Participant-Directed Services in Managed Long-Term Services and Supports Programs: A Five State Comparison

As of the beginning of 2013, 16 states had Medicaid managed long-term services and supports (MLTSS) plans available, with enrollment being either mandatory or voluntary for target populations that varied by state (e.g., elderly, younger adults with adult-onset disabilities, persons with intellectual or other developmental disabilities). In 13 of these states, MLTSS plan members were afforded the choice to “participant-direct” (PD) at least some home and community-based services. Based on five in-depth state case studies, state expectations regarding availability and take-up of PD services in MLTSS varied as did states’ methods of communicating these expectations to managed care organizations. Take up of PD options varied from a low of 1.2% in Arizona to a high of 24% in New Mexico. [37 PDF pages]

"

Acronyms

Executive Summary

Since 2004, the number of state Medicaid programs that have integrated health and long-term services and supports (LTSS) for elders and persons with disabilities through various types of managed care programs grew from eight to 16. During the same time period, the number of individuals receiving managed LTSS (MLTSS) increased from 105,000 to 389,000. Of the current 16 states that have MLTSS programs, 13 offer participant direction.

To gain a more thorough understanding of how MLTSS programs have implemented participant direction, researchers from the National Resource Center for Participant-Directed Services conducted an in-depth examination of participant-directed MLTSS (PD-MLTSS) programs in the following five states: Arizona, Massachusetts, New Mexico, Tennessee, and Texas. This examination revealed wide variation in:

1. State requirements for PD-MLTSS: Two states had specific contract language and two others reported relying on policy and procedure manuals or handbooks for communicating their requirements for PD-MLTSS. The fifth state (Massachusetts) prefers to avoid detailed documents and communicates its expectations orally to the managed care organizations (MCOs). Massachusetts state officials feel this allows for a better exchange of ideas.

2. How PD-MLTSS programs are developed and managed: Most of the MCOs in this study are given broad discretion in developing and managing PD-MLTSS service delivery options. This discretion means that there is variance in the features and flexibility of the PD-MLTSS program and the definition of "participant direction."

3. The numbers of participants enrolled in PD-MLTSS: Only Tennessee has clear enrollment expectations, and a review of the state's enrollment targets illustrates how the program has grown. Without such expectations, even states with a historical commitment to participant direction leave the future growth of PD-MLTSS to the discretion of MCOs (which may or may not be committed to such growth).

4. How PD-MLTSS quality is monitored: While all of the states required formal quality assurance and improvement plans prior to MLTSS implementation, few reported having specific quality performance measures or quality monitoring procedures for PD-MLTSS.

5. The roles and functions of financial management service (FMS) agencies: The FMS agencies across the five states fulfill the basic payment, management, and reporting functions. However, states varied in other types of FMS support (e.g., ongoing information and assistance, assistance with worker recruitment, providing training for participants and/or workers, etc.).

The implications of this five state examination include:

1. States play a major role in how PD-MLTSS is operationalized: How PD-MLTSS is shaped and operated in a given state is determined by either the presence or absence of state policies and procedures that emphasize participant direction in MLTSS.

2. There are examples of how the principles of managed care and participant direction can be integrated: Well-designed PD-MLTSS programs can achieve the common goals of MLTSS and participant direction: (1) the improvement of participant health and well-being; (2) the improvement of service satisfaction; and (3) the reduction of service costs).

3. Training for MCO service coordinators is vital: The low take-up rates for PD-MLTSS across the five states suggest the need for additional professional development training in participant direction. This observation was confirmed by a number of key informants across the five states. Throughout the five states the number and type of MCO staff varied greatly. Too often, MCO staff training was restricted to the mechanics of presenting the participant direction option to the member.

4. How PD-MLTSS is presented to participants is critical: Respondents across a couple of the states remarked that participants are overwhelmed with materials about participant direction and tend to be apprehensive. While each state required MCOs to have person-centered processes in place, it does not appear that all MCO service coordinators receive training on person-centered planning or participant direction that could reduce the participant's sense of feeling overwhelmed. Beyond training of professional staff, the use of peers is an undeveloped resource that could help people become comfortable with PD-LTSS.

5. PD-MLTSS would benefit from clarity in the roles and responsibilities of the different PD-MLTSS supports: In some states, this lack of clarity is due to the lack of specificity in state contracts with MCOs (or absence of policy and procedure manuals) regarding participant direction, participant-directed services, and necessary supportive services such as FMS. This lack of specificity can lead to important support functions not being readily available.

6. FMS is a key PD-MLTSS support element: The five states used the FMS agencies in various ways. In all states, the FMS provided traditional financial management support, but some states had the FMS provide other types of support as well.

7. MCOs would benefit from increased engagement from participants: The idea of involving participants in the design and evaluation of LTSS has been promoted for decades. Each of the states and MCOs in this study described various ways participant involvement is sought (e.g., public forums, town halls, member surveys, and advisory boards). However only New Mexico reported participant input specifically focused on PD-MLTSS.

1. Introduction

Participant-directed long-term services and supports (PD-LTSS) assists people of all ages, across all types of disabilities, to maintain their independence and determine for themselves what mix of services and supports works best for them. Since 2001, changes in federal law, regulation, and policy have promoted the growth of publicly funded PD-LTSS. All states have at least one publicly funded PD-LTSS program offering participants employer authority to select, hire, fire and manage individuals to help them with activities of daily living. Forty-three states have at least one program that allows individuals budget authority to manage not only their worker but also purchase other goods and services to help meet their needs.¹ PD-LTSS programs have demonstrated their effectiveness in reducing participants' unmet personal care needs, improving participant health outcomes, and increasing participant satisfaction when compared to traditional agency-directed service models.² Furthermore, research suggests that PD-LTSS programs can achieve cost savings by avoiding or delaying the need for institutional care.^{3, 4}

