A special examination of the Community Care Services Program

A Special Examination of the

January 2007

Community Care Services Program

by the
Georgia Department of Audits and Accounts
Russell Hinton, State Auditor

This Special Examination was conducted in response to a request by the Senate Appropriations Committee to review the Community Care Services Waiver Program (CCSP) administered by the Department of Community Health (DCH) through the Department of Human Resources (DHR). In its request, the Committee asked us to provide: A ten-year reconciliation, by fund source, of the number of CCSP slots funded by the General
Assembly and the actual number of slots funded by the program; An analysis of the expenditures per slot and an explanation of the cost components; A critique of the program's purpose, goals and outcome measures; and General comments on how efficient and effective the program has been over the past 10 years.
In response to the four parts of the request from the Senate Appropriations Committee, we found the following:
Ten-Year Reconciliation
It was not possible to fully accomplish a 10-year slot reconciliation and associated analyses of funding for the CCSP. In order to accomplish this, the following resource elements would be necessary:
A uniform understanding and consistent application of the term "slot" between the agencies and the General Assembly;
Appropriations documents that annually identify both the total amount of dollars and total number of slots funded by the General Assembly for the CCSP waiver; and
Agency records that annually reflect both the total amount of dollars expended and total number of slots filled in the respective waiver.
However, as listed below and more fully explained in the following paragraphs, we found that a reconciliation between total annual appropriations and total actual program activities was precluded because these necessary elements did not fully exist.
Appropriations documents did not address the total funding and total slots for the CCSP. Consequently, instead of comparing total slots and dollars funded by the General Assembly to total slots and dollars funded by the program, we compared annual incremental changes from the General Assembly to annual incremental changes in the program.
We found no correlation between the annual changes in the number of slots funded through appropriations and annual changes in the number of unduplicated recipients or active clients actually served.
We also found no consistent correlation between changes in the annual amount of funds appropriated and the annual change in the actual amount of program funds expended.

Special Examination Community Care Services Program

ii

Analysis of Expenditures Per Slot The average annual cost per unduplicated CCSP recipient has steadily increased during the ten-year period, from $4,140 in 1997 to $10,111 in 2006 (144%). The number of unduplicated recipients served has remained relatively stable, beginning at 12,607 in 1997 and ending at 12,596 in 2006.
An Explanation of the Cost Components In FY 2006, approximately 63% of CCSP expenditures were attributed to Personal Support Services, which are direct care services that include assisting recipients with eating, bathing, dressing, personal hygiene, and other activities of daily living. Most of the remaining CCSP expenditures were for Care Coordination (17%) and Alternative Living Services (12%). Personal Support Services have accounted for an increasing percentage of CCSP expenditures, growing from 42% in 1997. It is interesting to note that Care Coordination costs have increased from $9.9 million in 1997 to more than $21 million in 2006 (113%), though the program is serving essentially the same number of clients in 2006 as it did in 1997.
Critique of Purpose, Goals and Outcome Measures The purpose of the CCSP is largely established in federal statute, and broadly requires that the CCSP serve to decrease or avoid institutionalization and offer the appropriate services to CCSP eligible clients. This purpose is common to waivers in other Southern states. The CCSP purpose is further defined in O.C.G.A. and program materials, but these definitions are inconsistent with each other. It is unclear if CCSP clients must be aged and disabled or aged or disabled. It is also unclear whether an age threshold is appropriate to the CCSP program, but more than 20% of CCSP clients were under age 60 in 2006.
CCSP regulations and materials identify several goals, but do not establish measurable targets for these goals. Subsequent to the review, DHR provided twelve measures, but it is unclear the extent to which these measures should be considered goals. Further analysis is required to assess the appropriateness and validity of the information provided.
There are four outcomes reported in the CCSP annual report regarding length of stay, quality of services, client discharge status, and cost-effectiveness, only one of which can be used to measure the effectiveness of the program in comparison to its purpose. The length of stay measure clearly relates to the CCSP purpose. The quality measures include: provider assessments, which report on compliance and the effectiveness of services for a small sample of the CCSP population, but not in aggregate for the overall quality of CCSP services; client satisfaction surveys, which may not be generalizable to the CCSP population; and reviews of care plans, which review the completeness of client files and effectiveness of care plans for a small sample of the CCSP population, but not in aggregate for the overall quality of CCSP care plans or services. It is unclear how the client discharge status relates to institutional avoidance and the programs have not evaluated this measure in this respect. Agency staff were unable to explain how the key components of the cost-efficiency measures were calculated for any of the years in our review. Subsequent to the review, DHR provided twelve measures of program performance. Further analysis is required to determine the value and validity of the measures and associated data as related to the proper assessment of CCSP outcomes.

Special Examination Community Care Services Program

iii

General Comments on Efficiency and Effectiveness
DHR monitors the efficiency of the CCSP in its annual report, but DCH has not reported measures of efficiency to the federal government since 2002, and neither agency was able to explain or document these calculations from prior years. Alternatively, we attempted to compile information on other possible indicators of efficiency, such as the percentage of total CCSP expenditures attributed to administrative costs. Although we calculated this percentage at roughly 18%, the program lacks established benchmarks for use in determining whether such percentages indicate efficient operating levels. The average cost to coordinate care for each CCSP client has risen from $878 in FY 1997 to $1,780 in FY 2006, an increase of 103%. There is wide variance in the per-unduplicated client cost to coordinate care across the state. Finally, we discovered significant additional Medicaid outpatient costs attributable to CCSP recipients but paid outside of CCSP Medicaid claims.
It is unclear the extent to which the program has been effective in avoiding institutionalization or offering the needed services to CCSP clients. Without an appropriate and reliable framework of goals and outcomes, the effectiveness of the CCSP cannot be comprehensively assessed. However, we attempted to assess program effectiveness using the information available regarding length of stay, discharge status and the waiting list. The reported results regarding length of stay and client discharge status independent of other comparative data do not allow analysis of any resulting appropriate institutional avoidance. Finally, review of the management of the 12 regional waiting lists suggests that the CCSP program may not be effective in placing CCSP clients in a manner that is based solely on actual client needs. There are regional disparities regarding which clients are given priority on the CCSP waiting lists. To some degree, whether an applicant receives services is related as much to where he or she lives as to their level of need for services.

1
Table of Contents
Background
Examination Purpose ...................................................................................................................... 1 Explanation of Waiver Programs .................................................................................................... 1 CCSP Services and Program Structure ......................................................................................... 1 Scope and Methodology .................................................................................................................... 2
Results of Requested Analyses
10-Year Reconciliation of CCSP Slots Funded and Actual ................................................ 3 Analysis of the Expenditures per Slot .......................................................................................... 8 Explanation of the Cost Components ........................................................................................ 10 Critique of the Program's Purpose, Goals and Outcome Measures ............................... 12 General Comments on the Program's Efficiency and Effectiveness ............................... 17
Appendix A - Various Source Data ............................................................................................. 29 Appendix B - Sample CCSP Appropriations Language .................................................... 31 Appendix C Map of Area Agencies on Aging Regions..................................................... 35

Special Examination - Community Care Services Program

1

Examination Purpose
This Special Examination was conducted in response to a written request by the Senate Appropriations Committee to review the Community Care Services Program (CCSP) administered by the Department of Community Health (DCH) and the Department of Human Resources (DHR). In its request, the Committee asked us to provide:
A ten-year reconciliation, by fund source, of the number of CCSP slots funded by the General Assembly and the actual number of slots funded by the program;
An analysis of the expenditures per slot and an explanation of the cost components;
A critique of the program's purpose, goals and outcome measures; and General comments on how efficient and effective the program has been
over the past 10 years.
Explanation of Waiver Programs
In 1981, Congress established the Home and Community-based Waiver program, which allowed states to use Medicaid funds for services in recipients' homes and communities that would otherwise require nursing facility or institutional care. By offering these services in an alternative setting, the states are allowed to "waive" their obligation to provide the services in an institutional setting. Under federal regulations, neither the average nor the total cost of providing home and communitybased services may be more than the cost of providing care in an institution or nursing home. Each program is designed to help people who qualify for institutional care remain in the community or return to the community from nursing homes or hospitals. Participants are required to be eligible for Medicaid, although qualifying incomes and resources may be higher than those permitted in the regular Medicaid program.
Through a waiver with the federal Centers for Medicare and Medicaid Services (CMS), DCH agrees to provide services to Georgia's CCSP clients and obtain the federal matching funds. DHR operates the program and receives the state appropriations.
CCSP Services and Program Structure
Georgia's CCSP program was codified in state law in 1982, and according to Medicaid claims payment data, served 12,596 unduplicated recipients in FY 2006. Total expenditures for the program were $127.4 million. The CCSP program serves aged and/or physically-disabled clients. DHR contracts with 12 Area Agencies on Aging (AAA) to administer the CCSP locally. The AAAs provide care coordination either directly or via subcontracts with private care coordination companies.
The CCSP offers the following services to waiver recipients: service coordination (help with managing care needs and services); personal support services (i.e., assistance with daily living activities-- e.g., bathing, dressing, preparing meals, and housekeeping);

Special Examination - Community Care Services Program

2

home health services (nursing, home health aide, and occupational, physical, and speech therapy);
emergency response systems; respite care (caregiver relief); adult day health care; alternative living services (e.g., personal care home); and home delivered meals.
The CCSP program uses 12 separate waiting lists to prioritize admission into the program. As of September 20, 2006, there were 2,257 people waiting to receive CCSP services. A person may be selected to receive services based on their severity of need, the availability of informal/family support, their length of time on the waiting list, and the applicant's continued eligibility for the level of care provided in a nursing home.

