Two pathways, one regulation

Under 42 CFR 493.1253, every nonwaived laboratory test falls into one of two obligations before it can report patient results. An unmodified FDA-cleared or FDA-approved test, used exactly as the manufacturer's labeling describes, requires verification: the laboratory confirms that the manufacturer's claimed accuracy, precision, and reportable range hold up on its own instrument, operators, and patient population, and confirms the reference interval fits that population. A test that has been modified from its cleared labeling, or that was never FDA-cleared or approved at all, including a laboratory-developed test (LDT), requires validation: the laboratory establishes those performance specifications itself, because no manufacturer submission covers the version actually in use.

Classifying a method starts with one question: has anything about how this test is run changed from the cleared labeling? CMS guidance lists examples of what counts as a modification, including a different specimen type, a changed incubation time or temperature, a different dilution scheme, a substituted calibrator, an altered cutoff or endpoint calculation, or any changed procedural step. That list is illustrative, not exhaustive, so an ambiguous case still needs a documented judgment call, usually made with the laboratory director and quality team.

CLSI EP19 frames this the same way from a study-design angle: a laboratory that makes more than a minor change to a commercial method is, for evaluation purposes, acting as the manufacturer of that modified method and owes the same kind of performance work a manufacturer would.

One boundary needs to stay visible because it moves: FDA's 2024 rule that would have brought most LDTs under direct FDA device regulation was vacated by a federal court on March 31, 2025, and FDA formally removed the added language from its regulations on September 19, 2025. General FDA enforcement discretion toward LDTs has returned, and LDTs remain governed mainly through CLIA and applicable state law. That is the status as of this writing, not a fixed rule; a laboratory should re-check current FDA, CMS, and accreditor guidance rather than rely on this description indefinitely.

Before writing a single acceptance criterion, answer one question in writing, method status unchanged or modified, because that answer decides which whole set of studies the laboratory owes.