Since 2000, more states have begun to explore the integration of health and long-term services and supports (LTSS) for elders and persons with disabilities through various types of managed care programs. The rationale for such an integrated approach is to enable better care coordination, continuity of care when transitioning from acute to LTSS, and increased cost effectiveness of care delivery. Increased state use of managed LTSS (MLTSS) programs has also been suggested as a promising solution for state LTSS rebalancing efforts.⁵ Whatever the reason, from 2004 to 2012, the number of state Medicaid programs with MLTSS grew from eight to 16 and the number of individuals receiving MLTSS increased from 105,000 to 389,000. By 2014, the number of states projected to have MLTSS programs is 26.⁶ Of the current 16 states that have MLTSS programs, 13 offer participant direction.⁷

Even before the growth of publically funded participant-directed MLTSS (PD-MLTSS), the compatibility of the concepts of managed care and participant direction has been questioned. Early exploratory studies of managed care organizations (MCOs) found mixed attitudes towards PD-MLTSS. A 2002 study by Meiners and colleagues found that some MCOs saw participant direction as a means to improve service quality and efficiency, as well as increasing member independence. At the same time, Meiners et al. (2002) (and a companion study by Mahoney et al. in 2003) reported that some MCOs expressed concerns on whether participants were up to the task of managing their care.^{8, 9}

An early study remarked that MCOs were driven by characteristics critical for successful PD-MLTSS, not the least of which was the MCOs' focus on participant outcomes.¹⁰ An accompanying study of aging and disability experts expressed concerns that MLTSS would be dominated by a medical model perspective and thereby remove choice and control from participants.¹¹ One early commentator sums up these early studies with the observation that managed care and participant direction may be compatible, but "the devil is in the (program design) details."¹² In 2003, Kodner articulated some of the program design details that could promote a compatible relationship. These program elements included the use of a value-driven assessment process, structured opportunities for participant feedback and contribution to program design, implementation, evaluation, information and assistance services, and member and MCO staff training, among others.¹³

To gain a more thorough understanding of how MLTSS programs have implemented participant direction, researchers from the National Resource Center for Participant-Directed Services (NRCPDS) conducted an in-depth examination of PD-MLTSS programs in five states: Arizona, Massachusetts, New Mexico, Tennessee, and Texas. These states were included in this study for the following reasons:

• Arizona has a mature MLTSS program and is looking to expand participant direction for elders and persons with disabilities by promoting the Agency with Choice model where the participant shares responsibility with an agency for the hiring and management of their worker.

• Massachusetts has three decades of experience operating Medicaid PD-LTSS programs and has operated a managed care program with a participant-directed option since 2004.

• New Mexico is finalizing a comprehensive reform of its MLTSS with an employer authority and a budget authority PD-LTSS option beginning in January 2014. New Mexico's well-established budget authority program, Mi Via, is being incorporated under the new PD-MLTSS program for the Disabled and Elderly, Brain Injury, and Acquired Immunodeficiency Syndrome (AIDS) populations.

• Even though it only recently added MLTSS, Tennessee has approximately 15 years of experience contracting with MCOs to manage its Medicaid services. Tennessee has offered participant direction since 2010 and has the most extensive contract requirements for PD-MLTSS.

• Texas has operated an experienced PD-MLTSS program and intends to expand the participant-directed option statewide.

2. Methodology

In its examination of how MLTSS programs have implemented participant direction in these five states, NRCPDS researchers collected information from both primary and secondary sources. To get a better sense of state expectations and the basic operating parameters of PD-MLTSS, NRCPDS staff reviewed each study state's waiver application and existing contracts with MCOs to provide PD-MLTSS.¹⁴ The contract review informed the final development of the key informant interview guides.

NRCPDS staff conducted interviews in each state with representatives of five key stakeholder groups. The stakeholders included state program staff who oversaw the PD-MLTSS program, administrators and service coordinators from MCOs providing PD-MLTSS in the study state, Financial Management Service (FMS) agency administrators providing FMS services in the study state, and advocacy groups in each state. A complete list of stakeholders that participated in this study is listed in Appendix A, Stakeholder Organizations by State. The key informant interviews took place between May and July 2013. The majority of interviews were conducted via recorded phone calls that included the representative key informant (in some cases, two representatives) and two NRCPDS staff members. One NRCPDS staff member would guide the interview while the other would take notes. The interview notes were compared with the interview recording for accuracy. The written interview notes were then sent to the key informant(s) for review and accuracy.

In addition, NRCPDS staff conducted two in-person site visits in Massachusetts and Texas. Previously (as a part of its internal professional development process) NRCPDS staff had made site visits to Arizona, Texas, and Tennessee to learn more about PD-MLTSS operations in those states. Similar to the telephone interview procedure, two NRCPDS staff members conducted the visits. One NRCPDS staff member would guide the interview while the other would take notes. The written interview notes were then sent to the key informant(s) for review and accuracy.

After interviewing participant advocacy groups in the five study states, the NRCPDS felt that the majority of the interviewees lacked specific knowledge or personal involvement with PD-MLTSS, or in some cases, were direct service providers rather than "advocates." In order to ensure the participant advocacy voice was included in this study, the NRCPDS queried the National Participant Network (NPN), a national advocacy organization for participant-directing individuals, regarding their experiences with PD-MLTSS programs. The questions used in the query of NPN members were taken from the participant advocacy group key informant interview guide. Ninety-three members were contacted but only four (n=4) responded within the time allotted. These four responses are incorporated in the Participant Perspectives on PD-MLTSS section below.

3. Characteristics of Participant-directed Managed Long-term Services and Supports in the Five Study States

After a general overview of the PD-MLTSS program characteristics in the five study states, the following sections compare and contrast states in the following areas: (1) General Program Characteristics; (2) State Requirements for PD-MLTSS; (3) PD-MLTSS Quality Monitoring; (4) PD-MLTSS Program Features; (5) FMS; and (6) Participant Perspectives on PD-MLTSS.

