Inpatient Mental Health Services in Rural Areas: An Interregional Comparison
Morton 0. Wagenfeld,Western Michigan University
Harold F. Goldsmith, Substance Abuse and Mental Health Services Administration
Rockville, Maryland,
Diane Stiles,University of Maryland
Ronald W. Manderscheid, Substance Abuse and Mental Health Services Administration,Rockville,
Maryland
Abstract
It is well established that rural or nonmetropolitan areas suffer from a
deficit of mental health services relative to their urban or metropolitan
counterparts. Previous research, however, has not examined the different
kinds of nonmetropolitan areas. Applying a classification of nonmetropolitan
counties developed by the Economic Research Service of the U.S. Department
of Agriculture based on the 1980 Census to two 1983 NIMH data sets, the
distribution of inpatient mental health services is examined. It was found
that inpatient mental health services are less likely to be found in nonmetropolitan
areas. Where they do exist, they tend to be located in certain kinds of
counties and virtually absent in others. It also was found that where inpatient
services do exist, the non-federal-general hospitals were the most common
providers. It is concluded that using the USDA classification of nonmetropolitan
counties in conjunction with the metropolitan/nonmetropolitan status of
counties provided important insights into the availability of inpatient
mental health services to nonmetropolitan residents.
Introduction
A substantial body of literature has accumulated in the past several decades documenting the general health and human services deficits experienced by the more isolated rural components of nonmetropolitan areas relative to their more urban metropolitan counterparts (e.g., Ahearn, 1979; Bachrach, 1981, 1985; Baumhier, Derr, & Gage, 1973; Flax, Wagenfeld, Ivens, & Weiss, 1979;, Moscovice, 1989; Rosenblatt 1981; Wagenfeld, 1981; Wagenfeld & Gage, 1986; Wagenfeld & Ozarin, 1982). These reviews indicate that in almost all respects, nonmetropolitan areas, particularly the more rural parts, are poorer, have lower levels of education, and also exhibit higher levels of physical disability and a need for medical services.
The level and quality of mental health services are also considerably below that of metropolitan areas. Longest, Konan, and Tweed (1979), for example, demonstrated that "adequate [psychiatric] services" are more likely to be found in the more urban parts of metropolitan market areas and least likely to be found in isolated nonmetropolitan areas or fringe market areas of metropolitan communities. Similar results have been observed by Wagenfeld, Goldsmith, Stiles, and Manderscheid (1988), and Goldsmith, Wagenfeld, and Stiles (1988). More recently, Stuve, Beeson, and Hartig (1989), examining trends in Nebraska, noted a decline in full-time staff in nonmetropolitan CMHCs. They concluded that this portended a crisis in the centers' ability to respond effectively to the needs of rural residents. Not only are there fewer services in nonmetropolitan areas but evidence reviewed by Bachrach (1985) indicates that the population of these areas also utilize mental health services at lower rates than the populations of metropolitan areas. This paper attempts to sharpen the issue by analyzing the distribution of inpatient mental health services in nonmetropolitan areas in an alternative manner, one that the writers hope proves to be more useful.
The recent attention paid to the economic crisis in nonmetropolitan America - particularly its rural and agricultural sectors - (e.g, Barrett, 1985; NIMH, 1986; Wagenfeld, 1988, 1990) has put the service delivery and policy implications into sharp and urgent focus. There seems to be widespread agreement among both researchers and policymakers concerned with health issues (cf. Hersh & Van Hook, 1989; Hewitt, 1989; Moscovice, 1989; Patton, 1989) that an impediment to the resolution of the problem has been the lack of consensus about the nature of the independent variable. That is, what do we mean by "rural" or "nonmetropolitan?"
