Chapter 4
Utilization of Inpatient Beds

In Chapter 1, a future vision of the public mental health system was presented. This vision is based on a commitment to quality care and use of evidence-based services, an emphasis on providing services in the most integrated, community-based settings possible, and an operating policy to foster community integration for individuals in educational, employment, social, and recreational activities in their own neighborhoods. It is predicated on the public policy goal of maintaining the locus and continuity of care for individuals within their home communities. OMH's commitment to a new, statewide collaborative planning process that maximizes local input was described in Chapter 2. Chapter 3 provided an overview of the public mental health system that depicts who uses services across all regions, populations, and program types.

Relying on this framework, Chapter 4 presents the first step toward beginning a series of statewide policy-development conversations concerning the utilization of inpatient services within a recovery-oriented continuum of treatment, rehabilitative, and supportive services. The inpatient utilization issue is an appropriate starting point for these discussions for several reasons:

• Research indicates that evidence-based practices can decrease inpatient utilization and increase community tenure. Therefore, greater access to evidence-based practices has implications for future inpatient service capacity planning.
• Inpatient services represent the most expensive, restrictive, and intrusive level of care. The need for this level of care indicates a high level of distress and disruption to an individual's personal, family, and community roles. In addition, the long-term use of inpatient care must be considered with regard to the 1997 U.S. Supreme Court decision in Olmstead vs. L.C., which held that persons with mental disabilities have a right under the Americans with Disabilities Act to receive services in an integrated community setting when appropriate.
• For some time, the mental health community has maintained a position that development of supportive and preventive outpatient services can diminish the need for inpatient care. Commitment to a quality improvement agenda necessitates a careful examination of the degree to which this assumption is accurate, and is particularly relevant during an era of substantial growth in community-based treatment and support programs and housing opportunities.
• Stakeholder interest in public policy decisions regarding the appropriate capacity, location, and utilization of inpatient beds is very high. There is a broad spectrum of opinions on these issues and no commonly shared consensus.
• Inpatient capacity and utilization are correlated in geographic areas where there is State-operated capacity.

This Chapter first describes the historic and current utilization of inpatient beds within the context of the national perspective. It then presents a framework for further analysis using OMH's performance measurement system and some comparative examples of utilization, cost, and readmission outcomes. The Chapter concludes with observations about the data presented and their implications. This content represents a significant departure from previous plans, in that the data are made available for all operating auspices, with detailed county-level data presented in Appendix 5.

Comparing New York to the National Experience

Over the last 15 years, trends in inpatient psychiatric care in New York State have mirrored those of the nation: the steady decrease in the utilization of inpatient psychiatric care across the United States has resulted in closing and reorganizing state psychiatric hospitals. In August 2000, the National Association of State Mental Health Program Directors' Research Institute (NRI) published their findings that:

• In the United States, more state psychiatric hospitals were closed in the 1990s than in the 1970s and 1980s combined
• During the 1990s, 50% of states were reorganizing their state hospital systems
• From Fiscal Years 1993 to 1997, the number of patients in state-operated psychiatric hospitals decreased by 25%, while state psychiatric hospital expenditures decreased by only 4%
• In Fiscal Year 1997 for the nation as a whole, state mental health agency controlled expenditures for community mental health services exceeded state-operated inpatient services by over $2.5 billion

Figure 4-1 Mental Health Inpatient Days in State and General Hospitals per 1,000 General Population: 2001-2002

As shown in Figure 4-1, despite New York State's community integration efforts over the past years, the use of inpatient beds within both State and local sectors continues to be above that of other mid-Atlantic states. Although national discussions generally describe inpatient usage within both state-operated and general hospital sectors, of particular concern to state mental health directors is the utilization of state-operated inpatient beds, because this commitment of resources has historically been excluded from Federal cost sharing through Medicaid reimbursement.