3.1. General Program Characteristics

Four (Arizona, New Mexico, Tennessee, and Texas) of the five study states integrate medical (primary and acute), community-based LTSS, and institutional long-term care (intermediate care and skilled nursing facilities). Three of the states (Arizona, Tennessee, and Texas) also include behavioral health services. As noted in Appendix B, Table B1: State PD-MLTSS Program Overview, all five states offer an employer model and three states (New Mexico, Tennessee, and Texas) also offer a budget authority model. The following describes the states' extant environments and general program design elements:

• The majority of the states use Section 1115 waivers to implement PD-MLTSS: Perhaps because of its flexibility, three states (Arizona, Tennessee, and Texas) currently use Section 1115 to demonstrate innovation while maintaining budget neutrality. While New Mexico currently uses Sections 1915 (b) [Managed Care Waiver] and (c) [Home and Community-Based Services Waiver], they recently received approval to implement PD-MLTSS under Section 1115 and will begin to do so in January 2014. Massachusetts uses Sections 1915 (a) and (c) to offer voluntary participation.

• All five states contract with multiple MCOs: As described in Appendix B, Table B1: State PD-MLTSS Program Overview, each state currently contracts with between two and five MCOs. In four states (Arizona, Massachusetts, New Mexico, and Tennessee), the MCOs are not-for-profit and for-profit organizations. All the current MCOs contracted with Texas are for-profit organizations.

• All five states enroll elders and adults with disabilities in PD-MLTSS while four of the five carve out individuals with intellectual developmental disabilities (IDD): Only one of the state programs (Arizona) includes individuals with IDD in PD-MLTSS. What also makes Arizona unusual is that the state Developmental Disability Operating Agency is the MCO for the IDD population. Arizona's Medicaid program for people with IDD operates under a Section 1115 waiver that enables the Arizona Health Care Cost Containment System (AHCCCS) to be both the single state Medicaid agency and a state-operated, statewide managed care plan. A recent review indicates that this arrangement has worked well for the IDD population in Arizona.¹⁵

• Three of the five states offer both employer and budget authority PD-MLTSS: Arizona and Massachusetts only offer the employer authority model of participant direction while New Mexico, Tennessee, and Texas offer both employer and budget authority. Although Tennessee and Texas meet CMS requirements for offering budget authority, the budget authority model in both states operates under a number of restrictions. In Tennessee, budget authority is listed as an option but is used to a limited degree. In Texas, budget authority is limited to employment supportive items (e.g., fax machines, worker bonuses, health insurance, vacation pay, etc.). Among the five study states, only New Mexico currently offers a budget authority model as it is traditionally defined. In New Mexico, a participant who wants to self-direct their personal care attendant and is eligible under the state's Coordinated Long Term Services program transitions to the Mi Via waiver. Services offered in Mi Via include homemaker, chore, respite, Personal Assistance Services, etc. New Mexico will continue to offer both employer authority and budget authority options under its new MLTSS program, Centennial Care, as the Personal Care Option and Self-Directed Community Benefit, respectively. New Mexico IDD self-directing participants are carved out of Centennial Care and will continue to receive services under Mi Via.

• The five states vary in the types of services that can be self-directed by members: Two states (Massachusetts and Tennessee) define personal attendant services or personal care services as eligible to be directed by members. Texas offers more expanded options to include personal attendant and respite as well as nursing, physical therapy, occupational therapy, and speech or language therapy. Arizona allows members to self-direct certain skilled services. In addition to members directing their personal attendant services or personal care services, New Mexico's budget authority option allows participants to purchase both traditional Medicaid services and supports (e.g., adult day care, supportive employment, etc.) and Participant-Delegated Goods and Services (e.g., transportation, technology, household appliances, etc.).

• Some states have adjusted the Nurse Practice Act (NPA) to accommodate PD-MLTSS: Arizona, Tennessee, and Texas reported changes to the state NPA. Arizona indicated that its board of nursing modified the NPA so that a PD-MLTSS member could hire a non-skilled attendant to perform a limited set of non-invasive tasks. Arizona also indicated that there were no plans to extend this exemption, nor does the exemption apply for persons receiving PD-MLTSS under an Agency with Choice model. Texas reported a similar type of exemption for members in PD-MLTSS that met certain criteria. Tennessee indicated that, with primary care physician (PCP) approval, PD-MLTSS members could delegate medication management. New Mexico respondents conveyed that while New Mexico's NPA was not amended for persons in PD-MLTSS, there was a NPA provision for "certified medication aides" so if a PD-MLTSS member needed medication management assistance, his/her aide could perform this service provided they were certified.

• The numbers of participants who direct their own services vary in each state, however, in the majority of states, the number of PD-MLTSS participants is small and represents a very small percentage of MLTSS program members: Across the five states, the number of members reported enrolled in PD-MLTSS ranged from 300 (Arizona) to approximately 4,600 (Massachusetts). Currently, PD-MLTSS is available statewide in Arizona, Massachusetts, New Mexico, and Tennessee with PD-MLTSS becoming a statewide option in Texas beginning in 2014. The estimated PD-MLTSS take-up rates ranged from a low of 1.2% (Arizona) to a high of 24% (New Mexico). Most states did not have readily available information regarding how participant direction take-up rates in MLTSS compared to fee-for-service programs. This is an area for future research, comparing take-up rates for PD-LTSS before and after the advent of MLTSS. One state that had this information was Texas, where the participant direction take-up rate in its fee-for-service program was 8.2% (approximately 9,200 people out of an approximate 112,000) while it was only 2.5% in its MLTSS program (approximately 3,000 people out of an approximate 122,000). Part of the explanation for this difference is participant-directed is concentrated in the IDD waiver programs and this population is carved out of MLTSS in Texas. Texas expects the numbers of PD-MLTSS to grow as the state expands MLTSS to an additional 164 (mostly rural) counties in 2014.

