Metropolitan-nonmetropolitan differences in amount and type of mental health treatment☆☆☆★

Article Outline

• Abstract
• Mental health treatment
• Methods
  • Data source and sample
  • Measures
  • Statistical analysis
• Analysis
• Discussion
• References
• Copyright

Abstract 

This article examines the extent to which, conditional on receiving treatment, the type of care differs across metropolitan and nonmetropolitan areas. Using data from the Medical Expenditure Panel Survey (MEPS), the findings indicate that nonmetro residents who obtained mental health care (n = 2,381) have fewer mental health visits in a calendar year than their metro counterparts after adjusting for individual-level characteristics. Although observed rates of hospitalization and contact with physicians are higher in nonmetro areas than metro areas, this difference is attributable primarily to compositional differences between metro and nonmetro residents. Copyright 2003, Elsevier Science (USA). All rights reserved.

Investigations based on small-scale and regional data show that rural residents receive less mental health treatment than their urban counterparts (Fortney, Thill, Zhang, Duan, & Rost, 2001; Fox, Merwin, & Blank, 1995; Petterson, Hauenstein, Rovnyak, 2002; Rost, Zhang, Fortney, Smith, & Smith, 1998; Shea, Russo, & Smyer, 2000). The lower rural use of mental health services persists even when availability, accessibility, demographic, and need factors are controlled (Gale, 1993; Seivewright, Tyrer, Casey, & Seivewright, 1991; Rost et al., 1998). Other studies have documented the scarcity of mental health providers and facilities in rural settings (Wagenfeld, Murray, Mohatt, & DeBruyn, 1997; Yuen, Gerdes, & Gonzales, 1996). These findings are at odds with results from the National Comorbidity Study (NCS) showing no difference between rural and urban residents in first treatment for mental health problems (Kessler, Olfson, & Berglund, 1998).

Rural communities have fewer mental health resources than their urban counterparts (Howland, 1995; Merwin, Goldsmith, & Manderscheid, 1995; Yuen et al., 1996). Although more than 50 million individuals—20% of the U.S. population—live in nonmetropolitan communities, just 9% of all physicians, mainly generalists, practice in these communities (Rosenblatt & Hart, 1999). Hence, the majority of mental health care in rural areas is provided in the general health sector (Fortney et al., 2001; Leaf, Bruce, Tischler, Freeman, Weissman, & Myers, 1988; Wells, Manning, Duan, Newhouse, & Ware, 1986; Yuen et al., 1996). The problems of access to care are evidently exacerbated by a general reluctance of rural dwellers to seek out mental health care, caused in part by the stigma placed on the use of formal mental health services, and to lack of anonymity (Fox et al., 1995; Howland, 1995). Rural individuals frequently perceive the mental health system as hostile and often consider mental health problems the domain of family and church (Hill & Fraser, 1995; Bushy, 1998).

This article examines the extent to which, conditional on receiving treatment, the type of care differs across metropolitan and nonmetropolitan areas. To this end, observed and adjusted estimates are presented for 3 characteristics of treatment: the number of mental health visits in a calendar year, hospitalization rates, and contact with a medical doctor. Data from the Medical Expenditure Panel Survey (MEPS) is used.

Back to Article Outline

Mental health treatment 

There is no clear-cut definition of what constitutes mental health treatment. A common approach is to define it as the care received by individuals with an identified mental health condition. This definition is broad in that individuals with mental health problems obtain care in the health care sector from a wide variety of specialists as well as nonspecialists including primary care physicians. Moreover, many individuals obtain care outside the health care sector from self-help groups, religious professionals, and lay healers (Katz, Kessler, Frank, Leaf, Lin, & Edlund, 1999; Kessler et al., 1999). At the same time, the earlier-described definition is narrow in that many individuals report that they obtain mental health care even though they do not report a mental health problem. ECA and NCS studies indicate that about 44% of those who received specialized or nonspecialized mental health treatment did not have diagnosable mental health problems (Alegria, Bijl, Lin, Walters, & Kessler, 2000; Katz et al., 1997; Kessler, Frank, Edlund, Katz, Lin, & Leaf, 1997; Kessler et al., 1999; Narrow, Regier, Rae, Manderscheid, & Locke, 1993; Regier, Narrow, Rae, Manderscheid, Locke, & Goodwin, 1993). However, individuals without a disorder receive fewer treatment visits than those with a disorder. Persons with a diagnosable disorder account for about two thirds of all mental health visits, whereas those without a disorder account for the remaining third (Kessler et al., 1999; Narrow et al.).

