Measuring Outcomes of Multidimensional Geriatric Assessment Programs

Paul Stolee

Although frailty in older adults may be associated with an underlying loss of complexity in many physiologic systems,¹ the clinical conditions and geriatric syndromes²^,³ that are commonly present in frail older adults are often highly complex. This clinical complexity, including the presence of multiple interacting medical and social concerns, is the challenge and also the joy of geriatrics.⁴^,⁵

Geriatric services respond to this complexity with comprehensive approaches to assessment, multidisciplinary teams, and multidimensional interventions. Although there may be widespread agreement on the need for comprehensive, multidisciplinary, and multicomponent approaches, there is less agreement on the specific elements of these approaches. It is also not always clear which specific interventions or aspects of care (or combinations thereof) make a difference for an individual patient or for groups of patients—hence, the references to the black box of geriatrics.⁶^,⁷ Clinical complexity and comorbidity have often meant that frail older adults are excluded from many clinical trials,⁸ although there have been recent efforts to rectify this.⁹^–¹¹ This exclusion is problematic in terms of the interventions being tested and results of the studies, which are not relevant or generalizable to many frail older adult patients.¹²^,¹³ Multicomponent interventions have been found to be more effective than single-component interventions for frail older patients¹⁴ but these types of programs are much more difficult to evaluate in the context of clinical trials.¹² Allore and colleagues¹⁵ have made a distinction between statistical and analytic considerations and clinical considerations in the design of such trials. Statistical or analytic considerations would suggest that one specific intervention should target a single outcome or risk factor, the basis on which power calculations are generally undertaken.⁸ Clinically, however, it makes sense for interventions to target more than one outcome or risk factor, and many interventions are likely to have overlapping effects.¹⁵ For studies of interventions for frail older patients, clinical and analytic considerations are particularly at odds.

Given the heterogeneity of the patient population and the heterogeneity of clinical interventions, it is not surprising that evidence for the effectiveness of geriatric interventions has been hard to establish. Rubenstein and Rubenstein¹⁶ have closely observed this literature over the years and have pointed out a number of factors associated with an increased likelihood of demonstrating their effectiveness. These include appropriate targeting, more intensive interventions, control over longer term management, and a usual care control group. To this list, it is suggested here that an additional consideration be added, the selection of meaningful and responsive outcome measures. The selection of appropriate outcome measures for geriatric interventions is not straightforward and has been identified as a priority for research.¹⁷^,¹⁸ In the early 1990s, a working group of the American Geriatrics Society achieved a consensus on measures appropriate for measuring outcomes of geriatric evaluation and management units.¹⁹ The consensus statement recommended 12 physical outcomes, three psychological and social functioning outcomes, and 17 outcomes related to health care utilization and cost, reflecting concerns about future implementation and funding. The number and variety of these measures reflect the multidimensional nature of geriatric care as well as its potential system impact. Although all these measures may have relevance to specialized geriatric interventions, few, if any, of these measures would be relevant for all patients. The question therefore becomes how to achieve nonarbitrary dimensionality reduction from multidimensional interventions with multidimensional outcomes.

A more recent attempt to achieve a consensus on outcome measurement for older patients was undertaken by a U.S. National Institute on Aging (NIA) expert panel in 2001.²⁰ This working group was charged to “recommend the content of a core set of well-validated universal patient-centered outcome measures that could be routinely measured and recorded widely in health care delivery”²⁰ for older persons with multiple chronic conditions. This group recommended an initial composite measure, such as the SF-36²¹ or the Patient-Reported Outcomes Measurement Information System 29-item Health Profile (PROMIS-29)²² be used, with these results forming a basis for targeting additional follow-up measures. This approach has the potential to be more feasible in routine clinical practice, but still may require a fairly large array of outcome measures. The working group was unable to achieve consensus on appropriate follow-up measures in several important assessment domains, including disease burden, cognitive function, and caregiver burden. Also, despite an intention to recommend patient-centered measures, patients were not included in the consensus process nor were measures proposed to elicit patient preferences and values, which would be fundamental to a patient-centered approach.²³

Some of the challenges associated with measuring outcomes of multidimensional geriatric interventions can be gauged by reviewing the outcome measures used in randomized controlled trials (RCTs) of these interventions. Relevant studies were identified from selected major systematic reviews and meta-analyses, beginning with the seminal meta-analysis of comprehensive geriatric assessment services published by Stuck and associates in 1993.²⁴ Other reviews included a review of studies specifically focused on outpatient geriatric assessment,²⁵ two reviews of studies focused on preventive home visits,²⁶^,²⁷ and one review that specifically targeted multicomponent interventions.¹⁵ Collectively, these reviews reported results from 56 RCTs (see Appendix Table 38-1). Outcome measures were categorized into mortality, self-rated health, health care utilization, three assessment domains (physical function, cognitive function, and psychosocial outcomes), and an “other” category. These 56 studies are summarized as follows (see Appendix Table 38-1):

• Physical function was measured in 54 studies, using 77 different measures, of which 23 were statistically significant.

• Cognitive function was measured in 32 studies, using 12 different measures, of which six were statistically significant.

• Psychosocial function was measured in 39 studies, using 43 different measures, of which 13 were statistically significant.

• Self-rated health was measured in 18 studies, using nine different approaches, of which five were statistically significant.

• Health care utilization outcomes were measured in 46 studies, using 27 different measures, of which 26 were statistically significant.

• Other outcomes were measured in 32 studies, using 31 different measures, producing statistically significant results in 14 studies.

This review illustrates several points. Geriatric services were associated with statistically significant benefits in each category of outcome measure in at least some studies, but no category of outcome was significantly improved in all studies. None of the studies reported significant improvement in all the outcomes measured. The review also highlights the range of outcomes considered meaningful and plausible for geriatric services. Mortality is a clear end point and amenable to summation and comparison in meta-analyses, but is not necessarily the most meaningful outcome for programs serving a frail clientele for whom life expectancy is limited.⁸ Indicators related to health care utilization are of great relevance to the health care system and, although they may relate to an older person’s quality of life (e.g., for some older adults their quality of life may be higher in a community setting than in a long-term care home), these are at best indirect measures of quality of life from a patient’s perspective. Within each of the other domains, there is further evidence of heterogeneity; each domain has multiple aspects, and a large variety of instruments and approaches have been used to measure these. Even within the “other” category, an outcome such as falls is itself a multifactorial syndrome.¹⁵

Geriatric Assessment Outcomes and Quality of Life Measures

The assessment domains commonly measured in geriatric intervention studies can be seen as major components of quality of life. If outcomes commonly targeted by multidimensional geriatric interventions can be considered, collectively, as a reflection of quality of life as the overarching domain of importance, a sufficiently comprehensive quality of life measure could be a good choice as an outcome measure for common use in geriatric intervention studies. A candidate measure is the SF-36,²¹ or one of its variants with subsets of items, which has been very widely used as a health-related measure of quality of life.²⁸ Unfortunately, testing of its use with older adults has not been extensive,⁸ and results of these studies have suggested that the utility of this measure with older patients may be limited.²⁹^–³² A promising measure is the EQ-5D,³³ which quantifies an individual’s health-related quality of life into a single index value and provides a descriptive profile. It has proven to be a valid, reliable, and easy to use measure.³⁴^–⁴¹ However, it has also been shown to have limitations—predominantly, ceiling effects and poor sensitivity at the top of the scale.³⁶^,³⁷^,³⁹^,⁴²^–⁴⁴ A revised five-level version (EQ-5D-5L) has shown promise in addressing these limitations.⁴⁵^–⁴⁷ A few studies have tested the EQ-5D in populations that include older subjects⁴⁰^,⁴¹^,⁴⁸^,⁴⁹; further work in this area would be welcome.

