Team-Developed Care Plan and Ongoing Care Management by Social Workers and Nurse Practitioners Result in Better Outcomes and Fewer Emergency Department Visits for Low-Income Seniors

Problem Addressed

• Failure to provide evidence-based care, especially for vulnerable populations: A review of the care provided to over 345,000 Medicare beneficiaries between 1994 and 1996 found that they received indicated care less than two-thirds of the time for 16 of 40 indicators studied. African-Americans, low-income seniors, and residents of Federally defined Health Professional Shortage Areas were much less likely to receive indicated care.² A separate study of 24 evidence-based process-of-care measures (including measures related to primary and secondary prevention and disease-specific care for diabetes, heart failure, breast cancer, heart attack, pneumonia, and stroke) found that the percentage of patients receiving appropriate care ranged from 11 to 95 percent, with median performance of 69 percent.³ Another study concluded that "care for vulnerable elders falls short of acceptable levels for a wide variety of conditions."⁴
• Negative impact on health and costs: The failure to receive timely, evidence-based care can have a negative impact on physical and mental health and functional status, leading to increased morbidity and mortality and the need for expensive inpatient and ED care.¹ In fact, older adults with five or more chronic conditions account for 76 percent of Medicare expenditures, and per-capita health expenditures for those with 5 or more chronic conditions are 17 times higher than for those with no chronic conditions.⁵ 
• Limited use of existing models: Although a variety of system-level interventions to improve quality and outcomes in older adults exist (e.g., outpatient geriatric evaluation, collaborative interdisciplinary care), many of these models have been narrowly focused on a single disease, a specific subpopulation of older adults, and/or a single site of care. In addition, these interventions are often poorly integrated with primary care and mental health services and often fail to adequately manage care transitions.¹

Description of the Innovative Activity

• The support team: Trained, certified nurse practitioners and licensed clinical social workers with experience in geriatrics serve on the support team. These individuals are employed by the primary care practice. Three teams participated in the 2-year pilot program, each serving two practice sites.
• Population served: The program serves patients aged 65 and older who have an annual household income of less than 200 percent of the Federal poverty level, have had one or more primary care visits in the past 12 months, and who live in the local community and have access to a telephone. In the clinical trial,⁵ participants were on average 72 years old and over one-half (59 percent) were African-American. Trial participants had a high prevalence of chronic illness (e.g., 34 percent had diabetes mellitus) and geriatric syndromes (e.g., 22 percent had fallen in the prior 6 months). The study was not equipped to serve those who did not speak English.
• Initial, at-home assessment: The GRACE support team meets with patients (and family members if available) in the patient's home to conduct an initial, comprehensive geriatric assessment, including a medical and psychosocial history, medication review, functional assessment, and review of social support and advance directives. The team pays special attention to orthostatic vital signs, vision, hearing, gait and balance, affect, and mental status. In addition, the team performs a home safety evaluation.
• Consultation with larger interdisciplinary team to develop protocol-based care plan: After the assessment, the support team meets with a larger interdisciplinary team to develop an individualized care plan; the interdisciplinary team includes a geriatrician, pharmacist, physical therapist, mental health social worker, and community-based services liaison, all of whom work for the affiliated health system. The care plan is based on a set of program-specific protocols that have been adopted by GRACE for 12 conditions, including advance care planning, health maintenance, medication management, difficulty walking/falls, malnutrition/weight loss, visual impairment, hearing loss, dementia, chronic pain, urinary incontinence, depression, and caregiver burden. The protocols for each condition spell out specific interventions to be considered for implementation. For example, the difficulty/walking falls protocol contains 31 specific suggestions for further evaluation, management, consultation, and patient education, including 15 for the team to implement routinely and 16 additional recommendations that should be reviewed with the PCP. Each patient is proactively assessed for 10 of the 12 conditions (all patients receive the advance care planning and health maintenance protocols); for each condition present, the corresponding protocol recommendations, along with additional recommendations from the team, are activated.
• Face-to-face meeting with PCP to gain plan approval: Once the care plan is developed, the support team meets with the patient's PCP to review a computer-generated summary of the patient assessment and interdisciplinary team suggestions. Typically, the physician and team discuss one to three patients for 5 minutes each, with the goal of modifying, prioritizing, and gaining PCP approval of the care plan. Participation in these meetings is considered in the PCP's annual performance review.
• Team-led implementation, ongoing care coordination, and followup: Once approval has been secured, the support team collaborates with the PCP to implement the plan, including taking charge of ongoing care coordination and followup. Routine interventions based on the established protocols (often home and self-care interventions) can be implemented without the need for additional team review. Key steps in this process are outlined below:
  • Inhome followup visit: The support team meets with the patient in his or her home to review the care plan and patient goals.
  • Ongoing care coordination and case management: The support team provides ongoing care management and coordination of care across conditions, providers, and sites of care, primarily using face-to-face and telephone contacts with patients, family members/caregivers, and providers. During these calls and visits, the team encourages goal setting and self-care, teaches problem-solving skills, provides education using low-health-literacy materials that correspond to each GRACE protocol, prepares patients and physicians to address problems and team suggestions during office visits, and assists with transportation arrangements. Although the number, content, and timing of patient contacts will vary depending on patient needs, each patient receives a minimum of one phone contact each month. These monthly contacts provide an opportunity to check for and address any new problems, such as changes in medications, social supports, and/or living arrangements. Face-to-face home visits occur automatically after any ED visit or hospitalization. The social worker plays a critical role throughout this process, helping patients to access community-based resources (e.g., discounted fitness classes) and navigate the health and social services system.
  • Communication with providers: To enhance coordination of care across settings, the support team makes contact with other providers. For example, the team facilitates consultation by a specialized geriatric inpatient team whenever a program participant is hospitalized. Providers in the primary care and specialty clinics, ED, and hospital receive automated prompts via the EMR to contact the support team for information and assistance with followup and coordination of care.
  • Annual reassessment and followup visit: Each year, the support team repeats the comprehensive geriatric assessment, which in turn triggers another collaboration with the larger interdisciplinary team to develop a new individualized care plan.
  • Interdisciplinary team case reviews: The interdisciplinary team reviews each case at 3 and 6 weeks, and 3, 6, and 9 months after the initial and annual care-planning meetings. Additional reviews are held for any patient who has a major change in status or requires ED or inpatient care.
• EMR support: The nurse practitioners and social workers have access to the affiliated health system's EMR and a Web-based care management tracking tool developed specifically for the GRACE program. These tools serve as a critical support to the team in all of its activities, particularly with respect to facilitating communication and information transfer across team members and between the team and physicians.

