ESTRO 38 Abstract book

S450 ESTRO 38

implementation barrier. We develop an interactive web-based PDA based on user- centered design for better integration into the care pathway. In this approach, users are involved early and often in order to improve quality, reliability, usability of the PDA, and build ownership. Material and Methods PDA development consisted of five rounds of semi- structured interviews and usability tests with, urologists (n=8), radiation oncologists (n=4), nurses (n=2), general practitioners (n=8), ex-patients (n=19), usability experts (n=4) and male test volunteers (n=11). The PDA was continuously tested and refined based on user feedback. Interview topics included informational needs, PDA requirements and possible implementation barriers. We tested usability with surveys and heuristic evaluation by experts. Results Initial versions focused on clinical details (anatomy, complications, contra-indications) but as the rounds progressed the PDA was updated to reflect non-clinical factors that influenced patients’ choices: treatment logistics, the impact of side-effects on daily life and uncertainty. Patients tended to choose aggressive surgery due to fear and lack of knowledge, underlining the need for more balanced information, particularly on radiotherapy. Usability experts suggested navigational and visual changes, e.g. a sidebar menu, neutral color palette, more white space and less textual clutter to improve readability (Fig. 1 and 2).

models, stratified for propensity score strata, were used to compare the 2 populations in term of overall survival (OS), clinical progression free survival (cPFS), biochemical PFS (bPFS) and gastro-intestinal (GI) toxicity and genito- urinary (GU) toxicity. Propensity scores reflect the probability that a patient received MH or EH based on his baseline characteristics. Patients receiving MH were matched on a one-to-one basis with subjects receiving EH. Matching was performed based on nearest-neighbor matching. The analyses were applied on the matched cohorts and on the whole cohorts, stratifying the analyses by risk strata identified with the propensity scores (Table 1). Results The analysis included 227 PCa patients treated in 2007- 2015, receiving MH-IG-IMRT and 194 patients treated with EH-IG-IMRT in 2012-2015. Median age was 75 years (range: 50-89). Median follow-up was 43 months. A statistically significant difference in distribution regarding Gleason score (GS) and risk group was found between the 2 groups (p=0.004 and p=0.002, respectively). 226 patients with good matching characteristics and balance (113 men per group) were selected after propensity score matching.A multivariate Cox analysis performed on the pre-matched cohort showed that bPFS, cPFS and OS were not statistically different between the 2 treatment groups (p=0.439, p=0.317, p=0.906, respectively), after adjusting for prognostic variables and confounding factors (Figure A,B,C). These results were confirmed in the matched population analysis.As concerning GI toxicity, compared with EH, in MH acute G>1 events were significantly more (3% vs 9%, p=0.007). As concerning GU toxicity, compared with EH, in MH, acute G>1 events were registered in 11% vs 26% (p<0.001). Multivariate Cox regression models showed no significant differences between the cohorts in term of late GI and GU toxicity.Acute GU/GI toxicities in the matched cohort were confirmed to be worse in the MH (p=0.002 and p<0.001). The GU/GI toxicity detected at last follow-up were both in line with those emerged from the pre-matched analyses (p=0.43 and p=0.08). Conclusion Utilizing the Propensity Score Comparison there was no difference in terms of oncological outcome at a median follow-up of 43 months comparing the 2 RT modalities. Moreover, the 2 RT regimens were associated with similar toxicity profiles. PO-0855 Development and Validation of a Prostate Cancer Patient Decision Aid: Towards Participative Medicine A. Ankolekar 1 , B. Vanneste 2 , E. Bloemen 3 , J. Van Roermund 4 , E. Van Limbergen 2 , K. Van de Beek 4 , V. Zambon 5 , M. Oelke 6 , A. Dekker 1 , P. Lambin 7 , R. Fijten 1 , A. Berlanga 8 1 Maastro Clinic, Knowledge Engineering, Maastricht, The Netherlands ; 2 Maastro Clinic, Radiotherapy, Maastricht, The Netherlands ; 3 Fontys University of Applied Sciences, n/a, Eindhoven, The Netherlands ; 4 Maastricht University Medical Centre+, Urology, Maastricht, The Netherlands ; 5 Zuyderland Medical Center, Urology, Sittard, The Netherlands ; 6 St. Antonius-Hospital Akademisches Lehrkrankenhaus der WWU Münster, Urology, Gronau, Germany ; 7 Maastricht University, The D-Lab- Research Institute GROW, Maastricht, The Netherlands ; 8 Maastricht University, n/a, Maastricht, The Netherlands Purpose or Objective Prostate cancer is the second most prevalent cancer in men in the Western world. Insufficient information on the different treatment options often results in suboptimal treatment choices that could lead to poor quality of life and additional costs. A Cochrane meta-analysis proved the value of patient decision aids (PDAs) in improving patient knowledge and enabling values clarification to facilitate shared decision making, yet poor design is a major

The final PDA (www.beslissamen.nl) contains seven sections: 1. Introduction : a 2-minute video priming patients for using the PDA. 2. My treatment options : textual and visual information (diagrams, videos) on each treatment option (brachytherapy, external radiation, surgery and active surveillance).