While we speak of the diversity of rural (or nonmetropolitan) America, research and policy are frequently carried out on the basis of several incorrect assumptions. These incorrect assumptions have hampered the effective delivery of services and the formulation of policy solutions. The first is that "rural/urban" and "metropolitan/nonmetropolitan" are frequently regarded as interchangeable or equivalent concepts. The second is that persons living in open country or small towns are often seen to constitute a homogeneous entity. Patton (1989) has captured this quite well in noting:
Recently, Henry, Drabenstott, and Gibson (1986) urged that the economic
problems of nonmetropolitan and rural areas be appreciated in terms of the
complex and heterogeneous character of the region. As a first step toward
making informed policy decisions, there is a need to have data on both the
availability of resources and the need for services. These data also need
to be in a form that is meaningful. To do this, the open country and small
town environments that constitute nonmetropolitan America has, first to
be understood by taking into account its economic and social diversity.
This paper takes a step in that direction by examining the distribution
of inpatient mental health services in nonmetropolitan areas utilizing a
classification that explicitly builds on differences in the social and economic
fabric of this part of our society.
The Classification of Nonmetropolitan Areas
We began this paper by noting that while we speak of the diverse nonmetropolitan
areas, research and policy decisions are often carried out with instruments
that do not reflect this heterogeneity. To overcome this problem, the Economic
Research Service (ERS),using 1980 Census data, developed a classification
of counties that reflects the diversity of nonmetropolitan America (Bender,
Green, Hady, Kuehn, Nelson, Perkinson, & Ross, 1985; Ross, 1987) (See Footnote 1). The county is a useful unit of analysis because it is
a recognized geopolitical entity with stable boundaries with a wide variety
of socioeconomic and demographic data routinely available.
Before presenting this classification, it is necessary that the differences between metropolitan and nonmetropolitan counties and between rural and urban populations be clearly specified. First, metropolitan counties are those that are socially and economically integrated with a large city (usually 50,000 or more) on a daily basis. Nonmetropolitan counties are those that are not part of a metropolitan area. Within metropolitan and nonmetropolitan counties, people are considered to be urban residents if they live in a place that has at least 2,500 persons or in the densely settled areas around big cities. The remaining persons - those residing in very small towns or open country - are designated as rural residents (U.S. Bureau of the Census, 1983).
For the 2,443 counties that were classified as nonmetropolitan in 1980, all but 370 were classified into seven types using the ERS classification (see Table l). These types are based on type of economy and land use, as well as socioeconomic character of the population. In view of the commonly held notion that equates nonmetropolitan with agriculture, it is interesting to note that there are almost as many manufacturing-dependent counties (678) as farming-dependent (702). Moreover, the different types of nonmetropolitan counties are not randomly distributed. For example, farming-dependent counties are more likely to be found in the Midwest and Great Plains, mining in the West and Appalachia, persistent poverty in the Southeast, and federal lands in the West.
Parenthetically, it should be noted that over half (57.3%) of the counties could be classified as belonging exclusively to one group, an additional 22% belonged to two groups, and only 6% in three or more groups. Of the seven types, the poverty counties were most likely (83%) to be classifiable as another type. The authors did not consider this to be troublesome. In the final analysis, the utility of any classification depends on its ability to discriminate and provide useful information. As they noted (Bender et al, 1985):
At the beginning of this paper, we noted that nonmetropolitan areas are,
in almost all respects, socioeconomically disadvantaged relative to metropolitan
areas. The nonmetropolitan counties, however, display a great deal of variability.
As Bender et al. (1985) reported, in addition to their geographic dispersion,
nonmetropolitan counties differ with respect to a variety of socioeconomic
and demographic characteristics that are likely to be related to the need
for health and mental health services. Briefly, they noted that the proportion
of the population who report themselves as disabled ranges from a low of
less than 10% in farming-dependent and federal lands, to a high of 14.8%
in persistent poverty. Per capita income ranges from less than $5,000 in
persistent poverty to over $7,000 in mining-dependent counties. Proportion
of Black is lowest in federal lands (1.6%) and highest in persistent poverty
(2330/o. Mining-dependent and federal lands counties contain the lowest
proportion of persons 65 and over, while farming-dependent and destination
retirement counties contain the highest
In addition to sociodemographic variability, Bender et al. (1985) noted that nonmetropolitan counties are at different levels of risk for certain world changes. Changing levels of demand for agricultural or petrochemical products, for example, have had profound impact on farming or mining counties, while leaving unaffected retirement or government counties. Also, changes in the age structure of the population have had major impact on the retirement counties, while leaving unaffected federal lands and mining.
