Acute Care for Elders (ACE) Team Model of Care: A Clinical Overview

The Institute of Medicine (IOM) Reports of To Err is Human and Crossing the Quality Chasm have called for more interprofessional and coordinated hospital care. For over 20 years, Acute Care for Elders (ACE) Units and models of care that disseminate ACE principles have demonstrated outcomes in-line with the IOM goals. The objective of this overview is to provide a concise summary of studies that describe outcomes of ACE models of care published in 1995 or later. Twenty-two studies met the inclusion. Of these, 19 studies were from ACE Units and three were evaluations of ACE Services, or teams that cared for patients on more than one hospital unit. Outcomes from these studies included increased adherence to evidence-based geriatric care processes, improved patient functional status at time of hospital discharge, and reductions in length of stay and costs in patients admitted to ACE models compared to usual care. These outcomes represent value-based care. As interprofessional team models are adopted, training in successful team functioning will also be needed.


Introduction
The launching point for the quality movement in healthcare was the 1999 Institute of Medicine (IOM) Report To Err is Human, which concluded tens of thousands of Americans die annually from preventable errors in care [1]. In 2001, the follow-up report Crossing the Quality Chasm further described the gap between evidence-based healthcare and care delivered. This 2001 report acknowledged an aging population and thus, an increasing proportion of patients with multiple chronic conditions. The report also called for fundamental redesign of care delivery to include systems to train and support high performing patient-centered teams [2].
Forward thinking geriatric clinicians recognized the need for interprofessional team models of acute care prior to these IOM reports. The Acute Care for Elders (ACE) Unit is one of the better recognized examples. The ACE Unit model consists of several core components: (1) patientcentered care with proactive geriatric assessments, (2) nurse-driven care plans for the prevention and management of geriatric syndromes, (3) comprehensive care transition planning beginning at admission, and (4) medical care review with the goal to prevent iatrogenesis and incident geriatric syndromes. This care is delivered by an interprofessional team that conducts frequent (usually Monday-Friday) team meetings (ACE rounds) to develop the geriatric care plans for each patient. In most ACE Units a geriatrician or geriatric advanced nurse practitioner either participates in the team rounds and/or is an attending practitioner for patients. In addition, ideally the physical environment of the ACE Unit is modified to promote safe mobility and cognitive stimulation [3,4].
The first randomized controlled trial (RCT) of an ACE Unit was published in 1995 by Landefeld et al., and demonstrated the benefits of this care delivery redesign [4]. This publication was a seminal moment, launching ACE Unit development and research that continues today. With this in mind, we present here an overview of studies evaluating outcomes from ACE models and published in 1995 or later. We provide this overview with two primary objectives: (1) to provide a concise reference tool for use by clinicians and researchers working in the field of ACE team models of care, and (2) to assist in identifying opportunities for future ACE efforts, including scaling and dissemination strategies and future quality and research efforts.

Methods
Utilizing published criteria for fourteen recognized types of literature reviews, we present here a literature "overview," defined as "any summary of the medical literature that attempts to survey the literature and describe its characteristics." Overviews "can provide a broad and often comprehensive summation of a topic area" and the analysis may be "chronological, conceptual, or thematic" [5]. To complete this overview, we conducted a literature search using Pubmed, Embase, CINAHL, and Scopus with the following database-specific subject headings, key terms, and phrases: "Acute Care"; "Acute Care for Elders"; "Interprofessional Teams"; "Interdisciplinary Teams"; "Interprofessional Relations"; "Geriatrics"; "Aged"; "Aged 80 and older"; and "Elderly". After removing duplicates, two authors (KLF, KB) reviewed all titles and abstracts for studies satisfying our inclusion/exclusion criteria to ensure adherence to the ACE model (Table 1). This was followed by critical reading by four authors (KLF, KB, JV, DJ) of full text articles if insufficient information was provided in the title or abstract to determine relevance. Additionally, the reference lists of included retrieved systematic reviews and meta-analyses were also examined for other articles that would meet the inclusion criteria for this overview (Figure 1). In our final list of relevant articles, we did include one study from an ACE Unit accepting patients with a younger age threshold than traditional ACE units (60 years and over) [6]. Intervention delivered on a hospital unit to which patient was admitted for an acute illness Geriatric consult or co-management models of care Intervention included an outpatient component (i.e., home evaluation, follow-up in a geriatric clinic) Intervention designed to target only one diagnosis (i.e., delirium) and not comprehensive geriatric care Intervention delivered in a non-acute care unit setting (i.e., emergency department) or on a unit to which a patient was transferred for a rehabilitation focus after stabilization of an acute illness

Results
Our search strategy yielded twenty-two articles included in this overview. We present here the key findings from these studies in two categories of ACE models: (1) ACE Units, in which the care occurs for patients admitted to one geographically distinct hospital unit (19 studies), and (2) ACE Services, in which the team care is delivered by an admitting service caring for patients on more than one hospital unit (3 studies). Tables 2 and 3 are summaries of the key study characteristics and outcomes included in this overview.

