Abstract
The US opioid epidemic is national in scope, but many local solutions have been shown to have efficacy. Many nonprofit hospitals have the resources and infrastructure to lead these community-based efforts, but there is evidence that some organizations are not adopting opioid services as part of their community benefit requirements to assess and address critical community health needs. This paper assesses why hospitals do not address opioid abuse after completing a community health needs assessment. For a 20% random sample of nonprofit hospitals, a unique data set was constructed of hospital efforts to address opioid abuse using the most recent publicly available community health needs assessments and implementation strategies adopted by hospitals (calendar years 2015, 2016, 2017, or 2018). Multinomial logistic regression was used to assess the relationship between 5 different reasons hospitals cited for not addressing opioid abuse and both hospital and community characteristics. Results indicate that opioid abuse was not addressed by 32% (143) of hospitals in their formal implementation strategies. State community benefit laws, county overdose level, county poverty rate, hospital region, and hospital system membership all were significantly related to the reasons hospitals cited for not addressing opioid abuse as part of their community health engagement. Hospitals in communities with significant substance abuse needs and few institutional resources may need support to address opioid misuse and adopt treatment and harm reduction initiatives. Policies that support hospital–public health partnerships may be especially important to assist hospitals to address nonmedical or behavioral health needs in their communities.
Introduction
More than 700,000
The vast majority of hospitals in the United States are exempt from federal, state, and local taxes because of their nonprofit status. 4 Although private nonprofit hospitals have been required for some time to provide community benefits in exchange for their tax exempt status, these responsibilities expanded further in 2012 under the Patient Protection and Affordable Care Act (ACA). Specifically, hospitals have new reporting requirements and must conduct a community health needs assessment (CHNA) and develop an implementation plan to address identified needs. 5 This change reflects a broader shift in health care to a greater focus on social determinants of health 6 and has the potential to better position hospitals as key partners in responding to critical local health needs, such as opioid misuse.
Despite the national scope of the opioid crisis, local solutions have been shown to have efficacy in addressing substance use disorders. Nonprofit hospitals are well positioned to lead community-based efforts because of CHNA-related requirements that they engage in addressing community health needs and because opioid misuse is associated with a number of secondary health conditions that require acute medical care. 7 Emergency departments, in particular, are a likely point of access for opioid treatment, and there is tremendous potential for staging interventions in these settings to improve outcomes related to opioid misuse. The National Institute on Drug Abuse has issued guidelines for effective care in emergency departments, yet the recommended interventions are still not standard treatment in US hospitals. 8 Many nonprofit hospitals in the United States face barriers to engaging critical behavioral health needs in their communities, such as lack of expertise, financial resources, or dedicated personnel. 9 This study investigates why many nonprofit hospitals do not adopt interventions to address opioid misuse despite the fact that this issue is consistently identified as a top health need in communities surrounding nonprofit hospitals.
Methods
Data and sample
The data set for this study was constructed from several sources to explore how nonprofit hospitals are addressing substance use disorders in their communities. The research team identified a stratified random sample of US nonprofit hospitals from the American Hospital Association (AHA) 2015 Annual Survey. Drawn from the entire universe of US nonprofit general hospitals in 2015 (N = 2715), the team selected a 20% random sample of hospitals within each state, rounding up to the nearest whole number of hospitals, to construct the data set. After accounting for missing data, the total number of hospitals in the sample was 439. Only general nonprofit hospitals were included in this sample. Specialty hospitals and those serving primarily pediatric populations were excluded. The final sample was compared to the total population of US hospitals based on a number of characteristics including bed size, teaching status, system membership, rural/urban location, and critical access status; 2 were found to be highly similar. With 1 exception, the percentage of hospitals in each category were not significantly different between the sample and the larger AHA population (see Supplementary Table S1). In the largest category of bed size (≥400 beds), there was a higher percentage of hospitals in the sample than was in the larger AHA population.
