ICE Removes Reports on 33 Detention Deaths

Medical team rushing patient on a gurney through hospital corridor — Photo: Gorodenkoff / Shutterstock

When a custody death occurs, the truth lives in the paperwork; deleting the paperwork erases not only a record, but the leverage to fix what went wrong.

At a Glance

ICE briefly posted, then removed, internal mortality reviews that found widespread lapses in medical care in detention deaths since 2019.
Across 33 deaths reviewed, most concluded care fell outside safe practice, echoing years of independent findings about systemic medical deficiencies in immigration detention.
Mortality reviews are a required safety mechanism; suppressing them undermines accountability built into ICE’s own policy framework.
Causation in any single death can be complex, but patterns across cases—and rising death totals—demand transparent, facility-level remediation and public reporting.

What the deleted reviews show, and why they matter

U.S. Immigration and Customs Enforcement briefly published internal mortality reviews assessing medical care in custodial deaths spanning late 2019 through early 2026; the documents were then removed from the public site. Reporting on the corpus is consistent: the posted reviews covered 33 deaths, and a strong majority concluded that care was outside safe limits or deviated beyond them. These are ICE’s own postmortem evaluations—conducted to determine whether staff recognized red flags, followed protocols, and escalated appropriately. The finding pattern aligns with a decade of external analyses that have repeatedly identified failures in triage, chronic-disease management, mental health care, and emergency response inside detention facilities.

Mortality reviews are not mere narratives; they are the backbone of a safety system. In correctional and detention medicine, they function like aviation mishap boards: reconstruct the timeline, map decision points, flag protocol deviations, and recommend corrective actions. ICE policy codifies that purpose, requiring reviews after every detainee death and retention and sharing of the resulting reports with appropriate oversight entities. Removal from public view, after initial posting, impedes independent scrutiny and deprives families, advocates, and lawmakers of the only comprehensive contemporaneous accounting the system produces.

A long pattern, not an anomaly

The newly surfaced conclusions do not stand alone. Human Rights Watch’s 2016 and 2018 examinations of ICE-released death reviews documented substandard care and partial or delayed disclosure in a significant share of cases. Independent medical experts retained by Physicians for Human Rights and the ACLU likewise found that the majority of deaths they analyzed showed preventable lapses—delays in responding to abnormal vitals, failures to follow cardiac or mental health protocols, and inadequate chronic-disease management. Peer‑reviewed studies reach similar results: across published Detainee Death Reviews, roughly four out of five cases cited violations of ICE’s own performance-based standards for medical care, with multiple deficiencies per case.

Context matters for interpreting these patterns. Death causation in custody is multifactorial, and some detainees arrive with advanced disease or acute conditions that would challenge any system. But population-level signals—such as the concentration of deaths among relatively young adults without major comorbidities, as well as increases in total deaths and suicides in recent years—are hard to reconcile with a narrative of adequate care consistently delivered. Even outlets that emphasize causation complexity have flagged the public health red flags such patterns represent and the urgent need for timely medical and mental health interventions in custody.

How mortality reviews work—and where they break down

In ICE’s framework, a mortality review typically reconstructs the clinical trajectory: intake screening findings; sick-call requests and response times; medication continuity (for insulin, anticonvulsants, antihypertensives); housing decisions that intersect with suicide risk; and emergency recognition, from chest pain algorithms to CPR initiation. Each step is benchmarked against ICE’s Performance-Based National Detention Standards and relevant clinical guidelines. Deviations can be technical and still be fatal: a missed repeat blood glucose check in a brittle diabetic, a delayed EKG in chest pain, a failure to escalate persistent abnormal vitals, or gaps in suicide watch when housing moves occur. The repeated identification of such lapses across facilities and operators is precisely what elevates individual tragedies into a systemic signal.

Transparency is the feedback loop. The agency’s own directive envisions reports that drive remediation and are retained and shared to enable oversight. Public posting is not the entirety of accountability, but it is the only way external stakeholders can track whether recurring hazards—e.g., contract staffing shortfalls, sick call triage bottlenecks, pharmacy delays—are being fixed or merely described anew each year. Removing already-posted reviews interrupts that loop.

The facilities, the contractors, and the oversight gap

ICE detention is a networked system: some facilities are publicly run, many are operated by private prison contractors, and clinical services are a mix of ICE Health Service Corps and vendor staff. That fragmentation complicates accountability because the evidence trail—clinical charts, housing logs, camera footage, contract audits—sits in different hands. External investigations have shown that, under such conditions, incomplete disclosures and delayed release of reviews are commonplace, while independent physicians reviewing the same deaths often identify lapses that internal narratives either minimize or omit. It is not surprising that mortality review themes cluster around the predictable pressure points of a contractor model: staffing ratios, training on site-specific protocols, and continuity of medications for chronic conditions that cannot tolerate interruption.

The broader detention literature reinforces these practical observations. Clinical and public health researchers describe consistent deficits in access to timely care, coordination between custody and clinic operations, and robust mental health services; they also point to the structural barriers detainees face in reporting problems without retaliation and in navigating grievance systems effectively. None of that complexity absolves a system of the obligation to produce and preserve its own after-action reports—if anything, it makes those reports more essential.

Disturbing Pictures of Neglect in ICE Mortality Reviews https://t.co/7jpeLFNCIx
The federal government released dozens of mortality reviews of immigrants who died in detention in recent years. They show a uniform pattern of medical neglect.

— Michael Kemper (@OldGuy84) September 30, 2026

What meaningful accountability would look like

Fixing a pattern this entrenched does not require novel law; it requires execution against standards ICE already recognizes. Three steps would move the needle. First, restore and institutionalize public access: publish mortality reviews, in de-identified form as appropriate, within a defined window and keep them accessible. ICE’s policy backbone already contemplates timely preparation and sharing; the logical extension is durable public posting with a clear index and version control.

Second, operationalize learning at the facility level. Each mortality review should embed specific, time-bound corrective actions—staff training on chest pain and suicide risk protocols, pharmacy process redesigns to eliminate missed doses, staffing plans tied to detainee census—with follow-up audits to verify closure. The repeat appearance of near-identical deficiencies across reviews is the hallmark of a system that writes reports but does not fix workflows.

Third, create external checks that complement internal reviews. Independent clinical panels can be convened to re-review a stratified sample of deaths annually, focusing on high-signal categories like suicides, cardiac events, and diabetic emergencies. Their findings should be published alongside ICE’s reviews to triangulate where narratives align or diverge—an approach that HRW and medical NGOs have modeled, but that gains power when formalized.

The bottom line

Deaths in immigration detention are not rare anomalies; they are recurrent events in a large custodial system. ICE’s own mortality reviews—when visible—show that too many of those deaths traverse preventable error paths familiar to anyone who has studied correctional health: delayed recognition, protocol drift, medication gaps, and under-resourced mental health care. Removing those reviews from public view does not change their content. It only ensures the same findings will be written again next year, about someone else.