Scope and Methodology
The scope of this Special Examination was limited to the issues regarding the CCSP program cited in the Senate Appropriation Committee's request. Our methodology included:
Reviews of relevant O.C.G.A. sections, DCH and DHR annual reports, reports to the federal government, provider manuals, and waiver documents approved by CMS;
Analysis of Medicaid claims payment data; Review of budget, appropriations and financial documents; Review of DHR program data regarding slots filled; Appropriations information provided by the Governor's Office of
Planning and Budget (OPB); Review of academic and professional literature; Review of other states' waivers and reported data; and Interviews with state agency and subcontractor staff.
After systematically reviewing data sources to answer the questions, wide variances in reported data were identified. To the extent possible, efforts were made to use data with a clear and replicable methodology. Generally, Medicaid claims payment data generated by the Department of Audits and Accounts' (DOAA) Healthcare Information Analysis Group (HIAG) was used for analysis related to direct benefit expenditures and unduplicated recipients. Some of the information is relayed from reports provided by DCH or DHR program staff, including the annual 372 Reports of program utilization submitted to CMS, Aging and Information Management System (AIMS) data provided by DHR, the Client Health Assessment Tool (CHAT) data provided by DHR, and from the DCH and DHR Annual Reports. For a comparison of the information provided by the various sources, see Appendix A.
Comments regarding this review were received and discussed with the agencies and applicable revisions are reflected in this report.

Special Examination - Community Care Services Program

3

Results of Requested Analyses
A Ten-year Reconciliation, by Fund Source, of the Number of CCSP Slots Funded by the General Assembly and the Actual Number of Slots Funded by the Program
It was not possible to fully accomplish a 10-year slot reconciliation and associated analyses of funding for the CCSP. In order to accomplish this, the following resource elements would be necessary:
A uniform understanding and consistent application of the term "slot" between the agencies and the General Assembly;
Appropriations documents that annually identify both the total amount of dollars and total number of slots funded by the General Assembly for the CCSP waiver; and
Agency records that annually reflect both the total amount of dollars expended and total number of slots filled in the respective waiver.
However, as listed below and more fully explained in the following paragraphs, we found that a reconciliation between total annual appropriations and total actual program activities was precluded because these necessary elements did not fully exist. Appropriations documents did not address the total funding and total slots
for the CCSP. Consequently, instead of comparing total slots and dollars funded by the General Assembly to total slots and dollars funded by the program, we compared annual incremental changes from the General Assembly to annual incremental changes in the program. We found no correlation between the annual changes in the number of slots funded through appropriations and annual changes in the number of unduplicated recipients or active clients actually served. We also found no consistent correlation between changes in the annual amount of funds appropriated and the annual change in the actual amount of program funds expended.

Identifying Total Slots and Funds Appropriated
The total number of CCSP slots and dollars appropriated by the General Assembly cannot be identified in either the Appropriations Act or Budget in Brief (BIB) documents. However, the appropriations documents identify the annual incremental changes in "slots" and dollars appropriated by the General Assembly for CCSP. In most cases, the appropriations language refers to increases in "additional clients," or "slots," but some appropriations were for "enhancing services" for existing clients. We identified these incremental changes, but requested that the Governor's Office of Planning and Budget (OPB) also verify CCSP appropriations. OPB's analysis of incremental CCSP appropriations confirmed most of the items we identified, but diverged in some instances from the terminology found in the actual Appropriations

Special Examination - Community Care Services Program

4

Act, BIB or tracking document. The most significant variance was for two appropriations items: (1) In 1999, OPB identified 2,115 slots that were not identified in the other documents. (The other documents identified the appropriations, but did not specify the number of slots. See Appendix B for a sample of the appropriations language and OPB analysis). (2) In 2001, OPB labeled appropriations as "services for another 2,183 clients in CCSP," but other appropriations documents described these as "enhanced services" slots. For the reconciliation, we used the items identified by OPB except for these two discrepancies. Were these two figures used as increases in "slots" or "additional clients," the differences observed in Exhibit 1 would be larger.
After review of the available documents and discussions with DHR and OPB, it became clear that the use of the existing appropriations documents to identify appropriations for the purposes of reconciliation to actual activities is limited for the following reasons:
The calculation or identification of total appropriated slots and funding is not possible;
Appropriations documents sometimes identify "slots" and "additional clients" and at other times only the "expanding of services" which may or may not be intended to go to existing CCSP clients; and
DHR may be able to apply other allocated community health funds for CCSP.

Identifying Slots Funded by the CCSP Program
Beyond the limitations in the actual appropriations documents, the next obstacle in comparing the slots appropriated by the General Assembly to the slots actually filled by DCH and DHR exists in identifying the number of slots filled by the programs. In fact, prior to this project the term "slot" was not defined in any program records. During this review, CCSP personnel defined the term as the average projected cost per client for [a given] funding period, but they insisted that projected slot costs are applicable only during the period for which funds are allocated and are noncomparable (and therefore non-reconcilable) across different fiscal periods.
DHR tracks the number of CCSP slots filled in terms of "active clients," meaning the number of clients who received case coordination services in a given month as recorded in the Client Health Assessment Tool (CHAT). The number of unduplicated recipients denotes the number of unique persons that received CCSP direct services, as recorded in the Medicaid claims payment system.

Reconciliation of Slots Funded by the General Assembly and Funded by the CCSP
Despite the limitations identified above, we compared the incremental changes in appropriated slots to the actual slots funded by the program using the best available information. As seen in Exhibit 1, there is no correlation between the additional "slots"/"clients" appropriated by the General Assembly and the number of unduplicated persons served by the program. For example, in FY 1998, the General

Special Examination - Community Care Services Program

5

Assembly funded an additional 2,061 slots but the CCSP served 105 less people than the year before. Conversely, while the General Assembly funded 460 additional slots in FY 2005, the CCSP program served an additional 1,394 people.

Exhibit 1 Comparison of Annual Changes in Appropriated "Slots" and "Additional
Clients" and Actual Unduplicated Recipients

Change From Prior Year

Change in Unduplicated Recipients Compared to
Appropriated

Appropriated

FY

Slots/Additional

Clients1

Unduplicated Recipients2,3

1998 1999 2000 2001 2002 2003 2004 2005 2006

2,061 50 N/A N/A
1,000 822 84 460 279

(105) N/A N/A 952 38 (1,495) (833) 1,394 (1,178)

(2,166) N/A N/A N/A (962)
(2,317) (917) 934 (1,457)

Sources: (1) OPB (Identification of Changes in Appropriations); (2) 372 Reports (Unduplicated Recipients for 1998 and 2000); (3) DOAA HIAG (Unduplicated Recipients for 2001-2006) data
Notes: (A) Slots appropriated were identified as "slots" or "additional clients" by OPB analysis. (B) The 372 Report for 1999 was not provided; therefore, changes in persons served using unduplicated recipients is not available for 1999 and 2000. (C) OPB did not identify any changes in appropriations for 1997.

It should be noted that "enhanced services" slots were not compared here and their inclusion would expand the gap between persons served and slots appropriated. Moreover, if the "active client" figures, or the client figures reported in the CCSP annual reports were used for the years in which such data is available, a similar lack of correlation between changes in appropriated "slots" and "additional clients" would be observed.

Reconciliation of Annual Incremental Funds Appropriated by the General Assembly and Funds Expended by the CCSP
As Exhibit 2 illustrates, there is no consistent correlation between the annual changes in the amount of funds appropriated and the amounts expended by the program. For example, in FY 2004, although an additional $4 million was appropriated, program expenditures decreased by about $9.7 million, or roughly $13.8 million less than the appropriated amount. Conversely, in FY 2003, although there was $2.4 million less appropriated, actual

Special Examination - Community Care Services Program

6

expenditures increased by $9.8 million.

Exhibit 2 Comparison of Annual Changes Between Appropriations and Actual
Expenditures

Change From Prior Year

Change in Expenditures Compared to Appropriations

FY

Appropriated Funds1

Actual Expenditures2,3

1998

$9,068,400

$7,314,022

($1,754,378)

1999

$7,419,746

N/A

N/A

2000

$4,568,279

N/A

N/A

2001

$10,099,259

$8,127,313

($1,971,946)

2002

$17,623,443

$17,006,957

($616,486)

2003

($2,444,991)

$9,871,930

$12,316,921

2004

$4,088,438

($9,788,795)

($13,877,233)

2005

$11,164,312

$17,562,460

$6,398,148

2006

$2,508,591

$6,109,122

$3,600,531

Sources: (1)OPB (Appropriated Funds); (2) 372 Reports (Actual Expenditures for 1998 and 2000); (3)DOAA HIAG data

(Actual Expenditures for 2001-2006)

Notes: (A) It is inaccurate to relate the Appropriated Funds amounts to the Appropriated Slots numbers shown in Exhibit 1, as these funding amounts represent the total incremental funds appropriated to the CCSP, including those funds designated for something other than additional slots. For example, the $17 million change in 2002 appropriations includes $3.2 million for a rate increase to CCSP providers. (B) There are no changes in actual expenditures identified for FY 1999 and FY 2000 because DCH was unable to locate the 372 Report for FY 1999. (C) The change in appropriations identified by OPB for 2006 does not appear to account for a full federal match, which may explain much of the apparent difference between expenditures and appropriations. (D) OPB did not identify any changes in appropriations in 1997.

