County Context and Mental Health Service Utilization by Older Hispanics

County Context and Mental Health Service Utilization by Older Hispanics Abstract Background and Objectives Although older Hispanics experience high rates of depression, they tend to underuse mental health services. The study examined the association between county characteristics and mental health service use among older Hispanics, controlling for individual characteristics. Research Design and Methods The study used the 2008–2012 Medical Expenditure Panel Study and linked county-level data from the 2013–2014 Area Health Resources Files and the 2008–2012 Chronic Conditions Data Warehouse, using the Federal Information Processing Standard county code. The sample includes 1,143 community-dwelling Hispanics ages 60 years or older (Level 1) and 156 counties (Level 2) where the sample resides. The single dichotomous measure of mental health service utilization was based on whether or not the respondent met one or more of three conditions: (1) the respondent received care from a mental health professional, (2) received a service including mental health counseling or psychotherapy, or (3) received a service that was related to the International Classification of Diseases. Multilevel logistic regression analysis was used to examine the role of county context. Results The proportion of older adults and the existence of community mental health centers at the county-level were associated with mental health services use among this population. At the individual-level, education and mental health status were also associated with using mental health services. Discussion and Implications The county context plays an important role in understanding mental health services use among older Hispanics, indicating the need for intervention strategies at the county level. County context, Mental health services, Multilevel modeling, Older Hispanics Underutilization of mental services by older Hispanics is a growing concern in the United States. Underutilization may delay the treatment of mental health problems, and untreated problems may have a huge impact on the well-being of an individual. For example, untreated depression leads to disease burden and disability (Fuentes & Aranda, 2012) and may result in higher suicide and nonsuicide mortality (National Alliance on Mental Illness [NAMI], 2009). Although previous studies using national data (e.g., National Latino and Asian American Study) have advanced understanding of mental health service utilization among older Hispanics, they have not explored the socioenvironmental context in which this population accesses mental health services; most existing research on this topic has focused on individual-level variables. The number of adults aged 65 years and older in the United States with mental health problems is growing, and approximately 14%–20% of older adults are estimated to have mental illnesses such as depression and schizophrenia (Institute of Medicine [IOM], 2012). In addition, the overall population of older Hispanics is expected to grow from about 3.6 million (9% of the U.S. older population) in 2014 to 21.5 million (22% of the U.S. older population) in 2060 (Federal Interagency Forum on Aging-Related Statistics, 2016). It is likely that there will be a corresponding increase in the number of older Hispanics with mental health problems (Mackenzie, Pagura, & Sareen, 2010). A study using the National Latino and Asian American Survey found that 1 in 10 (9.7%) older Hispanics were diagnosed with anxiety disorders in the past 12 months, but only 5.5% used mental health services (Kim, Jang, Chiriboga, Ma, & Schonfeld, 2010). Another study, the 2005 California Health Interview Survey found that only 10.9% of older Hispanics received mental health services although 24% reported mental illnesses (Sorkin, Pham, & Ngo-Metzger, 2009). Older Hispanics are less likely to use professional mental health services compared to Whites because they face individual, cultural, linguistic, and systemic barriers to accessing mental health services as other ethnic minority populations do (Akincigil et al., 2012; SAMHSA, 2011). Older Hispanics have negative perceptions of seeking formal mental health services because of experiences of racial discrimination and stresses from immigration-related issues (Alvarez, Rengifo, Emrani, & Gallagher-Thompson, 2014). Education and economic resources are very important in utilization of mental health services (Gonzalez, Applewhite, & Barrera, 2015; Sorkin, Murphy, Nguyen, & Biegler, 2016). Older Hispanics underutilize mental health services in part because they are less likely to have a high school diploma and are more likely to live below the 150% poverty line compared to other racial and ethnic groups (Akincigil et al., 2012). Mental health service utilization by older Hispanics has been studied by many researchers, and reducing racial disparities in health care services has been identified as a national priority (Kim et al., 2013; U.S. Department of Health and Human Services [HHS], 2018). To understand facilitators of and barriers to the use of mental health services among older Hispanics, most research has focused on individual characteristics such as age, gender, education, income, and health and mental health status (Akincigil et al., 2012; Alvarez et al., 2014; Kim et al., 2010). A limited number of studies have focused on older adults as their study population to explain the relationships between socioenvironmental characteristics and mental health services utilization (e.g., Kaskie, Gilder, & Gregory, 2008; Kim et al., 2013; Wei, Sambamoorthi, Olfson, Walkup, & Crystal, 2005). Thus, this study aimed to examine mental health service utilization by older Hispanics. Theoretical Framework Employing Andersen’s (2008) Behavioral Model of Health Services Utilization as a theoretical framework, this study hypothesized that county contexts are associated with mental health service use by older Hispanics, controlling for individual characteristics. Andersen’s model is very useful for addressing this study’s research questions regarding mental health service utilization because it explains that the contextual predisposing (e.g., community demographic and social characteristics), enabling (e.g., supply of health care services), and need characteristics (e.g., depression rate) are associated with the use of health services (Andersen, 2008; Andersen & Davidson, 2007). Geographic differences may be an important contributing factor to racial and ethnic disparities in mental health (Kim et al., 2013). As Andersen and Davidson (2007) suggest, contextual characteristics, such as neighborhood sociodemographic characteristics, health care resources, and population health indices—in addition to individual characteristics—impact the use of mental health services and health outcomes. Health care resources such as mental health professionals and community mental health centers may be unequally allocated in different geographic areas (Cook, Doksum, Chen, Carle, & Alegría, 2013). In addition, older Hispanics are more likely to live in economically disadvantaged areas that have fewer health care resources and may have fewer opportunities to access mental health care (Cook et al., 2013). Defining Geographic Boundaries Defining a geographic boundary is important because geographic differences may be an important contributing factor to racial and ethnic disparities in mental health (Kim et al., 2013). This study used counties as a geographic boundary. Although census tracts or blocks may be more socioeconomically homogeneous than other geographic units, counties are more sociopolitically and geographically stable compared to census tracts or blocks (Singh, 2003). The counties are also the smallest geographical entity to provide data on varieties of health care and social services within states as well as to consistently provide health, population, and socioeconomic statistics over time (DeFranco, Lian, Muglia, & Schootman, 2008). Mental health care resources, such as number of mental health professionals and community mental health centers vary across counties, and there is more variability in mental health services use at the county-level compared to other area levels. Based on Andersen’s model and previous empirical studies, the purpose of the study is to examine what county context (i.e., county predisposing, enabling, and need factors) as well as individual-level characteristics are associated with the use of mental health services by older Hispanics. An important premise of this study is that an understanding of the relationship between county-level context and mental health services utilization in this population is key to developing effective intervention strategies at the macro level. Methods Sample The study used the 2008–2012 Medical Expenditure Panel Survey (MEPS) and linked county-level data from the 2013–2014 Area Health Resources Files (AHRF) and the 2008–2012 Chronic Conditions Data Warehouse (CCW). The MEPS is a national survey to estimate health care use, health care insurance coverage, and access to health care and quality for the U.S. civilian noninstitutionalized population (Agency for Healthcare Research and Quality, 2013). The study employed five-year MEPS individual-level data (panels 13–16) because most county data are based on 2010 census data or on 2008–2012 averages. Combining multiyear data has been used by researchers to provide a sufficient sample size for racial and ethnic groups (e.g., Choi, 2015; Jimenez, Cook, Bartels, & Alegría, 2013). This study also employed the 2013–2014 AHRF to provide information on county contexts; the AHRF contains information on county-level demographic and socioeconomic status, and health care resources such as primary care physician shortage areas and community mental health center. The third data set, the CCW, provided information on depression prevalence rates at the county level. This study selected 1,143 older Hispanics who were ages 60 years or older and who completed the survey by themselves from the MEPS. The MEPS was merged with the AHRF and the CCW using county identifiers (i.e., the Federal Information Processing Standard county code). Based on the data linkage, the level two sample of the study is 156 counties where 1,143 older Hispanics reside. Measures This study used individual characteristics from the MEPS and county characteristics from the AHRF and the CCW. Mental Health Services Utilization Mental health services utilization was defined as actual provider visits for mental health care. Using the Hospital Outpatient Visit Files and the Office-Based Medical Provider Visits Files of MEPS, there were three ways of identifying mental health service use: types of providers, types of treatments, and the International Classification of Diseases, ninth revision (ICD-9) codes. If at least one of the following conditions was met, it was considered a mental health service use: (1) the person received care from a psychologist or a social worker, (2) the type of treatment was psychotherapy or mental health counseling, or (3) a provider visit was coded with the clinical classification codes (CCC) associated with a mental health condition. The CCC are created by aggregating the ICD-9 codes to have clinically meaningful categories, and the CCCs of 650 through 670 refer to mental health conditions. Mental health services utilization was coded as 1 = yes or 0 = no. Individual Characteristics This study included individual characteristics suggested from Andersen’s (2008) model and previous literature. This study included individual predisposing characteristics (i.e., age in years, gender [female/male], married [yes], high school diploma [yes], foreign born [yes], language spoken at home [Spanish/English], and attitude toward health). As enabling characteristics, this study included health insurance (yes) and income, and also included health status from the Short Form 12 Version 2 (SF-12v2) (Ware, Kosinski, & Keller, 1996) and the Kessler Psychological Distress Scale (K6) (Kessler et al., 2002) as the need factor. All individual data were obtained from the MEPS. County Context To measure county context, this study used county predisposing (i.e., proportion of older adults, percent Hispanic population, metropolitan area (yes), percentage of those with high school diploma, unemployment rates) and county enabling (i.e., existence of community mental health center, primary care physician and mental health professional shortage area, mental health professionals per 100,000, and median home values) drawn from the AHRF. The study also included depression prevalence rates at the county-level from the CCW. Data Analysis Plan To answer the research questions, multilevel modeling analysis using Stata 14 software (StataCorp, 2015) was conducted to examine the role of county context on mental health services utilization by older Hispanics, controlling for individual characteristics. The MEPS uses a complex survey design that necessitated the need for weights: analyses, therefore, included the use of the Taylor series linearization method to provide an accurate estimation of the variance structure and standard errors, with the strata and the PSU specified (AHRQ, 2013). First, the unconditional model was tested to examine county-to-county variability of mental health service utilization. This can be done by calculating the unconditional intraclass correlation (ICC) (Bickel, 2007). Second, individual predisposing, enabling, and need factors (Level 1) were added to the model to examine the relationships between individual-level variables and mental health service use. Third, county contexts (Level 2) were added to the previous model to examine whether or not the county contexts are predictive of mental health service use among older adults, independent of individual characteristics. This study involved the use of restricted data (i.e., FIPS codes), so a proposal was submitted to and approved by the Agency for Healthcare Research and Quality (AHRQ) as well as a University Institutional Review Board in the mid-Atlantic area. Data analyses using the restricted data were conducted at the AHRQ Data Center in Rockville, Maryland. Results Table 1 presents the individual-level sample characteristics of 1,143 older Hispanics (a population estimate of 2,736,013 older Hispanics). Over half were women (55%) or married (53%); the average age of the sample was 69 years old (SD = 9.9). Less than half of the Hispanic sample (45%) earned a high school diploma. Almost two-thirds were born in a foreign country (63%) and spoke Spanish at home (60%). Regarding ethnicity, the majority of the sample was Mexican (53%), followed by Central or South American (13%), Puerto Rican (12%), Cuban (9%), Dominican (5%), and other (7%). About 6% of the sample had a mental health service visit during the data collection period. Table 1. Individual-level Hispanic Sample Characteristics Continuous variables Weighted M (SD) N = 2,736,013 Unweighted M (SD) N = 1,143 Unweighted range Age 69.28 (9.87) 68.56 (7.10) 60–85* Attitudes toward health 1.89 (1.17) 1.90 (0.87) 1–5 SF-12 (Physical health) 42.80 (16.06) 42.58 (11.88) 11.65–63.77 SF-12 (Mental health) 49.83 (14.08) 48.81 (10.58) 13.92–70.48 K6 4.29 (7.14) 4.68 (5.64) 0–24 Categorical variables Weighted n (%) Unweighted n (%) Mental health service visit 168,538 (6.16) 78 (6.82) Gende  Female 1,507,269 (55.09) 632 (55.29)  Male 1,228,743 (44.91) 511 (44.71) Race  White 2,618,090 (95.69) 1,091 (95.45)  Black 50,616 (1.85) 15 (2.19)  American Indian 29,275 (1.07) 13 (1.14) Married (yes) 1,453,643 (53.13) 636 (55.64) High school diploma (yes) 1,243,244 (45.44) 430 (37.62) Foreign born (yes) 1,728,339 (63.17) 778 (68.07) Language spoken at home  English 1,108,085 (40.5) 383 (33.51)  Spanish 1,627,927 (59.5) 760 (66.49) Income  Poor 583,318 (21.32) 270 (23.62)  Near poor 214,230 (7.83) 114 (9.97)  Low income 617,792 (22.58) 261 (22.83)  Middle income 833,389 (30.46) 333 (29.13)  High income 487,284 (17.81) 165 (14.44) Uninsured (yes) 371,277 (13.57) 190 (16.62) Continuous variables Weighted M (SD) N = 2,736,013 Unweighted M (SD) N = 1,143 Unweighted range Age 69.28 (9.87) 68.56 (7.10) 60–85* Attitudes toward health 1.89 (1.17) 1.90 (0.87) 1–5 SF-12 (Physical health) 42.80 (16.06) 42.58 (11.88) 11.65–63.77 SF-12 (Mental health) 49.83 (14.08) 48.81 (10.58) 13.92–70.48 K6 4.29 (7.14) 4.68 (5.64) 0–24 Categorical variables Weighted n (%) Unweighted n (%) Mental health service visit 168,538 (6.16) 78 (6.82) Gende  Female 1,507,269 (55.09) 632 (55.29)  Male 1,228,743 (44.91) 511 (44.71) Race  White 2,618,090 (95.69) 1,091 (95.45)  Black 50,616 (1.85) 15 (2.19)  American Indian 29,275 (1.07) 13 (1.14) Married (yes) 1,453,643 (53.13) 636 (55.64) High school diploma (yes) 1,243,244 (45.44) 430 (37.62) Foreign born (yes) 1,728,339 (63.17) 778 (68.07) Language spoken at home  English 1,108,085 (40.5) 383 (33.51)  Spanish 1,627,927 (59.5) 760 (66.49) Income  Poor 583,318 (21.32) 270 (23.62)  Near poor 214,230 (7.83) 114 (9.97)  Low income 617,792 (22.58) 261 (22.83)  Middle income 833,389 (30.46) 333 (29.13)  High income 487,284 (17.81) 165 (14.44) Uninsured (yes) 371,277 (13.57) 190 (16.62) Note: *Everyone aged 85 years and older was coded as 85 to maintain the confidentiality of the data. View Large Table 1. Individual-level Hispanic Sample Characteristics Continuous variables Weighted M (SD) N = 2,736,013 Unweighted M (SD) N = 1,143 Unweighted range Age 69.28 (9.87) 68.56 (7.10) 60–85* Attitudes toward health 1.89 (1.17) 1.90 (0.87) 1–5 SF-12 (Physical health) 42.80 (16.06) 42.58 (11.88) 11.65–63.77 SF-12 (Mental health) 49.83 (14.08) 48.81 (10.58) 13.92–70.48 K6 4.29 (7.14) 4.68 (5.64) 0–24 Categorical variables Weighted n (%) Unweighted n (%) Mental health service visit 168,538 (6.16) 78 (6.82) Gende  Female 1,507,269 (55.09) 632 (55.29)  Male 1,228,743 (44.91) 511 (44.71) Race  White 2,618,090 (95.69) 1,091 (95.45)  Black 50,616 (1.85) 15 (2.19)  American Indian 29,275 (1.07) 13 (1.14) Married (yes) 1,453,643 (53.13) 636 (55.64) High school diploma (yes) 1,243,244 (45.44) 430 (37.62) Foreign born (yes) 1,728,339 (63.17) 778 (68.07) Language spoken at home  English 1,108,085 (40.5) 383 (33.51)  Spanish 1,627,927 (59.5) 760 (66.49) Income  Poor 583,318 (21.32) 270 (23.62)  Near poor 214,230 (7.83) 114 (9.97)  Low income 617,792 (22.58) 261 (22.83)  Middle income 833,389 (30.46) 333 (29.13)  High income 487,284 (17.81) 165 (14.44) Uninsured (yes) 371,277 (13.57) 190 (16.62) Continuous variables Weighted M (SD) N = 2,736,013 Unweighted M (SD) N = 1,143 Unweighted range Age 69.28 (9.87) 68.56 (7.10) 60–85* Attitudes toward health 1.89 (1.17) 1.90 (0.87) 1–5 SF-12 (Physical health) 42.80 (16.06) 42.58 (11.88) 11.65–63.77 SF-12 (Mental health) 49.83 (14.08) 48.81 (10.58) 13.92–70.48 K6 4.29 (7.14) 4.68 (5.64) 0–24 Categorical variables Weighted n (%) Unweighted n (%) Mental health service visit 168,538 (6.16) 78 (6.82) Gende  Female 1,507,269 (55.09) 632 (55.29)  Male 1,228,743 (44.91) 511 (44.71) Race  White 2,618,090 (95.69) 1,091 (95.45)  Black 50,616 (1.85) 15 (2.19)  American Indian 29,275 (1.07) 13 (1.14) Married (yes) 1,453,643 (53.13) 636 (55.64) High school diploma (yes) 1,243,244 (45.44) 430 (37.62) Foreign born (yes) 1,728,339 (63.17) 778 (68.07) Language spoken at home  English 1,108,085 (40.5) 383 (33.51)  Spanish 1,627,927 (59.5) 760 (66.49) Income  Poor 583,318 (21.32) 270 (23.62)  Near poor 214,230 (7.83) 114 (9.97)  Low income 617,792 (22.58) 261 (22.83)  Middle income 833,389 (30.46) 333 (29.13)  High income 487,284 (17.81) 165 (14.44) Uninsured (yes) 371,277 (13.57) 190 (16.62) Note: *Everyone aged 85 years and older was coded as 85 to maintain the confidentiality of the data. View Large Table 2 presents the county-level characteristics for the 156 counties where the 1,143 older Hispanic respondents resided. In the 156 counties included in this sample, an average of 2 in 10 persons were Hispanic and an average of 19% of the county residents were older adults ages 60 years and over. About 9 in 10 (90%) of counties where the Hispanic sample resided were metropolitan counties, and the median home value was about $249,000. The average depression rate was 13% for all counties, and less than half of the counties (44%) had a community mental health center. On average, each county had 47 mental health professionals per 100,000 county residents. Table 2. County-level Characteristics (N = 156) Continuous variables M (SD) Range Ages 60 years and over (%) 18.99 (6.13) 9.24–56.19 Proportion of Hispanic (%) 21.75 (16.81) 2.30–90.60 High school diploma (%) 56.68 (6.71) 33.46–76.09 Unemployment (%) 8.35 (3.27) 2.80–28.20 Mental health professionals (per 100,000) 43.63 (43.71) 0–306.08 Median home value ($10,000) 24.92 (14.95) 5.73–82.73 Depression prevalence (%) 13.46 (2.87) 6.56–26.47 Categorical variables n (%) Metropolitan county (yes) 140 (89.74) Community mental health center (yes) 69 (44.23) PCP shortage area  None of the county 18 (11.54)  One or more parts of the county 78 (50.00)  The whole county 60 (38.46) Mental health professional shortage area  None of the county 22 (14.10)  One or more parts of the county 81 (51.92)  The whole county 53 (33.97) Continuous variables M (SD) Range Ages 60 years and over (%) 18.99 (6.13) 9.24–56.19 Proportion of Hispanic (%) 21.75 (16.81) 2.30–90.60 High school diploma (%) 56.68 (6.71) 33.46–76.09 Unemployment (%) 8.35 (3.27) 2.80–28.20 Mental health professionals (per 100,000) 43.63 (43.71) 0–306.08 Median home value ($10,000) 24.92 (14.95) 5.73–82.73 Depression prevalence (%) 13.46 (2.87) 6.56–26.47 Categorical variables n (%) Metropolitan county (yes) 140 (89.74) Community mental health center (yes) 69 (44.23) PCP shortage area  None of the county 18 (11.54)  One or more parts of the county 78 (50.00)  The whole county 60 (38.46) Mental health professional shortage area  None of the county 22 (14.10)  One or more parts of the county 81 (51.92)  The whole county 53 (33.97) Note: PCP = Primary Care Health Professional. View Large Table 2. County-level Characteristics (N = 156) Continuous variables M (SD) Range Ages 60 years and over (%) 18.99 (6.13) 9.24–56.19 Proportion of Hispanic (%) 21.75 (16.81) 2.30–90.60 High school diploma (%) 56.68 (6.71) 33.46–76.09 Unemployment (%) 8.35 (3.27) 2.80–28.20 Mental health professionals (per 100,000) 43.63 (43.71) 0–306.08 Median home value ($10,000) 24.92 (14.95) 5.73–82.73 Depression prevalence (%) 13.46 (2.87) 6.56–26.47 Categorical variables n (%) Metropolitan county (yes) 140 (89.74) Community mental health center (yes) 69 (44.23) PCP shortage area  None of the county 18 (11.54)  One or more parts of the county 78 (50.00)  The whole county 60 (38.46) Mental health professional shortage area  None of the county 22 (14.10)  One or more parts of the county 81 (51.92)  The whole county 53 (33.97) Continuous variables M (SD) Range Ages 60 years and over (%) 18.99 (6.13) 9.24–56.19 Proportion of Hispanic (%) 21.75 (16.81) 2.30–90.60 High school diploma (%) 56.68 (6.71) 33.46–76.09 Unemployment (%) 8.35 (3.27) 2.80–28.20 Mental health professionals (per 100,000) 43.63 (43.71) 0–306.08 Median home value ($10,000) 24.92 (14.95) 5.73–82.73 Depression prevalence (%) 13.46 (2.87) 6.56–26.47 Categorical variables n (%) Metropolitan county (yes) 140 (89.74) Community mental health center (yes) 69 (44.23) PCP shortage area  None of the county 18 (11.54)  One or more parts of the county 78 (50.00)  The whole county 60 (38.46) Mental health professional shortage area  None of the county 22 (14.10)  One or more parts of the county 81 (51.92)  The whole county 53 (33.97) Note: PCP = Primary Care Health Professional. View Large Table 3 presents analyses from multilevel logistic regression analyses to examine whether county-level characteristics were associated with mental health service utilization by older Hispanics, controlling for individual characteristics. First, the unconditional model was tested to examine the county-to-county variability of mental health service utilization. The ICC was 0.156, indicating that the use of mental health services differed from county to county; that is, 15.6% of mental health service utilization by older Hispanics can be explained by county membership, indicating that employing a multilevel modeling approach was appropriate. The Akaike information criteria (AIC) was 567.302, which was used as the baseline information criterion for model fit. Table 3. Multilevel Logistic Regression for Mental Health Service Utilization Parameter Model 1 Model 2 Individual only Individual and County OR SE p CIL CIH OR SE p CIL CIH Intercept 7.48 15.48 .331 0.13 431.98 0.62 2.4 .902 0 1,166.62 Level 1 (individual-level) Age 0.96 0.02 .113 0.92 1.01 0.96 0.02 .102 0.92 1.01 Female 1.93 0.62 .041 1.03 3.64 1.89 0.62 .051 0.996 3.58 Married 0.55 0.17 .055 0.3 1.01 0.61 0.19 .116 0.32 1.13 High school diploma 2.24 0.72 .012 1.19 4.2 2.17 0.71 .018 1.14 4.12 Foreign born 1.54 0.61 .28 0.7 3.36 1.53 0.64 .305 0.68 3.47 No English spoken in home 0.65 0.25 .249 0.31 1.36 0.65 0.25 .26 0.3 1.38 Attitudes toward health 0.85 0.16 .389 0.59 1.23 0.84 0.16 .367 0.58 1.22 Income 0.91 0.1 .399 0.73 1.13 0.9 0.1 .367 0.73 1.13 Uninsured 0.66 0.31 .373 0.26 1.65 0.74 0.35 .529 0.29 1.89 SF-12 (Physical health) 0.998 0.01 .911 0.97 1.03 0.999 0.01 .942 0.97 1.03 SF-12 (Mental health) 0.95 0.02 .004 0.91 0.98 0.95 0.02 .005 0.91 0.98 K6 1.07 0.03 .035 1.004 1.14 1.07 0.03 .032 1.01 1.14 Level 2 (county-level) Ages 60 years and over 1.11 0.06 .048 1 1.23 Proportion of Hispanic 1.01 0.02 .7 0.98 1.03 Metropolitan county 0.69 0.64 .678 0.11 4.26 High school diploma 0.98 0.06 .775 0.87 1.11 Unemployment 0.94 0.06 .361 0.83 1.07 Community mental health center 3.42 1.6 .009 1.36 8.57 Primary Care Health Professional shortage area 1.68 0.66 .185 0.78 3.62 Mental health professional shortage area 0.7 0.29 .391 0.31 1.59 Mental health professionals 0.999 0.01 .913 0.99 1.01 Median home value 1.03 0.02 .116 0.99 1.07 Depression prevalence 1.03 0.05 .537 0.94 1.14 Random parameters B SE CIL CIH B SE CIL CIH Residual variance 0.78 0.27 0.4 1.53 0.55 0.3 0.19 1.6 Model fit AIC 429.624 431.557 ICC 0.156 0.083 Parameter Model 1 Model 2 Individual only Individual and County OR SE p CIL CIH OR SE p CIL CIH Intercept 7.48 15.48 .331 0.13 431.98 0.62 2.4 .902 0 1,166.62 Level 1 (individual-level) Age 0.96 0.02 .113 0.92 1.01 0.96 0.02 .102 0.92 1.01 Female 1.93 0.62 .041 1.03 3.64 1.89 0.62 .051 0.996 3.58 Married 0.55 0.17 .055 0.3 1.01 0.61 0.19 .116 0.32 1.13 High school diploma 2.24 0.72 .012 1.19 4.2 2.17 0.71 .018 1.14 4.12 Foreign born 1.54 0.61 .28 0.7 3.36 1.53 0.64 .305 0.68 3.47 No English spoken in home 0.65 0.25 .249 0.31 1.36 0.65 0.25 .26 0.3 1.38 Attitudes toward health 0.85 0.16 .389 0.59 1.23 0.84 0.16 .367 0.58 1.22 Income 0.91 0.1 .399 0.73 1.13 0.9 0.1 .367 0.73 1.13 Uninsured 0.66 0.31 .373 0.26 1.65 0.74 0.35 .529 0.29 1.89 SF-12 (Physical health) 0.998 0.01 .911 0.97 1.03 0.999 0.01 .942 0.97 1.03 SF-12 (Mental health) 0.95 0.02 .004 0.91 0.98 0.95 0.02 .005 0.91 0.98 K6 1.07 0.03 .035 1.004 1.14 1.07 0.03 .032 1.01 1.14 Level 2 (county-level) Ages 60 years and over 1.11 0.06 .048 1 1.23 Proportion of Hispanic 1.01 0.02 .7 0.98 1.03 Metropolitan county 0.69 0.64 .678 0.11 4.26 High school diploma 0.98 0.06 .775 0.87 1.11 Unemployment 0.94 0.06 .361 0.83 1.07 Community mental health center 3.42 1.6 .009 1.36 8.57 Primary Care Health Professional shortage area 1.68 0.66 .185 0.78 3.62 Mental health professional shortage area 0.7 0.29 .391 0.31 1.59 Mental health professionals 0.999 0.01 .913 0.99 1.01 Median home value 1.03 0.02 .116 0.99 1.07 Depression prevalence 1.03 0.05 .537 0.94 1.14 Random parameters B SE CIL CIH B SE CIL CIH Residual variance 0.78 0.27 0.4 1.53 0.55 0.3 0.19 1.6 Model fit AIC 429.624 431.557 ICC 0.156 0.083 Note: AIC = Akaike Information Criteria; ICC = Intraclass Correlation; OR = Odds ratio; CIL: Confidence Interval Low; CIH: Confidence Interval High. The bold coding indicates a statistical significance at p < .05. View Large Table 3. Multilevel Logistic Regression for Mental Health Service Utilization Parameter Model 1 Model 2 Individual only Individual and County OR SE p CIL CIH OR SE p CIL CIH Intercept 7.48 15.48 .331 0.13 431.98 0.62 2.4 .902 0 1,166.62 Level 1 (individual-level) Age 0.96 0.02 .113 0.92 1.01 0.96 0.02 .102 0.92 1.01 Female 1.93 0.62 .041 1.03 3.64 1.89 0.62 .051 0.996 3.58 Married 0.55 0.17 .055 0.3 1.01 0.61 0.19 .116 0.32 1.13 High school diploma 2.24 0.72 .012 1.19 4.2 2.17 0.71 .018 1.14 4.12 Foreign born 1.54 0.61 .28 0.7 3.36 1.53 0.64 .305 0.68 3.47 No English spoken in home 0.65 0.25 .249 0.31 1.36 0.65 0.25 .26 0.3 1.38 Attitudes toward health 0.85 0.16 .389 0.59 1.23 0.84 0.16 .367 0.58 1.22 Income 0.91 0.1 .399 0.73 1.13 0.9 0.1 .367 0.73 1.13 Uninsured 0.66 0.31 .373 0.26 1.65 0.74 0.35 .529 0.29 1.89 SF-12 (Physical health) 0.998 0.01 .911 0.97 1.03 0.999 0.01 .942 0.97 1.03 SF-12 (Mental health) 0.95 0.02 .004 0.91 0.98 0.95 0.02 .005 0.91 0.98 K6 1.07 0.03 .035 1.004 1.14 1.07 0.03 .032 1.01 1.14 Level 2 (county-level) Ages 60 years and over 1.11 0.06 .048 1 1.23 Proportion of Hispanic 