Content uploaded by Greg Fairbrother
Author content
All content in this area was uploaded by Greg Fairbrother on May 31, 2021
Content may be subject to copyright.
Reducing EEG (Electroencephalogram) Electrode-induced
Skin Injury among Ambulatory EEG Monitored Patients: A
Non-randomized Interventional Study of Two Commonly
Used Cream-based Products for Electrode Application
Sumika Ouchida, MSN, NP, CNS
1
; Armin Nikpour, MBBS, BSc,
FRACP
1,2
; Greg Fairbrother, RN, Ph.D., MPH, BA
2,3
1
Comprehensive Epilepsy Service
Royal Prince Alfred Hospital, Camperdown
Sydney, Australia
2
Faculty of Medicine & Health
University of Sydney
Sydney, Australia
3
Sydney Research
Sydney Local Health District, Camperdown
Sydney, Australia
ABSTRACT. Ambulatory electroencephalography (AEEG) seeks to capture
inter-ictal epileptiform activity or paroxysmal events when patients are not in the
clinic. Skin inflammation is a common complication of prolonged EEG monitoring.
This non-randomized study aimed to investigate the performance of two commonly
used cream-based methods of electrode application in reducing electrode-induced
skin injury among patients undergoing AEEG monitoring. A non-randomized
interventional study was conducted from July to December 2019 in the
Neurosciences Ambulatory Care Unit at Royal Prince Alfred Hospital, Australia.
Patients were enrolled into two groups: i) Group T, which received Ten20®
Conductive Paste with Tensive® Conductive Adhesive Gel as the primary approach
to electrode application; ii). Group E, which received EC2⁺® Conductive Cream as
the primary approach to electrode application. Patients in Group T were enrolled in
the 1st and 3rd week of the month, and patients in Group E were enrolled in the 2nd
and 4th week for each month of the study. A total of 152 patients participated in this
study. Two sub-groups were established: those who were monitored for two days
(Group T; n = 36, Group E; n = 30) and those who were monitored for four days
Corresponding Author’s E-mail: sumika.ouchida@health.nsw.gov.au
Received: March 31, 2020. Accepted for publication: September 25, 2020.
The Neurodiagnostic Journal, 00: 1–17, 2020
© 2020 ASET – The Neurodiagnostic Society
ISSN: 2164-6821 print / 2375-8627 online
DOI: https://doi.org/10.1080/21646821.2020.1829894
1
(Group T; n = 43, Group E; n = 43). Significant (p < 0.05) differences indicating
greater inflammation in the Group E were noted for both Day 2 and Day 4
participants. Skin injury/inflammation was significantly less using the standard
method (Group T: Ten20® with Tensive® gel) when compared to EC2⁺® (Group E)
as the conductive material at the electrode site.
KEY WORDS. Ambulatory EEG monitoring, EEG electrodes, EEG quality,
skin assessment, skin injury.
INTRODUCTION
Electroencephalography (EEG) is a monitoring method to record electrical brain
activity. Recording inter-ictal epileptiform activity or paroxysmal events using ambula-
tory electroencephalography (AEEG) is an important diagnostic tool (Faulkner et al.
2012; Seneviratne et al. 2013; Tolchin et al. 2017). AEEG recording devices are portable,
inexpensive, and widely available (Lawley et al. 2015; Seneviratne and D’Souza 2019).
One of the disadvantages of prolonged EEG monitoring is skin irritation or skin injury at
the EEG electrode sites which can occur in 7.8 to 27.3% of inpatients, depending on the
duration of the monitoring (Drees et al. 2016; Moura et al. 2017). A 2017 observational
study of prolonged AEEG monitoring indicated that most participants (81.7%) had skin
irritation or injury at the electrode sites (Ouchida et al. 2019).
Two possible mechanisms may cause electrode-related skin injury (Stecker et al.
2006), Mechanical injury can occur during the process of skin preparation (the removal of
layers of scalp epidermis) and prolonged direct pressure from the electrodes on the head
during the monitoring and at the time of electrode removal. Chemical injury can also occur
during the process of applying conductive agents, causing skin irritation or contact
dermatitis.
A previous study of electrode-induced skin injury among AEEG patients which
used collodion glue showed an increase in skin inflammation between Day 2 and Day
4, and that certain risk factors (increasing age, fair skin color, dry skin texture and fine
hair texture) were associated with higher pressure injury scores across the majority of
the electrode positions (Ouchida et al. 2019).