Studies from ACE Units: Geriatric Processes of Care
Of the 19 studies examining ACE Units, five examine geriatric care processes and/or recognition of geriatric syndromes as outcome measures; two are RCTs [9,11,16,17,23]. Both RCTs found significantly increased documentation of geriatric syndromes in patients admitted to ACE compared to UC units [9,11]. In the Counsell et al., RCT, ACE demonstrated significant increased use of nurse-driven geriatric care protocols and earlier and more frequent consultations to social work and physical therapy.
Two non-randomized studies examined care processes in ACE models that did not require the in-person presence of a geriatric specialist [16,23]. Dr. Michael Malone and his team from the Aurora Health Care System have developed ACE Tracker software for use in several electronic medical record (EMR) systems. The ACE Tracker program summarizes in one report geriatric data items documented by various ACE team members (i.e., history of cognitive impairment, results of geriatric screens, presence of potentially inappropriate medications, consultation to social services or rehabilitation therapies). This unit-based ACE Tracker Report is then utilized by an "e-Geriatrician" who can participate in a unit's ACE rounds via conference call using the same electronic ACE Tracker data. In a pre/post study examining geriatric care processes, the e-Geriatrician model demonstrated significant reduction in the use of urinary catheters and increase in consultations for physical therapists on a surgical and medical unit, but no significant change in other processes measured (use of physical restraints, social service assessments, high risk medications) [16].
In a 2018 pilot study, Booth et al., evaluated the impact of a "Virtual ACE" unit intervention that also does not require geriatrician presence either in person or remotely. Team members from all disciplines on a hospital unit are trained in ACE principles, including use of care protocols stemming from standardized nurse assessments for cognitive impairment, delirium, function, and mobility. ACE Tracker capability also exists in this hospital and use of this tool for recognition of geriatric syndromes and interprofessional team communication was included in the training. In pre/post analysis Booth et al., report a significantly increased proportion of patients receiving geriatric screens for function and delirium on two orthopedic surgery units, signaling it may be possible to embed ACE care processes into the routine workflow of non-ACE units throughout a hospital [23].

Studies from ACE Units: Function and Mobility Outcomes
Likely because one of the primary goals of ACE Units was to prevent hospital acquired disability, eleven of the ACE Unit studies evaluated patient performance of activities of daily living (ADL), instrumental activities of daily living (IADL), mobility, and/or discharge to a post-acute care facility, either separately or as part of composite outcome measure [4,6,7,9,10,12,13,15,18,21,23]. Six of these studies were randomized and one was a secondary analysis of data collected from a prior RCT [4,7,9,10,12,18,21]. The landmark Landefeld et al., study randomized general medical patients age 70 and over to ACE and UC units at a teaching hospital and evaluated function as the primary outcome. Significantly more ACE patients improved performance of ADLs from baseline and admission to time of discharge compared to UC [4]. ACE patients also experienced significantly less post-acute care facility placement (14% vs. 22 %, p = 0.01). A second RCT from this same investigator group evaluated a new ACE Unit in a community hospital [9]. This study did not demonstrate the same benefits in ADL performance but did demonstrate significant improvement in the composite outcome of ADL decline or need for post-acute care facility placement. The authors noted the logistical challenges in the non-teaching environment due to lack of a physician representative present to participate in the ACE rounds. A third RCT by Barnes et al., also did not demonstrate improvements in function or post-acute care facility referral [18]. Of note, this RCT was conducted on the same ACE unit as in the Landefeld study and the authors noted some of the ACE care processes and environmental redesign were disseminated to UC units during the study period. Finally, a RCT from an ACE Unit in Sweden evaluated a composite outcome measure of death and/or severe ADL dependence and/or poor psychological well-being 3 months after discharge and found no significant difference in ACE versus UC [10]. Another non-USA randomized study of an ACE model for hip fracture patients also did not find a significant difference in a composite outcome that included mobility measures 6 months after surgery, except in subgroup analysis of cognitively impaired patients [12].
Several non-randomized studies did report significantly improved ADL performance at time of hospital discharge and/or reductions in post-acute care facility use. Two of these studies came from non-USA-based ACE Units that utilized geriatricians as the attendings [6,15] and one from implementing ACE care principles on a stroke unit [13]. In addition, the previously described Virtual ACE model reported increased patient mobilization [23]. In a study utilizing hospital units as the target for randomization, Borenstein et al. describe a unit-based care redesign intervention similar to the Virtual ACE approach but also included team communication notes with decision support in the EMR. The intervention units were also trained to conduct daily interprofessional team huddles facilitated by an in-person ACE physician advisor who was not directly caring for the patients. Analysis of data, which included patients who screened positive for defined geriatric risk factors, found medical complications were reduced and ACE patients had a significantly increased odds ratio of discharge to a facility [21]. Similar to Barnes et al., the authors noted that ACE care plans were also present to some extent in the comparison units [21].

