OE Journal

ORIGINAL ANALYSIS

3 reasons why P-values are dangerous

P-values continue to dominate orthopaedic research despite offering limited insight into what truly matters for clinical decision-making. A non-significant P-value does not confirm two treatments are equivalent; it often reflects inadequate sample size rather than true similarity. Even when significant, a P-value reveals nothing about the magnitude of a treatment effect and is highly sensitive to study size, allowing small differences in large trials—or large differences in small trials—to appear equally “significant.” Most importantly, statistical significance does not guarantee clinical relevance, as differences reaching conventional thresholds may be too small to influence practice.