-IV criteria for major depressive disorder DSM-IV criteria for major depressive disorder including the presence of either delusions or hallucinations Features of depressive illness, including the presence of either delusions or hallucinations, but does not meet full DSM-IV criteria for major depressive disorder DSM-IV criteria Clinical Characteristics Common admission behaviors Physically aggressive Agitation Verbally disruptive Resistive behavior Inappropriate sexual behavior Presentation of psychiatric illness" 3 Major depression without psychosis Minor depression without psychosis Major depression with psychosis Minor depression with psychosis Bipolar disorder Features of bipolar illness Organic psychosis Anxiety disorder Common admission medical comorbidities Atherosclerotic cardiovascular disease Hypertension Cerebrovascular disease Diabetes mellitus Hypothyroidism Ophthalmologic illness Features of bipolar illness Psychotic disorders Features of bipolar illness, but does not meet full DSM-IV criteria for bipolar disorder DSM-IV criteria a Numbers add to greater than because patients often had more than one admitting behavior, psychiatric illness, or active medical comorbidity. b Psychiatric diagnoses in addition to the patients' dementia diagnoses. Anxiety disorder DSM-IV criteria *DSM-IV = Diagnostic and Statistical Manual of Mental Disorders (4th ed.; American Psychiatric Association, 1994). on any number of psychotropic agents, 66 patients (26%) were discharged on one, 103 patients (41%) received two, 59 patients (24%) received three, and one patient (0.4%) received four. The most common combination of psychotropic agents consisted of an antidepressant and an antipsychotic agent. Twentyone patients (8%) were discharged without any psychotropic agents. It should be noted that on admission, only 20% of patients were being treated with two or more psychotropic agents as compared with 65% at the time of discharge. The medications prescribed at preadmission and discharge are summarized in Table 5. A significantly higher proportion of antipsychotic agents, antidepressant agents, and mood-stabilizing anticonvulsant agents were prescribed to patients at discharge from the program (p <.001). A significantly lower proportion of antianxiety agents were prescribed to patients at discharge from the program (p =.003). Comparison of Patients' Behavioral, Functional, and Cognitive Status Average admission, discharge, and improvement scores for the RACE, ACL, and FIM are presented in Table 6. The mean discharge RAGE score of was significantly lower than the mean admission RAGE score of 2 (p <.001). The observed reductions in mean RAGE scores reflect a clinically significant decrease. A complete elimination of the problem behaviors was observed in 38% of patients. The mean discharge ACL score of 3.5 was significantly higher than the mean admission ACL score of 3.3 (p <.001). The mean discharge FIM score of 24.9 was significantly higher than the mean admission FIM score of 22.4 (p <.001). These relatively small but significant increases in mean ACL and FIM scores indicate that behavioral management was achieved while maintaining and, in many cases, improving patients' cognitive and functional abilities. The outcomes in patients with different psychiatric diagnoses were evaluated. Of particular interest were Pharmacological Treatment Table 5. Pharmacological Treatment Prescribed Before Admission and at Discharge Before Admission Medication N (%) Discharge Medication N (%) Any antipsychotic agents Any antidepressant agents Any mood stabilizing anticonvulsant agents Any antianxiety agents a Two proportion comparisons between medications perscribed before admission and at discharge. 672 The Gerontologist

6 Table 6. Admission and Discharge RAGE, ACL, and FIM Scores Admission Discharge Improvement Measure N M SD M SD M SD P a Improvement Status (%) b RACE (Min 0, Max 61) ACL (Min 1, Max 6) Modified FIM Total (Min 7, Max 49) Self-care FIM (Min 5, Max 35) Social FIM (Min 1, Max 7) Leisure FIM (Min 1, Max 7) Notes: RACE = Rating Scale for Aggressive Behavior in the Elderly; ACL = Allen Cognitive Level Test; FIM = Functional Independence Measure; Min = minimum; Max = maximum. a Two-tailed, paired comparison f. tests for the difference between the scores at admission and discharge. b is defined as an improvement from admission to discharge (i.e., lower RAGE score, higher ACL and FIM scores). indicates no change from admission to discharge. is defined as a decline (i.e., higher RAGE score, lower ACL and FIM scores). For