Studies from ACE Units: Additional Outcomes
Several studies demonstrated other patient outcomes including but not limited to significantly improved self-rated overall health status [4] and health related quality of life [22]; improved patient, caregiver, and provider satisfaction [9]; and reduced complications of care [21]. One study, a RCT targeting older frail patients, evaluated mortality as the primary outcome and reported significantly reduced mortality at 3 and 6 months post-discharge from ACE compared to a UC unit [11]. A more recent study also targeted older frail patients and also noted reduced 3-month mortality [22].

Studies from ACE Units: Healthcare Utilization
In addition to the use of post-acute care facilities as summarized above, fifteen of the unit-based studies reported healthcare utilization (length of stay, readmissions, costs) either as primary or secondary outcome [6][7][8][10][11][12][13][14][15][16][18][19][20][21][22]. Ten of these demonstrated either significant reduction or trends toward reduced length of stay (LOS) and/or costs. More recent studies signal the ACE model may reduce costs even in the setting of short LOS. In a 2013 retrospective cohort study of hospitalist patients age 70 and over admitted to ACE versus UC hospitalist units, variable direct costs were significantly reduced despite the fact that both units had a LOS less than 4.3 days [20]. None of the retrieved studies evaluated 30-day readmissions as a primary outcome. Those that included readmissions as a secondary outcome reported either reduced or unchanged readmission rates even when LOS was reduced with ACE team care.

Studies from ACE Services
The literature search yielded three publications describing service-based ACE models (Table 3). In these models, patients admitted to one of these services received ACE-like care even if they were located on different units. The Mobile-ACE (MACE) study specifically states one of the aims is to deliver geriatric care "without the limitations of a physical unit." This model consisted of a geriatrician as the attending with a geriatric fellow, social worker, and clinical nurse specialist assigned to the service rather than a hospital unit. The MACE team conducted daily team meetings addressing geriatric care needs and early care transition planning. This model resulted in significantly reduced adverse events and LOS and an increase in care transitions preparedness as measured by the 3-item the Care Transitions Measure [26,27].
The remaining ACE service studies both utilized an intensive geriatric education experience delivered to general medical teams coupled with implementation of ACE care processes (geriatric screens, care protocols, and interprofessional team meetings). In Wald et al., this approach was implemented with Hospitalists to create the Hospitalist-ACE Service that admitted patients preferentially to 12 designated beds but also applied the ACE care and studied the impact in patients on any unit. The team conducted daily (Monday-Friday) 15-min team huddles to coordinate geriatric and care transition plans. This model demonstrated significantly increased recognition of geriatric syndromes and utilization of Do Not Resuscitate orders but did not demonstrate changes in LOS or costs in this program's first year in operation [24]. Yoo et al., implemented the same approach for general medicine services at an academic medical center. The intervention included daily geriatric assessments for function, cognition, delirium, medication reconciliation, and sleep disturbance and three times weekly 45-min interprofessional team meetings. The per-protocol analysis, which only included patients who received at least 80% of the daily geriatric screens, demonstrated reduced LOS (6.1 days vs. 6.8 days, p = 0.008) [25].

