College Mental Health Project

Depression is common and debilitating disorder which, among college students, is predictive
of significant problems including higher rates of dropout and suicidal behaviors. Despite
viable psychotherapeutic and pharmacological options, the majority of depressed college
students, like others in the general population, do not pursue treatment. Further, barriers
to care, both those involving practical issues and psychological factors, lead to high
attrition rates from treatment, resulting in modest effect sizes in effectiveness trials. In
reaction to these alarming figures and as a means of increasing accessibility and retention,
effective internet-based treatment for depression have been developed and tested. Despite
increased availability, response to internet-based treatment continues to vary substantially,
yet, controlled trials show that a meaningful proportion of patients who receive
internet-based therapy recover. Identifying individuals with a high likelihood of responding
to internet-based treatment would enable clinicians to target this inexpensive treatment only
to the patients with a high probability of responding; allowing more intensive treatments to
be reserved for patients who would not respond to internet-based therapy. The development of
a system to make this determination would represent a major advance and address an unmet
need.

ICare is an online depression treatment that has been adapted for college students (e.g.,
language used, problems discussed, embedded images). Prior work by the developers of ICare
have tested the internet-based treatment in diverse samples of depressed adults, and prior
meta-analytic research has demonstrated that psychological treatments for depression in
college students are as effective relative to trials carried out among depressed adults. The
overarching goal of the study is to: (a) test whether depressed college students utilize
ICare, (b) identify psychosocial and clinical characteristics that increase the likelihood of
ICare utilization, and (c) identify multivariate characteristics that predict treatment
response.