For the reconciliation of incremental dollars appropriated to incremental dollars spent shown in Exhibit 2, we compared all changes in CCSP appropriations identified, including rate increases, to all changes in CCSP expenditures in order to portray the clearest reconciliation possible. Changes in expenditures included changes in direct benefits, care coordination and state administrative costs. The changes in appropriations identified by OPB included the total state and federal funds except for one year (see note).

Fund Sources of CCSP Expenditures
Exhibit 3 below shows the growth in expenditures for the CCSP program over the ten-year timeframe.

Special Examination - Community Care Services Program

7

Exhibit 3 CCSP Total Expenditures by Fund Source 1,2,3

$140,000,000

$120,000,000

$100,000,000

Expenditures

$80,000,000

$60,000,000

$40,000,000

$20,000,000

$0 1997

1998

1999

2000

2001 2002 Fiscal Year

2003

2004

2005

2006

State Expenditures Federal Expenditures
Sources: (1) 372 Reports (Direct Benefits Expenditures: 1997, 1998 and 2000); (2) DOAA/HIAG (Direct Benefits Expenditures: 2001-2006); (3) DHR Annual Reports (Care Coordination and State Administrative Costs)

Note: DCH was unable to locate the 372 Report for 1999.

As seen in Exhibit 3 above, CCSP expenditures grew substantially over the ten-year period of review. Actual total annual expenditures (including the federal matching dollars) for CCSP grew from $52.1 million in 1997 to $127.3 million in 2006, or 144%. The annual state funds spent grew from $21.3 million in 1997 to $52.6 million in 2006, or 147%.
The costs presented here include total costs for CCSP services, Care Coordination, and state administrative costs. For the breakdown of these costs and direct services, see the discussion of cost-components beginning on page 10.

Special Examination - Community Care Services Program

8

An Analysis of the Expenditures per Slot and an Explanation of the Cost Components
Expenditures per Slot
The average annual cost per unduplicated CCSP recipient has steadily increased during the ten-year period, from $4,140 in 1997 to $10,111 in 2006 (144%). The number of unduplicated recipients served has remained relatively stable, beginning at 12,607 in 1997 and ending at 12,596 in 2006.

As discussed in the previous section, the CCSP does not track the number of slots filled, but instead tracks the number of "active clients," meaning those clients which the AAAs report as currently enrolled. DHR provided differing counts of CCSP clients based on various collection methodologies and annual report figures. In addition, there is a lack of documentation regarding the active clients and client counts found in the CCSP annual reports, there are "active clients" who receive little or no state funded services, and the vast majority of CCSP expenditures occur in the Medicaid claims payment system. Therefore, in Exhibit 4 we used unduplicated recipients as found in the Medicaid claims payment system for years in which the 372 Reports are unavailable. See Appendix A for a listing of each figure.
Exhibit 4 CCSP Unduplicated Recipients
16,000

15,000

14,000

13,000

12,000

11,000

10,000

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

Sources: (1) 372 Reports (1997, 1998 and 2000), DOAA HIAG (2001-2006) Note: DCH staff was unable to provide the 372 Report for 1999.

Special Examination - Community Care Services Program

9

Although the number of unduplicated recipients served in the CCSP waiver fluctuated within the ten-year period, the number served in 2006 (12,596) is similar to the number served in 1997 (12,607). There was an increase between 2000 and 2002, after which the number of CCSP clients served began to decrease. The same stability over the ten-year period is observed using the client numbers reported in the CCSP annual reports. In short, the CCSP program funded services for approximately the same number of people in 2006 as it did ten years ago, despite increases in appropriations, using either metric of clients served.
Using the cost per unduplicated CCSP recipient as found in Medicaid claims records, it is clear from Exhibit 5 below that CCSP average unduplicated recipient costs have steadily increased over the ten-year timeframe, despite the relative stability in the number of persons served.

Exhibit 5 Average Unduplicated CCSP Recipient Cost per Year 1,2,3

Average Cost

$12,000
$10,000
$8,000
$6,000
$4,000
$2,000
$0 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Fiscal Year
Sources: (1)372 Reports (Direct Benefits and Unduplicated Recipients: 1997, 1998 and 2000); (2) DOAA HIAG (Direct Benefits and Unduplicated Recipients: 2001-2006); (3) DHR (Care Coordination and State Admin Costs)
Note: DCH staff was unable to provide the 372 Report for 1999.
As seen in Exhibit 5 above, the average cost for CCSP unduplicated recipients has grown substantially over the ten-year timeframe, increasing from $4,140 in 1997 to $10,111 in 2006 (144%). Using the number of clients served as reported in the CCSP annual reports for the calculation, the per-client costs have increased 138% from 1997-2006. Since many CCSP clients also leave the program due to death, institutionalization and other reasons, the effect of "turnover" may significantly alter the picture of average client costs. Depending

Special Examination - Community Care Services Program

10

on the length of stay and utilization patterns for CCSP clients that leave the program, the use of unduplicated recipients may greatly understate the actual cost for CCSP clients who remain in the program for a full year or utilize more CCSP services. (See also the Efficiency section for further discussion).
According to OPB and program staff, appropriations have historically been requested in terms of additional slots or clients, and both have generally been defined as the average cost to provide services to an additional unduplicated client for the budget time period. However, our analysis of the number of CCSP clients that actually received the average dollar amount of services or more in 2006 showed that 11% of CCSP clients received more than twice the average cost, while 30% received less than half the average cost of services, and up to 15% of CCSP clients receive $2,000 or less of services.

Cost Components
In FY 2006, approximately 63% of CCSP expenditures were attributed to Personal Support Services, which are direct care services that include assisting recipients with eating, bathing, dressing, personal hygiene, and other activities of daily living. Most of the remaining CCSP expenditures were for Care Coordination (17%) and Alternative Living Services (12%). Personal Support Services have accounted for an increasing percentage of CCSP expenditures, growing from 42% in 1997. It is interesting to note that Care Coordination costs have increased from $9.9 million in 1997 to more than $21 million in 2006 (113%), though the program is serving essentially the same number of clients in 2006 as it did in 1997.

The main cost components are described below. For a complete listing of expenditures and percentages by cost components, see Appendix A:
Care Coordination and State Administrative Costs Care coordination costs are associated with the development and management of care plans for CCSP clients and include the administrative costs of the AAAs and/or the companies that contract with AAAs to provide case management. State administrative costs are those attributable to the CCSP unit within DHR's Division of Aging Services.
Alternative Living Services (ALS) ALS provides 24-hour supervision, medically-related personal care, nursing supervision and health-related support services in a state-licensed personal care home.
Personal Support Services (PSS) PSS provide clients with assistance with daily living activities (e.g., housekeeping, assistance with personal care, and relief of persons who normally provide care/oversight to the client).
Other Direct Services These include various health services offered to CCSP clients, including: Home Delivered Meals and Services, Adult Day Health, Emergency Response Services, Respite Care and Nursing.
Using data from the Medicaid claims payment system, the expenditures for CCSP services were analyzed by type of service. As seen in Exhibit 6, PSS, ALS and Care Coordination account for nearly all CCSP expenditures, though their percentage of the total expenditures has changed over time.

Special Examination - Community Care Services Program

11

Total Expenditures

$140,000,000

Exhibit 6 Cost Components in CCSP 1,2,3

$120,000,000

$100,000,000

$80,000,000

$60,000,000

$40,000,000

$20,000,000

$-
1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

Fiscal Year

Personal Support Services Alternative Living Services Other Direct Services

Care Coordination

State Administrative

Sources: (1)372 Reports (Direct Benefits: 1997, 1998 and 2000); (2) DOAA HIAG (Direct Benefits: 2001-2006); (3) DHR (Care Coordination and State Administrative Costs)

As seen in Exhibit 6 above, CCSP expenditures have increased substantially over the review period. It is interesting to note that Care Coordination costs have increased from $9.9 million in 1997 to more than $21 million in 2006 (113%), though the program is serving essentially the same number of clients in 2006 as it did in 1997.
PSS has increasingly accounted for CCSP total expenditures, growing from 42% in 1997 to 63% in 2006. The share of CCSP expenditures spent on ALS has declined, from 16% in 1997 to 12% in 2006. It is interesting to note that ALS is a 24-hour service provided in a state-licensed personal care home, instead of an individual CCSP client's home.

