Rationale for Early Diagnosis of Mild Cognitive Impairment (MCI) Supported by Emerging Digital Technologies.

Disease-modifying pharmacotherapies for Alzheimer's Disease (AD) are currently in late-stage clinical development; once approved, new healthcare infrastructures and services, including primary healthcare, will be necessary to accommodate a huge demand for early and large-scale detection of AD. The increasing global accessibility of digital consumer electronics has opened up new prospects for early diagnosis and management of mild cognitive impairment (MCI) with particular regard to AD. This new wave of innovation has spurred research in both academia and industry, aimed at developing and validating a new "digital generation" of tools for the assessment of the cognitive performance. In light of this paradigm shift, an international working group (the Global Advisory Group on Future MCI Care Pathways) convened to elaborate on how digital tools may be optimally integrated in screening-diagnostic pathways of AD The working group developed consensus perspectives on new algorithms for large-scale screening, detection, and diagnosis of individuals with MCI within primary medical care delivery. In addition, the expert panel addressed operational aspects concerning the implementation of unsupervised at-home testing of cognitive performance. The ultimate intent of the working group's consensus perspectives is to provide guidance to developers of cognitive tests and tools to facilitate the transition toward globally accessible cognitive screening aimed at the early detection, diagnosis, and management of MCI due to AD.

Early Detection of Mild Cognitive Impairment (MCI) in Primary Care.

Mild cognitive impairment (MCI) is significantly misdiagnosed in the primary care setting due to multi-dimensional frictions and barriers associated with evaluating individuals' cognitive performance. To move toward large-scale cognitive screening, a global panel of clinicians and cognitive neuroscientists convened to elaborate on current challenges that hamper widespread cognitive performance assessment. This report summarizes a conceptual framework and provides guidance to clinical researchers and test developers and suppliers to inform ongoing refinement of cognitive evaluation. This perspective builds upon a previous article in this series, which outlined the rationale for and potentially against efforts to promote widespread detection of MCI. This working group acknowledges that cognitive screening by default is not recommended and proposes large-scale evaluation of individuals with a concern or interest in their cognitive performance. Such a strategy can increase the likelihood to timely and effective identification and management of MCI. The rising global incidence of AD demands innovation that will help alleviate the burden to healthcare systems when coupled with the potentially near-term approval of disease-modifying therapies. Additionally, we argue that adequate infrastructure, equipment, and resources urgently should be integrated in the primary care setting to optimize the patient journey and accommodate widespread cognitive evaluation.

Early Detection of Mild Cognitive Impairment (MCI) in an At-Home Setting.

Emerging digital tools have the potential to enable a new generation of qualitative and quantitative assessment of cognitive performance. Moreover, the ubiquity of consumer electronics, such as smartphones and tablets, can be harnessed to support large-scale self-assessed cognitive screening with benefit to healthcare systems and consumers. A wide variety of apps, wearables, and new digital technologies are either available or in development for the detection of mild cognitive impairment (MCI), a risk factor for dementia. Two categories of novel methodologies may be considered: passive technologies (which monitor a user's behavior without active user input) and interactive assessments (which require active user input). Such examinations can be self-administered, supervised by a caregiver, or conducted by an informant at home or outside of a clinical setting. These direct-to-consumer tools have the potential to sidestep barriers associated with cognitive evaluation in primary care, thus improving access to cognitive assessments. Although direct-to-consumer cognitive assessment is associated with its own barriers, including test validation, user experience, and technological concerns, it is conceivable that these issues can be addressed so that a large-scale, self-assessed cognitive evaluation that would represent an initial cognitive screen may be feasible in the future.

Antidepressant prescribing in nursing homes: is there a place for tricyclics?