3.2. State Requirements for Participant-Directed Managed Long-Term Services and Supports

How states communicate their requirements for PD-MLTSS varies. Three states (Arizona, Massachusetts, and Texas) relied less on specific contract language regarding the development and management of PD-MLTSS. Two states (Arizona and Texas) indicated they have developed detailed policy and procedure manuals or handbooks. Those states without specific contract language reported that it was quicker and easier to make changes to manuals or handbooks than it was to contracts. Tennessee and New Mexico indicated having contract language outlining the MCOs responsibilities related to allowing someone to choose PD-MLTSS. Tennessee currently has the most detailed requirements regarding the development and management of PD-MLTSS. The PD-MLTSS section of Tennessee's contract is approximately 35 pages and covers all aspects of the program including policies, procedures, forms, processes, and reporting. The following are common state requirements (see Appendix B, Table B2: State Requirements for PD-MLTSS, for more information):

• Four of the five states require the MCO to offer PD-LTSS to all enrollees: Four of the five states (except Massachusetts) include specific language in their MCO contracts requiring MCOs to offer the PD-MLTSS option to all members at the time of enrollment and during reassessments. The fifth state (Massachusetts) prefers to avoid detailed documents and communicates its expectations orally to the MCOs. Massachusetts state officials feel this allows for a better exchange of ideas. New Mexico is implementing a policy that when someone is new to long-term care community-based services, they must first go into the agency-based benefit for 120 days and then they have the opportunity to choose participant direction. The State's rationale for this is new enrollees need time to adjust to all of the changes and choices associated with being a new MLTSS recipient. MCOs are required to document in the case file that participant direction was offered, whether the member chose to self-direct, and if the member selected employer authority, or employer and budget authority (if available). All of the five study states offer PD-MLTSS as a voluntary option and allow transitioning back to agency-based services any time the member desires. Tennessee goes farther by requiring MCOs to report the number and percent of those electing the participant-directed option, including the date referred to the FMS Agency.

• The majority of the states require person-centered processes: Arizona, New Mexico, and Tennessee all indicated requiring person-centered processes. Under its new Centennial Care program, New Mexico ensures person-centeredness is met through a new assessment regime tool called the Comprehensive Needs Assessment (CNA) for its budget authority option. Every participant will be assessed at the proper time for his/her specific needs. The CNA will then generate an individual budget amount tailored to each participant's specific needs instead of the current practice of categorizing the individual as having low, medium, or high needs.

• Tennessee was the only state with enrollment targets for PD-MLTSS: All of the states indicated optimism that PD-MLTSS enrollments would increase in the future. Texas state respondents believed that the expansion of PD-MLTSS to 164 mostly rural counties of the state would result in higher numbers of persons self-directing. It will be interesting to monitor this as one of the Texas MCO respondents made the observation that there are fewer participants self-directing in the rural areas of the state. The MCO respondents believed that the majority of participants in these rural counties speak Spanish and prefer using the local agencies and agency workers where Spanish is also the primary language. Respondents in Arizona hoped the expansion of its Agency with Choice option would increase participant direction enrollments. New Mexico indicated the expansion of its Medicaid program under the Section 1115 will increase enrollment in Centennial Care. Tennessee specifies increased enrollments as a quality benchmark in its contracts with MCOs (see Appendix C, Tennessee Enrollment Targets and PD-MLTSS Performance Measures, for Tennessee's enrollment targets).

• Training for members and direct service workers are available in each state but few require it: Each MCO interviewed conveyed having both member and direct service worker training available (and provides the training at the request of the member). Only Tennessee requires member and direct service worker training as conditions of eligibility in their program. In Arizona, member training is not required; however, direct service worker training is mandatory and includes Universal Precautions and Health Insurance Portability and Accountability Act privacy regulations. Additional training materials are available in the Arizona participant manual. Most state contracts stipulated the state is to review and approve training curriculum, but most trainings are conducted by the MCO and not the state.

• All the states require that PD-MLTSS participants have a back-up plan: Current PD-MLTSS programs require the development of a participant back-up plan to address instances when regularly scheduled workers are not available to provide critical services for the member. This requirement is conveyed to MCOs either in language found in the Request for Proposal (RFP), formally executed contracts, state Medicaid manuals, or state policies and procedures. Lack of precision in these documents, however, can lead to some role confusion as to who is responsible. For example, Tennessee's contract has the responsibility for creating a back-up plan assigned to the MCO service coordinator, FMS agency, and the participant. The detailed requirements for implementing back-up plans may also be a deterrent to expanding PD-MLTSS in some states. For example, Arizona goes so far as to require the MCO case manager to dispatch a back-up agency worker within two hours to provide a self-directing member services if the member notifies them that their worker is not available. As of the date of this research, this requirement is based on an Arizona state law. While ensuring that participants have adequate back-up plans is an important support for PD-LTSS, states need to be mindful that such a requirement does not either serve as a deterrent to case managers in presenting the option or make the responsible entity (e.g., the MCO, the FMS, or the state), that is held responsible to ensure back-up, look like "joint employers" of home care workers for purposes of the Fair Labor Standards Act.

• Three of the five states have specific reporting requirements related to PD-MLTSS: New Mexico, Tennessee, and Texas all require specific PD-MLTSS reports related to enrollment. Beyond enrollment, state requirements for PD-MLTSS reports vary. Tennessee currently has the most extensive set of PD-MLTSS performance indicators (see PD-MLTSS Quality Monitoring section). New Mexico reported that it will also have an extensive PD-MLTSS report under Centennial Care and is in the process of developing this in its PD-MLTSS workgroup. While Arizona does not require specific reports, it does have the ability to generate PD-MLTSS reports from the information provided by the MCOs.

3.3. Participant-Directed Managed Long-Term Services and Supports Quality Monitoring

All five states require the MCOs to submit formal quality assurance and improvement (QA/I) plans prior to program implementation. For the most part, MCOs cover the PD-MLTSS within the broader scope of their overall QA/I plans. The following information is also reported in Appendix B, Table B3: PD-MLTSS Quality Monitoring.

• Only Tennessee has specific performance indicators for PD-MLTSS: Four out of the five states did not report specific performance indicators (Tennessee was the sole exception). In addition to enrollment targets, Tennessee also requires MCOs to provide information on number of PD-MLTSS members with a representative, service utilization by type of service, number of referrals to FMS, and number of members who withdraw from PD-MLTSS (see Appendix C, Tennessee Enrollment Targets and PD-MLTSS Performance Measures, for Tennessee's full list of performance indicators).

