Targeted Use of Agencies for Personal Care Services Analysis of a Health Plan's Natural Experiment JULY 2023 AUTHORS Ann Hwang, MD, Bailit Health About the Author At the time this publication was written, Ann Contents Hwang, MD, was a senior consultant with Bailit Health. Bailit Health is a health policy consult- 3 Executive Summary ing firm d edicated t o a ssisting p ublic a gencies 3 Introduction with the design and management of health and human service programs. 4 Background 5 Analysis Methods Acknowledgments This paper summarizes an analysis prepared by Population Moss Adams. Moss Adams is a fully integrated Data Sources professional services firm dedicated to assisting Measures clients with growing, managing, and protecting Statistical Methods prosperity. The Moss Adams team producing the analysis used in this study includes Karl Rebay, 7 Results partner, and Justin Stewart, manager. Characteristics of the Study Population Utilization and Spending The author thanks the Health Plan of San Mateo for its assistance with this research: 8 Limitations Maya Altman, Chief Executive Officer (retired) 9 Learnings and Implications Luarnie Bermudo, Provider Services Director Liss Ieong, Director of Health Analytics Implications for Policy and Care Delivery Amy Scribner, Chief Health Officer 13 Appendices Ru Yan, Statistician 19 Endnotes About the Foundation The California Health Care Foundation (CHCF) is an independent, nonprofit philanthropy that works to improve the health care system so that all Californians have the care they need. We focus especially on making sure the system works for Californians with low incomes and for communities who have traditionally faced the greatest barriers to care. We partner with leaders across the health care safety net to ensure they have the data and resources to make care more just and to drive improvement in a complex sys- tem. CHCF informs policymakers and industry leaders, invests in ideas and innovations, and connects with changemakers to create a more responsive, patient-centered health care system. California Health Care Foundation www.chcf.org 2 Executive Summary This small observational study affirms the impor- tance of personal care services and demonstrates CalAIM (California Advancing and Innovating how disruptive changes to the provision of that Medi-Cal) and other initiatives to advance care for care can be for Medi-Cal enrollees. Medi-Cal stake- Medi-Cal enrollees present an important opportu- holders should thus plan carefully for any changes nity to improve the personal care services that help that impact the provision of personal care services, older adults and people with disabilities remain in particularly changes that decrease or limit exist- their homes and communities. For over a half million ing options. Importantly, this study suggests that Medi-Cal enrollees who need help with activities of although contract mode may be a more expensive daily living, personal care services are essential for model for delivering IHSS, in this selected popula- health, well-being, and independence. However, tion it may be associated with lower overall costs due ultimately, the effectiveness of these personal care to lower utilization of institutional long-term care services hinges on an individual's ability to secure and and lower rates of hospitalization. More broadly, in retain a reliable provider - which is difficult for many. the context of CalAIM, these findings suggest that agency-mediated personal care services could play In California, most personal care services are deliv- an important role as a cost-effective alternative to ered through the In-Home Supportive Services institutional long-term care. Given these findings (IHSS) Program. While most IHSS care recipients do and the critical role personal care services play in successfully hire an independent provider (often a keeping people in their homes and communities, family member or friend), many people who need Medi-Cal stakeholders should consider opportu- help at home struggle to find a reliable person to nities and policies to strengthen access to these help them.1 services, through all modes of delivery, for all con- sumers who need them. While the vast majority of IHSS recipients find, hire, and manage their own provider - a core and unique feature of the program's design since its inception - IHSS can also be provided through Introduction an agency that employs IHSS providers and con- CalAIM (California Advancing and Innovating Medi- nects them to care recipients. This delivery model Cal) is a far-reaching, multiyear plan to transform is known as contract mode IHSS. However, access California's Medi-Cal program and improve inte- to contract mode IHSS is very limited, with only one gration with other social services. CalAIM and the county in California currently offering this option. state's Master Plan for Aging seek to improve access to services and create a more equitable and inte- This report describes findings from an analysis of grated system of care for older adults and people the experiences of a small group of high-need, with disabilities. Strengthening access to personal high-cost Medi-Cal enrollees in San Mateo County care services should be an important priority in who were receiving contract mode IHSS but then the effort to improve care for Medi-Cal enrollees lost access due to policy changes. In this group, the who need long-term services and supports (LTSS). discontinuation of contract mode IHSS was associ- A study in Los Angeles County found that roughly ated with higher per-member use of institutional one of every 20 eligible In-Home Supportive long-term care, and more inpatient admissions. Services (IHSS) recipients did not have a provider These developments translated to higher spend- in January 2013, and eight months later, 17% of ing on selected categories of medical services and these recipients were still without a provider.2 A long-term services and supports. Targeted Use of Agencies for Personal Care Services www.chcf.org 3 2021 report by the state auditor noted that the short supply in many communities, the availability of number of recipients statewide who lacked care in-home care for Medi-Cal enrollees could also cre- increased from January 2015 through December ate capacity, potentially reducing prolonged inpatient 2019, from 33,000 to more than 40,000 