What published evidence supports Burna AI's approach to adverse event attribution?

The published literature on adverse event grading, attribution and patient self-reporting. Two reviewers identifying CTCAE toxicities in the same notes agreed at a kappa of 0.59 to 0.68 (Fairchild et al., 2020), and when the same adverse event recurred in the same patient, its attribution changed at least once in 31 to 36% of cases (Hillman et al., JCO 2010). Authors of a 2019 attribution workshop of regulators, sponsors and investigators described processes for determining causality as often sub-optimal, unreliable and inefficient (George et al., Cancer Treat Rev 2019). Between visits, structured patient self-reporting was associated with longer median survival, 22.5 vs 14.9 months (Denis et al., JAMA 2019), and fewer emergency visits, 34% vs 41% (Basch et al., J Clin Oncol 2016). Burna AI's own results come from the concordance study it is running as a Mayo Clinic Platform_Accelerate participant, which reads out in December 2026.