Special Examination - Community Care Services Program

12

A Critique of the Program's Purpose, Goals and Outcome Measures
Purpose
The purpose of the CCSP is largely established in federal statute, and broadly requires that the CCSP serve to decrease or avoid institutionalization and offer the appropriate services to CCSP eligible clients. This purpose is common to waivers in other Southern states. The CCSP purpose is further defined in O.C.G.A. and program materials, but these definitions are inconsistent with each other. It is unclear if CCSP clients must be aged and disabled or aged or disabled. It is also unclear whether an age threshold is appropriate to the CCSP program, but more than 20% of CCSP clients were under age 60 in 2006.
Code of Federal Regulations (CFR) title 42 section 441.303 establishes that the overall purpose of a Home and Community Based Services (HCBS) waiver program is to offer an array of home and community-based services that an individual needs to avoid institutionalization. A review of similar waivers in other Southern states suggests this purpose is common.
The CFR also sets out a number of requirements that must be met in order to receive approval and funding from the federal government, including safeguards for the health and welfare of recipients; financial accountability; evaluation of the needs of the recipients; and assurance that the recipient would qualify for institutionalization absent the waiver. The federal government periodically evaluates the waiver regarding institutionalization avoidance, services offered and the risk of institutionalization.
Various documents that govern the program are inconsistent regarding the CCSP-eligible population. Due to these inconsistencies, it is unclear whether the program is intended to serve both the elderly and the physically disabled, or if it is to serve physically disabled elderly persons only. Further, if the program is intended to serve elderly persons only, there is some inconsistency among documents that govern the program regarding how old recipients should be to qualify for services. The Community Care Services for the Elderly Act (O.C.G.A. 49-6-60) clearly identifies the program's target population as functionallyimpaired elderly persons age 60 and older. Other documents, such as the actual waiver agreement between CMS and DCH, suggest that the eligible population may be limited to 65 and older, but are also unclear as to any precise limitations on age or disability status.
The issue of whether the CCSP is intended to only serve persons 60 and older (or 65), and whether recipients must be physically disabled determines the appropriateness of key measures of the program's purpose and effectiveness. CCSP personnel indicated that the program has no age restriction, and that the intended client population is identified in the waiver agreement as 65 and older and physically disabled. However, an analysis of the active client list as of September 2006 shows that 20.8% of CCSP clients are younger than 60, while 28.0% are younger than 65. In addition, the CCSP annual report shows the percentage of clients who were under age 60 in 2005. Such reporting suggests that the actual client population served is not limited to those 60 (or 65) and older, regardless of such requirements in O.C.G.A. or the waiver agreement.
The CCSP reports the average client length of stay and services offered to CCSP clients in its annual reports. The 372 Reports that demonstrated the number of clients served and services delivered have not been submitted to CMS since 2002.

Special Examination - Community Care Services Program

13

Goals
CCSP regulations and materials identify several goals, but do not establish measurable targets for these goals. Subsequent to the review, DHR provided twelve measures, but it is unclear the extent to which these measures should be considered goals. Further analysis is required to assess the appropriateness and validity of the information provided.
Using the CCSP provider manual, CFR, CCSP annual reports, staff statements, and O.C.G.A., the following goals were identified and confirmed by CCSP program staff during the course of the review:
1. To establish a continuum of care so that functionally-impaired elderly persons age 60 and older may be assured the least restrictive environment suitable to their needs (O.C.G.A.);
2. To provide quality services consistent with the needs of the client (CCSP Provider Manual);
3. Improving the member's independence and safety in the community as long as possible (CCSP Provider Manual);
4. To involve the member (or the member's representative) in the provision and decision making process regarding member care (CCSP Provider Manual);
5. To demonstrate compassion for those served by treating members with dignity and respect while providing quality services (CCSP Provider Manual); and
6. To provide cost-effective services (CCSP Provider Manual and CFR).
Each of these goals fails to provide measurable targets for improvement, such as degrees of restrictive environments to avoid or quantified metrics of cost-effectiveness. Many of the goals identified simply restate the purpose of the CCSP program or requirements made by the federal government. None of the goals identified (besides those that restate the purpose) relate to the purpose of institutional avoidance.
Subsequent to our review, DHR provided 12 measures recording various assessments of program performance. However, it is unclear whether these measures should be considered goals as they do not set specific targets or baselines or record basic elements of program operations such as the number of active recipients. These measures are described in more detail in the following Outcome Measures section and require further review to assess their appropriateness and validity as goals. Appropriate goals would constitute a framework of measurable targets that could be used to assess overall programmatic performance in relation to the purpose of avoiding institutionalization and offering proper services. Such a framework could then be used to chart future improvements to the program.
Additionally, according to DHR, in November of 2006 a CCSP workgroup was created to "roadmap where we are, where we want to go, and determine how we will get there." This work-group was chartered in response to a new initiative by CMS designed to improve oversight of all HCBS waivers in the United States.

Special Examination - Community Care Services Program

14

Outcome Measures
There are four outcomes reported in the CCSP annual report regarding length of stay, quality of services, client discharge status, and cost-effectiveness, only one of which can be used to measure the effectiveness of the program in comparison to its purpose. The length of stay measure clearly relates to the CCSP purpose. The quality measures include: provider assessments, which report on compliance and the effectiveness of services for a small sample of the CCSP population, but not in aggregate for the overall quality of CCSP services; client satisfaction surveys, which may not be generalizable to the CCSP population; and reviews of care plans, which review the completeness of client files and effectiveness of care plans for a small sample of the CCSP population, but not in aggregate for the overall quality of CCSP care plans or services. It is unclear how the client discharge status relates to institutional avoidance and the programs have not evaluated this measure in this respect. Agency staff were unable to explain how the key components of the cost-efficiency measures were calculated for any of the years in our review. Subsequent to the review, DHR provided twelve measures of program performance. Further analysis is required to determine the value and validity of the measures and associated data as related to the proper assessment of CCSP outcomes.

Length of Stay
The CCSP program tracks and reports the average client length of stay. This measure could be used to relate to the purpose of avoiding institutionalization. For example, any changes in length of stay in the CCSP program could be linked to changes in the amount of institutionalization avoided. For discussion of the changes in length of stay in the CCSP, see the discussion of effectiveness beginning on page 21.

Quality of Services
Quality assurances are a requirement set forth in CFR, establishing conditions that must be met in order to receive federal funding. The CCSP program measures the quality of services provided to CCSP clients by
assessing providers; conducting client satisfaction surveys; and reviewing a sample of care plans administered by care coordination staff.
Reviews of DHR programs' providers are conducted by the DHR Office of Regulatory Services (ORS), but are not specific to CCSP providers. These reviews include investigations of complaints, incidents, and compliance with provider licensure requirements, but do not collect data in aggregate which could be used to portray the overall quality of CCSP services. The DHR Division of Aging Services (DAS) also reviews provider compliance, however, only providers of Adult Day Health services have received such compliance reviews. These reviews are largely indirect measures of administrative quality and are also not reported in aggregate for any CCSP services other than ADH.
DAS also conducts client satisfaction surveys; however, they do not compare the CCSP services delivered to any objective measures of quality. The presentation of the results for some of the Alternative Living Services (ALS) surveys combines the "somewhat satisfied"

Special Examination - Community Care Services Program

15

and "satisfied" responses, so that the only outcomes presented are clients who are "very satisfied" or "satisfied." Additionally, some of the surveys have low response rates (less than 5% of PSS clients), and the sample of site visits may not be representative of the larger CCSP population (less than 15% of ALS clients). Finally, the explanation of the sampling methodology used is unclear as to client selection, as well as the effect of adjustments made to the sample in the field. The client satisfaction surveys have only been conducted since 2004.
In general, the results of the most recent client satisfaction surveys appear positive, though 20% of the CCSP clients surveyed indicated that they were only sometimes satisfied with their PSS services. Additionally, 17% of the CCSP clients surveyed stated that they experienced problems with their PSS service in the last 30-60 days. It is also interesting to note that in 2002, CMS reviewed the CCSP program and suggested that utilization limits it observed imposed on PSS services may not be in compliance with the Olmstead decision, which stipulated that clients be given the choice of needed care in the community or the institutional setting. A recently completed review by CMS states that the CCSP met the federally required assurances for the waiver. This review was supported by agencysubmitted information and did not include evaluation of actual client records or onsite visits.
For the CCSP clients living in group homes and receiving Alternative Living Services, nearly all respondents indicated satisfaction with services. However, 2-3% of respondents stated that they are not satisfied with the way staff respond to their needs. CCSP clients receiving Adult Day Health (ADH) services responded with an overwhelmingly high satisfaction level (95-100%) for each item surveyed. Such high client satisfaction rates may be indications of CCSP effectiveness at providing ADH services, but are inconsistent with the results of client satisfaction surveys for PSS.
Finally, DHR and the regional Area Agencies on Aging annually review a sample of 10% of individual care plans and case files to evaluate whether the care plans are being updated in a timely fashion. While these reviews may correlate to timely delivery and quality of services for a small sample of CCSP clients, they are not used to demonstrate the overall quality of the care plans or services in CCSP.
Discharge Status
The CCSP program tracks and reports three reasons clients leave the program, including: death, institutionalization and other reasons (such as entering the hospice program, other waiver programs, loss of eligibility, and moving). It is unclear whether the relationship between discharges due to death and other reasons can be used to evaluate the CCSP program in terms of its purpose or goals. The incidence of discharge related to institutionalization could be evaluated in terms of institutional avoidance once in the program. However, the reporting of discharge status is presented as descriptive information only, and no analysis as to the meaning of the discharge status is performed by the program. A cursory review of this data shows mixed results and is discussed in the Effectiveness section.

