Opioid dependence and opioid related deaths are a public health problem which the United States Centers of Disease Control have declared an epidemic. While opioid agonist therapy for opioid addiction has been accepted as the most effective treatment for opioid dependence among academics, and office based buprenorphine treatment has been available in the Unites States for over 10 years, OB buprenorphine faces many barriers to widespread adoption. Empirical data on the geographic distribution of physicians able to prescribe buprenorphine and the prescribing patterns of those physicians show considerable unevenness in access and utilization of treatment services. Federal-level policies have recently been implemented to expand access to opioid agonist therapy, but the medium and long term impacts of these policy changes on individual outcomes, public health, and geographic access equity are not yet clear. This dissertation compares two recent federal level policies on expanding access to buprenorphine treatment: raising the regulatory limit on the number of patients a provider can treat (implemented July, 2016), and extending prescribing privileges to nurse practitioners and physician assistants (implemented February, 2017), using an empirically supported Agent Based Simulation model. Policies are assessed by a novel, at-a-glance, quantitative access equity metric: the Spatial Potential Access Gini Index, in addition to year-end treatment utilization, opioid overdose deaths, and the amount of illicit medication diversion. In the simulation, expanding access by increasing the patient limit did not result in more equitable spatial access, while extending prescribing to NPs and PAs increased both utilization and spatial access equity. This is likely due to empirically supported model assumptions that NPs and PAs providing primary care often serve in medically underserved areas including rural and remote regions. Extending prescribing to these practitioners opens up new treatment locations changing the spatial distribution of treatment opportunities. Changing patient limits does not change the overall spatial distribution of services, so spatial access equity does not change even if overall treatment supply gets better or worse. The primary contribution of this work is the Spatial Potential Access Lorenz Curve and the Spatial Potential Access Gini Index, measures that aggregate individual-level Spatial Potential Access Scores commonly used in health care geography to map and identify areas of access disparity within a region. The equitability of Spatial Potential Access is calculated by using the Lorenz Curve, which is commonly used to characterize the distribution of wealth or income in a society, from which a Gini Index is calculated. The Spatial Potential Access Gini Index allows for direct comparison of complex quantitative information about the geographic distribution of supply and demand in a region with other regions, or in response to policies that impact supply or demand within the region. The measure has potential applications in simulation studies on the spatial allocation of services, allowing equity assessment of policy alternatives, as well as in empirical work, allowing equity comparisons of different regions, or in hybrid studies in which policy experiments are conducted on data-rich maps.

The United States (U.S.) is in the midst of an opioid overdose epidemic. In the U.S., overdose deaths related to opioid exposure are the leading cause of accidental death, yet life-saving treatments, such as methadone or buprenorphine (opioid agonist therapy [OAT]), are underused. OAT underused is due, in part, to complex regulatory and health services delivery environments. Public health officials and policymakers have focused on expanding OAT access in the community (e.g. office-based buprenorphine treatment, and opioid treatment programs); however, an often-overlooked component of the treatment pathway is the acute care delivery setting, in particular hospitals. Opioid use disorder (OUD)-related hospitalizations are increasing, and incurring significant costs; care delivered in this setting is likely sub-optimal. This study examined hospital-based services for OUD using a conceptual framework based on an interdisciplinary review of policy, organizational behavior, systems science, economics, and health services delivery scholarship. The study's primary research question was: How do supply-side attributes influence hospital OAT delivery, health outcomes, and health services utilization for persons hospitalized with OUD? Supply-side attributes refer to the contextual elements inside and outside of a hospital that may be associated with hospital OAT delivery performance, such as social structures (e.g., hospital standards of care, societal values) and resources and technologies (e.g., hospital staffing, federal treatment policies). A mixed methods study described, explored, and identified how patients with OUD are cared for in the hospital and the barriers and facilitators to delivering OAT during hospitalization. The sequential mixed methods approach (i.e., qualitative followed by quantitative analyses) included analysis of 17 key informant interviews with addiction medicine physicians from 16 non-federal U.S. hospitals, 25 hospital guidance documents from 10 non-federal U.S. hospitals, and administrative data from 12,407 OUD-related hospital admissions from the Veterans Health Administration (VHA) health system. The findings from the study's three aims and 16 research sub-questions were integrated to reach seven conclusions: 1) OAT is underused in the hospital; 2) OAT delivery varies within and across hospitals; 3) OAT is used ineffectively; 4) non-OAT modalities are inappropriately used during and after hospitalization; 5) supply-side attributes inside and outside the hospital facilitate and impede hospital OAT delivery; 6) demand-side attributes facilitate and impede hospital OAT delivery; and 7) the hospital is an important service delivery mechanism in the OUD care continuum. The study's findings could be extrapolated to improve policy and practice by implementing education and health service delivery interventions through regulatory and allocative policy mechanisms focused on physicians, medical trainees, and hospital and health system administrators. Understanding how OAT delivery may be improved within the acute care delivery system is an important element to support efforts to curb the ongoing drug poisoning crisis.