The ERS classification has clearly demonstrated the considerable demographic and socioeconomic diversity in nonmetropolitan America. The proportion of the population in poverty, proportion of elderly or dependent persons, differential vulnerability to economic swings, and so on, in the different types of nonmetropolitan counties are all related to the need for services. Policy solutions also must recognize these geographic differences. As a first step toward appreciating the consequences of this diversity for mental health services, we examine the distribution of inpatient mental health services. This description can point the way toward future analytic research projects.
Methods
Previous studies of the distribution of mental health services have employed catchment areas (CAs) as their unit of analysis. Because these CAs were population-based aggregations of counties, as well as census tracts, minor civil divisions, or census county divisions, they frequently failed to conform to recognized geopolitical or ecological boundaries, this meant that comparison with health and other service data could not be made and that metropolitan areas could not be precisely separated from nonmetropolitan. Equally important comparisons across ecological units of radically different sizes are generally not considered appropriate because internal homogeneity can vary appreciably (Duncan, Cuzzort, & Duncan,1961).
In 1981, CAs were discontinued as a unit of geography for which the National Institute of Mental Health collected facility and service information. As a result, studies of the ecological distribution of mental health services have not been available. Recently, this problem has been overcome. We have been conducting analyses of the ecological distribution of inpatient mental health services in which the county has been used as the unit of analysis. Earlier papers considered county urbanization metropolitanization and trade area statuses (e.g., Goldsmith, Wagenfeld, & Stiles, 1988; Wagenfeld, Goldsmith, Stiles, & Manderscheid, 1988). Like the earlier papers, this one uses the county as the unit of analysis, thereby providing a precise designation of counties as metropolitan or nonmetropolitan. However, unlike the earlier papers, this paper emphasizes the diversity of nonmetropolitan counties. By using the ERS classification, it provides more detailed information on the location and type of inpatient mental health services in nonmetropolitan areas.
The inpatient psychiatric services data presented in this paper are from two sources: the 1983 Inventory of Mental Health Organizations (IMHO), conducted by NIMH - and the 1983 Inventory of General Hospital Mental Health Services (IGHMHS), a joint NIMH/American Hospital Association project (see Redick, Witkin, Atay, & Manderscheid, 1986; and Redick, Witkin, Atay, & Fell, 1987). This is the latest date for which the disaggregated information is available.
The analysis has been limited to inpatient mental health services because they are the only type of service in the 1983 IMHO and IGHMHS surveys for which accurate county-level estimates could be derived (see Goldsmith, Henderson, & Manderscheid, 1988, for more details). Since not all mental health facilities reporting inpatient psychiatric services reported the number of beds, statistical estimation techniques were employed to arrive at bed estimates for a county in a given county type (see Stiles & Goldsmith, 1988, for more details).
Inpatient Mental Health Services in Nonmetropolitan Areas
Before looking at differences in the distribution of inpatient mental health
services by ERS county type, it would be useful to note briefly some metropolitan/nonmetropolitan
differences. Earlier, we made the point that nonmetropolitan areas, particularly
the more rural parts, have consistently experienced a health services deficit
- including mental health - relative to their more urban metropolitan counterparts.
The extent of this can be seen for inpatient mental health services in Table 2.
While about one fourth of all counties have some inpatient services, almost two thirds of the 735 metropolitan counties have them. This is in contrast to only 12.8% of the more than 2,400 nonmetropolitan counties. The table also shows that these services are heavily concentrated in nonmetropolitan counties with large urban populations (20,000 persons or more). Slightly over one half of the most urban nonmetropolitan counties have inpatient services, as compared to only 11% for nonmetropolitan counties with urban populations between 2,500 and 20,000 and under 1% for the totally rural nonmetropolitan counties (those with urban populations under 2,500). Adjacency to a metropolitan area does not appear to be an important factor in differentiating counties. About 52% of nonmetropolitan counties adjacent to a metropolitan area have some inpatient mental health services, in contrast to 57% of the nonadjacent counties. This is an important point for policy development and is returned to later.