Figure 4-2 Inpatient Care for Adults in New York State (2003 data) vs All US States (1997 Data) per 100,000 Adult Population by Hospital Type

The NRI 2000 study had a particular focus on state psychiatric center experiences and summarizes state psychiatric hospital closures by state. It noted that Pennsylvania had 21 state psychiatric hospitals in 1955 and nine in 2000. California had 10 in 1958 and four in 2000. Illinois had 12 facilities in 1993 and four in 2000. Ohio had 25 in 1956 and had five located at nine sites in 2000. These and similar findings demonstrate that across the United States, the provision of psychiatric care has moved from state-operated institutions to community settings. States can no longer afford to operate and maintain costly inpatient facilities that are underutilized.

Figure 4-3: Inpatient Care for Children Under 18 Years Old in New York State (2003 data) vs All US States (1997 Data) per 100,000 Population under 18 by Hospital Type

Although most of the current interest at both national and state levels concerns the utilization and cost of state-operated inpatient psychiatric facilities, there is increasing attention at the national level to examination of the overall utilization of inpatient care across sectors. The study, Mental Health, United States, 2000, describes the numbers of inpatients under care nationally using adult and child age groupings and examining usage by the auspice of care.¹ Using State and local information systems, New York is able to compare these data with 2003 data on bed usage. In Figures 4-2 and 4-3, the numbers of adult and child psychiatric inpatients are measured by rates of 100,000 within the civilian population and displayed by auspice of inpatient program.

These graphs show that for inpatients under care, New York had more psychiatric inpatients per capita in 2003 than the nation as a whole had in 1997. In 2003, New York had 76.5 adult psychiatric inpatients per 100,000 adult civilian population as compared to a 1997 national rate of 45.6, and 28 child/adolescent inpatients per 100,000 civilian population under 18 years old as compared to a 1997 national rate of 15.7. In 2003, New York's general hospital usage for adults was nearly three times that of the United States in 1997 (38 to 13.6). New York's 2003 State hospital usage for adults also exceeded the 1997 national usage, however, this sector's difference (33.1 to 26.2) is not as dramatic as that of general hospital usage for adults. For children, however, New York's 2003 State hospital usage far exceeded the 1997 United States' rate (10.7 to 3.8), and the State's 2003 general hospital usage was also dramatically higher than the 1997 national rate (9.6 to 3.3). Actual data for these graphs are included in Appendix 5.

Figure 4-4 Average Daily Inpatient Census of New York State Psychiatric Centers vs Article 28 Hospitals*

Since 1995, New York State's bed usage in the State Psychiatric Centers and in general hospital units has been declining steadily. Figure 4-4 shows this decline, and Appendix 5 contains the average daily census data used to develop this graph. From 1995 to 2001, the decline in State Psychiatric Center census accounted for most of the combined census drop. State census declined by almost 4,000 (43%) from 9,550 to 5,557.

Figure 4-5: Reduction in Census at General Hospitals Between 1995 and 2001 by OMH Region

Average daily usage in general hospitals (Article 28) declined by 9% from 4,527 to 4,118. As described in Figure 4-5, the decline in census for general hospitals, while modest, has significant regional variation.

All regions showed decline in general hospital usage from 1995 to 2001. The Western region demonstrated the largest decline at 24%, followed by the other regions. Overall, the decline in bed usage from 1995 to 2001 in New York State has been substantial. On a typical day, the numbers of people using either State Psychiatric Center or general hospital inpatient beds declined during this period by 31%, from about 14,100 to just under 9,700. Despite these census reduction efforts, New York's use of state-operated beds continues to require further examination because it is higher than the total per capita bed usage in the United States and also higher than the per capita usage of state-operated beds in states with similar demographic features.

Figure 4-6 New York State Psychiatric Center Inpatients per 100,000 General Population Compared with Other Urban States and Large Urban Counties in New York*

New York's State-operated inpatient bed usage has been examined using nationally available data that reflect the unique population characteristics of the State. Since New York is a particularly dramatic mix of rural and urban population centers, this analysis compares State inpatient bed usage in two ways. First, the State's usage is compared to that of other states with similar demographic features. Because the comparative data sets available do not display information at the county level, Figures 4-6 and 4-7 show similarly urban states against New York State counties with high concentrations of urban population bases above 500,000. As can be seen in Figure 4-6, New York's bed usage per capita is higher than other, similar urbanized states such as California, Michigan, Texas, and Ohio.

