Introduction

Men with new localized prostate cancer (LPC) face a confusing set of care options. The purpose of our trial was to evaluate decisional conflict (DC) after access to decision support, the Personal Patient Profile-Prostate (P3P), or usual care (UC).

Methods

Men were randomized to P3P+UC or UC alone after responding to a baseline query of influential personal factors. The P3P, a tailored educational/coaching tool, was provided prior to the options review consult at 7 practices across the US. The low literacy DC scale (DCS) was administered at baseline and 1 month. One-month DCS (total score (TS) and 4 subscales) was compared by group using ANCOVA. Factors previously identified as influencing DC (age, education, partner status, working status, income, race, D&[prime]Amico risk levels, information resources, baseline/one-month decisional status, baseline DCS and study site) were assessed univariately and then adjusted in multivariable analysis. Backwards model selection was used and two-way interactions checked.

Results

392 (198 P3P, 194 UC) men were randomized. 309 (155 P3P, 154 UC) men returned 1 month DCS. No significant study group effect was indicated for the TS or subscales univariately. In multivariable analyses, P3P marginally reduced TS conflict (LSmean=3.19, P=0.06) and significantly reduced conflict related to being informed (LSmean =6.99, P=0.0004). There were significant group interactions with partner status for the TS and income for the informed subscale. Unpartnered P3P users had significantly lower total conflict compared to partnered men in both groups (P=0.0005) and UC unpartnered men (p=0.03). Among low income men, UC had significantly higher conflict on the informed subscale compared to P3P users (P<0.0001) and higher income men in both groups (P<0.0001). Overall, partnered and/or low income men had higher conflict in TS, as well as in 3 subscales. Men undecided at 1-mo had significantly higher conflict in TS and all subscales. Study site significantly impacted the TS, informed and value clarity subscales. Men who were working or used the Internet for information had significantly lower conflict in the value clarity subscale. Low D&[prime]Amico risk level was significantly associated with higher uncertainty.

Conclusions

The P3P is superior to UC with regard to informing men about LPC options. Our results have implications for who is at risk for greater DC. Men with lowest risk disease, and likely more options, are more uncertain. Men with partners/spouses and low income men are more likely to be conflicted. Decision support for partners is an important next step.

Funding

NIH 5R01NR009692