After comparing the sample to the larger population, the research team assessed each hospital's most recent CHNA and corresponding implementation plan. As part of the ACA, hospitals must complete new reporting requirements, including identifying top health needs in their communities with the help of primary and/or secondary data and outlining a strategy for addressing ≥1 needs in the subsequent reporting cycle. Between hospitals there is some variation in the structure of community benefit reports, although hospitals typically follow a specific template when preparing their reports. In the CHNAs, hospitals usually list the top community health needs (such as top 5) and in implementation strategies list these same needs, along with specific programs that will be implemented. If hospitals identify a need, they are not legally required to address it but are supposed to provide an explanation for why they are unable or unwilling to include a need in their implementation strategy.
Because hospitals must complete these reports every 3 years but were given the option to start their first round in 2012 or 2013, the reports ranged from 2015–2018. In the assessment of reports, the research team coded whether a hospital chose to address substance misuse in its implementation plan and, if it did not, the reason given for not doing so. Coding was completed by the second author and a research assistant. The 2 coders engaged in cross-coding for the first 20 community benefit reports and then selected random reports to compare coding and ensure reliability. The coders disagreed on only 3 hospitals. In these instances, they met to resolve any discrepancies in coding.
The research team combined information on whether hospitals addressed substance misuse with data from Community Benefit Insight 10 (a website tool developed by RTI International and the Milken Institute School of Public Health at The George Washington University that provides access to data on nonprofit hospital community benefit spending) and county overdose data from the Centers for Disease Control and Prevention (CDC). 11
Analytic strategy
The research team assessed the different reasons provided by hospitals for not adopting strategies to address substance use disorders compared to hospitals that did adopt such strategies. Hospitals that did not adopt substance misuse strategies fell into one of 5 groups: those that did not identify the need in their CHNA; those that lack resources or capacity; those that stated that the need does not align with their mission or expertise; those that stated the reason is being addressed by another organization. Despite the Internal Revenue Service (IRS) requirement to provide a rationale if a need will not be met by the hospital organization, not all hospitals provided explanations for not addressing an identified health need in their implementation plan. Therefore, a fifth category was included: “did not provide a specific reason.”
Multinomial logistic regression was used to assess the relationship between the reason given for not addressing substance misuse needs and a range of key hospital, community, and state characteristics, with relative risk ratios reported and interpreted as odds ratios. Hospital characteristics included system membership (0 = no, 1 = yes), religious affiliation (0 = no, 1 = yes), academic medical center (0 = no, 1 = yes), number of licensed beds within the hospital (categorized as <50, 50–199, 200–399, with >400 used as the reference group), and operating margin per bed (operating expenses divided by the number of licensed beds, estimated from Community Benefit Insight data).
Community characteristics included measures for poverty, the uninsured rate, and the unemployment rate, each of them based on county rates being placed into quartiles by Community Benefit Insight. Also included were the county drug overdose rate (obtained from the CDC), region of the country where the hospital is located (Northeast, Midwest and West, with South used as a reference group), and whether a hospital is located in a rural county (0 = no, 1 = yes). State-level factors included whether the hospital was located in state that had expanded Medicaid (0 = no, 1 = yes) and whether the hospital was located in a state that had enacted a community benefit law (0 = no, 1 = yes).
Results
Table 1 reports the descriptive statistics for the 439 nonprofit hospitals included in the sample. About 68% of the nonprofit hospitals in the sample (296 hospitals) adopted ≥1 strategies to address substance use disorders. Of the 143 hospitals that did not report adopting any strategies to address substance use disorders, 42 (or 10% of the sample) did not identify substance use disorders as a local health need in their CHNA; 6% felt the need was better addressed by another organization in their community; 5% of the sample cited a lack of resources or expertise as their reason; 3% said the need was not aligned with their mission; and the other 8% did not specify a reason for not implementing a strategy related to substance misuse.