However defined, there is wide consensus that mental disorders are both untreated and undertreated in the United States, despite the availability of effective treatments for most psychiatric disorders (Mental Health, A Report of the Surgeon General, 1999). Data collected in the United States and Europe for the Global Burden of Disease Study (Murray & Lopez, 1996) show that 80% of those with schizophrenia receive treatment whereas only 35% of those with depression, 25% of those with panic disorder, and 15% of those with obsessive-compulsive disorders receive treatment for their psychiatric disorder.

Back to Article Outline

Methods 

Data source and sample 

The MEPS (Cohen, 2000) is the third in a series of national surveys designed to provide estimates of the use of health services, medical expenditures, and sources of payment including insurance coverage. Interviews are conducted with one member of each family who reports on the health care experiences for the entire family. Beginning with 1996, a new household MEPS sample is drawn each year from the respondents to the National Health Interview Survey of the preceding year. The joint National Health Interview Survey and MEPS response rates were 77.7% for the first panel (initially interviewed in 1996) and 77.9% for the second panel (initially interviewed in 1997). After a preliminary contact, the panel is interviewed 5 times over the next 2.5 years. MEPS respondents are restricted to the civilian noninstitutionalized population of the United States. Thus, the sample does not include a substantial number of users of mental health services provided in long-term or military psychiatric facilities, residential settings, and jails and prisons, among other residence-based facilities. The MEPS has a complex multistage sample design that uses stratification, cluster sampling, and oversampling of certain population groups.

In this analysis we use pooled 1996 and 1997 calendar year information from the MEPS. The 1996 data is from respondents from the first panel (those who were interviewed initially in 1996) and the 1997 data pools respondents from the first panel with those from the second panel (initially interviewed in 1997). Of the 53,781 cases with positive sample weights, we excluded 16,747 cases that were below the age of 18 and 6,222 cases that were above the age of 64, as well as an additional 260 cases because of missing information on at least one variable used in our analysis. We excluded elderly respondents because of near-universal insurance coverage with Medicare in this group. After excluding these cases, our sample consisted of 30,966 person-year observations, with 12,426 observations from 1996 and 18,540 from 1997; 24,271 person-year observations are metropolitan residents and 6,695 person-year observations are nonmetropolitan residents.

Measures 

Self-reported mental health problems were ascertained at each MEPS interview. Primary respondents were asked to identify their own and other household members' physical and mental health problems, whether treated or untreated, in the reference period before the interview. These conditions were recorded verbatim by the interviewer and subsequently categorized by trained coders into International Classification of Diseases 9th edition (ICD-9-CM) codes. For our purposes, we classify as mental health conditions those with ICD-9 codes between 290 and 315 as well as codes of 797 (“senility without psychosis”) and V40 (“mental/behavioral problem”).

Our definition of mental health treatment is similar to that used by Zuvekas (2001) in a recent analysis of MEPS data. First, we defined treatment for a mental health problem as any visit to a health care provider in a calendar year that the respondent reported was for a mental health problem, as defined earlier. Second, respondents were asked how their visit was “best described,” and also asked whether during the visit they received specific treatments. For our purposes, we defined specialized treatment as all visits characterized as “psychotherapy or mental health counseling,” or “drug or alcohol treatment,” or “psychotherapy/counseling.” We also consider any visits to a nonphysician mental health specialist—psychologist, social worker, or counselor—as specialized treatment. Note that one limitation of the MEPS data is that it is not possible to distinguish psychiatrists from other physicians. Here we define any type of mental health treatment as either treatment for a mental health problem or specialized treatment in the calendar year.

