Associations Between Cross-Sector Partnerships and Local Health Department Participation in Population-Based Activities to Prevent Mental Health Conditions

Abstract

Objectives. To describe partnerships between US local health departments (LHDs) and community organizations and assess the relationship between the types of activities performed in these partnerships and LHD engagement in population-based activities to prevent mental health conditions.

Methods. Data were derived from 457 LHDs that responded to module 1 of the 2016 Profile Study conducted by the National Association of County and City Health Officials. These data were used to assess the presence of partnerships with community organizations and examine associations between the types of activities performed in such partnerships and LHDs’ participation in population-based activities to prevent mental health conditions.

Results. LHDs had higher odds of participating in population-based activities to prevent mental health conditions if they shared personnel or resources or had written agreements with mental health or substance use disorder providers, held regularly scheduled meetings with hospitals, or shared personnel or resources with community health centers. Odds were reduced if they exchanged information with community health centers or shared personnel or resources with faith-based organizations.

Conclusions. This study offers an improved understanding of how the types of activities performed in cross-sector partnerships affect LHDs’ participation in population-based activities to prevent mental health conditions, which is important as public policies, programs, and funding initiatives continue to encourage cross-sector partnership building.

Approximately 1 in 5 adults have a mental health condition, and rates of depression, suicide, and anxiety among youths have increased significantly in the past decade.^1,2 However, a majority of adults and children with mental health conditions are not using mental health services.¹

Population-based approaches to mental health, or nonclinical interventions and activities intended to improve mental health outcomes and the determinants of such outcomes (e.g., convening cross-sector stakeholders to coordinate mental health initiatives or advocate for public policies that improve population mental health), are gaining momentum as an opportunity to combat rising rates of mental health conditions.³ National initiatives such as the Affordable Care Act and prioritization of mental health as one of the 12 Healthy People 2020 objectives demonstrate substantial interest in integrating mental health into public health practice.^4–6 Consequently, several national initiatives have identified local health departments (LHDs) as a stakeholder group that could address population mental health by integrating population-based mental health approaches into their prevention and programming work.^4,7,8

LHDs are well suited to engage in population-based mental health primary prevention, or activities focused on targeting the population before mental health conditions occur, because approximately three quarters of these departments report doing community primary prevention work focused on nutrition and tobacco and half report doing work focused on physical activity, chronic disease, and unintended pregnancy.⁹ Despite momentum at the national level to encourage LHDs to address mental health,^10–16 less than a quarter of LHDs participate in population-based primary prevention work focused on mental health conditions.⁹

Research suggests that perceived organizational boundaries with other local organizations and agencies, as well as limited internal knowledge about mental health may be potential barriers for LHDs to participate in population-based activities to prevent mental health conditions (hereafter “mental health prevention”).¹⁶ In recent years, the role of LHDs has shifted significantly from addressing infectious disease and sanitation to addressing the social determinants of health^17,18 (e.g., transportation, housing, access to food).

Many programs and policies that address social determinants of health traditionally fall under the purview of other local agencies or departments. As a result, public health now includes nontraditional partners such as housing agencies, local planning and development agencies, and faith communities in addition to traditional health partners such as hospitals, health insurance companies, and community health centers.¹⁹ The new model for public health practice, Public Health 3.0 (outlined by the Office of the Assistant Secretary of Health in 2016), acknowledges the importance of nontraditional and traditional health partners as well as the downfall of the existing public health infrastructure by listing “cross-sector community partnerships” as one of 5 key recommendations for public health moving forward.^4,7

Cross-sector partnerships may provide an opportunity for LHDs to maximize use of community knowledge and resources by developing complementary strategies to address mental health in the populations they serve.^18,20,21 Previous studies leveraging theories from network science have shown that the numbers and types of organizations in each public health collaborative (a network of diverse organizations in a community with the intent to work on public health issues) vary according to the goal of the collaborative (e.g., chronic disease vs infant mortality).^19,22 Research suggests that partnerships that succeed in creating change in communities and systems need to be cross-sectoral, engaging partners at different levels and across different sectors (e.g., private industry, nonprofit organizations).^22,23 LHDs can strategically form relationships to improve their own structural capacity by engaging with organizations to leverage additional funding or access technical assistance.²⁴