ASET - The Neurodiagnostic Society (ASET 2016) developed skin safety
guidelines during EEG electrode application; however, there are few experimental
research findings which test competing approaches to minimizing electrode related
skin injury.
This non-randomized interventional study sought to answer the following ques-
tions: i) Is EC2⁺® conductive cream a superior approach to reducing skin injury than
Ten20® with Tensive® gel (our laboratory standard approach), among patients under-
going AEEG monitoring? ii) Does the use of EC2⁺® conductive cream yield
2 REDUCING EEG ELECTRODE-INDUCED SKIN INJURY
equivocal quality EEG recording data when compared with the use of Ten20® with
Tensive® gel? Study participants’ subjective tolerance of AEEG monitoring was also
explored.
METHODS
Study Design and Setting
A non-randomized interventional study was conducted from July to December 2019
in the Neurosciences Ambulatory Care Unit at Royal Prince Alfred Hospital. Patients
were enrolled into two groups: i) Group T (the laboratory standard approach), which
received Ten20® with Tensive® gel as the primary approach to electrode manage-
ment/care; ii). Group E (the Active approach), which received EC2⁺® conductive
cream as the primary approach to electrode management/care. Patients in Group
T were enrolled in the 1st and 3rd weeks for each month of the study, and patients
in Group E were enrolled in the 2nd and 4th weeks.
Participants
Patients were eligible for inclusion in the study if they were older than 16 years of age
and had been referred for AEEG monitoring for either two or four days. Patients were
excluded if they had skin irritation/inflammation or head lice on their scalp. Patients were
also excluded if they were allergic to Ten 20® conductive paste, Tensive® gel, EC2⁺®
cream, confused or agitated or unable to tolerate AEEG monitoring.
The hospital’s Institutional Review Board approved the study protocol. Written
informed consent was obtained from all patients or their guardian if the patients were
under 18 years of age or unable to give signed consent.
The sample size was calculated based on a published EEG electrode related skin
irritation rate of 27.3% (Drees et al. 2016). To have an 80% chance of detecting
a difference of 15% or greater at the 0.05 significance level, a target of 154 total
participants was needed (77 patients in each group).
Procedures
All procedures were conducted by two neurophysiology registered nurses who had
a minimum of two years of working as EEG technologists. Epileptologists and
a senior EEG technologist who completed a Graduate Diploma in Medical Science
(Clinical Neurophysiology) trained the neurophysiology nurses, and they received
weekly education throughout the study period. They also received comprehensive
education about the study procedures, especially EEG electrode application.
REDUCING EEG ELECTRODE-INDUCED SKIN INJURY 3
The two trained neurophysiology nurses assessed consecutive patients. Nurses
consented patients and completed a pre-application skin assessment form before
applying the EEG electrodes.
Skin color, skin texture, hair volume and hair texture were assessed by visual and
tactile inspection by the nurse who evaluated the patient at baseline. Sleep position
(left side, right side, prone or supine) & movement during sleep (frequent movement
or still) were assessed by asking the patient and/or their family members directly.
Further patient characteristics (age, gender), were also recorded at this time. Ten
patients were selected at random to participate in an inter-rater reliability study of
skin-, hair- and sleep-related patient characteristic assessment. Obtaining this data
involved the conduct of separate assessments by nurses who were blind to each
other’s assessments.
Skin condition assessment was done at all 23 sites of electrode application,
including the ground and reference electrodes on the day of electrode removal (Day
2 or 4) (Figure 1) and digital photographs of the patients’ scalps were taken. Inter-rater
reliability of skin assessment scoring was assessed by enrolling nurses blinded to the
EEG application method. They completed the skin assessment tool by blinded review
of unannotated photographs of 20 randomly selected patients for the usually hairless
Fp1 and Fp2 electrode sites. These nurses did not have knowledge of the study
allocation schedule, which was maintained by the research nurses.
The skin assessment tool yielded a score from 1–4, reflecting erythema presence/
severity, where 1 = minimal erythema, 2 = moderate erythema with sharply defined borders,
3 = intense erythema with or without edema and 4 = intense erythema with edema and
blistering/erosion (Figure 2). This skin assessment scale was used in our previous study
(Ouchida et al. 2019). Similar skin assessment procedures for electrode-related pressure
injury have been recently described (Drees et al. 2016; Moura et al. 2017).
