A recall doesn't close when the notice goes out

The strategy names a level before the recall starts

21 CFR 7.42 requires every recall to run under a strategy, developed by the firm or agreed with FDA, that states three things before the recall gets under way: the depth to which the recall will reach, whether public warning is needed, and the level of effectiveness checks that will be conducted. That third element isn't a wrap-up step tacked on once the recall communications have already gone out. It's set at the same time as the depth and the warning decision, against the same hazard the recall itself responds to, because the regulation treats verifying that the recall worked as part of planning the recall, not as an afterthought to it.

Five levels, and one of them is none at all

The strategy's effectiveness-check level runs on a defined scale, not a general commitment to “follow up.” Level A calls for contacting 100 percent of consignees. Level B sets a percentage above 10 and below 100, decided case by case against the specific recall. Level C fixes that figure at 10 percent, and Level D at 2 percent. Level E is on the same list, and it means no effectiveness checks at all — a real option the regulation makes available for the lowest-consequence actions, not a gap in the rule. A recall record that never states which of the five levels applies, and instead just describes “following up with customers,” has skipped the one decision Part 7 asks the strategy to make explicit.

A check verifies action taken, not delivery

What the check has to confirm is stated plainly: that consignees at the recall's own stated depth received the notification and took the appropriate action, not merely that a letter or email went out. Those are different findings. A notification can be sent, opened, and still sit un-acted-on while the recalled product stays right where it was, and a delivery receipt proves none of that. An effectiveness check that stops at confirming the notice reached its recipients has answered a question about the communication, not the question 7.42 actually asks about the product.

The firm runs the check, and FDA still watches it

The recalling firm ordinarily carries out its own effectiveness checks, with FDA assisting only where necessary. That doesn't leave the level or the results unreviewed. FDA monitors a recall's progress against the strategy the firm set, and a chosen level that isn't reaching the field — product still turning up on shelves, consignees who never respond — is grounds for the agency to press the firm toward a higher level or a different approach, not evidence that the original strategy simply ran its course. A recall record that treats its own effectiveness-check level as fixed once chosen, regardless of what the checks in progress are actually showing, has mistaken a starting plan for a finished one.

Where this meets the file

A recall-readiness record needs the effectiveness-check level stated as its own field alongside the depth and the public-warning decision, with a stated basis for why that level fits the hazard — and a place to record whether the checks in progress are actually confirming action taken, not just notice delivered. A recall-readiness checklist built around that structure, alongside this blog's own recall-classification piece and the correction-and-removal record it sits beside, is previewed in the launch catalog. If your program grades its own effectiveness checks differently, the shelf takes that correction directly.