1.01 0.02 .7 0.98 1.03 Metropolitan county 0.69 0.64 .678 0.11 4.26 High school diploma 0.98 0.06 .775 0.87 1.11 Unemployment 0.94 0.06 .361 0.83 1.07 Community mental health center 3.42 1.6 .009 1.36 8.57 Primary Care Health Professional shortage area 1.68 0.66 .185 0.78 3.62 Mental health professional shortage area 0.7 0.29 .391 0.31 1.59 Mental health professionals 0.999 0.01 .913 0.99 1.01 Median home value 1.03 0.02 .116 0.99 1.07 Depression prevalence 1.03 0.05 .537 0.94 1.14 Random parameters B SE CIL CIH B SE CIL CIH Residual variance 0.78 0.27 0.4 1.53 0.55 0.3 0.19 1.6 Model fit AIC 429.624 431.557 ICC 0.156 0.083 Parameter Model 1 Model 2 Individual only Individual and County OR SE p CIL CIH OR SE p CIL CIH Intercept 7.48 15.48 .331 0.13 431.98 0.62 2.4 .902 0 1,166.62 Level 1 (individual-level) Age 0.96 0.02 .113 0.92 1.01 0.96 0.02 .102 0.92 1.01 Female 1.93 0.62 .041 1.03 3.64 1.89 0.62 .051 0.996 3.58 Married 0.55 0.17 .055 0.3 1.01 0.61 0.19 .116 0.32 1.13 High school diploma 2.24 0.72 .012 1.19 4.2 2.17 0.71 .018 1.14 4.12 Foreign born 1.54 0.61 .28 0.7 3.36 1.53 0.64 .305 0.68 3.47 No English spoken in home 0.65 0.25 .249 0.31 1.36 0.65 0.25 .26 0.3 1.38 Attitudes toward health 0.85 0.16 .389 0.59 1.23 0.84 0.16 .367 0.58 1.22 Income 0.91 0.1 .399 0.73 1.13 0.9 0.1 .367 0.73 1.13 Uninsured 0.66 0.31 .373 0.26 1.65 0.74 0.35 .529 0.29 1.89 SF-12 (Physical health) 0.998 0.01 .911 0.97 1.03 0.999 0.01 .942 0.97 1.03 SF-12 (Mental health) 0.95 0.02 .004 0.91 0.98 0.95 0.02 .005 0.91 0.98 K6 1.07 0.03 .035 1.004 1.14 1.07 0.03 .032 1.01 1.14 Level 2 (county-level) Ages 60 years and over 1.11 0.06 .048 1 1.23 Proportion of Hispanic 1.01 0.02 .7 0.98 1.03 Metropolitan county 0.69 0.64 .678 0.11 4.26 High school diploma 0.98 0.06 .775 0.87 1.11 Unemployment 0.94 0.06 .361 0.83 1.07 Community mental health center 3.42 1.6 .009 1.36 8.57 Primary Care Health Professional shortage area 1.68 0.66 .185 0.78 3.62 Mental health professional shortage area 0.7 0.29 .391 0.31 1.59 Mental health professionals 0.999 0.01 .913 0.99 1.01 Median home value 1.03 0.02 .116 0.99 1.07 Depression prevalence 1.03 0.05 .537 0.94 1.14 Random parameters B SE CIL CIH B SE CIL CIH Residual variance 0.78 0.27 0.4 1.53 0.55 0.3 0.19 1.6 Model fit AIC 429.624 431.557 ICC 0.156 0.083 Note: AIC = Akaike Information Criteria; ICC = Intraclass Correlation; OR = Odds ratio; CIL: Confidence Interval Low; CIH: Confidence Interval High. The bold coding indicates a statistical significance at p < .05. View Large Second, individual predisposing, enabling, and need factors (Level 1) were added to the unconditional model to examine the relationships between individual-level variables and mental health service use (see Table 3). In Model 1, the AIC (429.624) showed improvement in the model fit compared to the unconditional model, but adding individual-level variables did not decrease the ICC (0.156). Model 1 identified four significant individual variables associated with mental health service utilization of older Hispanics: female, high school diploma, SF-12 mental health, and K6. Being a woman (odds ratio [OR] = 1.93) and having a high school diploma (OR = 2.24) increased the odds of using mental health services. People with a higher SF-12 mental health score (i.e., better mental health status) were less likely to use mental health services (OR = 0.95), while higher levels of psychological distress (K6) were related to increased odds of using mental health services (OR = 1.07). However, the variables of age, foreign born, language spoken at home, attitude toward health care, income, uninsured, and SF-12 (physical health) were not predictive of mental health service utilization. Third, Model 2 added 11 county-level characteristics to determine whether they were predictive of mental health service utilization among older Hispanics independent of individual characteristics. Compared to Model 1, the AIC did not show improvement (431.557) in Model 2, but adding county-level variables decreased the ICC (from 0.156 in Model 1 to 0.083 in Model 2). Model 2 identified several significant variables to explain mental health service utilization. The individual variables that were predictive of mental health service utilization in Model 1 remained significant and in the same direction in Model 2, with the exception of female (p = .051). Having a high school diploma (OR = 2.17), and higher levels of psychological distress (OR = 1.07) increased the odds of using mental health services; those with better SF-12 mental health status (OR = 0.95) were less likely to use a mental health service. Among the county-level characteristics, two variables were predictive of mental health service utilization by older Hispanics. The existence of a community mental health center increased the odds of using mental health services (OR = 3.42), and it had the largest odds ratio among the significant variables. In addition, living in a county with a higher proportion of older adults increased the odds of using mental health services among older Hispanics (OR = 1.11). However, the other variables—proportion of Hispanic, metropolitan county, high school diploma, unemployment rate, Primary Care Health Professional shortage area, median home value, and county mood disorder prevalence—were not predictive of mental health service utilization by older Hispanics. Discussion and Implications The findings of this study support that individual-level characteristics are essential to understanding mental health service use. Above all, better mental health status decreased the odds of using mental health services by older Hispanics. The findings were consistent with those from a systematic review of studies using Andersen’s model; greater perceived mental health needs were related to greater service utilization (Andersen, 2008; Babitsch, Gohl, & von Lengerke, 2012). The significant relationship between greater needs and greater service use makes intuitive sense: people with a mental health problem are more likely to seek out mental health treatment, although other factors may influence greater service use as well. The present study demonstrates that having a high school diploma was associated with increased use of mental health services. This finding highlights the role of education in the use of mental health services by older Hispanics (Gonzalez et al., 2015; Sorkin et al., 2016). Educational attainment of older Hispanics (45% were high school graduates in this study) is relatively low compared to other racial and ethnic groups. Older Hispanics might have experienced disadvantages regarding educational attainment during their early years; as a result, they may have limited employment opportunities and may have accumulated limited financial resources. Often combined with financial difficulties, lower levels of education may limit access to mental health services as well as adherence to evidence-based interventions and medications (Aranda, 2016). Also, education is closely related to health literacy, and low health literacy can impact obtaining appropriate health care. Those with a higher level of education may be more likely to be aware of symptoms and community resources, and as a result, may be more able to utilize a resource that they need for care. Hispanic adults tend to report lower levels of health literacy compared to other racial and ethnic groups (White, 2008). Therefore, disadvantaged education levels and low health literacy of older Hispanics should be addressed in future research and practice. The relationship between community resources and mental health service utilization by older Hispanics was also shown at the county level. Here, the existence of a community mental health center increased the odds of mental health services utilization by older Hispanics, compared to counties without a community mental health center. This finding was consistent with a previous study that reported that Hispanic adults living in a county with a community mental health center were more likely to use any mental health services in the past year (Cook et al., 2013). Community mental health centers may be particularly effective in providing culturally sensitive substance abuse and mental health services to their clients (Cummings, Cassie, & Trecartin, 2016). According to SAMHSA (2016), about 61% of community mental health centers provided treatment services in languages other than English, including Spanish (45.7%) and other languages (11.6%) in 2014. Although this study shows that older Hispanics living in counties with a community mental health center are more likely to report mental health services utilization, it is important to remember that this does not necessarily mean that they used mental health services at a community mental health center. Community mental health services may play an important role in mental health utilization by older Hispanics in offering affordable care. As shown in the sample characteristics, the Hispanic sample in this study was more likely to be uninsured (17%). This is in line with other studies that have found that, older undocumented Hispanics tend to lack health insurance. Earlier studies with Latino immigrants report that the undocumented Latinos were entirely uninsured (Perez & Fortuna, 2005), and that undocumented status is one of the biggest barriers to access health care services (Nandi et al., 2008). Thus, it is possible that older Hispanics are more likely to visit a community mental health center, instead of visiting other health service providers because of their insurance status. Indeed, community mental health centers have been discussed as one promising community safety net to serve underserved populations (Cook et al., 2013). For example, the U.S. Department of Health and Human Services requires that a community mental health center should provide at least 40% of its services to people who are not eligible for Medicare (American Medical Association [AMA], 2015). Future research should look into how the existence of community mental health centers influences mental health service utilization by older Hispanics. More detailed information on clients’ sociodemographic status (e.g., health insurance and undocumented status), staff (e.g., numbers of staff speaking Spanish and racial concordance), services provided (e.g., integrated mental health and health service in the center), proximity to a center, and a center’s funding structure would be helpful to understand mental health services provided to older Hispanics by community mental health centers. A higher proportion of older adults in a county increased the odds of mental health services utilization by older Hispanics. Auchincloss, Van Nostrand, and Ronsaville (2001) also found that a higher proportion of older adults in a neighborhood was associated with increased access to health care services. One possible explanation is that communities with a higher proportion of older adults have a different combination of available health services and health care facilities for older adults compared to counties with younger adults. Also, older adults living in counties with similar age groups may have a higher likelihood of obtaining health-related information from neighborhood peers through social interactions and building up their networks for health care services access. The mechanisms appear complex regarding positive links between mental health services utilization and a high proportion of older adults living in the same geographic area. Although communities with higher proportions of older adults may have more age-appropriate available health care services and more health care facilities, resulting from more demand for services (Andersen, Davidson, & Baumeister, 2014), we do not have information on community-level resources such as the characteristics of health care facilities, the number of mental health care professionals with similar cultural background to older racial and ethnic minorities and the existence of social service organizations serving older racial and ethnic minorities. Social interactions in the communities with a higher proportion of older adults also can be different from those with a lower proportion of older adults. Communities with a higher proportion of older adults are more likely to be stable and may have greater social networks and support resulting in positive health outcomes (Cagney, 2006). Thus, understanding characteristics of the communities with higher proportions of older adults will be helpful to examine how age composition affects mental health service utilization by older Hispanics. This study has limitations that should be noted. First, the narrow definition of mental health service utilization (i.e., an actual mental health professional visit) may misrepresent mental health service use among older Hispanics. They may prefer talking about their mental health concerns to spiritual providers and nonmental health professionals from the human service sector (Villatoro, Morales, & Mays, 2014). Another limitation is that this study could not include other plausible variables (e.g., acculturation, assimilation, discrimination, stigma, relationships between a provider and a patient) to explain mental health services use. Although this study includes a scale about attitudes toward health, older racial and ethnic minorities may have limited knowledge of mental health and/or may have culturally different beliefs about mental health services (Alvarez et al., 2014) that may not be captured on this scale. In addition, using counties as a unit of analysis may have limitations for increasing our understanding of mental health service utilization by older Hispanics. For example, about 52% of counties were partially mental health professional shortage areas. These patterns indicate that variability exists within counties. Therefore, future research should examine smaller units of analysis to better understand the impact of within-county variability. Despite its limitations, this study has several strengths. Above all, this study suggests that county contexts may be helpful to understand mental health service utilization among older Hispanics, even adjusting for individual-level characteristics. Most studies have focused on examining relationships between individual characteristics and mental health service utilization by older racial and ethnic minorities (e.g., Gonzalez et al., 2015; Jimenez et al., 2013). Although previous studies have explored the role of county or community characteristics to explain mental health service utilization, most have focused on the general population, not exclusively