OBJECTIVE: To deduce a model describing physicians' choice of antidepressants for treating elderly nursing home patients. METHODS: Subjects were geriatric residents of 137 skilled nursing facilities who regularly received an antidepressant medication for at least one month (n = 3,440, 28% of all residents). Reasons for prescribing antidepressants and physicians' diagnoses of depression and dementia were identified by medical record audit. Residents were grouped by dementia and antidepressant target symptoms (depression, or one or more non-psychiatric symptoms, i.e. insomnia, pain, incontinence, itching). RESULTS: Selective serotonin reuptake inhibitors (SSRIs) were prescribed preferentially over tricyclic antidepressants (TCAs) for treating depression in both demented and non-demented residents, but TCAs were nine times more likely to be prescribed for treating non-psychiatric target symptoms alone. When non-psychiatric target symptoms were present without depression or dementia, both amitriptyline and nortriptyline prescribing was increased, but amitriptyline appeared to be the antidepressant of choice. In all subgroups examined, its use was two to five times more prevalent when such symptoms were present. In patients with dementia, amitriptyline prescribing declined whether or not non-psychiatric target symptoms were present, but nortriptyline prescribing did not; nortriptyline was three times more likely than amitriptyline to be prescribed in the presence of dementia. CONCLUSIONS: Physicians prescribe anticholinergic TCAs principally to treat common non-depressive symptoms in nursing home residents, preferring SSRIs for uncomplicated depression and depression with dementia. They tend to avoid prescribing anticholinergic TCAs other than nortriptyline when they recognize a patient as demented. The data suggest that physicians employ a decision model for antidepressant prescribing that simultaneously recognizes the utility of TCAs in treating non-psychiatric symptoms and the anticholinergic vulnerability of older, especially demented, patients. Whether or not this model leads to optimal patient management requires further study.

Sleep disturbances in patients with Alzheimer's disease: epidemiology, pathophysiology and treatment.

Older adults represent an ever-growing proportion of the population of the industrialised nations, with a corresponding increase in the numbers of patients with dementing disorders. A common complaint in both normal aging and the dementias is that of significant sleep disturbance. The major causes of sleep disruption in aging and dementia include: (i) physiological changes that arise as part of normal, 'nonpathological' aging; (ii) sleep problems due to one of many physical or mental health conditions and their treatments; (iii) primary sleep disorders; (iv) poor 'sleep hygiene', that is, sleep-related practices and habits; and (v) some combination of these factors. Disrupted sleep in patients with dementia is a significant cause of stress for caregivers and frequently leads to institutionalisation of patients. It should be a target of clinical management when the goal is sustained home care, and when it is associated with disturbances of mood or behaviour. While the neuropathology of dementia can directly disrupt sleep, sleep disturbances in patients with dementia often have multiple causes that require systematic evaluation. Thorough assessment of associated psychopathology, day-time behaviour, medical disorders, medications, pain and environmental conditions is needed for optimal management. Differential diagnosis of a sleep problem in dementia is the basis of rational pharmacotherapy. However, patients with dementia are likely to be more sensitive than elderly persons without dementia to adverse cognitive and motor effects of drugs prescribed for sleep. Clinicians need to: (i) evaluate sleep outcomes when treating medical, psychiatric and behavioural disorders in older adults; (ii) be alert to emerging behavioural and environmental approaches to treatment; (iii) combine nonpharmacological strategies with drug therapies, when required, for added value; and (iv) avoid use of multiple psychotropic medications unless they prove essential to the adequate management of sleep disturbances.

Geriatric mental health services research: Strategic Plan for an Aging Population: Report of the Health Services Work Group of the American Association for Geriatric Psychiatry.

In November 1999, a working group of the American Association for Geriatric Psychiatry (AAGP) convened to consider strategic recommendations for developing geriatric mental health services research as a scientific discipline. The resulting consensus statement summarizes the principles guiding mental health services research on late-life mental disorders, presents timely and topical priorities for investigation with the potential to benefit the lives of older adults and their families, and articulates a systematic program for expanding the supply of well-trained geriatric mental health services researchers. The agenda presented here is designed to address critical questions in provision of effective mental health care to an aging population and the health policies that govern its delivery.

