Considering the State Costs and Benefits: In-Home Supportive Services Program

Considering the State Costs and Benefits: In-Home Supportive Services Program M AC TAylor • l e g i s l A T i v e A n A l y s T • JAnuAry 21, 2010 an LaO RepORt 2 LegisLative anaLyst’s Office an LaO RepORt ExEcutivE Summary The In-Home Supportive Services (IHSS) program provides care for over 430,000 recipi- ents, at an annual total cost of about $5.5 billion. The program, which is available to low-in- come elderly and disabled persons, provides various services to recipients in their own homes. Assistance is provided with tasks such as cleaning, meal preparation, bathing, grooming, and helping with medications. The IHSS caseload varies widely with regard to the number of hours of monthly service provided. Benefits of the IHSS Program. For many recipients, the program allows individuals to live at home rather than in an institutional setting (typically, a nursing home). By preventing—or at least delaying—the move to a nursing home, the program can save money for the state. This is because, for a given individual, the annual public sector costs of providing IHSS services is considerably less than the costs of a nursing home. Many other recipients, however, do not face institutionalization in the absence of IHSS services. In these cases, the public sector realizes costs, but there are still benefits for recipients. The program can enhance the quality of life by making it easier to live at home, and it reduces the time and financial burden on family and friends. Net Fiscal Impact on Public Sector. The net impact on the state and counties depends on the mix of the IHSS population—that is, what proportions of the caseload would be institution- alized and would not be institutionalized in the absence of the program. To explore this issue, we created a fiscal model that compares the cost of IHSS to the estimated cost of a long-term care system without IHSS. Our key findings are: ➢ Relative Risk for Institutionalization. Not surprisingly, our model estimates a much greater risk of entering a nursing home for those IHSS recipients who are the most el- derly and using the greatest number of hours. ➢ Net Costs to State and Counties. After accounting for both costs and savings to the state and counties, IHSS probably results in net costs. This is because the savings (in the form of avoided nursing home costs) are probably more than offset by the costs (to provide IHSS and related services) for those recipients who would not be institutional- ized in the absence of the program. ➢ Net Savings to the State. From the state’s perspective alone (not considering the coun- ties), IHSS may well result in net savings. (This is because the state receives a greater share of savings than it incurs in IHSS costs compared to counties.) Policy Implications. From a fiscal perspective, our findings indicate that the state maxi- mizes its net fiscal impact by targeting IHSS services to those recipients who are most likely to enter a nursing home in the absence of the program. This is the approach that the state took LegisLative anaLyst’s Office 3 an LaO RepORt in 2009-10 budget actions affecting the program. The state reduced services to those who are least likely to require institutionalization. Given the state’s continuing fiscal problems, we offer additional options for the Legislature to consider that can achieve state savings through in- creased targeting. We note that the Governor’s 2010-11 budget proposes to eliminate IHSS for all but the most impaired (13 percent of the caseload). 4 LegisLative anaLyst’s Office an LaO RepORt introduction The IHSS program provides in-home care considering the cost- for persons who cannot safely remain in their Effectiveness of iHSS own homes without such assistance. In order to Due to the increasing cost and demand for qualify for IHSS, a recipient must be aged, blind, the IHSS program, one major ongoing legislative or disabled and in most cases have income at concern has been whether the IHSS program has or below the level necessary to qualify for the saved the state money by reducing institutional Supplemental Security Income/State Supplemen- placement costs. This issue of whether IHSS tary Program (SSI/SSP). County social workers results in a net fiscal benefit to the public sector perform an assessment to determine the number is one of cost-effectiveness. Overall, the program of hours and type of services to authorize an would be considered cost-effective if the aggre- IHSS recipient. The recipient is responsible for gate amount the state spends on IHSS is equal to hiring and supervising a provider. Based on the or less than the amount the state would spend on submittal of timesheets, the providers are paid institutional placement and other services in the with a combination of state, federal, and county absence of IHSS. As we noted in our Analysis of funds. the 2006‑07 Budget Bill, the per-person, per-year cost of IHSS was $10,000 (at the time), while the Why Was iHSS created? per-person, per-year cost of institutional care was The IHSS program was established as a way about $55,000. This simple comparison of the to provide in-home domestic and personal care annual cost of these two types of care is mis- services to recipients who may otherwise be at leading, however, because it does not take into risk of nursing home placement. Another ben- account some key factors we will discuss later efit of the IHSS program is that it makes living at in this report. This report provides an analytical home easier for recipients, and reduces the time framework to consider the likelihood that indi- and financial burdens on their friends and family vidual IHSS recipients will enter a skilled nursing members. The rationale for providing IHSS was facility (SNF) if they lose IHSS, to compare the that in-home care, rather than institutional care, costs of these different types of care, and to ana- would increase the quality of life for program lyze the cost-effectiveness of IHSS with respect recipients and could potentially result in cost to state government. To this end, we created a avoidance for the state. However, as we will dis- fiscal model that compares the cost of IHSS to cuss later in this analysis, quantifying the extent the cost of a world without IHSS. to which IHSS has actually resulted in net state savings is challenging. LegisLative anaLyst’s Office 5 an LaO RepORt Background History of iHSS eligible for federal Medicaid funding. Program costs for these recipients are shared 65 percent Some version of the IHSS program has been by the state and 35 percent by the counties. The present in California since the 1950s. Until 1993, IHSS program administration costs are shared IHSS was funded through a combination of state, 50 percent by the federal government, 35 per- federal Title XX (Social Services Block Grant cent by the state, and 15 percent by the counties. funds), and county funds. In 1993, Chapter 939, The amount of the federal share in the IHSS Statutes of 1992 (AB 1773, Moore), directed the program (50 percent) is determined by a federal Department of Health Care Services (DHCS) to formula known as the federal medical assistance submit a state plan amendment (SPA) to the fed- percentage (FMAP). The American Recovery and eral Centers for Medicaid and Medicare Services Reinvestment Act of 2009 temporarily increased (CMS) to include a portion of the IHSS program the FMAP from 50 percent to 61.6 percent from as a service eligible for federal Medicaid funds October 2008 through December 2010. (known as Medi-Cal in California). As a result of Eligibility. To be eligible for IHSS, a person this SPA, some IHSS