on older adults (e.g., Cook et al., 2013; Stahler, Mennis, Cotlar, & Baron, 2009; Stockdale, Tang, Zhang, Belin, & Wells, 2007). Another strength is that this study used actual service visits obtained from health and mental health service providers to measure mental health service utilization. Mental health service utilization has been inconsistently operationalized in studies. Some studies define mental health service use as visiting specialty mental health professionals (e.g., Dobalian & Rivers, 2008; Ojeda & McGuire, 2006). Other studies ask whether respondents talk to professionals from the human service sector and spiritual providers to capture any efforts to receive mental health services (e.g., Fortuna, Porche, & Alegria, 2008; Kim et al., 2013). Depending on respondents’ memory may result in recall bias, and household respondents may not provide accurate answers. This study’s use of the MEPS, which draws its data from medical providers’ records, ensures that findings are not subject to recall bias. In sum, this study aimed to examine whether county contexts are associated with mental health service utilization by older Hispanics. Understanding county context where people reside is important because people use the mental health services that are in their communities. Both individual- and county-level characteristics affected mental health service utilization by older Hispanics. For example, this study identified that county characteristics (i.e., existence of community mental health centers and a higher proportion of older adults) increased mental health service utilization as did individual characteristics of higher levels of education and worse mental health status. In this way, this study identified different factors affecting mental health service use by older Hispanics, and will contribute to the existing literature that generally has focused on the role of individual characteristics to understand mental health service utilization. Understanding both individual- and county-specific characteristics associated with mental health service utilization is essential to develop tailored outreach programs for older Hispanics as well as to develop and evaluate mental health policy. Acknowledgements The author acknowledges the support of the Agency for Healthcare Research Quality (AHRQ) and the data analysis in this paper was conducted at the AHRQ data center. The author specially thanks Dr. Donna Harrington for her immeasurable source of support in this study. References Agency for Healthcare Research and Quality . ( 2013 ). MEPS HC-147 2011 full year consolidated data file . Rockville, MD : Author . Akincigil , A. , Olfson , M. , Siegel , M. , Zurlo , K. A. , Walkup , J. T. , & Crystal , S . ( 2012 ). Racial and ethnic disparities in depression care in community-dwelling elderly in the United States . American Journal of Public Health , 102 , 319 – 328 . doi: 10.2105/AJPH.2011.300349 Google Scholar CrossRef Search ADS PubMed Alvarez , P. , Rengifo , J. , Emrani , T. , & Gallagher-Thompson , D . ( 2014 ). Latino older adults and mental health: A review and commentary . Clinical Gerontologist , 37 , 33 – 48 . doi: 10.1080/07317115.2013.847514 Google Scholar CrossRef Search ADS American Medical Association . ( 2015 ). Mental health services . Retrieved from https://www.cms.gov/Outreach-and-Education/Medicare-Learning-Network-MLN/MLNProducts/Downloads/Mental-Health-Services-Booklet-ICN903195.pdf. Andersen , R . ( 1968 ). A behavioral model of families’ use of health services . (Research series. No. 25). Chicago, IL : The Center of Health Administration Studies . Andersen , R. M . ( 2008 ). National health surveys and the behavioral model of health services use . Medical Care , 46 , 647 – 653 . doi: 10.1097/MLR.0b013e31817a835d Google Scholar CrossRef Search ADS PubMed Andersen , R. , & Davidson , P . ( 2007 ). Improving access to care in America: Individual and contextual indicators . In R. Andersen , T. Rice & J. Kominski (Eds.), Changing the U.S. health care system: Key issues in health services policy and management (pp. 3 – 31 ). San Francisco, CA : John Wiley & Sons . Andersen , R. , Davidson , P. , & Baumeister , S . ( 2014 ). Improving access to care . In G. Kominski (Ed.), Changing the U.S. health care system: Key issues in health services policy and management (pp. 33 – 64 ). San Francisco, CA : Jossey-Bass . Aranda , M . ( 2016 ). Social work practice with older Latinos in the United States: A mental health perspective . In D. Kaplan & B. Berkman (Eds.), The Oxford handbook of social work in health and aging (pp. 527 – 533 ). New York, NY : Oxford University Press . Auchincloss , A. H. , Van Nostrand , J. F. , & Ronsaville , D . ( 2001 ). Access to health care for older persons in the United States: Personal, structural, and neighborhood characteristics . Journal of Aging and Health , 13 , 329 – 354 . doi: 10.1177/089826430101300302 Google Scholar CrossRef Search ADS PubMed Babitsch , B. , Gohl , D. , & von Lengerke , T . ( 2012 ). Re-revisiting Andersen’s behavioral model of health services use: A systematic review of studies from 1998-2011 . Psycho-social Medicine , 9 , 1 – 15 . doi: 10.3205/psm000089 Bickel , R . ( 2007 ). Multilevel analysis for applied research: It’s just regression! New York, NY : The Guilford Press. Cagney , K. A . ( 2006 ). Neighborhood age structure and its implications for health . Journal of Urban Health: Bulletin of the New York Academy of Medicine , 83 , 827 – 834 . doi: 10.1007/s11524-006-9092-z Google Scholar CrossRef Search ADS PubMed Choi , S . ( 2015 ). How does satisfaction with medical care differ by citizenship and nativity status?: A county-level multilevel analysis . The Gerontologist , 55, 735-747. doi: 10.1093/geront/gnt201 Cook , B. L. , Doksum , T. , Chen , C. N. , Carle , A. , & Alegría , M . ( 2013 ). The role of provider supply and organization in reducing racial/ethnic disparities in mental health care in the U.S . Social Science & Medicine (1982) , 84 , 102 – 109 . doi: 10.1016/j.socscimed.2013.02.006 Google Scholar CrossRef Search ADS PubMed Cummings , S. , Cassie , K. , & Trecartin , S . ( 2016 ). Community mental health centers . In D. Kaplan & B. Berkman (Eds.), The Oxford handbook of social work in health and aging (pp. 141 – 149 ). New York, NY : Oxford University Press . Google Scholar CrossRef Search ADS DeFranco , E. A. , Lian , M. , Muglia , L. A. , & Schootman , M . ( 2008 ). Area-level poverty and preterm birth risk: A population-based multilevel analysis . BMC Public Health , 8 , 316 . doi: 10.1186/1471-2458-8-316 Google Scholar CrossRef Search ADS PubMed Dobalian , A. , & Rivers , P. A . ( 2008 ). Racial and ethnic disparities in the use of mental health services . The Journal of Behavioral Health Services & Research , 35 , 128 – 141 . doi: 10.1007/s11414-007-9097-8 Google Scholar CrossRef Search ADS PubMed Federal Interagency Forum on Aging-Related Statistics . ( 2016 ). Older Americans 2016: Key indicators of well-being . Washington, DC : U.S. Government Printing Office . Fortuna , L. R. , Porche , M. V. , & Alegria , M . ( 2008 ). Political violence, psychosocial trauma, and the context of mental health services use among immigrant Latinos in the United States . Ethnicity & Health , 13 , 435 – 463 . doi: 10.1080/13557850701837286 Google Scholar CrossRef Search ADS PubMed Fuentes , D. , & Aranda , M. P . ( 2012 ). Depression interventions among racial and ethnic minority older adults: A systematic review across 20 years . The American Journal of Geriatric Psychiatry: Official Journal of the American Association for Geriatric Psychiatry , 20 , 915 – 931 . doi: 10.1097/JGP.0b013e31825d091a Google Scholar CrossRef Search ADS PubMed Gonzalez , J. M. , Applewhite , S. R. , & Barrera , I . ( 2015 ). Older Mexican Americans: Pathways to mental health service use . Social Work in Mental Health , 13 , 390 – 414 . doi: 10.1080/15332985.2014.927813 Google Scholar CrossRef Search ADS Institute of Medicine . ( 2012 ). The mental health and substance use workforce for older adults: In whose hands? Washington, DC : The National Academies Press . Jimenez , D. E. , Cook , B. , Bartels , S. J. , & Alegría , M . ( 2013 ). Disparities in mental health service use of racial and ethnic minority elderly adults . Journal of the American Geriatrics Society , 61 , 18 – 25 . doi: 10.1111/jgs.12063 Google Scholar CrossRef Search ADS PubMed Kaskie , B. , Van Gilder , R. , & Gregory , D . ( 2008 ). Community mental health service use by older adults in California . Aging & Mental Health , 12 , 134 – 143 . doi: 10.1080/13607860801942761 Google Scholar CrossRef Search ADS PubMed Kessler , R. C. , Andrews , G. , Colpe , L. J. , Hiripi , E. , Mroczek , D. K. , Normand , S. L. ,… Zaslavsky , A. M . ( 2002 ). Short screening scales to monitor population prevalence and trends in non-specific psychological distress . Psychological Medicine , 32 , 959 – 976 . doi: 10.1017/S0033291702006074 Google Scholar CrossRef Search ADS PubMed Kim , G. , Jang , Y. , Chiriboga , D. A. , Ma , G. X. , & Schonfeld , L . ( 2010 ). Factors associated with mental health service use in Latino and Asian immigrant elders . Aging & Mental Health , 14 , 535 – 542 . doi: 10.1080/13607860903311758 Google Scholar CrossRef Search ADS PubMed Kim , G. , Parton , J. M. , DeCoster , J. , Bryant , A. N. , Ford , K. L. , & Parmelee , P. A . ( 2013 ). Regional variation of racial disparities in mental health service use among older adults . The Gerontologist , 53 , 618 – 626 . doi: 10.1093/geront/gns107 Google Scholar CrossRef Search ADS PubMed Mackenzie , C. S. , Pagura , J. , & Sareen , J . ( 2010 ). Correlates of perceived need for and use of mental health services by older adults in the collaborative psychiatric epidemiology surveys . The American Journal of Geriatric Psychiatry: Official Journal of the American Association for Geriatric Psychiatry , 18 , 1103 – 1115 . doi: 10.1097/JGP.0b013e3181dd1c06 Google Scholar CrossRef Search ADS PubMed Nandi , A. , Galea , S. , Lopez , G. , Nandi , V. , Strongarone , S. , & Ompad , D. C . ( 2008 ). Access to and use of health services among undocumented Mexican immigrants in a US urban area . American Journal of Public Health , 98 , 2011 – 2020 . doi: 10.2105/AJPH.2006.096222 Google Scholar CrossRef Search ADS PubMed National Alliance on Mental Illness . ( 2009). Depression in older persons fact sheet . Retrieved from http://www.nami.org/Template.cfm?Section=By_Illness&template=/ContentManagement/ContentDisplay.cfm&ContentID=7515. Ojeda , V. D. , & McGuire , T. G . ( 2006 ). Gender and racial/ethnic differences in use of outpatient mental health and substance use services by depressed adults . The Psychiatric Quarterly , 77 , 211 – 222 . doi: 10.1007/s11126-006-9008-9 Google Scholar CrossRef Search ADS PubMed Pérez , M. C. , & Fortuna , L . ( 2005 ). Chapter 6. Psychosocial stressors, psychiatric diagnoses and utilization of mental health services among undocumented immigrant Latinos . Journal of Immigrant & Refugee Services , 3 , 107 – 123 . doi: 10.1300/J191v03n01_06 Google Scholar CrossRef Search ADS Singh , G. K . ( 2003 ). Area deprivation and widening inequalities in US mortality, 1969-1998 . American Journal of Public Health , 93 , 1137 – 1143 . Google Scholar CrossRef Search ADS PubMed Sorkin , D. H. , Murphy , M. , Nguyen , H. , & Biegler , K. A . ( 2016 ). Barriers to mental health care for an ethnically and racially diverse sample of older adults . Journal of the American Geriatrics Society , 64 , 2138 – 2143 . doi: 10.1111/jgs.14420 Google Scholar CrossRef Search ADS PubMed Sorkin , D. H. , Pham , E. , & Ngo-Metzger , Q . ( 2009 ). Racial and ethnic differences in the mental health needs and access to care of older adults in california . Journal of the American Geriatrics Society , 57 , 2311 – 2317 . doi: 10.1111/j.1532-5415.2009.02573.x Google Scholar CrossRef Search ADS PubMed Stahler , G. J. , Mennis , J. , Cotlar , R. , & Baron , D. A . ( 2009 ). The influence of neighborhood environment on treatment continuity and rehospitalization in dually diagnosed patients discharged from acute inpatient care . The American Journal of Psychiatry , 166 , 1258 – 1268 . doi: 10.1176/appi.ajp.2009.08111667 Google Scholar CrossRef Search ADS PubMed StataCorp . ( 2015 ). Stata (Version 14) [Software] . College Station, TX : Author . Stockdale , S. E. , Tang , L. , Zhang , L. , Belin , T. R. , & Wells , K. B . ( 2007 ). The effects of health sector market factors and vulnerable group membership on access to alcohol, drug, and mental health care . Health Services Research , 42 ( 3 Pt 1 ), 1020 – 1041 . doi: 10.1111/j.1475-6773.2006.00636.x Google Scholar CrossRef Search ADS PubMed Substance Abuse and Mental Health Service Administration [SAMHSA] . ( 2011 ). Leading change: A plan for SAMHSA’s role and actions 2011–2014 . Retrieved from http://store.samhsa.gov/shin/content//SMA11-4629/01-FullDocument.pdf. Substance Abuse and Mental Health Service Administration [SAMHSA] . ( 2016 ). National Mental Health Services Survey (N-MHSS): 2014: Data on mental health treatment facilities . HHS Publication No. (SMA) 16–5000. Rockville, MD : Author . U.S. Department of Health and Human Services . ( 2018 ). Health insurance and mental health services . Retrieved from https://www.mentalhealth.gov/get-help/health-insurance/. Villatoro , A. P. , Morales , E. S. , & Mays , V. M . ( 2014 ). Family culture in mental health help-seeking and utilization in a nationally representative sample of Latinos in the United States: The NLAAS . The American Journal of Orthopsychiatry , 84 , 353 – 363 . doi: 10.1037/h0099844 Google Scholar CrossRef Search ADS PubMed Ware , J. Jr , Kosinski , M. , & Keller , S. D . ( 1996 ). A 12-Item short-form health survey: Construction of scales and preliminary tests of reliability and validity . Medical Care , 34 , 220 – 233 . Google Scholar CrossRef Search ADS PubMed Wei , W. , Sambamoorthi , U. , Olfson , M. , Walkup , J. T. , & Crystal , S . ( 2005 ). Use of psychotherapy for depression in older adults . The American Journal of Psychiatry , 162 , 711 – 717 . doi: 10.1176/appi.ajp.162.4.711 Google Scholar CrossRef Search ADS PubMed White , S . ( 2008 ). Assessing the nation’s health literacy. Key concepts and findings of the National Assessment of Adult Literacy . Chicago, IL : American Medical Association . © The Author(s) 2018. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The Gerontologist Oxford University Press

County Context and Mental Health Service Utilization by Older Hispanics

The Gerontologist , Volume Advance Article – Apr 16, 2018

Loading next page...