The Mini-Cog: receiver operating characteristics with expert and naïve raters.

BACKGROUND: As elderly populations grow, dementia detection in the community is increasingly needed. Existing screens are largely unused because of time and training requirements. We developed the Mini-Cog, a brief dementia screen with high sensitivity, specificity, and acceptability. Here we describe the development of its scoring algorithm, its receiver operating characteristics (ROC), and the generalizability of its clock drawing scoring system. SAMPLE AND METHODS: A total of 249 multi-lingual older adults were examined. Scores on the three-item recall task and the clock drawing task (CDT-CERAD version) were combined to create an optimal algorithm. Receiver operating characteristics for seven alternatives were compared with those of the MMSE and the CASI using expert raters. To assess the CDT scoring generalizability, 20 naïve raters, without explicit instructions or prior CDT exposure, scored 80 randomly selected clocks as "normal" or "abnormal" (20 from each of four CERAD categories). RESULTS: An algorithm maximizing sensitivity and correct diagnosis was defined. Its ROC compared favorably with those of the MMSE and CASI. CDT concordance between naïve and trained raters was >98% for normal, moderately and severely impaired clocks, but lower (60%) for mildly impaired clocks. Recalculation of the Mini-Cog's performance, assuming that naïve raters would score all mildly impaired CDTs in the full sample as normal, retained high sensitivity (97%) and specificity (95%). CONCLUSION: The Mini-Cog algorithm performs well with simple clock scoring techniques. The results suggest that the Mini-Cog may be used successfully by relatively untrained raters as a first-stage dementia screen. Further research is needed to characterize the Mini-Cog's utility when population dementia prevalences are low.

Frequency of tau gene mutations in familial and sporadic cases of non-Alzheimer dementia.

BACKGROUND: Mutations in the tau gene have been reported in families with frontotemporal dementia (FTD) linked to chromosome 17. It remains uncertain how commonly such mutations are found in patients with FTD or non-Alzheimer dementia with or without a positive family history. OBJECTIVE: To determine the frequency of tau mutations in patients with non-Alzheimer dementia. PATIENTS AND METHODS: One hundred one patients with non-Alzheimer, nonvascular dementia, most thought to have FTD. Of these, 57 had a positive family history of dementia. Neuropathologic findings were available in 32. The tau gene was sequenced for all exons including flanking intronic DNA, portions of the 3' and 5' untranslated regions, and at least 146 base pairs in the intron following exon 10. RESULTS: Overall, the frequency of the tau mutations was low, being 5.9% (6/101) in the entire group. No mutations were found in the 44 sporadic cases. However, 6 (10.5%) of the 57 familial cases and 4 (33%) of the 12 familial cases with tau pathologic findings had mutations in the tau gene. The most common mutation was P301L. CONCLUSIONS: We conclude that tau mutations are uncommon in a neurology referral population with non-Alzheimer dementia, even in those with a clinical diagnosis of FTD. However, a positive family history and/or tau pathologic findings increase the likelihood of a tau mutation. There must be other genetic and nongenetic causes of FTD and non-Alzheimer dementia, similar to the etiologic heterogeneity present in Alzheimer disease.

The mini-cog: a cognitive 'vital signs' measure for dementia screening in multi-lingual elderly.