recipients became eligible for must be aged, blind, or disabled and usually have federal Medicaid funding through what is referred income at or below the SSI/SSP grant level ($845 to as the Personal Care Services Program (PCSP). per month for individuals as of October 2009). In 2004, DHCS and the Department of Social Ser- Those individuals with income in excess of this vices (DSS) submitted a waiver application to the grant level may still be eligible for IHSS with a CMS to make most of the remaining recipients eli- share of cost (SOC). An IHSS recipient with a gible for federal funding. Approval of this waiver, SOC must make an out-of-pocket monthly pay- now known as the IHSS Plus Option (IPO), was ment towards the receipt of IHSS services before granted in August 2004. the IHSS program pays the remainder of the cost Currently, IHSS consists of three compo- of their services. Eligibility for IHSS is generally nents—PCSP (about 92 percent of the casel- limited to individuals with no more than $2,000 oad), IPO (about 7 percent of the caseload), and in assets and couples with no more than $3,000 Residual (about 1 percent of the caseload). Thus, in assets (with certain exclusions for such assets about 99 percent of the caseload is eligible for as homes and vehicles). federal Medicaid funding. Application and Social Worker Assessment. How iHSS Works When a potential IHSS recipient applies for the pro- gram at a county office, the determination of their IHSS Program Funding. The IHSS program eligibility is a two-step process that takes into ac- is funded through a combination of state, county, count the applicant’s income and need for services. and federal Medicaid funds. For almost all IHSS Once a county worker verifies that an indi- recipients, 50 percent of program costs are paid vidual is financially eligible for IHSS, a county by the federal government, about 32.5 percent social worker visits the home of the recipient to by the state, and 17.5 percent by the county. determine whether there is a need for services. Only about 1 percent of IHSS recipients are not 6 LegisLative anaLyst’s Office an LaO RepORt To perform this assessment, the social worker of 2 on the “feeding” task may be authorized uses a uniform assessment tool to determine the to receive between 0.7 hours and 2.3 hours of number of hours for each type of IHSS service feeding per week. The corresponding range of for which a recipient qualifies in order to remain hours varies depending on the particular task safely in his/her own home. Figure 1 provides a being assessed. For example, meal preparation list of the types of services an IHSS recipient may services range from three to seven hours. Also, if be eligible to receive. an individual is assessed as having an FI ranking The uniform assessment tool, known as the of 1 for any given task, he or she will not receive hourly task guidelines (HTGs), assists the social any authorized hours for that task. The weighted worker in ranking the recipient’s impairment average of the FI rankings for each task is used to level on a five-point scale known as the func- create a total FI score. Although the HTGs pro- tional index (FI) ranking. Figure 2 (see next page) vide a standard tool, the assessment process is shows each of the potential FI rankings that may individualized. Social workers may, with written be assessed by a social worker, and what they justification, authorize hours above or below the mean for the impairment level of the recipient. range established by the HTGs. Each FI ranking corresponds to an established Assignment of Hours. Once a social worker range of service hours for a particular task. For has determined the number of hours to autho- example, a recipient who receives an FI ranking rize for a recipient, the recipient is notified of Figure 1 Examples of Services available to in-Home Supportive Services recipients tasks Examples domestic Services Cleaning, dusting, picking up, changing linens, changing light bulbs, wheelchair maintenance, and taking out garbage. Laundry Sorting, washing, hanging, folding, mending, and ironing. Shopping and Errands Purchasing groceries and putting them away, picking up prescriptions, and buying clothing. meal Preparation Planning menus, preparing food, and setting the table. meal cleanup Washing dishes and putting them away. Feeding Feeding. ambulation Assisting recipient with walking or moving in home or to car. Bathing, oral Hygiene, Bathing recipient, getting in or out of the shower, hair care, shaving, and grooming grooming. routine Bed Baths Sponge bathing the body. dressing Putting on/taking off clothing. medications and assistance Medication administration assistance; taking off/putting on, maintaining, with Prosthetic devices and cleaning prosthetic devices. Bowel and Bladder Bedpan/ bedside commode care, application of diapers, assisting with getting on/off commode or toilet. menstrual care External application of sanitary napkins. transfer Assistance with standing/ sitting. repositioning/ rubbing Skin Circulation promotion and skin care. respiration Assistance with oxygen and oxygen equipment. Protective Supervision Ensuring recipient is not harming themselves. LegisLative anaLyst’s Office 7 an LaO RepORt the number of hours they have been authorized regardless of their training, the type of services for each task. Using the HTGs, social workers they are providing, and their qualifications. may authorize between 1 and 283 total hours the iHSS Program is growing per month of IHSS services. Currently, recipients receive an average of about 85 hours of IHSS per The IHSS program is the fastest growing month. Recipients who receive over 195 hours of major social services program in California. service each month are considered to be severe- Between 1998-99 and 2008-09, IHSS General ly impaired. Fund expenditures grew at an average annual All Eligible Recipients Receive Services. rate of about 13 percent. This growth was due Once it has been determined that a recipient to the combined effect of an increase in the cost meets the eligibility criteria for IHSS, that indi- per case and an increase in the IHSS caseload. vidual is granted those IHSS services. As a result, In comparison, statewide General Fund spending there is no waiting list or cap on program enroll- increased by 4.8 percent annually over the same ment. time period. State Participation in Wages. County wages Growth in the IHSS Population. Figure 3 and benefits to IHSS workers range from $8.00 shows that the IHSS caseload has grown from per hour to $14.99 per hour. Currently, the state 208,400 in 1998-99 to about 430,000 in has a share in the cost of IHSS wages up to $9.50 2008-09. While the IHSS caseload has grown by per hour and, for benefits, up to $0.60 per hour. about 105 percent over this time period, the total Counties with wages and benefits above $10.10 population in California has only increased by split the additional cost with the federal govern- about 16 percent. ment. Prior to the 2009-10 February budget, Growth in the Cost Per Case. As shown in state participation in IHSS provider wages and Figure 4, in addition to the growth in the casel- benefits was $12.10 per hour. Although the state oad, the IHSS annual cost per case has increased participation in wages has recently been lowered from about $6,300 per case from all fund sourc- to $10.10 per hour, a federal judge issued an es in 1998-99 ($2,400 from the General Fund) injunction to stop the decrease in state participa- to about $13,000 per case in 2008-09 ($4,200 tion. As a result, despite current law, the state is from the General Fund). The increase in the cost still participating in com- bined wages and benefits Figure 2 of up to $12.10 per hour. Functional index rating Scale As noted above, Functional IHSS wages and benefit index impairment implications levels vary across coun- 1 Able to perform function without human assistance—independent. ties. Within a particular 2 Able to perform a function, but needs verbal assistance (reminding, encouraging). county, however, IHSS 3 Able to perform a function with some human, physical workers are paid—with assistance. very few exceptions— 4 Able to perform a function with substantial human assistance. 