on average hospital stays due to bottlenecks for post-acute care.7 each month.3 A recent statewide poll found that in the past 12 months, 14% of Californians tried to Funding and administration of IHSS in California get home-based support for themselves or a fam- relies on federal, state, and county government ily member. Of those who tried to get home-based participation, with all three levels of government support, 35% were unable to do so.4 contributing to program funding.8 The federal gov- ernment contributes just over half (54%) of costs. Much is still to be learned about how best to deliver The remaining nonfederal share of IHSS costs is personal care services. This report describes a covered by the state and counties, with counties natural experiment that occurred when an agency- paying for about 10% of IHSS costs in FY 2022–23.9 mediated mode of delivering IHSS was eliminated in a particular county, and examines the effects of that At the state level, the program is administered by change on cost and utilization for selected medical the California Department of Social Services and services and long-term services and supports. This the Department of Health Care Services. At the information may help inform policies that support county level, county human services departments and strengthen the availability of various modes of determine consumer eligibility for IHSS and assess personal care service delivery for Medi-Cal enroll- service needs. County IHSS public authorities estab- ees going forward. lish a registry of IHSS providers, investigate the qualifications and background of potential provid- ers, establish a referral system to connect providers and consumers, provide training for consumers and Background providers, and collectively bargain with IHSS work- California's flagship personal care services pro- ers over wages and benefits. gram, IHSS, had an average monthly caseload of more than 586,000 Medi-Cal enrollees in fiscal year California's predominant model of IHSS delivery (FY) 2021–22, providing in-home assistance with is independent provider mode, in which consum- essential activities such as feeding, dressing, bath- ers directly hire and manage their own personal ing, and toileting.5 (An overview of IHSS is available care aides. In contrast, in contract mode services, in the California Health Care Foundation fact sheet an agency hires and manages personal care aides. In-Home Supportive Services 101: Opportunities While not commonly used, this model may be help- and Challenges Under CalAIM.)6 IHSS has enabled ful in circumstances where consumers are unable to many Californians to remain in their homes, rather successfully find, hire, and manage their own aides. than in nursing homes, and upholds a commit- Contract mode services cost more on an hourly basis ment to self-direction, which prioritizes participant and tend to be reserved for those who are unable choice, control, and flexibility. to navigate independent provider mode. While a number of counties used to offer contract mode, In addition to supporting the preferences of many currently only one county (San Francisco) does.10 older adults and people with disabilities, home- based care is typically less costly than institutional During a recent federal demonstration program, care. In California, where nursing home beds are in IHSS became the responsibility of managed care plans in seven counties but was later pulled back California Health Care Foundation www.chcf.org 4 and administered once again by the counties.11 Who Might Use Contract Mode Services? This report examines the cost and utilization associ- David is a 55-year-old man with schizophrenia ated with enrollees of one Medi-Cal managed care and severe chronic obstructive lung disease. plan who were receiving contract mode IHSS but He can have intense paranoia and is reluctant to then lost access to it as a result of this transition. let anyone into his home. With contract mode This Medi-Cal plan operates under an arrangement services, the agency care manager helps support whereby it has responsibility for all Medi-Cal man- David to accept care from the agency's caregiv- aged care enrollees in the county. ers. All his caregivers have been trained to prompt him to take his medications and to allow basic personal care and some assistance to keep his apartment habitable. Before he had this support Analysis Methods from the agency, David was not able to success- fully retain his caregivers and had trouble keeping Population his apartment clean, putting him at risk for losing his housing. This study examines the experiences of Medi-Cal members in San Mateo County who were using Rosa is an 80-year-old woman who has lived contract mode IHSS and who lost access to contract alone for many decades. She has hyperten- mode services. Specifically, under the Coordinated sion and congestive heart failure, with impaired Care Initiative (CCI), Health Plan of San Mateo mobility after a stroke. She lives in an isolated area without public transportation. Her daughter lives (HPSM) (and other Medi-Cal plans in the seven CCI in another state and has been trying to help Rosa counties) assumed IHSS payment responsibility for find personal care aides, but the small number of its members in 2014. Over the next few years, a small daily hours she's been approved for, coupled with group of HPSM members eligible for IHSS could long travel distances, makes it challenging for Rosa receive IHSS through an agency (Homebridge). to find help. Through contract mode services, an agency is able to provide care to Rosa because the In HPSM's program, contract mode IHSS could be same caregiver can also serve other IHSS recipi- provided for people unable to self-direct their own ents who live near Rosa on the same day. personal care aides. These could include people Guang is a 72-year-old woman with moderate who had cognitive or behavioral health challenges, dementia. She lived with her husband until he coupled with a lack of family or social support. A died a few years ago, and she is estranged from small minority (no more than 5%) of IHSS recipients her adult children. Her dementia has worsened accessed contract mode. However, starting January such that she needs prompting and supervision 1, 2018, per state policy IHSS services were once for feeding and bathing, and help with grocery again carved