Special Examination - Community Care Services Program

16

Cost Efficiency
The CCSP program reports the cost-efficiency of the waiver as compared to the institutional alternative, however, as discussed above, this is a requirement of the waiver. In addition, the agencies have been unable to explain the methodology used to calculate the average institutional cost to which CCSP costs are compared. The 372 Reports submitted to the federal government are intended to demonstrate the cost-efficiency of the waiver, but these reports have not been submitted since 2002. The CCSP annual reports still present cost efficiency figures, but program staff have been unable to explain the methodology used to calculate these figures.
Additional Measures
Subsequent to the review, DHR provided information on 12 measures. Further analysis is required to fully evaluate these measures in terms of comprehensively assessing CCSP performance. Seven of the measures provided appear to be metrics of administrative operations and compliance, and may not necessarily correlate to CCSP outcomes. There may also be fallacies in the development and use of the other measures provided. For example, the "diversion rate" was presented as a measure of CCSP performance in avoiding institutionalization. However, this ratio is calculated by dividing the number of persons who chose CCSP by the number of persons eligible to choose CCSP. While the resulting ratio quantifies the number of CCSP eligibles who chose CCSP, it is instead presented as the number of clients "diverted" to CCSP. The diversion label implies that these clients would have entered an institutional facility, but does not take into account the number of applicants who may never choose to enter a nursing home. The use of such a measure may overstate the amount of savings and effectiveness attributable to the CCSP.

Special Examination - Community Care Services Program

17

General Comments on How Efficient and Effective the Program Has Been Over the Past 10 Years
Efficiency
DHR monitors the efficiency of the CCSP in its annual report, but DCH has not reported measures of efficiency to the federal government since 2002, and neither agency was able to explain or document these calculations from prior years. Alternatively, we attempted to compile information on other possible indicators of efficiency, such as the percentage of total CCSP expenditures attributed to administrative costs. Although we calculated this percentage at roughly 18%, the program lacks established benchmarks for use in determining whether such percentages indicate efficient operating levels. The average cost to coordinate care for each CCSP client has risen from $878 in FY 1997 to $1,780 in FY 2006, an increase of 103%. There is wide variance in the per-unduplicated client cost to coordinate care across the state. Finally, we discovered significant additional Medicaid outpatient costs attributable to CCSP recipients but paid outside of CCSP Medicaid claims.
The review team attempted to analyze the CCSP for efficiency measures, including:
1. Operational efficiency; 2. Comparison to other states; and 3. Comparison to the institutional alternative.

Operational Efficiency
The review team analyzed trends in the program's operational efficiency using four measures: (1) the overall program expenditures for administration and care coordination compared to expenditures in direct benefits; (2) care coordination costs per unduplicated client; (3) the regional differences in cost to coordinate care for CCSP clients; and (4) other Medicaid costs incurred for CCSP clients while in the waiver.
For overall programmatic operational efficiency, care coordination and state administrative costs made up 21% of total program expenditures in FY 1997 (a high for the review period) and 18% in FY 2006, with fluctuations as low as 17% in FY 2003. However, the decline in the percentage of overall CCSP expenditures spent operating the program does not appear to be due to increasing program efficiencies. Instead, total direct benefits expenditures have increased by a larger proportion.
The average cost to coordinate care per each client appears to have grown significantly. DHR staff provided data on the care coordination costs for CCSP clients, as these are reimbursed outside of the claims payment system via contracts at the regional level. Exhibit 7 shows the average cost to coordinate care per unduplicated client for the period FY 1997-2006. The cost has risen from $878 in FY 1997 to $1,780 in FY 2006, an increase of 103%. If the number of clients reported in the CCSP annual reports are used, the perclient care coordination costs increased by 98% over the ten year period.

Special Examination - Community Care Services Program

18

E x p e n d itu re s

Exhibit 7 Average per Unduplicated Client Cost for Care Coordination 1,2,3
$2,000 $1,800 $1,600 $1,400 $1,200 $1,000
$800 $600 $400 $200
$0 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006
Fiscal Year
Source: (1) DHR (Care Coordination Costs: 1997, 1998, 2000-2006); (2) 372 Reports (Unduplicated Recipients: 1997, 1998 and 2000); (3) DOAA/HIAG (Unduplicated Recipients: 2001-2006)
Note: DCH could not provide the 372 Report for 1999.
It is interesting to note that the Care Coordination costs per client have increased significantly over the ten-year timeframe, although approximately the same number of clients were served in 2006 as in 1997.
The ratios of overall CCSP operational efficiency and per-client operational efficiency do not illustrate the inconsistent operational costs across each of the 12 CCSP regions. DHR staff provided data regarding the number of unduplicated recipients in each region as well as the total care coordination dollars by region. Exhibit 8 shows the cost to coordinate care per unduplicated recipient, by AAA region, for FY 2005 and 2006. For a map of the location of each region, see Appendix C.

Special Examination - Community Care Services Program

19

Exhibit 8 Regional Care Coordination Costs Per Unduplicated Recipient (FY 2005--2006)

Area Agency on Aging
SOUTHEAST GEORGIA SOUTHWEST GEORGIA
ATLANTA REGION MIDDLE GEORGIA NORTHWEST GEORGIA CENTRAL SAVANNAH RIVER NORTHEAST GEORGIA HEART OF GEORGIA ALTAMAHA LOWER CHATTAHOOCHEE COASTAL GEORGIA GEORGIA MOUNTAINS SOUTHERN CRESCENT

2005 Costs
$1,775,737 $1,481,411 $3,841,319 $1,337,952 $2,197,420 $1,921,691 $1,186,772 $1,790,652 $1,232,431 $1,548,496 $1,268,144 $1,103,418

2005 Cost per
Unduplicated Recipient $1,106 $1,378 $1,519 $1,346 $1,355 $1,437 $1,487 $1,467 $1,425 $1,595 $1,358 $1,579

Rank
1 5 10 2 3 7 9 8 6 12 4 11

2006 Costs
$1,839,157 $1,262,885 $3,688,381 $1,358,008 $2,403,170 $1,915,455 $1,198,701 $1,833,700 $1,263,249 $1,646,607 $1,457,138 $1,184,519

2006 Cost per
Unduplicated Recipient $1,256 $1,287 $1,488 $1,509 $1,542 $1,592 $1,624 $1,639 $1,656 $1,657 $1,681 $1,765

Rank
1 2 3 4 5 6 7 8 9 10 11 12

Source: DHR AIMS Data Note: Figures reported here are based on care coordination costs and unduplicated recipients reported in the AIMS system, as these are the only regional care coordination data available.

As seen in the Exhibit above, there is wide variance in the per-unduplicated client cost to coordinate care across the state. The Southeast AAA had the lowest care coordination cost per unduplicated client in both FY 2005 and 2006, while the Coastal and Southern Crescent regions had the highest per client cost in 2005 and 2006, respectively. The range between the regional costs per unduplicated client was $489 in 2005 and $509 in 2006. Further exploration is needed to determine what, if any, effect possible differences in client turnover may have on the regional differences observed. For example, if the Atlanta region has a higher number of unduplicated recipients that do not receive a full year of services than other regions, the use of unduplicated recipients would skew the per client costs shown in Exhibit 8.
In addition to CCSP care coordination and administrative costs, there are substantial Medicaid expenditures for services CCSP clients received while on the waiver that are not billed in the CCSP program. These costs, such as those spent on home health services, durable medical equipment, physician services, and other services required by CCSP clients to remain in their community, were incurred to serve individuals while in the CCSP but were not part of the cost analyses on pages 10-11. It is unclear whether the program monitors these costs, since the 372 Reports requiring reporting of these additional costs have not been submitted since 2002. These additional Medicaid costs for CCSP clients, which did not include pharmacy or hospital costs, were substantial, ranging from $5.6 million in FY 2004 to $12.8 million in FY 2002. Six years of data on these additional costs are presented in aggregate in Exhibit 9. It is our opinion that the estimates shown in Exhibit 9 are conservative. This opinion is supported by an independent study conducted in 2004, which shows that additional Medicaid outpatient costs for CCSP clients were more than $22.1 million beyond their waiver costs. This study also shows that pharmacy and inpatient Medicaid costs for CCSP clients were an additional $82 million above their

Special Examination - Community Care Services Program

20

waiver costs.

Exhibit 9 Additional Medicaid Outpatient Costs for CCSP Recipients
(excluding including Pharmacy and Inpatient costs)

FY 2001 2002 2003 2004 2005 2006
Source: DOAA HIAG data

Cost $11,850,330 $12,813,430 $10,946,677 $ 5,555,699 $ 8,280,245 $10,520,556

Comparison to Other States
Efficiency comparisons to other states are currently impossible due to a lack of comparable data. DCH has not compiled and submitted to the federal government the 372 Reports that show waiver utilization and costs since 2002. Moreover, other states also have great flexibility in key programmatic determinants of cost, such as coverage and eligibility that make cross-state comparisons of limited applicability.