Figure 4-7 New York State Psychiatric Center Inpatients per 100,000 General Population Compared with Other Rural States and Rural Counties in New York*

Second, as shown in Figure 4-7, New York State inpatient bed usage is compared with states having high concentrations of counties with less than 100,000 in population. While State Psychiatric Center inpatient usage in large urban counties in New York State far exceeds usage in other states with large urban populations, usage in New York's rural counties is more similar to that of more rural states.

State-Operated Psychiatric Center Inpatient Bed Trends and Forecasts

An examination of New York State inpatient bed usage with national rates suggests that the State has a need to continue reducing inpatient bed usage. Table 4-1 shows the decline of State inpatient census and attendant workforce reductions that have occurred since the peak of inpatient usage in the 1950's. Although adult inpatient census has declined drastically from 93,197 in 1955, the number of adult centers has remained nearly the same (17 instead of 20). Since 1955, the average size of an adult psychiatric center has declined from about 5,200 patients to 250 patients in 2003. In addition, since 1955 the system has changed from having no small facilities (under 150 beds) to having seven that size during the past five years. Appendix 5 includes facility specific census, census trends, and surrounding service system capacity within each county catchment area served by each State facility.

Between 1986 and 1996, the adult inpatient census at State Psychiatric Centers decreased dramatically. This census decline was the direct result of policy initiatives regarding placement of geriatric, long-stay patients into nursing homes where they could receive more intensive and appropriate physical health related services, the introduction of a newer, more effective generation of anti-psychotic medications, the growth of the community based outpatient mental health system, and a shift in treatment emphasis from acute to intermediate care within the psychiatric center system. At the same time, the decline in beds for State inpatients has been balanced with significant growth in community housing, as described in Figure 4-8.

Figure 4-8 Changes in Inpatient and Community Residential Beds: 1995-2002

The adult inpatient census was approximately 4,366 on April 1, 2001, a decrease of 2,083 from 1997. This census level was within the range of 3,700 to 4,700 envisioned in OMH's 1997-2001 Five Year Plan. More recent statewide plans have documented that further census reductions have been modest, with virtually all of the census reduction related to "long-stay" individuals who had been inpatients for more than one year.

It should be noted that while New York has reduced the overall number of beds in the State system, it has not reduced community access for people who continue to require inpatient services. This is possible because OMH has significantly reduced the average inpatient length of stay by successfully integrating more long-stay inpatients back into community settings, thereby freeing up beds to serve multiple intermediate-stay individuals each year. New treatment services, changes in community care, support services, patient mix, and the extent and timing of new quality initiatives are all likely to continue this trend.

Over the past five years, admissions to State adult psychiatric centers have remained nearly constant and OMH assumes that this trend is likely to continue for the next three years. A direct consequence of this assumption is that no substantial change in adult short stay (under 1 year) census should be expected in the near future. For the past several years, the adult short stay census has been around 1,850, and it is expected to remain at this level. However, the OMH adult census is still projected to decline because there will be a continued reduction in the adult population whose length of stay is greater than one year. Since 1997, reduction in the total adult census is the result of the diminishing size of this group of inpatients. From 1997 to 2002, the annual rate of census reduction in this group ranged from 0 to 17%, and in the past year, reduction slowed to an annual rate of 4%. OMH projects that the census for the next three years will be 4,130 in 2004, 4,030 in 2005, and 3,980 in 2006.

This census reduction has been occurring with a concomitant emphasis on having community-based service options grow in local areas. As displayed in Figure 4-8, during the period of census reduction the State has continued to develop community living options for people who might otherwise need to use State-operated inpatient services on an extended basis and has been rapidly developing a range of community-based programs. Table 4-2 shows that OMH has added 315 new programs to its certified provider base between 1998 and 2002.

