12010 Background: Older patients often experience more toxicity from cancer treatments due to their functional age and co-morbidities. The G8 is a geriatric screening tool designed to identify patients who are appropriate for a comprehensive geriatric assessment (CGA) which will inform and optimize oncology therapeutic decision making, reduce toxicities and improve outcomes. Using a nurse-led assessment approach, we have implemented an EMR embedded G8 tool across our cancer center since June 2020. We aim to describe the characteristics of those patients who scored abnormal (≤ 14) on the G8. Methods: The demographic and clinical characteristics were collected from the EMR for those who were ≥ 65 and have completed the G8 screening tool from June 2020 to Jan 2022. Summary statistics such as means, standard deviations and proportions were reported. Comparisons of various endpoints between dichotomized (≤ 14 vs. > 14) G8 scores were performed using pairwise two-sample t-test, pairwise proportion test, Fisher’s exact test and/or Pearson’s chi-squared test. Selected stakeholders were elicited for their perspectives on the implementation process. Results: 544 patients age ≥ 65 completed G8 screening with a mean score of 13.4 (± 2.3). The mean age was 75.2 (± 7.0), and 55.7% were females and 78.3% were non-Hispanic White. The most frequent cancer type was breast cancer (21.3%), followed by lung/respiratory cancer (18.2%), and GI cancer (15.8%). The percentage of patients appropriate for a CGA (based on an abnormal G8 score) was 59% (n = 332), of which only 10% (n = 31) were referred. A spearman correlation of 0.4 (p < 0.001) between G8 scores and age was obtained. A statistically significant association between ECOG and G8 groups was found, but not for any comorbidities. The prevalence of current/former smoking and extreme polypharmacy (> 10) is significantly higher in the G8 abnormal group. Implementation process evaluation found new patients with cancer overwhelmed with information, a lack of patient education and provider engagement, and an unstandardized referral process to CGA, were barriers for CGA referral. Conclusions: The G8 is an effective tool identifying patients appropriate for a CGA. An EMR embedded G8 steadily improved the G8 utilization uptake but not referral for CGA at our institution. To further improve awareness, utilization and referral, we are redesigning a hybrid workflow integrating patient-mediated MyChart implementation and nurse navigator engagement. EPIC enhancements are being designed to inform the oncologist of the patient’s G8 score, CGA referral status and the final CGA report.