EMPIRICAL STUDIES
12303
Tympanic ear thermometer assessment of body temperature
among patients with cognitive disturbances. An acceptable
and ethically desirable alternative?
Lena Aadal RN, MScN, PhD(Head of Clinical Nursing Research)1, Lisbet Fog RN, MR(Nursing Specialist)2 and
Asger Roer Pedersen PhD(Statistican)1
1
Research Unit, Hammel Neurorehabilitation and Research Centre, Hammel, Denmark and 2H1, Hammel Neurorehabilitation and Research
Centre, Hammel, Denmark
Scand J Caring Sci; 2016
Tympanic ear thermometer assessment of body
temperature among patients with cognitive
disturbances. An acceptable and ethically desirable
alternative?
Aim: Investigation of a possible relation between body
temperature measurements by the current generation of
tympanic ear and rectal thermometers.
Background: In Denmark, a national guideline recom-
mends the rectal measurement. Subsequently, the rec-
tal thermometers and tympanic ear devices are the
most frequently used and first choice in Danish hospi-
tal wards. Cognitive changes constitute challenges with
cooperating in rectal temperature assessments. With
regard to diagnosing, ethics, safety and the patients’
dignity, the tympanic ear thermometer might comprise
a desirable alternative to rectal noninvasive measure-
ment of body temperature during in-hospital-based
neurorehabilitation.
Design: A prospective, descriptive cohort study. Consecu-
tive inclusion of 27 patients. Linear regression models
Introduction
Measurement of body temperature is a frequent proce-
dure and accuracy of the readings is pivotal for diagnos-
ing, treating and monitoring patients (1), because core
temperature is essential information about health and
sickness (2). Rectal temperature has historically been
considered a reference measure, but it is time-consuming
and difficult to obtain in certain patients (3).
Correspondence to:
Lena Aadal, Head of Clinical Nursing Research, Hammel
Neurorehabilitation and Research Centre, Voldbyvej, 15 8450
Hammel, Denmark.
E-mail: lena.aadal@hammel.rm.dk
© 2016 Nordic College of Caring Science
doi: 10.1111/scs.
were used to analyse 284 simultaneous temperature
measurements.
Ethics: Ethical approval for this study was granted by the
Danish Data Protection Agency, and the study was com-
pleted in accordance with the Helsinki Declaration 2008.
Results: About 284 simultaneous rectal and ear tempera-
ture measurements on 27 patients were analysed. The
patient-wise variability of measured temperatures was
significantly higher for the ear measurements. Patient-
wise linear regressions for the 25 patients with at least
three pairs of simultaneous ear and rectal temperature
measurements showed large interpatient variability of
the association.
Conclusion: A linear relationship between the rectal body
temperature assessment and the temperature assessment
employing the tympanic thermometer is weak. Both
measuring methods reflect variance in temperature, but
ear measurements showed larger variation.
Keywords: brain injury, rehabilitation, measurement,
body temperature, infrared tympanic thermometer.
Submitted 26 May 2015, Accepted 18 September 2015
It is difficult for patients with cognitive impairments to
cooperate during the rectal measurement procedure
because it penetrates their intimacy zone and they do
not understand the purpose of the procedure (4). Hence,
they may experience lack of control causing anxiety (5,
6), and there is a risk of patient pain and mucosa lesions
(7, 8). Additionally, the increased stimulation and
unpleasant experiences may induce reactions such as
anxiety, violence and agitated behaviour (5). However,
the rectal measurement procedure is common among
patients with acquired brain injury because elevated tem-
peratures can be based on the response to exogenous
stimuli requiring pharmacological interventions but also
post-traumatic hyperthermia, which may occur when
hypothalamus is damaged (9).
1
2 L. Aadal et al.

