Mind the gap: Physicians' assessment of patients' importance weights in localized prostate cancer

Background The management of localized prostate cancer is challenging because of the many therapeutic options available, none of which is generally acknowledged as superior to the others in every respect. The selection of the most appropriate treatment should therefore reflect patients' preferences. Objective The purpose of the following study was to pilot a new approach for investigating whether urologists who had previously provided patients with therapeutic advice actually knew their patients' importance weights concerning the relevant aspects of the treatments at issue. Method Participants were patients recently diagnosed with localized prostate cancer (n = 20), urologists (n = 10), and non-medical professionals (architects, n = 10). These last served as a control group for the urologists and were matched to them for age and gender. Patients' importance weights were elicited by two standard methods (Direct Rating and Value Hierarchy). Each urologist was asked to estimate (with Direct Rating) his/her patient's importance weights. The same task was performed by a corresponding architect, who never met the patient and knew only the patient's age. Univariate and bivariate statistical analyses were performed to investigate the association between importance weights as elicited from patients and as estimated by urologists and architects, as well as to assess whether such agreement was attribute-dependent. Results Participants found both elicitation methods easy to use. The correlation between patients' actual importance weights and urologists' estimates was poor and comparable to that obtained between patients and architects. This result did not depend on the attribute considered, with the sole exception of the attribute Effectiveness in curing the cancer, which was evaluated as the most important attribute by the majority of participants. Conclusion These findings demonstrate the feasibility of the employed methodology and highlight the need to support preference-sensitive decisions in clinical practice by facilitating the elicitation of patients' importance weights, as well as their communication to physicians.