Background: Patients with advanced cancer are recommended to receive palliative care (PC) early, but a limited PC specialty workforce limits feasibility. Primary PC, administered by providers outside the PC specialty, has been proposed as a way to increase access to PC. However, prior studies of primary PC have not demonstrated benefit, possibly due to lack of collaboration within the care team. We examine oncologist involvement in the largest primary PC study to date, where primary PC was delivered by oncology nurses. Methods: CONNECT (NCT02712229) was a cluster-randomized controlled trial of a primary PC intervention conducted at 17 community oncology clinics in Western Pennsylvania from 2016-2020. Selected clinic nurses (N=23) were trained to deliver up to 3 monthly primary PC visits, formulating a care plan after each visit. Nurses were asked to update the patient’s oncologist after each visit and to document this communication. We use descriptive statistics to examine level of oncologist involvement in these primary PC visits, based on nurse documentation. Results: Of 336 patients randomized to receive the primary PC intervention, 266 completed at least one study visit, and 233 (87.6%) had at least one visit where the nurse reported updating the oncologist afterwards. Of 674 total study visits, nurses reported updating the oncologist in 553 (82.0%) of them, most commonly in person (77.0%), followed by email (13.4%) and phone (9.6%). For in person or phone exchanges, mean length of interaction was 8.3 (SD 5.7) minutes, with a median of 5.0 minutes. For visits where communication occurred between the nurse and oncologist, the nurse reported presenting their specific care plan only 29.5% of the time (163/553 visits). In 96.9% (158/163) of cases, the oncologist was reported as receptive to the presented care plan, with no further details given in the 5 instances where the oncologist was reported as not receptive. When the oncologist was receptive to the care plan, it was reported that no changes were made by the oncologist in 94.3% (149/158) of cases. In the handful of cases where changes were made, changes involved ordering additional tests (3), changes to non-oncologic medications or therapeutics (2), involving other providers (1), changing oncologic treatment (1), and modifying the patient’s visit schedule (1). Conclusions: In a large trial of oncology nurse-led primary PC, oncologist involvement in care plans was not robust. More extensive team collaboration may be required in future primary PC interventions to increase the efficacy of primary PC. Clinical trial information: NCT02712229.