MODELS AND SYSTEMS OF GERIATRIC CARE
Kaiser Permanente Community Partners Project: Improving Geriatric Care Management Practices Susan M. Enguidanos, MPH,*† Nancy E. Gibbs, MD,‡ W. June Simmons, LCSW,* Karen J. Savoni, LCSW,‡ Paula M. Jamison, BA,* Lynn Hackstaff, ASCW,‡ Anne M. Griffin, MPH,* and David A. Cherin, PhD*§
This article describes a geriatric care management project that is testing whether geriatric care management plus a brief purchase of service (POS) intervention will lower medical costs, improve satisfaction with care, increase care plan adherence, and improve perceived quality of life. Kaiser Permanente members aged 65 and older who were eligible for geriatric care management and consented to participate in the study were randomized to one of four study groups: information and referral via mail, telephone care management, geriatric care management, or geriatric care management with POS capability. The POS intervention provides up to $2,000 of designated, paid services including in-home supportive services, transportation, respite, or medical equipment within the first 6 months of care management enrollment. Approximately 1,400 senior members were referred to the geriatric care management program, and 451 were randomly assigned to one of the four study groups. Those enrolled in the geriatric care management program were significantly more likely to be ethnic minorities and have lower income than the general Kaiser Permanente senior enrollment. Barriers encountered in implementing the POS intervention included establishing contractual agreements between Kaiser Permanente and private and community agencies, locating adequate and sufficient community agencies to provided needed services, monitoring service contracts, and delaying use of the POS benefit. J Am Geriatr Soc 51:710–714, 2003. Key words: geriatric case management; managed care organization; telephone case management; home- and community-based services
From the *Partners In Care Foundation, Burbank, California; †USC School of Social Work, University of Southern California, Los Angeles, California; ‡ Southern California Permanente Medical Group, Kaiser Permanente TriCentral Service Area, Bellflower, California; and §Department of Social Work, California State University Bakersfield, Bakersfield, California. This project was funded by a grant from the California HealthCare Foundation, Program for Elders in Managed Care Contract #98–374, from October 1998 through September 2002. Address correspondence to Susan Enguidanos, 101 S. First Street, Suite 1000, Burbank, CA 91502. E-mail: [email protected]
JAGS 51:710–714, 2003 © 2003 by the American Geriatrics Society
he population aged 65 and older is growing rapidly and is expected to double by 2030.1 In 1996, the average annual healthcare expenditure was more than $16,000 for an individual aged 85 and older.1 Several studies have been conducted to test alternate approaches to improving health status and reducing healthcare costs for the elderly through integration and coordination of community-based supportive services and medical care. Most notable are the San Francisco On-Lok Program for All-inclusive Care of the Elderly (PACE) and the Social Health Maintenance Organizations (S/HMOs). The PACE study revealed reductions in use of institutional care and rates of short-term hospital use from the overall use by the Medicare population.2 PACE almost exclusively served those dually eligible for Medicare and Medicaid and required that beneficiaries be nursing home eligible, limiting the applicability of this model to the majority of Medicare recipients. The S/HMO was implemented, in part, to determine whether additional coverage for long-term care with associated case management could produce sufficient savings to allow plans to offer chronic care benefits without increased cost to Medicare. The S/HMO plans have grown slowly because of difficulty in competing with other low- or no-premium Medicare HMOs in their regions. The program described in this paper is similar in concept to PACE and S/HMO, because its fundamental premise is that access to home- and community-based services will prevent physical decline, promote optimal health status and quality of life, and minimize use of more-expensive acute care services by frail elderly. In 1998, Kaiser Permanente (KP) TriCentral Service Area in partnership with Partners In Care Foundation and the Los Angeles County and City Area Agency on Aging (AAA), received a 4-year grant from the California HealthCare Foundation’s Program for Elders in Managed Care to test models of geriatric care management (GCM) and to determine whether GCM plus a brief payment intervention (up to $2,000 of designated, paid services such as for home supportive services, transportation, respite, or medical equipment) would lower medical costs, improve satisfaction with care, increase care plan adherence, and improve perceived quality of life. A specific aim of this project was to determine whether
MAY 2003–VOL. 51, NO. 5
improved collaboration between KP and AAA agencies would result in improved patient outcomes. PROGRAM OVERVIEW This project is being conducted in the KP TriCentral Service Area located in Southern California. The TriCentral Service Area consists of three KP medical centers (Baldwin Park, Bellflower, and Harbor City) and has a total member enrollment of 675,000, approximately 10% of whom are 65 and older. The GCM program was developed so that frail senior members could receive appropriate resources within KP and outside KP. A social worker, the assistant manager of the continuing care department, supervises the GCM program. Geriatricians working within the continuing care department provide medical oversight and clinical consultation. KP senior members access GCM services through two methods (Figure 1): the health status form (HSF) survey process and referral. Referrals may be made to GCM by a KP healthcare professional, community service agency, caregiver, or through self-referral. The HSF, piloted by KP Northwest Region3 as a self-report instrument to predict elderly members at risk for frailty in the coming year, was used to identify patients who may benefit from care management services. Information gathered via this form includes patient reports of current health conditions, functional abilities, recent falls, weight loss, medical service use, health behaviors, and demographic characteristics. In 1998, HSFs were mailed to the TriCentral region’s existing members aged 75 and older. Due to fiscal constraints, surveys were limited to members aged 75 and older because rates of frailty were found to be higher for this age group. HSFs are also sent to all new members aged 65 and older. Responses are scanned, and data are then electronically transferred to Southern California, where profiles for frail patients are generated and sent to the TriCentral GCM program.
