A landmark New England Journal of Medicine study analyzing 2.3 million emergency department visits has quantified what emergency physicians have long observed: psychiatric patients boarded in ERs for more than 12 hours face dramatically higher rates of medical harm. The findings are already shaping federal policy.
Emergency physicians have known for years that boarding psychiatric patients in emergency departments is harmful. They have said so in surveys, testified to it before legislative committees, and documented it in case reports. What they lacked was a large-scale, rigorously controlled study that could translate clinical observation into the kind of evidence that drives policy change. The New England Journal of Medicine study published May 1, 2026, provides exactly that.
The study, conducted by researchers at Harvard Medical School, the University of California San Francisco, and the Yale School of Medicine, analyzed 2.3 million emergency department visits by patients with primary psychiatric diagnoses across 312 hospitals in 28 states between 2021 and 2024. The dataset was drawn from the Healthcare Cost and Utilization Project's State Emergency Department Databases, supplemented by hospital-level data from the American Hospital Association Annual Survey.
The primary exposure variable was psychiatric boarding time — defined as the interval between medical clearance and either inpatient admission or discharge. The primary outcome was a composite of medical adverse events occurring during the ER stay, including medication errors, falls, pressure injuries, aspiration events, and cardiovascular events. Secondary outcomes included physical restraint use, chemical sedation use, and 30-day readmission rates.
The investigators used a rigorous instrumental variable approach to address confounding — specifically, the concern that sicker patients might both board longer and have more adverse events for reasons unrelated to boarding itself. The instrument used was hospital-level inpatient psychiatric bed availability on the day of the visit, which affects boarding time but is plausibly independent of individual patient severity.
Patients boarded for more than 12 hours had a 34% higher rate of the composite adverse event outcome compared to patients with boarding times under 4 hours (adjusted rate ratio 1.34, 95% CI 1.28–1.41, p<0.001). The relationship was dose-dependent: each additional 6 hours of boarding was associated with an approximately 9% increase in adverse event risk.
Patients boarded for more than 24 hours were 2.1 times more likely to be physically restrained at some point during their ER stay compared to patients with boarding times under 4 hours. The investigators note that restraint use itself is associated with adverse outcomes including aspiration, rhabdomyolysis, and cardiac arrest, suggesting a potential causal pathway between boarding and harm.
Medication errors — including wrong drug, wrong dose, and missed medication administration — were 1.8 times more common in patients with boarding times exceeding 12 hours. The investigators attribute this finding to the absence of standardized psychiatric medication management protocols in most emergency departments and the high nurse-to-patient ratios associated with prolonged boarding.
Patients with boarding times exceeding 12 hours had a 23% higher rate of 30-day emergency department readmission compared to those with shorter boarding times, suggesting that the quality of care and discharge planning during prolonged boarding visits is suboptimal.
The study's findings are particularly alarming given the scale of psychiatric boarding in the United States. The investigators found that the median psychiatric boarding time in their dataset was 8.4 hours — meaning that half of all psychiatric patients in the study spent more than 8 hours in the emergency department after medical clearance. Approximately 31% of patients boarded for more than 12 hours, and 14% boarded for more than 24 hours.
These figures reflect a chronic shortage of inpatient psychiatric beds that has worsened over the past two decades. The United States has approximately 11 psychiatric beds per 100,000 population, compared to a recommended minimum of 50 per 100,000. Between 2010 and 2024, the number of state psychiatric hospital beds declined by 14%, while the number of psychiatric emergency visits increased by 47%.
The result is a system in which emergency departments have become de facto psychiatric holding facilities — a function they are not designed, staffed, or equipped to perform. The NEJM study provides the most rigorous quantification to date of the patient harm that results from this structural mismatch.
The study was published on May 1, 2026, and was cited prominently in the preamble to the CMS behavioral health ER rule finalized three days later on May 4. CMS specifically referenced the 34% adverse event finding as evidence of the direct patient safety harm caused by inadequate behavioral health infrastructure in emergency departments.
The study has also been cited in congressional testimony supporting the EMTALA Modernization Act's behavioral health provisions, which require hospitals to document the clinical basis for psychiatric transfers and prohibit discharge of patients in acute psychiatric crisis solely on the basis of bed unavailability.
The American Psychiatric Association called the study "the most important piece of health services research on psychiatric emergency care published in the past decade," and called for immediate federal investment in inpatient psychiatric bed capacity as the only sustainable solution to the boarding crisis.
The investigators acknowledge several limitations. The study used administrative data, which may undercount adverse events that were not coded in billing records. The instrumental variable approach, while rigorous, relies on assumptions that cannot be fully verified. And the study cannot distinguish between adverse events caused by boarding itself versus adverse events caused by the underlying severity of illness that necessitated boarding.
Future research priorities identified by the investigators include prospective studies of specific adverse event mechanisms, evaluation of interventions designed to reduce boarding-associated harm (including dedicated psychiatric observation units and telepsychiatry-based disposition acceleration), and analysis of the cost-effectiveness of inpatient psychiatric bed expansion versus emergency department-based mitigation strategies.
ER Times summarizes high-impact emergency medicine research weekly. Subscribe for updates.