Oncologist use of the Adjuvant! model for risk communication: a pilot study examining patient knowledge of 10-year prognosis

Abstract

Background: Our purpose was to collect preliminary data on newly diagnosed breast cancer patient knowledge of prognosis before and after oncology visits. Many oncologists use a validated prognostic software model, Adjuvant!, to estimate 10-year recurrence and mortality outcomes for breast cancer local and adjuvant therapy. Some oncologists are printing Adjuvant! screens to use as visual aids during consultations. No study has reported how such use of Adjuvant! printouts affects patient knowledge of prognosis. We hypothesized that Adjuvant! printouts would be associated with significant changes in the proportion of patients with accurate understanding of local therapy prognosis.

Methods: We recruited a convenience sample of 20 patients seen by 2 senior oncologists using Adjuvant! printouts of recurrence and mortality screens in our academic medical center. We asked patients for their estimates of local therapy recurrence and mortality risks and counted the number of patients whose estimates were within +/- 5% of Adjuvant! before and after the oncology visit, testing whether pre/post changes were significant using McNemar's two-sided test at a significance level of 5%.

Results: Two patients (10%) accurately estimated local therapy recurrence and mortality risks before the oncology visit, while seven out of twenty (35%) were accurate afterwards (p = 0.125).

Conclusion: A majority of patients in our sample were inaccurate in estimating their local therapy recurrence and mortality risks, even after being shown printouts summarizing these risks during their oncology visits. Larger studies are needed to replicate or repudiate these preliminary findings, and test alternative methods of presenting risk estimates. Meanwhile, oncologists should be wary of relying exclusively on Adjuvant! printouts to communicate local therapy recurrence and mortality estimates to patients, as they may leave a majority of patients misinformed.

Figure 1

Study schema. Study schema showing chronological steps in this pre/post single-arm study.

Figure 2

Screenshot of the Adjuvant! software program. The two oncologists in this study used Adjuvant! to generate prognostic estimates tailored to patient information shown in the upper left of the screen. Oncologists printed screenshots showing patient recurrence (relapse) or mortality, selected at bottom right, for "no additional therapy", top right.