Authors: Aslanidis Th.1 MD, Grosomanidis V.1 MD, PhD, +Karakoulas K.1 MD, PhD, Chatzisotiriou A.2, MD, PhD
1Department of Anesthesiology and Intensive Care Medicine, AHEPA General University Hospital, Thessaloniki, Greece
2Laboratory of Physiology, Medical School, Aristotle University of Thessaloniki, Thessaloniki, Greece.
The electrical properties of the skin, also known as electrodermal activity (EDA), are considered as an indirect measure of autonomous nervous system. Along with that, the effects of noise-induced stress in intensive care units, is well explored. This study explores the noise-induced acute electrodermal activity changes in adult critical care patients and to compare these changes with cardiovascular effects of the same stress (noise) stimulus. Skin conductance variability, noise level, selected hemodynamic and respiratory parameters were monitored during 4 hour routine daytime intensive care nursing and treatment in an adult Intensive Care Unit. Average ambient noise levels during the time window (4 min) before the stimulation were 54.33(2.65) dB for Group A and 55.65(3.31) dB, while the noise stimulation was on average for Group A 70.8 (1.98) dB, and for Group B: 71.31(3.31) dB. EDA changes to noise stimulus were more distinct than hemodynamic and respiratory parameters. Yet, a weak relation was found between all EDA parameters and the particular noise level changes. Noise-induce stress causes more distinct EDA changes when measured immediately post stimulus. In addition, sedation level seems to affect the intensity of these changes. However, further studies are needed in to order to reach a definite conclusion.
It a more than 50 years, that noise has been identified as a potential source of allostatic load and consequently, stress1. Yet, and despite the fact that World Health Organization (WHO) recommends that the average background sound level in hospitals should not exceed 30dB2, the literature is full of evidence that noise in Intensive Care Units (ICU) typically exceeds that limit3-6. Moreover, the exposure to excessive noise level has a direct impact both on mortality and morbidity of the patients, and on the performance of cognitive tasks of critical care staff4-7.
There are several studies that record either cardiovascular or endocrine effects of noise-induced stress8-10. At the same time – albeit electrodermal activity (EDA) has long been used as stress monitor for several types of stimuli (such as painful and emotional) in perioperative enviroment11 – there is only one study of EDA changes as means of measuring ICU noise effect in healthy volunteers12.
The aim of the study is to measure acute EDA changes during noise stimulation in adult ICU patients and to compare these changes with cardiovascular effects of the same stress (noise) stimulus.
This prospective observational study was conducted at the adult general ICU, at AHEPA General University Hospital, Thessaloniki, Greece. Fifty four (54) measurements in critically ill patients under sedation, above 18 years old, were initially included in the study. Other inclusion criteria included administered mechanical ventilation>24h and constant sedation level under midazolam or propofol continuous intravenous infusion (c.i.v.). On the contrary, patients with Ramsay sedation score (RSS) 1, diagnosed or with history of hearing problems, psychiatric disorders, neurological diseases, neuro~ or myopathy, delirium, CNS or spinal cord injury, were excluded. Also as exclusion criteria were considered pregnancy, hemodynamic/respiratory instability, edema of the upper limbs (place of measurement) and the presence of sensitive electrical life-sustainable devices such as cardiac pace, renal replacement therapy devices, intra-abdominal aortal counterpulsion pump, extracorporal membrane oxygenation and artificial liver.
Skin conductance (SC) variability, noise level, selected hemodynamic and respiratory parameters were monitored during 4 hour routine daytime intensive care nursing and treatment. Measurements were divided into 2 categories according to sedation level: Group A- RSS 2-4 (na=10) and Group B –RAS 5-6 (nb =15). Another twenty nine (29) 4h-measurements were interrupted due to protocol violation or technical problems.
Med Storm Pain Monitor System (MED Storm® Innovation AS, Oslo, Norway) was used as SC monitor13. Three single use Ag/Cl electrodes were attached at the palmar surface of the hand: on the thenar eminence (current), on the hypothenar eminence (measurement) and just below 2nd and 3rd digits (reference). In order to minimize artifacts, the hand least likely to move, with no intravenous or intra-arterial lines was chosen.SC was measured by alternating current of 66Hz and an applied voltage of 50mV. SC parameters recorded were: absolute SC (in μS), peaks/sec or number of SC fluctuations per second (NSCF), the average peak (micro Siemens seconds – μSs), the rate of increase or decrease from the start to the end of the measurement window (rise time, in micro Siemens per second – μS/s), area huge peaks (μSs), area small peaks (μSs) and the larger of the two measures (referred as Area under curve- AUC, in μSs). Cut off for NSCF counting was >0.005, much more sensitive than the >0.02 μS used in relative pain monitoring literature11. Signal quality <80% was considered artifact and the measurement was also excluded.
