A video decision aid to improve acute myeloid leukemia patients’ illness understanding: Results of a pilot trial.

Abstract

7040 Background: Many acute myeloid leukemia (AML) patients harbor misunderstandings about their illness, overestimating both their likelihood of cure and risks of intensive therapies. Decision aids (DA) can improve illness understanding and reduce decisional conflict, but are not routinely used in AML. Methods: We developed an AML DA with input from patients, caregivers, clinicians, and laypersons, via the International Patient Decision Aids Standards (IPDAS) process. It includes 10 short animated videos with voiceovers, covering AML basics, etiology, outcomes, treatment paradigms, and risks/benefits of various treatment approaches. We enrolled 20 patients in a pilot feasibility and efficacy trial, with pre/post survey assessments of AML knowledge via an 18-item questionnaire, decisional conflict (Decisional Conflict Scale; DCS), anxiety (State Trait Anxiety Inventory, Short Form; STAI-6), and measures of DA usability and satisfaction. Results: Participants were a mean of 62.4 years old, 12 (60%) were male, 17 (85%) white, and 15 (75%) had newly-diagnosed disease. Mean time since AML diagnosis was 145 days (median 31; range 2-1092). 16 (80%) exhibited high-school-level understanding of medical terms per the REALM-SF, and participants on average exhibited moderate numeracy (mean score of 4.1 on the Subjective Numeracy Scale). All participants completed the study, exceeding our pre-determined feasibility threshold. AML knowledge scores generally improved, from a mean of 11.8 correct items on pre-test, to 15.2 on post-test assessment (p < 0.0001), with 80% of participants achieving improved scores. Struggles remained regarding patients’ understanding of the role that genetic tests play in AML care. There was no increase in anxiety after watching the videos, but decisional conflict was significantly reduced, from a mean of 28.5 at baseline to 22 in the post-test (p = .019). Participants reported high satisfaction and usability scores for the DA. Conclusions: Our AML decision aid exhibits favorable performance characteristics, with high satisfaction and usability, a marked increase in patient knowledge, and reduced decisional conflict. Further testing is warranted in a randomized trial. Clinical trial information: NCT03442452.