Median time from medical device recall initiation to FDA classification among recalls classified within 180 days, before and after April 2025: a data note with a privately pre-specified plan, openFDA enforcement data 2013 to 2026

Among US medical device recalls classified within 180 days of initiation, the median time to FDA classification was 37 days for 1,086 events initiated 2025-04-01 to 2026-05-23 and 41 days for 1,841 initiated 2023-04-01 to 2025-03-31: -4 days (95% bootstrap CI -5.5 to -3.0). Post hoc, the same design gives -5 days (-6.0 to -3.0) at April 2024, -11 to +21.5 days at April of 2016 to 2024, and 0, 0 and -5 days at the three boundaries passing the registered follow-up rule (2022 to 2024). The -4 days is thus of the size by which adjacent periods differ with no April 2025 event, not evidence of an effect of April 2025. This median can fall when classification slows: recalls taking over 180 days leave the estimand and are not yet visible for the post period. Post hoc, 229 unseen post events over 180 days (a post share of 18%, against 3.8% pre) would close the gap. Data: openFDA device enforcement export of 2026-11-19 (CC0), events posted by 2026-11-11, one 180-day cap counted from the export date. GAO-26-107619 used FDA's internal data to fiscal year 2024; no later public measure was found. The privately pre-specified plan was frozen before the download in a private repository, not a public registry, so a reader cannot verify its times; the AI sessions writing it held an earlier export, audited for counts, not lag. By the registered rule (95% interval inside plus or minus 14 days) the median did not change by 14 days or more. Post hoc, that margin passes at 8 of 9 placebo boundaries; the smallest margin passed is 5.5 days. Post hoc, quarterly medians had fallen to 36 and 34 days in 2024Q4 and 2025Q1. Post hoc, resampling months, weeks or firms gives all-classes intervals between -6.0 to -1.0 and -6.0 to -2.0. Secondary class contrasts: class II -4.5 days (-5.0 to -2.0), class I -3 days (-5.5 to -1.0). Post hoc, with firms, months or weeks resampled, the class I interval reaches zero (upper bounds 0.0 to +1.0). Termination times were registered but not computed. No causal claim is made; recalls not yet classified, a backlog openFDA lacks, are outside the data. The record holds the note (PDF and Markdown) and a supplement archive: the frozen plan and the gate files (internal references replaced), the deviations log, the analysis result summary, the download listing and record, the pre-outcome stop-rule counts, the outcomes file, the quarterly series and plot, the analysis code, the post hoc checks (post_hoc_checks.py, post_hoc_checks.json), the checks added after the referee review (post_review_checks.py, post_review_checks.json), the checks added after the second review round and the blind review (post_review_checks_r2.py, post_review_checks_r2.json), a manifest with the sha256 of every file (it also says that the data note supersedes the analysis result summary where they differ), and a trace of every number in the note. The source data are not redistributed; they are cited by export date, URL and sha256. Data: openFDA device enforcement data by the US Food and Drug Administration, CC0 1.0, export dated 2026-11-19. Plan, code, analysis and text by Claude (an AI model by Anthropic) for the author; no human has checked the figures; the human author is responsible for the deposit.

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-19
DOI
https://doi.org/10.5281/zenodo.24373431
Primary Topic
Healthcare Regulation
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Median time from medical device recall initiation to FDA classification among recalls classified within 180 days, before and after April 2025: a data note with a privately pre-specified plan, openFDA enforcement data 2013 to 2026

Keisuke Watanabe
Zenodo (CERN European Organization for Nuclear Research)
Healthcare Regulation
article

Median time from medical device recall initiation to FDA classification among recalls classified within 180 days, before and after April 2025: a data note with a privately pre-specified plan, openFDA enforcement data 2013 to 2026

Keisuke Watanabe
article en

Abstract

Among US medical device recalls classified within 180 days of initiation, the median time to FDA classification was 37 days for 1,086 events initiated 2025-04-01 to 2026-05-23 and 41 days for 1,841 initiated 2023-04-01 to 2025-03-31: -4 days (95% bootstrap CI -5.5 to -3.0). Post hoc, the same design gives -5 days (-6.0 to -3.0) at April 2024, -11 to +21.5 days at April of 2016 to 2024, and 0, 0 and -5 days at the three boundaries passing the registered follow-up rule (2022 to 2024). The -4 days is thus of the size by which adjacent periods differ with no April 2025 event, not evidence of an effect of April 2025. This median can fall when classification slows: recalls taking over 180 days leave the estimand and are not yet visible for the post period. Post hoc, 229 unseen post events over 180 days (a post share of 18%, against 3.8% pre) would close the gap. Data: openFDA device enforcement export of 2026-11-19 (CC0), events posted by 2026-11-11, one 180-day cap counted from the export date. GAO-26-107619 used FDA's internal data to fiscal year 2024; no later public measure was found. The privately pre-specified plan was frozen before the download in a private repository, not a public registry, so a reader cannot verify its times; the AI sessions writing it held an earlier export, audited for counts, not lag. By the registered rule (95% interval inside plus or minus 14 days) the median did not change by 14 days or more. Post hoc, that margin passes at 8 of 9 placebo boundaries; the smallest margin passed is 5.5 days. Post hoc, quarterly medians had fallen to 36 and 34 days in 2024Q4 and 2025Q1. Post hoc, resampling months, weeks or firms gives all-classes intervals between -6.0 to -1.0 and -6.0 to -2.0. Secondary class contrasts: class II -4.5 days (-5.0 to -2.0), class I -3 days (-5.5 to -1.0). Post hoc, with firms, months or weeks resampled, the class I interval reaches zero (upper bounds 0.0 to +1.0). Termination times were registered but not computed. No causal claim is made; recalls not yet classified, a backlog openFDA lacks, are outside the data. The record holds the note (PDF and Markdown) and a supplement archive: the frozen plan and the gate files (internal references replaced), the deviations log, the analysis result summary, the download listing and record, the pre-outcome stop-rule counts, the outcomes file, the quarterly series and plot, the analysis code, the post hoc checks (post_hoc_checks.py, post_hoc_checks.json), the checks added after the referee review (post_review_checks.py, post_review_checks.json), the checks added after the second review round and the blind review (post_review_checks_r2.py, post_review_checks_r2.json), a manifest with the sha256 of every file (it also says that the data note supersedes the analysis result summary where they differ), and a trace of every number in the note. The source data are not redistributed; they are cited by export date, URL and sha256. Data: openFDA device enforcement data by the US Food and Drug Administration, CC0 1.0, export dated 2026-11-19. Plan, code, analysis and text by Claude (an AI model by Anthropic) for the author; no human has checked the figures; the human author is responsible for the deposit.

Zenodo (CERN European Organization for Nuclear Research)
Healthcare Regulation
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon