TREATMENT DECISION MAKING IN LOW-RISK PROSTATE CANCER: RECRUITMENT AND DATA COLLECTION FEASIBILITY

Purpose: Physician preference for high-risk interventions over equally effective but lower-risk interventions is poorly understood, partly due to challenges in depending on extensive qualitative techniques to measure decision attributes. We sought to: 1) assess urologists’ tolerance for interview length; 2) develop a concise, optimally discerning semi-structured interview guide; and, 3) demonstrate the feasibility of recruiting practicing urologists.

Methods: We inquired about willingness to participate in interviews of varying duration during networking opportunities with urologists. We pilot tested a semi-structured interview guide based on standard ethnographic techniques in a sample of four urology residents. We exploited established networks of urologists and conducted face-to-face “cold” recruitment during a national professional society meeting with a goal of recruiting 40% of contacted participants.

Results: Urologists’ stated tolerance for interview length was 20 minutes. The pilot interview averaged 54 minutes to administer. Thus, transcripts were analyzed to eliminate duplicative questions and routine practice procedures. Questions labeled "difficult" were retained. Despite volunteering potential economic and organizational influences on treatment, participants reported sensitivity around overtreatment and physician-owned ancillary services. Participants suggested strategies to identify urologists' comfort with active surveillance. We restructured the interview guide as more open-ended with less specific probing, and to discriminate between active treatment and surveillance, rather than individual treatment modalities, resulting in a 40% reduction in questions and streamlining of item probes. We identified 5 essential questions which could be prioritized in the brief encounter, allowing additional time to explore participant-identified priorities.

We used the professional networks of four urologists to recruit 12 participants. We recruited three participants via direct request at the conference site. Three participants recruited an additional participant each. In total, recruitment offers were extended to 26 urologists, resulting in 16 (61.5%) completed interviews, including 5 in community practice. Two participants missed appointments, failing to complete interviews.

Conclusions: We demonstrated our ability to recruit practicing urologists to participate in decision-making research, to apply methods that reveal sensitive information, and to garner maximum information with least participant burden. Future work to map identified attributes of the decision making process to a comprehensive model of treatment decision-making is planned.