Is Screening for Delirium a Part of Your Facility Workflow?

At this year’s annual Summit in October, Dr. Steven Poser presented important information on the distinction between Neurologic and Psychiatric causes of dementia, which I highlighted in the November 1 edition of the WAVE.  

In that presentation, Dr. Poser stressed the importance of appreciating delirium as a subtle contributor to a BPSD (Behavioral and Psychiatric Symptoms of Dementia) event.  In fact the literature in the acute hospital setting supports the need for routine screening for delirium, since delirium is quite prevalent in our population, but the problem is commonly under-recognized by staff and providers, especially in the hypoactive form.  Traditionally, the CAM (Confusion Assessment Method) or CAM-ICU have been used to screen for delirium, but its use by most hospital teams hasn’t been well standardized, resourced, or implemented.  Of note, a CAM-based delirium screening tool is part of the Minimum Data Set (MDS), but there may be some question as to how reliably this is completed and reported. 

In the November 9 Annals of Internal Medicine, a Comparative Implementation of a Brief App-Directed Protocol for Delirium Identification by Hospitalists, Nurses, and Nursing Assistants, was reported.  This study from academic centers in Boston and Pennsylvania used a research reference standard delirium assessment (RSDA) over 2 days, and then compared this screening tool with the app.   Nursing Assistants (CNAs) completed an ultra-brief 2-item screen (UB-2) in an average of 62 seconds.  If 1 of these 2 questions was positive, their licensed nurse and hospitalist would complete the UB-CAM app based on the CAM Criteria.  This process took 104 seconds for nurses and 106 for physicians.  The completion rate by end users for the app was 97% This 2 step protocol had an overall identification accuracy of 89%.  This information automatically populated to a research data base, REDcap.  

This app isn’t yet available for general use, but it's apparent value caused me to look for another similar app with evidence for effectiveness.  The AT4 is an app developed by UK, Scotland.  It’s free, well-validated, brief, commonly used in Europe, and has improved detection of delirium by nurses in hospitals and other settings.  It can be found on IOS and Android platforms by searching for “Sign Decision Support”.  The screening tool is known as the 4 A’s Tool and measures Alertness, Attention, Abbreviated mental test, and Acute change. The tool scoring range is 0-12 with a positive screen indicated if > 3.  If positive in patients > 70 y/o with acute confusion, the specificity was 95% and sensitivity was 76%, making this a practical tool for initial assessment in time pressured environments.  In this report from Scotland, the estimated costs of the initial inpatient stay for patients with delirium were more than double those without delirium and these costs continued to be greater 12 weeks later.  More information on this tool is available online (link provided below.)

I have downloaded this app from Scotland, to my iPhone and explored its functionality.  Having worked with many UK physicians in NGO medical relief and education work in Albania, I’m not surprised by the precision, simplicity, and practicality reflected in this app.  Not only is there an easy to use 4AT calculator, but also up to date sections on delirium risk reduction, treatment, COVID associated delirium, and other resources.  

I agree with Dr. Poser that delirium may be a common contributor to problem behaviors and prove costly to patients and facilities.  Early detection and management mitigates these issues and will likely improve care plans and reduce some of the care chaos that has become more challenging during the pandemic.  

With this in mind, is it time for your facility to build delirium screening into your workflow beyond the screening contained in the MDS?


Annals of Internal Medicine:

4 A's Tool:

Share this post:

Comments on "Is Screening for Delirium a Part of Your Facility Workflow?"

Comments 0-5 of 0

Please login to comment