RAGE and FIM scores, and were determined by one full point higher or lower. For the ACL, and were determined by any higher or lower point score. the differences between depressed and bipolar patients (see Table 7). Both groups had similar mean scores on the ACL and FIM on admission (p >.63). We noted that bipolar patients had a significantly higher mean RACE score on admission (p =.04). The before-and-after comparison shows that the treatment significantly improved cognition and function in patients with depression while simply preserving cognition and functional status in patients with bipolar illness (ps <.001 and >.10, respectively). The reduction of agitated behavior was significant for both depressed and bipolar patients despite the higher mean RACE score on admission in the bipolar group (p <.001). Admission source was evaluated with respect to clinical characteristics and treatment (Table 8). Patients admitted from a home environment tended to have lower RAGE scores and higher total FIM scores than those admitted from a skilled nursing facility (p <.004). The before-and-after comparison indicates that dysfunctional behavior was effectively controlled and cognition improved in patients admitted from both a home environment and a skilled nursing facility (p <.02). However, functional status was found to significantly improve in patients admitted from a skilled nursing facility while simply being preserved in patients admitted from a home environment (ps <.001 and.09, respectively). Admission source was significantly associated with the number of psychotropic medications eventually required to achieve an optimal outcome (p =.03). Among the patients admitted from a home environment, 9% were treated with no psychotropic medications, 48% were treated with one, 26% were treated with two, and 17% were treated with three. Among the patients admitted from skilled nursing facilities, 7% were treated with no psychotropic medications, 19% were treated with one, 48% were treated with two, and 26% were treated with three or more. With respect to the entire study group, improvement in RAGE scores was observed in 90% of patients. Improvement in ACL scores was observed in 79% of patients and improvement in FIM scores was observed in 92%. Seventy percent of the patients demonstrated improvements on all three measures. Eighty-three percent of the total patient population either improved or remained the same on all three measures. Among patients demonstrating reduction in RAGE score, 89% improved or maintained their ACL scores, and 95% improved or maintained their FIM scores. Discussion We herein report on the outcomes of an inpatient behavior program in addressing the cognitive, as well as noncognitive, neuropsychiatric sequelae of dementing inness. Similar to the Mintzer et al. (1993) Vol. 39, No. 6,

7 Table 7. Global Outcomes for Patients With Depression 2 and Bipolar lllness b Admission Discharge Improvement Measure N M SD M SD M SD p 3 Improvement Status (%) RAGE Depression Bipolar ACL Depression Bipolar Modified FIM total Depression Bipolar Notes: The depression group consisted of all major depression and all minor depression patients; the bipolar group consisted of patients with bipolar disorder and patients with features of bipolar illness. RAGE = Rating Scale for Aggressive Behavior in the Elderly; ACL - Allen Cognitive Level Test; FIM = Functional Independence Measure. a Two-tailed, paired-comparison t tests for the difference between the scores at admission and discharge. program, our unit relies heavily on observation and treatment in a multidisciplinary environment with the goal of returning patients to the least restrictive environment that would ensure continuing behavioral control. In addition, we measured global parameters of function, including the behavioral, cognitive, and functional status of all study participants on both admission and discharge. We also attempted to more clearly define the nature of the treatable psychiatric dysfunction occurring in these patients as well as the impact of psychotropic therapies used as part of their treatment program. In identifying psychiatric diagnoses in this study sample, we found that 55% of patients had a major depressive illness or significant features of depression, although only 18% were being treated with antidepressant medications before admission. Features of bipolar illness were present in 23% of patients, with 5% meeting DSM-IV criteria for bipolar disorder. Although bipolar illness is not commonly reported in association with dementing illness (Shulman, 1997), our results indicate that it is a major contributor to the etiology of behavioral dysfunction leading to admission to this inpatient treatment program. Few of the patients in our study were being treated for bipolar illness at admission, evidenced in part by the fact that at the time of admission only 4 patients were being treated with mood stabilizing agents. In contrast, at discharge 102 patients received mood-stabilizing anticonvulsant agents as part of their treatment program. Combination psychotherapeutic regimens were found to be an essential aspect of treatment for a majority of our patients. At discharge from the program, combination psychotherapeutic regimens were used in 65% of patients, with the most common combination consisting of an antidepressant and an antipsychotic agent. Monotherapy was prescribed in only 26% of patients, with antipsychotics being used most frequently. Patients demonstrating evidence of a mixed mood state were often treated with three agents, including medications from the antidepressant, antipsychotic, and mood-stabilizing classes. Although it is clear that elderly people are a disparate group when it comes to the use of psychotherapeutic medications, we believe these results underscore the need to individualize treatment programs in these patients. In addition, these findings demonstrate that, contrary to common belief, "polypharmacy" can achieve desirable outcomes in patients who have shown themselves to be refractory to more conservative interventions. Improvements in global functioning were realized during the course of inpatient treatment as determined by performance on tests of behavioral, cognitive, and functional parameters. Interestingly, although dramatic improvements in behavior were seen in most patients, improvements in cognitive and functional conditions were more modest. We believe the improvements seen in the majority of patients in this study resulted from the effective treatment of comorbid psychiatric illness. The fact that improvements in behavior were more 674 The Gerontologist

8 Table 8. Global Outcomes for Patients Admitted From Home and Skilled Nursing Facilities (SNF) Admission Discharge Improvement Measure N M SD M SD M SD p a Improvement Status (%) RACE Home SNF ACL Home SNF Modified FIM Total Home SNF Notes: RAGE = Rating Scale for Aggressive Behavior in the Elderly; ACL = Allen Cognitive Level Test; FIM = Functional Independence Measure. a Two-tailed, paired-comparison t tests for the difference between the scores at admission and discharge. dramatic is consistent with the idea that cognitive and functional gains were secondarily related to improvements in behavior rather than primary effects of the prescribed treatments. In comparing clinical characteristics and outcomes based on the degree of structure in patients' daily lives at the time of admission, we chose to compare patients admitted from a home environment to patients admitted from a skilled nursing facility. Patients admitted from home were found to be more functionally independent and less behaviorally disruptive than their institutionalized counterparts. Treatment regimens required to achieve successful outcomes in patients admitted from home were less complex, often requiring the use of one or no psychotropic medication. Of interest, statistically significant improvements in functional status were realized only in patients admitted from a skilled nursing facility. Clearly, although institutionalized patients were found to be more advanced in their disease and require more complex management, their outcomes were largely comparable to patients with a lesser degree of dysfunction. In examining outcomes in the bipolar and depressed groups, we found that although improvements in behavior were comparable, outcomes with respect to cognitive and functional parameters were different. Depressed patients realized clinically and statistically significant gains in cognitive and functional status with treatment, whereas bipolar patients remained stable with respect to these outcomes. Patients with bipolar illness were also found to have higher RACE scores at admission compared with depressed patients. These findings are consistent with reports of elderly bipolar patients in other studies (Berrios & Bakshi, 1991; Shulman, 1997). The ACL was used to assess the cognitive level of the study population despite the less frequent use of this instrument in the dementia literature. We believe this test is more appropriate for this population than other widely known instruments