Background: Earlier and more effective management of cancer-related symptoms and treatment-related toxicities might improve the survival of patients receiving systemic anti-cancer therapy. One potential mechanism for improved outcomes could be prolonging time on therapy. We have previously reported on the successful real-world implementation of an electronic patient-reported outcomes (ePRO) tool in a community oncology practice. We now report on the impact of the tool on time on therapy for common anti-cancer treatments. Methods: We evaluated time on treatment across three community oncology practices that have implemented Expain, an EMR-integrated ePRO system which enables remote symptom monitoring during therapy. We compared drug-specific cohorts of patients that had been enrolled on the ePRO tool with a concurrent cohort of patients not enrolled on the ePRO platform. We examined time on treatment using a Kaplan Meier (KM) approach. Analysis was confined to patients with no prior administration of the parenteral therapeutic agent or a 90-day treatment-free washout period; and to patients with at least three months of potential follow-up. Analysis was only conducted for drugs with at least 80 patients in the Expain arm. Results: In each drug-specific cohort, the distribution of age, sex, and primary cancer diagnosis was similar in patients enrolled vs not enrolled on the ePRO (Mann-Whitney and Chi-square p > 0.05 except for the rituximab cohort which had more males in the ePRO cohort [67 vs 52%]). The impact of the ePRO-based symptom monitoring system on time on therapy was statistically significant for 5 agents. Conclusions: We confirm in the community setting the observation that digital symptom monitoring can increase time on therapy. We believe this is due to more effective management of treatment related toxicities, as well as more effective management of malignancy related or comorbid medical emergencies which might interrupt or lead to the discontinuation of therapy. We are evaluating the robustness of these results after using various methods to address any potential non-comparability of the cohorts. We are also examining the impact of prolonging time on treatment on survival.