• MCO monitoring of PD-MLTSS services varied across the five states: Arizona, Massachusetts, and Texas reported they monitor participants who opt for PD-MLTSS in the same manner they monitor members in MLTSS. Tennessee, however, uses technology to ensure PD-MLTSS compliance and quality for all home-based services and supports. The Tennessee MCO described the Tennessee Electronic Visit Verification (EVV) system as being very robust in tracking when workers arrive and depart and which tasks they performed. If the worker does not arrive within one hour of his or her scheduled time, the MCO service coordinator is notified immediately and the worker's timesheet is suspended. Some exceptions are provided, but the flexibility is typically limited to a four-hour window. Other Tennessee stakeholders noted that the EVV sometimes creates more work for the member and FMS agency as the worker and member may have made alternative service arrangements that did not correspond to the EVV schedule. As one stakeholder commented, the EVV is not set up for participant direction in the sense that every service needs to be scheduled at a specific day and hour. Because the EVV notifies the FMS when the worker visit is completed, thus initiating the process for worker reimbursement, the Tennessee MCO reported that the EVV is beneficial in that it reduces the need to contact the member and verifies service delivery.

• All states report having structured opportunities for participant input: Respondents from the states and the MCOs indicated a number of structured ways they seek participant input. All reported using member satisfaction surveys to gauge plan performance. However, none reported that their survey contained PD-MLTSS specific questions. All of the state respondents reported having an MLTSS advisory council and some states (Arizona and New Mexico) encourage MCOs to have their own MLTSS advisory councils. With the exception of New Mexico, none of the advisory councils are specifically focused on PD-MLTSS. In addition to advisory councils, Arizona, New Mexico, and Texas reported the use of community forums or "town meetings" as another avenue to solicit member input.

3.4. Participant-Directed Managed Long-Term Services and Supports Program Features

Most MCOs are given broad discretion in developing and managing PD-MLTSS service delivery options. This discretion means that the features and flexibility of the PD-MLTSS program and what is meant by "participant direction" can vary (see discussion of budget authority models in Tennessee and Texas above). Respondents in Tennessee and Texas remarked that the adoption of PD-MLTSS programs largely depends on the emphasis the state gives to participant direction. Only Tennessee indicated having an incentive from the state (e.g., enrollment targets--see above) to promote PD-LTSS among their members. The following describes some of the common program features of the PD-MLTSS programs in the five states (see Appendix B, Table B1: State PD-MLTSS Program Overview).

• In most states the MCO provides the information and assistance function in PD-MLTSS: Providing information and ongoing assistance to participant-directing individuals is a key supportive function of PD-MLTSS. In Arizona, Massachusetts, and New Mexico, this function is provided by the MCO service coordinator. Because Arizona sees PD-MLTSS as "just a different way to provide services," the MCO service coordinators typically have mixed caseloads of nursing home, assisted living, home and community-based agency-delivered, and PD-MLTSS members. Tennessee MCOs split the counseling function between the MCO service coordinator (who has initial responsibility to describe the program in general terms) and the support broker at the FMS (who has ongoing responsibility for assisting the member with selecting, training, and monitoring workers, and the development of the back-up plan). In Texas, the information and assistance function is performed by the FMS agencies (i.e., Consumer-Directed Service Agencies perform both financial management and information and assistance roles).

• The most common model of PD-MLTSS offered was employer authority (but the MCOs provide little support in finding workers): All of the MCOs interviewed in Arizona, Massachusetts, Tennessee, and Texas offered the employer authority model. None of the MCOs appeared to know of registries where members could find a worker. In Massachusetts and Texas, it was noted that such a registry would be beneficial in the more rural areas of the state. Massachusetts indicated the state maintained a general information website on hiring personal care attendants. In Tennessee, the FMS provider has initiated the development of a worker registry for assisting the member with worker recruitment. This is on a regional basis (East, West, and Middle Tennessee). Texas noted that the state's Consumer-Directed Service Agencies would assist members in placing newspaper ads and/or provide general guidance on where to obtain a worker.

• All the states allow family members to be paid workers, but some set restrictions or special conditions on legally-responsible family members or representatives: The most common worker restriction cited by the states was inability of the spouse or other legally-responsible representative to be hired as the participant's direct service worker. In some states, if the participant used a program representative, this person could not be hired as the member's direct service worker. Tennessee has an additional hiring limitation for non-spouse relatives/friends where the participant cannot hire a person who has lived with them within the past five years. While Arizona allows spouses to be hired as direct services workers, there are some restrictions that are typical in other participant-directed Medicaid programs. For instance, the paid services provided by a family member or spouse cannot be an activity that would ordinarily be performed by a family member; payment of spouses is limited to 40 hours per week; and spouses require additional monitoring, including a quarterly review of expenditures.

3.5. Financial Management Services

All five states require the MCOs to contract with state-approved agencies to provide FMS. The primary duties of an FMS agency include a payment function (providing payments on behalf of the participant to workers, agencies, or vendors for goods and services), a reporting function (generating expenditure reports for participants, MCOs, and state programs), and a management function (managing employer tax and insurance responsibilities). While the FMS agencies across the five states fulfill these functions, many provide other types of support as well (e.g., ongoing information and assistance (Tennessee), assistance with worker recruitment (Arizona, Tennessee, and Texas), providing training for participants and/or workers (Texas).

• FMS selection varies across the states: Across the five states there were two basic ways MCOs selected FMS entities with which to contract: (1) the state delegates the authority to each MCO to select and contract with FMS entities who have been approved as a Medicaid Provider by the state (Arizona, Massachusetts, and Texas); or (2) the state contracts directly with providers (New Mexico and Tennessee) and the MCO is required to contact with these entities as well. In Tennessee, the contract is a three-way contract signed by the state, MCO and FMS agency.

