Development of a brief electronic patient reported outcome (ePRO) tool to anticipate treatment delays: The CareTracker study.

397 Background: Cancer treatment delays are associated with decrements in breast and colorectal cancer survival. Patient navigation and automated health record warning systems are useful but can be resource-intensive and may identify delays retrospectively. We tested a brief ePRO tool, CareTracker, to proactively identify impending treatment delays for newly diagnosed breast and colorectal cancer patients. Methods: We recruited 100 patients with Stage I-III breast or colorectal cancer diagnosed < 6 weeks prior to enrollment who had not yet started treatment. Participants completed a baseline demographic survey, followed by weekly ePRO surveys delivered via email or text for up to 8 weeks or until treatment began. Participants reported treatment and other care (e.g., imaging, biopsy) since the prior survey. If treatment had not started, they selected barriers from a list, including a free text option. The primary endpoint was feasibility, defined as at least 70% of consented participants either reporting treatment within 56 days or completing at least four of eight weekly ePRO surveys. Secondary endpoints included the proportion not initiating treatment within 8 weeks, and the median ePRO-reported time from pathologic diagnosis to treatment initiation (TTI). To assess validity, we compared ePRO-reported TTI to manual medical record abstraction. Results: Median time from diagnosis to consent was 16 days (IQR 12-22 days). Among n = 100 consented participants, 3 were withdrawn due to screen failure (no diagnosis of invasive cancer) and 14 for non-completion of the baseline survey. Regarding feasibility, 79 participants (79% of consented participants and 95% of those who completed baseline) engaged with the CareTracker tool either by reporting treatment within 8 weeks (n = 74) or by answering at least 4 of 8 weekly surveys (n = 5), exceeding the feasibility boundary of 70%. Among participants evaluable for self-reported treatment delay, 5/79 (6.3%) reported no treatment by 8 weeks. The most commonly reported barriers included awaiting a future scheduled treatment and waiting for scheduling. Among those reporting treatment, the median TTI was 44.5 days by self-report (IQR 39.25-55) versus 39 days by medical record abstraction (IQR 30.25-56.5). Conclusions: Proactive monitoring of treatment delay in newly diagnosed cancer patients via an ePRO tool was feasible, and patient reported TTI closely matched medical records. However, the frequency of severe delay was relatively low, and data suggest that delays in our setting may be more driven by scheduling and system capacity than by patient-side barriers. Next steps include comparing engaged participants to system-wide demographics, and developing workflows to integrate ePRO treatment delay monitoring into institutional protocols for rapid access clinics, symptom monitoring and navigation of newly diagnosed cancer patients.