 
/lp/ou_press/county-context-and-mental-health-service-utilization-by-older-W90zKLQTZY
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.
ISSN
0016-9013
eISSN
1758-5341
D.O.I.
10.1093/geront/gny033
Publisher site
See Article on Publisher Site

Abstract

Abstract Background and Objectives Although older Hispanics experience high rates of depression, they tend to underuse mental health services. The study examined the association between county characteristics and mental health service use among older Hispanics, controlling for individual characteristics. Research Design and Methods The study used the 2008–2012 Medical Expenditure Panel Study and linked county-level data from the 2013–2014 Area Health Resources Files and the 2008–2012 Chronic Conditions Data Warehouse, using the Federal Information Processing Standard county code. The sample includes 1,143 community-dwelling Hispanics ages 60 years or older (Level 1) and 156 counties (Level 2) where the sample resides. The single dichotomous measure of mental health service utilization was based on whether or not the respondent met one or more of three conditions: (1) the respondent received care from a mental health professional, (2) received a service including mental health counseling or psychotherapy, or (3) received a service that was related to the International Classification of Diseases. Multilevel logistic regression analysis was used to examine the role of county context. Results The proportion of older adults and the existence of community mental health centers at the county-level were associated with mental health services use among this population. At the individual-level, education and mental health status were also associated with using mental health services. Discussion and Implications The county context plays an important role in understanding mental health services use among older Hispanics, indicating the need for intervention strategies at the county level. County context, Mental health services, Multilevel modeling, Older Hispanics Underutilization of mental services by older Hispanics is a growing concern in the United States. Underutilization may delay the treatment of mental health problems, and untreated problems may have a huge impact on the well-being of an individual. For example, untreated depression leads to disease burden and disability (Fuentes & Aranda, 2012) and may result in higher suicide and nonsuicide mortality (National Alliance on Mental Illness [NAMI], 2009). Although previous studies using national data (e.g., National Latino and Asian American Study) have advanced understanding of mental health service utilization among older Hispanics, they have not explored the socioenvironmental context in which this population accesses mental health services; most existing research on this topic has focused on individual-level variables. The number of adults aged 65 years and older in the United States with mental health problems is growing, and approximately 14%–20% of older adults are estimated to have mental illnesses such as depression and schizophrenia (Institute of Medicine [IOM], 2012). In addition, the overall population of older Hispanics is expected to grow from about 3.6 million (9% of the U.S. older population) in 2014 to 21.5 million (22% of the U.S. older population) in 2060 (Federal Interagency Forum on Aging-Related Statistics, 2016). It is likely that there will be a corresponding increase in the number of older Hispanics with mental health problems (Mackenzie, Pagura, & Sareen, 2010). A study using the National Latino and Asian American Survey found that 1 in 10 (9.7%) older Hispanics were diagnosed with anxiety disorders in the past 12 months, but only 5.5% used mental health services (Kim, Jang, Chiriboga, Ma, & Schonfeld, 2010). Another study, the 2005 California Health Interview Survey found that only 10.9% of older Hispanics received mental health services although 24% reported mental illnesses (Sorkin, Pham, & Ngo-Metzger, 2009). Older Hispanics are less likely to use professional mental health services compared to Whites because they face individual, cultural, linguistic, and systemic barriers to accessing mental health services as other ethnic minority populations do (Akincigil et al., 2012; SAMHSA, 2011). Older Hispanics have negative perceptions of seeking formal mental health services because of experiences of racial discrimination and stresses from immigration-related issues (Alvarez, Rengifo, Emrani, & Gallagher-Thompson, 2014). Education and economic resources are very important in utilization of mental health services (Gonzalez, Applewhite, & Barrera, 2015; Sorkin, Murphy, Nguyen, & Biegler, 2016). Older Hispanics underutilize mental health services in part because they are less likely to have a high school diploma and are more likely to live below the 150% poverty line compared to other racial and ethnic groups (Akincigil et al., 2012). Mental health service utilization by older Hispanics has been studied by many researchers, and reducing racial disparities in health care services has been identified as a national priority (Kim et al., 2013; U.S. Department of Health and Human Services [HHS], 2018). To understand facilitators of and barriers to the use of mental health services among older Hispanics, most research has focused on individual characteristics such as age, gender, education, income, and health and mental health status (Akincigil et al., 2012; Alvarez et al., 2014; Kim et al., 2010). A limited number of studies have focused on older adults as their study population to explain the relationships between socioenvironmental characteristics and mental health services utilization (e.g., Kaskie, Gilder, & Gregory, 2008; Kim et al., 2013; Wei, Sambamoorthi, Olfson, Walkup, & Crystal, 2005). Thus, this study aimed to examine mental health service utilization by older Hispanics. Theoretical Framework Employing Andersen’s (2008) Behavioral Model of Health Services Utilization as a theoretical framework, this study hypothesized that county contexts are associated with mental health service use by older Hispanics, controlling for individual characteristics. Andersen’s model is very useful for addressing this study’s research questions regarding mental health service utilization because it explains that the contextual predisposing (e.g., community demographic and social characteristics), enabling (e.g., supply of health care services), and need characteristics (e.g., depression rate) are associated with the use of health services (Andersen, 2008; Andersen & Davidson, 2007). Geographic differences may be an important contributing factor to racial and ethnic disparities in mental health (Kim et al., 2013). As Andersen and Davidson (2007) suggest, contextual characteristics, such as neighborhood sociodemographic characteristics, health care resources, and population health indices—in addition to individual characteristics—impact the use of mental health services and health outcomes. Health care resources such as mental health professionals and community mental health centers may be unequally allocated in different geographic areas (Cook, Doksum, Chen, Carle, & Alegría, 2013). In addition, older Hispanics are more likely to live in economically disadvantaged areas that have fewer health care resources and may have fewer opportunities to access mental health care (Cook et al., 2013). Defining Geographic Boundaries Defining a geographic boundary is important because geographic differences may be an important contributing factor to racial and ethnic disparities in mental health (Kim et al., 2013). This study used counties as a geographic boundary. Although census tracts or blocks may be more socioeconomically homogeneous than other geographic units, counties are more sociopolitically and geographically stable compared to census tracts or blocks (Singh, 2003). The counties are also the smallest geographical entity to provide data on varieties of health care and social services within states as well as to consistently provide health, population, and socioeconomic statistics over time (DeFranco, Lian, Muglia, & Schootman, 2008). Mental health care resources, such as number of mental health professionals and community mental health centers vary across counties, and there is more variability in mental health services use at the county-level compared to other area levels. Based on Andersen’s model and previous empirical studies, the purpose of the study is to examine what county context (i.e., county predisposing, enabling, and need factors) as well as individual-level characteristics are associated with the use of mental health services by older Hispanics. An important premise of this study is that an understanding of the relationship between county-level context and mental health services utilization in this population is key to developing effective intervention strategies at the macro level. Methods Sample The study used the 2008–2012 Medical Expenditure Panel Survey (MEPS) and linked county-level data from the 2013–2014 Area Health Resources Files (AHRF) and the 2008–2012 Chronic Conditions Data Warehouse (CCW). The MEPS is a national survey to estimate health care use, health care insurance coverage, and access to health care and quality for the U.S. civilian noninstitutionalized population (Agency for Healthcare Research and Quality, 2013). The study employed five-year MEPS individual-level data (panels 13–16) because most county data are based on 2010 census data or on 2008–2012 averages. Combining multiyear data has been used by researchers to provide a sufficient sample size for racial and ethnic groups (e.g., Choi, 2015; Jimenez, Cook, Bartels, & Alegría, 2013). This study also employed the 2013–2014 AHRF to provide information on county contexts; the AHRF contains information on county-level demographic and socioeconomic status, and health care resources such as primary care physician shortage areas and community mental health center. The third data set, the CCW, provided information on depression prevalence rates at the county level. This study selected 1,143 older Hispanics who were ages 60 years or older and who completed the survey by themselves from the MEPS. The MEPS was merged with the AHRF and the CCW using county identifiers (i.e., the Federal Information Processing Standard county code). Based on the data linkage, the level two sample of the study is 156 counties where 1,143 older Hispanics reside. Measures This study used individual characteristics from the MEPS and county characteristics from the AHRF and the CCW. Mental Health Services Utilization Mental health services utilization was defined as actual provider visits for mental health care. Using the Hospital Outpatient Visit Files and the Office-Based Medical Provider Visits Files of MEPS, there were three ways of identifying mental health service use: types of providers, types of treatments, and the International Classification of Diseases, ninth revision (ICD-9) codes. If at least one of the following conditions was met, it was considered a mental health service use: (1) the person received care from a psychologist or a social worker, (2) the type of treatment was psychotherapy or mental health counseling, or (3) a provider visit was coded with the clinical classification codes (CCC) associated with a mental health condition. The CCC are created by aggregating the ICD-9 codes to have clinically meaningful categories, and the CCCs of 650 through 670 refer to mental health conditions. Mental health services utilization was coded as 1 = yes or 0 = no. Individual Characteristics This study included individual characteristics suggested from Andersen’s (2008) model and previous literature. This study included individual predisposing characteristics (i.e., age in years, gender [female/male], married [yes], high school diploma [yes], foreign born [yes], language spoken at home [Spanish/English], and attitude toward health). As enabling characteristics, this study included health insurance (yes) and income, and also included health status from the Short Form 12 Version 2 (SF-12v2) (Ware, Kosinski, & Keller, 1996) and the Kessler Psychological Distress Scale (K6) (Kessler et al., 2002) as the need factor. All individual data were obtained from the MEPS. County Context To measure county context, this study used county predisposing (i.e., proportion of older adults, percent Hispanic population, metropolitan area (yes), percentage of those with high school diploma, unemployment rates) and county enabling (i.e., existence of community mental health center, primary care physician and mental health professional shortage area, mental health professionals per 100,000, and median home values) drawn from the AHRF. The study also included depression prevalence rates at the county-level from the CCW. Data Analysis Plan To answer the research questions, multilevel modeling analysis using Stata 14 software (StataCorp, 2015) was conducted to examine the role of county context on mental health services utilization by older Hispanics, controlling for individual characteristics. The MEPS uses a complex survey design that necessitated the need for weights: analyses, therefore, included the use of the Taylor series linearization method to provide an accurate estimation of the variance structure and standard errors, with the strata and the PSU specified (AHRQ, 2013). First, the unconditional model was tested to examine county-to-county variability of mental health service utilization. This can be done by calculating the unconditional intraclass correlation (ICC) (Bickel, 2007). Second, individual predisposing, enabling, and need factors (Level 1) were added to the model to examine the relationships between individual-level variables and mental health service use. Third, county contexts (Level 2) were added to the previous model to examine whether or not the county contexts are predictive of mental health service use among older adults, independent of individual characteristics. This study involved the use of restricted data (i.e., FIPS codes), so a proposal was submitted to and approved by the Agency for Healthcare Research and Quality (AHRQ) as well as a University Institutional Review Board in the mid-Atlantic area. Data analyses using the restricted data were conducted at the AHRQ Data Center in Rockville, Maryland. Results Table 1 presents the individual-level sample characteristics of 1,143 older Hispanics (a population estimate of 2,736,013 older Hispanics). Over half were women (55%) or married (53%); the average age of the sample was 69 years old (SD = 9.9). Less than half of