Despite some promising work in quality of life measurement, the development of any measure that could achieve wide acceptance has been hindered by the lack of a common conceptual or theoretical understanding of the meaning of quality of life and by lack of agreement on its constituent elements.⁵⁰ Spitzer has argued that the development of a gold standard measure is possible, even for a subjective construct such as quality of life: “We fail to have a Gold Standard…because no one has made it his or her primary objective to develop a Gold Standard either for measures of health status or for measures of quality of life…I believe Marilyn Bergner and her co-workers have a sufficiently long head start that they deserve support from all the rest of us.”⁵¹ Although Spitzer pointed to the work of Bergner on the Sickness Impact Profile⁵² as the best candidate for further development as a gold standard quality of life measure, Bergner turned out not to share this view: “The bitter truth is that there is no gold standard, there is unlikely ever to be one, and it is unlikely to be desirable to have one.”⁵³

Standardized Assessment Systems

Another approach that aims at providing a comprehensive assessment of health and social functioning is the use of a standardized assessment system, of which the interRAI minimum data set (RAI or MDS) assessment systems are the most prominent. The interRAI instruments are a comprehensive assessment and problem identification system developed by an international consortium of researchers.⁵⁴ The original interRAI assessment was developed for long-term care homes (MDS 2.0) in response to U.S. government regulations (Omnibus Budget Reconciliation Act of 1987) aimed at improving nursing home quality.⁵⁵ The interRAI home care assessment instrument (RAI-HC or MDS-HC)⁵⁶ has been developed for home care settings. Other versions have been developed for use in mental health, acute care, palliative care, and other settings.⁵⁷^–⁵⁹ RAI assessment items include personal items, referral information, cognition, communication and hearing, vision, mood, behavior, physical functioning, continence, disease diagnoses, preventive health measures, nutrition status, oral health, skin condition, environmental assessment, and formal and informal service use. Specific scales have been derived from RAI assessment items, including measures of activities of daily living (ADLs), cognitive impairment, depression, and pain.⁶⁰^–⁶³ Application of the RAI system has been linked with reduced institutionalization and functional decline.⁶⁴ The approach to data collection is one of best available information, which may be done by an interview or observation of the older adult via an interview of their caregiver (paid or unpaid) or through chart review. Although this approach may suggest the possibility of inconsistent data collection, it should be noted that there has been growing support for outcome measurement that incorporates a variety of perspectives, including self-report, proxy, and objective measures.⁸^,⁶⁵ Briefer screening tools have been developed as part of the RAI system, including the RAI contact assessment.⁶⁶ When articulated with the more comprehensive RAI assessments such as the MDS 2.0 and the RAI-HC, the RAI system could thus be seen as an alternative strategy to achieve the aims of the NIA working group mentioned earlier²⁰ (i.e., a screening tool followed by more in-depth assessment).

An important advantage of the interRAI system is that it allows for consistency in data collection across sites and across types of care settings; the various versions of the RAI instruments use similar questions and data collection approaches. This advantage is particularly strong when contrasted to the alternative practice of trying to achieve consensus on the battery of measures that should be used in clinical practice and outcome evaluation. Even if a particular group achieves consensus on a set of tools (e.g., as noted by Dickinson⁶⁷), another group is likely to agree on a different set (e.g., as noted by Pepersack⁶⁸), and it is unlikely that all members of either group will be consistent in their use of the prescribed measures.

A limitation of the interRAI assessment systems is the same as for other approaches aiming to achieve a comprehensive, multidimensional assessment; not all the assessment areas will point to relevant clinical outcomes for all patients, and it would still be necessary to identify the specific outcomes of interest for a specific intervention or for a specific patient. In the interRAI system, this is addressed to some extent through the use of triggers used to identify issues warranting further investigation, referred to as resident assessment protocols (RAPs)⁶⁹ or clinical assessment protocols (CAPs).⁷⁰

Individualized Outcome Assessment and Patient-Centered Care

The inadequacies of outcome measures have often been suggested as a possible explanation for negative or ambiguous results of intervention trials. This is illustrated in the following comments from several studies:

• “The fact that we observed no significant differences in the prevention of decline in activities of daily living or cognitive function in our study may be explained in several ways … (including) … insensitivity of our outcome measures to improvements that did occur.”⁷¹

• “There might, however, have been positive effects which we could not detect. Our measurements on the health state may not have been sensitive enough to show relevant effects.”⁷²

• “The outcome variables may have been wrongly chosen to measure the effects of this kind of program.”⁷³

• “Common measures of disability may be insensitive to change in the outpatient setting of the day hospital.”⁷⁴

• “For most of the published programs, efficacy was tested on questionable indicators (e.g., mortality, health services use), on a crude proxy for functional decline (e.g., admission to a nursing home) or using a global unresponsive measure of functional autonomy.”⁷⁵

• “It is possible that the measures we used to evaluate health-related quality of life lacked sufficient sensitivity.”⁷⁶

A point made clear in Appendix Table 38-1 is the lack of consensus and consistency in the selection of outcome measures for geriatric interventions. The heterogeneous and individualized nature of geriatric programs and their patients makes such a consensus unlikely. Williams⁷⁷ has argued strongly for the individualized nature of geriatric care:

It is clear, first, that there are immense individual differences among older people, more than at any earlier age, in virtually all types of characteristics—physical, mental, health, socioeconomic. Thus when we consider what quality of life means to an older person and what features of quality of care may contribute to that quality of life, we must arrive at highly individualized conclusions. This principle is of course recommended for all ages, but it may not be so essential in some aspects of earlier life as it is in the lives of older people.

One attempt to reflect the individualized nature of older adults in outcome measurement is the use of clinical judgment, with such measures as the Clinical Global Impression ⁷⁸ or Clinician Interview-Based Impression.⁷⁹ These approaches allow a clinically experienced rater to reflect individual characteristics and health concerns in an overall assessment of improvement. They provide a role for informed clinical judgment in outcome assessment, but do not provide details on the specific aspects of a patient’s health or quality of life that may have been improved as the result of an intervention.