The Process Framework for Public Health Collaboratives provides a way to measure how organizations work together to achieve specific public health outcomes.²³ Collective decision-making, resource sharing, and connectivity are crucial characteristics of public health partnerships that may lead to improved outcomes.²³ Several studies have explored the composition of public health collaboratives and key characteristics of successful partnerships, but no investigations to my knowledge have looked quantitatively at how specific types of activities (e.g., sharing personnel or resources, exchanging information) engaged in by LHDs and their collaborative partners influence LHDs’ population-based mental health prevention work.

This study sought to build on previous research by (1) describing the presence of partnerships between LHDs and community organizations and (2) assessing associations between the types of activities performed in partnership with community organizations and LHDs’ participation in population-based mental health prevention.

METHODS

This study incorporated data from 457 LHDs that responded to module 1 of the 2016 Profile Study, a Web-based survey conducted by the National Association of County and City Health Officials (NACCHO) that is widely used to characterize the function and structure of LHDs.^9,25 NACCHO maintains a comprehensive list of LHDs in the United States, and this list served as the sampling frame for the core 2016 Profile Study questionnaire (n = 2533). A total of 1930 LHDs completed the core questionnaire (response rate = 76%).

In addition to the core questionnaire, NACCHO selected 2 groups of LHDs, using stratified random sampling with strata defined by the size of the population served by the LHD, to receive one of 2 sets of supplemental questions (module 1 and module 2).⁹ Questions on partnerships and collaborations were part of module 1 (n = 484 LHDs; response rate = 80%).⁹ The Profile Study does not include local behavioral health departments or agencies.

LHDs were excluded from the present analysis if they did not respond to the item in the core questionnaire related to participation in population-based mental health prevention (n = 16) or the items in module 1 related to partnerships and collaborations (n = 11).

Dependent Variable

The primary dependent variable was LHD participation in population-based mental health prevention. Respondents were queried regarding “whether and how your LHD and other organizations provided [population-based mental health prevention] in your jurisdiction during the past year.” Five response options were provided (“performed by LHD directly,” “contracted out by LHD,” “provided by others in community independent of LHD funding,” “not available in community,” and “don’t know”). Consistent with how Profile Study service variables have been collapsed in previous research,^14,15 departments that responded either “performed by LHD directly” or “contracted out by LHD” were coded as participating in population-based mental health prevention.

Independent Variables

The primary independent variables were presence of partnerships with organizations and the type of activities performed in these organizational partnerships. LHDs were instructed as follows: “Check each way that your LHD has worked with each organization in the past year.” Organization types were “mental health/substance use disorder providers,” “hospitals,” “community health centers,” “colleges or universities,” “community-based nonprofits,” “local planning and developmental agencies,” “housing agencies,” “transportation,” “libraries,” and “faith communities.” LHDs were coded as having a relationship with an organization if they reported that they had “shared personnel or resources,” had a “written agreement,” had “regularly scheduled meetings,” or “exchanged information” with the organization in the past year. Binary variables were created to measure the presence of a partnership with each organization.

As noted, the secondary primary variable of interest was the type of activity in which the LHD engaged with the organization (sharing of personnel or resources, written agreement, regularly scheduled meetings, exchange of information, or no relationship). Binary variables (yes or no) were created to assess the types of activities performed between LHDs and organizations. An LHD could perform more than 1 type of activity with an organization.

Other LHD-level covariates were also included. Environmental variables included LHD population size (less than 25 000, 25 000–49 999, 50 000–99 999, 100 000–499 999, 500 000 or more), jurisdiction type (city, county, multicity, multicounty), and LHD governance classification (state government, local government, governed by both state and local authorities). Structural capacity was measured according to membership in a combined health and human services agency (yes or no).