With an estimated State-operated inpatient census of fewer than 4,000 by 2006, New York State must reconsider the role of State-operated inpatient care within the broader continuum of community-based treatment, rehabilitation, and support services available within each county. The benefits of maintaining the current State system infrastructure for the number of individuals served may not justify the associated costs. During this planning period, OMH intends to utilize information developed by the Commission for the Closure of State Psychiatric Centers (described in Chapter 2) to develop and implement long-term plans for the use of State operated inpatient facilities.

This new Commission will host a series of public hearings and will afford an opportunity for concentrated local-level planning efforts toward the most efficient use of resources given the needs and landscape of each community being served. During the period in which the Commission is in existence, selection criteria for making recommendations on State facility closures will be utilized to guide the development of closure recommendations. There is one facility where the case is so compelling that OMH is recommending moving forward with closure at this time. As presented in Table 4-3, the recommendations for closing the Middletown Psychiatric Center describe how some of these selection criteria can be used to analyze and develop closure recommendations that demonstrate fiscal accountability while maintaining the agency's commitment to quality care for all individuals who require services.

Use of a Performance Measurement Framework and Conceptual Model of a Continuum of Care

The data on inpatient usage in New York State speaks clearly to the need for an ongoing State and local capacity to routinely monitor and manage the use of inpatient services against community needs and the availability of other supportive resources. As described in Chapter 2, the new planning process will begin the concentrated State and local government effort to monitor care, and will be enhanced by recent decision-support tools made available to the State.

In 2001, the Center for Mental Health Services (CMHS) articulated the Federal government's commitment to quality improvement through issuance of three-year state grants for building performance measurement data infrastructures. These Data Infrastructure Grants (DIGS) assist all states in reporting on federally required performance measures. As a condition for receiving Federal Block Grant funding, states are expected to develop implementation plans and report yearly on specific measures. CMHS requires that the states, in developing their multi-year plans, select specific measures yearly to chart their own progress at systems development. It also requires that a state's use of selected measures is system-wide, and not restricted to block grant related activities. Appendix 5 contains the federally-required list of performance measures along with OMH's 2003 2004 implementation plan timetable. Further information about the national performance measurement initiative can be found at http://www.samhsa.gov/centers/cmhs/content/blockgrants.

OMH's performance measurement system, which will include but not be limited to the federally required measures, is a major quality improvement activity envisioned for the 2004-2008 planning period. The agency is addressing the domains in the CMHS required construct and is considering additional domains. During this developing process, it is anticipated that many stakeholders will offer ways to measure performance that are also useful. The State plans to begin the assessment of inpatient services by focusing on the following domains:

• Access
• Utilization
• Costs
• Outcomes

This 2004-2008 plan begins the public conversation on performance measurement with some preliminary examples of data describing performance in each of these domains. There will be additional data provided during the planning process. During the planning process, it may be helpful to consider measurement of these domains against the desired "system flow" of inpatient service usage depicted in Figure 4-9.

Figure 4-9 Conceptual model of inpatient access and utilization

This simple model provides a conceptual framework for public discussion on the role of inpatient care in supporting an adult's course of treatment and recovery from mental illness or a young person's experience with serious emotional disturbance. It assumes a continuing maintenance of effort to keep the individual's primary locus of care in home communities and in home settings. This is consistent with the findings from OMH's most recent Patient Characteristics Survey (PCS) that while receiving services, the majority of adults (62%) and children (77%) are living in private residences rather than in institutional settings.

This conceptual model assumes that inpatient services are but one of the services available within an individual's home community, and that this range of services constitute for that individual an appropriate continuum of care. The model assumes that in a new population-based planning process, the resources associated with inpatient services could be decentralized and allocated to communities on a need-driven, person-centered basis. This might mean that inpatient care could be managed according to both personal preferences and clinical needs determined through an objectively measured assessment of acuity. By combining the concepts of person-centered care with level of need based on acuity, even individuals with acute levels of need might be able to be served in settings consistent with their histories and preferences.