The individualized nature of geriatric care can also be addressed through individualized outcome measures. These could also be used to reflect individual patient preference, goals, and values in a manner consistent with a patient-centered care approach.⁸⁰ Individualized outcome measures allow for specific measurement domains to be selected that are most relevant for individual patients. Individualized measures can be used to generate clinical insights into the nature of the effects of geriatric interventions, and particularly into understanding the effects of Alzheimer disease treatment⁸¹:

To the extent that standard measures do not record ways in which important improvements or deteriorations occur, they miss an opportunity to enhance our understanding of what Alzheimer’s disease looks like when it gets better, and to provide clinical correlates of supposed pharmacologic changes. In this regard, I believe that the developments of individualized outcome measures may provide some useful insights into patterns of clinically important changes and heterogeneous disease conditions.⁸²

A number of fully or semi-individualized measures have been developed for use in a variety of settings.⁸³ The most widely known of these is likely Goal Attainment Scaling (GAS), which was proposed by Kiresuk and Sherman in the 1960s as a tool for evaluating human service and mental health programs.⁸⁴ GAS is an individualized goal setting and measurement approach that enables users to individualize goals to the needs, concerns, and wishes of a specific patient and to individualize the scale on which attainment of these goals is measured. GAS accommodates multiple individualized goals and also permits calculation of an overall score that enables comparisons among individuals or groups of patients. GAS differs from other individualized measures in two important respects. First, GAS allows for the individualization of the scales on which goals are measured, as well as the goals. Second, GAS requires a judgment to be made at the beginning of treatment on the level of goal attainment that will be considered to be a successful outcome—rather than, for example, subjectively rating achievement of outcome on 10-point scales, as in the Canadian Occupational Performance Measure.⁸⁵

Individual goals are scaled on a five-point rating scale of expected outcomes: −2, much less than expected; −1, somewhat less than expected; 0, expected level (program goal); +1, somewhat better than expected; and +2, much better than expected. The steps to construct a GAS follow-up guide are detailed in Box 38-1. An example follow-up guide is provided in Table 38-1. Goals can be weighted in terms of their relative importance, although equally weighted goals are generally recommended.⁸⁶ A summary goal attainment score (T score) allows comparison of outcomes for different patients or for groups of patients. GAS scores for a large group of patients are expected to have a mean of 50 and a standard deviation of 10. The GAS score can be calculated using a formula⁸⁴ or looked up in a table (e.g., see Zaza and coworkers’ study⁸⁷) if goals are unweighted. A standardized menu approach has been proposed as a means to facilitate goal setting.⁸⁸

Box 38-1

Developing a Goal Attainment Scaling (GAS) Follow-up Guide

1. Identify the issues that will be the focus of treatment.

• Focus on problems that are important to the patient and that the intervention is expected to change.

2. Translate the selected problems into goals—aim for at least three.

• It must be possible to observe or elicit the patient’s level of attainment on these goals at the time of follow-up.

3. Choose a brief title for each goal.

4. Select an indicator for each goal.

• The indicator is the behavior or state that clearly represents the goal and can be used to indicate progress in meeting the goal.

5. Specify the expected level of outcome for the goal.

• Predict the status of the patient on the selected goal at the end of treatment or at a prespecified follow-up time.

6. Describe the patient’s current status in relation to the goal indicator.

• Typically this is at the “somewhat less” or “much less” level.

• This is usually designated with a check mark on the guide.

7. Specify the remaining “somewhat better” and “somewhat less” than expected levels of outcome for the goal.

• These are more or less likely but still realistically attainable outcomes.

8. Specify the remaining “much better” and “much less” than expected levels of outcome.

• These are achievable, still realistic limits of the indicator.

• These represent outcomes that might be expected 5% to 10% of the time.

9. Repeat scaling steps for each of the goals.

• Try not to skip any of the five levels for each goal.

10. Although GAS is an individualized approach, descriptors, items, or scores from standardized measures may be useful in scaling some GAS goals.

11. On follow-up, rate the level for each goal that best reflects the patient’s current state. This is usually designated with an asterisk on the guide.

12. Determine the GAS follow-up score (see Table 38-1).

TABLE 38-1

Goal Attainment Scaling Follow-Up Guide

Attainment Level	Score	Mobility	Activities of Daily Living (ADLs)	Future Care Arrangements
Much better than expected	+2	>200 yards with walker, or independent with cane	Independent in ADLs and instrumental ADLs	Home with no need for home support
Somewhat better than expected	+1	Independent with walker (100-200 yards)*	Independent in ADLs and instrumental ADLs, except outside activities	Home with weekly home support
Expected level (program goal)	0	Independent with walker, limited distance (<100 yards)	Independent with ADLs, needs help with meal preparation, housework, and transportation*	Home with home support 2-3 times/wk*
Somewhat less than expected	−1	Walker with assistance✓	Dependent in ADLs, except dressing✓	Discharged to nursing home
Much less than expected	−2	Bedfast	Dependent in ADLs, including dressing	On rehabilitation unit > 7 wk✓
Comment				Patient does not wish nursing home placement

The first published use of GAS in geriatrics was in 1992.⁸⁹ Since then, the measurement properties of GAS in geriatric settings have been tested in a number of studies.⁹⁰ GAS has been found to have good interrater reliability (intraclass correlation coefficients of 0.87 to 0.93⁸⁹^,⁹¹^,⁹²) and to correlate with standardized measures such as the Barthel index and with global ratings.⁹² Of particular significance for outcome measurement in geriatrics is that GAS has consistently been found to be very responsive to change. This has been demonstrated in before and after studies, including a multisite study and in the context of a randomized controlled trial.⁹¹^–⁹⁶ GAS has been used as an outcome measure in randomized trials of a geriatric assessment team and an antidementia medication.⁹⁷^–⁹⁹ In both cases, GAS measured statistically significant benefits of the intervention. The clinical utility of GAS in geriatrics has been assessed using qualitative methods.¹⁰⁰

GAS is a measure that seems to be a particularly strong fit for the measurement needs and constraints of multidimensional geriatric interventions. It has potential as a research measure and a clinical tool. Although goal priorities may differ among patients, caregivers, and clinicians,¹⁰¹^,¹⁰² involving diverse perspectives can generate rich insights into the interventions that will most benefit older patients and into the effects of these interventions.

Reuben and Tinetti have recently recommended that a goal-oriented approach, such as GAS, be used for patient-centered outcome measurement.⁸⁰ They argued that such an approach facilitates decision making for patients with multiple conditions and aligns decision making with individual goals rather than universally desired health outcomes. This position can be contrasted with the aim of the NIA working group to achieve a set of universally applied measures.²⁰

In addition to the measurement of individual health outcomes, patient-centered care approaches are increasingly concerned with active engagement of patients in their care and with the measurement and understanding of patients’ experiences of care.¹⁰³^–¹⁰⁵ Measurement of patient experience can yield insights that are valuable in identifying priorities for quality improvement.¹⁰⁶ Older patients and their families often feel disengaged in their care, and greater understanding and measurement of their experiences may lead to improved quality and outcomes of geriatric services.¹⁰⁷^–¹⁰⁹

Conclusion

Measuring the outcomes of multidimensional geriatric interventions presents significant challenges. These challenges have resulted in frail older patients often being excluded from studies of interventions from which they might benefit and in potential benefits of geriatric interventions not being detected by the measures used. After 30 years of controlled trials in geriatrics, it seems unlikely or perhaps even inappropriate that consensus will be achieved on a set of standardized measures that will have wide applicability. Application of a universal set of outcome measures may also compromise efforts to achieve patient-centered care approaches that reflect individual patient preferences and needs. For multidimensional geriatric interventions, goal setting and outcome measurement thus need to balance the values of patient-centered care with the benefits of consistent data collection.

