Fig. 4. Sensitivity analyses addressing the association of BA.2.86 lineage detection with risk of emergency department presentation and hospital admission in the presence of differential misclassification of prior infection according to infecting lineage and clinical outcome.

We illustrate estimates of the adjusted hazard ratio of progression to A emergency department presentation and B hospital admission under analyses imputing “full” infection histories for cases under the assumption of differential misclassification of prior infection status. We consider multipliers of 1, 1.25, 1.5, 2, and 3 for the ratio of true to observed infections, first non-differentially among all cases ($\omega$), and for the relative ratio of true to observed infections comparing cases who evaded the indicated outcome versus those who experienced it ($\rho$) and comparing cases infected with BA.2.86-derived lineages to those infected with non-BA.2.86 lineages ($\theta$). Analyses include data from 46,067 eligible individuals throughout the study period. For all panels (A, B), points indicate maximum likelihood estimates, with surrounding lines delineating 95% confidence intervals; we generated estimates via Cox proportional hazards models. Gray bands illustrate 95% confidence intervals for estimates from the primary analysis (Table 3). Source data to replicate the figure are provided as a Source Data file (fig4_source.xlsx).