Comparison to the Institutional Alternative
Professional literature suggests that the methodological challenges in comparing CCSP costs to the institutional alternative are too complex to permit useful comparison. Regardless of this point, however, accurate efficiency comparisons to the institutional alternative are currently precluded as:
1. The 372 Reports showing this comparison have not been submitted since 2002; 2. For the years in which the relevant reports were submitted, DCH has not been
able to support or explain the methodology used to calculate the figures reported; 3. For the years in which actual claims payment data is available for comparison to expenditures reported to the federal government on 372 Reports (2001 and 2002), the numbers reported on 372 Reports are inconsistent with the claims payment data; and 4. DCH has reported varying figures for the cost comparison.
The efficiency of the waiver as compared to the institution is considered a requirement of the program by the federal government. According to the CFR Title 42 section 441.300, the annual average per capita expenditure estimate of the cost of home and community-based and other Medicaid services under the waiver must not exceed the estimated annual average per capita expenditures of the cost of services in the absence of a waiver. This requirement is further defined in the CCSP waiver document between DCH and CMS which states that Georgia will refuse to offer home and community-based services to any person for whom it can reasonably be expected that the cost of

Special Examination - Community Care Services Program

21

[waiver] services would exceed the cost of a nursing facility level of care.
Reviews of the literature regarding the cost comparisons of aging and physical disability waiver programs have found that developing a comprehensive methodology to identify and examine the multitude of costs associated with recipients of institutional care versus the costs of all the services received by recipients in their homes and communities has proven challenging and perhaps inappropriate. For instance, the case mix of individuals capable of being served in their homes may not correlate to the needs of individuals served in institutional settings. Additionally, individuals who receive care in their homes may also receive other services funded through Medicaid and local sources such as community support programs and acute medical care.
Despite concerns over the validity of comparing the reported average costs of institutional care to CCSP average client costs, a basic comparison of actual CCSP client costs to the reported figures is telling. In FY 2005, DCH reported the average nursing facility client cost as $23,106 in their annual report. The review team conducted an analysis of FY 2005 client benefit claims data to see if any CCSP clients received benefits at amounts greater than the DCH reported nursing facility cost. Our analysis showed that at least 180 clients (1.4%) received direct benefits totaling more than the DCH reported average nursing facility cost. In total, these clients incurred $4.7 million in CCSP waiver direct benefits. Approximately $550,000 spent on these CCSP clients was in excess of the DCH reported nursing facility cost. This analysis did not consider the costs associated with care coordination, state administration, or other non-waiver Medicaid costs, and thus probably underestimates the number of clients whose cost exceeded the nursing facility cost reported in DCH's annual report.
Finally, the current cost-monitoring structure may prevent the CCSP program from operating efficiently compared to the institutional alternative. The program reports that the 12 local Area Agencies on Aging are responsible for monitoring individual client costs. However, results of a survey of the AAAs indicate that some monitor client costs at a threshold that is too high to effectively prevent clients from receiving CCSP services at levels above the DCH reported nursing facility cost. Of the 11 that responded to our survey, 8 monitor client cost at $24,000 or higher, some as high as $36,000. The AAAs indicated that personnel from DCH and/or DHR have stated that the annual cost of nursing facility care ranges from $38,000-$43,000, a figure that is much higher than the one reported in the DCH annual report ($23,106, as mentioned in the previous paragraph). It should be noted that DCH personnel could not explain how nursing facility cost estimates were determined, and the program does not have a policy that prescribes the threshold against which the AAAs should monitor individual client costs.

Effectiveness
It is unclear the extent to which the program has been effective in avoiding institutionalization or offering the needed services to CCSP clients. Without an appropriate and reliable framework of goals and outcomes, the effectiveness of the CCSP cannot be comprehensively assessed. However, we attempted to assess program effectiveness using the information available regarding length of stay, discharge status and the waiting list. The reported results regarding length of stay and client discharge status independent of other comparative data do not allow analysis of any resulting appropriate institutional avoidance. Finally, review of the

Special Examination - Community Care Services Program

22

management of the 12 regional waiting lists suggests that the CCSP program may not be effective in placing CCSP clients in a manner that is based solely on actual client needs. There are regional disparities regarding which clients are given priority on the CCSP waiting lists. To some degree, whether an applicant receives services is related as much to where he or she lives as to their level of need for services.

Months Receiving Services

Effectiveness at Avoiding Institutionalization The CCSP annual reports show the average length of time (in months) that clients receive services. Exhibit 10 shows that the average length of time that clients have received CCSP services has risen during the period from 33 months in FY 1997 to 45 months in FY 2005, an increase of one year.
Exhibit 10 Length of Time that Clients Receive CCSP Services (on Average)
60
50
40
30
20
10
0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Fiscal Year
Source: CCSP Annual Reports
Increasing the time that clients receive services may be viewed as a proxy measure of institutional avoidance. However, these figures may be overstated as some literature suggests that people are generally more willing to accept home and community based services than they are to enter a nursing facility. Therefore, to presume that clients receiving CCSP services would have otherwise immediately entered a nursing facility (even though they are eligible) may overstate the time that clients avoid an institution. In short, it is unclear the degree to which the trends regarding length of stay reflect CCSP effectiveness.

Special Examination - Community Care Services Program

23

The program also tracks the discharge position of clients that stop receiving CCSP services. Specifically, the program tracks the percentage of clients that leave the waiver to enter a nursing facility, the percentage that die while in the program, and the percentage that leave for other reasons. Exhibit 11 shows the annual discharge position of clients who left the program for the period FY 1996-2005.

Exhibit 11 Client Discharge Position

100

% o f D isch arg ed C lien ts

80

60

40

20

0

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

Fiscal Year

Death Nursing Facility Other

Source: CCSP Annual Reports
These reported figures indicate that the percentage of discharged clients that have left the CCSP to enter a nursing facility has dropped from a high of 39% in FY 1996 to 30% in FY 2005. While this may suggest effectiveness in terms of decreasing institutionalization, review of the increasing percentage of clients that leave the program due to death suggests that the trend may be more complex. The percentage of CCSP clients who have left the CCSP due to death has risen from 38% to 44% during this period. It is not clear what these results mean, but they may suggest that the program is becoming more effective at avoiding institutionalization in that more clients are staying at home or in their communities in their final years as opposed to a nursing home. In sum, it is unclear what the trends in discharge status reflect regarding CCSP effectiveness.

Effectiveness in Terms of Management of the CCSP Waiting List
Review of the management of the CCSP waiting list could also be used to assess the effectiveness of the CCSP both in terms of avoiding institutionalization and providing needed services to CCSP eligible clients. During the period FY 2001-FY 2006, there have been significant fluctuations in the waiting list. For example, the waiting list reached a

Special Examination - Community Care Services Program

24

high of 5,560 at the end of FY 2003, but dropped significantly to 1,919 in FY 2005. As of September 2006, there were 2,257 applicants on the CCSP waiting list, an increase of approximately 338 from the previous fiscal year. During the six-year period, the waiting list has decreased by 941 applicants.
One might expect the persons served in the CCSP program to be directly and inversely related to changes in the waiting list. For example, when the waiting list decreases it is because the number of clients served increased, and vice versa. However, as seen in Exhibit 12, there appears to be no consistent correlation between the waiting list and the number of clients served. The trends observed in 2001, 2003, 2005, and 2006 support the supposition that CCSP program activities are directly and inversely linked to the waiting list. However, other years, such as 2002 and 2004 do not suggest such a relationship. Using the client figures as reported in the CCSP annual report also demonstrates a lack of consistent correlation between the changes in clients served and changes in the waiting list. CCSP staff have suggested that the waiting list may not be an effective tool in assessing the actual population in need of CCSP services, and have obtained the services of a CDC epidemiologist to conduct prevalence studies that would more accurately quantify potential CCSP clients who have not registered on the waiting list but are in need of CCSP services. In short, it appears that the assumption that increased CCSP activity will lead to a reduction in the waiting list may be inaccurate.

Exhibit 12 Unduplicated Clients, "Active Clients" and Applicants on the Waiting List

FY
2001 2002 2003 2004 2005 2006

Unduplicated Recipients1,2

Total Number

Annual Change

14,670

952

14,708

38

13,213

(1,495)

12,380

(833)

13,774

1,394

12,596

(1,178)

Waiting List3

Total Number
3,198 4,115 5,560 5,018 1,919 2,257

Annual Change
(237) 917 1,445 (542) (3,099) 338

Sources: (1) 372 Reports (Unduplicated Recipients: 2000); (2) DOAA HIAG (Unduplicated Recipients: 2001-2006); (3) DHR (Waiting List: 2000-2006)
Note: The 2006 waiting list figures are as of September. Waiting list data for prior years are from annual reports.