References/Related Articles

Contact the Innovator

Results

• Enhanced provision of evidence-based care: Data from the EMR and patient reports were evaluated to determine what impact the program had on Assessing Care of Vulnerable Elders quality indicators, which have been tied to better survival among community-dwelling vulnerable adults.⁷ This analysis showed that program participants were more likely than usual-care patients to have their condition(s) recognized or diagnosed, receive a specialty consultation, and be provided with appropriate intervention or treatment. Participants were also more likely to receive evidence-based interventions related to preventive care, continuity of care, medication use, and end-of-life care.¹ These high rates of adherence were driven by strong fidelity to the program's processes and resulting recommendations—for example, the interdisciplinary team meeting occurred within 30 days of enrollment for 85 percent of patients, an average of 5.3 GRACE protocols were activated for each patient, and adherence to GRACE interdisciplinary team suggestions was quite high (81 percent in the first year and 79 percent in the second year).
• Improved health status: Compared to the control group, participants exhibited significant improvements in four of eight measures of functional status (which was measured using the eight Medical Outcomes 36-Item Short Form), including general health, vitality, social functioning, and mental health, and on the Mental Component Summary score. No differences in activity of daily living status (which was assessed using items from the Assets and Health Dynamics of the Oldest-Old survey) were observed.⁶
• Fewer ED visits for all, fewer hospital visits for those most at risk: Over a 2-year period, ED visits for program participants were substantially lower than for usual-care patients. (Although ED visits during the first year exhibited no difference across groups, they were significantly lower for program participants in the second year.) Although hospital admission rates across the two groups were roughly the same, analysis of a subgroup of patients deemed to be at high risk of hospitalization found that program participants had significantly fewer hospital admissions in the second year than did those receiving usual care.⁶
• High levels of patient and provider satisfaction: Sixty-four percent of program participants rated their overall satisfaction with care as very good or excellent, roughly the same as those receiving usual care (62 percent).¹ A separate survey found that physicians were much more satisfied with the resources available to treat patients under the GRACE program than under usual care.¹
• Generally cost neutral, but lower costs for high-risk patients in third year: A recent analysis found that average (mean) total costs were not significantly different for enrollees than for those receiving usual care, both overall (i.e., for all participants) and for high-risk patients. For high-risk patients, however, costs declined in the third year (after the trial ended).⁸

Evidence Rating (What is this?)

