SHM Converge Scientific Abstract ePoster Gallery-Sleep Smart: Evaluating an Evidence-Based Approach to Decrease Nightly Disturbances

Pdf Summary

This study evaluated whether automated EHR-based risk scores can identify which hospitalized patients are low risk for overnight deterioration and therefore could be protected from unnecessary nighttime disturbances, versus high-risk patients who may need closer monitoring.<br /><br />Using data automatically exported from the electronic health record across six hospitals (2/28/22–5/15/23), the authors analyzed 359,409 inpatient non-ICU patient-days after filtering out observation, ED, and ICU patients. They compared three pre-bed risk scores: an institutional modified Modified Early Warning Score (jMEWS), the Epic Deterioration Index (DI), and the Cardiac Arrest Risk Triage (CART) score. Overnight outcomes were limited to events occurring between 11 PM and 8 AM, including new overnight medications, telemetry upgrade, ICU transfer, rapid response team activation, code, death, and composite outcomes. Manual validation of 37 records suggested the automated extraction was reliable.<br /><br />For the most severe outcome, code or death, DI performed best with an AUC of 0.93, compared with 0.89 for jMEWS and 0.86 for CART. For the broader composite outcome (excluding new medications), all three scores performed similarly, with AUCs around 0.73–0.74. Overnight risk increased sharply with higher DI scores; for example, patients with DI 25 had a much lower risk than those with DI 45.<br /><br />The authors conclude that the Epic Deterioration Index is the strongest of the tested scores for overnight risk stratification. Because it is widely available through common EHR integration, it could support clinical decision-making to reduce sleep disruption in low-risk patients while focusing monitoring and interventions on higher-risk patients.

Asset Subtitle

Alan Kubey

Meta Tag

Author List Alan Kubey, Caitlin R. Vander Neut, Jeffrey M. Riggio, Kelly McHugh, Melanie McArthur

Category Research

Concept Perioperative Mortality Risk Prediction Tool

Concept Overnight Deterioration

Concept Risk Stratification

Concept jMEWS

Concept Epic Deterioration Index

Distinguished Non-Finalist

Presenter Organization Thomas Jefferson University Hospital

Presenting Author Alan Kubey

Track Quality Improvement

Keywords

electronic health record

risk stratification

overnight deterioration

hospitalized patients

Epic Deterioration Index

Modified Early Warning Score

CART score

patient monitoring

clinical decision support

sleep disruption

Perioperative Mortality Risk Prediction Tool

Overnight Deterioration

Risk Stratification

jMEWS

Epic Deterioration Index