There is considerable overlap among these definitions. In our sample, we identified 2,381 cases with at least one visit of any type of mental health treatment during the calendar year and 1,773 cases with at least one visit of specialized treatment during the calendar year. Of those who received any type of mental health treatment, 1,146 obtained treatment for a reported mental health condition and specialized treatment; 608 reported a mental health condition but did not obtain specialized treatment; and 627 did not report a mental health condition but nevertheless obtained specialized treatment.

For each respondent who obtained any type of treatment, it is possible to determine the number of visits over the course of the calendar year. If a respondent was hospitalized, we counted each day in the hospital as a separate visit. Estimates of the mean number of visits are provided later. Because of the skewness of the distribution of this measure—with many respondents with just one visit and a long tail of a few respondents with many visits—we also use a categoric specification of the number of visits: 1, 2 to 4, 5 to 10, 11 to 20, and more than 20 visits. For the multivariate analysis, we follow Zuvekas (2001) and others and use a logarithmic transformation of number of visits.

We also examine 3 other aspects of mental health treatment. The first is whether or not the respondent was ever hospitalized in the course of the calendar year for treatment of a mental health problem. This is a dichotomous variable equal to one if hospitalized and zero otherwise. The second measure is equal to one if the respondent ever saw a medical doctor for treatment of a mental health problem and zero otherwise. The third measure is whether the respondent primarily saw a medical doctor for such treatment. For coding purposes, “primarily saw a doctor” is equal to one if more than half of the visits were with a medical doctor.

A metropolitan area is defined as the “core area (usually a county) containing a large population (densely settled) along with a set of adjacent communities that exhibit a high degree of economic and social integration with that core” (Butler and Beale, 1994). Self-reported mental health was obtained by asking, “How would you rate your overall mental health?” on a 5-point scale (responses ranged from “excellent” to “poor”). As reported in Table 1, there are relatively small metropolitan-nonmetropolitan differences in the proportions that report poor or fair mental health; however, a significantly smaller fraction of nonmetropolitan residents than metropolitan residents report excellent mental health, P < .01. The MEPS includes a parallel measure of general or physical health, also based on a 5-point scale ranging from excellent to poor. As reported in Table 1, nonmetropolitan respondents have slightly higher scores on this measure than metropolitan respondents, P < .01. Reports of perceived physical and mental health are highly correlated, r = .54, P < .001.

Table 1. Descriptive Statistics

Note. See text for definition of variables and sample.

In cases in which the respondent was not present at the interview, the primary respondents for the household survey were asked to report the mental health status of other household members. These proxy reports of overall mental health were obtained for a little less than 40% of the sample (Table 1). To control for potential bias, we included a dummy variable to flag these cases; “Self Report Flag” is equal to 1 for self-reports and 0 for proxy reports.

Our multivariate analyses included controls for several predisposing factors: sex, race/ethnicity, age, education, employment status, and marital status. Sex is coded 1 for women and 0 for men. We distinguished among non-Hispanic blacks, Hispanics, and non-Hispanic whites. This last group was primarily non-Hispanic white but included a few other respondents who are neither Hispanic nor black. Age was classified into 3 categories: 18 to 24, 25 to 44, and 45 to 64 years old. Education was a continuous variable equal to the number of years of schooling completed by the respondent. Employment was coded as one if the respondent was employed and zero otherwise. To minimize problems of endogeneity, all of these control variables were measured at the first interview during the calendar year. The descriptive results in Table 1 show that nonmetropolitan residents have less schooling, and are slightly less likely to be employed than their metropolitan counterparts. At the same time, nonmetropolitan residents are more likely to be married, and are older than metropolitan residents. Proportionately fewer non-Hispanic blacks and Hispanics live in nonmetropolitan areas.