Statistical Analysis

Bivariate analyses, including χ² tests and 2-tailed t-tests, assessed differences in covariates by LHD participation in population-based primary prevention. Logistic regression was used to examine associations between LHD participation in population-based primary prevention and (1) partnerships with community organizations and (2) the types of activities performed in organizational partnerships. All analyses were conducted in Stata version 15²⁶ using survey weights produced by NACCHO to be representative of all LHDs in the United States.^9,14 The study was completed in 2018.

RESULTS

Table 1 shows weighted descriptive statistics for LHD characteristics and partnership characteristics stratified by LHD participation in population-based mental health prevention (additional details on the stratified sample are provided in Table A, available as a supplement to the online version of this article at http://www.ajph.org). LHDs that did and did not participate in population-based mental health prevention differed significantly with respect to population size, membership in a combined health and human services agency, and average number of partnerships. A higher percentage of LHDs participating in population-based mental health prevention served populations of 50 000 to 99 999 residents (20.7% vs 14.2%; P = .001) and populations of 500 000 or more residents (11.4% vs 3.4%; P = .001).

TABLE 1—

Local Health Department (LHD) Characteristics and Cross-Sector Partnerships Stratified by Participation in Population-Based Activities to Prevent Mental Health Conditions: National Association of County and City Health Officials Profile Study, United States, 2016

LHD or Partnership Characteristic	All LHDs (n = 457), % or Mean (SD)	LHDs That Do Not Participate in Mental Health Prevention (n = 356), % or Mean (SD)	LHDs That Participate in Activities to Prevent Mental Health Conditions (n = 101), % or Mean (SD)	χ2 Pa
LHD population size				.001
< 25 000	40.1	43.0	27.2
25 000–49 999	21.4	22.2	18.2
50 000–99 999	15.4	14.2	20.7
100 000–499 999	18.2	17.2	22.5
≥ 500 000	4.9	3.4	11.4
Governance				.03
Governed by state government	19.2	19.4	18.2
Governed by local government	71.3	72.9	65.1
Governed by both	9.5	7.7	16.7
Part of local health and human services agency				.023
Yes	19.0	16.8	27.7
No	81.0	83.2	72.3
Jurisdiction type				.71
City	15.4	16.2	12.7
County	72.8	72.5	73.9
Multicounty	8.1	7.5	10.3
Multicity	3.7	3.8	3.1
Partnership with organization
Mental health/substance use disorder provider	78.9	75.4	93.9	<.001
Hospital	88.0	85.7	97.8	.004
Community health center	65.4	64.4	69.6	.39
University or college	65.7	63.3	75.4	.05
Community-based nonprofit	86.3	84.8	92.5	.09
Local planning or development agency	66.3	63.5	78.0	.013
Housing agency	54.7	51.3	69.2	.004
Transportation	50.0	46.8	62.4	.01
Library	66.8	64.3	77.5	.022
Faith-based organization	82.4	80.9	88.4	.12
No. of partnerships	7.7 (0.15)	7.5 (0.17)	8.8 (0.30)	< .001
Type of activity performed
Sharing of personnel/resources	33.4	30.3	46.5	.004
Written agreement	41.8	37.9	58.4	< .001
Regularly scheduled meetings	69.7	66.5	82.9	.003
Exchange of information	92.7	92.5	93.4	.7

Note. All estimates were calculated with nationally representative survey weights.

^aFor differences between LHDs that did and did not participate in activities to prevent mental health conditions.

LHDs that participated in population-based mental health prevention were more likely to be part of a combined local health and human services agency (27.7% vs 16.8%; P = .023) and less likely to be governed solely by a local government (65.1% vs 72.9%, P = .03). LHDs that participated in population-based mental health prevention had a significantly higher average number of partnerships than LHDs that did not (8.8 vs 7.5; P < .001).

The numbers and types of cross-sector partnerships for LHDs that participated in population-based mental health prevention differed significantly. LHDs that participated in population-based mental health prevention were more likely than those that did not to partner with hospitals (97.8% vs 85.7%; P = .004), mental health or substance use disorder providers (93.9% vs 75.4%; P < .001), universities (75.4% vs 63.3%; P = .05), local planning and development agencies (78.0% vs 63.5%; P = .013), housing agencies (69.2% vs 51.3%; P = .004), transportation (62.4% vs 46.8%; P = .01), and libraries (77.5% vs 64.3%; P = .022). Also, they were more likely to be involved in partnerships in which they shared personnel or resources (46.5% vs 30.3%; P = .004), had written agreements (58.4% vs 37.9%; P < .001), and had regularly scheduled meetings (82.9% vs 66.5%; P = .003).