For consistency of data collection and to provide comprehensive assessment information, there is a strong rationale to move toward standardized health information systems, such as the interRAI. In measuring outcomes, GAS is an effective, clinically useful, and patient-centered approach to addressing the challenges of outcome measures for heterogeneous, frail older adults.

Acknowledgments

I am grateful to Sarah Meyer and Miranda McDermott for assistance in reviewing background literature for the development of this paper.

Key Points

• Geriatric services respond to the clinical complexity of frail older adults with comprehensive approaches to assessment, multidisciplinary teams, and multidimensional interventions. This complexity presents challenges for evaluation and outcome measurement.

• A review of randomized controlled trials of multidimensional geriatric interventions has revealed a wide range of targeted outcomes and a lack of consensus on appropriate measures.

• Efforts to obtain consensus on appropriate outcome measures for geriatric interventions have resulted in recommendations for a wide range of measures, but have not achieved consensus on some important domains and have arguably not reflected a patient-centered approach.

• Identification or development of a widely accepted quality of life measure has been hindered by the lack of a common conceptual understanding of the meaning of quality of life and its constituent elements.

• Standardized assessment systems, such as those developed by interRAI, have shown promise in allowing consistent collection of comprehensive health information across care settings.

• The individualized nature of geriatric care of a heterogeneous population of older patients suggests a role for individualized and patient-centered measures, such as Goal Attainment Scaling.

Appendix

APPENDIX TABLE 38-1

Randomized Controlled Trials of Geriatric Interventions and Associated Outcome Measures

