Background: Older adults with cancer are more likely to fall than those without cancer, but the factors driving this increased risk are not fully understood. Falls may lead to injury, dependence, hospitalization, and death. Given the interventions available to reduce fall risk, identifying risk factors amenable to intervention for older adults with cancer is critical to provide targeted care and improve health outcomes. Methods: To examine factors associated with patient-reported falls within a 6 month follow up period, we analyzed data previously collected in a nationwide cluster randomized trial (ClinicalTrials.gov: NCT02107443; PI: Mohile, NCORP UG1CA189961). Patients were eligible if age ≥70, stage III/IV solid tumor or lymphoma with palliative treatment intent, and ≥1 geriatric assessment impairment (GA). A GA summary with tailored recommendations was given to oncologists in practices randomized to the intervention, but not usual care. We combined intervention and usual care groups and evaluated baseline risk factors for falls over a 6 month follow up including: prior falls, fear of falling (FOF), activity limitation due to FOF, activities of daily living, Short Physical Performance Battery, Timed Up and Go (TUG), Older Americans Resources Survey (physical health scale), cognition, polypharmacy, potentially inappropriate medications (PIM), and neurotoxic treatment agents. Incidence rate ratios (IRR) were estimated using generalized linear mixed models controlling for the study arm and practice site. Fully adjusted multivariable models were built for factors associated with follow up falls (p≤0.15) in bivariate. Results: Of 541 patients (mean age: 77, SD: 5.27), 140 (26%) patients had prior falls in the past 6 months. Over 6 months of follow up 467 (86%) had falls data for ≥ 1 follow up timepoint and 344 (64%) had complete follow up. Of those contributing any follow up data 103 patients (22%) reported at least one fall. In adjusted models prior falls, impaired TUG, and number of PIM were associated with higher incidence of falls over 6 months (see Table). Conclusions: Prior falls, TUG, and PIM are prospectively associated with falls among older adults with advanced cancer. These factors are feasible to assess and amenable to interventions such as rehabilitation or deprescribing. Future studies focused on implementation of fall risk reduction in the oncology setting are needed to determine the most effective ways to reduce fall risk in this vulnerable population. Clinical trial information: NCT02107443.

Adjusted* IRR for fall risk factors.

*Adjusted for age, gender, cancer type, neurotoxic treatment agents ^a≥13.5 seconds ^b≥5 prescription medications, any high-risk medication or creatinine clearance < 60 ml/min ^cBased on 2019 Beers criteria.