For example, people who have had significant experiences with service at an intermediate-care level with a set of trusted providers and clinicians might be able to reenter the system of care at that level rather than having to first access emergency or inpatient care. Currently, the system of care requires that every individual returns to the beginning of the continuum of care (emergency or inpatient) each time their mental health symptoms become exacerbated. This current access method is system-centered rather than person-centered and may create both frustration and lack of continuity for individuals and their families.

The conceptual model presented here implies that there could be a new "re entry at level of need" access method, which might mean that people could activate care directives when symptoms exacerbate and entirely avoid using either emergency or inpatient services. As an outgrowth of this concept, individuals at varying stages of their recovery process could have options beyond the existing inpatient hospitalization to include innovations such as peer operated crisis residences, intermediate care residences with treatment capacity provided in home-like, assisted living environments, and access to specialized diagnostic treatment centers for persons with mental health conditions who have been historically unresponsive to traditional therapies.

In addition, as the OMH planning process becomes population-based, it will be necessary to consider admission to inpatient services against historical patterns of usage by different groups. This new planning process could also include reviews of the respective roles of State and local inpatient services in the delivery of acute, intermediate, and extended stay levels of care. According to the 2001 PCS, there are several seemingly significant differences in the populations that use inpatient services. Table 4-4 describes several areas of difference among adults using State-operated psychiatric centers and general hospital inpatient units that need to be further explored and discussed during the planning process.

Examples of Inpatient Performance Measurement

Access to State and Local Beds for Adults and Children

A basic way to consider system performance in the area of access to services is to look at a system's capacity for delivering that service. For inpatient bed use, individuals and their families are unlikely to be able to use beds when needed if capacity is inconsistent with local need. Throughout this chapter, it has been shown that New York State has bed capacity in excess of the nation and other comparable states. It is useful to examine this excess against county-specific capacities in the State.

In Appendix 5, Tables 1 and 3 show the existing bed capacity for each county for adults and children. The county-level tables are all presented by showing capacity as a rate per 100,000 of the county's population. Examination of the county-level capacity level data shows that there is little in the way of a discernible pattern for bed allocation over time. During the new planning process, it is likely that counties and other stakeholders will be able to describe the factors that have led to their current capacity patterns. The set of descriptive statistics for both adults and children in Table 4-5 demonstrates the extreme degree of variability within these data.

Utilization of State and Local Children's Inpatient Beds

Another important measure for determining the degree to which inpatient services are being adequately provided to meet community need is to examine the degree to which the current bed capacity is actually being used. The concept, which is referred to here as "census," reflects the daily bed use for the inpatient units. In Appendix 5, Tables 2 and 4 display these data by county, showing that the rate per 100,000 of actual utilization of beds is also highly variable by county. The descriptive statistics in Table 4-5 demonstrate that making comparisons about county-level utilization may not be useful. Comparison in general will however show that utilization rates are lower than capacity and that, because utilization data is based on the county of residence of the individual using the service, there is in the lower Hudson River counties, significant out-of-county usage of both adult and children's inpatient services.

Cost of Inpatient Services

From a fiscal perspective, New York State continues to lead the nation in its commitment to provide funding for mental health programs such as inpatient care. Data compiled in 2003 by the National Association of State Mental Health Program Directors (NASMHPD) shows that since the mid-1990s, New York State mental health agency controlled expenditures exceeded those of any other state, both in total dollars and when adjusted for differences in population. For example, New York led the nation in 1993, spending almost $2.4 billion or $130 per capita on mental health services. In 2001, New York continued to lead the nation, spending over $3.3 billion or $175 per capita. Figure 4-10 shows how New York's 2001 per capita total expenditures compare to other mid-Atlantic states, the averages for all states in the region, and the national average for both community-based programs and state psychiatric hospitals.