In addition to the existence of an inpatient service, the number of beds also is an important consideration. "Number of beds" is a useful index of the amount of inpatient services available in a county. For metropolitan counties, the median number of beds per county is 28.6, while for nonmetropolitan it is 0. On this broad level then, the nonmetropolitan deficit seems evident.
Turning to the main focus of this paper, Table 2 also provides information indicating that the distribution of inpatient mental health services by type of nonmetropolitan county is not random. These services are most likely to be found in three types of counties: specialized government, manufacturing-dependent, and ungrouped ERS counties (23%, 21%, and 20%, respectively. The types of counties least likely to have services are the persistent poverty and farming-dependent. Counties in these types are the least urban (33% and 27% percent, respectively. It also should be noted that these counties also contain high concentrations of persons (Black and elderly) shown to have high need for mental health services.
The median number of beds/county for all nonmetropolitan counties and each type is 0, unlike metropolitan counties, which have a median of 29. This indicates that large amounts of inpatient mental health services do not exist in any ERS category. Some interesting variation occurs when one examines the average and the 90th percentile. Starting with the average, it ranges from a low of 4.7 in the farming-dependent, to a high of 55.2 in the specialized government counties. In the former, the average is about one fourth that of all nonmetropolitan counties while, in the latter, the average is over four times greater than the norm - that is, the average for all nonmetropolitan counties. The large average number of beds and small median number here reflects the existence of large facilities, such as state hospitals or Veterans Administration Medical Centers in a few counties and no beds in the vast majority of counties.
Because of the highly skewed nature of the distribution, neither the
mean nor the median provides clear insights into the nature of the distribution
of psychiatric inpatient services in nonmetropolitan counties. Accordingly,
data also are presented in Table 2 for the 90th percentile of beds for the
ERS categories. The 90th percentile identifies the amount of beds in the
top 10% of the nonmetropolitan counties most likely to have inpatient
services: that is, the ones with the most beds and, consequently, services.
Thus, while the median for all ERS categories with inpatient services is
0, the 90th is not. The almost complete lack of inpatient beds in the farming-dependent.
mining-dependent, and persistent poverty counties is indexed by their 90th
percentile being 0. Unlike these ERS categories, specialized government,
ungrouped, and manufacturing-dependent counties had the highest 90th percentiles:
62.8, 29.8, and 26.6 beds/county, respectively.
While the distribution of inpatient mental health services in nonmetropolitan
areas and the availability of beds have been discussed, the analysis has
not yet taken into account the kinds of organizations that are providers
of inpatient mental health services. Such additional information is essential
if one is to understand the kinds of services that are available in different
types of counties. That is because different types of organizations offer
different types of services and are available to different groups in a geopolitical
area.
To paraphrase Goldman, Taube, & Jencks (1987), all psychiatric beds are not created equal. They show a greater diversity than does the general medical sector. The authors noted that in moving from psychiatric units in general hospitals, to private psychiatric hospitals to state and county hospitals, one observes "a continuum of increasing length of stay, increasing frequency of patients with psychotic diagnoses, increasing frequency of civil and criminal commitments, and increasing frequency of permanently disabled patients" (S8). While their analysis did not include Veterans Administration Medical Centers, it seems reasonable to view these hospitals as lying closer on the continuum to the state and county hospital, while the Multiservice Mental Health Organization is more like the non-federal-general hospital. Additionally, the various hospitals also differ with respect to type of reimbursement, legal obligation to accept patients, diagnostic mix, and therapeutic modalities (McCarrick, Rosenstein, Milazzo-Sayre, & Manderscheid, 1988). Stated differently, mental health organizations differ with respect to the kinds of services provided and to the patients served.