• The number of FMS providers varies greatly across the states: New Mexico and Tennessee have a single FMS entity. Arizona (n=3) and Massachusetts (n=4) have a very small number of FMS agencies, while Texas has upwards of 400 FMS agencies. Texas has had an open period of allowing any interested entities to participate as an FMS agency. Entities interested in FMS must attend a three-day training and pass a test to show they have the required skills and knowledge. The three-day training and "certification" is run by the state. Once they are certified by the state, MCOs can contract with these FMS agencies. While there are benefits for states to allow multiple FMS agencies (e.g., allowing flexibility for the MCO and participant to select an FMS that is the right "fit" in terms of experience, and having an option in place if one FMS did not work out), there are drawbacks as well. Increased quantity does not always mean enhanced quality, and it is harder to monitor FMS activities. With the exception of Tennessee, MCOs do not closely monitor the tasks and performances of the FMS.

• FMS providers reported contracting with multiple MCOs: In all five states, MCOs were required to contract with any FMS provider approved by the state. So the majority of MCOs reported receiving FMS services from multiple providers. New Mexico and Tennessee were the only states with one approved FMS provider. States operating with more than one FMS providers can be challenged with monitoring oversight since universal standards are not required by the state.

3.6. Participant Perspectives on Participant-Directed Managed Long-Term Services and Supports

Six advocacy organizations representing various populations of LTSS users across the life-span were interviewed regarding their perspectives on PD-MLTSS. While five of the six advocacy groups were only able to discuss PD-MLTSS in very general terms, it was clear that few had either advocacy or personal involvement with the state or MCOs regarding PD-MLTSS. As a result, the perspectives from these five organizations were neutral or somewhat positive towards MLTSS. The one advocacy organization that was directly involved with participants in a PD-MLTSS program gave a vastly different perspective--one that felt participants' views were being solicited on a very limited basis and were marginalized in any discussion regarding the operations of PD-MLTSS.

In an effort to try and better understand the experiences of individuals in PD-MLTSS, NRCPDS asked the NPN, a national advocacy organization for individuals enrolled in PD-LTSS programs, to query their members regarding their experiences with PD-MLTSS programs. Ninety-three members were contacted but only four (n=4) responded within the timeframe. Overall, this small group felt they had little to no direct participation in the development of PD-MLTSS in their state.

According to these four NPN respondents, if they ever had a problem or question regarding MLTSS, they did not feel they had the ability to directly access either the state or the MCO. All of the respondents indicated that the initial orientation and enrollment was confusing, complex, and duplicative. When asked what would they improve, the respondents indicated that expanded participant representation on advisory groups to the state and MCO would enable a more direct pathway to provide input and greater participant voice in development, implementation and management of the system.

4. Study Limitations

Before discussing the study implications, the limitations of this study must be acknowledged. First, this report encompasses the perspectives of key stakeholders in five states. While the states were selected because of their historical experience with PD-LTSS, managed care, or both, the experiences of these states may not be generalizable. Second, neither time nor resources permitted interviews with all MCOs operating PD-MLTSS programs in the five states. Thus, the MCO perspectives presented above should not be seen as reflective of all MCOs in that state.

A final limitation is the paucity of participant perspectives on PD-MLTSS in the sections above. Again, given the time and available resources, a research design decision to capture the participant perspective via interviews with advocacy groups was made. While these interviews provided some general insights regarding PD-LTSS, their overall utility for understanding how MLTSS impacts the participant's experience was limited. While the study includes four participant perspectives, only two are from our five states (New Mexico and Texas) and cannot be seen as being representative of participants in those states or of participants in general. Despite its limitations, the study has identified both promising practices and areas of concern in PD-MLTSS that are discussed in the implications section below.

5. Study Implications

States play a major role in how PD-MLTSS is operationalized: How PD-MLTSS is shaped and operates in a given state is determined by either the presence or absence of state policies and procedures that emphasize participant direction in MLTSS. Because most of the states in this study possessed nominal language in their MLTSS contracts regarding participant direction, subsequent implementation of PD-MLTSS was delegated to MCOs who may, or may not, understand the philosophy or implementation of participant direction. Perhaps the clearest example is Tennessee's EVV system. From the state's perspective, it is verifying the receipt of services and the need for emergency back-up since the system notifies of worker no-shows. From the MCO perspective, the EVV makes the PD-MLTSS more efficient as it automates the verification of service delivery and reimbursement, thereby reducing unnecessary communication with the participant. But other non-MCO stakeholders described it as a tracking mechanism that takes control and flexibility away from participants when managing their workers' schedules. The issues are who is really in control and how can participant-directed supports and services be integrated into the management of coordinated health care and LTSS.

There are examples of how the principles of managed care and participant direction can be integrated: Early research and commentary on the compatibility of participant direction and managed care identified the possible disconnect between the goals of managed care entities (improve member health and well-being by providing efficient, coordinated, and cost-effective services) and participant direction (improve participant health and well-being by providing participants with meaningful choices and control over their services). The Commonwealth Care Alliance (CCA) in Massachusetts has been developing protocols that explicitly integrate PD-LTSS with MLTSS. CCA is working on how to integrate the participant's LTSS plan with the full care plan and how to involve the participant and their worker (with the participant's permission) into the total care planning process. Given the low turnover of participant-directed aides, involving the worker not only provides continuity of care but also can provide CCA an additional point of information to respond to health problems as they develop and possibly avoid emergency room visits and unnecessary hospital admissions. The Massachusetts NPA exemption also allows CCA to avoid bringing in an expensive nurse for simple issues like medication management (an area where other states have also amended their NPA). Based on its experiences, CCA is seeing tremendous reasons to support and grow participant direction, not the least of them is the CCA's work illustrating the common goals to MLTSS and PD-LTSS: (1) The improvement of participant health and well-being; (2) The improvement of service satisfaction; and (3) The reduction of service costs. Measuring the comparative effectiveness of participants enrolled in PD-MLTSS against those enrolled in only MLTSS against these measures should be of high interest for states and MCOs alike.

