LAUSR

e18531 Background: Prostate cancer patients generally have good survival outcomes, but some survivors have poor functional outcomes with persisting symptoms and poor quality of life (QoL). Socio-economic status (SES), access to specialist care and place of residence (including rural, remote areas) are associated with poor survival. These variables may also impact survivorship outcomes, including QoL. This study aimed to identify and visualise geographical disparities in QoL outcomes for prostate cancer survivors. Methods: We extracted complete data for 7690 patients, including patient-reported EPIC-26 QoL questionnaire results (12 months post-treatment), from the Prostate Cancer Outcomes Registry–Victoria (PCOR-VIC), a population-based clinical quality outcomes registry. Patient-matched geographic coordinates were obtained through the Victorian Cancer Registry. Scores from each of the 5 EPIC-26 functional domains were visualised using geospatial maps and correlated to the Index of Relative Socio-economic Advantage and Disadvantage (IRSAD) and geographical remoteness. A composite score condensing all 5 domains into a single number (range 5 – 20) was created for succinct visualisation of functional outcomes. We analysed hotspots using the Getis-Ord Gi* statistic. Results: Scores in all 5 functional domains were positively correlated to IRSAD, which remained an independent predictor of low functional score when controlling for age, disease stage, treatment modality and geographical remoteness (p < 0.001 for all). Geographical remoteness predicted low functional score in all domains (p < 0.05) except Urinary Irritative but was not statistically significant when adjusted for IRSAD. The constructed composite score was significantly correlated to IRSAD and remoteness in a univariate model (p < 0.001). In a multivariate model, remoteness was not an independent predictor of outcome, but the composite score rose by 0.13 for each IRSAD decile (p < 0.001), with patients in the bottom decile of IRSAD more likely to have composite scores below median ( < 14) than those in the top decile (OR 1.8 CI 1.4-2.2 p < 0.001). The composite score allowed visualisation of functional outcomes in a single figure and identification of geographic hotspots with poor outcome. Conclusions: Systematic collection of long-term patient-reported outcomes is feasible. Geographical disparities in QoL outcome following prostate cancer treatment correlated with SES and remoteness, and we note the relative importance of SES over remoteness. A single composite score for EPIC-26 showed face-validity. Our results will allow targeted efforts to further understand the drivers of these findings and improve equity in survivorship outcomes for prostate cancer survivors. The methods developed in this study can be extended and re-deployed to evaluate outcomes in other jurisdictions and cancer types.