out of managed care in CCI counties, shopping and light housekeeping. Because of her with responsibility for the program returned to the dementia, she is unable to find and manage her counties. At this juncture, in San Mateo County, own caregivers. Through contract mode services, a contract mode IHSS was no longer offered, and home care agency provides caregivers to support members using Homebridge were largely transi- her remaining in her home. tioned into independent provider mode IHSS.12 The population for this study included people used contract mode IHSS services for at least three continuously enrolled with HPSM for at least the months prior to the discontinuation of contract period from July 2017 through June 2018, who mode services at the end of December 2017. Targeted Use of Agencies for Personal Care Services www.chcf.org 5 The population included three subpopulations: the entire population showed directionally similar results to those presented here.) $ People eligible for and enrolled in both Medicare and Medi-Cal (dually eligible enroll- ees) who received both sets of benefits through Data Sources HPSM's Cal MediConnect integrated plan HPSM provided claims data with utilization and cost information for the study population, starting $ Dually eligible enrollees who received Medi- from the study start date (January 1, 2016) or their Cal benefits through HPSM but Medicare enrollment into IHSS, whichever was later, until their benefits elsewhere disenrollment from HPSM, death, or the end of the $ Medi-Cal–only HPSM members study period (December 31, 2019). Cost informa- tion was based on paid claims and did not include The initial population included 154 HPSM mem- non-claims-based payments or pharmacy rebates. bers. Members were excluded from further analysis Cost information included paid claims for IHSS, insti- if they received services from the Institute on Aging tutional long-term care (LTC), acute inpatient care, (IOA) after contract mode IHSS was discontinued post-acute skilled nursing facility (SNF) care, emer- at the end of December 2017. This exclusion was gency department (ED) services, and pharmacy. applied because IOA provides some services that are similar to those provided in contract mode IHSS. Plan responsibility for claims varied by Medicare Eighteen members were excluded for this reason. and Medi-Cal enrollment status, as detailed in Additionally, members with total costs that were Appendix A, Table A1. The cost analysis compo- greater than two standard deviations above the nents take into account claims fully paid by HPSM, mean were excluded. This was done to prevent the and vary based on Medicare and Medi-Cal enroll- effect of a small number of very high-cost members ment status (Table 1). While plan responsibility for from skewing the results. Eleven members were IHSS claims changed over the course of the study, excluded for this reason. The final study population the data source for IHSS claims was consistent, so thus included 125 members. (A separate analysis of they are treated in the same way throughout. Table 1. Service Categories Included in Analysis, for Each Subpopulation Based on Medicare and Medi-Cal Enrollment Status DUALLY ELIGIBLE ENROLLEES DUALLY ELIGIBLE ENROLLEES IN CAL MEDICONNECT WITH HPSM FOR MEDI-CAL (BOTH MEDICARE AND MEDI-CAL) ONLY AND RECEIVING MEDI-CAL–ONLY SERVICE CATEGORY WITH HPSM MEDICARE ELSEWHERE HPSM MEMBERS IHSS CHECK CHECK CHECK Institutional LTC CHECK CHECK CHECK Acute inpatient CHECK CHECK Post-acute SNF CHECK ED CHECK CHECK Pharmacy CHECK CHECK Source: Moss Adams data, 2022. Notes: ED is emergency department. HPSM is Health Plan of San Mateo. IHSS is In-Home Supportive Services. LTC is long-term care. SNF is skilled nursing facility. California Health Care Foundation www.chcf.org 6 Measures dually eligible enrollees included 72 people who This study describes health care utilization before received both their Medicare and Medi-Cal ben- and after the discontinuation of contract mode IHSS efits through HPSM's Cal MediConnect integrated at the end of 2017. Claims data were used to assess plan, and 32 people who received only their Medi- utilization and costs of IHSS, institutional LTC admis- Cal benefits through HPSM and received Medicare sions, acute inpatient admissions, post-acute SNF benefits elsewhere (i.e., through fee-for-service admissions, ED visits, and pharmacy. Costs were Medicare or another Medicare Advantage plan). based on paid claims and are reported in actual dol- In circumstances where individuals had a change lars; no adjustments were made for inflation. in enrollment status during the study period, they were assigned to the enrollment status in which they resided the longest during the study period. Statistical Methods The number of member-months for each year of the This was a descriptive study. Cost and utilization study are shown in Appendix A, Table A2. data were tabulated for the study population before and after the transition date. Due to the small sam- ple size, no statistical testing was completed. Utilization and Spending Utilization and spending were analyzed by service category (detailed results are provided in Appendix A, Tables A3 through A9). Per-member per-month Results (PMPM) utilization by service category showed an increase in utilization across all service categories Characteristics of the Study except pharmacy after the discontinuation of con- Population tract mode IHSS (Table 2). The study population was 72.8% female and had a mean age of 68.7 years (range: 23 to 101). Most of Per-member utilization of institutional LTC more the population was dually eligible for Medicare and than tripled in the year after the discontinuation of Medi-Cal, with just 21 Medi-Cal–only members. The contract mode IHSS and was more than five times Table 2. Per-Member Per-Month Utilization by Service Category YEARS BEFORE YEARS AFTER DIRECTION OF DISCONTINUATION DISCONTINUATION CHANGE AFTER OF CONTRACT MODE IHSS OF CONTRACT MODE IHSS DISCONTINUATION SERVICE CATEGORY UTILIZATION OF CONTRACT MODE METRICS 2016 2017 2018 2019 IHSS IHSS hours PMPM 39 70 82 79 ↑ Institutional LTC days per 1,000 599 468 1,478 2,349 ↑↑ member-months Acute inpatient admissions per 32 41 54 42 ↑ 1,000 member-months Post-acute SNF days per 1,000 492 473 663 581 ↑ member-months ED visits per 1,000 member-months 155 269 249 278 ↑* Pharmacy scripts PMPM 6.11 6.12 5.81 5.75 ↓ Source: Moss Adams analysis, 2022. Notes: ED is emergency department. IHSS is In-Home Supportive Services. LTC is long-term care. PMPM is per-member per-month. SNF is skilled nursing facility. *ED visits in 2018 and 2019 were about the same as ED visits in 2017 but slightly higher when compared with ED visits in 2016 and 2017. Targeted Use of Agencies for Personal Care Services www.chcf.org 7 higher in the subsequent year (compared with the Figure 1. Overall Weighted-Average PMPM Spending by Year year prior to the discontinuation of contract mode). IHSS hours also increased, which may reflect $3500 improved access for those who were successful in $3076 $2967 $3000 securing and retaining an IHSS provider. The final $2635 year of contract mode IHSS (2017) was marked $2500 by significant workforce shortages for the agency $2000 $1783 providing services, which may have contributed to enrollees not receiving all of the hours they were $1500 approved for. In addition, acute inpatient admis- $1000 sions and ED visits increased slightly, and pharmacy prescriptions declined slightly in the years following $500 the discontinuation of contract mode IHSS. $0 2016 2017 2018 2019 Overall weighted-average PMPM spending Years Before Years After Discontinuation of Discontinuation of increased from $1,783 and $2,635 in 2016 and Contract Mode IHSS Contract Mode IHSS 2017, respectively, to $3,076 and $2,967 in 2018 Source: Moss Adams analysis, 2022. and 2019, respectively (Figure 1). This increase was Note: PMPM is per-member per-month. driven by increases in PMPM spending on institu- tional LTC, post-acute SNF care, and acute inpatient care (spending by service category is shown in Medicare elsewhere). For this group, institutional Tables A3 through A8 in Appendix A). long-term care spending rose substantially after the discontinuation of contract mode IHSS, but this was To understand the overall observed change offset by even larger decreases in IHSS costs. in spending, total and PMPM spending were calculated separately for each of the three sub- For Medi-Cal–only enrollees, there was an increase populations: Cal MediConnect enrollees, dually in total and PMPM spending in 2018. Total and eligible enrollees receiving only Medi-Cal benefits PMPM spending declined from 2018 to 2019 but from HPSM and receiving Medicare elsewhere, and remained above 2017 levels. Medi-Cal–only enrollees. The results were calcu- lated separately for each of these groups because The PMPM results are summarized in Table 3 (see information for certain service categories was not next page); results for total spending, as well as available for all of them. detailed results by service category, can be found in Appendix A, Table A9. For Cal MediConnect enrollees, where HPSM was responsible for both Medi-Cal and Medicare ben- efits and costs, total spending and spending on a PMPM basis increased markedly from the 2016– Limitations 2017 period to the 2018–2019 period. This study examined the experience of a very small population. Results could be skewed by individu- Total PMPM spending by HPSM declined slightly als with extreme patterns of utilization. To address during the study period for dually eligible individu- this concern, members with the highest costs were als enrolled with HPSM for Medi-Cal only (receiving removed from the study population (a separate California Health Care Foundation www.chcf.org 8 Table 3. PMPM Spending by Subpopulation YEARS BEFORE DISCONTINUATION YEARS AFTER DISCONTINUATION OF CONTRACT MODE IHSS OF CONTRACT MODE IHSS SUBPOPULATION 2016 2017 2018 2019 Dually eligible enrollees in Cal Medi-Connect (both Medicare and $1,639 $3,032 $3,480 $3,636 Medi-Cal) with HPSM (n = 72) Dually eligible enrollees with HPSM for Medi-Cal only and receiving Medicare $2,017 $2,032 $1,965 $1,887 elsewhere (n = 32) Medi-Cal–only HPSM members (n = 21) $1,941 $2,103 $3,370 $2,283 Source: Moss Adams analysis, 2022. Notes: HPSM is Health Plan of San Mateo. IHSS is In-Home Supportive Services. PMPM is per-member per-month. analysis was conducted using the entire population without exclusions, and the results were direction- Learnings and ally similar to those presented in this paper). Even Implications within the remaining population, however, there This study describes the utilization and spending for can be large amounts of utilization driven by a small certain medical, pharmacy, and long-term services number of members. and supports of a population using IHSS, before and after the discontinuation of contract mode services. This study is observational in nature, and there was no control group. The observed trends that The high utilization and spending associated with showed increases in spending over the four-year the study population confirms that this is a high- period could thus reflect the increasing needs of need, high-cost population, similar to patterns seen a frail population, rather than solely the impact of in other studies.13 the discontinuation of contract mode services. The study also did not adjust for medical trends or infla- This study adds to prior research that has predomi- tion. Similarly, the observed trends could also be nantly compared outcomes related to independent influenced by policy shifts or other factors shaping provider mode versus contract mode.14 This study health care delivery or payment, such as the sever- differs in that it examines the experience of a pop- ity of the flu season of 2017–18. ulation specifically identified as benefiting from agency services, who then lost access to those The analyzed claims in this study varied depend- services. ing on whether enrollees were dually eligible for Medicare and Medi-Cal, and on whether they were Within a small population, this study demonstrates enrolled with HPSM for Medi-Cal only or for both an increase in overall PMPM spending during the Medicare and Medi-Cal. The study analyzed claims study period, with some variation between sub- provided by a health plan and thus did not capture populations. There were especially pronounced spending on services that the health plan was not increases in spending for institutional LTC admis- responsible for. For dually eligible enrollees who sions after the discontinuation of contract mode were receiving only Medi-Cal benefits from the services. For the study population as