Characteristics of Analytic Sample
SE, standard error.
Descriptive statistics also indicate that the average county overdose rate of 15.25 overdoses per 100,000 residents was similar to the average rates for the group of hospitals that adopted substance misuse strategies, those that lacked resources, those that felt their mission did not align with the need, and those that did not specify a reason (Table 1). The rate was slightly higher for hospitals that cited another organization as the reason for not adopting strategies to address substance misuse and notably lower for those that did not identify the need (Table 1).
Table 2 reports results from the multinomial logistic regression analysis showing the association between the different reasons hospitals cited for not addressing substance use disorders in their implementation plans and key institutional, community, and state characteristics. Hospitals located in counties with a higher overdose rate had lower odds of reporting that they did not address substance misuse because it was not identified in their CHNA as compared to hospitals located in counties with a lower overdose rate (OR: 0.88; 95% CI: 0.82, 0.95). Hospitals located in a state that had adopted a community benefit law had lower odds of reporting that they did not address substance misuse because it was not identified in their CHNA as compared to hospitals located in a state that had not adopted a community benefit law (Table 2). Hospitals located in the Northeast and West were less likely to report that they did address substance misuse because it was not identified in their CHNA as compared to hospitals located in the South (Table 2).
Hospital and Community Characteristics Associated with Whether a US Hospital Did Not Prioritize Opioid Use Services, 2015 (N = 439)
P < .05; ** P < .01.
CI, confidence interval; OR, odds ratio; SE, standard error.
Hospitals that are part of a health system had lower odds of citing lack of resources and capacity as the reason for not addressing substance misuse as compared to hospitals that are not part of a health system (OR: 0.36; 95% CI: 0.13, 0.94). Hospitals located in counties with a higher poverty level had higher odds of reporting that they did not address substance misuse because of a lack of resources and capacity compared to hospitals located in counties with a lower poverty level. Hospitals located in the Northeast had lower odds of citing resources and capacity as the reason that they did not address substance misuse as compared to hospitals located in the South (Table 2).
Hospitals located in a state that had adopted a community benefit law had lower odds of reporting that they did not address substance misuse because needs were being met by other organizations in their community as compared to hospitals located in a state that had not adopted a community benefit law (Table 2). There were no statistically significant associations between any of the hospital, community, and state factors considered and decisions to not address substance misuse because of alignment with mission or expertise, or when no reason was provided by hospitals.
Discussion
Study findings show that many hospitals are taking steps to address substance misuse within their community health implementation plans. These findings are important given concerns that community benefit regulations are not motivating hospitals sufficiently to address critical community health needs, especially those that are considered to be less clinical or nonmedical in nature such as behavioral health needs and social determinants of health. 12,13 However, nearly a third of hospitals in the sample did not address substance misuse in their community benefit efforts, despite the fact that their communities ranked it as a top health concern and available data that, nationally, we are currently in the midst of the most deadly drug epidemic in US history. 14 The present analysis offers insight into why hospitals do not take this issue on as a strategic priority and what types of interventions might shape future hospital decision making.
In many ways, study findings support the reasons given by the hospitals themselves, such as the lower overdose rate in communities where hospitals did not identify substance misuse as a need in their CHNA. In these cases, community stakeholders may be aware of the impact of substance misuse elsewhere, but recognize more urgent needs locally. This study also finds that hospitals in certain regions are more likely to identify substance misuse as a need in their CHNA. For some regions, such as the Northeast, this make sense as rates of drug overdose are higher. 1 The West, however, was more likely to identify a need as compared to the South, where a number of states have seen substantial increases in overdose deaths in recent years. 15 These findings might reflect different regional public health priorities as well as persistent challenges, such as structural or institutional stigma, 16 that limit institutional willingness to acknowledge and address substance misuse in their communities.