OBJECTIVES: The Mini-Cog, a composite of three-item recall and clock drawing, was developed as a brief test for discriminating demented from non-demented persons in a community sample of culturally, linguistically, and educationally heterogeneous older adults. SUBJECTS: All 129 who met criteria for probable dementia based on informant interviews and 120 with no history of cognitive decline were included; 124 were non-English speakers. METHODS: Sensitivity, specificity, and diagnostic value of the Mini-Cog were compared with those of the Mini-Mental State Exam (MMSE) and Cognitive Abilities Screening Instrument (CASI). RESULTS: The Mini-Cog had the highest sensitivity (99%) and correctly classified the greatest percentage (96%) of subjects. Moreover, its diagnostic value was not influenced by education or language, while that of the CASI was adversely influenced by low education, and both education and language compromised the diagnostic value of the MMSE. Administration time for the Mini-Cog was 3 minutes vs 7 minutes for the MMSE. CONCLUSIONS: The Mini-Cog required minimal language interpretation and training to administer, and no test forms of scoring modifications were needed to compensate for the extensive linguistic and educational heterogeneity of the sample. Validation in clinical and population-based samples is warranted, as its brevity and ease of administration suggest that the Mini-Cog might be readily incorporated into general practice and senior care settings as a routine 'cognitive vital signs' measure.

Effectiveness of nursing home staff as managers of disruptive behavior: perceptions of nursing directors.

Nursing directors of 899 geographically diverse skilled nursing facilities reported substantial limitations in the competence of staff at all levels in managing behaviorally disturbed patients and a broad-based need for improvements in skills. The authors propose an interdisciplinary team training model as an appropriate design for improving competence and promoting collaborative care.

Stage-specific prevalence of behavioral symptoms in Alzheimer's disease in a multi-ethnic community sample.

The authors extended previous studies of the stage-specific prevalence of behavioral pathology to members of ethnic minority groups. Behavioral symptoms and their relationship to severity of Alzheimer's disease (AD) were examined in 125 heterogeneous minority elderly patients interviewed with a modified CERAD protocol, with behavioral symptoms scored on the caregiver-rated BEHAVE-AD Rating Scale. Behavioral disturbances were extremely common, occurring in 98% of the sample; the most common was activity disturbances (89%), followed by paranoid and delusional ideation (72%), aggressivity (64%), anxieties and phobias (61%), depressive symptoms (50%), sleep disturbances (43%), and hallucinations (34%). As in white patients, overall behavioral symptoms were most frequent among patients with moderate and severe dementia. Preliminary evidence supports the possibility of ethnic differences in behavioral profiles; Blacks showed lower affective, anxiety, and sleep symptoms than Asians and Hispanics, and lower total BEHAVE-AD scores than Hispanics. Inquiry in larger, population-based samples will be needed to determine whether the ethnic differences in behavioral symptoms of AD found here are robust and replicable.

Thyroid hormones: positive relationships with cognition in healthy, euthyroid older men.

BACKGROUND: Although the association of clinical hypothyroidism with cognitive deficits is well known, the cognitive effects of thyroid hormones in euthyroid subjects are less studied and understood. The purpose of this study was to examine thyroid-cognition relationships in healthy, euthyroid older men. METHODS: We examined healthy men (N = 44, mean age = 72), excluding clinically hypothyroid/hyperthyroid or diabetic/hyperglycemic subjects and those with dementia, depression, CNS medications, or recent illness. Plasma samples obtained across a 24-hour period were pooled, then assayed for total thyroxine (TT4), total triiodothyronine (TT3), and T3 resin uptake. Free thyroxine index (FT4I) was calculated. A broad cognitive battery (including the Wechsler Adult Intelligence Scale-Revised [WAIS-R], the Dementia Rating Scale [DRS], and the Rivermead Behavioral Profile [PROFILE]) was administered to all subjects. RESULTS: Regression analyses controlling age and education showed TT4 and FT4I to have significant positive relationships with measures of overall cognition; TT4 accounted for 8% to 12% of the variance in omnibus cognitive measures such as WAIS Performance, WAIS Verbal score, and GLOBAL cognitive scores. CONCLUSIONS: Our findings suggest that within "normal" range of variation in plasma thyroid hormones, TT4 but not T3 positively associates with general cognition in healthy elderly men.

The Clock Drawing Test: utility for dementia detection in multiethnic elders.