From this perspective, using a rectal thermometer noninvasive core measurement of body temperature
poses diagnostic, ethical and safety problems if the proce- among patients with acquired brain injuries.
dure entails the need to restrain patients with cognitive
changes. To resolve this dilemma, the availability of clini-
cally validated noninvasive body temperature measure- Tympanic ear devices
ment methods is essential. The tympanic thermometer detects the radiation of infra-
red energy by the tympanic membrane and converts this
Background energy factor into a temperature reading of the arterial
blood in the carotid artery.
The hypothalamus plays a key role in the thermoregula-
Measuring the body temperature with an infrared tym-
tory function in humans and integrates thermal signals
panic thermometer has been debated because it has a rela-
received from sensors throughout the central nervous
tively low reproducibility compared with the rectal
system involving the spinal cord, midbrain, reticular for-
measurement (1, 9, 13, 15, 18). A number of studies have
mation and nervous vagus. Studies have shown that nor-
compared the various modes of thermometry (7, 11, 12,
mal temperature ranges between 36.2 and 37.5 °C (8)
17, 19–21). Taking the history of rectal temperature mea-
and undergoes a regular circadian fluctuation of 0.5–
surement as a reference measure into consideration, there
1.0 °C, where it is lowest late at night or early morning
are remarkably few studies of the accuracy and repeatabil-
and highest in the evening. Additionally, the normal
ity of infrared tympanic measurements with rectal ther-
body temperature shows interindividual variation (10)
mometry as reference in detecting and excluding fever (3,
that depends on weight, gender and age (2).
10, 16, 19, 22, 23). The studies of these early versions of
The measured body temperature depends on the type
the infrared tympanic thermometers emphasised risk of
of thermometer and the body site used for the measure-
misdiagnosis and questioned the accuracy of the measure-
ment (11). Shell and core temperature recordings are
ment, but the results were conflicting. None of the tested
used to measure body temperature in clinical practice.
thermometers were found to be appropriate alternatives to
The core temperature measurement requires invasive
the rectal thermometer (1, 16).
techniques in the oesophagus, bladder or pulmonary
In a recent study, Haugan et al. (16) found that the
artery (7, 12). Of these, the pulmonary artery catheter
infrared tympanic thermometer measured the temperature
measurement is considered the gold standard (8). How-
lower than rectal temperature and an indication of higher
ever, these measurement methods require advanced
variation in temperature measurements made in the ear.
measuring equipment which is unsuitable in clinical
However, this study used older versions of the tympanic
rehabilitation practice. Traditionally, the noninvasive rec-
thermometer released on the market in 2006 and 2007. In
tal site has been assumed to estimate the core tempera-
2012, a new and improved version, GeniusTM 2, was
ture (13, 14). Following this, axillary and oral readings
released by Covidien. In the sequential generations of
are adjusted to the rectal by adding 0.3 and 0.5 °C,
thermometers, improvements of geometry and algorithms
respectively. These adjustments are estimated on clinical
have been developed to ensure that the displayed result
experience, but the evidence for such relationships is not
reflects the tympanic temperature. Hence, the newest ver-
present (15). To the best of our knowledge, a more gen-
sion might meet the clinicians’ requested improvements of
eral linear relationship, indicating that a measured ear
repeatability in noninvasive temperature assessments.
temperature added a constant can estimate a temperature
In summary, cognitive disturbances are frequent
recorded in another site, has not been considered in the
among patients with acquired brain injury. Cognitive
literature.
changes constitute the challenges of cooperating in rectal
In Denmark, a national guideline recommends rectal
temperature assessments. With regard to diagnosing,
measurement. Consequently, rectal thermometers and
ethics and the patients’ safety, the tympanic ear ther-
tympanic ear devices are the most frequently used and
mometer might be a desirable alternative to rectal mea-
first choice in Danish hospital wards (3). Unlike this prac-
surement during in-hospital-based neurorehabilitation.
tice is noninvasive oral, axillary and tympanic ear mea-
The focus of this research study is to investigate a pos-
surement devices used in Norway for predicting core
sible relation between body temperature measurements
temperature. Here, the infrared tympanic thermometers
using the current generation of tympanic ear thermome-
are the most commonly used devices in clinical practice
ter and rectal temperature measurements.
(16). In relation to patients with cognitive disturbances,
the tympanic thermometer has been found acceptable in
Methods
predicting rectal temperatures in a population of elderly
who often demonstrate functional disturbances similar to The prospective, descriptive cohort study was carried out
patients with acquired brain injuries (3, 17). Hence, the in one of the wards at a neurorehabilitation hospital.
ear may provide an acceptable and ethically desirable Data were collected during 3 months.