We also included controls for enabling characteristics. Income-to-needs are calculated by dividing family income during the calendar year by the family's poverty line (based on family size and composition). The resulting percentages were grouped into 3 categories: poor or near poor (<125% of poverty line), low/middle (125% to 400% of poverty line), and high (400% of poverty line or higher). As shown in Table 1, nonmetropolitan respondents are more likely to be poor than metropolitan respondent. Health insurance is classified into 3 mutually exclusive categories, privately insured, publicly insured, and not insured. By using monthly insurance available in the MEPS we measured insurance as of January, the start of the calendar year. Individuals were considered privately insured if they were covered by any private insurance plan, including Medigap. Individuals were considered publicly insured if they were not privately insured and were covered by Medicare, Medicaid, Civilian Health and Medical Programs for the Uniformed Services or Civilian Health and Medical Programs for the Veterans Affairs. The uninsured were individuals without public or private insurance. Nonmetropolitan residents were less likely to have health insurance coverage than metropolitan residents, but slightly more likely to be covered by public insurance (Table 1).

Statistical analysis 

All analyses were performed using the svy or survey commands in STATA 7.0 (College Station, TX), which take into account the complex survey design of the MEPS, as well as the pooling of data across years. STATA uses linearization-based variance estimators that are appropriate for the design variables provided with the MEPS data. To test multidimensional hypotheses, we performed an adjusted Wald test, which uses the approximate F statistic (d − k + 1)W/(kd), where W is the Wald test statistic, k is the dimension of the hypothesis test, and d is the total number of sampled primary sampling units (P7SUs) minus the number of strata.

Back to Article Outline

Analysis 

The results in Table 2 show evidence of metropolitan-nonmetropolitan differences in mental health treatment rates, which compares proportions of persons receiving any type of mental health treatment and specialized treatment in a calendar year. These results are examined in greater detail in our companion report (Petterson, Hauenstein, & Rovnyak, 2002). Despite the poorer self-reported mental health of nonmetropolitan residents, just 6.9% obtained any type of mental heath care compared with 8.0% for their metropolitan counterparts, P < .05. The nonmetropolitan-metropolitan difference was substantially greater in the receipt of specialized care, 4.5% compared with 6.2%, P < .01.

Table 2. Metropolitan-Nonmetropolitan Differences in Mental Health Treatment by Reported Mental Health

Note. Nonmetropolitan-metropolitan difference: ^*P < .05.

Adjusted Wald Test is an overall test of the equality of treatment rates across metropolitan location by reported mental health (see text).

The probability of any mental health treatment increases as self-reported mental health declines: 2.9% of nonmetropolitan persons and 3.0% of metropolitan persons with reported excellent mental health obtained treatment, whereas nearly half of those with poor mental health obtained treatment. These rates are lower across all levels of reported mental health in nonmetropolitan areas compared with metropolitan areas. However, the metropolitan-nonmetropolitan disparity widens substantially as reported mental health changes from excellent to fair. A particularly striking finding is that just 23.5% of individuals in nonmetropolitan areas who reported fair mental health obtained any type of mental health treatment compared with 37.9% of their counterparts in metropolitan areas, P < .01. Likewise, nearly twice as many metropolitan residents with fair mental health obtained specialized treatment than their nonmetropolitan counterparts, P < .01. Interestingly, among the small number of respondents who report poor mental health (285 in metropolitan areas and 112 in nonmetropolitan areas), there is not a statistically significant difference in either the receipt of any type of mental health treatment or specialized treatment.

The results in Table 3 indicate that, on average, metropolitan residents have significantly more visits in a calendar year than their nonmetropolitan counterparts, P < .01. This difference also is evident in the distribution of visits. For instance, more than one third (33.7%) of nonmetropolitan residents report just one visit compared with 28.1% of metropolitan residents; at the other end of the distribution, 9.0% of nonmetropolitan residents report more than 20 visits compared with 13.1% of metropolitan residents. The results also show that there is not a statistically significant difference in likelihood of ever seeing a doctor (P = .15), but there is a significant observed difference in the likelihood of primarily seeing a doctor (P = .02). Specifically, 73.9% of nonmetropolitan residents obtained care primarily from a medical doctor compared with 68.5% of metropolitan residents. Finally, the Table 3 results show that nonmetropolitan residents are significantly more likely to be hospitalized for mental health problems than their metropolitan counterparts (P = .02).

