Transitional Care: Looking for the Right Shoes to Fit Older Adult Patients

Abstract

Potentially avoidable hospitalizations are associated with high costs and an increased risk for iatrogenic conditions in older adult patients. Although care managers may be aware of the common potential pitfalls that may arise in the transfer of patients to and from the hospital, defining best practice models has been difficult. Many current models of geriatric care have had little or no impact on lowering the rates of hospitalizations and rehospitalizations when formally studied. Health care reform legislation mandates initiatives involving new models of coordinated or guided care such as the medical home model and the accountable care organization. These new models too will face significant challenges in their attempt to provide the financial incentives and systematic changes needed to successfully address transitional care in older adults.

THE NEED TO ADDRESS TRANSITIONAL CARE

At a cost of $572 billion in 2011, Medicare spending is expected to grow by 6% annually through 2020 (Medicare Payment Advisory Commission, 2012). For a system based on “productivity,” it is estimated that more than one-fourth of all medical testing and procedures may be unnecessary (American College of Physicians, 2012; Sirovich, Wolshin, & Schwartz, 2011). Much of this unnecessary care occurs in the acute and postacute care setting. For example, 20% of all Medicare beneficiaries who are discharged from a hospital are readmitted within 30 days (Jencks, Williams, & Coleman, 2009). Most of these readmissions are potentially preventable (Jiang, Russo, & Barnett, 2009). Hospitalist care decreases patient lengths of stay but does not lower total costs (including inpatient and outpatient care) or the rate of hospital readmissions (Kuo & Goodwin, 2011; Lindenauer et al., 2007; Rachoin et al., 2012).

In addition to the costs, preventable hospitalizations and unnecessary institution-based subacute care may increase the risk of adverse events and long-term institutionalization (Covinsky, Pierluissi, & Johnston, 2011; Golden, Martin, da Silva, & Roos, 2011). The fragmented nature of the modern health care system results in omissions in care, exposure to unnecessary procedures, excessive use of medications, and care by health care professionals who are often unfamiliar with the complex medical history and psychosocial issues of the older adult patient (Wohlauer et al., 2012). Efforts to address the costs and iatrogenic conditions associated with our fragmented health care system are grouped under the term “transitional care.”

TRANSITIONAL CARE PRINCIPLES

Transitional care is defined as “a set of actions designed to ensure for the coordination and continuity of health care as patients transfer between different locations or different levels of care” (Coleman & Boult, 2003). In an effort to focus transitional care efforts, Coleman (2011) described four pillars of an effective care transitions program:

• Patient medication self-management

• A dynamic patient-centered record, the personal health record

• Timely primary and specialty posthospitalization physician follow-up care

• Patient education regarding the “red flags” (alerts) that indicate a worsening in condition and how to respond

The issue of medication self-management is aided by a thorough medical reconciliation. This process involves identifying the most accurate list of all medications a patient is taking including drug name, dosage, frequency, and route. The health care professional then compares that list against the physician discharge orders with the goal of providing the patient with a correct medication list (Home Health Quality Improvement Organization Support Center, 2007).

The development of a dynamic patient-centered record relies heavily on the adoption of an electronic medical record that can be easily accessed and modified by all health care professionals as patients move through the acute, outpatient, and postacute health care settings. Timely access to follow-up care is vital because half of all Medicare hospital readmissions did not have follow-up care with a physician (Jiang et al., 2009). Patient education regarding the red flags involves developing an emergency care plan for the patient to identify signs and symptoms of impending clinically significant changes. The plan should also reinforce which health care professionals to contact and when (Coleman, 2011).

An additional component of transitional care involves health care professionals and patients working together to develop patient-centered goals including a focus on palliative care and a discussion of advance directives (Ralston, 2012). Identifying the patient’s end-of-life goals may help patients avoid unnecessary interventional procedures and hospitalizations. The American College of Physicians has published guidelines to improve effective communication between hospitalists and primary care physician (Table 1; Ralston, 2012). Although much effort is directed toward teaching health care professionals about transitional care, its impact will be limited unless it is accompanied by major system-based change in the delivery of health care. The difficult question, however, is how does the system need to change?

TABLE 1.

Coordination of Care Among Primary Care Physicians and Hospitalists During the Transitional Care Process

For PCPs When making referrals, list the pertinent issues and the specific questions being asked in referral. Quickly send relevant information to those caring for the hospitalized patient. For Hospitalists Notify PCPs when patients are admitted so PCP can provide needed information. Provide PCPs with progress reports, notification of discharge, and prompt discharge summaries.

Note. PCPs = primary care physicians.