Study	Setting	Study Description (Duration, No. of Subjects)	Outcome Measures
Study	Setting	Study Description (Duration, No. of Subjects)	Physical Function	Cognitive Function	Psychosocial	Self-Rated Health	Mortality	Health Care Utilization	Other
Allen et al, 1986¹	IGCS (United States)	1 yr, N = 185; evaluated whether a geriatric consultation service (GCS) can provide additional input into patient care and strategies that improve compliance to this input	Katz Index of activities of daily living,² Older American’s Resources and Services (OARS), instrumental activities of daily living (IADLs) scale³	Pfeiffer short portable mental status questionnaire⁴	Center for Epidemiologic Studies Depression Scale (CES-D)⁵			Admitting service used, number of days in institution	Veterans Alcoholism Screening Test,⁶ time of yr of consultation, number of medical problems/patient, compliance rates of recommendation; direct discussion with house staff led to increased compliance in intervention group (P = .0030)
Alessi et al, 1997⁷	HAS (United States)	3 yr, N = 202; measured the process of comprehensive geriatric assessment (CGA) and determined: (1) major findings in CGA; (2) emergence of annual clinical yield of CGA; and (3) factors that affect patient adherence with recommendations	Oral health assessment,⁸ vision and hearing test,⁹ gait and balance assessment,¹⁰ functional status assessment,¹¹ hematocrit and glucose testing, urinalysis, fecal occult blood testing	Kahn-Goldfarb mental status questionnaire¹²	Social assessment, Geriatric Depression Scale (GDS)¹³				Percentage of ideal body weight,¹⁴ medication review,¹⁵ environmental assessment, adherence to recommendations; subjects more likely to adhere to recommendations involving referral to a physician than a nonphysician professional, for community service, or for recommendations involving self-care activities (P < .001)
Applegate et al, 1990¹⁶	GEMU (United States)	1 yr, N = 155; evaluated whether care for older patients in a geriatric assessment unit would affect their function, rate of institutionalization, and mortality	Self-reported ability to perform physical activities,¹⁷ performance on timed physical tests,¹⁸ self-reported ADLs showed significant improvement in study group than control in first 6 mo (P < .05)	Folstein Mini-Mental State examination (MMSE)¹⁹	CES-D⁵	Acute Physiology and Chronic Health Evaluation (APACHE) II score²⁰	Control group patients at lower risk of immediate nursing home placement had significantly higher mortality at 6 mo (95% CI, 1.2 to 15.2; P < .05); was no significance in higher risk stratum	After 6 wk, significantly fewer study patients living in an institution (P < .01); no significance at 6 mo; significantly fewer study patients institutionalized at 1 yr (P < .05), risk of nursing home admission 3.3 times higher in control group (95% CI, 2.6 to 3.8; P < .001), study group spent more days in rehabilitation than control group (P < .0001)
Beyth et al, 2000²¹	GEMU (United States)	6 mo; N = 325; studied the effectiveness of multicomponent management program of warfarin therapy and warfarin-related major bleeding in older patients	Recurrent venous thromboembolism, therapeutic control of anticoagulant therapy measured by patient-time approach²² and INR²³; intervention group within therapeutic range at each time period significantly more often than controls (P < .001)				No significant difference between groups		Bleeding Severity Index²⁴ showed significantly more incidence of bleeding in control group at 1, 3, and 6 mo (P = .0498)
Boult et al, 2001²⁵	OAS (United States)	18 mo, N = 568; studied the effectiveness and costs of geriatric evaluation management (GEM) in preventing disability	Bed disability days (BDDs), restricted activity days (RASs),²⁶ sickness impact profile (SIP): physical functioning dimension²⁷; treatment group lost less function after 12 and 18 mo (aOR = .67; 95% CI, 0.47-0.99), had fewer health-related restrictions in ADLs (aOR = .60; 95% CI = .37-.96)		GDS²⁸; treatment group less depressed at 12 mo (P < .01) and 18 mo (P < .01)	Individual questions on general health²⁹	No significant difference	Costs, Medicare expenditure, individual questions on use of nursing home and home health services; treatment group used less home health services (aOR = .60; 95% CI, 0.37-0.92)
Burns et al, 2000³⁰	OAS (United States)	2 yr; N = 98; aimed at comparing the effectiveness of long-term primary care management by interdisciplinary geriatric team	Katz Index,² instrumental ADL (IADL) deficits ³¹ significantly better in GEM group at 1 yr (P = .006),¹¹ study subjects showed improvement in Rand general well-being inventory³² (P = .001)	Study group showed increase in MMSE¹⁹ at 2 yr (P = .025)	Study group showed improvement on perceived global social activity (GSA)³³^–³⁵ (P = .001), on CES-D⁵ (P = .003), and on perceived global life satisfaction (GLS) scale³³^–³⁵ (P < .001) at 2 yr	Study subjects showed improvement on global health perception³³^–³⁵ (P = .001) at 2 yr	No significant difference	Number of days in institution, study subjects had smaller increases in number of clinical visits (P = .019) at 2 yr
Carpenter et al, 1990³⁶	HAS (United Kingdom)	3 yr, N = 539; tested benefits of regular surveillance on older adults living at home	Winchester Disability Rating Scale³⁶				No significant difference	Geriatric and psychogeriatric community support services, primary health care team contacts, use of community support services, control group spent 33% more days in an institution than study group (P = .03)	Falls doubled in control group but remained unchanged in study group (P < 0.05)
Clarke et al, 1992³⁷	HAS (United Kingdom)	3 yr, N = 523; tested the effect of social intervention in terms of mortality and morbidity on older adults living alone	ADLs³⁸	Measure of cognitive impairment and simple screening tool for dementia³⁹	Wenger scale (measure of support networks),⁴⁰ Wenger modification of the Philadelphia Geriatric Morale Scale,⁴⁰^,⁴¹ social contact score⁴²	Perceived health status significantly greater in treatment group^†	No significant difference
Cohen et al, 2002⁴³	GEMU/OAS (United States)	3 yr, N = 1388; assessed the effects of inpatient units and outpatient clinics on survival and functional status	Survival and quality of life with Medical Outcomes Study 36-Item Short-Form General Health Survey (MOS SF-36),⁴⁴^,⁴⁵ Katz ADLs,²^,⁴⁶ physical performance test,⁴⁷ positive effects on bodily pain at 12 mo in GEMU treatment group (P = .01)*				No significant difference	Use of health services, costs, GEMU treatment group experienced more days in hospital (P < .001)
Counsell et al, 2000⁴⁸	GEMU (United States)	3 yr, N = 1531; tested whether multicomponent intervention, called Acute Care for Elders, improved functional outcomes and the process of care in hospitalized older patients	Mobility index,⁴⁹ physical performance and mobility examination (PPME),⁵⁰ Charlson comorbidity score,⁵¹ IADLs,³¹ Katz Index² decline at 12 mo favored intervention group (P = .037); fewer intervention patients experienced composite outcome of ADL decline from baseline or nursing home placement at discharge (P = .027), persisted at 1-yr follow-up (P = .022)	Pfeiffer short portable mental status questionnaire⁴	CES-D (short form),⁵² physicians more often recognized depression in intervention group than in controls (P = .02), patient satisfaction with hospitalization⁵³ higher in intervention group (P = .001), along with caregiver satisfaction (P < .05)	Overall health status, APACHE II²⁰		Reason for hospitalization, time from admission to initiation of discharge planning, social work consultations, orders for bed rest, physical therapy consults, application of physical restraints, length of stay, costs, intervention; physicians significantly reported no difficulty getting treatment plans carried out (P = .010) and that they were often informed of useful information on discharge plans (P = .015); intervention nurses reported higher satisfaction with extent of care (P = .001) and extent to which issues were discussed (P = .001)	Medications