Table 2 shows adjusted associations between the presence of an organizational partnership and LHD participation in population-based mental health prevention. After adjustment for average number of partnerships, governance structure, population size, and LHD membership in a combined health and human services agency, partnerships with specific types of organizations were significantly associated with LHDs’ participation in population-based mental health prevention. The odds of participating in population-based mental health prevention were approximately 5 times higher among LHDs that had partnerships with mental health or substance use disorder providers (adjusted odds ratio [AOR] = 4.71; 95% confidence interval [CI] = 1.39, 15.97) or hospitals (AOR = 5.19; 95% CI = 1.21, 22.18) than among LHDs that did not have partnerships with these types of organizations.

TABLE 2—

Adjusted Associations Between Cross-Sector Partnerships and Local Health Department (LHD) Participation in Population-Based Activities to Prevent Mental Health Conditions: National Association of County and City Health Officials Profile Study, United States, 2016

Partnering Organization	AOR (95% CI)
Mental health/substance use disorder provider	4.71 (1.39, 15.97)
Hospital	5.19 (1.21, 22.18)
Community health center	0.74 (0.38, 1.43)
University or college	0.82 (0.36, 1.86)
Community-based nonprofit	1.07 (0.39, 2.99)
Local planning or development agency	1.55 (0.74, 3.23)
Housing agency	1.64 (0.83, 3.22)
Transportation	1.41 (0.69, 2.89)
Library	1.88 (0.90, 3.90)
Faith-based organization	0.72 (0.30, 1.70)

Note. AOR = adjusted odds ratio; CI = confidence interval. All estimates were calculated with nationally representative survey weights and mutually adjusted. Covariates included the log transformation of LHD population size, jurisdiction type, governance type, whether the LHD is part of a local health and human services department, and the number of partnerships the LHD has with other organizations.

Table 3 shows adjusted associations between types of activities performed in an organizational partnership and LHD participation in population-based mental health prevention. After adjustment for average number of partnerships, governance structure, population size, and LHD membership in a combined health and human services agency, specific types of activities performed with specific types of organizations were significantly associated with LHD participation in population-based mental health prevention. The odds of participating in population-based mental health prevention were 4 times higher among LHDs that shared personnel or resources with mental health or substance use disorder providers (AOR = 3.96; 95% CI = 1.38,11.36) and 3 times higher among LHDs that had written agreements with mental health or substance use disorder providers (AOR = 2.91; 95% CI = 1.07, 7.94) than among LHDs that did not perform such activities with these types of organizations.

TABLE 3—

Adjusted Associations Between Types of Activities Performed in Cross-Sector Partnership and Local Health Departments (LHDs) Offering Population-Based Activities to Prevent Mental Health Conditions: National Association of County and City Health Officials Profile Study, United States, 2016