Mental health conditions, particularly substance use disorders, are a growing public health threat that affect millions of Americans. Drug use disorders (DUDs), a subset of substance use disorders, are chronic conditions characterized by clusters of behavioral, cognitive, and physiological symptoms related to the use of opioids, hallucinogens, stimulants, cannabis, anxiolytics/hypnotics/sedatives, inhalants, multiple drugs, other drugs, or unknown drugs. The prevalence of DUDs among older adults ages 65 and older is growing. In 2021, almost 2 million older adults in the United States (U.S.) had a DUD. DUDs have a detrimental effect on health, especially among older adults with DUDs, who are more likely to have comorbid chronic conditions such as chronic obstructive pulmonary disorder, diabetes, and chronic heart failure than older adults without DUDs or other age groups with DUDs. The medications and symptoms of these comorbid conditions can interact with DUD-associated drugs and cause high risk for falls, delirium, medication interactions, chronic disease exacerbations, and acute emergency department (ED) utilization. Thus, in addition to the alarming growth in the prevalence of DUDs in older adults, this population faces additional risks for poor health and acute ED utilization.Despite similar prevalence of DUDs among rural and urban older adults, older adults in rural areas have decreased access to DUD care. In rural areas, there is lower availability of DUD specialists and treatment centers. In 2019, 1,149 rural counties did not have a provider who could prescribe buprenorphine (a medication for opioid use disorder), compared to only 57 urban counties. Rural residents must drive an average of 49.1 minutes to receive DUD treatment, compared to just 7.8 minutes for urban residents. Enhanced availability of DUD treatment in primary care settings may fill the gap in access to mental health and substance use care in rural communities as older adults seek primary care providers for other routine care. Prior studies have shown that patients who seek and receive pharmacologic or psychological interventions in primary care settings may have improved DUD outcomes such as lower rates of relapse, increased retention in treatment, and abstinence. The Affordable Care Act increased funding for rural primary care practices to deliver integrated mental and physical health care, which may improve outcomes for older adults with DUDs. Yet, many rural older adults with DUDs do not receive DUD screening and treatment in primary care. Presently, only 45% of patients are screened for DUDs in the community. Primary care providers often report that they do not have the confidence to address DUDs. While some research shows that organizational support such as DUD-specific training and clinical resources are key to improving primary care provider confidence in addressing opioid use disorder, there is very little research focused on the unique needs of rural primary care providers in delivering DUD services. Research on organizational factors that influence the availability and quality of DUD services in rural primary care practices is needed to bolster primary care capacity to serve older adults with DUDs in rural areas. Rural primary care practices increasingly rely on the growing nurse practitioner (NP) workforce to deliver care. Over one-quarter of the rural primary care workforce is NPs. Most research on the rural primary care NP workforce and DUD services focuses on NP prescribing of buprenorphine, a medication for opioid use disorder. Half of the new buprenorphine providers in rural areas are NPs, and NPs treat more patients with buprenorphine than physicians in rural areas. Despite NPs’ contribution to DUD care for rural patients, studies show that many NPs practice in challenging work environments that negatively impact their ability to deliver patient care. The NP work environment is often characterized by a lack of resources, autonomy, and support, which affects the delivery of patient-centered care, high-quality care, and health services utilization among patients with chronic conditions. Studies have suggested that work environment factors such as teamwork and autonomy may influence NP prescribing of buprenorphine. Yet, no studies have focused on NP work environments in primary care practices and how they affect care for older adults with DUDs in rural areas. The overall purpose of this dissertation is to produce evidence on enhancing the primary care NP workforce’s ability to deliver mental health and DUD services. The first study in this dissertation (Chapter 2) was a systematic review of the literature focused on the effectiveness of NP care for patients with mental health conditions (i.e., anxiety, depression, and substance use disorders) in primary care settings. These conditions were included to identify the state of the literature base focused on NP mental health care delivery in primary care settings, particularly focused on gaps in evidence related to NP care for rural older adults with DUDs. The second study in this dissertation (Chapter 3) conceptualized primary care provider confidence in addressing opioid use disorder to understand how to support providers (i.e., physician, NPs, and physician assistants) in delivering DUD care through policy, practice, and research innovations. The third study of this dissertation (Chapter 4) assessed the