© 2016 Nordic College of Caring Science

 Tympanic ear thermometer assessments 3

supplemented by surrogate informed consent from the

Participants
closest relative. Relatives were not allowed to give con-
Consecutive inclusion at admission to the ward used the sent for the patient, but their familiarity with the project
following criteria: was needed, in order to involve them in the interpreta-
tion of the patient’s expressions indicating violation in
relation to the intimate zones. Participation was
• Age ≥ 18 years.
voluntary, anonymity was preserved, and withdrawal
• Severe acquired brain injury despite aetiology.
was possible at any time without consequences for their
• Prescribed noninvasive assessments of body
care; no patient was excluded from the study. Ethical
temperature.
approval for this study was granted by the Danish Data
• Understanding Danish.
Protection Agency, and the study was completed in
• Informed consent from patient and closest relative.
accordance with the Helsinki Declaration (2008).
Exclusion criteria:
Permission to conduct the study was obtained from the
• Patient isolation.
rehabilitation hospital.
• Recent infection in the ear.
• Patients who refused rectal measurements verbally or
nonverbally, for instance by efforts to protect his or Material
her intimate zones by hands or garment.
Recorded data. All thermometers were labelled with the
patient’s name and used for all individual measurements.
Measurement As recommended by Coviden, the initial ear temperature
was measured in both ears and all later measurements
All data were measured by nurses, certificated health care
were conducted in the ear indicating the highest temper-
workers or nursing students who had accomplished a
ature. Temperatures were measured simultaneously using
training programme conducted by the investigators. The
tymphanic and rectal thermometers. Additional, data reg-
purpose of the programme was correct use and cleaning
istered were the patient’s age, sex, pulse rate and time of
of lenses at the tympanic infrared thermometers. Further-
assessment (2, 10, 22).
more, the staff learned how to fill out the data registra-
tion form. Subsequently, the investigators were available
Statistical analysis. Simultaneous ear and rectal tempera-
as supervisors. Tymphanic measurements were conducted
ture measurements were used to estimate a population
using GeniusTM 2 by Covidien, which meets the CEN and
level linear model for the prediction of rectal tempera-
ASTM technical standards in
0.1 °C (36.7–38.9). The
ture from measured ear temperature (linear mixed
GeniusTM 2 Peak Select System records and analyses up to
model with random subject specific coefficients; PROC
100 separate measurements and selects the most accurate
MIXED (SAS Institute Inc., Cary, NC, USA). Intraindi-
temperature in less than 2-seconds. Tympanic infrared
vidual standard deviations of both types of temperature
thermometry accesses the radiation of infrared energy by
measurements were compared using the Wilcoxon
the tympanic membrane to determine the body‘s temper-
signed rank test. For patients with at least three simul-
ature. In rectal measurements, Kivex Clinic blue rectal
taneous ear and rectal temperature measurements, indi-
thermometer was used. It has an accuracy of
0.1 °C. All
vidual linear models were estimated, and the estimated
temperatures were entered manually on the data registra-
slopes were compared between genders using the Wil-
tion sheet. Only a few minutes between infrared ther-
coxon rank sum test and tested for association with age
mometry and rectal recording was accepted. All
using Spearman’s rho.
thermometers were labelled and used for only one speci-
fic patient during the rehabilitation course. Temperature
measurements were conducted in accordance with the Results
medical prescription and the local clinical guideline, in In total, 290 rectal and 316 ear temperature measure-
order to collect data as similar to daily practice as ments were recorded for 27 patients. No simultaneous
possible. ear temperature measurement was obtained for 6 of the
290 rectal measurements. Hence, 284 simultaneous rectal
Ethics and ear temperature measurements from 27 patients
(nine women, 18 men, age range: 33–77 years, median
Rectal temperature measurements are prescribed as an
age: 62 years) were analysed (1–43 measurements pr.
ordinary intervention, and this study adds tympanic mea-
patient, median 8).
surement with a minor risk to violate the patient’s inti-
macy zone. Due to the patients’ vulnerability, informed The population model for converting measured ear tem-
verbal consent was obtained from the patient perature into a predicted rectal temperature was (Fig. 1)