a whole, this health plan, the costs of their medical care would increase in spending outweighed the reduction in largely have been paid for by their Medicare payer, and thus were not captured in this study. Targeted Use of Agencies for Personal Care Services www.chcf.org 9 IHSS costs associated with the transition from con- ranging from centralized data systems for locating tract mode to individual provider mode. and contacting potential providers, to changes in roles or payment structures that might attract more These findings suggest that contract mode is a care providers to IHSS. The recently established distinct and noninterchangeable mode of delivery permanent backup provider system is another for IHSS. Within the highly selected population example of a strategy to support individuals' access identified as benefiting from contract mode IHSS, to care.15 In addition, for people who may face more contract mode was not readily replaced by indi- challenges in finding a provider, whether due to the vidual provider mode services. The high utilization impacts of their illness or disability, or because of and spending after the discontinuation of contract their social circumstances, personal care services mode services also underscores the vulnerability of mediated through an agency appears to provide an this population to disruptions to care delivery, even important option. with careful transition planning. Fragmented policy and payment approaches fur- ther complicate efforts to ensure that individuals Implications for Policy and Care receive the form of assistance that they need. This is Delivery a version of the "wrong pocket problem": Counties This small observational study affirms the impor- contribute to the cost of IHSS but don't reap sav- tance of personal care services and demonstrates ings that might accrue to Medicare or Medi-Cal how disruptive changes to the provision of that care if more robust, but expensive, services are deliv- can be. Policymakers should thus plan carefully for ered. Thus, counties may not seek out investments any changes that impact the provision of personal in more expensive models of delivery of IHSS like care services, particularly changes that decrease or contract mode for specific subpopulations, even if limit existing options. those investments might result in overall, system- wide savings. Medi-Cal managed care plans would Further, this study describes increases in utiliza- benefit from reductions in institutional long-term tion and spending that were temporally associated care spending but don't directly pay for IHSS ser- with the discontinuation of contract mode services. vices, and the state bears only some of the costs of These preliminary findings support the need for the IHSS program because of federal and county deeper examination of not just whether but how contributions. to deliver personal care services to a high-need population. This examination should encompass Given this fragmentation, policymakers and payment, the delivery model, and the scope and Medi-Cal stakeholders should support the imple- quality of services received. Understanding which mentation of strategies to establish clear lines of components of the contract mode model are most sight across all entities involved in the payment valuable could help support replication of those and delivery of all long-term supports and services, components across the care delivery system. including public authorities, county agencies, health plans, and community partners. The new managed Ultimately, the effectiveness of IHSS hinges on care contract that takes effect in 2024 develops a an individual's ability to secure a reliable, high- framework for collaboration, with requirements quality IHSS provider. There are a variety of policy around care coordination and data sharing across and programmatic solutions that could support entities in each county. In addition, the Enhanced IHSS recipients in finding independent providers, Care Management benefit creates accountability California Health Care Foundation www.chcf.org 10 for case management across all services for people POLICY OPPORTUNITIES who are at risk of nursing home placement. In addi- To improve access to and delivery of personal tion to practical supports to foster collaboration, care services to Medi-Cal enrollees, stakeholders commitment from leaders across these entities will should consider taking the following actions: be needed to solve identified challenges and over- come barriers to collaboration. $ Examine current challenges and support opportunities to strengthen the ability of IHSS recipients to identify and secure providers In addition, the option of Community Supports under independent provider mode. under CalAIM presents an important opportunity to $ Support research into how best to pay for and expand the availability of agency-mediated personal deliver personal care services, including how to care services, as a complement to and augmenta- identify the delivery mode that works best for tion of existing personal care options. CalAIM's each consumer. Community Supports are services that Medi-Cal $ At the county level, support efforts to ensure managed care plans may provide in lieu of services clear lines of sight across all entities involved traditionally covered by Medicaid. CalAIM currently in the care of people with complex and over- includes a menu of 14 preapproved Community lapping medical, behavioral health, LTSS, and Supports, including personal care and homemaker health-related social needs. services, and caregiver respite. These services are $ Monitor and learn from the use of agency above and beyond any approved county IHSS services in CalAIM Community Supports, and hours and might be used when additional hours support relationship building between plans are required and IHSS benefits are exhausted, and agencies. during an IHSS waiting period, or to help mem- $ Explore policies to allow plans to collaborate bers who are not eligible to receive IHSS avoid a with IHSS public authorities to offer contract short-term stay in a skilled nursing facility. These mode IHSS for certain beneficiaries. services are provided through agencies. However, under Community Supports, contract mode cannot be provided in the way it was used in San Mateo County, as an alternative delivery mode for IHSS for care services might