Mobile phone apps for the prediction of prostate cancer: External validation of the Coral and Rotterdam apps

Abstract Objective To analyze the performance of two mobile phone apps—the Rotterdam prostate cancer risk app and the Coral app—in a cohort of patients undergoing prostate biopsies. Methods A consecutive series of men undergoing prostate biopsies were enrolled in two centers. Indications for prostate biopsy included abnormal prostate-specific antigen levels (PSA >4 ng/mL) and/or an abnormal digital rectal examination (DRE). Prostate cancer risk and high-grade prostate cancer risk were assessed using the Rotterdam prostate cancer risk app (iOS) and the Coral app (iOS). The usability of the apps was also assessed and compared using the Post-Study System Usability Questionnaire (PSSUQ) developed by IBM. Results Overall, 1682 patients with a median age of 68 (62–73) years were enrolled. The Rotterdam app outperformed the Coral app in the prediction of prostate cancer (AUC: 0.70 versus 0.631, p = 0.001) and of high-grade prostate cancer (0.75 versus 0.69, p = 0.001) (Fig. 1). PSSUQ data revealed that both Rotterdam and Coral applications were comparable in terms of usefulness (87% versus 83%, p = 0.708), information quality (74% versus 72%, p = 0.349), interface quality (79% versus 74%, p = 0.216) and satisfaction (76% versus 76%, p = 0.935), respectively. In terms of preferences, 26/50 (54%) preferred the Rotterdam app, while 24/50 (46%) preferred the Coral app. Conclusion In our experience the Rotterdam App outperformed the Coral App for the prediction of prostate cancer or high-grade cancer diagnosis. In particular we confirmed, using the Rotterdam app, that only one out of ten patients with a low Rotterdam score will harbor high-grade prostate cancer on biopsy.