Near-miss reporting is widely used in medication-safety improvement because events intercepted before patient harm can reveal vulnerabilities while preserving opportunities for organizational learning. Yet the observed number of near misses is not generated solely by the underlying frequency of unsafe events. It also depends on how often pharmacy work creates opportunities for error, whether an error encounters an available point of interception, whether it is detectable at that point, and whether a detected event is subsequently reported. These processes can vary across dispensing tasks, medicines, technologies, staffing arrangements, reporting systems, and organizations. Consequently, two pharmacies or two time periods can produce similar recorded near-miss counts through materially different safety configurations, while different counts can arise without corresponding differences in latent error generation. This methodological/conceptual article develops an Opportunity-Adjusted Near-Miss Framework that separates exposure to error opportunity, event occurrence, interception opportunity, detectability, reporting propensity, and the observed near-miss record. It specifies comparison conditions, classification requirements, interpretation constraints, and a validation agenda without proposing a universal adjustment coefficient. The central implication is that near-miss data become comparison-informative only when the opportunity base and observation process are sufficiently characterized for the inference being attempted. Near-miss counts remain valuable for learning, but their meaning depends on how pharmacy work makes errors possible, visible, interceptable, and reportable.
