Fig. 7

Average rating differences between pharmacist-reported logistical barriers (a ● = insufficient time, ■ = drug availability, ▲ = drug affordability, ▼ = provider access, ◆ = patient refusal) and ethical concerns (b ● = confidentiality breach, ■ = healthcare inequality, ▲ = moral hazard, ▼ = reduced autonomy, ◆ = value inefficiency). Filled shapes (●) correspond to current therapeutic responses while open shapes (○) correspond to anti-opioid vaccine responses. Colored shapes denote the five differences (●-○) in each matrix that were used to define the ‘reference equivalency’ model. Distribution of possible sums of differences in observed data and simulated random distribution for logistical barriers (c) and ethical concerns (d). Position of the sum of differences for the ‘reference equivalency’ model is denoted by the dashed lines, with subpanel zoomed in on the portion of the observed distribution where this value is found. Data plotted as average difference across all subjects (a, b) and counts of values falling within bins with a size of 0.1 (c, d)