Acute myeloid leukemia (AML) patients' understanding of prognosis and treatment goals: A mixed-methods study.
221 Background: Patients with AML face complex information about the risks and rewards of pursuing treatments of varying intensity. Little is known about what patients understand or value in the decision-making process. Methods: AML inpatients receiving induction chemotherapy completed weekly electronic surveys about their prognosis and treatment goals, along with a baseline semi-structured, qualitative interview exploring their understanding of illness. Their oncologists (MDs) completed baseline surveys about prognosis and treatment goals. We followed a mixed-methods approach to analysis, applying standard statistical methods to survey data, and a constant-comparative qualitative approach to the interview data to enrich our understanding of survey results. Results: We enrolled 13 dyads (a patient and MD pair). Mean patient age was 65.5, and all had high-risk disease due to either age >60, complex cytogenetics, secondary AML, or relapsed disease. At baseline, mean MD-rated chance of cure was 27% (SD 17.6), while patients’ rating was 54% (SD 34.7; p=0.02). Only 2 patients’ estimates of cure matched their MD’s rating, yet MDs rated their patients’ prognostic understanding as high (mean 7.2 on an 11-point scale). MDs reported an average of 3.1 available treatment options, but patients recalled just 1.5. Most MDs gave a specific treatment recommendation (11 of 13), and most patients received the recommended treatment (10 of 11). Agreement about treatment goal was markedly worse than expected by chance (kappa -0.41; 95% CI -0.88-0.07). Qualitative analysis suggests that patients often viewed treatment decisions as binary “life-or-death” propositions, rather than choices between options with differing goals and intensities. Patients also significantly underestimated the risks of induction chemotherapy. Conclusions: AML patients receiving induction chemotherapy have a poor understanding of their prognosis, treatment goals, and risks of induction chemotherapy, but their oncologists are not aware of this. These findings suggest the need for an intervention to improve patient understanding of their illness and treatment options.