5 Cannot perform the function with or without human assistance. the same hourly wage 8 LegisLative anaLyst’s Office an LaO RepORt per case is primarily due Figure 3 to increasing wages for In-Home Supportive Services Caseload IHSS providers and an Continues to Grow increase in the average number of authorized 450,000 IHSS hours per case. 400,000 Figure 4 illustrates that the growth of the 350,000 General Fund portion of 300,000 the IHSS cost per case slowed from 2003-04 250,000 through 2004-05. This 200,000 was due primarily to the 150,000 previously mentioned 2004 federal waiver ap- 100,000 proval that made most 50,000 individuals in the Re- sidual program eligible 1998-99 2000-01 2002-03 2004-05 2006-07 2008-09 for federal Medicaid funds, thereby offset- ting some General Fund Figure 4 costs. Since 2005-06, the In-Home Supportive Services Cost Per Case average rate of General Also Growing Fund growth in the cost per case has been about $14,000 4.1 percent. Other Funds 12,000 a closer Look at General Fund iHSS recipients 10,000 The IHSS caseload is 8,000 very diverse. There are some very frail recipients 6,000 who receive the maxi- mum amount of hours, 4,000 and some recipients with less severe disabilities 2,000 and fewer hours. As a result, some IHSS recipi- 1998-99 2000-01 2002-03 2004-05 2006-07 2008-09 ents rely more heavily on LegisLative anaLyst’s Office 9 an LaO RepORt IHSS services than other recipients. Below, we each month (see Figure 7). Moreover, nearly highlight some of the characteristics that contrib- 60 percent of IHSS recipients receive less than ute to the diversity of the IHSS caseload. 80 hours of care each month (about 18 hours of Recipient and Provider Relationship. Cur- care per week). rently, there are about 376,000 IHSS individual Some IHSS Recipients are Developmentally providers statewide. As shown in Figure 5, Disabled. About 35,000 (nearly 9 percent) of almost two-thirds of IHSS recipients receive care IHSS recipients are developmentally disabled. A from a provider who is related to them. More- developmental disability is defined as a disabil- over, about 46 percent of IHSS recipients receive ity attributable to mental or physical impairment care from either their own parent, spouse, or that originates before an individual is 18 years adult child (defined as a “close relative” for pur- old, and that is expected to continue indefi- poses of this report). In about half of cases, IHSS nitely. Developmental disabilities include, but providers live in the same home as the IHSS are not limited to, mental retardation, cerebral recipient. palsy, epilepsy, autism, and other disabling con- Age of IHSS Recipients. Although the IHSS ditions related to mental retardation. Within the program serves a wide age range of recipients, IHSS population, developmental disabilities are as shown in Figure 6, the majority of recipients most common among the young recipients with are elderly. While almost 60 percent of IHSS high hours. recipients are over the age of 65, only about Figure 5 5 percent of the total Most Providers and Recipients Are Relatives IHSS population is under the age of 18. Authorized Service Hours. Within the IHSS Relatives Provider Not Related caseload, some recipi- ents have high impair- Close ment levels and a high number of authorized hours, and others have low impairment levels and a low number of hours. Although recipi- ents may receive up to 283 hours, only a small percentage of recipients Other actually receive more than 200 hours of care 10 LegisLative anaLyst’s Office an LaO RepORt Community services Figure 6 for the developmentally Most of the In-Home Supportive Services disabled population are Population Is Elderly overseen by the Depart- (Age of Recipients) ment of Developmental Services (DDS). These 40% services are provided 35 through 21 regional centers (RCs) located 30 throughout the state. 25 The RCs are responsible for eligibility determina- 20 tions and client assess- ments, the develop- 15 ment of an individual 10 program plan, and case management. In gen- 5 eral, RCs only pay for services after individuals 0 to 6 7 to 18 19 to 44 45 to 64 65 to 79 80+ have maximized their access to so-called Figure 7 “generic” services (those services provided at the Number of In-Home Supportive Services Hours Varies local level by counties, (Percentage of Recipients According to Monthly Authorized Hours) cities, school districts, and other agencies), 30% such as IHSS. Diverse Caseload 25 Results in Diverse Ser- vices. As demonstrated 20 by the above data, the IHSS caseload is not 15 uniform. The program serves over 430,000 10 different recipients with varying service autho- 5 rizations and needs. Some recipients rely 0-25 26-55 56-79 80-119 120-159 160-199 200-283 more heavily on the program than others. LegisLative anaLyst’s Office 11 an LaO RepORt coSt-drivErS For iHSS and rELatEd ProgramS Is IHSS Cost-Effective to State Government? tional care is higher than the average annual As we discuss below, a simple comparison of the cost per person of community-based programs. lower yearly costs of IHSS to the higher yearly Specifically, in 2007-08, the average yearly cost cost of a SNF does not take into account other of a SNF was over $50,000 per person while factors which may contribute to the overall cost the average cost of IHSS was about $12,000 per of IHSS. Below we discuss how the costs of these person. However, these comparisons in many two types of long-term care compare. cases understate the true cost of keeping some- one in the community. Some IHSS recipients, How the costs of iHSS and for example, may be receiving multiple services, institutional care compare such as a home-delivered meals or case manage- California currently administers a continuum ment services, that add to the overall cost of their of long-term care programs available to seniors community-based care. and those with disabilities. As shown in Figure 8, Differences in Average Time in Care. Anoth- the IHSS program is part of California’s long-term er key factor to consider in comparing the costs care continuum. Most long-term care programs of institutional care and community-based care may be classified as either community-based or is the time an individual spends in each type of institutional programs. care. On average, individuals in institutional care Differences in Average Yearly Cost of Care. settings are there for shorter periods of time. As The average annual cost per person of institu- seen in Figure 9, for all age groups, the average time in care is longer on IHSS than in a SNF. As a result, the length of time recipients spend in Figure 8 IHSS should be considered when comparing the Program costs vary in california’s total cost of providing care for a recipient in the Long-term care continuum community rather than in a SNF. 2007‑08 Total Costs average annual other Factors affecting relative costs Program cost Per Participant Not All IHSS Recipients Would Otherwise community-Based care Enter a SNF. In the absence of IHSS, what would Linkagesa $2,012 Multipurpose Senior 3,454 happen to program recipients? Some clearly Services Program would be