© 2016 Nordic College of Caring Science

4 L. Aadal et al.

Figure 1 The 284 simultaneous rectal and ear temperature Figure 2 Patient-wise standard deviations of measured temperatures
measurements (dots) and the estimated population model for for the 25 patients with at least three pairs of simultaneous ear and
converting measured ear temperature into a predicted rectal rectal temperature measurements.
temperature (solid line). The grey area indicates the 95%-prediction
uncertainty associated with the prediction, and the dotted line is the
identity line (agreement between ear and rectal temperature). pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
37:3
1:96  0:0863 ¼ 37:3
0:58 ¼ ½36:7; 37:9

Predicted rectal temperature ¼ 37:3 The interpretation is that at measured ear temperature
of 37 °C, where the predicted rectal temperature equals
þ0:35  ðMeasured ear temperature  37Þ
37.3 °C, the 95% prediction interval [36.7; 37.9] contains
with 95%-prediction uncertainty limits given by
95% of all measured rectal temperatures. This prediction
uncertainty increases as the measured ear temperature

1:96
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi increases/decreases from 37 °C (Fig. 1).
 0:0863 þ 0:0233  ðMeasured ear temperature  37Þ Patient-wise standard deviations of measured tempera-
tures could be calculated for 25 patients, which had at
Hence, with measured ear temperature 37 °C, the least three pairs of simultaneous ear and rectal tempera-
predicted rectal temperature is 37.3 °C, and for each ture measurements. For the remaining two patients, only
increment/decrement in measured ear temperature of a single pair of temperature measurements was available.
1 °C above/below 37 °C, the predicted rectal tempera- The patient-wise variability of measured temperatures
ture increases/decreases by 0.35 °C. In particular, the was significantly higher for the ear measurements
difference between predicted rectal and measured ear (p = 0.0025, Fig. 2).
temperature depends on the measured ear temperature. Patient-wise linear regressions for the 25 patients with
At approximately 37.5 °C measured ear temperature, at least three pairs of simultaneous ear and rectal temper-
the predicted rectal temperature is the same. Con- ature measurements showed large interpatient variability
versely, the predicted rectal temperature is larger/smaller of the association (Table 1).
than the measured ear temperature for measured ear The association was found to be statistically significant
temperatures above/below 37.5 °C. This is due to the (slope coefficient different from zero) in only 6 of the 25
nonunity slope coefficient of 0.35 in the prediction cases, which in part can be explained by the small num-
equation (Fig. 1). Although the estimated slope coeffi- ber of measurements for many patients, although non-
cient is subject to estimation uncertainty (95%-confi- significant associations were also found for several
dence interval [0.25; 0.45]), it is, however, significantly patients with 10 or more measurements (6 of 9 patients,
different from unity (p < 0.0001) at the usual statistical 67%). The estimated slope coefficient varied considerably
significance level of 5%. Hence, the best fitting predic- between patients (range: [1.00; 0.91], median: 0.16),
tion line is not parallel to the identity line (dashed line indicating quite different temperature associations
in Fig. 1) in this patient population providing evidence between them, and was not associated with age (Spear-
that measured ear temperature cannot be converted into man’s q = 0.17, p = 0.4045) or gender (p = 0.6368).
a measured rectal temperature by adding a fixed tem- The 95% prediction uncertainties ranged 0.09–1.07 (me-
perature value. dian: 0.44), and the coefficient of determination, R2, ran-
The 95% prediction uncertainty also depends on the ged 0.002-0.70 (median: 0.15), which indicates weak
measured ear temperature. It is smallest at measured ear relationship between the two temperatures in many
temperature 37 °C, where the 95% prediction uncer- patients. By inspection of the individual linear regressions,
tainty limits are: different types of associations could be identified (Fig. 3).