Table 3. Observed Metropolitan-Nonmetropolitan Differences in Type of Mental Health Treatment

Table 4 displays estimates of metropolitan-nonmetropolitan differences in the number of visits (logged) for model alternative specifications.

Table 4. Multivariate Regression Analysis of the Relationship Between Metropolitan Location and Number of Visits (Logged)

Note. t statistic in parentheses (adjusted for complex survey design, see text). Nonmetropolitan-metropolitan difference: ^†P < .10.

Controls: Self-reported mental health (excellent, very good, good, fair, and poor) and proxy report of mental health. Demographic characteristics include physical health sex, age, race, region, highest grade completed, married, employed. Income/needs (<125%, 125% to 400%, and 400% of poverty line or higher); insurance (privately insured, publicly insured, and not insured).

The baseline model includes no controls, the second model adds self-reported mental health and an indicator variable for proxy reports, the third model adds demographic variables and the final model adds measures of income-to-needs ratio and insurance status. The results show that after adjusting for covariates, metropolitan residents who obtained care in a calendar year were still more likely to have more visits than their nonmetropolitan counterparts. The decline in the magnitude of the coefficient—from 0.215 in model A to 0.141 in model D—indicates that some, but not all, of the metropolitan-nonmetropolitan difference is attributable to compositional differences across metropolitan location. In the final model, this difference is marginally significant (P = .07).

Table 5 displays parallel estimates from nested logistic regression models for (1) whether the respondent was ever hospitalized in the calendar year, (2) whether the respondent ever saw a medical doctor, and (3) whether the respondent primarily saw a medical doctor.

Table 5. Multivariate Logistic Analysis of the Relationship Between Metropolitan Location and Ever Hospitalized, Ever Saw A Doctor, and Primarily Saw A Doctor, Odds Ratios

Note. Nonmetropolitan-metropolitan difference: ^*P < .10. T-statistic in parentheses (adjusted for complex survey design, see text).

Controls: Self-reported mental health (excellent, very good, good, fair, and poor) and proxy report of mental health. Demographic characteristics include physical health, sex, age, race, region, highest grade completed, married, employed. Income/needs (<125%, 125% to 400% and 400% of poverty line or higher); insurance (privately insured, publicly insured, and not insured).

The controls in each model are the same as in the earlier analysis. Odds ratios are reported for ease of presentation. Consistent with the results reported in Table 3, the estimates presented for the baseline model show that the metropolitan residents were less likely to be hospitalized and less likely to primarily see a doctor; the estimate metropolitan-nonmetropolitan difference for ever seeing a doctor is not statistically significant (P = .20). After controlling for covariates, particularly the demographic factors in model C, the magnitude of the difference declines and the estimates for ever hospitalized and primarily saw a doctor are no longer statistically significant. In results not reported here, the main covariate accounting for the decline in the metropolitan coefficient are levels of schooling.

Back to Article Outline

Discussion 

There are observed metropolitan-nonmetropolitan differences in the type of mental health treatment received by nonelderly adults. Persons in nonmetropolitan areas who obtain any type of treatment have fewer visits than their metropolitan counterparts. They also are more likely to ever see a medical doctor for their care and especially more likely to primarily see a medical doctor. Finally, nonmetropolitan residents are more likely to be hospitalized.

The multivariate results indicate that the metropolitan-nonmetropolitan difference in number of visits for mental health treatment persists after adjusting for a number of covariates. This result is consistent with the findings that nonmetropolitan residents are less likely to obtain specialized treatment than metropolitan residents (see Table 2, Petterson, Hauenstein, & Rovnyak, 2002). This follows from the fact that specialized treatment involves more extended treatment, and more visits, than nonspecialized treatment.