The previous discussion has highlighted the differences among inpatient
psychiatric facilities. It follows, then that the location of a particular
type of psychiatric facility is not an accurate index of availability, accessibility,
or appropriateness of the service to a county's residents. A state and county
hospital or a Veterans Administration HospitaL for example, serve administratively
designated areas that encompass part or all of a state or, in the case of
the latter, several states, and is legally obligated to accept all eligible
patients. These organizations, also are more likely to provide long-term
care for chronic patients. In contrast, the non-federal-general hospital
and private psychiatric hospitals, along with the multiservice mental health
organizations, are more likely to provide a range of services to persons
residing in the counties in which they are located or the residents of adjacent
counties.
Table 3 provides
information on the organizations that provide services within the several
ERS categories. Like the previous table, it presents two kinds of data:
metropolitan/nonmetropolitan and ERS county type differences for the nonmetropolitan
counties. For both the metropolitan and the nonmetropolitan counties, the
major provider of inpatient mental health services is the non-federal-general
hospital, followed by state and county hospitals. One half of the metropolitan
counties have these general hospital psychiatric units. In contrast, only
8.2% of the nonmetropolitan counties have such services. Similarly, one
in four metropolitan counties has a state and county hospital as compared
to only 3% of the nonmetropolitan counties. The general pattern is the same
for the other types of providers. In the aggregate, while they are not a
major provider of services, private psychiatric hospitals - both for-profit
and not-for-profit - are an almost exclusively metropolitan phenomenon.
The situation, however, may be changing. Stuve, et al. (1989) noted that
the number of private psychiatric beds in Nebraska's nonmetropolitan areas
increased from 0 to 172 between 1981 and 1988. Anecdotal accounts confirm
this.
Looking at the ERS county type comparisons for all of the categories, the non-federal-general hospital is most likely to be the most important provider of inpatient services. The few private psychiatric hospitals that operate in nonmetropolitan areas are concentrated in the more urbanized manufacturing-dependent counties. Farming-dependent and persistent poverty counties have a dearth of inpatient services. If we exclude state and county hospitals and Veterans Administration Medical Centers - essentially providers of long-term care - we can see that there are virtually no inpatient services in these counties. One percent of the former, and less than 2% of the latter are likely to have any acute-care services in the county.
Discussion
This paper has, for the first time, applied a classification of rural areas to a 1983 data set on the distribution of inpatient mental health services. Analysis of the data again documents the point that with respect to the location of inpatient mental health services, nonmetropolitan counties are significantly underserved relative to nonmetropolitan. Further, we have shown that the services tend to be located in the more urban counties and are virtually absent in the more rural counties. The counties with services are more likely to have larger urban populations.
Having looked at where inpatient services are provided, how much is available, and what organizations provide these services, we then looked at what organizations provide services to the different types of nonmetropolitan areas.The non-federal-general hospital was the most common provider of services in virtually all of the ERS county types. However, non-federal-general hospitals were never more than 16% of any county type and were present in only 8% of the nonmetropolitan counties. Finally, with respect to the farming-dependent and persistent poverty counties, we emphasized their particularly low level of services.
As we noted earlier, the location of a particular type of psychiatric facility is not an accurate index of availability, accessibility, or appropriateness of the service to the county's residents. Barriers to utilization may exist with respect to cost of services, military status, or age restrictions. Additionally, state and county hospitals and Veterans Administration Medical Centers, on the one hand, and non-federal-general and specialty psychiatric hospitals and Multiservice Mental Health Organizations, on the other, differ in important ways in terms of service provided and patients served. The latter also are more likely to provide services to persons residing in the counties in which they are located or the residents or adjacent counties.
In sum, our analysis has shown that inpatient mental health beds are unlikely to be available to the counties where nonmetropolitan people reside, particularly if they reside in poverty or farming areas. It is likely that they will have to travel some distance. Further, where services are sited, they are not distributed in a uniform fashion. Rather, they are concentrated in those counties with specialized government functions. In view of the recent concern with the crisis in U.S. agriculture and our long-standing national concern with the problems of poverty, it is ironic that these counties are the least likely to have inpatient services.
