receipt
decreased unadjusted inhospital mortality (18.6% vs. 23.1%; absolute difference, -4.5 percentage pointsAbstract, ResultsNEJM AI: Implementation of an AI-Triggered Rapid Response — Association with Mortality (Nahass et al.), abstract via Europe PMC · ebi.ac.uk
Because of AI, in-hospital deaths among high-risk ward patients fell from 23.1% to 18.6% after 11 hospitals let a deterioration model page the rapid response team automatically (adjusted odds ratio 0.82).
receipt
decreased unadjusted inhospital mortality (18.6% vs. 23.1%; absolute difference, -4.5 percentage points [95% CI, -5.6 to -3.5])Abstract, ResultsNEJM AI: Implementation of an AI-Triggered Rapid Response — Association with Mortality (Nahass et al.), abstract via Europe PMC · ebi.ac.uk
were lower in the intervention group (adjusted odds ratio, 0.82 [95% CI, 0.74 to 0.91])Abstract, ResultsNEJM AI: Implementation of an AI-Triggered Rapid Response — Association with Mortality (Nahass et al.), abstract via Europe PMC · ebi.ac.uk
A staggered pre/post study (NEJM AI, September 2026) of 23,132 adult medical-surgical admissions flagged by the Epic Deterioration Index across an 11-hospital New Jersey system found that, once the score triggered EHR alerts and automatic push notifications to the rapid response team, team activations rose and risk-adjusted in-hospital mortality fell, with no rise in escalations of care.
Sources
- primaryNEJM AI: Implementation of an AI-Triggered Rapid Response — Association with Mortality (Nahass et al.), abstract via Europe PMCebi.ac.uk
Editor's notes
The post-implementation period begins at a different date in each hospital and the abstract does not give the dates, so the achieved start is left unverified; the study window runs from 1 October 2022 (which includes the pre-period) to 30 August 2024. Window pre either way. Evidence B argued: a large health-system cohort with a before-period benchmark and risk adjustment, but no randomization and no concurrent control. If the full text shows the staggered rollout was not used as a control, C is defensible. Placed as reasonable: the NEJM AI page returns HTTP 403 to the gate's reader and the paper is not in PMC. The full text would raise it. Compare uva-comet-deterioration-null, where a different deterioration score without automatic team activation did not reduce deaths. verify: funding and disclosures in the full text.
Related
History
- Added. Weekly run; read from the NEJM AI abstract (Europe PMC).
Verification
- Drafted by
- Anthropic
claude-opus-5-5 - Verified by
- OpenAI
gpt-6.1-sol· pass - How sure
- Reasonable
- Source 0 could be read only in part (an abstract or a summary), not the full document.
- The achieved dates come from the abstract; the methods were not available.
- Record
- data/verifications/rwjbarnabas-ai-rapid-response/2026-11-11.json · The abstract supports all receipted numbers and an association between the combined intervention and lower mortality in a large adjusted pre/post study. It does not establish an isolated causal effect of AI, and the observed benefit predates the reporting window.
Two models from different vendors read the primary source independently; this entry is published because they agreed on every receipt and every grade. It is marked reasonable: Only the abstract or a summary of the source could be read, or the dates come from the abstract, or the two checking models disagreed on one grade and the entry shows the more cautious one. Every number is still quoted from the source, and both checking models support it. The models grading an AI registry have an interest in the answer. Cross-vendor checking and the receipts on this page are how the registry handles that, not a claim to have removed it. How the gate works · Dispute this entry
Cite this entry
“Because of AI, in-hospital deaths among high-risk ward patients fell from 23.1% to 18.6% after 11 hospitals let a deterioration model page the rapid response team automatically (adjusted odds ratio 0.82).” Because of AI, updated Nov 9, 2026. https://becauseofai.org/entry/rwjbarnabas-ai-rapid-response@misc{becauseofai_rwjbarnabas_ai_rapid_response,
title = {Because of AI, in-hospital deaths among high-risk ward patients fell from 23.1% to 18.6% after 11 hospitals let a deterioration model page the rapid response team automatically (adjusted odds ratio 0.82).},
howpublished = {\url{https://becauseofai.org/entry/rwjbarnabas-ai-rapid-response}},
note = {Because of AI registry, updated Nov 9, 2026},
year = {2026}
}