Figure 4-10 Fiscal Year 2001 SMHA-Controlled Per Capita Expenditures By Type in Mid Atlantic States*

Reflecting national trends, there has been a significant, concurrent shift in emphasis away from state psychiatric hospital-based care to community based programs. NASMHPD data show that in 1993, New York spent $87 per capita on inpatient care and less than $44 per capita on community-based care. By 2001, per capita spending on inpatient care had dropped by half to $53, while per capita spending on community-based care more than doubled to $117. Much of this shift in spending in New York was driven by reinvesting savings derived from the reduction in State inpatient capacity into community-based programs. The State will continue to examine the underlying reasons why its public mental health inpatient usage and costs are so significantly different from national averages while maintaining a commitment to the highest possible quality of care.

Figure 4-11 Percent of Clients Served vs Percent of Expenditures

Figure 4-11 provides an indication of why the issue of inpatient bed utilization is important for OMH by describing the relative distribution of expenses compared with utilization of key service areas. It shows that while relatively few New Yorkers are using inpatient services, the costs for these services account for more than half of the agency's expenditures. For example, an OMH retrospective review of 2002 Medicaid data revealed that 1% (3,000) of the total number of individuals (259,000) accessing licensed mental health services accounted for 14% of the total mental health Medicaid costs for these services. The primary service driving these expenditures was hospital inpatient utilization. This is significant to consider against the framework of policy commitments to bring more effective, evidence-based interventions to individuals in their own neighborhoods through an improved array of community-based services. As New York State continues its efforts to realize this shift in the quality of and locus for effective interventions, it must look for ways of realigning resources to support this commitment. Appendix 6 contains a list of all revenues and expenditures for the State by all major program categories.

Patterns of Usage: Costs Associated with Inpatient Care

It is also useful to examine one important variable within costs for inpatient care - the degree of frequency with which individuals use the service. At present, OMH is not routinely able to analyze individual-level outcomes based on either the diagnosis or level of functioning of individuals. With the increasing use of the Child and Adult Integrate Reporting System (CAIRS) described elsewhere in this Plan, it will be possible to determine more information about the patterns of inpatient usage for persons deemed to have multiple and intense service needs. At this time it is possible, however, to analyze important information about this potential cohort of individuals by examining patterns of multiple inpatient service usage over a given year's period of time. This examination might infer that individuals with high rates of inpatient usage may have service needs that are not easily met through traditional courses of treatment either in the inpatient or community care settings.

One Outcome of Using Inpatient Services: Rates of Readmission Within 30 and 180 Days
There are several potential, key ways to think about both the patterns of using inpatient services and intended outcomes. As OMH considers domains for a performance measurement system, it is useful to look at just one of the most basic measures of both appropriateness and outcomes: the degree to which individuals who use inpatient services return to inpatient services within relatively short periods of time. The Federal indicators for time frames of readmission call for measurement at 30 and 180-day intervals. Additional work will be done during the development of the agency's performance measurement system to review readmission data by level of care (acute inpatient, intermediate, and extended care). OMH will also be producing reports on admissions and discharges to these levels of care by all relevant auspices, geographic distributions, and population groups, and plans to examine population-and sector specific readmission rates in the future.

Figure 4-12 30 Day Readmission Rates for Medicaid Recipients to Inpatient Settings by Region 1997 to 2001

Using data from the DOH Medicaid Management Information System, Table 4 6 presents rates of readmission to inpatient psychiatric units statewide and by auspice among mental health recipients who were discharged during calendar year 2001. The following graphs display rates of readmission to inpatient services in all service sectors by 30 and 180-day measures also using data from the system. Figure 4-12 displays the 30-day rates of readmission to inpatient services for Medicaid recipients by OMH region from 1997 to 2001. Figure 4-13 displays the 180-day rates of readmission to inpatient services for Medicaid recipients by OMH region from 1997 to 2001. Appendix 5 contains more specific county readmission profiles.

Figure 4-13 180 Day Readmission Rates for Medicaid Recipients to Inpatient Settings by Region 1997 to 2001

Readmission Patterns Among Users of Inpatient Services
The use of inpatient services over time is not presented with the judgment that repeated use represents either an ineffective course of care during inpatient treatment or ineffective outpatient care. The intent of displaying information on readmission rates is to provide the mental health community with management information, which may be useful for planning and monitoring purposes. To further the intended use of these data, it is helpful to provide one additional way of looking at the information on inpatient service usage - a look at the patterns of use by individuals, and their specific episodes of inpatient treatment.

