Multicentre implementation of a quality improvement initiative to reduce delirium in adult intensive care units: An interrupted time series analysis

J Crit Care. 2024 Jun:81:154524. doi: 10.1016/j.jcrc.2024.154524. Epub 2024 Jan 10.

Abstract

Purpose: The ABCDEF bundle may improve delirium outcomes among intensive care unit (ICU) patients, however population-based studies are lacking. In this study we evaluated effects of a quality improvement initiative based on the ABCDEF bundle in adult ICUs in Alberta, Canada.

Material and methods: We conducted a pre-post, registry-based clinical trial, analysed using interrupted time series methodology. Outcomes were examined via segmented linear regression using mixed effects models. The main data source was a population-based electronic health record.

Results: 44,405 consecutive admissions (38,400 unique patients) admitted to 15 general medical/surgical and/or neurologic adult ICUs between 2014 and 2019 were included. The proportion of delirium days per ICU increased from 30.24% to 35.31% during the pre-intervention period. After intervention implementation it decreased significantly (bimonthly decrease of 0.34%, 95%CI 0.18-0.50%, p < 0.01) from 33.48% (95%CI 29.64-37.31%) in 2017 to 28.74% (95%CI 25.22-32.26%) in 2019. The proportion of sedation days using midazolam demonstrated an immediate decrease of 7.58% (95%CI 4.00-11.16%). There were no significant changes in duration of invasive ventilation, proportion of partial coma days, ICU mortality, or potential adverse events.

Conclusions: An ABCDEF delirium initiative was implemented on a population-basis within adult ICUs and was successful at reducing the prevalence of delirium.

Keywords: Critical care; Delirium; Implementation science; Interrupted time series analysis; Patient care bundles.

Publication types

  • Clinical Trial
  • Multicenter Study
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Alberta / epidemiology
  • Critical Care
  • Delirium* / epidemiology
  • Delirium* / prevention & control
  • Humans
  • Intensive Care Units
  • Interrupted Time Series Analysis
  • Quality Improvement*