The neurophysiology nurses interviewed all eligible patients and collected demo-
graphic information prior to the application of electrodes.
Application of EEG Electrodes
Step 1 (Both Groups)
The nurse measured the patient’s head according to the International 10/20 system of
electrode placement and marked the 23 sites of electrode position with an “X” using
a colored pencil. A cotton-tipped applicator was used to apply a small amount of
NuPrep® Skin Prep Gel (Weaver and Company, Auroa, CO, USA) to the center of the
marked “X”; this was rubbed with the side of the cotton-tipped applicator for five seconds in
one direction to remove dead skin cells and oils from the scalp.
4 REDUCING EEG ELECTRODE-INDUCED SKIN INJURY
Step 2
Group T (Ten20® with Tensive® gel): After skin preparation, the nurse applied Ambu®
Neuroline disposable EEG electrodes (Ambu A/S, Bellerup, Denmark) onto the scalp by
using Ten 20 conductive paste™ (Weaver and Company, Auroa, CO, USA) and then
a minimal amount of Tensive® adhesive gel (Parker laboratory, Fairfield, NJ, USA), as
Ten 20 cream does not have adhesive properties. Hypafix tape (BSN Medical Australia and
New Zealand, VIC, Australia) was used to secure the electrodes. The electrode wires were
connected to the recording device (Siesta 802, Compumedics, Charlotte, NC, USA) to
FIG. 1. EEG electrode positions.
1= Minimal erythema 2= Moderate erythema
with sharply defined
borders
3= Intense erythema
with or without oedema
4= Intense erythema
with oedema and
blistering/erosion
FIG. 2. Pressure injury scale.
REDUCING EEG ELECTRODE-INDUCED SKIN INJURY 5
check the impedance of the electrodes (aiming for less than 5 kΩ). The value of impedance
used in this study was based on the laboratory protocol, which was based on the Association
of Neurophysiology Technology of Australia’s EEG Guidelines (ANTA 2018) and the
American Clinical Neurology Society’s Long-Term Monitoring for Epilepsy Guidelines
(ACNS 2008). Mölnlycke Tubifast® tubular bandage (Mölnlycke Health Care AB,
Goteborg, Sweden) was used for bandaging. This tubular bandage provides equal and
consistent pressure on the scalp of the patients.
Group E (EC2⁺®): After the skin preparation, the nurse applied Ambu®
Neuroline disposable EEG electrodes onto the scalp by using EC2⁺® conductive
cream (Natus Manufacturing Ltd, Galway, Ireland). Tensive® adhesive gel was not
used as EC2⁺® cream has both adhesive and conductive properties. Hypafix tape was
used to secure the electrodes. Mölnlycke Tubifast® tubular bandage was placed on the
patient’s scalp to secure the electrodes.
Post-electrode Application
If the patient returned to the clinic due to electrode dislodgement or experiencing
painful sensations at the site of the electrodes, the bandage was removed to assess the
skin condition at the electrode site(s). If the impedance of the electrodes was higher
than 5 kΩ, or if there were artifacts in the recording, the affected electrodes were
removed. If there were any signs of skin irritation (redness, blisters, abrasion), the
electrode was repositioned 0.5 cm away from the original location, to allow for the
skin to recover. If no signs of skin irritation were seen, the electrodes were re-glued at
the original site. Then the procedure steps as outlined in Step 2 above, were followed.
Patients who received four days of AEEG monitoring were required to return to
the clinic on Day 2 post-application for repositioning of the frontal pole electrodes
(Fp1 & Fp2) to prevent or minimize skin irritation.
Day of EEG Electrode Removal
Two nurses checked the impedances and completed the EEG quality data scale.
This scale measured impedance, artifact presence, electrode dislodgment and record-
ing quality. The scale yielded a score from 0–5, where 0 indicated poor quality and 5
indicated high quality. A total score (0–5) was calculated to indicate the overall
quality of each EEG recording. The questions were as follows:
i. Were there any electrode artifacts (pop or 50 Hz) in the EEG recording on
the day of EEG electrode removal (Day 2 or Day 4)? (Yes = 0, No = 1)
ii. Was the impedance of all electrodes on the last day of the AEEG recording
less than 5 kΩ? (Yes = 1, No = 0)
iii. Were any of the electrodes dislodged? (Yes = 0, No = 1)
6 REDUCING EEG ELECTRODE-INDUCED SKIN INJURY
iv. Compared to day 1 of AEEG recording, how was the quality of the last day
AEEG recording? (Good = 2, Fair = 1, Poor = 0)
v. Total score is the sum of all scores. 0 indicates poor quality and 5 indicates
high quality.