the Hispanic sample (45%) earned a high school diploma. Almost two-thirds were born in a foreign country (63%) and spoke Spanish at home (60%). Regarding ethnicity, the majority of the sample was Mexican (53%), followed by Central or South American (13%), Puerto Rican (12%), Cuban (9%), Dominican (5%), and other (7%). About 6% of the sample had a mental health service visit during the data collection period. Table 1. Individual-level Hispanic Sample Characteristics Continuous variables Weighted M (SD) N = 2,736,013 Unweighted M (SD) N = 1,143 Unweighted range Age 69.28 (9.87) 68.56 (7.10) 60–85* Attitudes toward health 1.89 (1.17) 1.90 (0.87) 1–5 SF-12 (Physical health) 42.80 (16.06) 42.58 (11.88) 11.65–63.77 SF-12 (Mental health) 49.83 (14.08) 48.81 (10.58) 13.92–70.48 K6 4.29 (7.14) 4.68 (5.64) 0–24 Categorical variables Weighted n (%) Unweighted n (%) Mental health service visit 168,538 (6.16) 78 (6.82) Gende  Female 1,507,269 (55.09) 632 (55.29)  Male 1,228,743 (44.91) 511 (44.71) Race  White 2,618,090 (95.69) 1,091 (95.45)  Black 50,616 (1.85) 15 (2.19)  American Indian 29,275 (1.07) 13 (1.14) Married (yes) 1,453,643 (53.13) 636 (55.64) High school diploma (yes) 1,243,244 (45.44) 430 (37.62) Foreign born (yes) 1,728,339 (63.17) 778 (68.07) Language spoken at home  English 1,108,085 (40.5) 383 (33.51)  Spanish 1,627,927 (59.5) 760 (66.49) Income  Poor 583,318 (21.32) 270 (23.62)  Near poor 214,230 (7.83) 114 (9.97)  Low income 617,792 (22.58) 261 (22.83)  Middle income 833,389 (30.46) 333 (29.13)  High income 487,284 (17.81) 165 (14.44) Uninsured (yes) 371,277 (13.57) 190 (16.62) Continuous variables Weighted M (SD) N = 2,736,013 Unweighted M (SD) N = 1,143 Unweighted range Age 69.28 (9.87) 68.56 (7.10) 60–85* Attitudes toward health 1.89 (1.17) 1.90 (0.87) 1–5 SF-12 (Physical health) 42.80 (16.06) 42.58 (11.88) 11.65–63.77 SF-12 (Mental health) 49.83 (14.08) 48.81 (10.58) 13.92–70.48 K6 4.29 (7.14) 4.68 (5.64) 0–24 Categorical variables Weighted n (%) Unweighted n (%) Mental health service visit 168,538 (6.16) 78 (6.82) Gende  Female 1,507,269 (55.09) 632 (55.29)  Male 1,228,743 (44.91) 511 (44.71) Race  White 2,618,090 (95.69) 1,091 (95.45)  Black 50,616 (1.85) 15 (2.19)  American Indian 29,275 (1.07) 13 (1.14) Married (yes) 1,453,643 (53.13) 636 (55.64) High school diploma (yes) 1,243,244 (45.44) 430 (37.62) Foreign born (yes) 1,728,339 (63.17) 778 (68.07) Language spoken at home  English 1,108,085 (40.5) 383 (33.51)  Spanish 1,627,927 (59.5) 760 (66.49) Income  Poor 583,318 (21.32) 270 (23.62)  Near poor 214,230 (7.83) 114 (9.97)  Low income 617,792 (22.58) 261 (22.83)  Middle income 833,389 (30.46) 333 (29.13)  High income 487,284 (17.81) 165 (14.44) Uninsured (yes) 371,277 (13.57) 190 (16.62) Note: *Everyone aged 85 years and older was coded as 85 to maintain the confidentiality of the data. View Large Table 1. Individual-level Hispanic Sample Characteristics Continuous variables Weighted M (SD) N = 2,736,013 Unweighted M (SD) N = 1,143 Unweighted range Age 69.28 (9.87) 68.56 (7.10) 60–85* Attitudes toward health 1.89 (1.17) 1.90 (0.87) 1–5 SF-12 (Physical health) 42.80 (16.06) 42.58 (11.88) 11.65–63.77 SF-12 (Mental health) 49.83 (14.08) 48.81 (10.58) 13.92–70.48 K6 4.29 (7.14) 4.68 (5.64) 0–24 Categorical variables Weighted n (%) Unweighted n (%) Mental health service visit 168,538 (6.16) 78 (6.82) Gende  Female 1,507,269 (55.09) 632 (55.29)  Male 1,228,743 (44.91) 511 (44.71) Race  White 2,618,090 (95.69) 1,091 (95.45)  Black 50,616 (1.85) 15 (2.19)  American Indian 29,275 (1.07) 13 (1.14) Married (yes) 1,453,643 (53.13) 636 (55.64) High school diploma (yes) 1,243,244 (45.44) 430 (37.62) Foreign born (yes) 1,728,339 (63.17) 778 (68.07) Language spoken at home  English 1,108,085 (40.5) 383 (33.51)  Spanish 1,627,927 (59.5) 760 (66.49) Income  Poor 583,318 (21.32) 270 (23.62)  Near poor 214,230 (7.83) 114 (9.97)  Low income 617,792 (22.58) 261 (22.83)  Middle income 833,389 (30.46) 333 (29.13)  High income 487,284 (17.81) 165 (14.44) Uninsured (yes) 371,277 (13.57) 190 (16.62) Continuous variables Weighted M (SD) N = 2,736,013 Unweighted M (SD) N = 1,143 Unweighted range Age 69.28 (9.87) 68.56 (7.10) 60–85* Attitudes toward health 1.89 (1.17) 1.90 (0.87) 1–5 SF-12 (Physical health) 42.80 (16.06) 42.58 (11.88) 11.65–63.77 SF-12 (Mental health) 49.83 (14.08) 48.81 (10.58) 13.92–70.48 K6 4.29 (7.14) 4.68 (5.64) 0–24 Categorical variables Weighted n (%) Unweighted n (%) Mental health service visit 168,538 (6.16) 78 (6.82) Gende  Female 1,507,269 (55.09) 632 (55.29)  Male 1,228,743 (44.91) 511 (44.71) Race  White 2,618,090 (95.69) 1,091 (95.45)  Black 50,616 (1.85) 15 (2.19)  American Indian 29,275 (1.07) 13 (1.14) Married (yes) 1,453,643 (53.13) 636 (55.64) High school diploma (yes) 1,243,244 (45.44) 430 (37.62) Foreign born (yes) 1,728,339 (63.17) 778 (68.07) Language spoken at home  English 1,108,085 (40.5) 383 (33.51)  Spanish 1,627,927 (59.5) 760 (66.49) Income  Poor 583,318 (21.32) 270 (23.62)  Near poor 214,230 (7.83) 114 (9.97)  Low income 617,792 (22.58) 261 (22.83)  Middle income 833,389 (30.46) 333 (29.13)  High income 487,284 (17.81) 165 (14.44) Uninsured (yes) 371,277 (13.57) 190 (16.62) Note: *Everyone aged 85 years and older was coded as 85 to maintain the confidentiality of the data. View Large Table 2 presents the county-level characteristics for the 156 counties where the 1,143 older Hispanic respondents resided. In the 156 counties included in this sample, an average of 2 in 10 persons were Hispanic and an average of 19% of the county residents were older adults ages 60 years and over. About 9 in 10 (90%) of counties where the Hispanic sample resided were metropolitan counties, and the median home value was about $249,000. The average depression rate was 13% for all counties, and less than half of the counties (44%) had a community mental health center. On average, each county had 47 mental health professionals per 100,000 county residents. Table 2. County-level Characteristics (N = 156) Continuous variables M (SD) Range Ages 60 years and over (%) 18.99 (6.13) 9.24–56.19 Proportion of Hispanic (%) 21.75 (16.81) 2.30–90.60 High school diploma (%) 56.68 (6.71) 33.46–76.09 Unemployment (%) 8.35 (3.27) 2.80–28.20 Mental health professionals (per 100,000) 43.63 (43.71) 0–306.08 Median home value ($10,000) 24.92 (14.95) 5.73–82.73 Depression prevalence (%) 13.46 (2.87) 6.56–26.47 Categorical variables n (%) Metropolitan county (yes) 140 (89.74) Community mental health center (yes) 69 (44.23) PCP shortage area  None of the county 18 (11.54)  One or more parts of the county 78 (50.00)  The whole county 60 (38.46) Mental health professional shortage area  None of the county 22 (14.10)  One or more parts of the county 81 (51.92)  The whole county 53 (33.97) Continuous variables M (SD) Range Ages 60 years and over (%) 18.99 (6.13) 9.24–56.19 Proportion of Hispanic (%) 21.75 (16.81) 2.30–90.60 High school diploma (%) 56.68 (6.71) 33.46–76.09 Unemployment (%) 8.35 (3.27) 2.80–28.20 Mental health professionals (per 100,000) 43.63 (43.71) 0–306.08 Median home value ($10,000) 24.92 (14.95) 5.73–82.73 Depression prevalence (%) 13.46 (2.87) 6.56–26.47 Categorical variables n (%) Metropolitan county (yes) 140 (89.74) Community mental health center (yes) 69 (44.23) PCP shortage area  None of the county 18 (11.54)  One or more parts of the county 78 (50.00)  The whole county 60 (38.46) Mental health professional shortage area  None of the county 22 (14.10)  One or more parts of the county 81 (51.92)  The whole county 53 (33.97) Note: PCP = Primary Care Health Professional. View Large Table 2. County-level Characteristics (N = 156) Continuous variables M (SD) Range Ages 60 years and over (%) 18.99 (6.13) 9.24–56.19 Proportion of Hispanic (%) 21.75 (16.81) 2.30–90.60 High school diploma (%) 56.68 (6.71) 33.46–76.09 Unemployment (%) 8.35 (3.27) 2.80–28.20 Mental health professionals (per 100,000) 43.63 (43.71) 0–306.08 Median home value ($10,000) 24.92 (14.95) 5.73–82.73 Depression prevalence (%) 13.46 (2.87) 6.56–26.47 Categorical variables n (%) Metropolitan county (yes) 140 (89.74) Community mental health center (yes) 69 (44.23) PCP shortage area  None of the county 18 (11.54)  One or more parts of the county 78 (50.00)  The whole county 60 (38.46) Mental health professional shortage area  None of the county 22 (14.10)  One or more parts of the county 81 (51.92)  The whole county 53 (33.97) Continuous variables M (SD) Range Ages 60 years and over (%) 18.99 (6.13) 9.24–56.19 Proportion of Hispanic (%) 21.75 (16.81) 2.30–90.60 High school diploma (%) 56.68 (6.71) 33.46–76.09 Unemployment (%) 8.35 (3.27) 2.80–28.20 Mental health professionals (per 100,000) 43.63 (43.71) 0–306.08 Median home value ($10,000) 24.92 (14.95) 5.73–82.73 Depression prevalence (%) 13.46 (2.87) 6.56–26.47 Categorical variables n (%) Metropolitan county (yes) 140 (89.74) Community mental health center (yes) 69 (44.23) PCP shortage area  None of the county 18 (11.54)  One or more parts of the county 78 (50.00)  The whole county 60 (38.46) Mental health professional shortage area  None of the county 22 (14.10)  One or more parts of the county 81 (51.92)  The whole county 53 (33.97) Note: PCP = Primary Care Health Professional. View Large Table 3 presents analyses from multilevel logistic regression analyses to examine whether county-level characteristics were associated with mental health service utilization by older Hispanics, controlling for individual characteristics. First, the unconditional model was tested to examine the county-to-county variability of mental health service utilization. The ICC was 0.156, indicating that the use of mental health services differed from county to county; that is, 15.6% of mental health service utilization by older Hispanics can be explained by county membership, indicating that employing a multilevel modeling approach was appropriate. The Akaike information criteria (AIC) was 567.302, which was used as the baseline information criterion for model fit. Table 3. Multilevel Logistic Regression for Mental Health Service Utilization Parameter Model 1 Model 2 Individual only Individual and County OR SE p CIL CIH OR SE p CIL CIH Intercept 7.48 15.48 .331 0.13 431.98 0.62 2.4 .902 0 1,166.62 Level 1 (individual-level) Age 0.96 0.02 .113 0.92 1.01 0.96 0.02 .102 0.92 1.01 Female 1.93 0.62 .041 1.03 3.64 1.89 0.62 .051 0.996 3.58 Married 0.55 0.17 .055 0.3 1.01 0.61 0.19 .116 0.32 1.13 High school diploma 2.24 0.72 .012 1.19 4.2 2.17 0.71 .018 1.14 4.12 Foreign born 1.54 0.61 .28 0.7 3.36 1.53 0.64 .305 0.68 3.47 No English spoken in home 0.65 0.25 .249 0.31 1.36 0.65 0.25 .26 0.3 1.38 Attitudes toward health 0.85 0.16 .389 0.59 1.23 0.84 0.16 .367 0.58 1.22 Income 0.91 0.1 .399 0.73 1.13 0.9 0.1 .367 0.73 1.13 Uninsured 0.66 0.31 .373 0.26 1.65 0.74 0.35 .529 0.29 1.89 SF-12 (Physical health) 0.998 0.01 .911 0.97 1.03 0.999 0.01 .942 0.97 1.03 SF-12 (Mental health) 0.95 0.02 .004 0.91 0.98 0.95 0.02 .005 0.91 0.98 K6 1.07 0.03 .035 1.004 1.14 1.07 0.03 .032 1.01 1.14 Level 2 (county-level) Ages 60 years and over 1.11 0.06 .048 1 1.23 Proportion of Hispanic 1.01 0.02 .7 0.98 1.03 Metropolitan county 0.69 0.64 .678 0.11 4.26 High school diploma 0.98 0.06 .775 0.87 1.11 Unemployment 0.94 0.06 .361 0.83 1.07 Community mental health center 3.42 1.6 .009 1.36 8.57 Primary Care Health Professional shortage area 1.68 0.66 .185 0.78 3.62 Mental health professional shortage area 0.7 0.29 .391 0.31 1.59 Mental health professionals 0.999 0.01 .913 0.99 1.01 Median home value 1.03 0.02 .116 0.99 1.07 Depression prevalence 1.03 0.05 .537 0.94 1.14 Random parameters B SE CIL CIH B SE CIL CIH Residual variance 0.78 0.27 0.4 1.53 0.55 0.3 0.19 1.6 Model fit AIC 429.624 431.557 ICC 0.156 0.083 Parameter Model 1 Model 2 Individual only Individual and County OR SE p CIL CIH OR SE p CIL CIH Intercept 7.48 15.48 .331 0.13 431.98 0.62 2.4 .902 0 1,166.62 Level 1 (individual-level) Age 0.96 0.02 .113 0.92 1.01 0.96 0.02 .102 0.92 1.01 Female 1.93 0.62 .041 1.03 3.64 1.89 0.62 .051 0.996 3.58 Married 0.55 0.17 .055 0.3 1.01 0.61 0.19 .116 0.32 1.13 High school diploma 2.24 0.72 .012 1.19 4.2 2.17 0.71 .018 1.14 4.12 Foreign born 1.54 0.61 .28 0.7 3.36 1.53 0.64 .305 0.68 3.47 No English spoken in home 0.65 0.25 .249 0.31 1.36 0.65 0.25 .26 0.3 1.38 Attitudes toward health 0.85 0.16 .389 0.59 1.23 0.84 0.16 .367 0.58 1.22 Income 0.91 0.1 .399 0.73 1.13 0.9 0.1 .367 0.73 1.13 Uninsured 0.66 0.31 .373 0.26 1.65 0.74 0.35 .529 0.29 1.89 SF-12 (Physical health) 0.998 0.01 .911 0.97 1.03 0.999 0.01 .942 0.97 1.03 SF-12 (Mental health) 0.95 0.02 .004 0.91 0.98 0.95 0.02 .005 0.91 0.98 K6 1.07 0.03 .035 1.004 1.14 1.07 0.03 .032 1.01 1.14 Level 2 (county-level) Ages 60 years and over 1.11 0.06 .048 1 1.23 Proportion of Hispanic 1.01 0.02 .7 0.98 1.03 Metropolitan county 0.69 0.64 .678 0.11 4.26 High school diploma 0.98 0.06 .775 0.87 1.11 Unemployment 0.94 0.06 .361 0.83 1.07 Community mental health center 3.42 1.6 .009 1.36 8.57 Primary Care Health Professional shortage area 1.68 0.66 .185 0.78 3.62 Mental health professional shortage area 0.7 0.29 .391 0.31 1.59 Mental health professionals 0.999 0.01 .913 0.99 1.01 Median home value 1.03 0.02 .116 0.99 1.07 Depression prevalence 1.03 0.05 .537 0.94 1.14 Random parameters B SE CIL CIH B SE CIL CIH Residual variance 0.78 0.27 0.4 1.53 0.55 0.3 0.19 1.6 Model fit AIC 429.624 431.557 ICC 0.156 0.083 Note: AIC = Akaike Information Criteria; ICC = Intraclass Correlation; OR = Odds ratio; CIL: Confidence Interval Low; CIH: Confidence Interval High. The bold coding indicates a statistical significance at p < .05. View Large Table 3. Multilevel Logistic Regression for Mental Health Service Utilization Parameter Model 1 Model 2 Individual only Individual and County OR SE p CIL CIH OR SE p CIL CIH Intercept 7.48 15.48 .331 0.13 431.98 0.62 2.4 .902 0 1,166.62 Level 1 (individual-level) Age 0.96 0.02 .113 0.92 1.01 0.96 0.02 .102 0.92 1.01 Female 1.93 0.62 .041 1.03 3.64 1.89 0.62 .051 0.996 3.58 Married 0.55 0.17 .055 0.3 1.01 0.61 0.19 .116 0.32 1.13 High school diploma 2.24 0.72 .012 1.19 4.2 2.17 0.71 .018 1.14 4.12 Foreign born 1.54 0.61 .28 0.7 3.36 1.53 0.64 .305 0.68 3.47 No English spoken in home 0.65 0.25 .249 0.31 1.36 0.65 0.25 .26 0.3 1.38 Attitudes toward health 0.85 0.16 .389 0.59 1.23 0.84 0.16 .367 0.58 1.22 Income 0.91 0.1 .399 0.73 1.13 0.9 0.1 .367 0.73 1.13 Uninsured 0.66 0.31 .373 0.26 1.65 0.74 0.35 .529 0.29 1.89 SF-12 (Physical health) 0.998 0.01 .911 0.97 1.03 0.999 0.01 .942 0.97 1.03 SF-12 (Mental health) 0.95 0.02 .004 0.91 0.98 0.95 0.02 .005 0.91 0.98 K6 1.07 0.03 .035 1.004 1.14 1.07 0.03 .032 