Measurement window of interest was 5 sec and 5sec after sound stimulation, provided that a) 4min before and 1 min after the stimulus there was no other stimulus of any kind (i.e. pain) and that b) noise stimulus was minimum 10dB higher than the baseline recorded before and c) with minimum duration (about 2 sec). In order to ensure the observational character of the study, and to waive any possible ethical considerations, noise stimulation (referred as noise “event”) was product of the daily nursing/treatment routing inside ICU environment and not artificial deliberately-created noise stimulation. Only those noise “events” that were within the aforementioned frames, were included for further analysis (in total 52 for both groups). Noise level was measured at distance 30 cm from the head of the patient via Sound Level Meter GM13656 (Shenzhen Jumaoyuan Science & Technology® Co., China)14.
The rest of the parameters were monitored via Bedside Monitor BSM 9101K and Monitor CNS 9601 (Nihon Kohden® Ltd., Japan);and included: heart rate (HR), Systolic (SAP), diastolic (DAP) and mean arterial pressure (MAP), number of ventricular premature contractions (VPC), electocardiographic ST wave deviation in II lead (ST II) and respiratory rate (RR). Since the above were used in the literature8-10 as possible measures of stress, recordings were used as measure of comparison with SC parameters.
In selected measurements (Goup A: 9, Group B: 22) Bispectral index monitor (BIS) (Covidien®, USA) was also in place.
Data analysis was performed with MS Office Excel 2007 and Rstudio v.0.99.903. Descriptive statistics are presented as mean, standard deviation (SD). Two comparison designs were followed: one examined acute changes before/after the noise stimulus and one that examined the range of change between the 2 groups. Shapiro-Francia normality test is performed for the parameters of interest and then paired t-test or Wilcoxon signed ranked test is calculated. Results are presented as p value (Confidence Interval –CI). Statistical significance for p is set to p<0.05 and CI level at 95%. Finally, correlation between noise stimulus level and EDA changes were also investigated.
General characteristic of patients in each group of measurements is illustrated in Table 1. Different averages of APACHE II score, Extended Glasgow Outcome Score (GOSE) and PaO2/FiO2 are partially explain the different sedation level.
Table 1.General characteristics of the patients included finally in each group.
|Group Α||Group B||Group Α||Group B|
|N measurem||10||15||APACHE II||15.4(1.55)||19.6(1.66)|
|Sex||♂ =10,♀=0||♂ =9, ♀=6||SOFA||6.3(0.9)||7.9(0.4)|
|Weight (kg)||90.6(15,1)||89.95(12.6)||t (oC)||37.2(0.3)||37.1(0.4)|
|ΒMI( kg/m2)||28(1.65)||30.3(0.85)||PaO2/FiO2||294(69.3)||230 (81.8)|
Presented form: mean (SD), rounded to the nearest decimal.
During recording time, 17 noise “events” occurred in Group A and 35 in Group B that met inclusion criteria for further analysis. Average ambient noise levels during the time window (4 min) before the stimulation, the level of noise stimulation and the time window of the recording (1 min after) are displayed in table 2.
Table 2. Noise levels (dB) in the 2 groups during the period before and after the stimulus.
|Group A||Group B|
|After (during recording)
|After (during recording)
|Comparison before and during recording
|Comparison before and during recording
|0.945 [ -3.2,2.5]||0.72 [ -1.5,2.01]|
*Wilcoxon rank sum test with continuity correction
The changes caused from the stimulus in every group are illustrated in Tables 2 and 3 (See Supplement File Figure 1).
Table 3. Changes before and after noise stimulation in Group A (before/after).
|Group A (RSS 2-4), n=17|
|Area Huge Peak (μSs)||0.111(0.3)||0.557(0.806)||0.0011||[-0.8,-0.1]|
|Area Small Peak (μSs)||0||0.019(0.039)||0.0137||[-0.08,-0.01]|
|Average rise time||0.004(0.03)||0.022(0.003)||0.1005||[-0.06,0.005]|
|SC average (μS)||8.005(4.09)||8.07(4.14)||0.0093||[-0.14,-0.006]|
*Wilcoxon signed ranked test with continuity correction (paired),** CI 90%
*** CI 60, +selected patients (n=9)
Table 4. Changes before and after noise stimulation in Group B (before/after).
|Group B (RSS 5-6),n=35|
|Area Huge Peak (μSs)||0||0.036(0.069)||0.0018||[-0.12,-0.03]|
|Area Small Peak (μSs)||0||0.002(0.005)||0.08||NA|
|Average rise time||0||0.002(0.004)||0.001||NA|
|SC average (μS)||8.2(3.96)||8.13(9.95)||4.9e-5||[-0.009,-0.003]|
*Wilcoxon signed ranked test with continuity correction (paired),** CI 90%
*** CI 60%, +selected patients (n=22)
The percentile range of change (Δ%) for the majority of EDA changes in Group B are larger than in Group A, although sedation level is less in the first Group. Only SC average level change (Δ%) is +1.63% in group A and -0.85% in group B. For the rest of the EDA parameters Δ% is between 46-382% and for Group B (where applicable) > 900%. (Figure 1)
On the contrary the percentile range of change (Δ%) of the hemοdynamic parameters (HR, SAP,DAP,MAP) is between 0.9-1.7% for Group A and 0-1.89% for Group B, for BIS Δ% is -0.34 and 3.35% respectively and only RR(respiratory rate) Δ% in Group A is considerably large (+32%).