Patient tolerability of AEEG monitoring was also compared across the two
groups. Patients were asked to complete a five-item questionnaire about the comfort
associated with AEEG monitoring. Four questions were posed to participants. These
were:
i. How strongly did you feel the presence of a foreign object on your head?
ii. How strongly did the tactile sensations (itchiness or aching) on your head
attract your attention?
iii. How strongly would you like to remove the EEG electrodes from the head
during the testing?
iv. How strong was your feeling of being embarrassed while having the object
on your head in the presence of other people?
A six-point Likert scale was offered for each item, where 0 = no discomfort and
5 = very high level of discomfort. The highest possible total score achievable was 20
and the lowest score possible was 0.
After all EEG electrodes were removed, each participant’s scalp was digitally
photographed. Examples of scalp injury in all five scalp domains are included in Figure 3.
FIG. 3. Images of skin injury on the five scalp domains.
REDUCING EEG ELECTRODE-INDUCED SKIN INJURY 7
Data Analysis
The two groups (T and E) were compared for characteristics and outcomes using
univariate parametric (one-way analysis of variance) and non-parametric (Mann-Whitney
U and chi-square) tests. To test inter-rater reliability for skin/hair type assessment and EEG
data quality assessment (n = 10 participants), the Kappa statistic was utilized for dichot-
omous variables (McHugh 2012) and Spearman correlation analysis was utilized for ordinal
variables (Schober et al. 2018). To test inter-rater reliability for the primary outcome (skin
inflammation), the Kappa statistic was employed alone in a larger sample of n = 20
randomly selected participants, and for the Fp1 and Fp2 sites only, as these sites have the
most considerable burden of inflammation (Ouchida et al. 2019).
RESULTS
From July to December 2019, 158 patients were enrolled in the study. Figure 4
outlines the flow of patients through the study. Patients were enrolled in two groups:
Group T (81 patients) and Group E (77 patients).
A total of 152 patients were included for outcome analysis. Six patients discon-
tinued involvement in the study due to high levels of discomfort (See Figure 4: Patient
flow diagram). Comparisons on characteristics and outcome were conducted between
Groups T (Ten20® with Tensive® gel) and E (EC2⁺®) for patients who had the
AEEG removed at Day 2 of monitoring (n = 66; of which Group T = 36 and Group
E = 30) and for patients who had the AEEG removed at Day 4 (n = 86; of which
Group T = 43 and Group E = 43). Sample characteristics are summarized in Table 1.
No significant differences were noted between the two groups on demographics, skin,
hair and sleep-related descriptors.
Table 2 outlines a comparison between groups for Day 2 and Day 4 samples.
Significant (p < 0.05) differences indicating greater inflammation in Group E were noted
for frontal and midline regions for both Day 2 and Day 4 participants. Significantly
increased inflammation in Group E was also noted for the temporal region at Day 2 (but
not Day 4) and for the parietal and occipital regions at Day 4 (but not Day 2). This difference
might be influenced by the mild difference in mean age between patients who had their
electrodes removed on Day 2 vs Day 4, where Day 4 patients were younger. This difference
was not statistically significant though may be of practical importance. We also cannot
exclude the possibility that slight differences in nurse-driven application practices might
have influenced inflammation rates. It remains the case though, that only two highly trained
nurses were utilized in the study protocol, and they often worked together on the one patient
(e.g. one worked on the left side of the patient’s head and one worked on the right).
Table 3 outlines self-reported patient comfort scoring and indicates that no by-
group differences were noted for either Day 2 or Day 4 groups. Table 4 outlines
results for data quality and indicates no by-group differences for either Day 2 or Day 4
8 REDUCING EEG ELECTRODE-INDUCED SKIN INJURY
groups. Table 5 explores regional inflammation score by sleep-related characteristics
for the group who had the AEEG removed at Day 4. There was an association
between the habit of sleeping on the side, with higher inflammation in the frontal
region (F7/F8). Still sleep movement and side sleeping were associated with higher
inflammation in temporal and midline regions.