placed in a SNF. However, our analysis Alzheimer’s Day Care 5,043 Resource Centera indicates that, given the diversity in the caseload, Adult Day Health Care 10,482 it is likely that a significant portion of recipients In-Home Supportive 12,287 would not otherwise require institutional care. Services institutional care Although some IHSS recipients currently meet Skilled Nursing Facilities $51,100 the eligibility criteria for SNF placement, and a Funding for these programs was eliminated as of October 2009. are therefore deemed SNF-certified, meeting 12 LegisLative anaLyst’s Office an LaO RepORt this threshold is not an eligibility requirement IHSS because some recipients first enter a hospi- for IHSS program participation. Moreover, no tal, or some other facility, before entering a SNF.) current data is available on the percentage of Recipients May Enroll in Multiple Programs. IHSS recipients who meet the eligibility criteria Recipients who receive IHSS are usually receiving for SNF placement. Notably, exit data indicate assistance through other state programs, such as a that each year about 5,500 IHSS recipients (or cash grant through the SSI/SSP Program. Compar- about 1.3 percent of the total number of IHSS ing only the costs of IHSS and nursing home care recipients) enter SNFs directly from IHSS. (This understates to some degree the cost of providing number, however, understates the actual number care for the individual in the community. of recipients who eventually end up in a SNF after PErSPEctivES on tHE coSt-EFFEctivEnESS oF iHSS Whether IHSS results in overall state sav- earlier, some program recipients would likely ings on long-term care is difficult to determine. end up in institutional care, while others Such an analysis requires a combination of data would not. However, it is difficult to estimate collection, development of key assumptions, the number of recipients that would be most and fiscal modeling. Below, we provide an likely to enter a SNF if the IHSS program did analytical framework for considering whether Figure 9 the operation of the Average Time in In-Home Supportive Services Much IHSS program results in Longer Than for Skilled Nursing Facilities a net fiscal benefit for the state. We describe Years of Service In-Home Supportive Services 5.0 the purpose of our Skilled Nursing Facility 4.5 fiscal model, our key 4.0 data sources, our main assumptions, and our 3.5 methodology. (We also 3.0 discuss research related 2.5 to this subject in the 2.0 box on the next page.) 1.5 1.0 Purpose of 0.5 the model Developing a Fiscal 0 to 6 7 to 18 19 to 44 45 to 64 65 to 79a 80+a (65 to 84 (85+ SNF) Model. In the absence Age Group SNF) of IHSS, as we noted aThe average length of stay data is stratified slightly differently for IHSS and skilled nursing facilities. LegisLative anaLyst’s Office 13 an LaO RepORt not exist. This is a key factor in evaluating the of recipients entering a SNF in the absence of cost-effectiveness of the program. IHSS. This cost estimate can then be compared To address this question, we created a model to our forecast of the “business as usual” cost of that, based on certain assumptions, estimates the continuing the IHSS program without changes potential cost of nursing home care in the ab- to determine the net cost-effectiveness of the sence of IHSS. As shown in Figure 10, the model IHSS program. Although this model compares assumes that, over time, all current and future costs based on different assumptions of SNF IHSS recipients would otherwise (1) enter a SNF, entry rates in the absence of IHSS, it does not (2) receive care from friends and family at no predict or estimate the actual SNF entry rate in government cost, or (3) receive (if eligible to do the absence of IHSS. This is because data are not so) an increased level of care from DDS. Since available to predict the behavioral response to the exact percentage of recipients who would the elimination of IHSS. shift to SNF care is unknown, the model allows LAO Hypothesis: Factors Increasing Like- us to test various assumptions as to the percent- lihood of SNF Entry. Given the difficulty in age of recipients who would enter a SNF in the predicting which recipients would be likely to absence of IHSS. The model estimates the Gen- enter a SNF in the absence of IHSS, we created eral Fund cost of the long-term care system for a working hypothesis of which IHSS recipients any given assumption concerning the percentage would be most likely to enter a SNF in the ab- P r c -B c revious esearch on ommunity ased are The research we reviewed as part of our analysis of the cost-effectiveness of community- based care is limited and offers mixed findings. Some studies, such as one performed by researchers at the University of California, San Francisco, consider the per year cost of the community-based program and the SSI/SSP grant and compare those costs to the costs of institutional care programs. On this basis they con- clude that community-based programs cost less per person annually than institutionalization. However, these studies do not estimate the number of recipients who would enter a skilled nursing facility (SNF) in the absence of community-based services, and therefore do not deter- mine whether in-home care results in overall governmental savings. Another study, sponsored by the U.S. Department of Health and Human Services, points out that recipients of home and community-based care services may never have entered a SNF even in the absence of alternatives. Rather than reducing costs, these recipients added to the overall cost of providing long-term care in the community. A common theme among the research is the importance of targeting community-based care to those most likely to enter an institution in order to increase the cost-effectiveness of in-home care programs. Another common theme of the research is that providing care in the community often increases the quality of life for elderly clients. 14 LegisLative anaLyst’s Office an LaO RepORt sence of IHSS. Our hypothesis was based on in- because they die, or because they no longer formation provided by state and county officials; meet eligibility requirements. A DSS computer experiences accompanying social workers on system documents where recipients “went” upon IHSS recipient assessments and reassessments; exiting IHSS. We obtained this IHSS data for a and public testimony at legislative hearings by six-month period. These data were stratified by program recipients, providers, advocates, and age, provider type, and the number of autho- county representatives. Essentially, we concluded rized IHSS hours. We specifically considered the that the recipients most likely to enter a SNF in characteristics of the IHSS recipients who died, the absence of IHSS would be those who were entered a SNF, or were admitted to a hospital (1) older, (2) received high levels of authorized over this six-month period. This allowed us to IHSS care, and (3) received services from a pro- determine which types of IHSS recipients were vider who was not a close relative. most likely to enter a SNF. We concluded that these recipients were the most frail among the key data Sources IHSS caseload. We collected data in a form that allowed SNF Data From the DHCS. The DHCS for the testing of our hypothesis from DSS provided data on the average length of stay and and DHCS. Below, we describe the main data average cost of SNF placements by the age of the sources. recipient. From this data, we were able to better Stratifying the Current IHSS Caseload. To understand differences in the average length of better understand the differences among Figure 10 IHSS recipients and Overview of the LAO Model: Flow of Recipients how they differ in their In the Absence of In-Home Supportive Services use of the program, we obtained data that gave us a “snapshot” of the current com- Enter a skilled nursing facility. position of the IHSS caseload. The casel- oad was stratified by Rely on resources of age, the relationship All current and future family and friends IHSS recipients. of the provider to the (no government cost). recipient, and the number of authorized IHSS hours. Rely on increased developmental Current IHSS “Ex- disability services. its.” Some recipients exit IHSS because they enter an institution, LegisLative anaLyst’s Office 15 an LaO RepORt stay and cost of SNF care based on the age of frail recipients will enter a SNF in the absence the recipient. of IHSS. For example, if the model is testing a scenario where 50 percent of recipients would key Features of the model enter a SNF in the absence of IHSS, those who Below, we describe the assumptions we were frail would enter SNFs at a higher rate than made related to the length of time a recipient 50 percent, while the less frail would enter at a would remain in the SNF, the characteristics of rate lower than 50 percent. the recipients at the highest risk of SNF place- Comparing Costs at Equilibrium. For pur- ment, and the costs and growth over time. poses of this analysis, we focused on the cost-ef- Time in Care. This analysis assumes that fectiveness of IHSS in the longer term—once the those who would enter a nursing home in the model has reached equilibrium in about seven absence of IHSS will stay there for the current years. Under the model, the costs of SNF place- average length of stay in a SNF. In other words, ment for the current IHSS caseload would be if the model is testing the cost-effectiveness high in the early years. However, for purposes of of IHSS based on a scenario that assumes that this analysis, we believe it is more appropriate to 50 percent of IHSS recipients would enter a SNF compare the long-term costs of a world without in the absence of IHSS, all 50 percent that enter IHSS to the long-term costs of IHSS as it operates a SNF are assumed to have stayed for the aver- today. age length of stay for each age group. Base-Year Costs and Growth. In general, Relative Propensity to Enter a SNF. The exit the costs in our model for IHSS and SNF care data was used to determine the characteristics of are based on 2007-08 General Fund and county IHSS recipients who were most likely to enter a expenditures. In the out-years of the analysis, the SNF relative to other recipients. Figure 11 shows costs were kept constant and only adjusted for the relative propensity of an IHSS recipient to en- caseload growth. ter a SNF based on two factors: age and num- Figure 11 ber of IHSS authorized in-Home Supportive Services recipients hours. The darkest shad- most at risk of SnFa Placement ing in Figure 11 indicates (By Age and IHSS Monthly Hours) that older recipients with Hours of care a high number of autho- age 0-79 80-200 200+ rized hours are the most risk of likely to enter a SNF. 0 to 6 SnF Entry 7 to 18 Lowest The exit data was 19 to 44 used to inform and build 45 to 64 the model. Under any 65 to 79 Highest 80+ scenario, the model a Skilled Nursing Facility. assumes that a higher percentage of the more 16 LegisLative anaLyst’s Office an LaO RepORt additional costs and Savings SSI/SSP Savings. About 86 percent of IHSS included in the model recipients receive a monthly SSI/SSP grant pay- ment. When IHSS recipients enter institutional Treatment of the Developmentally Disabled care, their SSI/SSP grant is significantly reduced. Population. Our model assumes that, in the The model accounts for these savings in SSI/SSP. absence of the availability of IHSS services, all IHSS recipients who are developmentally dis- Some Factors not accounted abled would incur increased DDS costs, rather For in the model than enter a SNF. We assumed that each hour As previously noted, because this model is of lost IHSS services would be replaced with an a simplification of the real world, it is based on hour of services purchased through the RCs at an many assumptions. As a result, it is important to increased cost to the state. acknowledge several factors that we were unable Costs for Accidents. We recognize that, in to build into our analysis. These include such the absence of IHSS, recipients may not imme- factors as the impact of the program on the qual- diately enter a SNF. Some recipients may instead ity of life of recipients. For more details about have an accident or other episode that eventually the factors that were excluded from our analysis, results in their placement in a SNF. As a result, please see the box on the next page. potential costs for accident-related injuries for some recipients are factored into the analysis. ScEnario anaLySiS As explained above, to test the cost-effective- Below, we present the results of our analysis ness of IHSS, we used our model to estimate the under different potential scenarios. cost of nursing home placement versus the cost results of the modeling of IHSS under various scenarios. Figure 12 (see page 19) shows how IHSS cases (including cases Whether IHSS is cost-effective in the ag- of recipients with and without developmental gregate depends on the percentage of recipients disabilities) flow through the model. As explained who would likely enter a SNF in the absence of earlier, under our model, all IHSS recipients who IHSS. Because the state and counties both have are developmentally disabled (about 9 percent of a share in the cost of the program, we looked the total IHSS population) would receive in- at both cost-effectiveness from the perspectives creased developmental services, rather than enter of (1) the state and county funds combined and institutional care. The percentage of IHSS recipi- (2) the state General Fund alone. Below, we pres- ents who do not have developmental disabilities ent the results of our model from both perspec- and who enter a SNF depends on the particular tives under several different scenarios. scenario. For example, Figure 12 shows a scenar- Scenario Analysis From the Combined State io where it is assumed that 50 percent of the non- and County Perspective. After accounting for developmentally disabled IHSS recipients would both costs and savings to the state and counties, eventually enter a SNF if IHSS did not exist. our model showed that the break-even point for LegisLative anaLyst’s Office 17 an LaO RepORt L o m imitations of ur odeL The model discussed in this report has limitations and it does not account for all of the po- tential costs and benefits of In-Home Supportive Services (IHSS). Below, we provide examples of some of the factors that were not accounted for in our model. ➢ Quality of Life for Recipients. Although we recognize that one of the primary benefits of the IHSS program is that it may increase the quality of life for program recipients and their families, we are unable to quantify this benefit in the model. As a result, this analy- sis is only focused on the fiscal costs and benefits of IHSS. ➢ Potential Costs in Other Programs for IHSS Recipients. Some IHSS recipients receive other services in the community that add to the cost of their care, such as housing ben- efits, home-delivered meals, and case-management services. Due to data limitations, we are not able to quantify the extent to which these other programs are utilized by IHSS recipients and the related costs. ➢ Cost-Effectiveness of IHSS for the Federal Government. Our analysis measures the cost- effectiveness of IHSS from the standpoint of the state. It does not measure whether the IHSS program is