Epstein et al, 1990⁵⁴	HAS (United States)	1 yr, N = 600; studied effectiveness of consultative geriatric assessment and follow-up for ambulatory patients	Physical examination,⁵⁵ new diagnoses, functional impact of patient diagnosis, Katz Index,² OARS (IADLs),⁵⁶ SIP⁵⁷	MMSE¹⁹ showed significantly better cognitive function at 3 mo than controls (P < .05); those > 80 yr improved more than those who were younger (P < .05)	Social support, social activities,⁵⁸ coping style, emotional health adapted from RAND Health Insurance Study,⁵⁹ satisfaction⁶⁰ showed significant benefits for those in lowest quintile of functional health at 1 yr (P < .05)	Changes in health status, overall perceived health with adapted RAND⁶¹	No significant difference	Nursing home placement, incidence of hospitalization, costs, length of stay, office visits, use of diagnostic tests	Medications, nutrition, economic issues, environmental issues
Fabacher et al, 1994⁶²	HAS (United States)	1 yr, N = 254; examined the effectiveness of preventive home visits in improving health and function in older adults	Physical examination, health behavior inventory, gait and balance assessment,⁶³ Katz Index,⁶⁴ IADLs³¹ significantly higher in intervention group at 1 yr (P < .05)	MMSE¹⁹	GDS²⁸			Intervention group had significantly increased likelihood of having primary care physician at 1 yr (P < .05)	Environmental hazards, falls, immunization rates significantly improved in intervention group at 1 yr (P < .05); nonprescription drug use increased significantly for control group at 1 yr (P < .05)
Fretwell et al, 1990⁶⁵	IGCS (United States)	6 mo, N = 436; assessed whether early interdisciplinary geriatric assessment could prevent mental, physical, and emotional decline without increasing hospital stay or costs	Katz Index⁶⁶	MMSE⁶⁷	Zung Self-Rating Depression Scale (SDS),⁶⁸ treatment groups emotional function improved (P =.045) at 6 wk		No significant difference	Costs, number of days in institution
Gayton et al, 1987⁶⁹	IGCS (Canada)	6 mo, N = 222; evaluated effects of interdisciplinary geriatric consultation team in an acute care hospital	Barthel Index,⁷⁰ level of rehabilitation scale (LORS)⁷¹	Pfeiffer short portable mental status questionnaire⁴			No significant difference	Health care use, number of days in institution, place of residence at discharge
Gilchrist et al, 1988⁷²	GEMU (United Kingdom)	22 mo, N = 222; tested efficacy of an orthopedic geriatric unit in managing older women with proximal femoral fractures	General medical assessment, hip and chest X-ray; more patients in study group found to have new medical disorders than those in control (95% CI, 3.4 to 28.5; P < .025)	Mental function⁷³^,⁷⁴			No significant difference	Placement of patients, length of hospital stay
Gunner-Svensson et al, 1984⁷⁵	ICGS (Denmark)	11 yr, N = 343; assessed whether social medical intervention would help avoid relocation in nursing homes	Unspecified questions on somatic symptoms, functions, activities	Unspecified questions on mental condition, with emphasis on dementia	Unspecified questions on communication		No significant difference	Housing, medical contact, help in illness, relocations significantly differed in favor of intervention group for women > 80 yr, old (P < .05)	Diet, demographic information (age, gender, marital status)
Hall et al, 1992⁷⁶	HAS (Canada)	3 yr, N = 167; evaluated a local health program (LTC program of the British Columbia Ministry of Health) to assist frail older adults living at home	ADLs, chronic disease		Memorial University Happiness Scale,⁷⁷ UCLA Loneliness Scale,⁷⁸ Social Readjustment Rating Scale,⁷⁹ social support	MacMillan Health Opinion Survey,⁸⁰ health locus of control (HLC)⁸¹	Significantly higher survival rates for those in treatment group at 3 yr (P = .054)	At 2 yr, significantly more of treatment group remained at home (P = .02), and at 3 yr (P = .04)	Smoking, alcohol consumption, nutrition, number of prescription medications
Hansen et al, 1992⁸²	HHAS (Denmark)	1 yr, N = 344; evaluated nurse- and physician-led follow-up model of home visits to older patients after discharge from hospital	General medical data		Unspecified social data		No significant difference	Number of days in institutions, number of readmissions to hospital, intervention patients were admitted to nursing home significantly less than controls (P < .05) at 1-yr follow-up
Harris et al, 1991⁸³	GEMU (Australia)	1 yr, N = 267; aimed at testing differences in medical management and clinical outcome between a designated geriatric assessment unit and two general medical units	ADLs,⁸⁴ radiology and pathology tests, discharge diagnosis	MMSE¹⁹			No significant difference	Accommodation prior to hospitalization, length of admission, accommodations following discharge	Procedures performed, medications on admission and discharge showed that patients in the GAU discharged on fewer drugs (P < .04)
Hebert et al, 2001⁸⁵	HAS (Canada)	1 yr, N = 503; tested efficacy of multidimensional program aimed at functional decline of older adults	Functional Autonomy Measurement System (SMAF),⁸⁶ hearing		General well-being schedule,⁸⁷^,⁸⁸ Social Provisions Scale⁸⁹		No significant difference	Admissions, use of health services	Medications, risk of falls
Hendrikson et al, 1984⁹⁰	HAS (Denmark)	3 yr, N = 572; measured the effects of preventative community measures for older adults living at home			Social services, intervention group received more home help (P < .05)		Significantly more deaths in control group than intervention group (P < .05)	Contact with GPs, admissions into nursing home, significantly more medical calls registered to control group (P < .05), significant reduction in hospital admissions in intervention group (P < .01)
Sorensen and Sivertsen, 1988⁹¹	HAS (Denmark)	3 yr, N = 585; tested effectiveness of sociomedical intervention aimed at relieving unmet medical and social needs of older adults	ADLs and IADLS⁹²		Quality of life	Self-rated health	No significant difference	Practical help received, need for more help, number of institutionalizations
Hogan and Fox, 1990⁹³	IGCS (Canada)	1 yr, N = 132; conducted trial of geriatric consultation team in acute care setting	Improved Barthel Index⁹⁴ at 1 yr in intervention group (P < .01)	Mental status scale⁹⁵			Intervention group had improved 6-mo survival (P < .02) at 4 mo	Number of days in institution, living arrangements post discharge, referrals to hospital services
Hogan et al, 1987⁹⁶	IGCS (Canada)	1 yr, N = 113; assessed effectiveness of GCS on outcomes related to hospital stay	Barthel Index⁹⁴	Improvement in metal status score⁹⁵ in intervention group (P < .01)			Lower short-term death rates in intervention group (P < .05)	Costs, number of days in institution, number of referrals to community services at discharge higher in intervention group (P < .005), referrals to hospital services, intervention group more likely to receive in-hospital physiotherapy and occupational therapy (P < .025, P < .005, respectively)	Falls, treatment group received fewer medications at discharge (P < .05)
Inouye et al, 1999⁹⁷	GEMU (United States)	2 yr, N = 852; evaluated a multicomponent strategy for prevention of delirium in hospitalized older patients	IADLs,³¹ Katz Index,² Jaeger vision test, Whisper test,⁹⁸ APACHE II²⁰	Confusion assessment method,⁹⁹ MMSE,¹⁹ digit span test,¹⁰⁰ modified Blessed dementia rating scale¹⁰¹^,¹⁰² showed significantly less incidence of delirium (P = .02), total number of days of delirium (P = .02) and total number of episodes (P = .03) in intervention group					Adherence to intervention
Jensen et al, 2003¹⁰³	ICGS (Sweden)	45 wk, N = 362; assessed effectiveness of a multifactorial program for prevention of falls and injury on older adults with high and low levels of cognition	Hearing and vision, Barthel ADL Index,⁷⁰^,¹⁰⁴ mobility interaction fall chart,¹⁰⁵ DiffTUG (measures ability to walk and carry a glass of water)¹⁰⁶	MMSE,¹⁹ concentration					Environmental hazards, medications, falls [number of residents sustaining falls, number of falls, and time to occurrence of first fall significantly longer in high MMSE intervention group (P < .001)], fall-related injuries using Abbreviated Injury Scale¹⁰⁷ showed increased injuries in low MMSE control group (P = .006)
Kennie et al, 1988¹⁰⁸	IGCS (United Kingdom)	18 mo, N = 144; assessed whether collaborative care between orthopedic surgeons and geriatric physicians could reduce various outcome measures in women with femoral fractures	Katz Index,² ADLs significantly better in treatment group (P = .005)	Pfeiffer short portable mental state questionnaire⁴	Social prognosis¹⁰⁹			Significantly fewer discharges of patients in treatment group to NHS or private nursing care (P = .03), length of stay in hospital shorter in control group^†
