Acute Care for the Elderly: A Literature Review

Abstract

Traditionally, acute medical care has been insufficient to meet the complex care needs of frail older adults. The purpose of this study was to evaluate the effectiveness of Acute Care for the Elderly (ACE) units at improving hospitalization outcomes for adults older than 65 years of age. A review of the literature was performed, focusing on randomized controlled trials, clinical trials, reviews, and meta-analyses from 1990 to 2008. This review revealed ACE to be associated with positive global outcomes (eg, cost, length of stay, readmission rates, utilization, rehabilitation, cognition, function, patient/staff satisfaction). Furthermore, some studies may point to a decreased incidence of delirium and polypharmacy. Though larger studies with consistent operational definitions and replicative studies are needed, the literature presents compelling evidence that warrants further investigation of ACE as a valuable alternative paradigm of acute geriatric care. (Population Health Management 2010;13:219–225)

Introduction

O lder adults aged ≥65 years, account for 48% of inpatient hospitalization days. Among the cohort aged ≥65 years, those older than 85 years of age are 2 times more likely to require hospitalization.¹ As those older than age 85 remain the fastest growing population, projected to comprise 25% of the US population by the year 2050, traditional acute care practices will have to evolve to fit the unique needs of the frail elderly in the years to come. Older adults are highly susceptible to the hazards of acute hospitalization, as imposed bed rest, urinary catheter utilization, excessive environmental stimulation, and other characteristics of inpatient care often deplete frail elders of already limited physiologic reserves. Though patients may recover from the illnesses that precipitate hospitalization, they are often less functionally independent and require higher levels of care or nursing home placement after discharge.

The Acute Care for the Elderly (ACE) was developed to disrupt the trajectory of functional decline often experienced by geriatric patients who are admitted to acute hospital wards. The ACE model was initiated to challenge the preconceived notion that elder patients would require posthospital rehabilitation to overcome deconditioning associated with acute hospitalization. Prevention of functional decline, “prehab,” was the focus of the interdisciplinary approach of care for older patients. The first ACE, developed at University Hospitals of Cleveland, was funded by the John A. Hartford Foundation in 1990.^2

–8 This original intervention included the following 4 components: a specialized environment, patient-centered care, medical review, and interdisciplinary team plans of care.⁵ This unit was developed to disrupt the presumably unavoidable trajectory of functional decline of geriatric patients who are admitted to acute hospital wards.

Studies of the early ACE units demonstrated improved functional outcomes and decreased placement in long-term care. Since this time, additional studies of ACE have demonstrated similarly favorable results regarding the following: costs of care, lengths of stay (LOS), patient/provider satisfaction, maintenance of functional status, and favorable discharge disposition. Significant reductions in delirium, polypharmacy, and readmissions to the hospital have also been reported as demonstrable benefits of ACE for hospitalized patients admitted from the community. The purpose of the current review is to determine whether this model contributes to positive patient care outcomes (ie, rehabilitative, cognitive, utilization, functional, pharmacological, and patient/staff satisfaction) for acutely hospitalized older adults compared to traditional medical care.

Characteristics of ACE

The concept of specialized geriatric care for acutely ill older adults is not new, and was attempted for years prior to the initiation of the ACE model.⁵ Results were mixed, probably because of the heterogeneity of study interventions and measured outcomes. Geriatric Evaluation and Management (GEM) units are geared for carefully targeted patients who are selected to transfer from acute care units to receive prolonged care to improve their functional status. The GEM model of care has been well associated with improved hospitalization outcomes, but extensive expenditures that relate to prolonged hospitalization stays are thought to make this model cost prohibitive for nonfederal hospitals.⁷ Interventions to support elder care were historically directed to Geriatric Assessment Units (GAUs) for post-acute care patients, or to geriatric consultation services for acute geriatric syndromes.⁵ GAUs differ from GEMs in that they are more similar to the ACE paradigm of care.⁷ They are units that utilize a multidisciplinary approach through consultation with a geriatrician lead team to provide specialized, comprehensive geriatric care.⁵ In contrast to ACE, however, GAUs lack the environmental modifications, nurse lead care, and formal, daily interdisciplinary team meetings that are considered the cornerstone of the ACE paradigm.⁷

Interdisciplinary teams usually consist of a combination of geriatricians, nurse case managers, nurses, a nutritionist, social workers, physical therapists, an occupational therapist, a pharmacist, and, of course, the patient and his or her family. Teams can vary according to patient needs and often may include other disciplines as well, such as pastors and psychiatrists. The purpose of an interdisciplinary team is to work collaboratively and in an integrated manner toward the goals that were agreed upon for the patient. The team members meet on a regular basis (often daily in an ACE unit) to assess the patient's health status, discuss individual team members' concerns, and develop a plan of care for the remainder of the hospital stay as well as discharge. Most teams undergo some formal training and team-building exercises to learn how to function and communicate effectively. The leader of the team varies depending on the patient's needs.