cost-effective for the federal government or society as a whole. ➢ A Change in the Average Length of Stay in a Skilled Nursing Facility (SNF). This analy- sis assumes that recipients would have the same average length of stay in a SNF in the absence of IHSS as they have in the current long-term care system where IHSS exists. To the extent that the existence of IHSS has an impact on the average length of stay in a SNF, the impact is not recognized in this analysis. Nevertheless, we used our model to test the effects of a 10 percent increase in the average length of stay. This change had a minimal impact on the results and did not impact our final conclusion. ➢ The Impact of the American Recovery and Reinvestment Act (ARRA) on the Federal Medical Assistance Percentage. As noted earlier, the federal share in the cost of IHSS and SNFs has been temporarily increased pursuant to the federal ARRA. This report does not account for this temporary reduction in the state’s cost in these programs. This is be- cause this analysis considers the long-term cost-effectiveness of IHSS, while the General Fund relief from ARRA is temporary. ➢ The Most Recent IHSS Program Reductions and Proposals. In recent months, the Legis- lature has taken action to reduce the costs in the IHSS program (we provide more infor- mation on the IHSS reductions included in the 2009‑10 Budget Act later in this report). In general, the actions aim to maintain IHSS for the most impaired recipients, and eliminate, or reduce, services for the least disabled. The potential savings associated with these proposals are not included in this analysis. 18 LegisLative anaLyst’s Office an LaO RepORt IHSS cost-effectiveness occurs when 58 percent IHSS, they do not have a share of cost in SNFs. of the non-developmentally disabled popula- In other words, if less than 58 percent of IHSS re- tion enters a SNF in the absence of IHSS. In this cipients enter a SNF in the absence of IHSS, the “break-even” scenario each dollar of combined program would not result in a net fiscal benefit state General Fund and county funds invested in to the state and counties combined. IHSS saves one dollar of state General Fund SNF In addition to the 58 percent break-even costs. There are no corresponding county savings scenario, Figure 13 shows the fiscal effects because, while counties have a share of cost in for nursing home entry rates under alternative scenarios—in which the Figure 12 SNF entry rate for IHSS Illustration of Scenario Where 50 Percent of IHSS recipients who are not Recipients Eventually Enter a SNF developmentally dis- abled is either 38 percent or 78 percent (20 per- Enter skilled nursing centage points above facility (SNF). All current and future (50%) and below the estimated IHSS recipients break-even point). These without developmental disabilities. percentages do not (91% of total) Resources of family include the developmen- and friends (no government cost). tally disabled recipients, (50%) who would receive increased developmental disability benefits rather All current and future Increased IHSS recipients with than enter a SNF. The fig- Developmental developmental Services ure also summarizes the disabilities. (100%) (9% of total) net fiscal impact under each scenario in year seven. Scenario Analysis From the Perspective of Figure 13 the State Alone. When Summary of model results for different Scenarios only considering the (General Fund and County Funds, in Millions) state General Fund costs Percent Shifted to SnFa of IHSS, the break-even 38% 58% 78% point for cost-effective- Total SNF and developmental disability costs $2,880 $3,822 $4,787 ness is significantly lower with no IHSS. than for the state and Baseline costs of IHSS. 3,822 3,822 3,822 county combined. Our net cost (-)/Savings (+) -$924 — $965 a Percent of non-developmentally disabled shifted to skilled nursing facility (SNF). model shows that for LegisLative anaLyst’s Office 19 an LaO RepORt the state General Fund, the break-even point for The Implications of a Cost-Effective Sce- IHSS cost-effectiveness occurs when 32 percent nario for the State and Counties. Under a state/ of the non-developmentally disabled population county break-even scenario, 58 percent of the enters a SNF in the absence of IHSS. total caseload that is not developmentally dis- Figure 14 is similar to Figure, 13 except it abled would enter a SNF facility in the absence shows the fiscal effects for nursing home entry of IHSS. rates from the perspective of the state General A scenario with overall SNF entry rates at Fund alone under alternative scenarios—in 58 percent means different entry rates for the which the SNF entry rate for IHSS recipients who various subgroups within the IHSS caseload. This are not developmentally disabled is 16 percent, is because the exit data we collected indicates 32 percent, or 48 percent. (A 50 percent increase that certain recipients are more likely to enter a and decrease compared to the estimated break- SNF than others. For example, this break-even even point.) scenario would mean that 100 percent of the non-developmentally disabled recipients over the What do the Break-Even Points imply age of 65 with over 200 hours of IHSS services With respect to SnF Entry rates? would enter a SNF in the absence of the pro- Because there is no way to know how gram. Additionally, it means that 12 percent of individuals and state programs would actually recipients under age six with under 80 hours behave in a world without IHSS, it is difficult to of authorized services would enter a SNF in determine whether IHSS is cost-effective for state the absence of IHSS. Moreover, this means that government. The key question presented by our 81 percent of all recipients over the age of 80, model is whether the IHSS program prevents at and 22 percent of those under the age of six least 58 percent (or 32 percent from the perspec- would enter a SNF in the absence of IHSS. Lastly, tive of the state alone) of the recipients who are if 58 percent of recipients enter a SNF in the not developmentally disabled from entering a absence of IHSS, our model estimates that about SNF, on average. 43 percent (91,000) of recipients with under To evaluate this question, we compared the 2.5 hours of IHSS service per day (80 hours per percentage of certain subgroups that shifted from month) would enter a SNF. IHSS to a SNF under our model to what we as- Figure 14 sumed would shift under Summary of model results for different Scenarios our working hypothesis. (General Fund, in Millions) In other words, we sub- jected our model to a sort Percent Shifted to SnFa 16% 32% 48% of “reality check” to see if it was reasonable. Below, Total SNF and developmental disability costs $1,184 $2573 $3,356 with no IHSS. we describe the aspects Baseline costs of IHSS. 2,573 2,573 2,573 of our model that helped net cost (-)/Savings (+) -$732 — $783 to inform this perspective. a Percent of non-developmentally disabled shifted to skilled nursing facility. 