McEwen et al, 1990¹¹⁰	HAS (United Kingdom)	20 mo, N = 296; tested effectiveness of nurse-run screening program	ADLs, McMaster health index,¹¹¹ functional and problem evaluation interview¹¹²		Nottingham health profile,¹¹³ Philadelphia Morale Scale,¹¹⁴ significantly better in test group with respect to attitude in own ageing (P < .01) and loneliness (P < .05) at 20-mo follow-up, emotional reaction (P < .05) and isolation (P < .01) perceived to be worse in control group at 20-mo follow-up		No significant difference	Contact with health and social services	Compliance with medication
Melin and Bygren, 1992¹¹⁵	HHAS (Sweden)	17 mo, N = 249; assessed impact of primary home care intervention program on patient outcomes after discharge from a short-stay hospital	Modified Katz Index,⁶⁶ IADLs¹¹⁶^,¹¹⁷ increased at follow-up in study group (P = .04), medical disorders declined in study group at 6-mo follow-up (P < .001); indoor walking,¹¹⁸ outdoor walking¹¹⁸ significantly improved in study group (P = .03)	MMSE¹⁹^,¹¹⁹	Social function ratings on activities attended, contacts made during preceding week significantly higher in study group (P = .01) at 6-mo follow-up			Number of admissions to short- term care and rehabilitative care hospitals, number of inpatient care days and outpatient care days showed that study group spent more days in home care than controls (P < .001) but fewer days in long-term hospital care than controls (P < .001)	Number of medications increased in control group at 6 month follow-up (P = .02)
Newbury et al, 2001¹²⁰	HAS (Australia)	2 yr, N = 100; measured effectiveness of nurse-led health assessment of older adults living independently at home	Hearing and vision, physical condition, Barthel Index,⁷⁰ mobility	MMSE¹⁹	Unspecified social factors, SF-36 Quality of Life Questionnaire¹²¹ and GDS-15¹³ showed significant improvement in intervention group at 1 yr (P = .032, .05, respectively)	Self-rated health	No significant difference	Housing, admission to institutions	Medication, compliance, vaccinations, alcohol and tobacco use, nutrition, number of problems in each group, number of participants with problems, number of self-reported falls showed significant improvement in intervention group (P = .033)
Pathy et al, 1992¹²²	HAS (United Kingdom)	3 yr, N = 725; evaluation of case- finding and surveillance program of older patients at home	Townsend score¹²³		Nottingham Health Profile,¹¹² Life Satisfaction Index¹²⁴	Self-rated overall health significantly higher in intervention group (P < .05)	Significantly lower in intervention group (P = .05)	Use of services [domiciliary visits less frequent in intervention group (P < .01), contact with GP, podiatrist, and attendance allowance], questions about meals on wheels and home help, hospital admissions did not differ but duration of stay shorter in younger intervention group (P < .01)
Powell and Montgomery, 1990¹²⁵	GEMU (Canada)	3 mo, N = 203; studied effectiveness of inpatient geriatric unit at a hospital	Functional activity	Cognitive function improved between discharge and home visit in intervention group^†	Depression, life satisfaction		Fewer patients died in intervention group^†	Length of stay higher in intervention group but overall admissions lower^†
Reuben et al, 1999¹²⁶	OAS (United States)	15 mo, N = 363; tested effectiveness of outpatient CGA coupled with an adherence intervention	NIA lower extremity battery,¹²⁷ functional status questionnaire,¹²⁸ MOS SF-36¹²⁹^,¹³⁰ showed change scores for treatment group in physical functioning (P = .021); RAS, and BDD¹³¹ significantly lower in treatment group (P = .006); physical performance test⁴⁷ showed treatment effect (P = .019)	MMSE,¹⁹ mental health summaries showed significant treatment effect (P = .006)	Patient satisfaction questionnaire,¹³² Perceived Efficacy in Patient-Physician Interaction scale,¹³³ treatment group benefited on social functioning scale (P = .01), and emotional well-being (P = .016) at 15 mo	Treatment group reported less pain¹²⁹ than control group (P = .043)	No significant difference		Falls
Rubenstein et al, 1984¹³⁴	GEMU (United States)	2 yr, N = 123; assessed effectiveness of geriatric evaluation unit at improving patient outcomes	IADLs,³¹ personal self-maintenance scale³¹ showed significant improvement in study patients (P < .01); almost five times as many new diagnoses were made in study group than in controls (P < .001)	Kahn-Goldfarb Mental Status Questionnaire¹²	Philadelphia Geriatric Morale Scale¹¹⁴ showed significant improvement at 1-yr follow-up for study patients (P < .05)		Mortality significantly higher in control group at 1-yr follow-up (P < .005)	Utilization costs, initial placement at discharge, significantly more study patients discharged to their home than controls (P < .05), more than twice as many controls discharged to nursing home (P < .05), study patients underwent more specialized screening examinations and consultations than controls (P < .001), at 1-yr follow-up, controls averaged more than twice as many nursing home days (P < .05)	Medications
Rubin et al, 1992¹³⁵	HHAS (United States)	1 yr, N = 200; studied effectiveness of GEM program of health care charges and Medicare	Medical history, Katz Index,⁶⁶ IADLs	Sensory and communication abilities,¹³⁶ MMSE⁵⁶	Social history, affective, and behavioral status¹³⁶			Experimental group significantly more likely to receive home health care than control (P < .01), control group had significantly greater inpatient charges (P < .03) and Medicare reimbursement (P < .005)	Medications
Rubin et al, 1993¹³⁷	OAS (United States)	1 yr, N = 200; assessed effectiveness of outpatient GEM on physical function, mental status, and well-being	Katz Index,⁶⁶ IADLs⁴⁶ showed greater improvement and less decline in treatment group at 1 yr (P = .038)		Life Satisfaction Index-Z (LSI-Z)¹³⁸	Self-perception of health status (OARS)⁵⁶ significantly higher for treatment group (P = .006); perceived less decline in health (P = .007) and less activity limitations (P = .024)	No significant difference	No significant differences between groups on long-term nursing placements
Shaw et al, 2003¹³⁹	IGCS (United Kingdom)	1 yr, N = 274; determined effectiveness of multifactorial intervention after falls in older patients with cognitive impairments and dementia	General medical examination, mobility assessment,¹⁴⁰ assessment of walking aids, feet, and footwear¹⁴¹				No significant difference	Fall-related attendance at accident and emergency department, fall-related hospital admissions	Number of falls, time to first fall, injury rates, medications, environmental hazards¹⁴²
Silverman et al, 1995¹⁴³	OAS (United States)	1 yr, N = 442; studied process and outcomes of outpatient CGAs	ADLs,³^,¹⁴⁴ Barthel Index,⁹⁴ urinary and bowel incontinence identified significantly more in study group (P < .0001)	MMSE,¹⁹ Clinical Dementia Rating (CDR) scale,¹⁴⁵ cognitive impairment identified significantly more in study group (P < .0001)	Measures of social support, patient satisfaction with care,¹⁴⁶ clinical depression and anxiety sections of diagnostic interview schedule (DIS)¹⁴⁷^,¹⁴⁸ showed significantly lessened anxiety in study group at 1 yr (P = .036); depression identified significantly more in study group (P = .0004)	Self-perceived health status		Nursing home institutionalizations	Changes in participants status, caregiver stress¹⁴⁹ significantly less at 1 yr in study group (P = .002)
Strandberg et al, 2001¹⁵⁰	OAS (Finland)	5 yr, N = 400; determined effectiveness of multifactorial prevention program for composite major cardiovascular events in older adults with atherosclerotic disease	General medical examination, cardiovascular tests (blood pressure, heart rate, and 12-lead resting ECG), blood tests, physical function,¹⁵¹ clinical events	Consortium to Establish a Registry for Alzheimer disease tool (CERAD)¹⁵²	Health-related quality of life using the 15D,¹⁵³^,¹⁵⁴ Zung questionnaire		No significant difference	Health care resource use, hospitalizations, permanent institutionalization
