Delirium detection in routine clinical care: two basic processes

This blog is on an area that is very close to my heart: delirium detection in clinical practice. A tweet from a geriatrician illustrated the issue perfectly:

‘ “Delirium is common, yet under-recognised” Isn’t it worrying!? We knew that for decades. What can be done to increase recognition of delirium on admission and during hospital stay?’

Great question!

We would all like to have delirium detected in our patients when they first present and also if delirium arises during the hospital admission. I’d add that delirium arising in patients in long-term care facilities is another challenge, of course closely related to detecting new-onset delirium in hospital patients.

So we need to look at two processes:

(1) detection of delirium at first presentation, during care transitions, and at other times when delirium is first suspected, that is, during certain episodes of care - this can be termed episodic testing

(2) monitoring for new-onset delirium in inpatients, that is, a regularly-conducted process sensitive to post-admission delirium - this can be termed delirium monitoring

What approaches might be considered?

For (1) we need a tool that is sensitive and specific enough to get you close to a diagnosis (no tool can give a diagnosis - it’s always a matter of clinical judgement). In my classification of delirium tools I included the following tests with these features in this list:

These tests involve a combination of bedside assessment including cognitive testing and interview, and documentation of change or fluctuation. Such tests are not suited to doing more than once per day, because of their duration and the practice effects of cognitive tests.

For (2) we need tools with different characteristics - those that sensitive enough, but also do not involve multiple cognitive testing or interview processes. Such monitoring tools rely much more on observation of various delirium features. Many are designed for use by nurses. Some are scored in real-time at the bedside (eg. RADAR, NEWS2) and others are scored at the end of a shift. Here is a list:

The Royal College of Physicians issued recent guidance (March 2020) for use in the UK National Health Service (NHS) on use of these two basic delirium detection processes. The monitoring process is the National Early Warning Score - 2 or NEWS2. This includes a trigger for new confusion, and there is also a level of consciousness item:

This is part of the standard physiological observations chart, so would typically be performed 4 times a day in an acute ward. That would mean that, potentially, delirium is being screened for several times per day.

The Royal College of Physicians guidance then states that if NEWS2 suggests possible delirium, then a 4AT is performed.

This paired of assessments could work very well together as a way of improving detection of new onset delirium in inpatients. We just need now to do some evaluations and see what education and training is required to make it work.