A mini EEG for the detection of delirium superimposed on dementia; A proof of concept study.

Delirium, also known as acute confusional state, poses unique challenges in
patients with cognitive impairment or clinically manifest dementia. It causes
immediate and extensive suffering in patients, distress in caregivers, and
health care staff. While the diagnosis of delirium can be difficult in
otherwise healthy older persons, detection of delirium superimposed on
pre-existing cognitive impairment or clinically manifest dementia (DSCID) is
extremely challenging. In this patient population it is difficult to accurately
distinguish the cardinal signs of delirium, i.e. newly occurring changes in
attention and cognition and a fluctuating clinical course, from signs of
pre-existing cognitive impairments or dementia. Moreover, from a clinical point
of view, distinguishing subtle changes in orientation, memory, executive
functions and behavior from pre-existing cognitive impairments is extremely
difficult, if not to say impossible given the currently provided diagnostic
scales. As a direct consequence of these diagnostic difficulties patients tend
to be misdiagnosed and thus mistreated, and are often prescribed psychotropic
medications without adequate assessment or more appropriate non/pharmacological
counseling. Therefore, reliance on clinical assessment for the diagnosis of
DSCID may be suboptimal.
These considerations have sparked researchers to increase diagnostic accuracy
by using more objective information that rely on electroencephalography (EEG).
Classical literature, dating back 60 to 70 years, strongly suggests already
that electroencephalography (EEG) can distinguish between delirium and
non-delirium using EEG frequency band parameters. More recent research confirms
these findings by investigating whether maintenance of activation (eyes open)
assessed by EEG discriminates delirium in association with dementia from
dementia without delirium and cognitively unimpaired elderly subjects.
However, standard EEG recordings, applying 25 electrodes, taking 20 minutes or
more, in a neurophysiology lab with specialized recording facilities, are not
feasible in DSCID populations. Recent studies by the group of prof. Slooter et
al. in postoperative patients have shown that it is possible to reliably detect
delirium using only a four electrode EEG recording (reference electrode, and
three recording electrodes), of less than 10 minutes with bedside, automatic
processing based on the relative power in de delta and theta frequency band. A
pilot study in the target population (VUMC-METC 2015.408) documented that
wearing this tiny EEG device was not associated with any additional discomfort
compared to mere clinical observation.
In our current study, this promising, innovative bedside mini EEG monitor with
automatic processing will be evaluated in a proof of concept study for the
early and reliable detection of delirium in patients with cognitive impairment
or clinically manifest dementia. Ultimately an improved diagnosis of delirium
in these patients will facilitate early treatment and prevention of subsequent
cognitive deterioration.

Detection of delirium superimposed on pre-existing cognitive impairment or
clinically manifest dementia (DSCID) is extremely challenging. As a direct
consequence of these diagnostic difficulties patients tend to be misdiagnosed
and thus mistreated.
In our current study, an innovative bedside mini EEG monitor with automatic
processing will be evaluated in a proof of concept study for the early and
reliable detection of delirium in patients with cognitive impairment or
clinically manifest dementia. Ultimately an improved diagnosis of delirium in
these patients will facilitate early treatment and prevention of subsequent
cognitive deterioration.

The study is a Prospective cohort observational study.

All patients together with their legal representatives, who belong to the above
defined source population, and who fulfil inclusion criteria, but none of the
exclusion criteria, will be informed about this study by a member of the
research team. The patients, or their legal representative, will be asked to
sign informed consent. After informed consent is received, the patient will be
given a study number based on chronological order of admission.

The first measurement (T1) takes place as soon as possible, yet at least one
week after
admission. Both a 5 minutes eyes-open, and 5 minutes eyes-closed EEG-monitor
recording will be conducted, while the patient is observed for signs of
discomfort, using the Discomfort Scale * Dementia of the Alzheimer Type
(DS-DAT). This will be used as baseline measure.

Within the same hour of this recording, after removing the electrodes, a
standardized clinical screening according to the Confusion Assessment Method
for the Intensive Care Unit (CAM-ICU) and the Observational Scale of Level of
Arousal (OSLA), and will be recorded on digital video. Patients will be
examined for presence or absence of a possible delirium using the DSM-V
criteria by a member of the research team (blinded for the mini EEG monitor
result). Finally a clinical conclusion 1) delirious, 2) possibly delirious, 3)
not delirious will be documented.
To overcome an order effect of activation, the patients will be randomized for
order of assessments: half of the participants will first undergo the mini EEG
after which the CAM-ICU will be assessed and the other half will first be
assessed according to the CAM-ICU, after which they will undergo the mini EEG.
When patients have a delier during their stay in the clinic, they will undergo
a maximum of three more measurements (T2,T3,T4=within measurements), at least
seven days after T1. All participating patients will have a last measurement at
about one week before discharge (T5=end measurement). Discharge takes place
when there is no longer any susceptibility of possible delirium.

The researcher who will be analyzing the mini EEG registrations will be blinded
for the classification made by the delirium experts, as described below, and
the delirium experts will be blinded for the results of the mini EEG monitor.
Furthermore, all treating doctors and nursing specialists will be blinded for
the results of the mini EEG monitor.

From our previous study we learned that EEG recordings in delirious patients
are feasible without problems. The delirium monitor will make use of only four
electrodes instead of 25, and is designed to have optimal usability.

A maximum of 5 measurements with the EEG device will take place. Each EEG
measurement takes place for 10 minutes maximal. The patients will be observed
while wearing the EEG device (see Figure 1). The patient will be asked to keep
his/her eyes open for 5 minutes followed by 5 minutes eyes closed. There will
be no forcing, yet a neutral and natural situation will be created. Thereafter,
the headband with four electrodes will be removed. The length of the total
observation period will be taylor-made, depending on patients* (dis)comfort.
Rest periods will be introduced while necessary. Measurements will be stopped
earlier when signs of large discomfort are observed.
There will be a group-related benefit, as development of an objective tool for
delirium detection superimposed on dementia may improve treatment of delirium
and therefore outcome.