Source. From Ralston, F. (2012). Making transitions better for patients. Philadelphia, PA: American College of Physicians. Retrieved from http://www.kevinmd.com/blog/2012/01/acp-making-transitions-patients.html

IMPACT OF CURRENT MODELS OF CARE

Although many anecdotal reports and quality improvement projects exist that highlight successful transitional care efforts, evidence-based reviews and best practices are difficult to identify. Three recent systematic reviews of transitional care studies all found insufficient evidence for any specific interventions that lowered overall costs or decreased the risk of rehospitalization (Hansen, Young, Hinami, Leung, & Williams, 2011; Hesselink et al., 2012; Prvu Bettger et al., 2012). Most of the reviewed studies had a quality improvement focus rather than an experimental design. Comparisons among studies are difficult because there is much heterogeneity in the study populations analyzed. The heterogeneity and poor description of the individual interventional components further hampers the ability to develop best practices. The ability to generalize rehospitalization research to the care of community-based older adults is further limited because these studies exclude high-risk patients who have dementia, have no caregiver, are non-English speakers, are geographically remote, are unable to use telephone, or refuse to consent for a research study (Golden, Tewary, Dang, & Roos, 2010).

Research studies of geriatric care coordination models demonstrate mixed results regarding decreasing hospital readmission rates and health care expenditures (Golden, Tewary, et al., 2010). In fact, an analysis by the Congressional Budget Office of 10 Medicare demonstration projects found little or no decrease in admissions (Nelson, 2012). A further analysis of these projects identified four programs that were able to lower hospital readmission rates by 8%–33% among “high-risk” older adults. Interventions used by these four programs are listed in Table 2. The ability to replicate these potentially beneficial components to “real-world” settings may be limited because of the high variability of care management training, processes, and funding (Golden, Roos, Silverman, & Beers, 2010). Despite the subgroup analysis showing decreases in rehospitalizations, none of these programs generated net savings to Medicare.

TABLE 2.

Features of Medicare Coordinated Care Demonstration Programs That Cut Hospital Admissions of High-Risk Patients

Telephone calls to patients and frequent in-person meetings Occasionally meeting in person with health care providers Acting as a communications hub for providers Delivering evidence-based health care education to patients Providing medication management Providing comprehensive transitional care after hospitalizations

Source. From Nelson, L. (2012, January). Lessons from Medicare’sdemonstration projects on disease management, care coordination, and value-based payment. Washington, DC: Congressional Budget Office.

The use of home telemonitoring in older adult patients with multiple health issues, likewise, found no impact on hospital admissions or emergency department visits (Takahashi et al., 2012).

Advanced practice nurse–led transitional care models have been studied in patients with heart failure (Naylor et al., 2004; Stauffer et al., 2011). Although these clinical research studies demonstrated a decrease in hospital readmissions, the cost of these programs may exceed any potential cost savings. The Care Transitions Intervention uses nurses and social workers as “transitional care coaches” to assist older adults on hospital discharge. By reinforcing the four pillars of transitional care, several studies have shown a decrease in hospital readmission rates (Coleman, Parry, Chalmers, & Min, 2006; Voss et al., 2011).

Other efforts involve the development of coordinated longitudinal care models.

The Program of All-Inclusive Care of the Elderly (PACE) and the Geriatric Resource for Assessment and Care of Elders (GRACE) were both developed as clinical models to provide coordinated longitudinal care for indigent homebound older adults. PACE provides transportation to bring patients to an adult day care setting that is staffed by an interdisciplinary team of health care professionals. PACE programs receive a fixed Medicare and Medicaid reimbursement. This model has been shown to significantly decrease hospitalizations and Medicare costs. However, Medicaid reimbursement is 86% higher than projected by fee-for-service estimates (Institute of Medicine [IOM], Committee on the Future Health Care Workforce, 2008). In addition, there maybe a selection bias among the patients who elect to enroll in this program because PACE enrollment requires that all care be provided/coordinated through the PACE physician (Golden, Roos, et al., 2010). Thus, patients who require large amounts of specialty care or those who are highly satisfied with their specialist physicians may be less likely to enroll in this program.

GRACE provides home-based integrated care for poor older adults (Counsell et al., 2007) through home assessments and care management by both an advanced registered nurse practitioner and a social worker. The GRACE model also uses an electronic medical record, Web-based care management tools, and the integration of pharmacy, mental health, home health, community, and inpatient geriatric services. This program demonstrated a lower incidence of emergency department visits. Hospital admission rates were unchanged except for the group at highest risk for rehospitalization. Total costs were unchanged.

Concierge medicine has arisen as a private pay model for physicians to provide greater access and more personal care to a smaller number of patients. In some concierge practices, the patient pays an annual fee for a specified bundle of services (e.g., annual exam, specified number of office visits, enhanced physician access). If additional physician services are needed, then the patient would pay an out-of-pocket fee for each service. From a public health perspective, concierge medicine only provides services to a small number of patients. Therefore, it has the potential to siphon off capacity that might limit the overall availability of primary care physicians to the general population.

Additional efforts to lower rehospitalizatons in the frail older adult largely have focused on nursing home residents (Ouslander & Berenson, 2011). Interventions to Reduce Acute Care Transfers (INTERACT) is an intervention designed to reduce nursing home hospitalizations (Ouslander et al., 2011) through the adoption of protocols that assist nursing home staff in early identification, assessment, and communication about changes in resident status. This intervention lowered hospital admissions in 25 nursing homes by 17%. A more recent study involving the use of INTERACT found lower (nonsignificant) hospitalization rates among those facilities that adopted the protocols (Tena-Nelson et al., 2012). Efforts to lower the rate of hospitalization and rehospitalization among nursing home residents is limited by the strong financial incentives and medical–legal pressures to transfer patients between hospitals and long-term care facilities (Mor, Intrator, Feng, & Grabowski, 2010). In addition, because of the conflicting payor interests long-term and acute care facilities, nursing homes are financially disincentivized to provide high cost medical care.