Finally, we explored possible relation between the change in noise stimulation and the change of the other parameters. However, a weak relation was found between all EDA parameters and the particular noise level changes. In addition, the relation is stronger in Group B than in Group A (Table 5, Supplement Figure 2 and 3.). The relation does not change significantly if certain clusters of ΔSPL are examined [ΔSPL∈(10,15dB) or ΔSPL∈(16,20dB) or ΔSPL>20dB].
Table 5. Correlation coefficient (τ) between range of noise stimulus (ΔSPL=noise level after-noise level before) and the different EDA parameters (AHP-area huge peaks, AsmP- area small peaks, AvRT- average rise time).
*p (two-tailed) <0.05, **p (two tailed) <0.01
There are several studies measuring noise levels in ICU environment3-7. Almost all of them report sound levels above 50bB with maximum peaks up to 101dB16-17. Its impact upon both staff and patients’ mental, emotional and physical health status and their ability for effective communication is negative. Sleep deprivation is probably the commonest result reported in patients. The latter can contribute to delirium, cognitive function impairment and to general morbidity and mortality. However, noise can also produce direct detrimental effects other than sleep disturbances, like cardiovascular and endocrine abnormalities8-10, 18-20. The major noise sources identified by the previous studies vary; yet staff conversations and alarms seem to be the more disturbing for ICU patiens19-20.
Electrodermal activity of the skin is based upon the electrical properties of the skin, which in their turn, are controlled by the autonomic nervous system (ANS). Thus; EDA monitoring is an indirect measuring of the activation of ANS, and therefore a good monitor mean for the effect of various stress-inducing factors (e.g. pain, emotional arousal, etc)21. There are several studies of EDA in perioperative environment22, yet very few of them are coming from ICU environment and none of them examines the effect of noise levels to EDA activity. Available reports are studies conducted mainly in psychology and psychiatry field23-26. Yet, due to diversity of the design, no assumption can be made regarding other types of populations.
The present study does reveal a relation, though weak, between short –time sound stimulus and acute EDA changes in sedated ICU patients. In comparison with others, “traditionally” physical measures of stress, like e.g. cardiovascular parameters, EDA seems to be more accurate. Interestingly, EDA changes are found to be more pronounced in patients under deeper sedation. The latter may be explained by the fact that lighter sedation levels allow better accommodation to / better perception of environmental conditions. Thus; an auditory stimulus is not perceived as stress –induced stimulus; and as a result, EDA changes are smaller.
However, those findings do not come without some compromises, the most important of which was the lack of similar literature. Other limitations: i) the physical characteristics of every sound stimulus was not controlled, so as to examine the effect of sound stimuli in real environment. The same sound level may have different frequencies, which may cause different effects. ii) Although the vast majority of the patients included in the present study were diagnosed with acute respiratory failure, beside exclusion criteria, every other diagnosis was included in the study. iii) Recording time was restricted to minimum so as to limit any other co-founding factors in a complex surrounding like ICU. As a consequence, no study was performed regarding possible adaptation to stimulus. iv)Finally, the authors examined EDA changes due to sound stimuli and not due to auditory- stimuli perceived as pain; hence, cut off for NSCF counting was lower than the limit used in most EDA studies in perioperative environment11.
Even with the above restrictions, the study highlights another perspective in the study of noise-induced stress in ICU. Future investigation, probably in more controlled samples, is needed in order to define the exact role of noise-induced stress in ANS activation. Comparison of EDA measurements with other stress monitoring means Saliva cortisol or heart rate variability have been already studied in other populations8, 10,27.
Noise-induce stress causes more distinct EDA changes when measured immediately post stimulus. Sedation level seems to affect the perception of the stimulus and thus, affect the EDA changes. However, further studies are needed in to order to reach a definite conclusion.
The authors wish to thank Dr. Maria-Giannakou Peftoulidou, director of the ICU and Prof. Dimitrios Vasilakos, director of the Department in which the study took place; and the medical and nursing staff of the unit for their assistance.