such as the Mini- Mental State Examination (MMSE; Folstein, Folstein, & McHueh, 1975) because, unlike the MMSE, which, in general, measures domains of cognitive performance, the ACL measures function related to activities of daily living. Kehrberg, Kuskowski, Mortimer, and Shoberg (1992) reported that persons unable to respond to the MMSE were able to be evaluated with the ACL. In addition, the ACL has been found to correlate significantly with the MMSE in previous studies (Heying, 1985; Kehrberg et al., 1992; Wilson, Allen, McCormack, & Burton, 1989). It is interesting that despite the diverse nature of this study population, clear improvements in most outcome measurements were found. Differences in individual outcomes for specific patients can be explained by the fact that patients included in the study had a variety of dementia diagnoses with a large range of behavioral, cognitive, and neurologic problems. They were also medically complex, with an average of 2.3 active comorbid medical conditions noted at the time of admission. One would expect these factors to influence individual patients' outcomes. Vol. 39, No. 6,

9 There are several limitations of this study. One is the lack of a control group. For example, it is wellknown that a significant placebo effect can be noted when the behavioral effects of psychotropic therapies are studied in mentally ill patients in a randomized, controlled fashion. Another limitation is that data collection did not include the patient's severity and type of dementia, which may be related to the outcomes of the treatment. For example, bipolar illness in elderly patients has been attributed to focal brain lesions occurring as the result of such conditions as cerebral infarcts and tumors (Shulman, 1997). We could not determine if the bipolar illness seen in this study was attributable only to focal brain disease or if primary degenerative conditions, such as Alzheimer's disease, were also responsible. This study also did not follow patients after discharge; therefore, we were not able to determine if the gains realized within the inpatient program were maintained in the outpatient setting or whether these improvements resulted in additional tangible benefits. This study may also be limited by a form of investigator bias in that the admission and discharge assessments were conducted by the treatment team. The results of our study indicate that inpatient geriatric behavioral treatment using a longitudinal, multidisciplinary approach can effectively evaluate and treat dysfunctional behaviors in dementia patients while at the same time preserving, if not enhancing, their cognitive and functional abilities. Because of the evolving and progressive nature of dementing illness, further research is needed to determine if inpatient geriatric behavior programs, such as the one described in this study, can result in long-term benefits in patients following discharge (i.e., reduction in the frequency of additional medical illness and hospital readmission rates). Further research is also needed to more clearly define the nature, course, and optimal treatment of bipolar illness in dementia patients. References Allen, C. K. (1982). Independence through activity: The practice of occupational therapy (psychiatry). The American Journal of Occupational Therapy, 36, Allen, C. K. (1985). Occupational therapy for psychiatric diseases. Measurement and management of cognitive disabilities. Boston/Toronto: Little, Brown. Alldn, C. K., Earhart, C. A., & Blue, T. (1992). Occupational therapy treatment goals for the physically and cognitively disabled. Rockville, MD: American Occupational Therapy Association, Inc. American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author. Berrios, C. E., & Bakshi, N. (1991). Manic and depressive symptoms in the elderly: Their relationships to treatment outcome, cognition and motor symptoms. Psychopathology, 24, Birchmore, T., & Clague, S. (1983). A behavioural approach to reduce shouting. Nursing Times, 79, Brody, E. M., Kleban, M. H., Lawton, P., & Silverman, H. A. (1971). Excess disabilities of mentally imparied aged: Impact of individualized treatment. The Gerontologist, 11, Cohen-Mansfield, J. (1995). Assessment of disruptive behavior/agitation in the elderly: Function, methods, and difficulties. Journal of Geriatric Psychiatry Neurology, 8, Cornbleth, T. (1977). Effects of a protected hospital ward area on wandering and nonwandering geriatric patients. Journal of Gerontology, 32, Dresse, A., Marechal, D., Scuvee-Moreau, J., & Seutin, V. (1994). Towards