Researchers developed and tested a tool to help men with newly diagnosed prostate cancer view predicted treatment decisions based on similar patients from a registry. A report in European Urology described the tool and outcomes of its use.
The web-based Michigan Urological Surgery Improvement Collaborative (MUSIC) is a quality improvement consortium of urology practices that maintains a prospective registry of men with prostate cancer. The researchers used registry data from 45 MUSIC urology practices, including 5,016 men, from 2015 to 2017 to develop and validate a random forest machine learning model. A total of 2,527 men with prostate cancer tested the tool. Researchers analyzed how often the program predicted the treatment choice these men would pick for themselves, and it turned out to be highly accurate (area under the curve = 0.81). Among the entire cohort of 7,543 men, 45% underwent radical prostatectomy, 30% surveillance, 17% radiation therapy, 5.6% androgen deprivation, and 1.8% watchful waiting.
We've teamed up with @PeterMacCC as one of the trial sites for #Navigate – an online resource for men navigating the treatment options for #ProstateCancer. Find out more: https://t.co/Kz9QKawROL pic.twitter.com/KsD0Hn035h
— The Alfred (@AlfredHealth) November 20, 2018
— Dana-Farber (@DanaFarber) April 20, 2017
However, the researchers note that this tool is only meant for men with early-stage tumors and disease that has not spread to other parts of the body.
Source: European Urology