benefit care recipients who are those unable to hire and retain an independent pro- unable to thrive in independent provider mode vider. Nonetheless, the relationships established IHSS or need a bridge to it. Managed care plans between plans, counties, and service agencies, and might have an aligned interest in collaborating with the alignment of incentives for plans, could create counties to provide contract mode services for this a platform for broader expansion of contract mode subset of the plan's members. This could occur services beyond CalAIM. As of November 2022, through a financing mechanism in which managed 97 plans by county and 53 of the 58 counties in care plans might be allowed to pay the difference California have elected to offer both the personal between contract mode and independent provider care and homemaker services and the caregiver mode, in order to offer contract mode to certain respite Community Supports by January 2024.16 specific populations. Given the observations from this limited study, Ultimately, personal care services are, at their core, Medi-Cal stakeholders could also consider where intensely personal. While independent provider and how targeted use of contract mode personal services are understandably the preferred and Targeted Use of Agencies for Personal Care Services www.chcf.org 11 predominant model for most personal care recipi- ents, a one-size-fits-all approach can leave behind those who have difficulty managing their own care. Figuring out how to get people the personal care services they need - such as by strengthening the systems supporting independent provider mode IHSS, increasing the availability of contract mode IHSS, and augmenting IHSS with the Community Supports personal care services in CalAIM - will be essential to help the growing aging population in California thrive in their homes and communities. California Health Care Foundation www.chcf.org 12 Appendix A. Additional Data Plan Responsibility for Claims This study analyzed claims data provided by Health Plan of San Mateo (HPSM) for the study population. Utilization and cost information was obtained from paid claims for In-Home Supportive Services (IHSS), insti- tutional long-term care (LTC), acute inpatient care, post-acute skilled nursing facility (SNF) care, emergency department (ED) services, and pharmacy. Plan responsibility for claims in each of these service categories varied by Medicare and Medi-Cal enrollment status, as shown below in Table A1. Table A1. Plan Responsibility for Claims by Service Category, for Each Subpopulation Based on Medicare and Medi-Cal Enrollment Status (N = 125) DUALLY ELIGIBLE ENROLLEES DUALLY ELIGIBLE ENROLLEES IN CAL MEDICONNECT (BOTH WITH HPSM FOR MEDI-CAL MEDICARE AND MEDI-CAL) ONLY AND RECEIVING MEDI-CAL–ONLY HPSM SERVICE CATEGORY WITH HPSM (N = 72) MEDICARE ELSEWHERE (N = 32) MEMBERS (N = 21) IHSS Yes Yes Yes Institutional LTC Yes Yes Yes Acute inpatient Yes Partial: only crossover Yes claims* that are not consistently submitted Post-acute SNF Yes Partial: only crossover No: There is no distinction in claims that are not Medi-Cal between short-term consistently submitted skilled and long- term custodial care in a skilled nursing facility; thus, this category does not apply. ED Yes Partial: only crossover Yes claims that are not consistently submitted Pharmacy Yes Partial: includes claims for Yes Medicaid-covered drugs, with crossover claims for Medicare-covered drugs Source: Moss Adams data, 2022. Notes: In circumstances where members had a change in enrollment status during the study period, they were assigned to the enrollment status in which they resided the longest during the study period. ED is emergency department. HPSM is Health Plan of San Mateo. IHSS is In-Home Supportive Services. LTC is long-term care. SNF is skilled nursing facility. *"Crossover claims" are claims for a member who is eligible for both Medicare and Medi-Cal, where Medicare pays a portion of the claim and Medi-Cal is billed for any remaining deductible and/or coinsurance. Targeted Use of Agencies for Personal Care Services www.chcf.org 13 Member Months by Subpopulation The study population included three subpopulations: (1) people eligible for and enrolled in both Medicare and Medi-Cal (dually eligible enrollees) who received both sets of benefits through HPSM, (2) dually eligible enrollees who received Medi-Cal benefits through HPSM but Medicare benefits elsewhere, and (3) Medi- Cal–only HPSM members. Table A2 shows the number of people in each subpopulation, as well as the corresponding number of member-months associated with each subpopulation in each study year. In circum- stances where individuals had a change in enrollment status during the study period, they were assigned to the enrollment status in which they resided the longest during the study period. Table A2. Member-Months Included in Analysis, by Subpopulation and Study Year (N = 125) YEARS BEFORE DISCONTINUATION YEARS AFTER DISCONTINUATION OF CONTRACT MODE IHSS OF CONTRACT MODE IHSS SUBPOPULATION 2016 2017 2018 2019 Dually eligible enrollees in Cal MediConnect (both Medicare and 794 862 857 743 Medi-Cal) with HPSM (n = 72) Dually eligible enrollees with HPSM for Medi-Cal only and receiving Medicare 347 372 372 323 elsewhere (n = 32) Medi-Cal–only HPSM members (n = 21) 212 222 227 216 Total, all subpopulations (N = 125) 1,353 1,456 1,456 1,282 Source: Moss Adams analysis, 2022. Notes: HPSM is Health Plan of San Mateo. IHSS is In-Home Supportive Services. Observed Utilization and Spending for Long-Term Services and Supports (In-Home Supportive Services and Institutional Long-Term Care) In-Home Supportive Services (IHSS) Utilization and Spending Utilization of in-home supports and services, as measured by IHSS hours received per member per month, increased during the study period. However, the expenses associated with IHSS decreased after 2017, likely due to the lower payment rates associated with independent provider mode relative to contract mode (Table A3). Table A3. In-Home Supportive Services Utilization and Spending (N = 125) YEARS BEFORE DISCONTINUATION YEARS AFTER DISCONTINUATION OF CONTRACT MODE IHSS OF CONTRACT MODE IHSS IHSS UTILIZATION AND SPENDING METRICS 2016 2017 2018 2019 Total number of hours 53,003 101,513 118,909 100,963 Number of hours PMPM 39 70 82 79 Total spending $1,142,888 $2,264,401 $1,801,132 $1,602,533 Spending PMPM $845 $1,555 $1,237 $1,250 Source: Moss Adams analysis, 2022. Notes: IHSS is In-Home Supportive Services. PMPM is per member per month. California Health Care Foundation www.chcf.org 14 Institutional Long-Term Care (LTC) Utilization and Spending The study population had high and increasing utilization of institutional long-term care during the study period. LTC admissions increased from 15 in 2016 to 37 in 2019. The total number of institutional LTC days dramatically increased after the transition to independent provider mode: from 810 in 2016 and 682 in 2017, to 2,152 in 2018 and 3,012 in 2019. The increase in utilization was accompanied by an even larger increase in expenditures on an absolute and per-member per-month basis, with total spending increasing from $57,595 in 2016 to $726,894 in 2019. This reflects an increase in per-member per-month costs from $43 and $56 in 2016 and 2017, respectively, to $316 and $567 in 2018 and 2019, respectively (Table A4). Table A4. Institutional Long-Term Care Utilization and Spending (N = 125) YEARS BEFORE DISCONTINUATION YEARS AFTER DISCONTINUATION OF CONTRACT MODE IHSS OF CONTRACT MODE IHSS INSTITUTIONAL LTC UTILIZATION AND SPENDING METRICS 2016 2017 2018 2019 Total number of admissions 15 25 31 37 Number of admissions per 1,000 member- 11 17 21 29 months Total number of days 810 682 2,152 3,012 Number of days per 1,000 member-months 599 468 1,478 2,349 Total spending $57,595 $81,686 $460,182 $726,894 Spending PMPM $43 $56 $316 $567 Source: Moss Adams analysis, 2022. Notes: IHSS is In-Home Supportive Services. LTC is long-term care. PMPM is per member per month. Observed Utilization and Spending for Medical Services (Acute Inpatient, Post-Acute Skilled Nursing Facility, Emergency Department, and Pharmacy) Acute Inpatient Utilization and Spending Utilization and spending for acute inpatient care were assessed in the combined Cal MediConnect and Medi- Cal–only subpopulations. Dually eligible enrollees receiving Medicare elsewhere were not included in this assessment, because HPSM claims for acute inpatient admissions for this population include only crossover claims, which are not consistently submitted. The number of acute inpatient admissions per member per month was higher in the 2018–2019 period than in the preceding two years. Spending associated with these admissions was also higher on a per-member per-month basis in 2018–2019 than in 2016–2017, rising from $447 and $626 in 2016 and 2017, respectively, to $1,187 and $743 in 2018 and 2019, respectively (Table A5). Targeted Use of Agencies for Personal Care Services www.chcf.org 15 Table A5. Acute Inpatient Utilization and Spending, Cal MediConnect and Medi-Cal–Only Enrollees (n = 93) YEARS BEFORE DISCONTINUATION YEARS AFTER DISCONTINUATION OF CONTRACT MODE IHSS OF CONTRACT MODE IHSS ACUTE INPATIENT UTILIZATION AND SPENDING METRICS 2016 2017 2018 2019 Total number of admissions 32 44 58 40 Number of admissions per 1,000 member- 32 41 54 42 months Total spending $450,027 $678,642 $1,286,832 $712,378 Spending PMPM $447 $626 $1,187 $743 Source: Moss Adams analysis, 2022. Notes: IHSS is In-Home Supportive Services. PMPM is per member per month. Post-Acute Skilled Nursing Facility (SNF) Utilization and Spending Trends in utilization of and spending on post-acute SNF care were assessed only for the Cal MediConnect population, as post-acute SNF is not categorized separately in Medi-Cal, and HPSM claims for SNF for dually eligible individuals enrolled with HPSM for Medi-Cal only would consist only of crossover claims, which are inconsistently submitted. In the Cal MediConnect population, there was an upward trend in post-acute SNF utilization across the four- year period, with an increase in post-acute SNF admission days from 492 and 473 per 1,000 member-months in 2016 and 2017, respectively, to 663 and 581 per 1,000 member-months in 2018 and 2019, respectively. Per-member per-month spending on post-acute SNF care also trended upward during the study period, from $272 and $256 before the transition, to $415 and $337 after (Table A6). Table A6. Post-Acute Skilled Nursing Facility Utilization and Spending, Cal MediConnect Enrollees (n = 72) YEARS BEFORE DISCONTINUATION YEARS AFTER DISCONTINUATION OF CONTRACT MODE IHSS OF CONTRACT MODE IHSS POST-ACUTE SNF UTILIZATION AND SPENDING METRICS 2016 2017 2018 2019 Total number of admissions 15 21 24 19 Number of admissions per 1,000 member- 19 24 28 26 months Total number of days 391 408 568 432 Number of days per 1,000 member-months 492 473 663 581 Total spending $215,598 $220,291 $355,281 $250,578 Spending PMPM $272 $256 $415 $337 Source: Moss Adams analysis, 2022. Notes: IHSS is In-Home Supportive Services. PMPM is per member per month. SNF is skilled nursing facility. California Health Care Foundation www.chcf.org 16 Emergency Department Utilization and Spending Data for ED utilization and spending were examined for the Cal MediConnect and Medi-Cal–only popula- tions. The population of dually eligible members enrolled with HPSM for Medi-Cal but receiving Medicare elsewhere was not included in this assessment, because HPSM claims for ED services would reflect only crossover claims, which are not consistently submitted. ED visits and spending on a PMPM basis trended higher from the 2016–2017 period to the 2018–2019 period. However, utilization and spending were particularly high in 2017, which was the year with the highest number of total ED visits as well as the highest total ED spending (Table A7). Table A7. Emergency Department Utilization and Spending, Cal MediConnect and Medi-Cal–Only Enrollees (n = 93) YEARS BEFORE DISCONTINUATION YEARS AFTER DISCONTINUATION OF CONTRACT MODE IHSS OF CONTRACT MODE IHSS ED UTILIZATION AND SPENDING METRICS 2016 2017 2018 2019 Total number of visits 156 292 270 267 Number of visits per 1,000 member-months 155 269 249 278 Total spending $18,625 $37,584 $34,073 $37,306 Spending PMPM $19 $35 $31 $39 Source: Moss Adams analysis, 2022. Notes: ED is emergency department. IHSS is In-Home Supportive Services. PMPM is per member per month. Pharmacy Utilization and Spending Data for pharmacy utilization and spending were examined for the Cal MediConnect and Medi-Cal–only populations. The population of dually eligible individuals enrolled with HPSM for Medi-Cal and receiving Medicare elsewhere