Presentation Authors: Zhiyu Qian*, Sylvia Lambrechts, Lorna Kwan, Christopher Saigal, Los Angeles, CA
Introduction: Patients diagnosed with low-risk prostate cancer face the choice between undergoing active treatments (AT) and active surveillance (AS). AUA recommends shared decision-making with attention to the personal preferences of patients in this context. However, little is known about the relative influence of patient preferences compared to other factors, such as physician recommendation, on treatment decisions. We aimed to characterize the impact of patient preferences in relation to other clinical factors in the decision-making process of low-risk prostate cancer patients.
Methods: Patients with low-risk prostate cancer seen at UCLA Health were offered a decision aid that used conjoint analysis to quantify their personal preferences for relevant outcomes prior to consults. Quantified preferences were then studied via latent class analysis (LCA). An electronic chart review was conducted to record treatment decisions, physician recommendation, comorbidities, and other clinical parameters. The association of each factor in the decision-making process of patients was evaluated in univariate and multivariate analyses.
Results: 80 low-risk prostate cancer patients were included in this analysis. After consultation, 46% chose AS, 30% chose AT, and 24% were undecided. Univariate analysis showed relationship status, physician recommendation, and physician type were associated with different treatment decisions. LCA identified 2 clusters of representative preference profiles, where Cluster 1 more strongly valued longevity than Cluster 2 and preferred AT to AS. Multivariate analysis showed patients were 10.3 times more likely to prefer AT if they received an AT recommendation from an urologist. Patients with normal BMIs were 4.3 times more likely to favor active treatment. The identified patient preference profile was not a significant predictor of treatment choice in the multivariate analysis.
Conclusions: The strongest impact that recommendations from urologists carry among all other factors in patient decision-making is consistent with prior literature. Cluster analysis revealed the existence of distinct preference profiles associated with respective treatment patterns. The insignificant impact of patient preference in treatment decision raises concerns of medical paternalism. However, same finding can be explained by physicians already incorporated patient preferences in their recommendations, leading to masking of effects. Future work is needed to clarify this finding.
Source of Funding: This project is supported UCLA Health.