20 LegisLative anaLyst’s Office an LaO RepORt The Implications of a Cost-Effective Scenar- 49,000) of recipients with less than 2.5 hours of io for the State Alone. Under a state-only break- service each day would enter a SNF. even scenario, 32 percent of the total caseload Model Results Are Generally Consistent that is not developmentally disabled would enter With Our Hypothesis. The model results are a SNF. generally consistent with our initial hypothesis. In This break-even scenario means that 63 per- other words, under all scenarios, older recipients cent of the non-developmentally disabled recipi- with a high number of hours have higher SNF ents over the age of 80 with over 200 hours of entry rates than younger recipients with fewer IHSS services would enter a SNF in the absence hours. Although generally consistent with our hy- of IHSS. Looking at the other end of the age and pothesis, some elements of the results are some- hour spectrum, 6 percent of recipients under the what different than we would have expected. For age of six with fewer than 80 hours of services example, we expected that recipients with rela- would enter a SNF in the absence of IHSS. In this tive providers would enter SNF at a lower rate scenario, 44 percent of recipients over the age of than recipients with other providers. However, 80, and 12 percent of recipients under the age of the results did not seem to demonstrate a sig- six would enter a SNF. Finally, if 32 percent of re- nificant difference between the SNF entry rates cipients enter a SNF in the absence of IHSS, our of those with close relative providers and those model estimates that about one-quarter (about with other providers. FindingS Whether IHSS results in a net fiscal ben- However, iHSS may Well Be cost- efit or is cost-effective to the state depends on Effective for the State general Fund whether the costs and benefits are counted When only considering the effect on the from the perspective of the state and counties state General Fund, it is very possible that IHSS is combined or for the state General Fund alone. cost-effective in the aggregate. Our reality check Below, we describe our findings related to the of the model results confirms that it is reasonable cost-effectiveness of IHSS from these two differ- to believe that, in the absence of IHSS, 32 per- ent perspectives. cent or more of non-developmentally disabled IHSS recipients would enter a SNF. This find- iHSS is Probably not cost-Effective ing is based on our best judgment and program For State and counties combined knowledge of IHSS. We recognize that others When considering state and county costs could reasonably arrive at a different conclusion combined, we find that IHSS is probably not regarding the cost-effectiveness. cost-effective in the aggregate. This is because, based on our reality check of the model results, iHSS may not Be cost-Effective we find it unlikely that in the absence of IHSS, For all recipients 58 percent or more of the non-developmentally The IHSS program serves a diverse set of disabled IHSS recipients would enter a SNF. recipients with different needs. For those at the LegisLative anaLyst’s Office 21 an LaO RepORt greatest risk of institutional placement, IHSS is other words, even if a developmentally disabled usually cost-effective. For some current IHSS recipient is not likely to enter a SNF absent the recipients, however, the evidence suggests that IHSS program, it is more cost-effective to provide the provision of IHSS services is not likely to be that recipient with IHSS services rather than an the difference between living in the community increase to the level of developmental disability or living in a SNF. These are the recipients with services that they were already recieving. a low likelihood of entering a SNF even in the cost-Effectiveness is not the absence of IHSS. Providing IHSS services to Sole Purpose of iHSS those recipients is not a cost-effective practice, from the state’s perspective. Each dollar invested As we have noted throughout this report, in IHSS for recipients who would never have en- the existence of IHSS serves several purposes. In tered a SNF in the absence of the program adds some cases it may delay or prevent SNF place- to the overall cost of long-term care, without ments, and in other cases it may make the lives avoiding costs for SNF care. Although the provi- of program recipients easier and reduce the sion of these services may increase the quality of caretaking responsibilities of friends and rela- life for those recipients, it adds to the overall cost tives. As a result, whether IHSS is cost-effective of providing care in the community. to state government should not be the sole basis The exception to this is the developmentally for evaluating the merits of the program. Instead, disabled caseload. Although developmentally the Legislature should consider both the cost- disabled recipients may not enter a SNF as a avoidance potential of IHSS and the enhanced result of the loss of IHSS services, the increased quality of life for all recipients, including those cost of the developmental disability services they who may be at minimal risk of SNF placement in would instead obtain from RCs would exceed the the absence of the program. state’s savings from the elimination of IHSS. In PoLicy imPLicationS: targEting incrEaSES coSt-EFFEctivEnESS In examining the issue of the IHSS program’s ductions included in the 2009-10 budget (dis- cost-effectiveness, we have dealt with spending cussed in the nearby box) move toward a more in the aggregate. Our findings also indicate that, targeted approach to providing IHSS. regardless of the program’s cost-effectiveness, on Below, we discuss further approaches which average, the state can take incremental steps to target IHSS services to those who are most increase the program’s relative cost-effectiveness. likely to enter a SNF in the absence of IHSS. We For instance, targeting any service reductions to note that these targeting strategies are subject those recipients with the lowest chance of enter- to federal approval and may require federal law ing a SNF (younger recipients and those with changes and/or additional waivers. Additionally, fewer hours of care) would increase the overall some proposed changes could create an incen- cost-effectiveness of IHSS. To this end, the re- tive for recipients to request reassessments and 22 LegisLative anaLyst’s Office an LaO RepORt file appeals, potentially offsetting some of their only a 3 to receive authorization for bathing as- fiscal benefit. Lastly, we note that because IHSS sistance. By increasing the minimum ranking for may be enhancing the quality of life and alleviat- qualification, services would be targeted to those ing caretaking responsibilities for recipients and with the highest impairment levels for those families, the loss of services for those recipients particular tasks. We note that the use of FI rank- who do not enter a SNF may result in additional ings and scores as a method for targeting IHSS hardships for families. services is currently being challenged in court. However, there are other ways to target services, increase minimum threshold for such as making certain services only available Qualifying for iHSS Services to recipients who are authorized to receive a One option the Legislature could consider to certain minimum number of hours, or requiring better target the program would be to raise the SNF certification to qualify for particular services. minimum threshold for qualifying for IHSS ser- a new Service delivery approach vices. For example, each task could have a differ- ent minimum FI ranking that would be required The Legislature could alter the delivery of to receive the particular task. Recipients could IHSS services based on the level of recipient be required to have a FI ranking of at least 4 to need. As noted earlier, the IHSS program consists receive authorization for dressing assistance, but of a diverse caseload with varying levels of need. r B c a c a o m ecent udget and ourt ctions ouLd ffect ur odeL The 2009‑10 Budget Act includes several significant changes to In-Home Supportive Ser- vices (IHSS). Specifically, the budget eliminates domestic and related services for recipients with functional index (FI) rankings of less than 4 for domestic and related tasks. Additionally, the budget eliminates all IHSS services