Over the years, modifications have been implemented by various institutions to facilitate the implementation of ACE in additional hospital settings. In 2003, Jayadevappa et al surveyed 16 ACE units and utilized a stepwise regression method to determine the basic characteristics of ACE units throughout the country.⁹ Most ACE units comprised 10–20 beds and had interdisciplinary teams.⁹ No mention was made of advanced practice nurses in this survey study, but the original model heavily emphasized the role of advanced practice nursing in the successful implementation of ACE interventions.^5,8,10

Methods

A comprehensive review of the literature was performed utilizing PubMed, Ovid, and Scopus databases. Search terms included the following keywords: “ACE units,” “ACE models,” and “acute care for the elderly.” Search limitations included randomized controlled trials (RCTs), clinical trials, reviews, and meta-analyses between the years of 1990–2008. Selected articles were from English language publications and were limited to the adult population ≥65 years of age who received acute hospital care on ACE units or ACE and general medicine wards.

The search yielded 462 findings, which included 421 irrelevant articles that did not mention the ACE model of care in the studies. Twenty-one of the remaining articles were excluded for the following reasons: 2 were not English language, 3 were not retrievable, and 16 were duplicate studies. Consequently, 20 studies were deemed to be the most relevant for the purposes of this review, but it is important to note the lack of heterogeneity in outcome variables and operational definitions within these studies. This review comprised the following: 5 RCTs; 3 literature reviews; 5 descriptive studies; 1 nonrandomized, retrospective, case-control design; 1 non-concurrent cohort study; 2 interventional studies; 2 survey studies; and 1 case study.

Data extraction and synthesis

Because ACE is still at a very early stage of development in the literature, the homogeneity of applicability of RCTs and clinical trials is limited. Studies were interestingly diverse in design and outcome variables of interest; however, a lack of replication of existing RCTs hindered attempts to confirm or dispute previous findings. Consequently, all studies that contributed to the measure of ACE for acutely hospitalized elders were utilized (Table 1).

Table 1.