Table 6 outlines the results of inter-rater reliability sub-studies for hair/skin
assessment at baseline, EEG data quality and skin inflammation scoring. The results
of these analyses suggest high reliability for EEG data quality, medium-high relia-
bility for hair/skin assessment and high reliability for skin inflammation scoring.
FIG. 4. Recruitment flow diagram.
REDUCING EEG ELECTRODE-INDUCED SKIN INJURY 9
Table 1. Participants characteristics.
Group T
Day 2
(n = 36)
Group E
Day 2
(n = 30)
Difference
Day 2
P
Group T
Day 4
(n = 43)
Group E
Day 4
(n = 43)
Difference
Day 4
P
Age Mean 50.3
SD 16.8
Mean 48.4
SD 21.5
ANOVA: 0.70
F = 0.15
Mean 47.0
SD 19.5
Mean 41.4
SD 17.3
ANOVA: 0.16
F = 2.0
Gender 61.1% Female 57.7% Female X
2
= 0.13
0.72
41.9% Female 60.5% Female X
2
= 2.98
0.08
Skin color 66.7% Fair
27.8% Medium
5.6% Dark
66.7% Fair
23.3% Medium
10.0% Dark
X
2
= 0.55
0.76
67.4% Fair
27.9% Medium
4.7% Dark
69.8% Fair
27.9% Medium
2.3% Dark
X
2
= 0.35
0.84
Skin texture 41.7% Dry
58.3% Moist
36.7% Dry
63.3% Moist
X
2
= 0.17
0.68
53.5% Dry
46.5% Moist
46.5% Dry
53.5% Moist
X
2
= 0.42
0.52
Hair volume 80.6% Full
11.1% Partial
8.3% Bald
83.3% Full
13.3% Partial
3.3% Bald
X
2
= 0.76
0.69
74.4% Full
20.9% Partial
4.7% Bald
86.0% Full
14.0% Partial
0% Bald
X
2
= 2.96
0.23
Hair texture 58.3% Fine
35.4% Thick
8.3% N/A
66.7% Fine
30.0% Thick
3.3% N/A
X
2
= 2.32
0.36
55.8% Fine
39.5% Thick
4.7% N/A
53.5% Fine
46.5% Thick
0% N/A
X
2
= 0.43
0.51
Sleep side 38.9% Right
44.4% Left
13.9% Prone
2.8% Supine
43.3% Right
43.3% Left
3.3% Prone
10.0% Supine
X
2
= 3.50
0.32
41.9% Right
32.6% Left
14.0% Prone
11.6% Supine
44.2% Right
41.9% Left
4.7% Prone
9.3% Supine
X
2
= 2.60
0.45
Movement sleep 25.0% Still
75.0% Frequent
33.3% Still
66.7% Frequent
X
2
= 0.55
0.46
37.2% Still
62.8% Frequent
30.2% Still
69.8% Frequent
X
2
= 0.47
0.49
10 REDUCING EEG ELECTRODE-INDUCED SKIN INJURY
Table 2. Skin inflammation score at Day 2 and Day 4 by Region.
Group T
Day 2
(n = 36)
Median
[Mean]
Group E
Day 2
(n = 30)
Median
[Mean]
Difference
(Mann-Whitney U Test)
Day 2
Z P
Group T
Day 4
(n = 43)
Median
[Mean]
Group E Day 4
(n = 43)
Median
[Mean]
Difference
(Mann-Whitney U Test) Day 4
Z P
Frontal (F3, F7, Fp1, Fp2, F4, F8)
(Scoring Range: 0–24)
4
[4.6]
6.5
[6.9]
–2.7 0.006** 3
[4.1]
6
[6.4]
–3.2 0.002**
Temporal (T3, A1, T5, T4, A2, T6)
(Scoring Range: 0–24)
0.5
[1.7]
2.5
[3.0]
–2.6 0.008** 0
[1.7]
1
[2.0]
–0.9 0.39
Parietal (C3, P3, C4, P4)
(Scoring Range:0–16)
0
[0.7]
0
[0.8]
–0.6 0.53 0
[0.3]
0
[1.1]
–2.9 0.004**
Midline (GRD, Fz, REF, Cz, Pz)
(Scoring Range: 0–20)
1
[2.1]
3
[3.1]
–2.5 0.014* 1
[1.8]
2
[2.0]
–2.2 0.028*
Occipital (O1, O2)
(Scoring Range 0–8)
0
[0.3]
0
[0.3]
–0.8 0.42 0
[0.0]
0
[0.2]
–2.0 0.048*
*P < 0.05, **P < 0.01
REDUCING EEG ELECTRODE-INDUCED SKIN INJURY 11
Table 3. Patient comfort at Day 2 and Day 4.