a pharmacological approach of Alzheimer's disease based on the molecular biology of the amyloid precusor protein (AFP). Life Sciences, 55, Drevets, W. C, & Rubin, E. H. (1989). Psychotic symptoms and the longitudinal course of senile dementia of the Alzheimer type. Biological Psychiatry, 25, Evans, D. A., Funkenstein, H. H., Albert, M. S., Scherr, P. A., Cook, N. R., Chown, M. J., Hebert, L. E., Hennekens, C. H., & Taylor, J. O. (1989). Prevalence of Alzheimer's disease in a community population of older persons. JAMA, 262, Folstein, M. F., Folstein, S. E., & McHugh, P. R. (1975). Mini-Mental State A practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research, 12, Granger, C. V. (1990). Health accounting: Functional assessment of the long-term patient. In F. Kottke (Ed.), Krusen's handbook of physical medicine and rehabilitation (4th ed., pp ). Philadelphia: W. B. Saunders. Hamilton, B. B., Laughlin, J. A., Fiedler, R. C, & Granger, C. V. (1994). Interrater reliability of the 7-level Functional Independence Measure (FIM). Scandinavian Journal of Rehabilitation Medicine, 26, Heying, L. M. (1985). Research with subjects having senile dementia. In C. K. Allen (Ed.), Occupational therapy for psychiatric diseases: Measurement and management of cognitive disabilities (pp ). Boston: Little, Brown. Jenike, M. A. (1989). Geriatric psychiatry and psychophamacology. St. Louis: Mosby-Year Book. Kehrberg, K. L., Kuskowski, M. A., Mortimer, J. A., & Shoberg, T. D. (1992). Validating the use of an enlarged, easier-to-see Allen Cognitive Level Test in geriatrics. Physical & Occupational Therapy in Geriatrics, 70(3), Knopman, D. S., & Sawyer-DeMaris, S. (1990). Practical approach to managing behavioral problems in dementia patients. Geriatrics, 45(4), Lachs, M. S., Becker, M., Siegal, A. P., Miller, R. L., & Tinetti, M. E. (1992). Delusions and behavioral disturbances in cognitively impaired elderly persons. Journal of the American Geriatric Society, 40, Livingston, G. (1994). The scale of the problem. In A. Burns & R. Levy (Eds.), Dementia (pp ). London: Chapman and Hall. McKhann, G., Drachman D., Folstein, M., Katzman, R., & Stadlan, E. M. (1984). Clinical diagnosis of Alzheimer's disease: Report of the NINCDS- ADRA Work Group under the auspices of the Department of Health and Human Services Task Force on Alzheimer's disease. Neurology, 34, Mintzer, J., Lewis, L., Pennypacker, L., Pitner, J., Simpson, W., Bachman, D., Wohlreich, G., Meeks, A., Hunt, S., Herman, K., et al. (1994). A new approach for the management of acute psychiatric disorders in elderly demented patients. Journal - South Carolina Medical Association, 90, Mintzer, J., Lewis, L., Pennypacker, L., Simpson, W., Bachman, D., Wohlreich, G., Meeks, A., Hunt, S., & Sampson, R. (1993). Behavioral intensive care unit (BICU): A new concept in the management of acute agitated behavior in the elderly demented patients. The Gerontologist, 33, Patel, V., & Hope, R. (1992). A rating scale for aggressive behaviour in the elderly The RAGE. Psychological Medicine, 22, Pollak, N., Rheault, W., & Stoecker, J. L. (1996). Reliability and validity of the FIM for persons aged 80 years and above from a multilevel continuing care retirement community. Archives of Physical Medicine Rehabilitation, 77(10), Reisberg, B., Ferris, S. H., Franssen, E., Jenkins, E. C, & Wisniewski, K. E. (1989). Clinical features of a neuropathologically verified familial Alzheimer's cohort with senile onset Alzheimer's disease and etiopathogenic implications. In Alzheimer's disease and related disorders (pp ). New York: Alan R. Liss. Schneider, L. S. (1996). New therapeutic approaches to Alzheimer's disease. Journal of Clinical Psychiatry, 57(Suppl. 14), Shulman, K. I. (1997). Disinhibition syndromes, secondary mania and bipolar disorder in old age. Journal of Affective Disorders, 46, Streim, J., Rovner, B., & Katz, I. (1996). Psychiatric aspects of nursing home care. In J. Sadavoy, Joel, Lazarus, Lawrence, Jarvik, Lissy, & Grossberg (Eds.), Comprehensive review of geriatric psychiatry (pp ). Washington, DC: American Psychiatric Press. Velligan, D. I., Bow-Thomas, C. C, Mahurin, R., Dassori, A., & Erdely, F. (1998). Concurrent and predictive validity of the Allen Cognitive Levels Assessment. Psychiatry Research, 80(3), Wilson, D. S., Allen, C. K., McCormack, G., & Burton, G. (1989). Cognitive disability and routine task behaviors in a community-based population with senile dementia. Occupational Therapy Practice, 7(1), Received July 7, 1998 Accepted August 31, The Gerontologist