Stuck et al, 1995¹⁵⁵	HAS (United States)	3 yr, N = 414; evaluated effect of in-home CGA and follow-up of older adults	Geriatric Oral Health Assessment Index,⁸ balance and gait,¹⁵⁶ vision and hearing,⁹ treatment group required less assistance in basic ADLs¹¹ (P =.02) at 3 yr; IADLs,¹¹ combined basic and instrumental activities¹¹⁷^,¹⁵⁷	Kahn-Goldfarb mental status questionnaire¹²	GDS,¹³ extent of social network and quality of social support¹⁵⁸			Costs, admissions to acute care hospital, short-term nursing home admissions, significantly more visits to GP among intervention group (P = .007), permanent nursing home admission higher among control group (P = .02)	Medications, environmental hazards, percentage of ideal body weight¹⁵⁹
Stuck et al, 2000¹⁶⁰	OAS (Switzerland)	3 yr, N = 791; examined effects of preventive home visits with annual multidimensional assessments on functional status and institutionalization between high- and low-risk older persons	Gait and balance performance,⁶³ ADLs and IADLs¹¹; intervention group at low baseline less dependent in IADLs (95% CI, 0.3 to 1.0; P = .04)	MMSE¹⁹	GDS¹³	Self-perceived general health,¹⁶¹ self-reported chronic conditions		Costs, permanent nursing home admission higher in high-risk intervention group (P = .02)	Medication use¹⁵
Teasdale et al, 1983¹⁶²	GEMU (United States)	1 yr, N = 124; assessed whether a geriatric assessment unit using multidisciplinary team approach affected patient placement outcomes					No significant difference	Source of admission, placement at discharge, location 6 mo postadmission, location of patient after discharge, mean length of stay significantly higher in intervention group (P < .001)
Thomas et al, 1993¹⁶³	IGCS (United States)	1 yr, N = 120; tested effectiveness of inpatient geriatric consultation team	Functional Assessment Inventory (FAI),¹⁶⁴ physical and activities scales, Katz scale²		FAI, psychological and social scale¹⁶⁴		Significantly more patients died in control group at 6 mo (P = .01)	Referrals to community service, number of postdischarge GP visits, discharge destination, number of days in institution, control group had significantly more readmissions (P = .02)	FAI—economic scale¹⁶⁴
Timonen et al, 2002¹⁶⁵	OAS (Finland)	9 mo, N = 68; studied effects of multicomponent training program focused on strength training after hospitalization	Strength and physical performance (walking speed and Berg Balance scores) significantly improved after the intervention in the study group at 3 mo (P < 0.05). Isometric hip abduction and walking speed significantly improved after the intervention in the study group at 9 mo (P < 0.05).					Community services use	Medication
Tinetti et al, 1994¹⁶⁷	HAS (United States)	1 yr, N = 301; evaluated effect of multiple risk factor reduction on incidence of falls	Presence of chronic disease, ADLs,³¹ vision¹⁶⁸ and hearing,¹⁶⁹ Sickness Impact Profile (ambulation and mobility subscales),²⁷ risk factor for balance impairment reduced in intervention group at 1 yr (P = .003), impairment in balance and bed to chair transfers reduced (P = .001), impairment in toilet transfers reduced (P = .05)		Depressive symptoms¹⁷⁰		No significant difference	Costs, hospitalizations, number of hospital days, intervention group received more home visits (P < .001)	Room by room number of hazards for falling, falls efficacy scale¹⁷¹; at 1 yr, control group fell significantly more (P = .04), intervention group significantly reduced number of medications (P = .009)
Toseland et al, 1997¹⁷²	OAS (United States)	2 yr, N = 160; investigated effectiveness of an outpatient GEM team by examining changes in health status, health care utilization, and costs	SF-20,¹⁷³ (FIM)¹⁷⁴^–¹⁷⁶				No significant difference	Outpatient utilization (visitation of UPC/GEM clinic, medicine clinic, surgery clinic, emergency room, and total clinic visits), inpatient utilization (number of hospital admissions, hospital days of care, nursing home admissions and nursing home days of care); GEM patients used significantly fewer emergency room services (P ≤ .05), GEM patients used significantly more total outpatient clinic services (P ≤ .01), costs (total inpatient costs, total outpatient costs, nursing home costs, institutional costs, total health care costs); significantly more outpatient cost in GEM patients over 2 yr (P ≤ .05)
Tucker et al, 1984¹⁷⁷	OAS (New Zealand)	5 mo, N = 120; assessed effectiveness of day hospital in geriatric service	Significant increase in Northwick Park ADL index¹⁷⁸ at 6 wk for intervention group (P = .002)	Cognitive function¹⁷⁹	Zung index,⁶⁸ intervention group showed improved mood at 5 mo (P =.011)			Domiciliary services, day hospital costs one third more than alternative
Tulloch and Moore, 1979¹⁸⁰	OAS (United Kingdom)	2 yr, N = 295; evaluated effects of geriatric screening and surveillance program on older adults	Screening for medical disorders found significantly greater incidence in study group compared to controls (P < .01); greater proportion of medical problems unrecognized in control group (P < .001)					Rate of hospital admission, outpatient referrals significantly higher in study group (P < .01), time spent in hospital less for the study group than controls (P < .01)	Socioeconomic problems
van Haastregt et al, 2000¹⁸¹	HAS (Netherlands)	18 mo, N = 316; assessed whether multifactorial program of home visits reduces falls and mobility impairments in older adults	Physical health, control scale, and mobility range scale of SIP 68,¹⁸²^,¹⁸³ number of physical complaints, Frenchay daily activities¹⁸⁴^,¹⁸⁵	Mental health section of RAND-36¹⁸⁶^,¹⁸⁷	Social functioning,¹⁸⁸ psychosocial functioning	Perceived health by RAND-36,¹⁸⁶^,¹⁸⁷ perceived gait problems			Falls efficacy scale,¹⁷¹^,¹⁸⁹ falls, medications, environmental hazards¹⁹⁰
van Rossum et al, 1993¹⁹¹	HAS (Netherlands)	3 yr, N = 580; tested effectiveness of preventive home visits to older adults	Self-rated functional state, hearing and vision problems	Memory disturbances¹⁷⁹	Self-rated well-being,¹⁹² loneliness,¹⁹³ modified Zung index⁶⁸	Self-rated health	No significant difference	Costs, use of community and institutional care
Vetter et al, 1984¹⁹⁴	HAS (United Kingdom)	2 yr, N = 1286; evaluated effectiveness of health visitors on older population of urban (Gwent) and rural (Powys) towns	Townsend score¹²³		Mental disability,¹⁹⁵^,¹⁹⁶ use of social contacts, self-rated quality of life		Significantly more deaths in Powys (P < .01)	Use of medical and social services, Gwent intervention group used podiatrists significantly more than Powys group (P = .02), significantly more home visits for Gwent intervention group (P = .005)	Availability of caregiver, composition of household, type and quality of housing, participants in Gwent attended more lunch clubs than Powys (P < .05)
Vetter et al, 1992¹⁹⁷	HAS (United Kingdom)	4 yr, N = 674; assessed whether health visitors reduced incidence of fractures in older adults	Townsend score,¹²³ medical condition, assessment and improvement of general muscle tone						Falls and fractures, nutrition, medications, environmental hazards
Wagner et al, 1994¹⁹⁸	HAS (United States)	2 yr, N = 1559; tested multicomponent program to prevent disability and falls in older adults	Fitness test, hearing and vision, control group worsened in RAS¹⁹⁹ (P < .05), BDD¹³¹ (P < .01) and MOS²⁰⁰^,²⁰¹ (P = .05) at 1-yr follow-up			Self-rated health and practices questionnaire			Environmental hazards, alcohol consumption, medications; significantly fewer members of intervention group reported falling than control (difference = 9.3%; CI, 4.1%-14.5%) at 1-yr follow-up
Williams et al, 1987²⁰²	OAS (United States)	1 yr, N = 117; evaluated whether team-oriented assessment can improve traditional health care approaches	Functional status and medical diagnoses²⁰³^,²⁰⁴		Social supports²⁰³^,²⁰⁴			Health care use, degree of client satisfaction with evaluation, health service use behaviors
Winograd et al, 1991²⁰⁵	IGCS (United States)	1 yr, N = 197; studied effect of inpatient multidisciplinary geriatric consultation service on health care utilization and functional and mental status	Physical Self-Maintenance Scale, ADLs, IADLs	MMSE¹⁹	Philadelphia Geriatric Morale Scale¹¹⁴			Health care use
Yeo et al, 1987²⁰⁶	OAS (United States)	18 mo, N = 205; compared effects of two models of outpatient care on functional health and subjective well-being	SIP⁵⁷ showed significantly less functional decline in intervention patients (P = .029) and its physical dimension (P = .011)		Zung Self-Rating Depression Scale (SDS),⁶⁸^,²⁰⁷ Life Satisfaction Index-A (LSI-A)²⁰⁸ Affect Balance Scale (ABS),²⁰⁹ psychosocial dimension of SIP⁵⁷	Self-rated health measure²¹⁰^,²¹¹	No significant difference