Partnering Organization	Sharing of Personnel/Resources, AOR (95% CI)	Written Agreement, AOR (95% CI)	Regularly Scheduled Meetings, AOR (95% CI)	Exchange of Information, AOR (95% CI)
Mental health/substance use disorder provider	3.96 (1.38,11.36)	2.91 (1.07, 7.94)	0.63 (0.28, 1.41)	0.95 (0.43, 2.10)
Hospital	2.43 (0.43, 2.10)	1.06 (0.51, 2.21)	2.25 (1.12, 4.52)	2.24 (0.86, 5.83)
Community health center	4.55 (1.40,14.74)	1.00 (0.37, 2.69)	0.72 (0.28, 1.81)	0.32 (0.15, 0.70)
University or college	0.30 (0.09, 1.04)	0.74 (0.30, 1.80)	1.41 (0.55, 3.60)	0.71 (0.29, 1.70)
Community-based nonprofit	0.81 (0.24, 2.72)	2.26 (0.85, 5.98)	2.34 (0.90, 6.08)	1.15 (0.41, 3.25)
Local planning or development agency	0.38 (0.09, 1.58)	0.65 (0.11, 3.91)	0.58 (0.24, 1.40)	0.87 (0.40, 1.88)
Housing agency	1.59 (0.30, 8.37)	0.95 (0.15, 6.03)	1.09 (0.40, 2.96)	1.21 (0.57, 2.59)
Transportation	0.80 (0.16, 3.91)	2.40 (0.70, 8.28)	1.13 (0.44, 2.89)	0.91 (0.44, 1.89)
Library	0.35 (0.07, 1.82)	0.33 (0.03, 3.31)	0.64 (0.22, 1.86)	1.31 (0.55, 3.12)
Faith-based organization	0.13 (0.02, 0.81)	0.44 (0.96, 2.01)	0.46 (0.18, 1.17)	1.28 (0.34, 4.76)

Note. AOR = adjusted odds ratio; CI = confidence interval. All estimates were calculated with nationally representative survey weights and mutually adjusted. Covariates included the log transformation of LHD population size, jurisdiction type, governance type, whether the LHD is part of a local health and human services department, and the number of partnerships the LHD has with other organizations.

The odds of participating in population-based mental health prevention were 4.5 times higher among LHDs that shared personnel or resources with community health centers (AOR = 4.55; 95% CI = 1.40, 14.74) than among LHDs that did not do so. However, the odds of participating in population-based mental health prevention decreased by 68% among LHDs that exchanged information with community health centers (AOR = 0.32; 95% CI = 0.15, 0.70). The odds of participating in population-based mental health prevention were 2.25 times higher among LHDs that had regularly scheduled meetings with hospitals (AOR = 2.25; 95% CI = 1.12, 4.52) than among LHDs that did not have such meetings. The odds of participating in population-based mental health prevention decreased by 87% among LHDs that shared personnel or resources with faith-based organizations (AOR = 0.13; 95% CI = 0.02, 0.81).

DISCUSSION

The purpose of this study was to describe LHD cross-sector partnerships and assess associations between the types of activities performed in these partnerships and LHD participation in population-based mental health prevention. The study’s results suggest that (1) there are differences in cross-sectoral partnerships between LHDs that do and do not participate in population-based mental health prevention and (2) the types of activities performed in partnership with specific organizations are significantly associated with LHD participation in population-based mental health prevention.

National initiatives support the integration of mental health into public health.^4–6 For this reason, clear expectations for how LHDs should engage in population-based mental health prevention are crucial. Public Health 3.0 and the 10 Essential Public Health Services encourage LHDs to embrace a population-based approach, which means that these departments should not necessarily be responsible for providing treatment for mental health conditions but rather should be responsible for promoting mental health. Examples of population-based mental health approaches that LHDs could consider include conducting community training on mental health or trauma-informed practice, leveraging electronic health record data to monitor population-level mental health, and including mental health as a potential health impact when evaluating proposed policies or planning decisions.³ If these population-based mental health approaches and others are to be successfully implemented by LHDs, cross-sector partnerships need to be leveraged.

In the present study, LHDs that participated in population-based mental health prevention were more likely to be part of a combined health and human services agency than LHDs that did not. This result, in line with the findings of Purtle et al., suggests that even in communities with local behavioral health departments, LHDs still have a crucial role to play in population-based mental health prevention.¹⁶

Local behavioral health agencies have traditionally focused on secondary and tertiary prevention, using their funding and resources to ensure access to services for individuals and families.¹⁶ LHDs can complement the current efforts of local behavioral health departments by focusing on primary prevention activities aimed at improving population mental health (e.g., supporting policies to increase access to green space or monitoring population-level mental health indicators through existing health surveys). Policies, programs, and accrediting boards should consider how the presence of a local health and human services agency might influence the delineation of responsibility for population-based mental health prevention work in a community.