In 2001, there were 66,659 psychiatric inpatients who were discharged from all inpatient settings combined. This includes people using State Psychiatric Centers (1,683), private hospitals (3,571), general hospitals (60,985), and residential treatment facilities for children (420). For 12,187 (18%) of these episodes, the individual's discharge was followed by an inpatient readmission within 30 days of the discharge. In total, 7,434 persons accounted for these readmissions, but some persons were readmitted more frequently than others. Figure 4-14 presents the number of readmissions in 2001 per person. Approximately 5,200 persons were readmitted once, accounting for 5,200 (42%) readmissions; 1,300 persons were readmitted twice, accounting for 2,600 (21%) readmissions; and 45 persons were admitted over 10 times during the year, accounting for 637 (5%) of the readmissions. Expressed differently, 13% of the persons readmitted accounted for 35% of the inpatient readmissions.

Figure 4-14 Identifying Frequent Users of Inpatient Services

Part of the anticipated statewide planning discussions will focus on ways to further analyze these data, including an understanding of the course of service delivery associated with repeated inpatient admissions. It is necessary to conduct additional analysis of these rates and patterns of use because their underlying factors are not readily discernible. There are many reasons for these data being what they are and their relative use as performance measures is still in its early stages of development.

Discussion and Questions for Consideration

Inpatient Care

The information presented in this Chapter and its accompanying appendices shows that there are disparities in the use of inpatient care among population groups, providers, service sectors, regions, and counties. Factors influencing the differences in use are not necessarily obvious. A number of questions, some of which are outlined below, need to be considered before the appropriateness of the observed differences can be assessed. These questions and the issues that they raise can be discussed and prioritized as part of the planning process.

• What type of variation is there across counties by region and by comparable county? Are there some counties in the different regions or counties of similar demographics with similarly high or low rates? If so, do they share similar service system characteristics?
• Are there any other demographic or service variables that might account for the observed differences or similarities in inpatient use by region, county size, and utilization group? For example, the correlation of factors associated with severe mental illness (poverty rates, unemployment, education levels, percent Caucasian, etc.) with inpatient use should be considered.
• Is the pattern of inpatient use by region and county demography similar for all types of inpatient care (i.e., State Psychiatric Centers and general hospital units)? Are there particular characteristics associated with general hospital units that may influence frequency and duration of service use?
• What are the patterns of use and capacity in other 24-hour mental health care settings (e.g., community residences, supported housing, adult homes, transitional housing, and family care)? Does the availability of these options influence inpatient service use?
• Do inpatient rate levels suggest overuse, under use, or misuse of this most costly service and if they do, how can the State and county governments address these while continuing to build capacity for community based treatment, rehabilitative, and support services? Are there benchmarks for assessing the appropriateness of readmission and do these benchmarks need to be adjusted for differences in level of care, auspice, geography or population sub-groups? Do specific individuals or cohorts of individuals account in large part for some of these differences?
• Are there reliable prediction models of psychiatric inpatient use? How can such models be used in developing a person-centered continuum of community-based care where both need and personal preference can be concurrently and equitably addressed?

Notes

1. Mental Health, United States, 2000. Ronald W. Manderscheid, Ph.D. and Marilyn J. Henderson, M.P.A. (Eds.) U.S. Department of Health and Human Services, Substance Abuse and Mental Health Services Administration, Center for Mental Health Services. (Rockville, MD: 2001). Retrieved December 19, 2003 from http://www.mentalhealth.org/publications/allpubs/SMA01-3537/default.asp
2. *     Data Sources:  Additions and Residents at End of Year, State and County Mental Hospitals by Age and Diagnosis, by State, United States, 2000

Comments or questions about the information on this page can be directed to the Office of Planning.