The study is part of a thesis project, approved by AHEPA General University Hospital Research Committee and by No 16/09-07-2013 General Assembly of Special Composition of Medical School, Aristotle University of Thessaloniki (Ref. No.8220/10-07-2013) and archived in National Archive of PhD Theses (NID:43587).
- Dieroff Hg. Audiometric stress tests in selection of workers in noisy factories. HNO 1959; 16(7):161-5.
- Falk SA, Woods NF. Hospital noise-levels and potentials health hazards. N Eng J Med. 1973; 289(15):774-81.
- Kam PC, Kam AC. Thompson JF. Noise pollution in anesthetic and intensive care environment. Anesthesia 1994;49(11):982-6.
- Tsara V. Nena E, Serasli E,. Noise levels in Greek hospitals. Noise Health 2008; 10(41):110-112.
- Konkani A, Oakley B. Noise in hospital intensive care units- a critical review of a critical topic. J Crit Care 2012;(27)5:522.e1-9
- Hu RF, Hegadoren KM, Wang XY, et al. An investigation of light and sound levels on intensive care units in China. Aust Crit Care 2016; 29(2):62-7.
- Filus W, lacerda AB, Albizu E. Ambient noise in Emergency Rooms and its Health Hazards. Int Arch Otorhinolaryngol 2015;19(3):205-9.
- Kirschbaum C. Hellhammer D. Noise and stress- salivary cortisol as non-invasive measure of allostatic load. Noise Health 1999; 1(4):57-65.
- Ising H, Babisch W, Kruppa B. Noise-induced endocrine effects and cardiovascular risk. Noise Health 1999;1 (3):37-48.
- Hall JS, Aisbett B, Tait JL, et al. The acute physiological stress response to an emergency alarm and mobilization during the day and at night. Health Noise 2016; 18(82):150-6.
- Aslanidis T. Electrodermal activity: Applications in perioperative care. Int J Med Res Health Sci 2014;3(3):687-95.
- Günther AC, Schandl AR, Berhardsson J, et all. Pain rather than induced emotions and ICU sound increases skin conductance variability in healthy volunteers. Acta Anaesthesiol Scand 2016; 60(8):1111-20.
- Med Storm Monitor User Manual v1.0 English MA001-25, Part Number 4001, Med Storm Innovation AS, Oslo,2010
- Sound Level Meter GM1356.[Internet] Available from : http://benetechco.com/en/products/sound-level-meter-gm1356.html.
- Johansson L, Bergom I, Waye Person K, et al. The sound environment in ICU patient room – A content analysis of sound levels and patients experiences. Crit Care Nurs. 2012; 28(5):269-79.
- Lawson N, Thompson K, Saunders G, et al. Sound intensity and noise evaluation in a critical care unit. Am J Crit Care. 2010; 19(6):88-98.
- Qutub HO, El-Said KF. Assessment of ambient noise levels in the intensive care unit of a university hospital. J Family Community Med. 2009;16(2):53-57
- Putz -Maidl C, MacAndrew SN, Leske JS. Noise in ICU: Sound levels can be harmful. Nurs Crit Care 2014;9(5):29-35
- Hiu X, Kang J, Mills SG. Clinical review: The impact of noise on patients’ sleep and the effectiveness of noise reduction strategies in intensive care units. Crit Care 2009;13:208.
- Wenham T, Pittard A. Intensive Care Unit Environment. BJA Education. 2009; 9(6):178-83..
- Bucsein W. Electrodermal Activity, 2nd ed. New York: Springer Science + Business Media. 2012.
- Aslanidis T. Electrodermal Activity: Applications in perioperative care. Int J Med Res Health Sci.2014;3(3):687-695.
- Rolton WT, Goodale J, Pfefferbaum A. Auditory event-related potentials and electrodermal activity in medicated and unmedicated schizophrenics. Biol Psychiatry 1991; 29(6):585-99.
- Koelsch S, Kilches S, Steinbeis N, et al. Effects of unexpected chords and of performer’s expression on brain responses and electrodermal activity. PLoS One 2008;3(7):e2631.
- Stevens S, Gruzelier J. Electrodermal activity to auditory stimuli in autistic, retarded, and normal children. J Autism Dev Disord. 1984;14(3):245-60.
- Shibagaki M, Yamanaka T. Attention of hyperactive preschool children–electrodermal activity during auditory Percept Mot Skills 1990;70(1):235-42.
- Green A, Jones AD, Sun K, Neitzel RL The Association between Noise, Cortisol and Heart Rate in a Small-Scale Gold Mining Community-A Pilot Study. Int J Environ Res Public Health. 2015;12(8):9952-66.
Authors Aslanidis Th., Grosomanidis V., +Karakoulas K., Chatzisotiriou A. have no conflicts of interest or financial ties to disclose.
Aslanidis Theodoros MD,
4 Doridos street, PC 54633, Thessaloniki, Greece,