COMMUNITY PARTNERS PROJECT
ELIGIBILITY CRITERIA All patients referred for GCM program services are assessed for program eligibility. The HSF screening focuses on identifying frail seniors who may meet criteria for GCM program services. The GCM program eligibility criteria target only those seniors with unmet healthcare and home and community service needs. Many of the frail patients identified through the HSF have no unmet needs that may be served through the GCM program. To be eligible for GCM, a KP member must be aged 65 and older and meet the criteria listed in Table 1. Using an assessment protocol, a research assistant contacts the patient/family via telephone and assesses eligibility for program services. If the member is eligible for GCM and consents to participate in the study, he or she is randomized to one of four study groups: information and referral, telephone care management, GCM, or GCM with purchase of service (POS) capability. INTERVENTIONS The four levels of interventions were selected because they represent the full continuum of care management practices. The information and referral intervention reflects the current minimum level of care management practice within KP. Information and Referral Assistance Participants randomized to this intervention receive a package containing a list of referrals accommodated to their specific needs and geographical location and the telephone number to local AAA services. Telephone care managers assembled this package based on the needs identified through the telephone interview and send it to the patient by mail. Telephone Care Management This is a short-term telephone-based intervention conducted by bachelor’s level social workers. The telephone
Table 1. Geriatric Care Management Program Eligibility Criteria Kaiser Permanente member must be aged 65 or older and meet at least one of the following four criteria: • have had more than three emergency room visits or hospitalizations in the past 12 months • used more than two medical office visits in past 3 months • have one or more IADL or ADL deficiencies as measured by the Katz index of ADLs4 • have cognitive impairment as measured by the telephone interview for cognitive status5 In addition, the member must meet one of the following criteria • does not have a caregiver to provide assistance with ADL or IADL deficiencies • have severe behavior problems as measured by a shortened version of the memory and behavior problems checklist6 • have a caregiver reporting stress/burden as measured by a shortened version of the burden interview7
Figure 1. Flow chart of geriatric care management study.
IADL instrumental activity of daily living; ADL activity of daily living.
ENGUIDANOS ET AL.
care manager calls the patient within 48 working hours of the referral. Before contacting the patient, the telephone care manager reviews KP databases for history of service use and recent illnesses, diagnoses, and current medications. Next, the telephone care manager contacts the patient and primary caregiver and confirms initial assessment information and discusses and facilitates needed services, referrals, and other needs with the patient/caregiver and encourages access to these services. These may include KP and community-based referrals. A follow-up letter is sent to the patient/caregiver listing the referrals given via telephone. At 2 to 4 weeks after initial contact, the telephone care manager conducts a telephone follow-up with the patient and/or caregiver. For each case, approximately four to five telephone contacts are made with patients, family members, community service agencies, and KP service providers over a 4-week period. The total time spent on each case varies according to the extent of the patients’ needs (an average of 60 minutes is spent on each case). The initial call is approximately 25 minutes long. Telephone case managers have an average caseload of 30 patients and provide brief one-time-only assistance and referrals to approximately 80 additional patients each month. Geriatric Care Management A care manager who is a registered nurse or master’s level social worker provides this intervention. Cases are assigned based on geographical area rather than by care manager discipline. The GCM intervention is more comprehensive than telephone care management because it includes in-home visits and ongoing coordination, monitoring, and follow-up for 8 months on average. In most instances, upon assignment of a new case, the care manager conducts an in-home assessment. This face-to-face assessment verifies and expands upon the telephone eligibility assessment by assessing environmental conditions and home safety. Based on information collected in both assessments, a care plan is developed to address health-related and psychosocial needs. The care plan identifies specific problems, interventions, and goals and includes target dates for problem resolution. The initial care plan is reviewed at case conference team meetings with the GCM team, which includes a geriatrician and the assistant manager of the continuing care department, and is then reviewed with the member and his or her caregiver. The assistant manager of the continuing care department is a social worker responsible for providing administrative and clinical oversight for the GCM program. The geriatrician may provide care recommendations, such as alternate treatment options, potential medication adjustments, and other information that would enhance the development and implementation of the care plan. Geriatricians also interface with the primary care physician when needed to assist in meeting care plan goals. Additionally, they provide office or in-home (on rare occasions) consultation if further geriatric consultation is warranted and the patient and the primary care physician are in agreement. After approval of the care plan, the care manager interacts with the patient and the family to assist them in negotiating the KP system to address the identified needs or to access community services such as home-delivered
