RESEARCH: Care Management For Older Adults
Care management programs have become more widely adopted as health systems try to improve the coordination and integration of services across the continuum of care, especially for frail older adults. Several models of care suggest the inclusion of registered nurses (RNs) and social workers to assist in these activities. In a 2018 national survey of 410 clinicians in 363 primary care and geriatrics practices caring for frail older adults, we found that nearly 40 percent of practices had no social workers or RNs.
However, when both types of providers did work in practice, social workers were more likely than RNs to be reported to participate in social needs assessment and RNs more likely than social workers to participate in care coordination. Physicians’ involvement in social needs assessment and care coordination declined significantly when social workers, RNs, or both were employed in the practice.
Care management programs have become more widely adopted in the past decade as health systems try to reduce costly utilization and foster better coordination of services across the continuum of care for complex, frail, or seriously ill patients.1 The Centers for Medicare and Medicaid Services (CMS) recognizes complex chronic care management as a critical component of primary care and in 2017 instituted new codes under the Medicare Physician Fee Schedule for these services furnished to Medicare patients with multiple chronic conditions.
Early care management programs focused on patients who used a high volume of services or whose care was in the highest tier of costs, especially frail older adults who are at increased risk for poor health outcomes, hospitalization, falls with injury, and mortality and others with complex comorbid medical conditions.2–5 Health systems have also had increased incentives to develop or contract with such programs in an era of bundled and other value-based payment systems. Recent reports indicate that most accountable care organizations (ACOs) and certified patient-centered medical homes now include some sort of care or case management.6 Care management programs are also intended to support the efforts of primary care clinicians and geriatricians, who are in short supply in many regions of the United States as the health system faces the increased care demands of an aging population.
Registered nurses (RNs) have played a major role in developing and leading care management programs in hospitals and health systems, consistent with meeting the health care needs of patients with complex chronic illness, frailty, or comorbidity at hospital-to-community transitions. Leading researchers in primary and geriatric care teams have also promoted clinician partnerships with RNs in community practice.7,8 Increasing interest in assessing social needs (financial, housing, and food security) and behavioral health needs has engendered the increased inclusion of social workers, community health workers, and navigators to assess and address social and behavioral health needs.9–12 In one study of fifteen ACOs, all but one of them included RN care managers, while eleven had social workers and nine had navigators or community health workers.5 Demonstrating cost savings in care management programs has been challenging: Savings are typically thought to come with decreased use of emergency and hospital services, but programs may be labor and staffing intensive. Still, many care management programs have clear benefits for patients in managing care arrangements; assisting with care coordination at transitions in care; and providing behavioral, educational, and community resource supports.13,14
While several contemporary models for developing primary care teams recommend integrating RNs and social workers in primary care and geriatrics practices,7,8,10 fewer than 10 percent of all RNs and social workers were employed in ambulatory care office settings in 2016.15 Differently educated and licensed health professionals and community health workers bring different competencies to the needs of a population with serious illness or disability. Physicians, nurse practitioners (NPs), physician assistants (PAs), RNs, and social workers are all expected by professional credentialing organizations to have competencies in assisting patients with managing complex chronic illness, ensuring smooth care transitions to and from and coordination with hospitals, and assessing social issues that may pose barriers to adhering to care plans and accessing needed services (for example, structural racism, housing insecurity, family status, isolation, and financial stress).10
Patients with complex illness face a bewildering array of specialty and primary care providers, as well as home and long-term care services and supports. Care arrangements may be especially confusing for frail elderly people with high health care needs—who, despite making up only 4 percent of the Medicare population, account for 43.9 percent of total potentially preventable spending.16 Given the growing public and private investment in care management programs, we wondered how and whether these activities are accomplished at the practice level, and how they may vary by practice staffing.
As part of a 2018 national survey of primary care and geriatrics clinicians who care for frail older adults, we explored how practices allocate staff to care management activities, including complex chronic care management, coordination at care transitions to or from hospitals, and assessing social issues in the home. We focused this work on frail elders as part of a two-year project to explore effective teams and optimal staffing for that population. We report here new data obtained from physicians and NPs in primary care and geriatrics on the roles of different licensed and unlicensed staff types to accomplish care management activities. We consider the implications of different staffing models for the work and education of clinicians in nursing, social work, and medicine, as well as other health care staffing.
STUDY DATA AND METHODS
Data were from the 2018 Survey of Primary Care and Geriatric Clinicians, a national cross-sectional survey of 410 clinicians in 363 practices that provide care to older adults ages sixty-five and over. The complete survey instrument is available online.17 The fieldwork was conducted by Mathematica Policy Research in March-August 2018.