The multivariate results for hospitalization rates and the likelihood of ever or primarily seeing a doctor are more ambiguous. They show that to a considerable extent metropolitan-nonmetropolitan differences in these facets of treatment are explained by different characteristics of metropolitan and nonmetropolitan residents, particularly education.

This study has several limitations that must be addressed in future research. First, as noted earlier, the MEPS is restricted to the civilian, noninstitutionalized population and thus does not include many users of mental health services who reside in long-term residence or military facilities, including psychiatric hospitals, prisons, and jails. In addition, the dichotomous measure of rurality available in the MEPS—metropolitan location—does not capture considerable heterogeneity among localities (Ricketts, Johnson-Webb, 1997; Ricketts, Johnson-Webb, & Taylor, 1998). This study also is limited in its reliance on proxy self-reports of mental health and mental health conditions.

Back to Article Outline

References 

1. Alegria M, Bijl R, Lin E, Walters E, Kessler RC. Income differences in persons seeking outpatient treatment for mental disorders: A comparison of the United States with Ontario and the Netherlands. Archives of General Psychiatry. 2000;57(4):383–391
   • View In Article
   • MEDLINE
   • CrossRef
2. Bushy A. Health issues of women in rural environments: An overview. Journal of the American Medical Women's Association. 1998;53(2):53–56
   • View In Article
3. Butler MA, Beale CL. Rural-Urban Continuum Codes for Metro and Nonmetro Counties, 1993. USDA Economic Research Service Report AGES 9425 Washington, DC: U. S. Department of Agriculture; 1994;
   • View In Article
4. Cohen SB. Sample Design of the 1997 Medical Expenditure Panel Survey Household Component. MEPS Methodology Report #11. AHCPR Pub. No. 01-0001 Rockville, MD: Agency for Health Care Policy and Research; 2000;
   • View In Article
5. Fortney J, Thill J, Zhang M, Duan N, Rost K. Provider choice and utility loss due to selective contracting in rural and urban areas. Medical Care Research and Review. 2001;58(1):60–75
   • View In Article
6. Fox JC, Merwin E, Blank M. Characterizing the de facto mental health service system in the rural south: A review of essential data for health care reform. Health Care for the Poor and Underserved. 1995;6(4):434–468
   • View In Article
7. Gale BJ. Psychosocial health needs of older women: Urban versus rural comparisons. Archives of Psychiatric Nursing. 1993;7(2):99–105
   • View In Article
   • MEDLINE
   • CrossRef
8. Hill CE, Fraser GJ. Local knowledge and rural mental health reform. Community Mental Health Journal. 1995;31(6):553–568
   • View In Article
   • MEDLINE
   • CrossRef
9. Howland RH. The treatment of persons with a dual diagnoses in a rural community. Psychiatric Quarterly. 1995;66(1):33–49
   • View In Article
10. Katz SJ, Kessler RC, Frank RG, Leaf P, Lin E, Edlund M. The use of outpatient mental health services in the United States and Ontario: The impact of mental morbidity and perceived need for care. American Journal of Public Health. 1997;87(7):1136–1143
    • View In Article
    • MEDLINE
    • CrossRef
11. Kessler RC, Frank RG, Edlund M, Katz SJ, Lin E, Leaf P. Differences in the use of psychiatric outpatient services between the United States and Ontario. New England Journal of Medicine. 1997;8(8):551–557
    • View In Article
12. Kessler RC, Olfson M, Berglund PA. Patterns and predictors of treatment contact after first onset of psychiatric disorders. American Journal of Psychiatry. 1998;155(1):62–69
    • View In Article
13. Kessler RC, Zhao S, Katz SJ, Kouzis AC, Frank RG, Edlund M, et al. Past-year use of outpatient services for psychiatric problems in the National Comorbidity Survey. American Journal of Psychiatry. 1999;156(1):115–123
