BACKGROUND

Women of lower socioeconomic status (SES) with early-stage breast cancer are more likely to report lower doctor-patient communication, lower involvement in decision-making, and higher decision regret than are women of higher SES.

OBJECTIVE

To compare patient-reported outcomes for 2 paper-based conversation aids among women with early-stage breast cancer, particularly those of lower SES.

METHODS

Semistructured interviews with early-stage breast cancer patients of lower SES to ensure the acceptability of the Decision Quality Instrument (DQI), our primary outcome measure. Three-arm cluster-randomized controlled superiority trial of 2 conversation aids (Picture Option Grid and Option Grid) vs usual care among women with early-stage breast cancer with clinician-level randomization. Decision quality (primary outcome) was measured as 3 subscales: breast cancer knowledge, decision process, and treatment concordance. Semistructured interviews with a purposive sample of women in intervention arms, surgeons, and various stakeholders to learn about aid implementation.

RESULTS

From 39 interviews, we made the following changes: (1) added a glossary, (2) added 2 questions and an open text question in the concordance subscale, (3) reworded the treatment intention question, and (4) revised the knowledge instructions. Sixteen surgeons saw 571/622 consented patients (51 became ineligible after baseline). The average patient age was 59 years, 67.1% were of higher SES, and 44.8% had inadequate health literacy. Compared with patients receiving usual care (n = 257), participants in the Picture Option Grid arm (n = 248) had higher knowledge (0.27; 95% CI, 0.01-0.53), higher decision process (1.18; 95% CI, 0.23-2.13), lower decision regret (−22.32; 95% CI, −44.65 to −0.528), and more observed (24.71; 95% CI, 5.93-43.49) and patient-reported (0.17; 95% CI, 0.03-0.31) shared decision-making (SDM). Compared with patients receiving usual care, participants in the Option Grid arm (n = 66) had higher decision process (0.82; 95% CI, 0.01-1.62), better care coordination (0.65; 95% CI, 0.03-1.27), and more observed SDM (28.9; 95% CI, 8.0-49.9). Subgroup analyses suggested that, compared with usual care, the Picture Option Grid had more impact among women of lower SES and lower health literacy. There were no differences between the 2 aids. There was high between-surgeon variation across outcomes. All 73 interviewees agreed that the aids should be used routinely. Patients believed the aids were a part of usual care and were interested in receiving the aids electronically before their consultation. Surgeons indicated they would choose which aid to use depending on patient literacy.

CONCLUSIONS

Paper-based conversation aids improved knowledge, decision process, observed and patient-reported SDM, and coordination of care. Compared with usual care, the Picture Option Grid had a larger impact among patients of lower SES and health literacy than did the Option Grid. Between-surgeon variation suggests the need for standardized physician-level training with the conversation aids in future implementation studies.

LIMITATIONS

Some participants reported attributes of higher SES. Surgeon-level randomization led to arm imbalance. Recruitment was 50% lower than planned and disproportionately low for women of low SES. Loss to follow-up was higher than expected. The interviews on conversation aid implementation were not conducted in the context of an implementation study.