This study revealed that LHDs that participated in population-based mental health prevention were more likely than those that did not to collaborate with both traditional health partners (e.g., mental health and substance use disorder providers, community health centers, hospitals) and nontraditional partners (e.g., local planning and development agencies, housing agencies, transportation agencies). In the fully adjusted model, LHDs were more likely to participate in population-based mental health prevention if they shared personnel or resources, had written agreements, or had regularly scheduled meetings with traditional health partners. Previous research has shown that organizations in public health collaboratives exhibit preferences for partnering with similar organizations^19,27 and that traditional health partners are most trusted and valued by other members of the collaborative.

Traditional health partnerships often have the most impact because they bring together community organizations that can mutually benefit from improvements in health outcomes.²⁴ For example, health care providers (e.g., hospitals and community health centers) may be motivated to work with LHDs on population-based mental health prevention to minimize mental health conditions in their patient populations. Such mutual benefits might be one reason that partnerships with traditional health organizations were stronger predictors of LHDs participating in population-based mental health prevention than partnerships with nontraditional organizations.

This study suggests that the type of activity performed in partnership with a specific community organization, rather than simply the presence of a community partnership, influences LHD participation in population-based mental health prevention. For example, the presence of a partnership with a community health center had no significant impact on the odds of an LHD engaging in population-based mental health prevention. However, partnerships with community health centers in which LHDs shared personnel or resources or exchanged information had a significant impact on odds of participating in population-based mental health prevention.

The odds of participation in population-based mental health prevention were increased if LHDs shared personnel or resources or had a written agreement with mental health or substance use disorder providers. The responsibility and role of LHDs in population mental health have shifted significantly over the past decade.⁷ A lack of internal knowledge or internal experience in population-based mental health approaches may compel LHDs to form partnerships focused on shared personnel or resources with organizations that have this expertise (e.g., mental health providers).^4,16 This has important implications for policies and programs and suggests that capacity building and technical assistance for population-based mental health activities may be crucial aspects of encouraging LHD engagement in population-based mental health prevention.

Faith-based organizations provide an array of health-related services and programming to their communities.^28,29 However, LHDs that shared personnel or resources with such organizations were less likely to participate in population-based mental health prevention. Previous research involving the 2005 Profile Study showed that a similar percentage of LHDs had partnerships with faith-based communities (83% vs 82.4% in this study) and that sharing personnel or resources was the least common type of activity performed in these partnerships.³⁰ Although faith-based organizations may be playing a significant role in the health of their community, it remains unclear whether their efforts are synergistic with other local public health efforts. Further research is needed to understand how faith-based organizations contribute to population-based mental health prevention in their communities.

Limitations

This study is not without limitations. For example, the data were self-reported by LHD staff members, which is relevant with respect to the partnership and collaboration questions. Unlike questionnaires designed to assess networks or collaborations (i.e., network science), data were not collected from the organizations with which LHDs partnered. As a result, I was unable to verify that LHDs’ relationships with other organizations were reciprocal.

In addition, there is potential for different interpretations of certain questions owing to a lack of explicit definitions on the part of NACCHO. For example, NACCHO did not provide a definition or example of “population-based mental health prevention,” which could lead to underreporting or overreporting of LHDs’ participation in this type of work. Also, no definitions were provided for the different types of partnerships (e.g., partnerships involving an exchange of information), and this could lead to different interpretations of what each type of partnership implies.

Finally, the sample size was limited in terms of the types of activities performed with some types of organizations (e.g., libraries, faith-based organizations). Although this study makes an important contribution to science by demonstrating the importance of the various types of activities performed by LHDs in partnership with community organizations, further research is needed to better understand the role of nontraditional health organizations in supporting population-based mental health prevention in their communities.

Public Health Implications

Collaborations between LHDs and traditional health partners might be a solution to fill gaps in LHDs’ resources and improve their capabilities to participate in population-based mental health prevention. The outcomes of cross-sector partnerships could be improved by promoting meaningful types of partnership activities (e.g., resource sharing, regular meetings). The specific type of activity performed in a cross-sector partnership, not simply the presence of such a partnership, is important to consider as public policies and funding initiatives continue to encourage collaboration as a means of addressing population mental health.