Date of Award
Spring 2023
Document Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
Department
Public Health
First Advisor
Kershaw, Trace
Abstract
Background: Co-occurring alcohol and opioid use are an under-addressed factor associated with increased drug overdose mortality. There are many challenges in the treatment of co-occurring alcohol and opioid use with most individuals in need of substance use treatment not receiving treatment services. Those that do have a high likelihood of returning to use, with resumption of use following treatment episodes increasing risk for overdose. Contextual factors have been recognized as triggers of cravings to return to use. Few interventions work to address the effect these factors have on individuals in early recovery, with locations serving as a particularly challenging trigger. Research suggests that raising awareness of such location-based triggers can support individuals in developing coping strategies and maintaining recovery. mHealth interventions have the potential to provide real-time support to individuals in triggering locations. Following the IDEAS framework for mHealth intervention design, this dissertation uses both quantitative and qualitative methods to build the knowledge base around the challenges facing individuals engaging in co-occurring alcohol and opioid use as well as potential intervention strategies that can support them in engaging in safer use behaviors. Methods: To start, this dissertation describes the demographics and spatial relationships of opioid overdose associated with co-occurring alcohol and opioid use. In the first part of Chapter 1, social network analysis techniques were used to visualize the detection of both alcohol and opioids in opioid overdose deaths in Connecticut from 2010 through 2020 (based on data from the Connecticut Office of the Chief Medical Examiner), with Chi-square analyses and Bonferroni correction for pairwise comparisons used to determine how combinations of alcohol and other drugs varied by demographic characteristics. The second part of Chapter 1 uses spatial analysis techniques and regression models to determine how the spatial distribution of alcohol outlets relates to alcohol detection in opioid overdose deaths in 2019 and 2020 at both the individual and ecological levels. Information about alcohol outlets was obtained from the Department of Consumer Protection while information about census tracts was obtained from the 2018 American Community Survey. Moving on to intervention development, in Chapter 2, thematic analysis was used to explore how 30 individuals in substance use treatment in Connecticut and Georgia used smartphone technology in support of their recovery as well as their attitudes towards the use of tracking technology for research and intervention purposes. In Chapter 3, another round of qualitative interviews was conducted with 31 individuals reporting co-occurring alcohol and opioid use in both Connecticut and Georgia. Thematic analysis according to Marlatt and Gordon’s cognitive behavioral model was used to understand how these individuals experienced location-based triggers and how they responded to these triggers. Themes and participant quotes related to these responses informed the generation of messaging for an mHealth intervention. Finally, in Chapter 4, the first round of testing was conducted on the functional prototype of the intervention, e-Covery, in order to determine its preliminary usability, feasibility, and acceptability through both a quantitative survey and a qualitative interview at the end of a seven-day trial period among ten participants in Connecticut reporting problem levels of both alcohol and opioid use. This was supplemented by a pair of focus groups with twelve individuals who provide services to people who use substances in order to understand how the intervention can be integrated into existing treatment, harm reduction, and recovery support interventions. Results: In the first part of Chapter 1, alcohol was detected in nearly a third (30.7%) of the opioid overdose deaths occurring from 2010 to 2020 in Connecticut, most commonly in combination with fentanyl. Among opioid overdose decedents, alcohol was most commonly detected among male (33.1%) decedents, Black decedents (36.8%), decedents over the age of 50 (36.1%), and those without confirmed or suspected injection (32.2%). Alcohol detection in opioid overdose deaths appears to have increased slightly over the period under investigation. Opioid overdose deaths among Black and Latinx were observed to increase over time. Results from this analysis revealed that the combination of fentanyl and stimulants was detected among Black and Latinx decedents more commonly than among White decedents and that these drugs and their combination was detected more frequently over time. These trends persisted even after limiting analysis to focus on the combination of fentanyl and stimulants excluding the detection of any other opioids. Returning our focus to the co-occurrence of alcohol and opioid use, in the second part of Chapter 1, distance to the nearest alcohol outlet was associated with increased odds of alcohol detection in opioid overdose deaths at the individual level, though this relationship did not remain significant after controlling for compositional factors such as age or sex. At the ecological level, the density of alcohol outlets in each census tract was positively associated with the number of opioid overdose deaths where alcohol was detected even after controlling for the social vulnerability index of the census tract. This association did not remain significant when analysis was limited to those alcohol outlets selling alcohol for off-premises consumption (e.g., package stores, supermarkets). In Chapter 2, individuals in substance use treatment described how they used their smartphones to buy and/or sell drugs and subsequently modified their use of smartphone technology as they entered treatment and recovery. While in treatment and early recovery, participants described using their smartphones to access all four forms of social support: affiliational, emotional, informational, and instrumental. When asked for their opinions about the use of smartphone tracking technology in research, participants expressed little concern about these components. Participant concerns focused on the security of their information and the potential burden of responding to study questions. Participants identified how participation in research involving tracking technology could have potential therapeutic effects on their own recovery. In Chapter 3, individuals reporting co-occurring alcohol and opioid use described their experiences with and responses to location-based triggers. Participants described location-based triggers as interrelated and associated with multiple people, locations, events, or emotions. Participants commonly identified ‘home’ as a trigger, one that is particularly difficult to cope with. Participants most commonly employed avoidance as a coping strategy. Finally, participants described an expectation that substance use was associated with more time spent outside of ‘home’ while recovery was associated with limited travel and more time spent at home. Finally, in Chapter 4, five themes emerged from thematic analysis of interviews with individuals who tested the e-Covery app: 1) the app was easy to use; 2) the app had therapeutic potential; 3) there were some minor challenges with app functionality; 4) participants wanted the app to be more engaging and interactive; and 5) participants were interested in being kept up-to-date around app development. These themes were largely corroborated by the quantitative survey data. Thematic analysis of the focus groups with service providers echoed some of the participant themes; service providers: 1) saw the therapeutic potential of the app; 2) felt that the app could be more engaging and interactive; and 3) wanted the app to have greater accessibility. Conclusions: This dissertation research details the involvement of alcohol in opioid overdose deaths in Connecticut and highlights the role of the built environment in co-occurring alcohol and opioid use, emphasizing the importance of interventions that account for the role of such environmental factors and that specifically target individuals engaged in such patterns of polysubstance use. This dissertation then details the process of developing an app-based geographic momentary intervention according to the IDEAS framework for mHealth intervention development. The first set of exploratory qualitative interviews integrated the perspectives of individuals in substance use treatment, describing how they already use smartphones to connect with myriad forms of social support and their attitudes towards the use of tracking technology. The second set of interviews contributed to the design of the intervention, highlighting key aspects of the experience of location-based triggers as well as messages and strategies to respond to such triggers. Finally, we tested the intervention with a small set of individuals reporting co-occurring alcohol and opioid use to assess its usability, feasibility, and acceptability. In sum, this research identifies the e-Covery app as an evidence-based intervention with high feasibility and acceptability and the potential to support individuals reporting co-occurring alcohol and opioid use in managing location-based triggers.
Recommended Citation
Viera, Adam, "Developing e-Covery, an App-based Intervention to Support Individuals With Co-occurring Problem Alcohol and Opioid Use" (2023). Yale Graduate School of Arts and Sciences Dissertations. 991.
https://elischolar.library.yale.edu/gsas_dissertations/991