© 2016 Nordic College of Caring Science

 Tympanic ear thermometer assessments 5

Table 1 Patient characteristics (gender, age) and individual estimated linear models for converting measured ear temperature into a predicted
rectal temperature for the 25 patients with at least three pairs of simultaneous ear and rectal temperature measurements (n). The rows are sorted
in descending order of the number of measurements. The estimated linear models are presented by the predicted rectal temperature (At 37° C)
at measured ear temperature 37° C, the slope (Slope) of the linear model with 95%-confidence interval and p-value for the null hypothesis of
zero slope, and strength of the linear relationship quantified by the 95%-prediction uncertainty (95%-PU) and the coefficient of determination
(R2). The coefficient of determination can be interpreted as the fraction of the variability in the measured rectal temperatures that can be
explained by variability in the measured ear temperatures through the linear model.

Patient ID Gender Age n At 37 °C Slope 95%-PU R2

15 Female 39 43 37.3 [37.1; 37.5] 0.20 [0.01; 0.42], p = 0.0654 0.86 0.08
12 Male 63 34 37.3 [37.2; 37.4] 0.52 [0.28; 0.75], p = 0.0001 0.48 0.39
3 Female 63 25 37.4 [37.2; 37.6] 0.10 [0.51; 0.31], p = 0.6279 0.55 0.01
32 Male 63 19 37.2 [37.1; 37.2] 0.07 [0.08; 0.21], p = 0.3285 0.22 0.06
18 Male 58 16 37.4 [37.2; 37.6] 0.69 [0.36; 1.02], p = 0.0005 0.35 0.59
36 Male 57 15 37.4 [37.2; 37.5] 0.63 [0.38; 0.88], p = 0.0001 0.45 0.70
28 Female 61 13 37.3 [37.1; 37.5] 0.19 [0.27; 0.65], p = 0.3787 0.46 0.07
24 Male 63 11 37.3 [37.1; 37.6] 0.16 [0.39; 0.72], p = 0.5225 0.40 0.05
35 Male 68 10 37.0 [36.7; 37.3] 0.19 [1.03; 0.65], p = 0.6152 0.46 0.03
22 Female 56 9 37.0 [36.9; 37.2] 0.80 [0.32; 1.27], p = 0.0055 0.25 0.69
31 Male 66 9 37.0 [36.6; 37.4] 0.48 [0.06; 0.90], p = 0.0302 0.33 0.51
1 Male 33 8 37.4 [37.0; 37.7] 0.27 [0.46; 1.00], p = 0.3966 0.55 0.12
9 Male 69 8 37.1 [36.8; 37.4] 0.09 [0.34; 0.52], p = 0.6116 0.44 0.05
21 Male 49 8 36.7 [36.3; 37.2] 0.79 [0.15; 1.42], p = 0.0234 0.47 0.60
11 Male 77 7 37.2 [36.6; 37.8] 0.68 [0.37; 1.72], p = 0.1565 0.93 0.36
17 Male 65 7 36.8 [36.7; 36.9] 0.05 [0.05; 0.16], p = 0.2533 0.09 0.25
2 Female 68 6 37.4 [36.8; 37.9] 0.64 [0.84; 2.13], p = 0.2943 0.95 0.27
7 Female 39 5 37.7 [37.2; 38.2] 0.91 [0.24; 2.06], p = 0.0857 0.66 0.68
10 Female 61 5 37.1 [36.3; 37.9] 0.21 [1.78; 1.35], p = 0.6925 0.36 0.06