1.01 1.14 Level 2 (county-level) Ages 60 years and over 1.11 0.06 .048 1 1.23 Proportion of Hispanic 1.01 0.02 .7 0.98 1.03 Metropolitan county 0.69 0.64 .678 0.11 4.26 High school diploma 0.98 0.06 .775 0.87 1.11 Unemployment 0.94 0.06 .361 0.83 1.07 Community mental health center 3.42 1.6 .009 1.36 8.57 Primary Care Health Professional shortage area 1.68 0.66 .185 0.78 3.62 Mental health professional shortage area 0.7 0.29 .391 0.31 1.59 Mental health professionals 0.999 0.01 .913 0.99 1.01 Median home value 1.03 0.02 .116 0.99 1.07 Depression prevalence 1.03 0.05 .537 0.94 1.14 Random parameters B SE CIL CIH B SE CIL CIH Residual variance 0.78 0.27 0.4 1.53 0.55 0.3 0.19 1.6 Model fit AIC 429.624 431.557 ICC 0.156 0.083 Parameter Model 1 Model 2 Individual only Individual and County OR SE p CIL CIH OR SE p CIL CIH Intercept 7.48 15.48 .331 0.13 431.98 0.62 2.4 .902 0 1,166.62 Level 1 (individual-level) Age 0.96 0.02 .113 0.92 1.01 0.96 0.02 .102 0.92 1.01 Female 1.93 0.62 .041 1.03 3.64 1.89 0.62 .051 0.996 3.58 Married 0.55 0.17 .055 0.3 1.01 0.61 0.19 .116 0.32 1.13 High school diploma 2.24 0.72 .012 1.19 4.2 2.17 0.71 .018 1.14 4.12 Foreign born 1.54 0.61 .28 0.7 3.36 1.53 0.64 .305 0.68 3.47 No English spoken in home 0.65 0.25 .249 0.31 1.36 0.65 0.25 .26 0.3 1.38 Attitudes toward health 0.85 0.16 .389 0.59 1.23 0.84 0.16 .367 0.58 1.22 Income 0.91 0.1 .399 0.73 1.13 0.9 0.1 .367 0.73 1.13 Uninsured 0.66 0.31 .373 0.26 1.65 0.74 0.35 .529 0.29 1.89 SF-12 (Physical health) 0.998 0.01 .911 0.97 1.03 0.999 0.01 .942 0.97 1.03 SF-12 (Mental health) 0.95 0.02 .004 0.91 0.98 0.95 0.02 .005 0.91 0.98 K6 1.07 0.03 .035 1.004 1.14 1.07 0.03 .032 1.01 1.14 Level 2 (county-level) Ages 60 years and over 1.11 0.06 .048 1 1.23 Proportion of Hispanic 1.01 0.02 .7 0.98 1.03 Metropolitan county 0.69 0.64 .678 0.11 4.26 High school diploma 0.98 0.06 .775 0.87 1.11 Unemployment 0.94 0.06 .361 0.83 1.07 Community mental health center 3.42 1.6 .009 1.36 8.57 Primary Care Health Professional shortage area 1.68 0.66 .185 0.78 3.62 Mental health professional shortage area 0.7 0.29 .391 0.31 1.59 Mental health professionals 0.999 0.01 .913 0.99 1.01 Median home value 1.03 0.02 .116 0.99 1.07 Depression prevalence 1.03 0.05 .537 0.94 1.14 Random parameters B SE CIL CIH B SE CIL CIH Residual variance 0.78 0.27 0.4 1.53 0.55 0.3 0.19 1.6 Model fit AIC 429.624 431.557 ICC 0.156 0.083 Note: AIC = Akaike Information Criteria; ICC = Intraclass Correlation; OR = Odds ratio; CIL: Confidence Interval Low; CIH: Confidence Interval High. The bold coding indicates a statistical significance at p < .05. View Large Second, individual predisposing, enabling, and need factors (Level 1) were added to the unconditional model to examine the relationships between individual-level variables and mental health service use (see Table 3). In Model 1, the AIC (429.624) showed improvement in the model fit compared to the unconditional model, but adding individual-level variables did not decrease the ICC (0.156). Model 1 identified four significant individual variables associated with mental health service utilization of older Hispanics: female, high school diploma, SF-12 mental health, and K6. Being a woman (odds ratio [OR] = 1.93) and having a high school diploma (OR = 2.24) increased the odds of using mental health services. People with a higher SF-12 mental health score (i.e., better mental health status) were less likely to use mental health services (OR = 0.95), while higher levels of psychological distress (K6) were related to increased odds of using mental health services (OR = 1.07). However, the variables of age, foreign born, language spoken at home, attitude toward health care, income, uninsured, and SF-12 (physical health) were not predictive of mental health service utilization. Third, Model 2 added 11 county-level characteristics to determine whether they were predictive of mental health service utilization among older Hispanics independent of individual characteristics. Compared to Model 1, the AIC did not show improvement (431.557) in Model 2, but adding county-level variables decreased the ICC (from 0.156 in Model 1 to 0.083 in Model 2). Model 2 identified several significant variables to explain mental health service utilization. The individual variables that were predictive of mental health service utilization in Model 1 remained significant and in the same direction in Model 2, with the exception of female (p = .051). Having a high school diploma (OR = 2.17), and higher levels of psychological distress (OR = 1.07) increased the odds of using mental health services; those with better SF-12 mental health status (OR = 0.95) were less likely to use a mental health service. Among the county-level characteristics, two variables were predictive of mental health service utilization by older Hispanics. The existence of a community mental health center increased the odds of using mental health services (OR = 3.42), and it had the largest odds ratio among the significant variables. In addition, living in a county with a higher proportion of older adults increased the odds of using mental health services among older Hispanics (OR = 1.11). However, the other variables—proportion of Hispanic, metropolitan county, high school diploma, unemployment rate, Primary Care Health Professional shortage area, median home value, and county mood disorder prevalence—were not predictive of mental health service utilization by older Hispanics. Discussion and Implications The findings of this study support that individual-level characteristics are essential to understanding mental health service use. Above all, better mental health status decreased the odds of using mental health services by older Hispanics. The findings were consistent with those from a systematic review of studies using Andersen’s model; greater perceived mental health needs were related to greater service utilization (Andersen, 2008; Babitsch, Gohl, & von Lengerke, 2012). The significant relationship between greater needs and greater service use makes intuitive sense: people with a mental health problem are more likely to seek out mental health treatment, although other factors may influence greater service use as well. The present study demonstrates that having a high school diploma was associated with increased use of mental health services. This finding highlights the role of education in the use of mental health services by older Hispanics (Gonzalez et al., 2015; Sorkin et al., 2016). Educational attainment of older Hispanics (45% were high school graduates in this study) is relatively low compared to other racial and ethnic groups. Older Hispanics might have experienced disadvantages regarding educational attainment during their early years; as a result, they may have limited employment opportunities and may have accumulated limited financial resources. Often combined with financial difficulties, lower levels of education may limit access to mental health services as well as adherence to evidence-based interventions and medications (Aranda, 2016). Also, education is closely related to health literacy, and low health literacy can impact obtaining appropriate health care. Those with a higher level of education may be more likely to be aware of symptoms and community resources, and as a result, may be more able to utilize a resource that they need for care. Hispanic adults tend to report lower levels of health literacy compared to other racial and ethnic groups (White, 2008). Therefore, disadvantaged education levels and low health literacy of older Hispanics should be addressed in future research and practice. The relationship between community resources and mental health service utilization by older Hispanics was also shown at the county level. Here, the existence of a community mental health center increased the odds of mental health services utilization by older Hispanics, compared to counties without a community mental health center. This finding was consistent with a previous study that reported that Hispanic adults living in a county with a community mental health center were more likely to use any mental health services in the past year (Cook et al., 2013). Community mental health centers may be particularly effective in providing culturally sensitive substance abuse and mental health services to their clients (Cummings, Cassie, & Trecartin, 2016). According to SAMHSA (2016), about 61% of community mental health centers provided treatment services in languages other than English, including Spanish (45.7%) and other languages (11.6%) in 2014. Although this study shows that older Hispanics living in counties with a community mental health center are more likely to report mental health services utilization, it is important to remember that this does not necessarily mean that they used mental health services at a community mental health center. Community mental health services may play an important role in mental health utilization by older Hispanics in offering affordable care. As shown in the sample characteristics, the Hispanic sample in this study was more likely to be uninsured (17%). This is in line with other studies that have found that, older undocumented Hispanics tend to lack health insurance. Earlier studies with Latino immigrants report that the undocumented Latinos were entirely uninsured (Perez & Fortuna, 2005), and that undocumented status is one of the biggest barriers to access health care services (Nandi et al., 2008). Thus, it is possible that older Hispanics are more likely to visit a community mental health center, instead of visiting other health service providers because of their insurance status. Indeed, community mental health centers have been discussed as one promising community safety net to serve underserved populations (Cook et al., 2013). For example, the U.S. Department of Health and Human Services requires that a community mental health center should provide at least 40% of its services to people who are not eligible for Medicare (American Medical Association [AMA], 2015). Future research should look into how the existence of community mental health centers influences mental health service utilization by older Hispanics. More detailed information on clients’ sociodemographic status (e.g., health insurance and undocumented status), staff (e.g., numbers of staff speaking Spanish and racial concordance), services provided (e.g., integrated mental health and health service in the center), proximity to a center, and a center’s funding structure would be helpful to understand mental health services provided to older Hispanics by community mental health centers. A higher proportion of older adults in a county increased the odds of mental health services utilization by older Hispanics. Auchincloss, Van Nostrand, and Ronsaville (2001) also found that a higher proportion of older adults in a neighborhood was associated with increased access to health care services. One possible explanation is that communities with a higher proportion of older adults have a different combination of available health services and health care facilities for older adults compared to counties with younger adults. Also, older adults living in counties with similar age groups may have a higher likelihood of obtaining health-related information from neighborhood peers through social interactions and building up their networks for health care services access. The mechanisms appear complex regarding positive links between mental health services utilization and a high proportion of older adults living in the same geographic area. Although communities with higher proportions of older adults may have more age-appropriate available health care services and more health care facilities, resulting from more demand for services (Andersen, Davidson, & Baumeister, 2014), we do not have information on community-level resources such as the characteristics of health care facilities, the number of mental health care professionals with similar cultural background to older racial and ethnic minorities and the existence of social service organizations serving older racial and ethnic minorities. Social interactions in the communities with a higher proportion of older adults also can be different from those with a lower proportion of older adults. Communities with a higher proportion of older adults are more likely to be stable and may have greater social networks and support resulting in positive health outcomes (Cagney, 2006). Thus, understanding characteristics of the communities with higher proportions of older adults will be helpful to examine how age composition affects mental health service utilization by older Hispanics. This study has limitations that should be noted. First, the narrow definition of mental health service utilization (i.e., an actual mental health professional visit) may misrepresent mental health service use among older Hispanics. They may prefer talking about their mental health concerns to spiritual providers and nonmental health professionals from the human service sector (Villatoro, Morales, & Mays, 2014). Another limitation is that this study could not include other plausible variables (e.g., acculturation, assimilation, discrimination, stigma, relationships between a provider and a patient) to explain mental health services use. Although this study includes a scale about attitudes toward health, older racial and ethnic minorities may have limited knowledge of mental health and/or may have culturally different beliefs about mental health services (Alvarez et al., 2014) that may not be captured on this scale. In addition, using counties as a unit of analysis may have limitations for increasing our understanding of mental health service utilization by older Hispanics. For example, about 52% of counties were partially mental health professional shortage areas. These patterns indicate that variability exists within counties. Therefore, future research should examine smaller units of analysis to better understand the impact of within-county variability. Despite its limitations, this study has several strengths. Above all, this study suggests that county contexts may be helpful to understand mental health service utilization among older Hispanics, even adjusting for individual-level characteristics. Most studies have focused on examining relationships between individual characteristics and mental health service utilization by older racial and ethnic minorities (e.g., Gonzalez et al., 2015; Jimenez et al., 2013). Although previous studies have explored the role of county or community characteristics to explain mental