Hospitals that are part of a health system faced fewer issues related to adopting new programs related to substance misuse because of a lack of resources and capacity compared to hospitals that are not part of a health system. System resources may facilitate institutional ability to address critical community health needs, especially those that are outside of their existing expertise. Hospitals also were less likely to address substance misuse as rates of local poverty increased, suggesting that the local environment influences decision making, particularly as hospitals weigh addressing a number of public health needs that are associated with poverty. Residents of poorer communities may be the most likely to depend on outreach by local hospitals for such services and may have few other options if the local hospital is unable to provide them. In short, hospitals may need increasing support to take on critical health needs in vulnerable communities where hospital involvement would seemingly matter the most.
A finding worth exploring further is the relationship between state community benefit policies and hospitals believing that substance misuse would be better addressed by another type of organization. Past research suggests that hospitals in states with community benefit policies in place report more activities and greater spending on community benefit. 17,18 These findings raise the question of whether community benefit policies at the state level encourage adoption of certain types of activities over others, which has not yet been investigated systematically. It is possible, for example, that state community benefit laws encourage hospitals to engage in activities in which they have traditional expertise, such as health promotion and chronic disease prevention, while shifting the responsibility for behavioral health needs and social determinants of health to other community organizations. Though all nonprofit hospitals are required by federal law to contribute to their communities in some way, additional policies at the state level may encourage hospitals to act, but not necessarily address all critical public health needs.
Limitations
There are several limitations to this study related to the analysis of existing data on hospital decision making. The research team is relying primarily on self-reported reasons for why hospitals do not address opioid abuse in their communities. These reports are completed by different personnel at each organization and it is possible that they do not fully capture the barriers faced by an organization when developing new programs to address opioid abuse. There also were a number of hospitals that did not specify a reason for not addressing opioid abuse, despite the IRS requirement to provide a written explanation. It is not possible to know why these hospitals did not address opioid abuse, which limits understanding of hospital decision making. Because CHNAs must be completed by hospitals every 3 years, and were first required in either 2012 or 2013, the team had to include hospital data from 4 different years when collecting the most recent community benefit documents. Hospitals were either on their second or third round of post-ACA CHNAs and implementation plans that may have introduced unanticipated variation between the hospitals were included. The fact that hospital reports spanned different years also made it difficult to assign individual hospitals with the county-level overdose rate for the specific year in which hospitals were deciding whether to address opioid abuse. The county overdose rate may have varied slightly across years and the team is unable to assess whether year-to-year changes in the county overdose rate may have influenced decisions to address opioid abuse.
Conclusion
The fact that nearly a third of nonprofit hospitals are not addressing substance misuse, even as a majority of these hospitals identified it as a priority need in their CHNA, suggests that hospitals may need more support to tackle the most challenging health needs facing their communities. Hospitals are uniquely positioned to address substance misuse because of their interaction with individuals with substance use disorders in the emergency room and inpatient setting. 19 Hospitals also play a broader role as anchor institutions in many communities affected by opioid misuse and serve as high-value partners in addressing critical health needs. 20 Study findings suggest specific points of intervention if local, state, and federal policies are to shift hospital behavior upstream and toward pressing health issues that may transcend the traditional mission and expertise of hospitals. In particular, additional support should be provided to small independent hospitals and institutions that are operating in vulnerable communities. In areas where hospital community engagement stands to have the most impact on community health improvement, policies should provide support for hospitals to partner with other local organizations to build capacity and address the most critical health needs identified by their communities.
Footnotes
Acknowledgments
We wish to acknowledge Alexandra Wainwright and Ciara Martin for their assistance constructing the data set. We also wish to acknowledge the Honors Tutorial College at Ohio University for supporting a summer research assistantship related to hospital programs to address opioid abuse.
Author Disclosure Statement
The authors declare that there are no conflicts of interest.
Funding Information
Funding for this study was provided by the Ohio University Honors Tutorial College.
Supplementary Material
Supplementary Table S1
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