Conclusions and Implications
We noted earlier the plea for a common typology of rural ar. that would
enable researchers to make comparisons across different data base. Using
the ERS classification, our descriptive analysis of the distribution of
inpatient mental health services suggests that it has some potential to
fill this need. This classification was based on land use and economic characteristics
of counties and permitted us to identify areas with and without inpatient
mental health services. 'This ability to differentiate in an area of inpatient
service delivery availability suggests that its use be expanded in order
to begin to develop a common data base.
This paper has shown, once again, the metropolitan/nonmetropolitan services
gap in the distribution of inpatient mental health services. We have gone
a step further, though, in applying the ERS classification and demonstrating
that certain nonmetropolitan areas suffer from greater deficits than others.
We showed that certain types of counties are virtually without these services.
In part, the distribution of inpatient psychiatric services appears to be
associated with urban concentrations (urbanization).
To repeat an earlier caveat: This paper was limited to the location and
quantity of inpatient services and the nature of theproviders. The data
do not permit any discussion of mental health services utilization. The
next step is research that builds on this approach and examines patterns
of access and utilization.
References
Ahearn, M. (1979, July). Health care in rural America . Washington, DC: Economics, Statistics, and Cooperative Service, USDA, Agriculture Information Bulletin No. 428.
Bachrach, L. L. (1981). Human services in rural areas.- An analytic
review. Human Services
Monograph Series No. 22. Rockville, MD- Project SHARE.
Bachrach, L L. (1985). "A sociological perspective." In L R. Jones & R. R. Pariour (Eds.), Psychiatric services for underserved rural populations, pp. 5-26. New York: Brunner/Mazel.
Barrett, J. (1985). Mending a broken heartland. Alexandria, VA: Capitol Publications.
Baumheir, E. C., Derr, J. M, & Gage, R. W. (19731 Human services in rural America: An assessment of problems, policy, and research. Denver: Center for Research and Development, University of Denver.
Bender, L. D., Green, B. L., Hady, T. F, Kuehn, J. A, Nelson,
J. K., Perkinson, L. B.,
& Ross, P. J. (1985, September). The diverse social and economic
structure of nonmetropolitan America . Washington, DC: Economic
Research Service, USDA,
Rural Development Research Report No. 49.
Duncan, 0. D, Cuzzort, R. P, & Duncan B. (1961). Statistical geography.-
Problems in
analyzing area data. Glencoe, IL: Free Press.
Flax,J.W, Wagenfeld,M.O, Ivens,R.E, & Weiss,R.J. (1979). Mental health and rural America: An overview and annotated bibliography. DHEW Publication No. (ADM) 78-753. Rockville, MD: National institute of Mental Health.
Goldman, H. H, Taube, C. A, & Jencks, S. F. (1987). The organization of the psychiatric inpatient services system. Medical Care, 25, Supplement, S6-S21.
Goldsmith, H. F, Henderson, P, & Manderscheid, R. W. (1988). Feasibility of county level analysis based on the 1983-1984 Inventory of Mental Health Organizations and the 1983-1984 Inventory of General Hospital Mental Health Services. Unpublished report. Rockville, MD: Survey and Reports Branch, National Institute of Mental Health.
Goldsmith, H. F., Wagenfeld, M. 0., & Stiles, D. F. (1988, August). Ecological patterns of inpatient mental health service in nonmetropolitan and rural areas. Presented to the Annual Meeting of the Rural Sociological Society, Athens, GA.
Henry, M., Drabenstott, M., & Gibson, L. (1986, July/August). A changing
rural America. Economic Review, Federal Reserve Bank of Kansas City,
pp. 23-41.
Hersch, A., & Van Hook, R. (1989). A research agenda for rural health
services. Health Services Research, 23, 1053-1064.
Hewitt, M. (1989, July). Defining "rural" areas: Impact on health care policy and research. Unpublished staff paper. Health Programs. Office of Technology Assessment. Congress of the United States. Washington, D.C: US Government Printing Office.
Longest, J., Konan, M., & Tweed, D. (1979). A study of deficiencies and differentials in the distribution of mental health resources in facilities. Analytic and Special Studes Reports, Series B, No. 15. DHEW Publication No.(ADM) 79-517. Rockville, MD: National Institutes of Mental Health.