MAY 2003–VOL. 51, NO. 5
meals, transportation services, personal care, and homemaking services. The care manager also interacts closely with the patient’s primary care physician, who oftentimes initiates the referral to GCM, to ensure continuity of care and coordination of all services. A summary of the care manager’s assessment is filed in the patient’s paper medical chart. After the approval of the care plan, the care manager informs the physician of the patients’ enrollment in GCM services and areas of concern and coordinates needed medical intervention with the patient and the physician. The primary care physician’s time involved in each case varies; the physician may spend as little as a few minutes reviewing the assessment information to several hours conducting follow-up appointments, making telephone calls, charting orders, and consulting with family members. The GCM intervention usually includes at least one in-home visit, several follow-up calls or visits with the patient or caregiver, and much coordination between community and KP service providers. The care manager spends approximately 20 hours on each case over an 8month period, most of which takes place during the first 3 months of care management. About one-third of the time is spent interacting directly with the patient, in person or via telephone, another 4 to 5 hours is spent coordinating with community-based services and resource providers, and the remaining time is spent communicating with primary care physicians and coordinating KP services. The average caseload per GCM is 80 clients. GCM with POS Capability This intervention includes all of the elements of the GCM intervention with the addition of up to $2,000 of designated, paid services available within the first 6 months of geriatric care management enrollment. Provision of these funds is intended to facilitate initial implementation of the care plan and overcome barriers of access to home- and community-based services. The goals of care plan implementation are to stabilize or improve health status or prevent decline. Care managers are expected to use AAA or other local and state-funded public programs to implement care plans where possible and appropriate. If the member is not eligible for these programs or if a waiting list exists POS funds may be used to provide designated services from a list of approved agencies or vendors. POS-designated services are listed in Table 2. Time spent coordinating the case is approximately 4 to 6 hours longer than in GCM without POS. This extra time is primarily used to implement and monitor POS services. DESCRIPTIVE RESULTS From May 2000 to September 2001, more than 1,400 senior members were referred to the GCM program; 156 (10.7%) refused the initial assessment, and 35 (2.5%) could not be reached. Of those assessed, 50.4% were eligible for the study and 11.7% declined to participate. Four hundred fifty-one were randomized: 98 to information and referral, 113 to telephone care management, 117 to GCM, and 123 to GCM and POS. All members receiving services were aged 65 and older, with a mean age of 79. Fifty-three percent of those served were white, 20% African American, 21% Latino,
MAY 2003–VOL. 51, NO. 5
Table 2. Designated Services for Purchase of Service Intervention Patients enrolled in the geriatric care management (GCM) plus purchase-of-service group were eligible to receive up to $2,000 of designated, paid services available within the first 6 months of GCM enrollment. These services include: • in-home supportive services, such as personal care, homemaker/chore, home delivered meals, minor home repairs and maintenance, heavy cleaning, legal assistance, and money management • equipment, such as noncovered medical equipment and special communication devices • transportation • respite care, such as a personal care attendant or homemaker, or out-of-home care, such as board and care facilities, skilled nursing homes, adult day care, or Alzheimer’s center • other special needs that are critical for maintaining or improving health status or preventing a crisis or decline
and 3.4% Asian. Approximately 45% were widowed and 42% married. The majority (67%) of those receiving program services were female, and 53% reported having three or more existing health problems. Although more than two-thirds of those served had at least a high school education, the average income was low (80% reported their income to be under $20,000). Seventy-nine percent reported living in their own home or apartment. The GCM program serves a statistically significantly higher portion of ethnic minorities (47% vs 38.1%; P .001) and a larger segment with low income (80% vs 31% report having an income under $20,000; P .001) than the total TriCentral Service Area enrollment of members aged 65 and older. A review of care plans indicated that telephone care managers identified an average of 5.05 problems at the initial assessment and 6.69 interventions to address these problems. The average number of problems identified for GCM patients was 5.28, with a mean of 8.03 interventions/referrals. Although the GCM plus POS group had similar number of problems identified (mean 5.75), more interventions (mean 9.48) were identified to address these problems than with telephone care management and GCM. (Those receiving information via mail did not have care plans developed.) Final data collection and data analysis will be completed and available by fall 2003. PROBLEMS AND LIMITATIONS Several challenges were encountered in the intervention design being tested in this study. One area was in the POS benefit. Because KP had not previously been a purchaser of community-based long-term care services for members, a new infrastructure had to be developed to implement this aspect of the study. In the first year of the study, there were system obstacles in establishing contractual agreements between KP and significantly smaller private and community agencies. The complexity of the KP TriCentral