For some patients (Fig. 3a,b), the difference between
the two temperatures appeared constant (independent of
the measured ear temperature), although with different
signs (Fig. 3a: rectal > ear; Fig. 3b: rectal < ear), whereas
for others (Fig. 3c,d), the measured rectal temperature
remained fairly constant, while the measured ear temper-
ature varied 1.5–2 °C. For some (Fig. 3e,f), the individual
temperature relationship resembled the one found on the
population level (Fig. 1). Hence, for many patients, the
population-based prediction model (Fig. 1) will be biased.
Discussion
The performance of the tympanic measurements in this
study may reflect a number of factors. Our results must
be treated with caution because of the small number of
patients, none of whom were highly febrile. The mea-
surement of body temperature was conducted by tym-
panic ear and rectal thermometers. Both of these
methods are associated with uncertain accuracy influ-
enced by factors such as lack of training or poor equip-
ment handling, and correct and observer-independent
use of infrared thermometry can be challenging in a clin-
ical setting (22), which may explain the observed higher
variability in ear temperature measurements. Many vari-
ables involved in the use of tympanic thermometers com-
bine to produce measurements that may be too
unreliable for a neuroscience ICU (24). Data are collected
in clinical practice and may be influenced by incorrect
use, incorrect registration or the soiled lenses of the ther-
mometers. These critical factors need to be addressed in
educating staff and critique of studies using specific
assessment devices.
Although the rectal measurement is considered the
reference standard in this study, we acknowledge that it
may be imperfect in many ways in our context. Inva-
sive temperature measurement is neither suitable for
routine clinical practice, nor easily applicable in patients
with cognitive disturbances. Measurements from differ-
ent areas of the body may provide different results and
should not be directly compared, even if taken at the
same time (22). This study showed a weak linear rela-
tionship between body temperature by rectal assessment
and temperature assessment applying the tympanic ther-
mometer both on individual and group level. This is
consistent with other studies which found no evidence
for adjusting one site to another (15) and large varia-
tions between sites due to intraindividual variation.
(15). It is claimed that ear temperature is on average
0.2–1.3 degrees below rectal temperature (25). However,
we found high interpersonal variation in the correlation
between sites. In one case (Fig. 2a), the rectal measure-
ments were higher than the tympanic, while these were
lower in another case (Fig. 2b). A review showed that

© 2016 Nordic College of Caring Science

6 L. Aadal et al.

(a) (b)

(c) (d)

(e) (f)

Figure 3 Different types of associations identified between measured ear and rectal temperatures. (a & b): Similar difference between the two
temperatures at all measured ear temperatures although with different signs (A: rectal>ear, B: rectal<ear). (c & d): Almost constant rectal
temperatures associated with ear temperatures varying 1.5-2° C. (e & f): Individual temperature relationships similar to that found on the
population level (Fig. 1).

infrared tympanic thermometry had low sensitivity, but
high specificity in detecting and excluding fever compared
with rectal measurements (22). Hence, simultaneous
baseline measurements are required in clinical practice, in
order to identify temperature incensement of diagnostic
relevance when different sites are needed. Taking the
diversity of our findings into consideration, the need for
accurate and continuous temperature screenings in a
neurorehabilitation setting is challenged. Despite mea-
surement site prescriptions or care, it may be recom-
mended not to base decisions on a single vital sign, but a
holistic picture of the patient’s clinical condition.
Viewed independently, both measuring methods reflect
variance in temperature, but ear measurements showed
larger variation. Large variations are found between sites
due to intraindividual variation (15). The variation can

