The difference between treatment efficacy demonstrated under trial conditions and effectiveness realized in routine care is often described as an efficacy–effectiveness gap. That shorthand is useful but analytically incomplete. After prescribing, several distinct processes can alter whether an efficacious treatment becomes an effective real-world intervention: treatment may never be initiated; persistence may end early; implementation of the regimen may be irregular; clinically necessary monitoring may not occur or may be uninformative; new information may fail to produce appropriate dose or treatment adaptation; access may be interrupted; care may become fragmented; and outcomes may be incompletely observed or incorrectly attributed. These processes operate at different times, leave different data signatures, and imply different remedies. They also should not be assumed to produce a monotonic loss of benefit. Appropriate treatment switching, risk-responsive monitoring, dose adjustment, restoration of access, or improved coordination can preserve benefit or reduce harm. This Current Opinion argues that economic and outcomes research should represent post-prescribing effectiveness as a sequence of mechanism-specific processes rather than as a single adherence penalty or fixed efficacy multiplier. The practical implication is that models should distinguish changes in treatment exposure from failures of care response and from distortions in outcome ascertainment. Doing so clarifies causal responsibility, improves correspondence between real-world data and decision models, and makes assumptions about preventability and value more transparent.