Beyond the issues in whether CCSP activities have an effect on the waiting list, there are other issues regarding management of the waiting list itself. It is important to note that there is not one centrally managed waiting list for the CCSP. Instead, there are 12 separate waiting lists managed at the regional level by the AAAs. As such, each waiting list has a distinct subpopulation of CCSP applicants competing for limited regional funds.
We reviewed the management of the regional waiting lists in terms of the days applicants were waiting for CCSP services in each region. The average wait time, as well as the

Special Examination - Community Care Services Program

25

number of clients served and applicants on the waiting list is shown by region in Exhibit 13 below.
Exhibit 13 Persons Waiting for CCSP Services Compared to Active Clients,
by Region

Region

Applicants Active Current Avg. # of Waiting Clients Days on Wait List

SOUTHERN CRESCENT

135

569

337

NORTHEAST GEORGIA

248

711

331

SOUTHWEST GEORGIA

166

817

291

MIDDLE GEORGIA

185

762

269

CENTRAL SAVANNAH RIVER

178

1,069

265

HEART OF GEORGIA ALTAMAHA

217

1,235

239

GEORGIA MOUNTAINS

248

893

237

ATLANTA REGION

617

2,442

233

SOUTHEAST GEORGIA

152

1,517

144

NORTHWEST GEORGIA

85

1,393

62

COASTAL GEORGIA

22

919

50

LOWER CHATTAHOOCHEE

4

597

50

Source: DHR Data (CHAT system as of September 2006)

As seen in Exhibit 13, some regions may be more effective at moving applicants more quickly from their waiting list into the CCSP than others. For example, potential CCSP clients in the Southern Crescent and Northeast regions appear to wait much longer before getting into the program than do their counterparts in other regions. Conversely, in the Lower Chattahoochee region, which has a very low number of waiting applicants, the average wait time was 50 days. It should be noted that these days are the number of days on average that eligible applicants on the waiting list have been currently waiting as of September 2006. There are many factors that may contribute to this regional variance, such as higher caseload, more or less effective enrollment procedures at one region than another, and differences in the CCSP eligible population.
Finally, CCSP effectiveness can be measured in terms of placing the applicants with the most need in the program. CCSP policy requires that, in instances when budget constraints prevent new client admissions, the AAAs use need scores to prioritize regional waiting lists until funds become available. CCSP policy states that applicants with the highest need scores from within their respective AAA region are to be the first referred for admission into the program once funds become available for the respective AAA region.
Using the CCSP waiting list as of September 2006, the review team analyzed how the actual prioritization of applicants on the 12 regional waiting lists compared to a hypothetical prioritization of all applicants on a combined single statewide list. Exhibit 14 shows results of the analysis for the highest need score applicant in each AAA region. For each applicant, their corresponding statewide ranking is also shown.

Special Examination - Community Care Services Program

26

Exhibit 14 Comparison of Regional Highest Need Score Applicant vs. Statewide
Waiting List Need Prioritization

Highest Need Score Applicant in Region 1
NORTHWEST GEORGIA ATLANTA REGION
GEORGIA MOUNTAINS SOUTHWEST GEORGIA SOUTHEAST GEORGIA CENTRAL SAVANNAH RIVER HEART OF GEORGIA ALTAMAHA
MIDDLE GEORGIA COASTAL GEORGIA SOUTHERN CRESCENT NORTHEAST GEORGIA LOWER CHATTAHOOCHEE

State Rank 2
1 2 3 6 15 25 41 56 74 76 139 280

Sources: (1) AAA Waiting Lists (Highest ranked applicants in regions: September 20, 2006); (2) DOAA Analysis (Highest ranked applicants in state)
Note: State rank is based on the same Determination of Need (DON-R) data used by the CCSP to assess applicants and by the AAAs to rank them regionally. These scores are derived from self-reported information via a telephone survey that is standardized and used statewide.

This "regional ranking vs. statewide ranking" comparison shows that applicants who rank very high when compared to applicants within their own AAA region may not rank high when compared to applicants statewide. As seen in Exhibit 14, the highest need score applicants in some regions also are the highest need score applicants in the state, such as the Northwest, Atlanta, and Georgia Mountains regions. However, some applicants are prioritized in the region well above their statewide need score, such as the applicants in the Lower Chattahoochee and Northeast regions.
Only the highest need score clients in each region are shown in Exhibit 14, but the variance between AAA rank and statewide rank becomes much wider further down the regional AAA waiting list. For example, the Lower Chattahoochee and Coastal regions have the two shortest waiting lists, and both regions' top five applicants rank very low when compared to the state's entire CCSP waiting list population. In one instance an applicant is ranked 2,165th out of 2,257 total applicants in the state, but is prioritized as fourth on the list in his/her respective region.
These results suggest that the program's current method of regionally prioritizing applicants may permit lower-need applicants access to CCSP services while more needy applicants continue to wait. Thus, to some degree, whether an applicant receives services is related as much to where he or she lives as to their level of need for services. In sum, the CCSP may not be effectively reducing the waiting list equally among the regions or based on the actual needs of all applicants.

Special Examination - Community Care Services Program

27

It should be noted, however, that applicants with extremely high need scores (60+) had not been on the waiting list for more than one year. Exhibit 15 shows the results of an analysis of waiting list applicants as of September 2006 grouped according to their need score and the time that they have already spent on their respective regional waiting lists.

Exhibit 15 Waiting List Applicants: Need Scores and Waiting Times

Needs-Score

Time on Waiting List

10-19 20-29 30-39 40-49 50-59 60-69 70 or More

> 1 Month 1 25 68 84 43 10 1

1-2 Months 0 26 89 79 31 7 1

3-6 Months 4 71
198 170 54 11
3

6-12 Months 6
122 342 265 80
8 0

12-24 Months 1 69
152 115 44
0 0

24-36 Months 0 8 10 10 2 0 0

> 36 Months 0 12 18 12 5 0 0

Source: AAA Waiting Lists (September 20, 2006) Note: A need score of 15 is required to qualify for CCSP services.

Although applicants with the highest need scores do not appear to wait longer than one year, this analysis does suggest that a "bottle-necking" of applicants with relatively high need scores (30-59) occurs. In fact, 368 (16%) applicants with need scores equal to 30 or more (twice the threshold score to qualify for services) have been on the waiting list more than one year. Further exploration is needed to determine if the high need applicants are actually moved into the CCSP program, or if they exit the waiting list for other reasons. Such analysis would be difficult, as the program does not maintain a record of the previous need status of clients once they enter the program.

Special Examination - Community Care Services Program

28

Appendices
Appendix A - Various Source Data ........................................ . . . . p. 29 Appendix B - Sample CCSP Appropriations Language . . . . . . . . . . . . . p. 31 Appendix C Map of Area Agencies on Aging Regions........................p. 35

Special Examination--Community Care Services Program

29

Appendix A

DIRECT BENEFIT EXPENDITURES

Source
DCH 372 Reports to CMS
DCH Annual Report DCH Expenditure Analysis (8/10/06)
DCH "Long Sheet Budget"
CCSP Annual Report DHR's Aging Information
System Data DOAA Data

1997 $ 41,118,978 41,553,886
43,575,243

1998 $48,004,650 50,622,108
53,067,456

1999 54,078,796
61,616,111

2000 $64,013,750 63,047,498
64,394,696 71,289,084

2001 $74,782,339 69,560,855
70,073,064 78,200,923
68,884,478

2002 $87,386,454 84,094,843
84,759,424 84,738,390 83,431,704 84,396,716

2003
88,801,642 87,976,416 91,344,093 85,717,353 72,510,990 94,673,955

2004
84,451,067 89,620,587 85,520,070 88,866,386 73,788,911 84,280,557

2005
97,697,732 96,918,083 95,602,910 93,956,571 87,000,168 99,194,107

2006
100,064,365 110,861,849 104,937,702

CARE COORDINATION EXPENDITURES

Source
CCSP Annual Report DHR's Aging Information
System Data

1997 $ 9,869,415

1998 $10,828,851

1999 $12,126,890

STATE ADMINISTRATIVE EXPENDITURES

2000 $13,734,959
9,392,542

2001 $16,471,682 13,240,231

2002 $17,806,115 14,529,956

2003 $17,806,114 17,768,064

2004 $18,174,150 18,072,404

2005 $ 20,700,359 20,685,443

2006 21,050,970

Source CCSP Annual Report

1997 $1,203,657

TOTAL EXPENDITURES

1998 $672,571

1999 $698,969

2000 $722,857

2001 $1,242,719

2002 $1,403,005

2003 $1,035,747

2004 $1,336,010

2005 $1,371,880

2006

Source CCSP Annual Report

1997 $54,648,315

1998 $64,568,878

1999 $74,441,970

2000 $85,746,900

2001 $95,915,324

2002 $103,947,510

2003 $104,559,214

2004 $108,376,546

2005 $116,028,810

2006 -

UNDUPLICATED RECIPIENTS

Source DCH 372 Reports to CMS DCH Annual Report CCSP Annual Report DOAA Data Waiver Document (estimate) CHAT Data (active clients)

1997 12,607 12,413 14,185

1998 12,502 13,497 14,194

1999
13,815 15,228

2000 13,718 12,274 14,848

2001 15,574 14,943 16,873 14,670

2002 15,418 14,930 16,653 14,708
12,578

2003
13,514 14,687 13,213 15,000 12,292

2004
12,581 14,099 12,380 15,500 10,566

2005
13,980 15,830 13,774 16,000 11,485

2006
14,534 12,596 17,000 11,707

COST COMPONENTS

$ in Millions / (% of Total CCSP Expenditures Billed in the Waiver)

Category

1997

1998

1999

2000

2001

2002

2003

PSS

$21.65 (41.5%) $26.43 (44.4%)