© 2016 Nordic College of Caring Science

 Tympanic ear thermometer assessments 7

partly be explained by the tympanic measurements instan-
taneously responding to changes in the set point tempera-
ture (15), which is considered to reflect the core
temperature (7). Due to the fact that tympanic membrane
and the hypothalamus share blood supply, the tympanic
measurements reflect the temperature of the arterial blood
affecting the thermoregulatory centres (8). The rectal tem-
perature is more stable because of its position far from the
hypothalamus in a sealed area with low emission of heat.
It lags behind changes at other sites especially during rapid
temperature changes for at least 20 minutes (2).
Conclusion
A linear relationship between the rectal body tempera-
ture assessment and the temperature assessment employ-
ing the tympanic thermometer is weak. Hence, there is a
substantial uncertainty in estimating and predicting tem-
peratures between the two sites in a population, but also
among individuals. Both instruments reflect the core
temperature, but the variation in ear measurements was
higher.
Relevance to clinical practice
The tympanic thermometer appears to be an immediate
alternative to the rectal thermometer method among
patients with cognitive disturbances, but the assessments
need to be supplemented by other vital data in detecting
and excluding temperature variation. Given the wide-
spread use of infrared tympanic thermometer, further
documentation of diagnostic accuracy and repeatability of
newer models used in a clinical setting is needed.
Author contributions
Aadal contributed to study conception/design; data analy-
sis; drafting of manuscript; critical revisions for important
intellectual content; and supervision. Fog contributed to
study conception/design, data collection and critical revi-
sion of manuscript. Roer Pedersen contributed to study
conception/design; data analysis; statistical expertise; and
critical revisions for important intellectual content.
Funding
Sources of grants, equipment or drugs was not received.
Ethical approval
Ethical approval for this study was granted by the Danish
Data Protection Agency, and the study was completed in
accordance with the Helsinki Declaration 2008.

References
1 Sund-Levander M, Grodzinsky E.
Time for a change to assess and evalu-
ate body temperature in clinical prac-
tice. Int J Nurs Pract 2009; 15: 241–9.
2 Sund-Levander M, Grodzinsky E.
Assessment of body temperature
measurement options. Br J Nurs
2013; 22: 880, 882-8.
3 Smitz S, Van de Winckel A, Smitz
MF. Reliability of infrared ear ther-
mometry in the prediction of rectal
temperature in older inpatients. J
Clin Nurs 2009; 18: 451–6.
4 Romsland GI. Kognitiv svikt og hjer-
nens rolle. Kognitiv svikt (Cognitive
defecits and the role of the brain.
Cognitive defecits), 1st edn. 2011,
Gyldendal Akademisk, Oslo.
5 Becker C. Nursing care of the brain
injury patient on a locked neurobe-
havioral unit. Rehabil Nurs 2012; 37:
171–5.
6 Romsland GI, Grimen H, Ingstad B.
Experiences of crisis after acquired
brain injury. An ethnographic study.
V
ard i Norden 2012; 32: 9.
7 Gasim GI, Musa IR, Abdien MT, Adam
I. Accuracy of tympanic temperature
measurement using an infrared tym-
panic membrane thermometer. BMC
Res Notes 2013; 6: 194.
8 Costa EQ, Guedes J, Baptista JS.
Core body temperature evaluation:
suitability of measurement proce-
dures. Vienne 2012; 3(Appendix 1 to
number 3): 33–44.
9 McIlvoy L. Fever management in
patients with brain injury. AACN Adv
Crit Care 2012; 23: 204–11.
10 Holst I. Temperaturregulering (Tem-
perature adjustment). In Sygeplejebo-
gen (Book of Nursing), 1st edn
(Pedersen S ed.), 2009, 3 printing
Gads Forlag ed., Blackwell Publishing
Asia Pty Ltd, Copenhagen, Denmark,
Chap 7.
11 Rubia-Rubia J, Arias A, Sierra A,
Aguirre-Jaime A. Measurement of
body temperature in adult patients:
comparative study of accuracy, relia-
bility and validity of different devices.
Int J Nurs Stud 2011; 48: 872–80.
12 Moran JL, Peter JV, Solomon PJ,
Grealy B, Smith T, Ashforth W,
Wake M, Peake SL, Peisach AR.
Tympanic temperature measure-
ments: are they reliable in the criti-
cally ill? A clinical study of measures
of agreement. Crit Care Med 2007; 35:
155–64.
13 Sund-Levander M, Grodzinsky E,
Loyd D, Wahren LK. Errors in body
temperature assessment related to
individual variation, measuring tech-
nique and equipment. Int J Nurs
Pract 2004; 10: 216–23.
14 Edling L, Carlsson R, Magnuson A,
Holmberg H. Rectal thermometers
still best for body temperature mea-
surement. Comparative study of
tympanic and oral thermometers.
L€akartidningen 2009; 106: 2680–3.
15 Sund-Levander M, Grdzinsky E. Ac-
curacy when assessing and evaluat-
ing body temperature in clinical
practice: time for a change?. Ter-
mologi international, European Associa-
tion of Thermology (EAT) 2012; 22: 25.
16 Haugan B, Langerud AK, Kalvoy H,
Froslie KF, Riise E, Kapstad H. Can
we trust the new generation of infra-
red tympanic thermometers in