Context of the Innovation

Planning and Development Process

• Adoption of target conditions: Several champion PCPs and opinion leaders within the system reviewed a set of geriatric conditions that had previously been identified by an expert panel as being optimal targets for quality improvement. These individuals came to consensus on 12 conditions to be targeted.
• Protocol development: Program-specific protocols were developed for each of the 12 conditions based on several different published practice guidelines developed by various professional and specialty organizations and agencies.
• Development of educational materials: Patient education and self-management materials were developed for seniors with limited literacy skills. Materials were written at a fifth-grade level.
• Development of Web-based tracking system: Although the practice had its own EMR, a separate Web-based tracking system for the GRACE program was developed to provide summary sheets for team rounds and "to-do" lists. The GRACE support team shared progress toward implementation of the care plan with PCPs using printouts from this system and summary notes entered in the shared EMR.
• Training: Nurse practitioners and social workers each complete a 12-session training program (with meetings held once a week) on implementing the GRACE protocols and working as part of an interdisciplinary team.
• Expansion to other at-risk populations: A shorter (3 to 6 month) version of the same basic approach has been adopted for use with two other at-risk populations—those without a regular PCP who present at the ED/hospital or contact the call center (this is known as the "Bridge to Primary Care" program), and 64-year-olds who are about to become eligible for Medicare (this is known as the "Bridge to Medicare" program). Beginning in 2010, the GRACE model will also be used for patients enrolled in a new Medicare Advantage Plan at Wishard Health Services.

Resources Used and Skills Needed

• Staffing: To handle a caseload of 250 patients, the following staff are needed: two full-time equivalent nurse practitioners and social workers (i.e., each support team can handle 125 cases); one full-time equivalent administrative assistant; and 0.2 full-time equivalents (i.e., 1 day a week) of a geriatrician, pharmacist, physical therapist, mental health social worker, community-based services liaison, and practice manager. The health system also has a Director for Senior Connections, who supports the GRACE program by staying connected to local resources available in the community.
• Costs: The cost of the program is roughly $1,000 per patient per year, roughly 10 percent of which is reimbursable by Medicare for nurse practitioner home visits. Costs will likely be less without the expenses associated with conducting a clinical trial. In addition to the salary and benefits for the staff cited above, other expenses include travel (e.g., mileage reimbursement to the support teams), pager and cell phones, home visit bags, and office supplies. This cost estimate does not include office space and administrative overhead.

Funding Sources

National Institute on Aging; Wishard Health Services; Nina Mason Pulliam Charitable Trust

Tools and Other Resources

• The program Web site and GRACE protocols are located at: http://medicine.iupui.edu/IUCAR/research/grace.asp
• Information on the Assessing Care of Vulnerable Elders measure set can be found at: http://rand.org/health/projects/acove/acove3/
• Information on the Medical Outcomes 36-Item Short Form measure set can be found at: http://www.sf-36.org/
• Information on the Assets and Health Dynamics of the Oldest-Old measure set can be found at: http://hrsonline.isr.umich.edu/meta/sho_meta.php?hfyle=qnaires

Getting Started with This Innovation

• Consider target population carefully if cost savings is a critical objective: The program led to fewer hospitalizations only in the highest-risk patients. Given that outpatient visits also increased, a broadly targeted program could potentially result in an increase in overall costs, while a more targeted approach focused on high-risk patients is much more likely to be cost neutral or yield overall savings. In addition to performing a formal cost analysis, GRACE investigators are conducting further analysis to determine the program's impact on several subgroups, including patients with depression and comorbidities and those with geriatric syndromes.
• Integrate with primary care: Although any organization, including a health plan, could adopt the GRACE model, the case management program must be integrated with the primary care practice. The team itself can be within the practice or external to it, but the focus needs to be on identifying and treating geriatric medical and psychosocial problems, care coordination, and facilitating access to needed care.

Sustaining This Innovation

• Align financial incentives: The savings generated under the GRACE model will typically accrue to the payer (e.g., Medicare, Medicaid, or a health plan affiliated with either or both of these programs). Under traditional reimbursement systems, providers, including both hospitals and physician practices, will have little or no incentive to implement this kind of program because they carry the expense of the program but do not share in potential savings.
• Maintain relations with community-based resources: As noted, Wishard's Director for Senior Connections plays an important support role by maintaining trusting, collaborative relationships with local agencies that provide a wide array of vital services to program participants. Low-income seniors are much more dependent on community resources than are more affluent and well-educated populations.
• Seek ongoing funding sources: Although Medicare provides some reimbursement for program services, and a managed care organization may be able to fund the program out of the cost savings generated, provider-sponsored programs operating under traditional reimbursement will likely need outside financial support to maintain the initiative over time. For example, Wishard used the GRACE model as a foundation to apply to the Centers for Medicare & Medicaid Services to operate as a Medicare Advantage plan.

Use By Other Organizations