Training for MCO service coordinators is vital: In Arizona, where PD-MLTSS is considered one among many service options, there is little incentive on behalf of MCO service coordinators (who are managing mixed caseloads across the LTSS continuum) to promote PD-MLTSS. Across all five states, service coordinators acknowledged that arranging PD-MLTSS involves extra "upfront" time working with the participant. Couple this additional effort with the expectation that the Arizona service coordinator is responsible if the participant's back-up plan fails, it is not surprising that Arizona has the lowest PD-MLTSS take-up rate (1.2%) among the five study states. Tennessee stakeholders talked about how the take-up rate for PD-MLTSS depends on the service coordinator where some service coordinators have 25% of their caseload self-directing while others have zero. While Arizona illustrates a structural barrier, Tennessee suggests the need for additional professional development training in participant direction. Across the five states the amount and type of training of MCO staff varied greatly. Too often, staff training was restricted to the mechanics of presenting the participant direction option to the member.

How PD-MLTSS is presented to participants is critical: Respondents across three of the states (New Mexico, Tennessee, and Texas) remarked that participants are overwhelmed with materials about self-direction and tend to be apprehensive of becoming an employer. While each state required MCOs to have person-centered processes in place, it does not appear that all service coordinators receive training in person-centered planning or participant direction that could reduce the participant's sense of feeling overwhelmed. Beyond training of professional staff, the use of peers is an undeveloped resource that could help people become comfortable with PD-LTSS. Because individuals who newly find themselves in need of LTSS are often overwhelmed, in 2014 New Mexico will institute under Centennial Care a 120-day adjustment period when the individual will receive agency-based services. At the end of the 120 days the individual will then be presented the option of PD-MLTSS. Such an arrangement begs a number of questions: What if the person already has a participant-directed plan in mind? After receiving services from an agency for four months, how likely is the participant to change their services to self-direction? Person-centered planning needs to start at the beginning. While New Mexico's adjustment period is well intentioned, it may be best as an option and not a requirement for all new enrollees. According to members of the NPN, people learn best by actually doing and perhaps agency-delivered services can help while a person is developing their participant-directed plan and have agency-based services available in the interim--so there is no gap in services--but they should not have to be mandatory for four months.

States expressed that commitment to participant direction is important to program growth: Across all five states the numbers of participants enrolled in PD-MLTSS is low. Few states could compare enrollment numbers between PD-MLTSS and MLTSS or the take-up rates for participant direction before and after MLTSS was implemented. Only Tennessee has clear enrollment expectations and a review of the state's enrollment targets, which illustrates how the program has grown (see Appendix C, Tennessee Enrollment Targets and PD-MLTSS Performance Measures). Without such expectations, even states with a historical commitment to participant direction leave the future growth of PD-MLTSS to the discretion of MCOs (who may or may not be committed). While such an approach has worked in Texas (where non-MCO stakeholders lauded MCOs for promoting PD-MLTSS even though the state appears to be indifferent) it is not a guaranteed approach in other states. Even in Tennessee, non-MCO stakeholders believed many more members could be enrolled in PD-MLTSS. One Tennessee stakeholder suggested that participant direction is the default program for eligible participants so that participants are required to "opt out" of PD-MLTSS rather than to "opt in." The service coordinator presents them with their options and if they do not want to choose participant direction they actually have to sign off to say they do not want to choose it. Such a policy changes the dynamic of the presentation of options.

PD-MLTSS would benefit from clarity of the roles and responsibilities of the different PD-MLTSS supports: In some states this lack of clarity is due to the lack of specificity in state contracts with MCOs (or absence of policy and procedure manuals) regarding participant direction, participant-directed services, and necessary supportive services such as FMS. This lack of specificity can result in important support functions not being readily available. An example of this was the lack of knowledge about worker registries for PD-MLTSS members. Even though every MCO offers an employer authority option of PD-MLTSS, no one knew whether a worker registry existed in the state to assist participants in finding a worker. Many MCOs assumed the state or the FMS agency would assist in this area (or express the need for such a registry but were uncertain that it was their role to provide one). This small point depicts perhaps a larger issue in PD-MLTSS, namely that unless it is delineated as a specific role or responsibility in a RFP, contract, or policy and procedure manual, it depends on the interpretation of the MCO, FMS, or the participant as to who is responsible. This can lead to confusion and frustration on behalf of all three parties. Another example is the existence of overlapping roles and responsibilities. For example, in Tennessee, like most states with PD-MLTSS, if an individual wants to select the participant-directed option, it is a decision that is made with the MCO service coordinator. Afterwards, they are assigned a support broker with the state FMS provider, leaving the participant with a support broker from the FMS and a services coordinator from MCO. As one Tennessee key informant noted, the existence of a support broker and a service coordinator can lead to some confusion--which is very understandable--about whom to ask questions of and who provides direct guidance to the participant.

FMS is a key PD-MLTSS support element: The five states used the FMS entities in various ways. In all states the FMS provided traditional financial management support, but some states had the FMS provide other types of support as well. For example, in Tennessee the FMS provides ongoing information and assistance to the member regarding participant direction. In Arizona, Tennessee, and Texas the FMS provides assistance with worker recruitment. In Texas the FMS provides training for members and/or their workers. The number of FMS entities in the five state sample ranged from one (New Mexico and Tennessee) to approximately 400 (Texas). One negative to PD-MLTSS in Texas is the number of FMS the MCO has to contract with. This may make it difficult to monitor quality of service provision. Texas is very committed to participant direction but the system to deliver PD-MLTSS is complex with hundreds of different partners.

MCOs would benefit from increased engagement from participants: The idea of involving participants in the design and evaluation of LTSS has been promoted for decades. Each of the states and MCOs in this study described various ways participant involvement is sought (e.g., public forums, town halls, member surveys, and advisory boards). However no one reported that their survey contained PD-MLTSS specific questions. While all of the state respondents reported having an MLTSS advisory council and some states (Arizona and New Mexico) encourage MCOs to have advisory councils. With the exception of New Mexico, none of the advisory councils are specifically focused on PD-MLTSS. Given the nature of PD-MLTSS, the general lack of meaningful participant engagement is a major shortcoming. As the title of a 2011 study on the relative advantages of an advisory committee that is committed to the participant direction, "it's not so simple" to engage participants given limitations in time and resources.¹⁶ Despite the limitations, participant engagement is seen as an avenue for better PD-MLTSS program design and improvement and improved member satisfaction.