Evidence Table for Acute Care for the Elderly (ACE) Unit versus Usual Care

Source	Study Type	Target Population	Study Goals	ACE Study End Points
	Cost
Covinsky et al¹⁴ 1997	RCT (Cleveland, OH)	Acutely hospitalized medical patients with a mean age of 80 years	To compare the cost of acute medical care between ACE and usual care	Higher initial cost of careDecreased total cost of careDecreased LOSDecreased hospital readmissionIncreased caregiver satisfaction
Covinsky et al² 1998	RCT (Cleveland, OH)	Hospitalized older adults aged ≥70 years	To determine if ACE contributed to decreased hospitalization costs and improved functional outcomes versus usual care	Decreased LOSIncreased initial costDecreased total costsDecreased functional decline
Counsell et al¹⁵ 2000	RCT (Cleveland, OH)	Hospitalized, community-dwelling adults aged ≥70 years	To determine if ACE would improve functional outcomes without increasing costs	Neutral impact on cost of careNeutral impact on LOS Decreased NH placement Decreased functional decline Increased patient/provider satisfaction
Asplund et al¹¹ 2000	RCT (Umea, Sweden)	Hospitalized older adults aged ≥70 years assigned to acute geriatrics-based ward vs. general medical wards	To determine if a specialized, geriatric approach shortened the length of stay and decreased discharges to nursing home facilities	Decreased LOS Decreased NH placement. Neutral impact on cost Neutral impact on functional status Neutral impact on readmissions Neutral impact on polypharmacy
Jayadevappa et al¹² 2006	Retrospective case-control design (Philadelphia, PA)	Hospitalized older adults with a mean age of 79.6 with an admitting diagnosis of UTI or CHF	To determine if ACE decreased cost of care and improved hospital outcomes when compared to usual care	Decreased hospital costs Decreased LOS Decreased acute care readmissions Decreased NH placement
	Interdisciplinary Teams
Kresevic et al⁸ 1998	Descriptive Study (Akron, OH)	Hospitalized older adults aged ≥65 years	To define and describe an IDT approach to caring for hospitalized older adults on ACE	IDT approach/model defined
Krevesic et al¹⁰ 1998	Descriptive Study (Cleveland, OH)	Hospitalized older adults aged ≥65 years	To describe a patient-centered approach to acute geriatric care	Patient-centered model described
Palmer⁶ 1995	Descriptive Study (Cleveland, OH)	Hospitalized older adults aged ≥65 years	To describe a model of care (ACE) that utilized an IDT approach to promote functional status	IDT model defined
Benedict et al¹⁶ 2006	Descriptive Study (Akron, OH)	Hospitalized older adults aged ≥65 years	To describe ACE as a model for multidisciplinary management of older adults and the role of clinical nurse specialists (CNS) within this context	Increased decision making at point of care More interventions to meet patient needs Increased use of EBM practices Increased IDT collaboration Increased provider satisfaction Decreased hospital readmissions
	Functional Outcomes
Palmer et al⁵ 1994	Literature Review (Cleveland, OH)	Hospitalized older adults aged ≥65 years	To determine which characteristics contributed to improvement of functional decline for hospitalized older adults on ACE	Description of ACE characteristics Description of ACE model
Landefeld et al³ 1995	RCT (Cleveland, OH)	Hospitalized older adults aged ≥70 years and at teaching hospital; randomized to a specialized geriatric unit versus usual care	To determine whether ACE contributed to improvement of functional mobility and performance of activities of daily living when compared to usual care	Decreased cost of care Decreased LOS Decreased NH placement at discharge Decreased functional decline
Palmer et al⁴ 1998	Literature Review (Cleveland, OH)	Hospitalized older adults aged 65 plus years.	To determine which standards of care and nurse initiated strategies contribute to prevention and potential restoration of functional decline for hospitalized older adults and develop a model of care	Description of ACE model of care
Kresevic et al⁸ 1998	Literature Review (Cleveland, OH)	Hospitalized older adults aged ≥65 years	To determine which standards of care and nurse initiated strategies contribute to prevention and potential restoration of functional decline for hospitalized older adults and develop a model of care	Description of ACE model of care
Miller¹⁷ 2002	Case Study (Newark, NJ)	Older adults aged ≥65 years with multiple comorbid conditions and acute hospitalization	To describe an experience of 1 elderly patient hospitalized for pneumonia and transferred for care to an ACE unit	Case Study revealed: Reduced delirium Decreased functional decline Decreased polypharmacy Reversed unnecessary NH placement
Palmer et al⁷ 2003	Descriptive Study (Cleveland, OH)	Hospitalized older adults aged ≥65 years	To define and describe an IDT approach to caring for patients on ACE	IDT model described
Allen et al¹³ 2003	Interventional Study (Akron, OH)	Hospitalized older adults aged ≥65 years and admitted to a stroke unit	To determine if the intervention of the ACE Stroke Unit (ACE SU) contributed to decreased LOS/mortality, and increased functional status of acute stroke patients versus usual care	SU model described Decreased LOS Increased provider satisfaction Decreased hospital readmissions
	Pharmacological Outcomes
Edwards et al²⁰ 2003	Non-concurrent cohort study (Winston-Salem, NC)	Hospitalized older adults aged ≥65 years and admitted to ACE	To determine if potentially inappropriate medications (PIMs) were more or less likely to be prescribed or discontinued on ACE units	Decreased use of PIMs Increased incidence of discontinued PIMs
	Delirium
Flaherty et al¹⁹ 2003	Observational/Interventional Study (St. Louis, Missouri)	Patients on ACE unit with mean age of 82.6 years admitted to a 4-bed, restraint-free unit	To determine whether the ACE model for delirium would promote care and decrease incidence of delirium for hospitalized older adults at risk	Description of ACE model of delirium Decreased prevalence/duration of delirium Decreased LOS
	Staffing/Characteristics
Siegler et al¹⁸ 2002	Survey Study (New York)	ACE units throughout the country that care for hospitalized, older adults aged ≥65 years	To determine what staffing and continuous quality improvement data were utilized at ACE sites	Staffing description on existing ACE units
Jayadevappa et al⁹ 2003	Survey Study (Philadelphia, PA)	Operational ACE Units (16) since the 1990s	To determine both the prevalence and characteristics of ACE units, as well as the characteristics of hospitals that were associated with these units	Characteristics of existing ACE units

CHF, congestive heart failure; EBM, evidence-based medicine; IDT, interdisciplinary teams; LOS, length of stay; NH, nursing home; RCT, randomized controlled trial; UTI, urinary tract infection.