McCarrick, A.K., Rosenstein, M.J., Milazzo-Sayre, L. J., & Manderscheid, R.W. (1988). National trends in the use of psycholtherapy in psychiatric inpatient settings. Hospital and Community Psychiatry, 39, 835-841.
Moscovice, I. (1989). Rural hospitals: A literature synthesis and health services research agenda. HSR: Health Services Research, 23, 892-930.
National Institute of Mental Health. (1986). Summary report on policy forum on rural stress. Unpublished report. Rockville, MD: Office of Policy Analysis and Development.
Patton, L. (1989). Setting the rural health services research agenda: The congressional perspective. Health Services Research, 23, 1005-1051.
Redick, R.W., Witkin, J.J., Atay, J.E., & Manderscheid, R. W. (1986). Specialty mental health organizations, United States, 1983-1984. National Institute of Mental Health, Series CN, No. 11. DHEW Publication No. (ADM) 86-1490. Washington, DC: Superintendent of Documents.
Redick, R.W., Witkin, M.J., Atay, J.E., & Fell, A.S. (1987). Separate psychiatric services in non-federal general hospitals, United States, 1983. Statistical Note No. 185. Rockville, MD: National Institute of Mental Health.
Rosenblatt, R.A. (1981). Health and health services. In A.H. Hawley & S.M. Mazie (Eds.), Nonmetropolitan America in transition, pp.614-642. Chapel Hill: University of North Carolina Press.
Ross, P.J. (1987). Remarks on the development of a policy-oriented classification of nonmetropolitan couties. In A. E. Luloff (Ed.), Rural people and places: A symposium on typologies, pp. 177-180. Publication No. 47, University Park, PA: Northeast Regional Center for Rural; Development.
Stiles, D., & Goldsmith, H.F. (1988). Estimating the number of inpatient psychiatric beds in counties. Unpublished manuscript. Survey and Reports Branch, Division of Biometry and Applied Science. Rockville, MD: National Institute of Mental Health.
Stuve, P., Beeson, P., & Hartig, P. (1989). Trends in the rural community mental health work force: A case study. Hospital and Community Psychiatry, 40, 932-936.
U.S. Bureau of the Census. (1989). Statistical Abstract of the United States: 1989 (104th ed.). Washington, DC: U.S. Government Printing Office.
Wagenfeld, M.O. (Ed.) (1981). Perspectives on rural mental health. New Directions for Mental Health Services Series, No. 9. San Francisco: Jossey-Bass.
Wagenfeld, M.O. (Ed.) (1988). Mental health and the crisis of rural America. (Special Issue). Journal of Rural Community Psychology, 9, 3-92.
Wagenfeld, M.O. (1990). Mental health and rural America: Review of a decade. Journal of Rural Health, 6, 507-522.
Wagenfeld, M.O., & Gage, L.W. (1986, September). Mental health services in rural areas: Origins, progress and prospects. Presented to the Annual Meeting of the American Public Health Association, Las Vegas, NV.
Wagenfeld, M.O., & Ozarin, L.D. (1982). Serving the underserved through rural mental health programs. In H.C. Schulberg & M. Killilea (Eds.), The modern practice of community mental health (pp. 467-485). San Francisco: Jossey-Bass.
Wagenfeld, M.O., Goldsmith, H.F., Stiles, D., & Manderscheid, R.W. (1988). Inpatient mental health services in metropolitan and nonmetropolitan counties. Journal of Rural Community Psychology, 9, 13-28.
Table 1
Classification of Nonmetropolitan Counties*
| Type | Definition | Number of counties |
| Farming-dependent | >20% income from agriculture | 702 |
| Manufacturing- dependent | >30% labor and proprietor income from manufacturing | 678 |
| Mining-dependent | >20% labor and proprietor income from mining | 200 |
| Specialized government | >25% labor and proprietor income from government | 315 |
| Persistent poverty | Persistent low income county | 242 |
| Federal lands | >33% land federally owned | 247 |
| Destination retirement | >15% '70-'80 net immigration persons >60 years | 515 |
| Ungrouped | 370 |
*Adapted from Bender, et al. (1985), and Ross (1987).