© 2016 Nordic College of Caring Science

8 L. Aadal et al.
clinical practice? J Clin Nurs 2013;
22: 698–709.
17 Onur OE, Guneysel O, Akoglu H,
Aydin YD, Denizbasi A. Oral, axil-
lary, and tympanic temperature mea-
surements in older and younger
adults with or without fever. Eur J
Emerg Med 2008; 15: 334–7.
18 Duberg T, Lundholm C, Holmberg H.
€
Orontermometer inte fullgott alterna-
tiv till rektaltermometer. J€amf€ orande
studie visar stora avvikelser i temper-
atur mellan o €ronm€atningar. (Ear
thermometers are not a satisfactory
alternative to rectal thermometers.
The present study shows large devia-
tion between ear measurements).
L€akartidningen 2007; 104: 1479–82.
19 Dzarr AA, Kamal M, Baba AA. A com-
parison between infrared tympanic
thermometry, oral and axilla with rec-
tal thermometry in neutropenic adults.
Eur J Oncol Nurs 2009; 13: 250–4.
20 Mangat J, Standley T, Prevost A,
Vasconcelos J, White P. A compar-
ison of technologies used for estima-
tion of body temperature. Physiol
Meas 2010; 31: 1105–18.
21 Sener S, Karcioglu O, Eken C, Yaylaci
S, Ozsarac M. Agreement between
axillary, tympanic, and mid-forehead
body temperature measurements in
adult emergency department patients.
Eur J Emerg Med 2012; 19: 252–6.
22 Holte TO, Vandvik PO, Elvsaas IØ,
Norderhaug IN. The diagnostic accu-
racy of infrared tympanic, oral,
axillary and temporal thermometry,
compared with rectal readings when
identifying fever in adult hospitalized
patients. 2014, The Norwegian Kno
wledge Centre for the Health Services,
http://www.kunnskapssenteret.no/en/
publications/the-diagnostic-accuracy-
of-infrared-tympanic-oral-axillary-
and-temporal-thermometry-compa
red-with-rectal-readings-when-iden-
tifying-fever-in-adult-hospitalized-pa-
tients (last accessed 26 May 2015).
23 Edling L, Carlsson R, Magnuson A,
Holmberg H. Rektaltermometern fort-
farande b€ast f€ or temperaturm€atning.
J€amf€orande studie med o € ron- och
muntermometrar. (Rectal thermome-
ters are continuously the best for tem-
perature measurement. Sustaining
study with ear- and oraltermometry.
L€akartidningen 2009; 106: 2680–3.
24 Dunleavy KJ. Which core body tem-
perature measurement method is most
accurate? Nursing 2010; 40: 18–9.
25 Sund-Levander M, Forsberg C, Wah-
ren LK. Normal oral, rectal, tym-
panic and axillary body temperature
in adult men and women: a system-
atic literature review. Scand J Caring
Sci 2002; 16: 122–8.
© 2016 Nordic College of Caring Science