Most transitional care initiatives have been applied in an environment with little incentive for health care professionals and health care administrators to change the status quo (Katz, 2011). In fact, transitional care efforts to lower hospitalizations and rehospitalizations may even be viewed by some as a financial threat. Lacking such support, improving transitional care remains an uphill battle.

EMERGING MODELS FOR THE POST-HEALTH CARE REFORM ENVIRONMENT

Recent federal health care reform provides strong incentives for system-based change in the delivery of transitional care to patients following a hospitalization (One Hundred Eleventh Congress of the United States of America, 2010). Various models are mandated in the Health Care Affordability Act. The two that have received the most attention are the medical home model and the accountable care organization (ACO). Additional models are listed in Table 3 (Jiang et al., 2009; Kocher & Adashi, 2011).

TABLE 3.

Innovative Programs Mandated in the Health Care Reform Legislation

National Pilot Program on Payment Bundling Large scale 5-year voluntary effort will test the bundling of Medicare payments into a single comprehensive fee for an episode of care. Fee covers all health care entities before (3 days), during, and after (30 days) hospitalization.

Hospital Readmission Reduction Program Penalizes hospitals with greater than expected readmission rates for Medicare patients treated for congestive health failure, myocardial infarction, or pneumonia. Beginning October 2012, Medicare payments will decrease by 1%, 2% in 2013, and 3% in 2014. Expands in later years list of conditions that can result in penalties.

Adapted from “Hospital Readmissions and the Affordable Care Act: Paying for Coordinated Quality Care,” by R. P. Kocher & E. Y. Adashi, 2011, Journal of the American Medical Association, 306 (16), pp. 1794–1795.

The medical home model emphasizes the concepts of (a) physician care that is continuous, comprehensive, and with an extended access, (b) “whole person orientation” that addresses all of the patient’s health care needs, (c) coordination of care, and (d) focus on quality and savings. Despite its intentions, the widespread implementation of the medical home model faces significant challenges (Hoff, 2010) because there is no consensus on the specific processes that define a medical home. The costs to implement the program may also outweigh any potential savings (Maeng et al., 2012; Peikes et al., 2012). Similarly, it was unclear whether this model decreases the risk of hospitalization and rehospitalization (Peikes et al., 2012).

ACOs are provider networks organized to deliver low-cost, high-quality care. The provider networks may consist of multispecialty physician practices, hospitals, and/or health plans. The incentive for provider participation is the extra money the ACOs can receive (“shared savings”) if it can decrease Medicare costs lower than specific benchmarks and meet compliance with 33 clinical quality measures. In theory, the cost savings of an ACO will be achieved through an emphasis on seamless communication, quality improvement, and care coordination/disease management (Fisher, Staiger, Bynum, & Gottlieb, 2006). An important difference from health maintenance organizations (HMOs) is that patients are not required to stay in the ACO network and may go “out of network” without any increased personal costs.

Currently, there is no clear consensus on the ideal composition of the provider network. The integration of hospitals and physicians may not necessarily lead to improved quality of care or lower costs (Cuellar & Gertler, 2006). Our best guess is that it will depend on local market conditions. ACO structures will by necessity adapt to unique (or specific) local conditions, and, for that reason, any successes or failures may be difficult to generalize.

Not surprisingly, the potential impact of ACOs on costs and outcomes is also unclear. A review of the initial 10 demonstration programs revealed that five showed no savings. The other five saved $32 million with $16 million coming from one program (Correia, 2011). The potential shared savings revenue from Medicare may not compensate for the loss of income from decreased hospital admissions and the start-up costs for the ACO. In addition, there is little information about how ACOs will be expanded to rural settings or unurbanized areas where a large number of frail older adults may reside.

CONCLUSION

Addressing transitional care can only be addressed through dramatic changes to the Medicare fee-for-service system. Unfortunately, there is no clear evidence-based consensus about the proper care management model(s) that can decrease preventable hospitalizations and lower Medicare costs. Given the large heterogeneity of the older adult population regarding functional impairment, medical comorbidities, and psychosocial issues, it is unlikely that a “one size fits all” solution will work. Health systems will need to have a range of transitional care programs available to select the ones that most closely meet the needs of the older adult patient.

In the next several years, various new integrated health care models will be studied. The financial sticks and carrots attached to the health care reform models will allow for the vertical integration of longitudinal clinical care across multiple health care settings. The concept of care coordination will change from the current model from that of a defined case manager working in isolation to overcome the barriers of the health care system into a team effort involving members drawn from across the entire longitudinal spectrum of care.

Transitional care models will be evaluated for their impact on costs and for their effect on health outcomes. The inclusion of quality of care indicators as outcomes measures is meant to help minimize the potential conflict between profit incentives and efforts to decrease clinical services for patients. Further studies are needed to determine the quality indicators that are most appropriate for frail older adult patients.