OBJECTIVE - To examine methods for identifying and recruiting prostate cancer patients prior to initiating treatment to gain insight into the treatment decision process and avoid recall bias.
STUDY DESIGN AND SETTING - One recruitment strategy involved providers and nursing staff approaching patients in community and academic urology clinics. The second recruitment strategy used electronic pathology reports to identify newly diagnosed cases in real time in an integrated health system. Our recruitment goal was to have patients complete the survey about the decision-making process prior to initiating therapy.
RESULTS - The two recruitment methods yielded different response rates. Of the 226 eligible participants approached in urology clinics, 187 (83%) returned a completed baseline survey. Of the 1177 surveys mailed to potentially eligible participants at KPNC, 617 (52%) returned a completed baseline survey. The number of surveys completed prior to treatment was 125 (67%) for the clinic recruitment approach and 437 (71%) for the electronic medical record approach. Younger participants and patients with less aggressive clinical characteristics were more likely to complete the survey before initiating treatment. Other patient demographic and clinical factors were not associated with the timing of survey return.
CONCLUSIONS - Recruiting newly diagnosed patients prior to initiating treatment is feasible with both approach methods. The use of electronic medical records to identify subjects was more cost efficient, although it results in a lower response rate.