for recipients with FI scores (the average of all of the individ- ual FI ranks) of less than 2. We note that for both of these FI reductions, the Legislature included exceptions for individuals who may have had low ranks or scores but had overall hours of over 120 per month or who received certain services. Finally, the enacted budget includes several anti-fraud activities that are estimated to result in significant savings. These changes have not been incorporated into the scenario analysis results shown in Figures 13 and 14. This is because the caseload dynamic and cost data are not available at this time to update this analysis to reflect these recent policy changes. Additionally, at the time of this analysis, a federal judge had issued injunctions that have prevented the state from implementing both of the FI reductions. Although there is significant uncertainty about the impacts of these proposed changes to IHSS, we believe that the changes would make IHSS more cost-effective. The Governor’s 2010-11 budget includes a proposal to eliminate all IHSS services for re- cipients with FI scores of less than 4. This would discontinue IHSS for about 87 percent of IHSS recipients and is estimated to result in state savings of $650 million General Fund. LegisLative anaLyst’s Office 23 an LaO RepORt As a result, it may make sense to construct a pro- who would monitor the condition of the recipi- gram that recognizes the variation in the needs of ent and provide case-management services. The different segments of the caseload, and provides intent of monitoring this group would be to trans- different types of services accordingly to each fer them to tier one or two if their condition met segment. For example, under this approach, the qualification levels of those tiers. the more frail recipients could continue receiv- Although our example bases the tiers on ing IHSS, while other recipients could instead the number of authorized hours, the Legislature receive funds to purchase goods and services to could consider basing the tiers on other factors assist them in remaining in their own homes. related to the level of recipient need—such as As shown in Figure 15, a new service deliv- the recipient’s FI score or whether they receive ery approach to IHSS could establish different developmental disability services or are certified tiers of available services based on the recipi- for SNF placement. Additionally, for purposes of ent’s overall level of need. Recipients in the first illustration, our example includes three tiers. The tier would be the most frail recipients (in this Legislature could consider creating additional example, the severely impaired with over 195 service tiers. authorized hours per month) for whom services Such a major reform to the IHSS program would remain unchanged. The middle tier of which would require legislative input on many recipients could receive a variation on cash details, and likely require significant changes and counseling services, with authorization to to the current Medicaid State Plan and waiver spend a set allocation of funds on the purchase agreements with the federal government. How- of goods and services, such as in-home care or ever, we believe that a tiered approach to service home modifications, such as a wheelchair ramp, delivery would effectively target resources to to assist them in remaining in their own home. those with the highest risk of SNF placement. (The text box provides more information about the cash and counsel- ing approach.) Tier two Figure 15 recipients would also Example of a new tiered approach to receive increased case delivering in-Home Supportive Services management and other affected recipients iHSS Policy change assistance from their Severely-impaired Tier One—Recipients No change. IHSS social worker. The recipients (195+ with at least 195 hours of care per authorized hours per third tier of recipients month). month. would receive no IHSS Non-severely impaired Tier Two—Recipients Provide (1) cash for pur- services and no alloca- recipients (less than with between 80 and chase of goods and ser- 195 hours of care per 194 authorized hours vices and (2) increased tion of funds. Instead, month). per month. case management. these recipients would Tier Three—Recipients No IHSS or cash assis- with between 1 and tance. Increased social receive quarterly visits 79 authorized hours worker case management. from a social worker per month. 24 LegisLative anaLyst’s Office an LaO RepORt Potential Savings range from the low millions of dollars to about $400 million annually. However, our analysis These types of targeting strategies would indicates that there is a limit to this approach. If likely result in long-term savings to the state. The these reductions are taken so far as to shift a sig- options discussed above may be implemented nificant number of recipients from IHSS to SNFs, individually or in combination. The amount the potential savings could be more than offset of savings that could be achieved under each by larger SNF costs. policy, or combination of policies, would vary depending upon implementation, but could concLuSion The IHSS program is the fastest-growing General Fund in the aggregate, there are some major social services program. Evaluating the IHSS recipients for whom the investment in IHSS cost-effectiveness of IHSS is a complicated task may have no effect on the state’s SNF costs. In that raises many issues. As we have pointed out this report, we have identified some ways that in this report, whether IHSS may result in a net the Legislature could target IHSS services to fiscal benefit to the state depends on which cost those most likely to enter a SNF in the absence and benefit perspective is being considered—the of the program. Given the state’s severe fiscal dif- state and counties combined or the state General ficulties, we recommend that the Legislature con- Fund alone. Additionally, there are other non- sider targeting IHSS program services to those fiscal benefits that are not captured in this fiscal with the highest risk of SNF placement in order cost-effectiveness analysis. to achieve significant additional state savings. Although our model demonstrates that the program may well be cost-effective to the state c c o s s ash and ounseLing an Ption in ome tates Some states have implemented a cash and counseling model of self-directed, in-home care. The cash and counseling program provides recipients with a monthly sum of available funds, based on the cost of the hours of in-home services that they would otherwise have been authorized to receive. Recipients have more flexibility in the use of these funds than they would in a program like In-Home Supportive Services (IHSS). They can use these monthly sums to set wage levels and hire a provider, install wheelchair ramps, and purchase goods that make it easier to remain at home—expenditures not permitted now under IHSS. Cash and counseling recipients work directly with county officials to craft spending plans, write checks, and handle payroll taxes. LegisLative anaLyst’s Office 25 an LaO RepORt 26 LegisLative anaLyst’s Office an LaO RepORt LegisLative anaLyst’s Office 27 an LaO RepORt LAO Publications This report was prepared by Ginni Bella Navarre, and reviewed by Todd Bland. The Legislative Analyst’s Office (LAO) is a nonpartisan office which provides fiscal and policy information and advice to the Legislature. To request publications call (916) 445-4656. This report and others, as well as an E-mail subscription service, are available on the LAO’s Internet site at www.lao.ca.gov. The LAO is located at 925 L Street, Suite 1000, Sacramento, CA 95814. 28 LegisLative anaLyst’s Office