LAUSR

Carlsson et al have brought forward the concern that any “decision aid that does not improve uptake of active surveillance ... is severely limited.” First, one should understand the context of the original trial design, in which men with localized prostate cancer—not just low-risk tumors—were randomized to Personal Patient Profile–Prostate (P3P) versus usual care. Two-thirds of the men had intermediateor high-risk prostate cancer. The trial demonstrated that men randomized to P3P had reduced decisional conflict in comparison with those receiving usual care. Uptake of active surveillance is not applicable among men with higher risk prostate cancer. Second, Carlsson et al view decisional conflict as related to “feeling sure” about a treatment decision. This is oversimplified. The Decisional Conflict Scale (DCS) measures 5 dimensions of decision-making: feeling uncertain, feeling uninformed, feeling unclear about values, feeling unsupported, and effective decision-making. Decision support interventions based on the Ottawa Decision Support Framework, which provides theoretical justification for DCS domains, improve knowledge, increase accurate risk perceptions, and increase congruence between values and treatment. The Feeling Informed subscale contributes the most information to the DCS. The authors argue that patients pursuing active surveillance with reservations would have high decisional conflict. This is not a reasonable argument because uncertainty alone does not define the construct. If reservations persist but P3P clarifies preferences and provides validated information, decisional conflict can be reduced. The potential value of P3P goes beyond making patients “feel sure” about their treatment decision. Third, a key aspect of P3P (and other decision aids) is the lack of a value assigned to any treatment option. This is at the crux of shared decision-making. Rather, the intent is to provide consistent information about the pros and cons of treatment options while informing the provider of patient preferences related to potential outcomes from each option (both good and bad). Decision aids facilitate shared decision-making and are in line with guideline-concordant management of men with prostate cancer. Our post hoc analysis explored whether P3P use was associated with increased uptake of active surveillance among men with lower risk prostate cancer. On the basis of our analysis, it was not. However, recruitment for this trial occurred from 2013 to 2016, which predated the guidelines that Carlsson et al highlight as evidence that active surveillance for prostate cancer is not preference-sensitive. At the time of this randomized trial, existing guidelines that were validated in 2011 stated that “active surveillance ... brachytherapy ... radiotherapy, and radical prostatectomy are appropriate ... treatment options for the patient with low-risk localized prostate cancer.” That statement allows for preference sensitivity. Perhaps our findings would have been different if the trial data had been collected after 2017. A standalone decision aid will not align patients with the “best” treatment. Providers caring for patients with low-risk prostate cancer should try to rectify modifiable patient-related barriers to the adoption of active surveillance, such as cancer-specific anxiety. Our analysis demonstrated that physicians wielded considerable influence over the decision to pursue active surveillance, with men who did not have active surveillance offered to them naturally being less likely to pursue it. Interventions geared toward optimizing use of active surveillance should not just be directed at patients. Instead, we should also dispel myths held by providers (eg, Black men and anxious men should not pursue active surveillance) that inhibit broad adoption of active surveillance for patients with low-risk prostate cancer.