PI name and affiliations Andrew Auerbach, MD; Professor, Division of Hospital Medicine, University of California San Francisco (UCSF), Christine Ritchie, MD; Professor, Division of Geriatrics, UCSF

Potential Co-investigators ; Neil Sehgal, PhD; Assistant Professor, Health Services Administration, University of Maryland, Kevin O’Leary MD, Northwestern University, Kathryn Huber MS University of Arizona School of Medicine

Program overview/introduction This project seeks to develop a national overview of geriatrics and palliative care service delivery models for older patients in hospitals, and link these hospital and service-level factors to patient outcomes such as readmission, length of stay, discharge to SNF, and use of potentially inappropriate medications at discharge (by BEERS criteria). This knowledge will be critical in guiding healthcare system changes aimed at providing high-value care for this growing high-need and high-cost population.

Key clinical questions or evidence gaps: How to provide optimal care to the growing number of older patients admitted to United States hospitals is a subject of substantial debate. While models such as the Acute Care of the Elderly (ACE) unit-based models are considered a gold standard (1,2), they are difficult to adopt due to practical requirements such as dedicated space, availability of specialized personnel such as geriatricians, and the ongoing costs of maintaining these programs.  Robust single-site study evidence supports ACE models (3), but few have studied the variability in their implementation or how and whether components are used broadly.

As an alternative (or precondition, in some cases) to full adoption of the ACE model, other programs have been developed to encourage adoption of best practices in the care of elderly patients. These models include the Nurses Improving Care for Health System Elders (NICHE), the Hospital Elder Life Program (HELP), as well as geriatric needs tailored programs provided through inpatient pharmacy services, hospitalists, or in the context of unit based care models (e.g. shared rounding). Few national data exist to describe the prevalence of these variants and even fewer exist to describe how and whether these programs are associated with differences in outcomes of relevance to older patients in the hospital.

The overall goal of this study is to develop a national overview of the prevalence of geriatrics-tailored inpatient services and to determine whether particular models are associated with differences in processes (e.g. use of potentially inappropriate medications), or outcomes (e.g. length of stay, bed falls, discharge to hospice, ICU utilization).

Aims and Hypotheses 

Aim 1: To carry out a phone administered site level survey assessing presence of geriatrics-tailored services among a broadly representative group of US hospitals part of the HOMERuN CRG.

Hypothesis 1a: That prevalence of ACE units will be relatively low compared to less intensive models (such as shared rounding models, or early ambulation models).

Hypothesis 1b: That hospital-level factors (such as presence of hospitalists, or of palliative care services) will be associated with higher adoption of geriatrics-tailored services.

Aim 2: To, using administrative data collected as part of PCORnet efforts, define whether variations in Aim 1 structures of care are also associate with variations in processes and outcomes of importance to hospitalized elders

Hypothesis 2: That, among patients 65 or older, presence of more geriatric services (fully adopted ACE models vs. partial vs. none ) will be associated with:

• Lower rates of administration of BEERS-criteria defined inappropriate medications at discharge
• Shorter length of stay
• Fewer 30 day readmissions
• Fewer ICU days
• Less ICU use among patients who die in hospital.
• Higher likelihood of being discharged home

Any preliminary data  Kathryn Huber has been piloting our survey in a separate cohort of hospitals; these data are being aggregated now.

Study design, including study subjects (patients and/or providers) and setting, comparator groups, data sources, outcomes, analysis plan, power and sample size, limitations, and timeline The study design is a cross-sectional phone- and email-administered survey of Hospital Medicine CRG leads,  Geriatrics Division Chiefs, Chief Medical Officers (CMOs),  or Vice-Presidents for Medical Affairs (VPMAs) at HOMERuN CRG sites. . Results from this survey will be linked (at the site level) to anonymized patient-level data (except for site identifiers) patient-level administrative data collected from each site as part of PCORnet data-sharing activities. We will then test whether the presence of specific features known to improve outcomes of elders and medically complex patients (such as those in an ACE unit) are associated with reduced risk for in-hospital death, readmission, discharge to SNF or discharge on medications that may be inappropriate for elders.  

Using our survey data, we will use simple statistics to define prevalence of ACE unit features across hospital sites. We will use hierarchical multivariable models to determine the association between care at each hospital with features as defined by our site survey and patient-level outcomes. These analyses will utilize data collected from our survey and administrative data sources.

Characteristics of sites who might participate The study seeks to develop a national overview, and all CRG sites will be invited to participate in Aim 1 work.   Patient-level data will focus on adults admitted to PCORnet hospitals during the study time period (7/1/15-6/30/17), and may be limited based on PCORnet data availability

Potential funders with RFA/RFP and due dates : This research program is likely to be of interest to NIA, as well as several foundations.

REFERENCES

1. Fox MT, Persaud M, Maimets I, et al. Effectiveness of acute geriatric unit care using acute care for elders components: a systematic review and meta-analysis. Journal of the American Geriatrics Society 2012;60:2237-45.

2. Fox MT, Sidani S, Persaud M, et al. Acute care for elders components of acute geriatric unit care: systematic descriptive review. Journal of the American Geriatrics Society 2013;61:939-46.

3). Landefeld CS, Palmer RM, Kresevic DM, et al. A randomized trial of care in a hospital medical unit especially designed to improve the functional outcomes of acutely ill older patients. New England Journal of Medicine 1995; 332:1338-44.