Discussion
Our review of the twenty-two studies retrieved revealed several themes. The majority of studies are not randomized trials, and those that are occurred predominantly in the first decade of ACE Unit research. One possibility for this limited number of RCTs may be the logistical inability to "hold beds" on designated hospital units due to continually high patient census levels. Non-randomized studies have the associated limitations and therefore any findings and conclusions should be considered within this context. Service related studies are difficult to perform, leading to the disparate study types our search revealed. The lack of standardized and uniform outcome measures or metrics for ACE Units further limits the ability to compare the effectiveness of the various ACE interventions. Many of the ACE studies we found were implementation or feasibility studies and fit into the category of pragmatic trials. Glasgow et al., summarizes the impetus behind the call for increasing the number of pragmatic studies, defined as trials "designed to answer the question of whether a program works under usual conditions, compared to explanatory trials that answer the question if an intervention works under ideal conditions" [28].
We also observed the methodology by which ACE care is implemented has begun to expand. Most commonly, ACE Units continue to care for general medical patients with a unit-based interprofessional team and in-person involvement from a geriatric trained clinician. However, likely in part due to the improved outcomes demonstrated by ACE studies and the increasing numbers of geriatric patients on all units, ACE models are increasingly being deployed for subspecialty populations and operationalized on units and services with and without direct geriatrician participation. We included these studies of ACE care in subspecialty patient populations (stroke, orthopedic surgery) not necessarily for direct comparison to studies of general medical patients, but rather to reveal the dissemination and possible benefits of the ACE model for older adults throughout a hospital.
The mixed results observed in some of the outcomes across studies may be related to the ACE intervention implemented, study design limitations, the outcome measurement tools, or other factors. For example, several studies evaluated outcomes within the first year of operation of a new ACE program. Given the time required to develop high-performing teams in systems accustomed to functioning in silos, repeat evaluation of these models may be beneficial now that time has elapsed. The two studies that demonstrated longer LOS from ACE were both from non-USA hospitals over 15 years ago. Overall the ACE studies in which the usual care LOS was six or more days demonstrated LOS reductions. Cost savings can most easily be attained by reducing LOS. However, ACE Units have also demonstrated reduced costs of care even when LOS is short. This perhaps signals the role of ACE in delivering care that is aligned with patient goals, therefore avoiding unnecessary interventions. Some ACE studies reported increased formal consultation to allied health services such as nutrition and rehabilitation therapies. Another possible future outcome for study may be appropriateness of these referrals, given these resources are often limited and the mandate for stewardship of resources in today's healthcare system. The mixed results in the post-discharge function outcomes measured may signal another opportunity for future intervention and investigation. Additionally, given the barriers of saving beds discussed above, we suspect many ACE Units today admit geriatric patients with varied potential for clinical and functional improvement. As one example, the University of Alabama at Birmingham ACE Unit does not utilize any ACE specific admission criteria. This has resulted in an ACE team caring for geriatric patients ranging from fully independent in ADLs to those receiving end-of-life care. Thus, the primary focus of this ACE Unit has become the delivery of "goal-aligned care". This heterogeneous mix of patients with disparate goals will likely require alternative research approaches and measurement tools.
Finally, a common theme in the ACE literature is the prerequisite for training healthcare professionals in the basic principles of geriatric medicine. Given the national shortage of fellowship trained geriatricians, attention to equipping all healthcare providers with this knowledge is essential to disseminating ACE care and achieving the ultimate goal to establish Age-Friendly Health Systems [29]. National institutions and funding mechanisms are required to support this education. Examples of such support include the Nurses Improving Care for Healthsystem Elders institution [30] and the Health Resources and Services Administration funded Geriatric Workforce Enhancement Programs [31]. The basic training of all healthcare providers in geriatrics could then shift the role of fellowship trained geriatricians to lead more system-level initiatives to further promote age-friendly care delivery [32]. Extending beyond geriatric education, working in teams has been recognized as a mechanism for improving quality and care coordination for patients of all ages. Future opportunities may also include merging patient safety teamwork curriculum such as TeamSTEPPS ® (Team Strategies & Tools to Enhance Performance & Patient Safety) with ACE team training. Available from the Agency for Healthcare Research and Quality [33], TeamSTEPPS ® use has resulted in sustained improvements in patient safety [34]. In addition to the challenges of providing geriatric education to all healthcare providers, other barriers hospitals face in developing ACE Units include lack of hospital-based geriatric providers with protected time and effort to lead program development, challenges in securing dedicated unit space, and the need for some start-up funding, despite published evidence of return on this investment. While dedicated ACE Units will likely remain the core model, the future for ACE care, and a means to address development barriers, may reside in the current trend toward embedding ACE care processes into existing hospital systems and workflows.

Conclusions
The patient characteristics and outcome measures in these 22 ACE studies are varied. While we have summarized key variables that may influence outcomes (i.e., mean age, mortality rates), conducting an in-depth systematic review and meta-analysis is beyond the scope of this literature overview. Rather, we conducted this overview with the goal to concisely summarize ACE studies to serve as a reference tool for clinicians and researchers working to improve and expand the implementation of team care for all hospitalized elders. Despite our attempt to be thorough in our search we suspect there are additional ACE unit studies that were not retrieved with our strategy. We also did not conduct an in-depth rating of the quality of each study retrieved. Overall, the majority of the studies demonstrated improved quality with simultaneous reductions in measures of inefficiency (LOS) that were cost-neutral or cost-saving. These outcomes represent value-based care and are aligned with the goals the IOM called for almost two decades ago.
Funding: This research received no external funding.