relationship between the NP work environment and rural NP practices’ capacity to serve older adults with DUDs. Finally, the fourth study (Chapter 5) examined the impact of the NP work environment in rural primary care practices on ED utilization among older adults with DUDs. Findings from the studies included in this dissertation can be used to improve primary care delivery in rural areas and inform policy and research innovations to address the growing public health threat of DUDs among rural older adults. Materials and Methods This dissertation utilized novel materials and methods to inform the development of policy, practice, and research innovations that enhance the primary care NP workforce’s ability to deliver mental health and DUD services. The following presents the study design of each individual chapter to achieve the study’s aims. Study 1 | Chapter 2: This systematic review of the evidence addressed the effectiveness of NP care for patients with mental health conditions in primary care settings. We followed Joanna Briggs Institute (JBI) Manual for Evidence Synthesis for systematic reviews of effectiveness and reports results according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The review protocol is published with PROSPERO (ID = CRD42021269816). Study 2 | Chapter 3: This concept analysis followed the Walker and Avant framework to conceptualize primary care provider confidence in addressing opioid use disorder. Study 3 | Chapter 4: The goal of this study was to assess the impact of the NP work environment on the capacity of rural primary care practices to serve older adults with DUDs in rural areas. To achieve this aim, we conducted a secondary data analysis of cross-sectional NP survey data from 2018-2019 merged with Medicare claims from 2018. We conducted our statistical analysis using fractional logistic regression. Study 4 | Chapter 5: The goal of this study was to examine the association between the NP work environment and ED utilization among older adults with DUDs in rural areas. To achieve this aim, we conducted a secondary data analysis of cross-sectional NP survey data from 2018-2019 merged with Medicare claims from 2018. We conducted our statistical analysis using multilevel logistic regression. Conclusions This dissertation produced evidence that enhances the primary care NP workforce’s ability to deliver mental health and DUD services. In Chapter 2, the systematic review of the literature identified a lack of high-quality research focused on NP primary care for patients with mental health conditions, particularly among patients with substance use disorders in rural areas. Chapter 3 conceptualized primary care provider confidence in addressing opioid use disorder and found that teamwork, organizational culture, resources, and support may uniquely influence provider confidence, which is associated with positive patient outcomes, increased delivery of and access to opioid use disorder services, and provider attitude changes. In Chapters 4 and 5, our studies found that the NP work environment in rural primary care practices is associated with increased practice capacity to serve older adults with DUDs and reduced ED utilization in this population. These dissertation findings provide important insights that will allow policymakers, practice administrators, and researchers to invest in productive efforts to enhance the primary care NP workforce’s ability to deliver mental health and DUD services.

The dissertation is intended to guide the selection and implementation of health department-led interventions with a long-term goal of reducing opioid overdose deaths. This dissertation is comprised of three aims. First, a narrative review describes models of buprenorphine treatment, summarizes retention in buprenorphine treatment, and includes descriptions of how each study defined and measured retention in treatment to aid cross-study comparisons. Trends in buprenorphine retention, as well as heterogeneity in trends, are described, and sub-analyses examine the role clinician experience in inducting patients on buprenorphine treatment may play in promoting retention. Lastly, the effect of a specific intervention—academic detailing—aimed at reducing the prescribing of opioid analgesics is measured using methods aimed at isolating the impact of a policy intervention that occurred at the same time as the detailing campaign. Overall, this dissertation finds a lack of consistency in how retention in buprenorphine treatment is measured, which precludes easy identification of the most effective models and interventions for retention in buprenorphine treatment. Additionally, significant variation in buprenorphine treatment retention and trends in buprenorphine treatment was observed, suggesting opportunities for improvement. From 2015 to 2019, retention in buprenorphine increased among New York City residents; however, in 2019, the predicted prevalence of retention for three months was 52.7% and 34.6% for six months, below rates observed in other studies. Lastly, this dissertation should prompt public health officials to reconsider using academic detailing campaigns to decrease opioid analgesic prescribing, given that decreases in prescribing practices were only observed following one detailing campaign, which coincided with a policy change, and did not occur following another campaign, which took place two years after the policy change.