APPENDIX A. STAKEHOLDER ORGANIZATIONS BY STATE¹⁷

Notes

1. Sciegaj, M., & Selkow, I. (2011). Growth and Prevalence of Participant Direction: Findings from the National Survey of Publically Funded Participant-Directed Services Programs. Retrieved from http://web.bc.edu/libtools/details.php?entryid=340.

2. Carlson, B.L., Foster, L., Dale, S.B., & Brown, R. (2007). Effects of Cash and Counseling on Personal Care and Well-Being. Health Services Research, 42(1p2): 467-487.

3. Dale, S.B., & Brown, R.S. (2007). How does Cash and Counseling affect costs? Health Services Research, 42(1p2): 488-509.

4. Doty, P., Mahoney, K.J., & Sciegaj, M. (2010). New State Strategies to Meet Long-Term Care Needs. Health Affairs, 29(1): 49-56.

5. Engguist, G., Johnson, C., Lind, A., & Palmer-Barnette, L. (2010). Medicaid-Funded Long-Term Care: Toward More Home-and Community-Based Options. Center for Health Care Strategies, Incorporated. Retrieved from http://www.chcs.org/usr_doc/LTSS_Policy_Brief_.pdf.

6. Saucier, P., Kasten, J., Burwell, B., & Gold, L. (2012). The Growth of Managed Long-Term Services and Supports (MLTSS) Programs: A 2012 Update. Prepared for the Centers for Medicare & Medicaid Services (CMS) by Truven Health Analytics. Retrieved from http://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Delivery-Systems/Downloads/MLTSSP_White_paper_combined.pdf.

7. NRCPDS (2013). National Program Map of Participant-Directed Programs. Retrieved from http://web.bc.edu/libtools/insights-publications.php.

8. Meiners, M., Mahoney, K., Shoop, D., & Squillace, M. (2002). Consumer Direction in Managed Long-Term Care: An Exploratory Survey of Practices and Perceptions. The Gerontologist, 42(1): 32-38.

9. Mahoney, K., Meiners, M., Shoop, D., & Squillace, M. (2003). Cash and Counseling and Managed Long-Term Care. Case Management Journal, 4(1): 18-22.

10. Kodner, D.L., Mahoney, K., & Raphael, T. (1997). Managed care and consumer-directed care: Are they compatible. Managed Care and Aging, 4(4): 3-4.

11. Simon-Rusinowitz, L. Bochniak, A.M., Mahoney, K.J., Marks, L.N., & Hecht, D. (2000). Implementation issues for consumer-directed programs: A survey of policy experts. Generations, 24(1): 34-40.

12. Stone, R. (1997). Consumer advocacy and managed care. Managed Care and Aging, 4(3): 3-4.

13. Kodner, D.L. (2003). Consumer-directed services: Lessons and implications for integrated systems of care. International Journal of Integrated Care, 3(17): 1-7.

14. Crisp, S., Sciegaj, M., DeLuca, S., & Mahoney, K. (2013). Participant Direction in Home and Community-Based Services: Summary of Selected Provisions from Integrated Care RFPs and Contracts. Integrated Care Resource Center.

15. Kodner, D.L. (2011). Medicaid Managed Care for People with Intellectual Developmental Disabilities--Revisiting the Experiences of Arizona, Michigan, Vermont, and Wisconsin. Prepared for the Arthur Webb Group. Retrieved from http://arthurwebbgroup.com/pdfs/medicaid_managed_care.pdf.

16. McGaffigan, E. (2011). It's not so Simple: Understanding Participant Involvement in the Design, Implementation, and Improvement of Cash and Counseling Programs. Retrieved from http://scholarworks.umb.edu/cgi/viewcontent.cgi?article=1054&context=doctoral_dissertations.

17. Crisp, S., Sciegaj, M., DeLuca, C., Mahoney, K.J. (2013). Participant Direction in Home and Community-Based Services: Summary of Selected Provisions from Integrated Care RFPs and Contracts. Integrated Care Resource Center.

18. Tennessee Contract: Quality Benchmark #5 Section 2.9.8.13.1.5. http://www.tn.gov/tenncare/forms/middletnmco.pdf.

19. Tennessee Contract: Section 2.30.6.5. http://www.tn.gov/tenncare/forms/middletnmco.pdf.

Appendices

Appendix A. Stakeholder Organizations by State

Appendix B. Participant-directed Managed Long-term Services and Supports Program Tables

Appendix C. Tennessee Enrollment Targets and Pd-mltss Performance Measures

Tennessee developed distinct benchmarks, standards, and quality outcomes for PD-MLTSS programs. Quality benchmarks specified in Tennessee's contract¹⁸ require the three MCOs (United HealthCare, Amerigroup, and Volunteer State Health Plan) to increase the number of participant-directing members each year over the life of the contract. The contract specifies the level of increase from one year to the next. For example, Tennessee's United Healthcare contract specifies that the MCO is measured against the following goals:

Year	# Participant-Directing
2011	450
2012	750
2013	1,000
2014	1,250
2015	1,400
2016	1,500

In Tennessee, the FMS and MCO routinely forward the following reports to the state: (1) number of members electing the participant-directed option including their name, social security number, and phone number; (2) number of members who disenrolled and reason for the action; (3) reports and investigations of critical incidents and results of follow-up; (4) number of fair hearings requested; (5) number of scheduled visits with self-directing members; (6) number of late or missed home assessment visits; (7) maximum and average time from FMS referral to commencement of participant direction, including the date the member was referred to the FMS; (8) number and percent of members enrolled in participant direction who appointed a representative to manage the program on their behalf; (9) number and percent of members receiving participant -directed services by type of service (attendant care, companion care, homemaker, in-home respite, or personal care); (10) total number of members who do not wish to receive traditional home and community-based services from contract providers pending enrollment into participant direction; and (11) the total number of days per member that home and community-based services have not been received.¹⁹