Date of Graduation


Document Type


Degree Type



School of Pharmacy


Pharmaceutical Systems and Policy

Committee Chair

Erin L. Winstanley

Committee Member

Kimberley M. Kelly

Committee Member

Traci LeMasters

Committee Member

Virginia G. Scott

Committee Member

Ian R. H. Rockett

Committee Member

Gordon S. Smith


Previous research has estimated that 20–30% of individuals with opioid use disorder (OUD) have a history of both suicide attempt and unintentional opioid overdose. In 2018, more than 1.9 million adults 18 years and older had an OUD; while 1.4 million adults attempted suicide that same year (SAMHSA, 2019). The key to understanding this relationship is examining behaviors that precede these potentially fatal behaviors. One such precursory behavior is self-injury. Self-injurious behaviors (SIB) are one of the most important risk factors for future suicide (Cavanagh et al.,2003). This study has three primary objectives. The first is to conduct a systematic literature review to better understand the relationship between SIB, overdose and/or suicidal behaviors (i.e. suicidal ideation, suicide planning, or attempt) among people with OUD. Results from the systematic review were used to inform the second and third aims of this study. The second aim was to evaluate the feasibility of a standardized screen for suicide and overdose among patients receiving addiction treatment. The combined results of the systematic review and mPSS helped shape the final aim of this study. A cross-sectional study conducted in a subset of subjects who screened positive on the mPSS were used to assess clinical characteristics. Participants were categorized into three mutually exclusive groups: 1) unintentional overdose(s) (OD), 2) suicidal ideation or suicide attempt(s) (SI/SA), and 3) suicidal ideation or suicide attempt and unintentional overdose (SI/SA/OD). Multinomial logistic regression models were used to determine whether adverse childhood experiences (ACE), self-injurious behaviors, substance use history, overdose history, and past year stressful life events were differentially associated with history of OD/SI/SA. All subjects categorized as having an unintentional OD only, reported having ever used cocaine, while only 72% of SI/SA subjects had ever used cocaine (p=0.02). Subjects in the unintentional OD group were more likely to have used heroin in the 24-hours preceding their most recent overdose (92.3%) compared to SI/SA (30.7%) and SI/SA/OD (75.0 %). In the multivariable model subjects with history of SI/SA had higher ACE scores and higher mean episodes of SIB than either of the other two participant groups. Findings from this study support simultaneously screening for suicide and overdose in OUD inpatient treatment settings. Among patients who screen positive for all three outcomes, SI/SA/OD, there appears to be greater clinical severity. More thorough evaluation of drugs involved in overdose and history of self-injury may help distinguish future risk and better inform treatment planning. Collectively this information has implications for resource needs, treatment, and prevention. However, additional research is needed to determine whether screening improves provision of services and reduces self-injurious behavior.

Embargo Reason

Publication Pending