COMMUNITY PARTNERS PROJECT
regional administrative and financial systems demanded intricate procedures in procuring contractual agreements with service agencies, which often delayed onset of services and, in some cases, necessitated substitution of service providers. In addition, the vast geographical region (approximately 2,000 square miles) covered by the KP TriCentral Service Area required establishing several contracts for each service type with agencies located in the various communities throughout the service area. Finally, some of the smaller agencies providing localized services worked on a cash-only basis, which further complicated the use of POS funds. There were external and internal barriers to establishing POS contracts. In some areas, it was difficult to purchase a service even with the availability of funds because vendors of a given type did not exist in all geographic areas. Moreover, many of the contractual service agencies failed to provide the contracted services, particularly inhome support services. Oftentimes, care managers found that home care aides did not appear on scheduled days, yet the agency billed these services. Care managers spent substantial time tracking the delivery of services to ensure that contracts were honored and needed services were received. Initially, the POS benefit was designed so that it had to be used within 1 year of initiating GCM services. Once accessed, the care manager had 3 months to spend the $2,000 benefit for POS services. The POS benefit was originally designed to eliminate patient-service-use barriers such as patients’ saving money for a crisis situation rather than using their money for needed home and communitybased services. However, the time frame imposed by the POS structure contributed to care managers postponing use of the benefit for fear that the patient might have a more-critical need for POS funds later. To eliminate this saving of funds by care managers, the structure of the POS benefit was changed. Care managers now have 6 months from the day the patient enters care management services to access POS benefits. The maximum POS benefit remains at $2,000. The second limitation of the study lies in the lack of delineation of the care management intervention before initiation of the study. As a result, the care management intervention was not consistently implemented between care managers and in all patients, and practice varied considerably between practitioners. Efforts are currently underway to strengthen the care management intervention through implementation of standardized protocols and procedures. DIRECTIONS FOR FUTURE INNOVATIONS The results from the PACE and S/HMO study, outcomes from other studies of GCM practice, and insights that have been gained from the implementation of this project suggest that further research in the field of GMC is needed. The scope of GMC practice must be expanded from the existing primary focus on service coordination to include theoretically driven interventions that hold more promise for effecting long-term change in health-related behaviors and improving outcomes. Research is needed to test interventions that integrate cognitive and behavior change theories into GCM practice, building upon the traditional care management goals of service coordination,
ENGUIDANOS ET AL.
comprehensive assessment, and care plan development. These models should be multimodal, combining telephone services, office visits, and in-home assessment when indicated. In addition, further studies need to be conducted to determine which patient variables are most associated with positive patient outcomes for each intervention type. This would enable program developers to link patients with the intervention most correlated with positive outcomes based on patient characteristics, assets, and needs. REFERENCES 1. Federal Interagency Forum on Aging-Related Statistics. Older Americans 2000: Key Indicators of Well-Being [On-line]. Available at www.agingstats.gov/ chartbook2000/default.htm Accessed August 2000.
MAY 2003–VOL. 51, NO. 5
2. Wieland D, Lamb VL, Sutton SR et al. Hospitalization in the program of allinclusive care for the elderly (PACE): Rates, concomitants, and predictors. J Am Geriatr Soc 2000;48:1373–1380. 3. Brody KK, Johnson RE, Ried D. Evaluation of a self-report screening instrument to predict frailty outcomes in aging populations. Gerontologist 1997;37: 182–191. 4. Katz S, Akpom C. A measure of primary sociobiological functions. Int J Health Serv 1976;6:496–507. 5. Brandt T, Spencer M, Folstein M. Telephone interview for cognitive status. Clin Nurs Res 1996;5:185–198. 6. Teri L, Truax P, Logsdon R et al. Assessment of behavioral problems in dementia: The revised memory and behavior problems checklist. Psych Aging 1992;7: 622–631. 7. Zarit SH, Zarit JM. The Memory and Behavior Problems Checklist and the Burden Interview. University Park, PA: Pennsylvania State Gerontology Center, 1990.