BACKGROUND: Disproportionate increases in dementia morbidity in ethnic minorities challenge established screening methodologies because of language and culture barriers, varying access to health services, and a relative paucity of cross-cultural data validating their use. Simple screening techniques adapted to a range of health and social service settings would accelerate dementia detection and social and health services planning for demented minority elders. METHODS: The effectiveness of the Clock Drawing Test (CDT) for dementia detection was compared with that of the Mini-Mental State Examination (MMSE) and the Cognitive Abilities Screening Instrument (CASI) in community-dwelling elders of diverse linguistic, ethnic, and educational backgrounds. Subjects (N = 295) were tested at home in their native languages (English, n = 141; another language, n = 154). An informant-based clinical dementia history and functional severity index derived from the Consortium to Establish a Registry for Alzheimer's Disease (CERAD) protocols were used to classify subjects as probably demented (n = 170), and probably not demented (n = 125). RESULTS: All tests were significantly affected by education (p < .001) but not by primary language (p > .05). Sensitivities and specificities for probable dementia were 82% and 92%, respectively, for the CDT; 92% and 92% for the MMSE; and 93% and 97% for the CASI for subjects completing each test. However, in poorly educated non-English speakers, the CDT detected demented subjects with higher sensitivity than the two longer instruments (sensitivity and specificity 85% and 94% for the CDT, 46% and 100% for the MMSE, and 75% and 95% for the CASI). Moreover less information was lost due to noncompletion of the CDT than the MMSE or CASI (severe dementia or refusal: CDT 8%, MMSE 12%, and CASI 16%). CONCLUSIONS: Overall, the CDT may be as effective as the MMSE or CASI as a first-level dementia screen for clinical use in multiethnic, multilingual samples of older adults. Its brevity (1-5 minutes), minimal language requirements, high acceptability, and lack of dependence on specialized testing materials are well adapted for screening of non-English-speaking elderly persons in settings where bilingual interpreters are not readily available and screening time is at a premium.

CD4 and CD8 counts are associated with interactions of gender and psychosocial stress.

OBJECTIVE: This study examined relationships of gender, psychosocial stress/distress (caregiving, hassles, depressed mood), and the relative percentage and absolute cell counts of CD4 and CD8 cells in two samples of older adults (mean age = 69.4)--spouse caregivers of persons with Alzheimer's disease (N = 78) and age- and gender-matched spouses of nondemented controls (N = 72). METHODS: Counts and percentages of CD4 and CD8 cells and psychosocial variables were assessed twice (Time 1, Time 2) over a 15- to 18-month period. Several covariates were examined in the analyses, including body mass index (BMI), medication use, alcohol use, exercise, and illness history. RESULTS: Caregiver men had fewer CD4 cell counts at Times 1 and 2 than did control men (p < .05). At Times 1 and 2, both CD8 cell counts and percentages were positively associated with hassles in men (p < .05), but not in women. Although interactions of hassles and gender were present for CD8 percentages at both times, interactions and main effects were not present for CD4 percentages at either time. When the ratio of CD4 to CD8 levels was analyzed, hassles by gender interactions were present at both Times 1 and 2-hassles were negatively associated with the CD4/CD8 ratio in men (p < .05), but unrelated in women. From Time 1 to Time 2, change analyses showed that increases in hassles scores were associated with decreases in CD4 counts (p < .05), whereas increases in Hamilton Depression Scores were related to increases in both CD8 counts and percentages (p < .05). CONCLUSION: Caregiver status, hassles, and depressed mood had cross-sectional and/or longitudinal associations with CD4 and CD8 counts, but such relationships occurred primarily in men. Moreover, absolute cell counts were more related to psychosocial factors than were percentages.

Psychiatric consultation in the nursing home. A survey of six states.