Table 2
Counties with Inpatient Services, and Estimated 50th (Median) and 90th
Percentiles and Average Number of Beds per County in 1983 According to
Metropolitan Status, Adjacency of Nommetropolitan to Metropolitan Areas,
and
ERS Classification of Nonmetropolitan Counties
Total number of counties |
Percent of counties w/inpatient services |
50th Percentile of beds (Median) |
90th Percentile of Beds |
Average number of beds |
| Total | 3,137 | 24.7 | 0.0 | 57.0 | 69.8 |
| Metropolitan counties | 735 | 63.4 | 28.6 | 788.4 | 244.8 |
| Nonmetropolitan counties | 2,402 | 12.8 | 0.0 | 15.0 | 16.5 |
|
|||||
|
292 | 54.1 | 10.5 | 44.1 | 47.8 |
|
147 | 51.7 | 6.0 | 58.4 | 56.8 |
|
145 | 56.6 | 13.0 | 35.4 | 38.8 |
|
1,325 | 10.9 | 0.0 | 10.0 | 19.3 |
|
785 | 0.6 | 0.0 | 0.0 | 0.2 |
| ERS country classification | |||||
|
702 | 2.7 | 0.0 | 0.0 | 4.7 |
|
678 | 21.4 | 0.0 | 16.6 | 23.3 |
|
200 | 9.5 | 0.0 | 0.0 | 5.9 |
|
315 | 22.9 | 0.0 | 62.8 | 55.2 |
|
242 | 3.3 | 0.0 | 0.0 | 6.3 |
|
247 | 13.0 | 0.0 | 10.0 | 6.1 |
|
515 | 12.0 | 0.0 | 12.0 | 11.2 |
|
370 | 20.3 | 0.0 | 29.8 | 27.1 |
Table 3
Percentage of Counties With a Specified Type of Mental Health Organization
Providing Inpatient Psychiatric Services in 1983 According to Metropolitan
Status, Size of the Urban Population of ERS Classification of Nonmetropolitan
Counties
Percentage of counties with at least one facility type/percent of counties with inpatient services with at least one facility type |
Number of counties |
State and county hospitals |
Private psychiatric hospitals |
Non-federal general hospitals w/separate psychiatric service |
VA medical hospitals |
Multi service facilities* |
| Total | 3,137 | 7.1 | 4.8 | 19.1 | 6.1 | 3.7 |
| Metropolitan counties | 735 | 21.1 | 18.8 | 54.4 | 15.7 | 12.9 |
| Nonmetropolitan counties | 2,402 | 2.9 | 0.5 | 8.2 | 3.1 | 0.9 |
| Size of county urban populations | ||||||
|
292 | 6.2 | 3.1 | 44.1 | 13.4 | 3.4 |
|
1,325 | 3.8 | 0.3 | 5.1 | 2.6 | 0.8 |
|
785 | 0.3 | 0.0 | 0.3 | 0.1 | 0.0 |
| ERS county classification | ||||||
|
702 | 1.0 | 0.0 | 1.0 | 1.0 | 0.1 |
|
678 | 3.8 | 1.5 | 15.9 | 7.3 | 1.3 |
|
200 | 1.0 | 0.0 | 7.5 | 2.0 | 0.0 |
|
315 | 7.3 | 1.3 | 13.0 | 3.5 | 2.5 |
|
242 | 1.2 | 0.0 | 1.7 | 0.4 | 0.0 |
|
247 | 1.2 | 0.0 | 1.7 | 0.4 | 0.0 |
|
370 | 4.8 | 0.7 | 13.2 | 5.3 | 1.4 |
|
370 | 4.8 | 0.7 | 13.2 | 5.3 | 1.4 |
Footnotes
In conformity with one of its developers (Ross, 1987), we refer to it as the "ERS classification".
Correspondence concerning this article should be addressed to Morton 0. Wagenfeld, (wagenfeld@wmich.edu), Department of Sociology, Western Michigan University, Kalamazoo, MI, 49008-5189.
Original Journal Pages: 3-19.