How does ICE define 'in custody' for death reports and how does that affect counts?
Executive summary
ICE’s public reporting treats “in custody” deaths as those covered by its 2021 Directive on Notification, Review, and Reporting Requirements for Detainee Deaths and by the agency’s detainee-death reporting process, triggering internal notification timelines and public reports [1] [2]. That official frame produces a clear, auditable stream of “in-custody” death reports, but external reviews and researchers say the definition’s application — combined with practices like releasing medically vulnerable people shortly before death and limits on external oversight — creates gaps between deaths reported by ICE and deaths associated with recent detention [3] [4] [5].
1. How ICE formally treats “in custody” in its own rules
ICE’s Directive 11003.5 (the 2021 Notification, Review, and Reporting Requirements) is the operative guidance that governs which deaths are officially treated as detainee deaths: the directive requires the Field Office Director to report any detainee death quickly and mandates follow-up reporting and public notice, with internal notification timelines such as a 12-hour alert to senior custody officials [2] [1]. ICE also posts news releases and death reports on its Detainee Death Reporting page and states it will notify Congress and stakeholders consistent with the DHS appropriations law requiring public reporting of in‑custody deaths [6] [1].
2. The legal and procedural reporting deadlines that shape counts
Congressional appropriations language set deadlines for ICE to make in‑custody death reports public (requirements cited in DHS appropriations and summarized in ICE press material), and ICE says it has complied by posting reports since 2018 and following the Directive’s timelines for initial and complete reporting [6] [3] [1]. Those statutory and directive-driven clocks create the formal dataset that researchers and watchdogs use when they tally “ICE detention” deaths [3].
3. Where the official frame can miss deaths: releases and post‑release mortality
Independent researchers and clinicians analyzing ICE death reports have documented cases in which medically complex people were released from custody shortly before dying, and those post‑release deaths did not always appear in ICE’s official in‑custody tallies, indicating reported deaths can underestimate total mortality linked to recent detention [3]. Public compilations and commentators have long alleged that releases timed close to death can reduce the count of “in‑custody” fatalities; Wikipedia’s summary of public claims reiterates concerns that releasing detainees shortly before death can allow those deaths to be omitted from ICE’s official list [4].
4. Oversight reviews, watchdogs, and competing narratives
DHS Office of Inspector General reviews and advocacy reports have both scrutinized ICE’s death investigations and systemic issues in detention medical care, finding problems in investigations, transparency, and health care delivery that complicate confidence in reported counts [7] [5]. The ACLU and allied groups have documented instances where investigatory reports omit critical facts and have urged stronger public access to inspections and mortality reviews, arguing that gaps in transparency make it harder to reconcile ICE’s reported “in‑custody” deaths with independent accounts [5] [8].
5. Political reporting and alternative tallies
Senate and congressional offices and policy institutions often publish alternative counts and analyses — for example a Senate office compiled a count of 36 ICE-reported deaths during a recent yearlong review period — illustrating that even when relying on ICE releases, different analysts produce divergent totals depending on inclusion criteria and cutoffs [9]. Think tanks and news outlets have also reported varied totals for deaths “in ICE custody” in recent years, reflecting the same reliance on ICE’s definitional boundaries and public releases [10].
6. Bottom line on how the definition affects the numbers
Because ICE’s official definition — operationalized in Directive 11003.5 and ICE’s press/reporting practice — triggers mandatory reporting only for deaths that occur while an individual is technically detained, any death that occurs after release or while in the custody of another agency is less likely to be included in ICE’s formal counts; studies and watchdogs show this produces a systematic potential undercount of mortality linked to detention experiences [2] [1] [3]. Independent oversight reports and advocacy investigations document both preventable medical failures inside facilities and instances where agency practices or reporting gaps reduce public visibility into the full mortality burden associated with recent ICE custody, creating a contested evidence base that policymakers and researchers must navigate [7] [5] [3].