Results

Lengths of stay/hospital costs

The ACE model was reported to reduce hospital acute care days.^2,3,11
–13 The average reduction for inpatient LOS was typically 1 day compared to similar cohorts treated on general medicine wards. Only 1 study demonstrated neutral results in the determination of comparative LOS for ACE unit and general medicine unit patients and none showed increased LOS.¹² Factors that contributed favorably to resource utilization also included increased discharges to home versus long-term care, decreased acute care stays, and reduced hospital readmissions. Subsequently, despite higher initial costs, ^2,13,14 the total cost of care for patients on ACE units was statistically significant and demonstrably less when compared to usual care.^3,12

Hospital readmission

Hospital readmission data yielded statistically significant, positive outcomes for ACE units. The ACE model is well reported in the literature to reduce overall costs and LOS; trends toward reduced acute hospital readmissions are also reported.^{2,12,13,15,16} Only 2 studies reported neutral findings regarding readmissions for the intervention group.^3,14 Decreased readmission rates combined with a decreased length of stay contribute to reduced overall health care costs. As a result, patients do not require more frequent acute care admissions.¹² Data from these studies suggest that patients discharged from ACE were presumably more physically, functionally, and socially prepared to leave the hospital than patients who were treated on general medical units.

Nursing home placement

Nursing home placement is more often associated with caregiver burnout than patient disposition.¹⁰ The ACE model likely decreases caregiver burden by maintaining function and has been associated with decreased institutionalization compared to usual care on general medicine wards.^{2,3,11,14,15,17} Most RCTs included in this review (3 of 5) evaluated the prevalence of discharges to nursing homes and reported statistically significant reductions in long-term care placement among the ACE Unit intervention group.^3,11,15 Landefeld et al reported a significant reduction in discharges to long-term care among patients who received ACE Unit intervention versus usual care (14% versus 22%, respectively).³

Functional decline

Prevention of functional decline is one of the major thrusts of ACE units. The majority of studies cited the intervention group to have significantly less functional decline in activities of daily living.^2,3,15,17 Landefeld et al reported statistically significant findings of overall health status at discharge for patients assigned to the ACE intervention (P < 0.001), which most notably included improvement in the ability to perform instrumental activities of daily living (P = 0.06) and improved ambulation (P = 0.04).³ Asplund et al determined the ACE intervention group to have similar or neutral results compared to general medical care.¹¹

Delirium

The ACE model of care was designed to reduce the risk of functional decline, delirium, and iatrogenic illness^{5,6,8

–13,16
–18}; 3 studies (ie, 1 RCT, 1 intervention study, 1 case study) evaluated the effect of ACE on acutely hospitalized older adults at risk for or diagnosed with delirium.^3,4,13 Asplund et al reported mixed results of the prevalence of delirium for patients admitted to acute geriatric wards versus medical wards. Initially, the prevalence was less in the intervention group (4.3% versus 5.0%), but later development of delirium was more prevalent in this group (3.3% versus 1.9%).¹¹ Flaherty et al performed an intervention study to decrease the prevalence and duration of delirium. Specific modifications for this intervention included the following: environmental modifications, interdisciplinary team meetings, focused geriatric principles of care, avoidance of physical and chemical restraints, and utilization of sitters.¹⁹ This study yielded statistically robust reductions of delirium prevalence and duration for older adults cared for on ACE units versus general medicine wards.¹⁹ Miller reported a positive outcomes case study that demonstrated reversibility of functional decline, delirium, and a potentially inappropriate discharge to long-term care.¹⁷

Polypharmacy

Polypharmacy is a well-known contributor to acute hospitalization, functional impairment, and cognitive decline, but few studies have evaluated the effects of ACE intervention on inpatient pharmacologic practices (eg, prevalence of potentially inappropriate medications for elders [PIMs], polypharmacy, prevalence of unnecessary medications). Results yielded 2 positive outcome studies^17,20 and 2 neutral studies^11,12 that examined the effect of ACE versus usual care on medicine wards regarding polypharmacy. A noncohort, retrospective, non-RCT hypothesized that the ACE model would be less associated with polypharmacy and more associated with medication reduction and avoidance of PIMs when compared with usual care; however, these results were not statistically significant.²⁰ Another study, an RCT, determined that patients in the intervention group had a lower mean number of drugs when compared to patients on medical wards (4.8 versus 5.2 with a 95% CI).¹⁹ Though this was quantitatively a marginal reduction in the number of prescriptions between groups, this trend persisted up to 3 months post hospital discharge for the intervention group.

Patient/provider satisfaction

Four of the 18 studies included in this review evaluated patient and provider satisfaction with ACE intervention inpatient care. Two of 5 RCTs reviewed compared ACE units to general medical wards and specifically measured patient, provider, or caregiver satisfaction between groups.^2,12 One study with strong methodology, by Counsell et al, surveyed 1531 patients who were divided into 2 groups: an intervention group (n = 767) and a usual care group (n = 764).¹² The patients who received care on ACE units, as opposed to general medical units, and participated in surveys 1 month post hospital discharge reported superior overall satisfaction with care compared to prior hospitalization experiences (40% versus 26%, P < 0.001).

Provider surveys also yielded favorable satisfaction scores for the ACE intervention group versus usual care.^13,15,16 For example, nurses reported being very satisfied that geriatric patients were receiving needed care (64% versus 27%).¹⁵ Of the 2 descriptive studies, surveys demonstrated improved provider satisfaction rates for nurses and physicians in the intervention group versus usual care.^13,16 Surprisingly, 70% of physicians surveyed rated geriatric care on an ACE Unit to be 4.6 on a Likert scale (with 5 being “high satisfaction” and 1 being “low satisfaction”).¹⁶ Additionally, all included studies that specified satisfaction scores for patients, caregivers, and families identified ACE as superior compared to usual care.^4,13,15,17

Discussion

In this review of the literature, we attempted to ascertain an estimate of the prospective value of expansion and further implementation of the ACE paradigm of care. The results of studies of the ACE model have been primarily positive, with a few neutral global outcomes (Table 2). All studies reported reduced total costs of care for patients who received interventional care.^2,3,11,14,15 Lengths of stay were not reported to increase ^14,15,19 and were more often reduced for hospitalized older adults compared to general medical care.^2,3,11
–13 Despite decreased costs of care and LOS, readmissions to acute care hospitals did not increase; readmissions were predominantly lower for patients discharged from ACE units.^3,11

–14

Table 2.

Overall Summary of Studies: Acute Care for the Elderly (ACE) versus Usual-Care

Outcome Variables	Increased	Decreased	Neutral
Cost^*	2,14	11,15	11,15
Length of stay		2,3,11–13	12
Hospital readmission		12,13	3,11
Nursing home placement		2,3,11,14,15,17
Functional decline		2–4,15,17	11
Delirium		13,19	3

Increased initial costs were offset by decreased LOS, resulting in decreased total cost of care

Most studies report reduced functional decline for acutely ill patients discharged from the intervention group, and this is increasingly important in the wake of the new methodology of paying for performance based on quality measures defined by Medicare. Frail elders cost hospitals more, have extended LOS, and are more likely to require more frequent hospital readmissions if their global needs are not addressed prior to discharge. The traditional approach to acute care is inadequate to support these needs; although patients may recover from illnesses that precipitate hospitalization, they are often less functionally independent and require a higher level of care upon discharge.

There were mixed results concerning the reduction and prevalence of delirium:^4,11,19 only 1 study was designed to specifically address delirium as a primary outcome variable,¹⁹ and neither of the 2 remaining studies focused on delirium as a primary outcome variable of interest.^4,11 Nonetheless, existing evidence suggests that the ACE model contributes to reductions in polypharmacy.^11,12,17,20 Finally, satisfaction scores for patients, providers, and caregivers serve as a reflection of quality care that is potentially unbiased by researcher opinion or statistical analysis. All studies that reported satisfaction survey results reported superior evaluations for the ACE intervention group compared to the usual care group.^4,12,16,17

Limitations of this study include the scarcity in duplicated results and homogeneity in operational definitions and outcome variables associated with this relatively new model of care. This underscores the need for replication in future research to confirm or dispel significant findings. Additionally, publication bias is another potential weakness of literature reviews, as neutral results are less likely to be published in journals than impressive positive or negative outcome studies. More studies are needed to delineate prescribing practices on ACE units and to identify how these practices contribute to the functional prognosis of acutely ill older adults.