The authors examined availability, characteristics, and perceived adequacy of psychiatric consultation in nursing homes, as reported by directors of nursing, who returned 899 questionnaires. Thirty-eight percent of nursing home residents were judged to need a psychiatric evaluation; current frequency of consultation was rated as adequate by half of nursing directors. Nearly two-thirds reported that psychiatrists adequately provided diagnostic and medication recommendations; however, advice on nonpharmacologic management techniques, staff support, and dealing with staff stress and family conflicts was largely viewed as inadequate. Findings suggest that perceived need for psychiatric services is far greater than the level actually provided. Overall, more attention must be directed to identifying incentives for psychiatrists to practice in nursing homes, determining clinical effectiveness of mental health services, and examining effects of alternative payment mechanisms on level of care.

Depression and Chronic Obstructive Pulmonary Disease: Treatment Trials.

Chronic obstructive pulmonary disease (COPD) affects over 16 million people in the United States and is a major cause of disability and death worldwide. Its prevalence and mortality are increasing disproportionately among the elderly, women, African-Americans, persons of lower socioeconomic status, and the populations of developing countries in which tobacco is aggressively marketed. In contrast to other major chronic diseases such as heart disease and cancer, medical treatments for COPD have not made decisive inroads into its morbidity or death rates over the last 20 years, resulting in continuing efforts to reduce disability in patients with established disease. Depression is a source of increased disability in COPD, and, as in other chronically ill patient populations, is often unrecognized and untreated in the primary and specialty care sectors. Nearly half of all patients experience some depressive symptoms and at least one-fifth have had one or more major depressive episodes, frequently of long duration. Evidence from randomized controlled trials supports the thesis that patients with mild depression improve with multidisciplinary rehabilitation, whereas patients with major depression may require specific pharmacotherapy to achieve significant improvement in mood disorder and day-to-day function. In addition to its impact on disability, depression may contribute indirectly to the etiology and progression of COPD through its relationship to addictive smoking. Mood disorder in adolescence and early adulthood contributes to early smoking and failure to quit, even after the onset of respiratory disease in later life. Patients with a history of major depression are more likely to fail in smoking cessation programs and to develop a major depressive episode when they do stop. This relationship calls for psychiatrically informed intervention models to improve long-term abstinence rates. The functional impairments associated with COPD are themselves potential promoters of depressive morbidity and chronicity, acting through complex causal pathways. Progressive hypoxia due to respiratory insufficiency leads to structural brain changes and neurocognitive deficits that impair day-to-day function and reduce adaptive potential; and oxygen therapy, as now practiced, offers minimal neurocognitive and mood benefits to most patients. Limited data from studies of experimental hypoxia in animals suggest that relatively mild lack of oxygen impairs the function and plasticity of critical neurotransmitter systems implicated in both cognition and mood, although current practice standards withhold oxygen therapy until late in the course of disease when the damaging effects of hypoxia on the brain have become well established. Neuropsychiatric approaches to the prevention, delay, and treatment of brain dysfunction should be a primary objective of research to improve patient outcomes. A comprehensive relational model that links pulmonary disease, hypoxia, neurocognitive impairment, and structural brain disease with depression provides a useful framework for the design of such studies. The near-term research agenda should include three components: (1) practical methods for improving physician and patient recognition of depression and neurocognitive impairment as targets for intervention; (2) additional trials of standard antidepressant treatment approaches for both major and minor depression; and (3) tests of the hypothesis that late-onset depression in patients with COPD is a marker for the presence of neurocognitive deficits and structural brain changes. The long-range research agenda must aim at preventive interventions designed to forestall brain deterioration. Controlled clinical trials of supplemental oxygen in patients with mild hypoxia and minimal cognitive deficits are needed to determine whether early treatment can reverse or moderate decline, reduce the incidence and chronicity of depression, and improve response to antidepressant treatment. Novel neuroprotective therapies such as antioxidant supplementation and modulation of monoaminergic neurotransmission, coupled with overall improvements in long-term respiratory disease management that minimize episodes of increased systemic oxidative stress, should be considered for multisite trials designed to define optimal treatment and prevention.