was not included in this assessment, because HPSM claims for this category of service consist of crossover claims, which are not consistently submitted. Pharmacy utilization, as measured by the number of filled prescriptions per member per month, declined after the end of contract mode, from 6.11 in 2016 to 5.75 in 2019. Pharmacy spending per member per month also declined over this period, from $525 in 2016 to $495 in 2019 (Table A8). Table A8. Pharmacy Utilization and Spending, Cal MediConnect and Medi-Cal–Only Enrollees (n = 93) YEARS BEFORE DISCONTINUATION YEARS AFTER DISCONTINUATION OF CONTRACT MODE IHSS OF CONTRACT MODE IHSS PHARMACY UTILIZATION AND SPENDING METRICS 2016 2017 2018 2019 Total number of scripts 6,147 6,629 6,302 5,512 Number of scripts PMPM 6.11 6.12 5.81 5.75 Total spending $527,952 $553,672 $540,999 $474,582 Spending PMPM $525 $511 $499 $495 Source: Moss Adams analysis, 2022. Notes: IHSS is In-Home Supportive Services. PMPM is per member per month. Targeted Use of Agencies for Personal Care Services www.chcf.org 17 Summary of Total and Per Member Per Month (PMPM) Spending by Subpopulation and Service Category Total and PMPM spending were calculated separately for each of the three subpopulations: dually eligible enrollees in Cal MediConnect (both Medicare and Medi-Cal) with HPSM, dually eligible enrollees receiving only Medi-Cal benefits from HPSM and receiving Medicare elsewhere, and Medi-Cal–only HPSM members. For each subpopulation, total and PMPM spending for each service category are shown in Table A9. Table A9. Total and PMPM Spending by Subpopulation and Service Category (N = 125) YEARS BEFORE DISCONTINUATION YEARS AFTER DISCONTINUATION OF CONTRACT MODE IHSS OF CONTRACT MODE IHSS 2016 2017 2018 2019 SUBPOPULATIONS AND SERVICE CATEGORIES Total PMPM Total PMPM Total PMPM Total PMPM DUALLY ELIGIBLE ENROLLEES IN CAL MEDICONNECT (BOTH MEDICARE AND MEDI-CAL) WITH HPSM (N = 72) IHSS $169,502 $213 $1,233,590 $1,431 $990,189 $1,155 $878,807 $1,183 Institutional LTC $39,321 $50 $29,419 $34 $207,463 $242 $474,463 $639 Acute inpatient $420,314 $529 $644,395 $748 $974,843 $1,138 $673,541 $907 Post-acute SNF $215,598 $272 $220,291 $256 $355,281 $415 $250,578 $337 ED $10,786 $14 $17,319 $20 $13,697 $16 $14,918 $20 Pharmacy $445,700 $561 $468,366 $543 $441,181 $515 $409,483 $551 Totals for this subpopulation $1,301,221 $1,639 $2,613,380 $3,032 $2,982,654 $3,480 $2,701,790 $3,636 DUALLY ELIGIBLE ENROLLEES WITH HPSM FOR MEDI-CAL ONLY AND RECEIVING MEDICARE ELSEWHERE (N = 32) IHSS $681,650 $1,964 $703,821 $1,892 $519,466 $1,396 $431,437 $1,336 Institutional LTC $18,274 $53 $52,267 $141 $211,352 $568 $177,968 $551 Totals for this subpopulation $699,924 $2,017 $756,088 $2,032 $730,818 $1,965 $609,405 $1,887 MEDI-CAL–ONLY HPSM MEMBERS (N = 21) IHSS $291,736 $1,376 $326,991 $1,473 $291,477 $1,284 $292,289 $1,353 Institutional LTC $0 $0 $0 $0 $41,367 $182 $74,463 $345 Acute inpatient $29,713 $140 $34,247 $154 $311,989 $1,374 $38,837 $180 ED $7,839 $37 $20,265 $91 $20,376 $90 $22,388 $104 Pharmacy $82,252 $388 $85,306 $384 $99,818 $440 $65,099 $301 Totals for this subpopulation $411,540 $1,941 $466,809 $2,103 $765,027 $3,370 $493,076 $2,283 Source: Moss Adams analysis, 2022. Notes: The totals for PMPM dollar amounts may not add up exactly due to rounding. ED is emergency department. HPSM is Health Plan of San Mateo. IHSS is In-Home Supportive Services. LTC is long-term care. PMPM is per member per month. SNF is skilled nursing facility. California Health Care Foundation www.chcf.org 18 Endnotes 1. Lucy Rabinowitz Bailey et al., The 2023 CHCF California 15. California's New IHSS Backup Provider System: What You Health Policy Survey, California Health Care Foundation Need to Know (PDF), Justice in Aging, December 14, 2022. (CHCF), February 16, 2023; and In-Home Supportive Services 16. CalAIM Community Supports Spotlight: Personal Care and Program (PDF), California State Auditor, February 2021. Homemaker Services and Respite Services (PDF), California 2. Fei Wu, Provider Retention and Turnover in the In-Home Department of Health Care Services, November 3, 2022. Supportive Services Program: Statistical and Geo-Spatial Analyses (PDF), County of Los Angeles, Department of Public Social Services, revised March 2016. 3. In-Home Supportive Services Program, California State Auditor. 4. Bailey et al., The 2023 CHCF California Health Policy Survey. 5. 2023-24 Governor's Budget: Caseload Projections (PDF), California Department of Social Services, accessed April 3, 2023. 6. Athena Chapman and Elizabeth Evanson, In-Home Supportive Services (IHSS) 101: Opportunities and Challenges Under CalAIM, CHCF, May 16, 2023. 7. Laurel Beck and Landon Gibson, Anticipating Changes in Regional Demand for Nursing Homes, Public Policy Institute of California, November 2016; and Addressing San Francisco's Vulnerable Post-Acute Care Patients: Analysis and Recommendations of the San Francisco Post-Acute Care Collaborative (PDF), Hospital Council of Northern and Central California, 2018. 8. The 2022-23 Budget: In-Home Supportive Services (PDF), Legislative Analyst's Office, February 2022. 9. In-Home Supportive Services (IHSS): Legislative Briefings (PDF), California Department of Social Services, December 2022. 10. Mark Burns (executive director, Homebridge) to California State Master Plan for Aging Taskforce, LTSS Subcommittee, Memorandum Re: Advocacy for Inclusion of Contract Mode IHSS Access for Counties in Master Plan Recommendations (PDF), December 13, 2019. 11. "2017-18 Budget: The Coordinated Care Initiative: A Critical Juncture," Legislative Analyst's Office, February 27, 2017. 12. A small group of individuals enrolled in a pilot program run by the Institute on Aging continued to receive agency personal care services. As described later in this section, they were excluded from the final study population. 13. Rebecca J. Gorges, Prachi Sanghavi, and R. Tamara Konetzka, "A National Examination of Long-Term Care Setting, Outcomes, and Disparities Among Elderly Dual Eligibles," Health Affairs (Millwood) 38, no. 7 (July 2019): 1110–18. 14. Clark A. Veet, Mary E. Winger, and Suzanne M. Kinsky, "Professional Agency vs Consumer Directed Care Workers: Outcomes in Managed Care," Health and Social Care in the Community 30, no. 4 (July 2022): 1562–67. Targeted Use of Agencies for Personal Care Services www.chcf.org 19