    • View In Article
14. Leaf PJ, Bruce ML, Tischler GL, Freeman DH, Weissman MM, Myers JK. Factors affecting the utilization of specialty and general medical mental health services. Medical Care. 1988;26:9–26
    • View In Article
    • MEDLINE
    • CrossRef
15. Mental Health, A Report of the Surgeon General. Available http://www.surgeongeneral.gov/1999; Accessed January 13, 2003
    • View In Article
16. Merwin EI, Goldsmith HF, Manderscheid RW. Human resource issues in rural mental health services. Community Mental Health Journal. 1995;31(6):525–537
    • View In Article
    • MEDLINE
    • CrossRef
17. Murray CJ, Lopez AD. Evidence-based health policy—lessons from the global burden of disease study. Science. 1996;274(5288):740–743
    • View In Article
    • MEDLINE
    • CrossRef
18. Narrow WE, Regier DA, Rae DS, Manderscheid RW, Locke BZ. Use of services by persons with mental and addictive disorders: Findings from the National Institute of Mental Health Epidemiologic Catchment Area Program. Archives of General Psychiatry. 1993;50(2):95–107
    • View In Article
    • MEDLINE
19. Petterson SM, Hauenstein EJ, Rovnyak VG. Metro-Nonmetro differences in Mental Health Treatment. Unpublished manuscript 2002;
    • View In Article
20. Regier D, Narrow WE, Rae D, Manderscheid R, Locke B, Goodwin F. The de-facto US Mental and Addictive Disorders Service System: Epidemiologic area catchment area prospective one-year prevalence rate of disorders and services. Archives of General Psychiatry. 1993;50:85–92
    • View In Article
    • MEDLINE
21. Ricketts TC, Johnson-Webb KD. What is “Rural” and How to Measure “Rurality”: A Focus on Health Care Delivery and Health Policy. Technical Issues Paper prepared for the Federal Office of Rural Health Policy Health Resources and Service Administration; 1997;
    • View In Article
22. Rosenblatt RA, Hart LG. Physicians and rural America. In:  Ricketts TC editors. Rural Health in the United States. Cambridge: Oxford University Press; 1999;p. 38–51
    • View In Article
23. Rost K, Zhang M, Fortney J, Smith J, Smith GR. Rural-urban differences in depression treatment and suicidality. Medical Care. 1998;36(7):1098–1107
    • View In Article
    • MEDLINE
    • CrossRef
24. Seivewright H, Tyrer P, Casey P, Seivewright N. A three-year follow-up of psychiatric morbidity in urban and rural primary care. Psychological Medicine. 1991;21(2):495–503
    • View In Article
    • MEDLINE
    • CrossRef
25. Shea DG, Russo PA, Smyer MA. Use of mental health services by persons with a mental illness in nursing facilities: Initial impacts of OBRA87. Journal of Aging and Health. 2000;12(4):560–578
    • View In Article
    • MEDLINE
    • CrossRef
26. Wagenfeld MO, Goldsmith HF, Stiles D, Manderscheid RW. Inpatient mental health services in metropolitan and nonmetropolitan communities. Journal of Rural Community Psychology. 1988;9:13–28
    • View In Article
27. Wagenfeld MO, Murray JD, Mohatt DF, DeBruyn JC. Mental health service delivery in rural areas: Organizational and clinical issues. NIDA Research Monograph. 1997;168:418–437
    • View In Article
    • MEDLINE
28. Wells KB, Manning WG, Duan N, Newhouse JP, Ware JE. Sociodemographic factors and the use of outpatient mental health services. Medical Care. 1986;24(1):75–85
    • View In Article
    • MEDLINE
    • CrossRef
29. Yuen EJ, Gerdes JL, Gonzales JJ. Patterns of rural mental health care: An exploratory study. General Hospital Psychiatry. 1996;18(1):14–21
    • View In Article
    • Abstract
    • Full-Text PDF (913 KB)
    • CrossRef
30. Zuvekas SH. Trends in mental health services use and spending, 1987-1996. Health Affairs. 2001;20:214–224 spaces
    • View In Article