$45.04 (57.4%) $47.92 (55.3%) $59.47 (57.4%) $68.86 (60.7%)

ALS

$8.27 (15.8%) $9.27 (15.6%)

$13.31 (17.0%) $14.82 (17.1%) $17.22 (16.6%) $16.91 (14.9%)

HDM

$0.89 (1.7%)

$1.24 (2.1%)

$1.64 (2.1%)

$1.76 (2.0%) $2.78 (2.7%) $3.67 (3.2%)

ADH

$1.91 (3.7%)

$2.09 (3.5%)

$1.25 (1.6%)

$2.55 (2.9%) $2.96 (2.9%) $3.24 (2.9%)

ERS

$1.20 (2.3%)

$1.31 (2.2%)

$1.57 (2.0%)

$1.76 (2.0%) $1.91 (1.8%) $1.87 (1.6%)

Respite

$4.23 (8.1%)

$4.97 (8.4%)

$1.10 (1.4%)

-

-

$0.03 (0%)

Nursing

-

-

$0.01 (0%)

-

-

-

HDS Medicaid Home Health Care Coordination State Administrative

$0.13 (0.3%) $2.84 (5.4%) $9.87 (18.9%) $1.20 (2.3%)

$0.11 (0.2%) $2.58 (4.3%) $10.83 (18.2%) $0.67 (1.1%)

$0.09 (0.1%) -
$13.73 (17.5%) $0.72 (0.9%)

$0.07 (0.1%) $0.06 (0.1%) $0.08 (0.1%)

-

-

-

$16.47 (19%) $17.81 (17.2%) $17.81 (15.7%)

$1.24 (1.4%) $1.40 (1.4%) $1.04 (0.9%)

TOTALS

$52,192,050

$59,506,072

$78,471,566 $86,598,879

Sources: 372 Reports: (1996, 1997,1998 and 2000); DOAA HIAG (2001-2006)

$103,605,832

$113,515,816

2004 $62.22 (60.0%) $14.15 (13.6%)
$3.34 (3.2%) $2.80 (2.7%) $1.67 (1.6%) $0.03 (0%) $0.03 (0%) $0.03 (0%)
-
$18.07 (17.4%)
$1.34 (1.3%) $103,688,971

2005 $74.42 (61.4%) $15.32 (12.6%)
$3.88 (3.2%) $3.61 (3.0%) $1.80 (1.5%) $0.06 (0.1%) $0.05 (0%) $0.04 (0%)
-
$20.69 (17.1%)
$1.37 (1.1%) $121,251,431

2006 $80.20 (63.0%) $14.90 (11.7%)
$4.07 (3.2%) $3.88 (3.0%) $1.69 (1.3%) $0.09 (0.1%) $0.07 (0.1%) $0.04 (0.1%)
-
$21.05 (16.5%)
$1.37 (1.1%) $127,360,552

Special Examination--Community Care Services Program

31

Appendix B

Below is an analysis of CCSP appropriations for "slot" enhancements and provider rate increases. It was produced by reviewing analysis provided by OPB, BIB documents, and tracking documents.

FY

OPB Interpretation of Language

1997

N/A

N/A

Language in BIB
N/A N/A

Language in Tracking Document
To add 461 slots under the Community Care for the Elderly Program
AMENDED: To provide for the transfer of funds for the Division of Mental Health for additional slots under the CCSP.

1998

Adds $3,896,400 for additional 2,061 clients to the CCSP

use $11,890,934 in Indigent Care Trust Funds for 2,061 additional slots under the Community Care for the Elderly Program and for community based services to 932 low income elderly who are not Medicaid eligible

To add $3,896,400 in Indigent Care Trust Funds for 2,061 additional slots under the Community Care for the Elderly Program and for community based services to 932 low income
elderly who are not Medicaid eligible

1999

Expand services to 2,115 clients in the Community Care Services Program

add funds to increase the number of slots in the community care for the elderly program

To add funds for additional slots in the community care for the elderly program

(FY99 supplemental budget added an additional 50 clients $25,150 in state funds)

Expand services to 125 clients in the CCSP

To expand services to 125 clients in the CCSP

2000

Expand services to 375 clients in the CCSP

Add funds to expend services to 375 clients in the CCSP

To add funds to expand services to 375 clients in the CCSP (includes annualizing the 125 clients listed above.)

10% rate increase for family and group model alternative living service providers ($680,000 state; $1,720,212 total)

Provides a 10% rate increase for Alternative Care providers in the CCSP. ($680,000)

To provide a 10% rate increase for Alternative Care providers in the CCSP. ($680,000)

FY

OPB Interpretation of Language

2001

Services for another 2,183 clients in CCSP.

Language in BIB
Add services for another 2,183 clients in CCSP

Language in Tracking Document
To expand services for another 2183 clients in the CCSP for Medicaid eligible elderly

2002

Additional 2,000 clients who are Medicaid eligible. {NOTE: In the FY 2002 supplemental budget, this enhancement was cut by 1,000 slots and $4,174,655 state funds; $9,268,155 total funds. The FY 2003 appropriation
(below) restored 3/4 of this cut.}

Provide Home and community based services to an additional 2,000 elderly clients who are Medicaid
eligible in the CCSP. AMENDED - Reduce funds for HCBS Program, resulting in funding 1,500 of the 2,000 new slots originally funded in FY 2002. AMENDED Reduces funding to the CCSP and continues to phase
in 1,000 of the 2,000 new slots funded in FY 2002.

To provide Home and community based services to an additional 2,000 elderly clients who are Medicaid eligible in the
CCSP. AMENDED - To reduce funds for HCBS Program, resulting in funding 1,500 of the 2,000 new slots originally funded in FY 2002. AMENDED - To reduce funding to the CCSP and continues to phase in 1,000 of the 2,000 new slots
funded in FY 2002.

4% rate increase for CCSP Providers. ($1,294,597 state; $3,187,092 total)

Funds a 4% rate increase for CCSP Providers. ($1,294,597 state; $3,187,092 total)

To fund a 4% rate increase for providers in the CCSP. ($1,294,597 state; $3,187,092 total)

2003

Partially "restore" FY02 enhancement by phasing in 822 slots instead of 2,000 slots.

Partially restore FY 02 enhancement funding by phasing in 822 new slots for the CCSP

To partially restore FY 02 enhancement funding by phasing in 822 new slots for the CCSP

2004

Provide services to 84 nursing home clients in the CCSP
Adjust for the increased cost of the provision of services in the CCSP.

Provides CCSP services to 84 NH residents transferring into the community
Adjust for the increased cost of the provision of services in the CCSP.

To provide CCSP services to 84 NH residents transferring into the community
To adjust for the increased cost of the provision of services in the CCSP.

2005

Adjust for the increased cost of the provision of services in the CCSP.

Adjust for the increased cost of the provision of services in the CCSP.

To adjust for the increased cost of the provision of services in the CCSP.

Adds funds for an additional 460 slots in the CCSP.

Provide funds for an additional 460 slots in the CCSP

To provide funds for an additional 460 slots in the CCSP

2006

Transfer funds from DCH to CCSP in DHR for 79 individuals moving to the community.
Fund 200 slots for CCSP

Transfer funds from DCH to CCSP in DHR for 79 individuals that desire to move from the NH into the
community.

To transfer funds from DCH to CCSP in DHR for 79 individuals that desire to move from the NH into the community.

Fund an additional 200 slots for elderly clients on the To fund an additional 200 slots for elderly clients on the waiting

waiting list that are Medicaid eligible and meet NH level

list that are Medicaid eligible and meet NH level of care

of care requirements for the CCSP

requirements for the CCSP

Special Examination--Community Care Services Program

33

FY

OPB Interpretation of Language

Language in BIB

Language in Tracking Document

To provide a rate increase for Alternative Living Services in Provide a rate increase for Alternative Living Services

CCSP. ($691,900 state; $691,900 total)

in CCSP. ($691,900 )

To provide a rate increase for Alternative Living Services. ($691,900 state; $691,900 total)

2007

Fund 1,000 CCSP slots

Provide funds for an additional 1,000 slots in the CCSP.

To provide funds for an additional 1,000 slots in the CCSP.

Appendix C
The 12 Planning and Service Areas of the CCSP are shown below.

2 1

3

5

4

7

8

6 9 12

10 11

Legend 1. Northwest Georgia AAA 2. Georgia Mountains AAA 3. Atlanta Region AAA 4. Southern Crescent AAA 5. Northeast Georgia AAA 6. Lower Chattahoochee AAA 7. Middle Georgia AAA 8. Central Savannah River AAA 9. Heart of Georgia Altamaha AAA 10. Southwest Georgia AAA 11. Southeast Georgia AAA 12. Coastal Georgia AAA

This report was performed by the Program Audits and Management Support Section of the Healthcare Audits Division. We assess the state's healthcare systems and programs to improve operations and reach management solutions. Our reviews focus on healthcare expenditures, best practices, and data analysis, and are requested by agency commissioners, the governor's office, and the state legislature.
For additional information or for copies of this report call 404-656-9148 or see our website:
http://www.audits.state.ga.us/internet/searchRpts.html
Select Healthcare from the drop-down menu