METHODS: Investigators developed two multi-item instruments to assess decision quality for patients with early stage breast cancer facing local and systemic therapy decisions. The goals of the instruments are to assess whether patients made informed decisions that reflect their values. Items included decision-specific knowledge questions, decision-specific values assessment, and treatment undergone. Each instrument has been revised based on cognitive interviewing (n=6), expert review (n=31) and field-testing with breast cancer survivors (n=62). The field-testing was a mailing study that included the Decisional Conflict Scale. After a second mailing to non-responders, a 65% response rate was achieved (40 survivors for each decision).

RESULTS: The participants' median knowledge scores were 40% and 50% for the surgery and systemic therapy decisions, respectively. Median scores for the Informed sub-scale of the Decisional Conflict Scale (DCS) were 12 (out of 15) for both decisions. There was no correlation between the decision-specific knowledge score and the Informed subscale of the DCS for surgery (Pearson r=0.09, p=n.s.) or systemic therapy (Pearson r=0.10 p=n.s.). Patients varied in their opinions about the importance of preserving a breast and avoiding side effects of treatments, but were fairly uniform in assessing the importance of minimizing recurrence and “doing everything possible to treat the cancer.” Patients who felt strongly about keeping their breast were almost four times as likely to have breast-conserving surgery than those who did not feel strongly (OR 3.75). However, patients who were uniformed (knowledge score of 40% or less) were less likely to have made decisions concordant with their values (OR 1.25). Overall, 43% made high quality decisions (both well informed and concordant with the participant's values), and 19% made low-quality decisions(poorly informed and not concordant with their values).

CONCLUSION: Many problems in health care today arise from the poor quality decision-making. Formal methods of decision analysis and decision quality assessment are often not clinically feasible. The pilot work provides evidence of content validity for the decision quality instruments. Further evaluation and refinement are necessary to assess of criterion and predictive validity and reliability. The proposed instruments provide a practical means to assess decision quality for breast cancer patients.