Footnotes

Author Disclosure Statement

Dr. Ahmed and Ms. Pearce disclosed no financial conflicts of interest. Dr. Ahmed's salary is supported by a Geriatric Academic Career Award (GACA grant).

References

Palmer

, Meldon

. Acute care. Hazzard

, Blass

, Halter

, Ouslander

, Tinetti

. Principles of Geriatric Medicine & Gerontology, 5th. New York: McGraw-Hill, 2003.

Covinsky

, Palmer

, Kresevic

et al. Improving functional outcomes in older patients: Lessons from an acute care for elders unit. Jt Comm J Qual Improv, 1998; 24:63–76.

Landefeld

, Palmer

, Kresevic

, Fortinsky

, Kowal

. A randomized trial of care in a hospital medical unit especially designed to improve the functional outcomes of acutely ill older patients. N Engl J Med, 1995; 332:1338–1344.

Palmer

, Counsell

, Landefeld

. Clinical intervention trials: The ACE unit. Clin Geriatr Med, 1998; 14:831–849.

Palmer

, Landefeld

, Kresevic

, Kowal

. A medical unit for the acute care of the elderly. J Am Geriatr Soc, 1994; 42:545–552.

Palmer

. Acute hospital care of the elderly: Minimizing the risk of functional decline. Cleve Clin J Med, 1995; 62:117–128.

Palmer

. Acute care for elders: Practical considerations for optimizing health outcomes. Dis Manag Health Outcomes, 2003; 11:507–517.

Kresevic

, Counsell

, Covinsky

et al. A patient-centered model of acute care for elders. Nurs Clin North Am, 1998; 33:515–527.

Jayadevappa

, Bloom

, Raziano

, Lavizzo–Mourey

. Dissemination and characteristics of acute care for elders (ACE) units in the United States. Int J Technol Assess Health Care, 2003; 19:220–227.

10.

Kresevic

, Holder

. Interdisciplinary care. Clin Geriatr Med, 1998; 14:787–798.

11.

Asplund

, Gustafson

, Jacobsson

et al. Geriatric-based versus general wards for older acute medical patients: A randomized comparison of outcomes and use of resources. J Am Geriatr Soc, 2000; 48:1381–1388.

12.

Jayadevappa

, Chhatre

, Weiner

, Raziano

. Health resource utilization and medical care cost of acute care elderly unit patients. Value Health, 2006; 9:186–192.

13.

Allen

, Hazelett

, Palmer

et al. Developing a stroke unit using the acute care for elders intervention and model of care. J Am Geriatr Soc, 2003; 51:1660–1667.

14.

Covinsky

, King

Jr , Quinn

et al. Do acute care for elders units increase hospital costs? A cost analysis using the hospital perspective. J Am Geriatr Soc, 1997; 45:729–734.

15.

Counsell

, Holder

, Liebenauer

et al. Effects of a multicomponent intervention on functional outcomes and process of care in hospitalized older patients: A randomized controlled trial of Acute Care for Elders (ACE) in a community hospital. J Am Geriatr Soc, 2000; 48:1572–1581.

16.

Benedict

, Robinson

, Holder

. Clinical nurse specialist practice within the Acute Care for Elders interdisciplinary team model. Clin Nurse Spec, 2006; 20:248–251.

17.

Miller

. Acute care of the elderly units: A positive outcomes case study. AACN Clin Issues, 2002; 13:34–42.

18.

Siegler

, Glick

, Lee

. Optimal staffing for Acute Care of the Elderly (ACE) units. Geriatr Nurs, 2002; 23:152–155.

19.

Flaherty

, Tariq

, Raghavan

, Bakshi

, Moinuddin

, Morley

. A model for managing delirious older inpatients. J Am Geriatr Soc, 2003; 51:1031–1035.

20.

Edwards

, Harrison

, Davis

. Potentially inappropriate prescribing for geriatric inpatients: An acute care of the elderly unit compared to a general medicine service. Consult Pharm, 2003; 18:37–42, 47–39.