Group T
Day 2
(n = 36)
Median
Group E
Day 2
(n = 30)
Median
Difference
(Mann-Whitney U
Test)
Day 2
Z P
Group T
Day 4
(n = 43)
Median
Group E
Day 4
(n = 43)
Median
Difference
(Mann-Whitney
U Test)
Z P
Feeling
How strongly did you feel the presence of a foreign
object on your head?
(Scoring Range 0–5)
3 4 –0.9 0.37 3 3 –0.4 0.71
Sensations
How strongly did you feel tactile sensations - itchiness
or achiness?
(Scoring Range 0–5)
2 3 –0.5 0.64 3 3 –1.2 0.22
Wish to remove
How strongly did you want to remove the EEG
electrodes from your head?
(Scoring Range 0–5)
5 5 –0.3 0.77 4 4 –0.7 0.47
Embarrassment
How strong was your feeling of being embarrassed
when with other people?
(Scoring Range 0–5)
2 2 –0.3 0.77 2 1 –1.2 0.25
Patient Comfort Total
(Scoring Range 0–20)
11 12 –0.08 0.44 11 11 –0.46 0.64
12 REDUCING EEG ELECTRODE-INDUCED SKIN INJURY
Table 4. Data quality at Day 2 and Day 4.
Group T
Day 2
(n = 36)
Group E
Day 2
(n = 30)
Difference
P
Group T
Day 4
(n = 43)
Group E
Day 4
(n = 43)
Difference
P
Artifact = Y 61.1% 76.7% X
2
= 1.8
0.18
83.7% 67.4% X
2
= 3.1
0.08
Impedance less than
5kΩ = Y
38.9% 50.0% X
2
= 0.8
0.37
7.0% 14.0% X
2
= 1.1
0.29
Dislodge = Y 30.6% 26.7% X
2
= 0.1
0.73
46.5% 58.1% X
2
= 1.2
0.28
Quality 1. Poor: 11.1%
2. Fair: 38.9%
3. Good: 50.0%
1: 16.7%
2: 50.0%
3: 33.3%
X
2
= 1.9
0.39
1: 32.6%
2: 44.2%
3: 23.3%
1: 51.2%
2: 27.9%
3: 20.9%
X
2
= 3.4
0.18
Total score for data
quality
(Scoring range: 0–5)
Median: 3 Median: 2 Mann-Whitney 0.45 U
Test: z = −0.8
Median: 2 Median: 1 Mann-Whitney 0.55 U
Test: z = −0.6
REDUCING EEG ELECTRODE-INDUCED SKIN INJURY 13
DISCUSSION
The purpose of this non-randomized interventional study was to compare electrode
site inflammation associated with a newer cream-based method of supporting con-
ductivity at the electrode site (EC2⁺®; Group E) with those of our laboratory standard
method (Ten20® with Tensive® gel; Group T), on Day 2 and Day 4 of AEEG
monitoring. Quality of recorded EEG data and self-reported patient comfort were
also compared.
Interestingly, our laboratory standard method (Group T: Ten20® with Tensive® gel)
performed significantly better than EC2⁺® in terms of electrode site inflammation. This
result applied to the patients who had the AEEG removed on Day 2 and those who had it
Table 5. Inflammation at Day 4 by sleep-related characteristics [n = 86 (Group T = 43; Group E = 43)].
Inflammation greater: Frontal Temporal Parietal Midline Occipital
Sleep on side [not prone or supine] (ANOVA) 0.039* NS NS NS NS
Sleep movement is still [not frequent] (ANOVA) NS P = .029* NS 0.068# NS
NS = Not significant; *P < 0.05; #Of practical importance (0.05 < P < 0.10)
Table 6. Inter-rater reliability results.
