Description

Individuals living with likely incurable cancer (ILLIC) are a heterogeneous, growing subpopulation of cancer survivors who live with cancer as a chronic relapsing disease. As a result of their transitions through multiple lines of cancer therapy and prognostic uncertainty, ILLIC have unique survivorship care needs. Principal among these is the need for depression treatment. Up to half of ILLIC report depressive symptoms with negative sequalae including lower quality of life, reduced adherence to anti-cancer therapies, suicidal ideation, and desire for hastened death. Numerous trials and meta-analyses have documented that evidence-based psychosocial treatment improves depression outcomes for ILLIC. However, multilevel barriers, including transportation issues, stigma, and a scarcity of oncology mental health providers, limit access. Thus, ILLIC need feasible, accessible evidence-based depression treatment options3.

Consistent with Commission on Cancer accreditation standards, short depression screeners (e.g., PHQ-2) are routinely administered in oncology settings with results recorded in structured Electronic Health Record (EHR) fields. Despite widespread screening adoption, treatment referral rates remain low (10-50%) across cancer centers. To address this depression screening vs. treatment referral gap, screening data can be used to proactively (i.e., remotely, outside an encounter) link survivors in need of depression treatment to scalable options. While depression screening data can be readily used for proactive identification (ID), as noted by NCI, there is a critical need to develop methods to identify and enumerate ILLIC. Data necessary to determine curability likelihood (e.g., advanced stage, metastatic), are typically recorded in unstructured EHR fields, necessitating labor-intensive, manual chart review to identify ILLIC. To realize the goal of delivering scalable evidence-based depression care for ILLIC, accurate, automated approaches to identify ILLIC are needed.

Self-guided digital mental health interventions (DMHIs) can be paired with proactive ID to create a scalable depression treatment delivery model. Our team recently developed "Moodivate" as a DMHI-based approach to deliver Behavioral Activation, an evidence-based first-line depression treatment for cancer survivors. In a pilot that informs this R01, we: 1) gathered stakeholder feedback and tailored Moodivate for the unique needs of ILLIC, 2) developed infrastructure and refined the approach for proactive ID of ILLIC with depression, and 3) conducted a pilot RCT (N=15) to evaluate feasibility, acceptability, and preliminary efficacy of a proactive ID + DMHI approach. In our RCT, ILLIC with depressive symptoms were proactively identified via structured (depression) and unstructured (ILLIC) EHR data, remotely enrolled, and randomized to proactive ID + DMHI (Moodivate tailored for ILLIC) or proactive ID + usual care (UC). Our preliminary data show that this care delivery model is feasible (60% of eligible patients accrued; Moodivate mean rating of excellent on the System Usability Scale), acceptable (70% used Moodivate continuously for one month), and may improve depression (40% reported a clinically meaningful improvement). Importantly, a sustainable treatment model must also address chronic evidence-to-practice gaps. Thus, implementation outcomes and determinants of the proactive ID + DMHI approach must be concurrently evaluated across multiple care delivery levels to enhance future adoption.