health service utilization, most have focused on the general population, not exclusively on older adults (e.g., Cook et al., 2013; Stahler, Mennis, Cotlar, & Baron, 2009; Stockdale, Tang, Zhang, Belin, & Wells, 2007). Another strength is that this study used actual service visits obtained from health and mental health service providers to measure mental health service utilization. Mental health service utilization has been inconsistently operationalized in studies. Some studies define mental health service use as visiting specialty mental health professionals (e.g., Dobalian & Rivers, 2008; Ojeda & McGuire, 2006). Other studies ask whether respondents talk to professionals from the human service sector and spiritual providers to capture any efforts to receive mental health services (e.g., Fortuna, Porche, & Alegria, 2008; Kim et al., 2013). Depending on respondents’ memory may result in recall bias, and household respondents may not provide accurate answers. This study’s use of the MEPS, which draws its data from medical providers’ records, ensures that findings are not subject to recall bias. In sum, this study aimed to examine whether county contexts are associated with mental health service utilization by older Hispanics. Understanding county context where people reside is important because people use the mental health services that are in their communities. Both individual- and county-level characteristics affected mental health service utilization by older Hispanics. For example, this study identified that county characteristics (i.e., existence of community mental health centers and a higher proportion of older adults) increased mental health service utilization as did individual characteristics of higher levels of education and worse mental health status. In this way, this study identified different factors affecting mental health service use by older Hispanics, and will contribute to the existing literature that generally has focused on the role of individual characteristics to understand mental health service utilization. Understanding both individual- and county-specific characteristics associated with mental health service utilization is essential to develop tailored outreach programs for older Hispanics as well as to develop and evaluate mental health policy. Acknowledgements The author acknowledges the support of the Agency for Healthcare Research Quality (AHRQ) and the data analysis in this paper was conducted at the AHRQ data center. The author specially thanks Dr. Donna Harrington for her immeasurable source of support in this study. References Agency for Healthcare Research and Quality . ( 2013 ). MEPS HC-147 2011 full year consolidated data file . Rockville, MD : Author . Akincigil , A. , Olfson , M. , Siegel , M. , Zurlo , K. A. , Walkup , J. T. , & Crystal , S . ( 2012 ). Racial and ethnic disparities in depression care in community-dwelling elderly in the United States . American Journal of Public Health , 102 , 319 – 328 . doi: 10.2105/AJPH.2011.300349 Google Scholar CrossRef Search ADS PubMed Alvarez , P. , Rengifo , J. , Emrani , T. , & Gallagher-Thompson , D . ( 2014 ). Latino older adults and mental health: A review and commentary . Clinical Gerontologist , 37 , 33 – 48 . doi: 10.1080/07317115.2013.847514 Google Scholar CrossRef Search ADS American Medical Association . ( 2015 ). Mental health services . Retrieved from https://www.cms.gov/Outreach-and-Education/Medicare-Learning-Network-MLN/MLNProducts/Downloads/Mental-Health-Services-Booklet-ICN903195.pdf. Andersen , R . ( 1968 ). A behavioral model of families’ use of health services . (Research series. No. 25). Chicago, IL : The Center of Health Administration Studies . Andersen , R. M . ( 2008 ). National health surveys and the behavioral model of health services use . Medical Care , 46 , 647 – 653 . doi: 10.1097/MLR.0b013e31817a835d Google Scholar CrossRef Search ADS PubMed Andersen , R. , & Davidson , P . ( 2007 ). Improving access to care in America: Individual and contextual indicators . In R. Andersen , T. Rice & J. Kominski (Eds.), Changing the U.S. health care system: Key issues in health services policy and management (pp. 3 – 31 ). San Francisco, CA : John Wiley & Sons . Andersen , R. , Davidson , P. , & Baumeister , S . ( 2014 ). Improving access to care . In G. Kominski (Ed.), Changing the U.S. health care system: Key issues in health services policy and management (pp. 33 – 64 ). San Francisco, CA : Jossey-Bass . Aranda , M . ( 2016 ). Social work practice with older Latinos in the United States: A mental health perspective . In D. Kaplan & B. Berkman (Eds.), The Oxford handbook of social work in health and aging (pp. 527 – 533 ). New York, NY : Oxford University Press . Auchincloss , A. H. , Van Nostrand , J. F. , & Ronsaville , D . ( 2001 ). Access to health care for older persons in the United States: Personal, structural, and neighborhood characteristics . Journal of Aging and Health , 13 , 329 – 354 . doi: 10.1177/089826430101300302 Google Scholar CrossRef Search ADS PubMed Babitsch , B. , Gohl , D. , & von Lengerke , T . ( 2012 ). Re-revisiting Andersen’s behavioral model of health services use: A systematic review of studies from 1998-2011 . Psycho-social Medicine , 9 , 1 – 15 . doi: 10.3205/psm000089 Bickel , R . ( 2007 ). Multilevel analysis for applied research: It’s just regression! New York, NY : The Guilford Press. Cagney , K. A . ( 2006 ). Neighborhood age structure and its implications for health . Journal of Urban Health: Bulletin of the New York Academy of Medicine , 83 , 827 – 834 . doi: 10.1007/s11524-006-9092-z Google Scholar CrossRef Search ADS PubMed Choi , S . ( 2015 ). How does satisfaction with medical care differ by citizenship and nativity status?: A county-level multilevel analysis . The Gerontologist , 55, 735-747. doi: 10.1093/geront/gnt201 Cook , B. L. , Doksum , T. , Chen , C. N. , Carle , A. , & Alegría , M . ( 2013 ). The role of provider supply and organization in reducing racial/ethnic disparities in mental health care in the U.S . Social Science & Medicine (1982) , 84 , 102 – 109 . doi: 10.1016/j.socscimed.2013.02.006 Google Scholar CrossRef Search ADS PubMed Cummings , S. , Cassie , K. , & Trecartin , S . ( 2016 ). Community mental health centers . In D. Kaplan & B. Berkman (Eds.), The Oxford handbook of social work in health and aging (pp. 141 – 149 ). New York, NY : Oxford University Press . Google Scholar CrossRef Search ADS DeFranco , E. A. , Lian , M. , Muglia , L. A. , & Schootman , M . ( 2008 ). Area-level poverty and preterm birth risk: A population-based multilevel analysis . BMC Public Health , 8 , 316 . doi: 10.1186/1471-2458-8-316 Google Scholar CrossRef Search ADS PubMed Dobalian , A. , & Rivers , P. A . ( 2008 ). Racial and ethnic disparities in the use of mental health services . The Journal of Behavioral Health Services & Research , 35 , 128 – 141 . doi: 10.1007/s11414-007-9097-8 Google Scholar CrossRef Search ADS PubMed Federal Interagency Forum on Aging-Related Statistics . ( 2016 ). Older Americans 2016: Key indicators of well-being . Washington, DC : U.S. Government Printing Office . Fortuna , L. R. , Porche , M. V. , & Alegria , M . ( 2008 ). Political violence, psychosocial trauma, and the context of mental health services use among immigrant Latinos in the United States . Ethnicity & Health , 13 , 435 – 463 . doi: 10.1080/13557850701837286 Google Scholar CrossRef Search ADS PubMed Fuentes , D. , & Aranda , M. P . ( 2012 ). Depression interventions among racial and ethnic minority older adults: A systematic review across 20 years . The American Journal of Geriatric Psychiatry: Official Journal of the American Association for Geriatric Psychiatry , 20 , 915 – 931 . doi: 10.1097/JGP.0b013e31825d091a Google Scholar CrossRef Search ADS PubMed Gonzalez , J. M. , Applewhite , S. R. , & Barrera , I . ( 2015 ). Older Mexican Americans: Pathways to mental health service use . Social Work in Mental Health , 13 , 390 – 414 . doi: 10.1080/15332985.2014.927813 Google Scholar CrossRef Search ADS Institute of Medicine . ( 2012 ). The mental health and substance use workforce for older adults: In whose hands? Washington, DC : The National Academies Press . Jimenez , D. E. , Cook , B. , Bartels , S. J. , & Alegría , M . ( 2013 ). Disparities in mental health service use of racial and ethnic minority elderly adults . Journal of the American Geriatrics Society , 61 , 18 – 25 . doi: 10.1111/jgs.12063 Google Scholar CrossRef Search ADS PubMed Kaskie , B. , Van Gilder , R. , & Gregory , D . ( 2008 ). Community mental health service use by older adults in California . Aging & Mental Health , 12 , 134 – 143 . doi: 10.1080/13607860801942761 Google Scholar CrossRef Search ADS PubMed Kessler , R. C. , Andrews , G. , Colpe , L. J. , Hiripi , E. , Mroczek , D. K. , Normand , S. L. ,… Zaslavsky , A. M . ( 2002 ). Short screening scales to monitor population prevalence and trends in non-specific psychological distress . Psychological Medicine , 32 , 959 – 976 . doi: 10.1017/S0033291702006074 Google Scholar CrossRef Search ADS PubMed Kim , G. , Jang , Y. , Chiriboga , D. A. , Ma , G. X. , & Schonfeld , L . ( 2010 ). Factors associated with mental health service use in Latino and Asian immigrant elders . Aging & Mental Health , 14 , 535 – 542 . doi: 10.1080/13607860903311758 Google Scholar CrossRef Search ADS PubMed Kim , G. , Parton , J. M. , DeCoster , J. , Bryant , A. N. , Ford , K. L. , & Parmelee , P. A . ( 2013 ). Regional variation of racial disparities in mental health service use among older adults . The Gerontologist , 53 , 618 – 626 . doi: 10.1093/geront/gns107 Google Scholar CrossRef Search ADS PubMed Mackenzie , C. S. , Pagura , J. , & Sareen , J . ( 2010 ). Correlates of perceived need for and use of mental health services by older adults in the collaborative psychiatric epidemiology surveys . The American Journal of Geriatric Psychiatry: Official Journal of the American Association for Geriatric Psychiatry , 18 , 1103 – 1115 . doi: 10.1097/JGP.0b013e3181dd1c06 Google Scholar CrossRef Search ADS PubMed Nandi , A. , Galea , S. , Lopez , G. , Nandi , V. , Strongarone , S. , & Ompad , D. C . ( 2008 ). Access to and use of health services among undocumented Mexican immigrants in a US urban area . American Journal of Public Health , 98 , 2011 – 2020 . doi: 10.2105/AJPH.2006.096222 Google Scholar CrossRef Search ADS PubMed National Alliance on Mental Illness . ( 2009). Depression in older persons fact sheet . Retrieved from http://www.nami.org/Template.cfm?Section=By_Illness&template=/ContentManagement/ContentDisplay.cfm&ContentID=7515. Ojeda , V. D. , & McGuire , T. G . ( 2006 ). Gender and racial/ethnic differences in use of outpatient mental health and substance use services by depressed adults . The Psychiatric Quarterly , 77 , 211 – 222 . doi: 10.1007/s11126-006-9008-9 Google Scholar CrossRef Search ADS PubMed Pérez , M. C. , & Fortuna , L . ( 2005 ). Chapter 6. Psychosocial stressors, psychiatric diagnoses and utilization of mental health services among undocumented immigrant Latinos . Journal of Immigrant & Refugee Services , 3 , 107 – 123 . doi: 10.1300/J191v03n01_06 Google Scholar CrossRef Search ADS Singh , G. K . ( 2003 ). Area deprivation and widening inequalities in US mortality, 1969-1998 . American Journal of Public Health , 93 , 1137 – 1143 . Google Scholar CrossRef Search ADS PubMed Sorkin , D. H. , Murphy , M. , Nguyen , H. , & Biegler , K. A . ( 2016 ). Barriers to mental health care for an ethnically and racially diverse sample of older adults . Journal of the American Geriatrics Society , 64 , 2138 – 2143 . doi: 10.1111/jgs.14420 Google Scholar CrossRef Search ADS PubMed Sorkin , D. H. , Pham , E. , & Ngo-Metzger , Q . ( 2009 ). Racial and ethnic differences in the mental health needs and access to care of older adults in california . Journal of the American Geriatrics Society , 57 , 2311 – 2317 . doi: 10.1111/j.1532-5415.2009.02573.x Google Scholar CrossRef Search ADS PubMed Stahler , G. J. , Mennis , J. , Cotlar , R. , & Baron , D. A . ( 2009 ). The influence of neighborhood environment on treatment continuity and rehospitalization in dually diagnosed patients discharged from acute inpatient care . The American Journal of Psychiatry , 166 , 1258 – 1268 . doi: 10.1176/appi.ajp.2009.08111667 Google Scholar CrossRef Search ADS PubMed StataCorp . ( 2015 ). Stata (Version 14) [Software] . College Station, TX : Author . Stockdale , S. E. , Tang , L. , Zhang , L. , Belin , T. R. , & Wells , K. B . ( 2007 ). The effects of health sector market factors and vulnerable group membership on access to alcohol, drug, and mental health care . Health Services Research , 42 ( 3 Pt 1 ), 1020 – 1041 . doi: 10.1111/j.1475-6773.2006.00636.x Google Scholar CrossRef Search ADS PubMed Substance Abuse and Mental Health Service Administration [SAMHSA] . ( 2011 ). Leading change: A plan for SAMHSA’s role and actions 2011–2014 . Retrieved from http://store.samhsa.gov/shin/content//SMA11-4629/01-FullDocument.pdf. Substance Abuse and Mental Health Service Administration [SAMHSA] . ( 2016 ). National Mental Health Services Survey (N-MHSS): 2014: Data on mental health treatment facilities . HHS Publication No. (SMA) 16–5000. Rockville, MD : Author . U.S. Department of Health and Human Services . ( 2018 ). Health insurance and mental health services . Retrieved from https://www.mentalhealth.gov/get-help/health-insurance/. Villatoro , A. P. , Morales , E. S. , & Mays , V. M . ( 2014 ). Family culture in mental health help-seeking and utilization in a nationally representative sample of Latinos in the United States: The NLAAS . The American Journal of Orthopsychiatry , 84 , 353 – 363 . doi: 10.1037/h0099844 Google Scholar CrossRef Search ADS PubMed Ware , J. Jr , Kosinski , M. , & Keller , S. D . ( 1996 ). A 12-Item short-form health survey: Construction of scales and preliminary tests of reliability and validity . Medical Care , 34 , 220 – 233 . Google Scholar CrossRef Search ADS PubMed Wei , W. , Sambamoorthi , U. , Olfson , M. , Walkup , J. T. , & Crystal , S . ( 2005 ). Use of psychotherapy for depression in older adults . The American Journal of Psychiatry , 162 , 711 – 717 . doi: 10.1176/appi.ajp.162.4.711 Google Scholar CrossRef Search ADS PubMed White , S . ( 2008 ). Assessing the nation’s health literacy. Key concepts and findings of the National Assessment of Adult Literacy . Chicago, IL : American Medical Association . © The Author(s) 2018. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

The GerontologistOxford University Press

Published: Apr 16, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off