% agreement
(n = 2 blind raters)
Reliability statistic
(Strength of agreement) P
EEG data quality
(n = 10 patients)
Artifact (yes/no) 90 Kappa = 0.78 (Moderate) 0.011
Impedance (yes/no) 100 Kappa = 1.0 (Perfect) 0.002
Electrode dislodgement (yes/no) 100 Kappa = 1.0 (Perfect) 0.002
Quality (poor/fair/good) 70 Spearman’s Rho = 0.76 (Strong) 0.011
Total score (range 0–5) 60 Spearman’s Rho = 0.93 (Strong) <0.0001
Skin type/hair type assessment
(n = 10 patients)
Skin color (fair/medium/dark) 90 Spearman’s Rho = 0.75 (Strong) 0.013
Skin texture (dry/moist) 80 Kappa = 0.55 (Weak) 0.053
Hair volume (full/partial loss/bald) 90 Spearman’s Rho = 0.99 (Strong) <0.0001
Hair texture (fine/thick) 100 Kappa = 1.0 (Perfect) 0.002
Skin inflammation assessment
(n = 20 patients)
Fp1 electrode site (1–4 scale) 85 Kappa = 0.78 (Moderate) <0.0001
Fp2 electrode site (1–4 scale) 90 Kappa = 0.84 (Strong) <0.0001
14 REDUCING EEG ELECTRODE-INDUCED SKIN INJURY
removed on Day 4. Most domains of electrode position (Frontal: F7, F3, Fp1, Fp2, F4, F8,
Temporal: T3, A1, T3, T4, A2, T6, Parietal: C3, P3, C4, P4, Midline: GRD, Fz, REF, Cz,
Pz, Occipital: O1, O2) showed better performance using Ten20® with Tensive® gel, with
frontal and midline domains showing the most consistency across the Day 2 to Day 4
divide.
Despite skin inflammation results favoring Ten20® with Tensive® gel, the patient
comfort self-reports did not show any differences between the groups, i.e. Group E patients
did not have a worse subjective experience. The patient comfort self-reports do show quite
a high level of discomfort overall though. This result has been shown previously by
Ouchida et al. (2019). Although one group clearly had less inflammation, all the participants
rated the experience at a high level of discomfort. The inter-rater reliability results for
inflammation assessment suggest that the tools used in this study were reliable.
EEG data quality scale measured artifact presence (pop, 50/60 Hz), impedance of
electrodes on the last day of recording, presence of electrode dislodgments and
recording quality comparing the Day 1 recording to the last day of recording. EEG
data quality did not differ between the groups, suggesting that conductive gel choice
was not a factor at play in relation to data quality. The decision to use an impedance
value of less than 5 kΩ was based on both Australian and North American guidelines.
We will however consider using a <10 kΩ value in future clinical studies (Sinha et al.
2016).
The cost associated with the use of either of the two product options was very
similar (~A$29 per person). Associations between inflammation level and patient
characteristics (demographic, hair and skin-related) have recently been reported by
Ouchida et al. (2019). In the current study, a new variable of interest, sleep habit,
was recorded. Sleeping on the side was linked to frontal region inflammation. This
link may be related to the positioning of the F7 & F8 electrodes which are in the
frontotemporal region. It should be allowed that this variable is likely to be a sub-
optimal proxy measure of sleep positioning as it is a self-report, rather than an
objective report by an observer (e.g. partner). Most commercially available sleep
position sensors are body strap sensors which our patients have reported to be
uncomfortable to wear on an all-day basis. Night-only use is not recommended in
our clinical protocol as it would require a high degree of patient comprehension of,
and compliance with, the strapping-on and strapping-off procedures. If wireless
sleep position sensors become available in the future, these would ideally be used.
The study confirms earlier findings reported by Ouchida et al. (2019) that skin
inflammation at the electrode site remains a significant burden for patients. Explorations
of the use of hydrogel electrodes may now be warranted. These have advantages such as
increased adhesion, smaller skin pressure points, sterility and disposability. A recent study
by El Ters et al. (2018) showed promising results with regards the use of hydrogel
electrodes on preterm neonates undergoing EEG. A randomized study of electrode type
(e.g. mixed hydrogel and disposable vs disposable) in the context of AEEG would be the
REDUCING EEG ELECTRODE-INDUCED SKIN INJURY 15
ideal next research step, as we seek to identify the best method to reduce skin inflamma-
tion at electrodes sites during AEEG monitoring.
CONCLUSION
Ten20® with Tensive® gel outperformed EC2⁺® cream as the conductive mate-
rial of choice in terms of the extent of skin inflammation associated with its use.