The survey was designed to measure how practices with primary care and geriatrics physicians and NPs organize and deliver care to older adults. The survey was developed following an analysis of twenty-two site visits and the results of twenty focus groups with physicians, RNs, NPs, and social workers; frail elders; and caregivers in five US regions in 2017 (data not yet published). We measured individual and team functions through key-informant interviews, team meetings, and observations of practice and clinical operations. Among the sites visited, we observed a range of staffing models, most notably in six care management programs that worked with primary care and geriatrics practices. We observed programs that were almost exclusively led and staffed by nurses, others that were exclusively led and staffed by social workers, and many practices that had neither RNs nor social workers.
The questionnaire was drafted and reviewed by an interprofessional, interdisciplinary panel of experts in medicine, nursing, and social work, and it also received a quality review by Mathematica Policy Research. The draft versions underwent cognitive and pilot testing before survey administration. A key element of the survey development was a list of health services and assessments that are part of a comprehensive assessment of primary care and geriatrics patients (see text below in Analysis).17 For each of these services and assessments, we asked who in the practice provided the service. Clinicians could indicate whether multiple staff worked on an assessment or activity.
SAMPLE AND FIELDWORK
We selected a nationally representative random sample of practices that employed primary care or geriatrics physicians or NPs from a file maintained by SK&A, a sampling vendor. The file is a national database of more than 760,000 physician practices that contains staffing and practice variables, including clinical specialties of physicians and NPs. We sampled in six strata by the presence of physician and NP (physician only, physician and NP, or NP only) and specialty (primary care or geriatrics). Our initial sample had 1,000 clinicians in 761 practices. On five occasions we attempted to contact these clinicians by postal mail, alternating full questionnaires and reminder postcards. We used Priority Mail and initially enclosed prepaid incentives of $40 per clinician; we escalated the incentives to $75 in subsequent contacts. Clinicians who were no longer working at the sampled practice or specialty were excluded as ineligible, as were those who did not provide patient care to adults ages sixty-five and older or whose sampled practice had closed. The practice-level response rate was 60.1 percent.
We examined the relationship between practice staffing type and the provision of three care management activities: an assessment of social issues, isolation, and financial stress; care coordination at transitions to and from hospitals; and complex chronic care management. For each of these activities, we asked clinicians, “In your practice, who typically provides this service to all patients age 65 and over?” Response options were “MD,” “NP,” “PA,” “RN,” “social worker,” “other licensed professionals,” “other non-licensed staff,” and “no one/not applicable.”
We grouped clinician responses by practice staffing type, a variable created to describe four types of staffing mix: both RNs and social workers, RNs but not social workers, social workers but not RNs, and neither RNs nor social workers. For this purpose, NPs were not coded as RNs but were grouped with physicians.
We used chi-square tests to compare activities in these differently configured practice groups.
We also examined, by practice staffing type, hours spent by clinicians in different communication activities that are related to care coordination and social needs assessment, including communication time with patients, other team members, other providers, and family caregivers. We further asked respondents: “If your practice had the resources to hire additional personnel within the next month, what type(s) of professional or staff member would help increase the quality of the care your practice provides to frail older adults?”
The data shown in all exhibits were weighted to adjust for nonresponse and stratification. In some practices, when a physician and an NP were both sampled, the practice weight in these practices assigned a 0.5 weight to the report of each clinician. Analyses were conducted using SAS, version 9.4.
Our study had several limitations. First, the data were self-reported in a national random sample. All sample surveys are subject to sampling error and to possible errors of response or non response.
Second, neither the sample frame nor the questionnaire included data on patient-centered medical home certification or membership in an ACO.
Third, most of the survey content pertained to the provision of care to patients ages sixty-five and older and might not reflect practice with younger patients.
Fourth, responding clinicians were reporting on the roles of themselves and others, and the reports may be more reliable as they pertain to the clinician’s personal tasks and roles.
Fifth, the group of practices with social workers but not RNs was small (unweighted n=19, weighted n=15), and estimates should be viewed with caution.
Clinicians in 39.7 percent of the practices worked with neither an RN nor a social worker; clinicians in 40.4 percent worked with an RN but not a social worker; clinicians in 4.5 percent worked with a social worker but not an RN, and clinicians in 15.4 percent worked with both.
Tests of significance shown in the exhibit compare all four practice staffing types. There were no significant differences in the responses of clinicians in the differently staffed practices with respect to frail elderly patient encounter volume, time spent with their teams in formal meetings or huddles, or time spent communicating with informal caregivers. Responses did differ significantly by practice type in hours spent on communication with other providers about care coordination; clinicians in practices with neither RNs nor social workers reported spending the highest number of hours.
Staffing types differed significantly by payer type: Clinicians in practices employing both an RN and a social worker were more likely to report a higher share of funding by Medicaid; those with an RN and no social worker or with neither an RN nor a social worker reported a higher share of Medicare fee-for-service funding.
This open-access article taken from https://www.healthaffairs.org/doi/10.1377/hlthaff.2019.00030 in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license.