Psychiatric assessments of nursing home residents under OBRA-87: should PASARR be reformed? Pre-Admission Screening and Annual Review.

OBJECTIVE: As part of nursing home practice reforms, OBRA-87 mandates formal psychiatric assessments (PASARR) of nursing home residents suspected of having mental disorders, a responsibility it delegates individually to states. We describe the initial year of implementation of the PASARR process in King County, Washington, and characterize the mental disorders and mental health services needs of nursing home residents referred for psychiatric screening. DESIGN: Cross-sectional study. SETTING: The 54 Medicare-certified King County nursing homes (total beds = 7013). PARTICIPANTS: All patients referred for psychiatric evaluation under PASARR (n = 510). MEASUREMENTS: A systematic, multidimensional evaluation including a semistructured psychiatric diagnostic examination, validated measures of cognitive dysfunction, depression, and global psychopathology, functional variables relevant to need for nursing home care, and selected mental health services indicators. RESULTS: Fewer than 10% of all nursing home residents were referred for psychiatric evaluation. A primary mental illness, evenly divided between psychoses and mood disorders, was found in 60% of the sample, and a psychiatric disorder associated with dementia or mental retardation was found in 25%. Six percent had complex neuropsychiatric features defying classification, and 4% had no mental disorder. Other disorders, such as substance abuse, were rare. Cognitive impairment and global psychopathology were prevalent in all diagnostic groups, and depressive symptoms were common even in patients without affective diagnoses. Eighty-eight percent of the sample were appropriately placed, based on their needs for daily care. Fifty-five percent had unmet mental health services needs. CONCLUSIONS: The PASARR referral process detected a group of seriously mentally ill, functionally disabled patients, most of whom required the level of care that nursing homes provide. Depressed and psychiatrically impaired dementia patients were underrepresented in the referral pool as measured against widely accepted prevalence figures for mental disorders in nursing home populations. The PASARR process as currently configured appears to be most efficient in identifying schizophrenic patients, who represent a small minority of nursing home residents, and the skewed sample it generates fails to provide an adequate basis for estimating overall mental health services needed in nursing homes. The PASARR process should be altered to improve referral rates for depressed and behaviorally disturbed dementia patients.

The impact of OBRA-87 on psychotropic drug prescribing in skilled nursing facilities.

OBJECTIVES: This study examined the impact of regulations established by the Omnibus Budget Reconciliation Act of 1987 (OBRA-87) on prescriptions for psychotropic drugs, and on research on their use in nursing homes. METHODS: Data were collected on drugs prescribed for residents of 39 skilled nursing facilities over the four-year period from 1989 to 1992, bracketing the implementation of OBRA-87 in the fall of 1990. Changes in prescribing patterns were analyzed by drug class, specific target medications and doses, number of drugs prescribed, and multidrug combinations. To determine the effect of OBRA-87 on research, peer-reviewed journals were searched for the number and content of publications on psychotropic drug use in skilled nursing facilities between 1980 and 1996. RESULTS: The number of prescriptions for antipsychotics, sedative antihistamines, and sedative-hypnotics decreased significantly, while prescribing of anxiolytics increased. Qualitative, but not quantitative, shifts occurred in prescriptions for antidepressant drugs, the most frequently used psychotropic medications in all years. Rates of psychotropic polypharmacy remained stable. The number of data-based publications on psychotropic drug use in nursing homes increased after implementation of OBRA-87, but few were related to the effectiveness of drug treatment. CONCLUSIONS: Implementation of OBRA-87's nursing home regulations was associated with reductions in use of drugs specifically targeted by this legislation and was a potent stimulus to research, an unanticipated benefit of legislative action. Increased use of anxiolytics, persistent prescribing of anticholinergic antidepressants, enthusiastic adoption of new agents despite a limited research database involving frail patients, and the paucity of new studies reporting data on clinical effectiveness suggest a need for targeted research on treatment outcomes to improve the care of this population.