Despite experiencing more inflammation, patients in the Ten20® with Tensive® gel
(Group T) did not report higher discomfort, suggesting that the high-burden sub-
jective experience of AEEG is not impacted upon by skin inflammation level alone.
ACKNOWLEDGMENTS
We are thankful to the neurophysiology nurses of Neuroscience Ambulatory Care
Unit, Royal Prince Alfred Hospital, who supported and completed this study.
DISCLOSURE STATEMENT
The authors reported no potential conflict of interest.
DISCLAIMER
EC2⁺® conductive cream is currently unavailable for clinical use in the United
States of America.
REFERENCES
ACNS. 2008 Mar. Guideline twelve: guidelines for long-term monitoring for Epilepsy. Am Clin Neurophysiol
Soc. [Internet]. [accessed 2020 July 19]. https://www.acns.org/pdf/guidelines/Guideline-12.pdf.
ANTA. 2018. EEG guideline version 2. Australian Neurophysiological Technology Association. [Internet].
[accessed 2020 July 19]. https://www.anta.org.au/wp-content/uploads/2018/11/ANTA-Inc-EEG-
Guideline-v2-2018-1.pdf.
ASET - The Neurodiagnostic Society. 2016. Position statement on skin safety during EEG procedures –
a guideline to improving outcome. Neurodiagn J. 56(4):296–300.
Drees C, Makic MB, Case K, Mancuso MP, Hill A, Walczak P, Limon S, Biesecker K, Frey L. 2016. Skin
irritation during video-EEG monitoring. Neurodiagn J. 56(3):139–150.
El Ters NM, Mathur AM, Jain S, Vesoulis ZA, Zempel JM. 2018. Long term electroencephalography in
preterm neonates: safety and quality of electrode types. Clin Neurophysiol. 129(7):1366–1371.
Faulkner HJ, Arima H, Mohamed A. 2012. The utility of prolonged outpatient ambulatory EEG. Seizure. 21
(7):494–495.
Lawley A, Evans S, Manfredonia F, Cavanna AE. 2015. The role of outpatient ambulatory electroence-
phalography in the diagnosis and management of adults with epilepsy or nonepileptic attack disorder:
A systematic literature review. Epilepsy & Behav. 53:26–30.
McHugh M. 2012. Interrater reliability: the kappa statistic. Biochemia Medica. 22(3):276–282.
16 REDUCING EEG ELECTRODE-INDUCED SKIN INJURY
Moura LMVR, Carneiro TS, Moura VF, Blodgett CS, Cohen J, Guanci MM, Hoch DB, Hsu J, Cole AJ,
Westover MB. 2017. cEEG electrode-related pressure ulcers in acutely hospitalized patients. Neurol
Clin Pract. 7(1):15–25.
Ouchida S, Nikpour A, Fairbrother G, Senturias M. 2019. EEG electrode-induced skin injury among adult
patients undergoing ambulatory EEG monitoring. Neurodiagnostic J. 59(4):219–231.
Schober P, Boer C, Schwarte L. 2018. Correlation coefficients: appropriate use and interpretation. Anesth
Analg. 126(5):1763–1768.
Seneviratne U, D’Souza WJ. 2019. Chapter 10 - ambulatory EEG. Handb Clin Neurol. 160:161–170.
Seneviratne U, Mohamed A, Cook M, D’Souza WJ. 2013. The utility of ambulatory electroencephalogra-
phy in routine clinical practice: A critical review. Epilepsy Res. 105:1–12.
Sinha SR, Sullivan L, Sabau D, San-Juan D, Dombrowski KE, Halford JJ, Hani AJ, Drislane FW,
Stecker MM. 2016. American Clinical Neurophysiology Society guideline 1: minimum technical
requirements for performing clinical electroencephalography. J Clin Neurophys. 33(4):303–307.
Stecker MM, Patterson T, Netherton BL. 2006. Mechanisms of electrode induced injury. Part 1: theory. Am
J of Electroneurodiagnostic Technol. 46(4):315–342.
Tolchin B, Lee JW, Pavlova M, Dworetzky BA, Sarkis RA. 2017. Diagnostic yield of ambulatory EEGs in
the elderly. Clin Neurophysiol. 128(7):1350–1353.
REDUCING EEG ELECTRODE-INDUCED SKIN INJURY 17
