Ph.D., Clinical Psychology, University of Wisconsin-Madison, in progress
M.S., Clinical Psychology, University of Wisconsin-Madison, 2019
B.S., Psychology, Davidson College, 2016
Follow me on Twitter, Google Scholar, and ResearchGate
Research Interests
My research pursues precision mental health goals to improve treatment for individuals with substance use disorders. Precision mental health seeks to guide treatment selection by harnessing characteristics likely to predict treatment success for a specific patient. Previous precision mental health research has been largely unable to account for real-world complexities and to generalize models to new patients and across populations not well-represented in the model development sample. Contemporary machine learning approaches are well-suited to address these limitations because they can accommodate high-dimensional data and extract reliable, generalizable prediction signals. These approaches can be used not only to select initial treatments that differ between individuals but also to provide continuing care that adapts within an individual over time.
Adaptive, continuing care necessitates frequent, in-situ, long-term monitoring to identify moments of greatest risk and pinpoint the factors increasing risk in-the-moment for a specific individual. To intervene in these moments, we need just-in-time interventions that can be deployed when personally relevant. Digital therapeutics hold promise for delivering evidence-based treatments, tools, and supports in response to momentary needs.
Benefits of personalized treatments may include not only improved effectiveness but also increased responsiveness to individual differences and social contexts that affect healthcare disparities. Digital therapeutics specifically may help to address inequities in mental healthcare by making treatment more accessible and acceptable. These tools can be made better still by involving community members and leaders in treatment development so that available supports are effective for and acceptable to the specific individuals who will use them.
My ultimate program of research will harness individual differences to guide not only initial treatment selection but also ongoing care that adapts as symptoms, situations, and risk fluctuate. I hope to incorporate prediction models into tools like digital therapeutics to identify moments of risk throughout recovery and provide accessible, timely, personally relevant support.
Current Projects
Machine learning-assisted treatment selection for smoking cessation (Dissertation, supported by NRSA Fellowship F31 DA056144)
Cued insight into lapse risk among individuals in early recovery from alcohol use disorder
Selected Publications
Wyant K*, Sant’Ana SJ*, Fronk GE, & Curtin JJ (under review). Machine learning models for temporally precise lapse prediction for alcohol use disorder. Preprint
Wyant K, Moshontz H, Ward S, Fronk GE, & Curtin JJ (in press). Acceptability of personal sensing among people with alcohol use disorder: Observational study. JMIR mHealth and uHealth. Preprint
Fronk GE*, Hefner K*, Gloria R*, & Curtin JJ (2022). Central stress response among heavy marijuana users. Psychology of Addictive Behaviors, 36(8), 1023-1035. DOI
Schultz ME*, Fronk GE*, Jaume-Felicios N, Magruder K, & Curtin JJ (2022). Stressor-elicited craving and smoking during a smoking cessation attempt. Journal of Psychopathology and Clinical Science, 131(1), 73-85. DOI
Fronk GE, Sant’Ana SJ, Kaye JT, & Curtin JJ (2020). Stress allostasis in substance use disorder: Promise, progress, and emerging priorities in clinical research. Annual Review of Clinical Psychology, 16, 401-430. DOI
Bradford DE, Fronk GE, Sant’Ana SJ, Magruder KP, Kaye JT & Curtin JJ (2018). The need for precise answers for the goals of precision medicine in alcohol dependence to succeed. Neuropsychopharmacology, 43(9): 1799-1800. DOI
*These authors contributed equally as co-first authors.