Learning Center
Plans & pricing Sign in
Sign Out

Comparison of Temperature Measurement Devices in Post


									     Comparison of Temperature Measurement
       Devices in Post Anesthesia Patients
                            Georgita Tolbert Washington, RN, MSN, CCRN, CCNS,
                                     Joetta Lynn Matney, RN, BSN, MBA

                     A descriptive correlational study was used to evaluate the correla-
                     tion/agreement of oral and axillary temperature measurements to
                     patient core temperatures obtained in the OR. Data collectors
                     recorded oral or axillary patient temperature from 752 patients on
                     admission and discharge from the PACU. Results indicated that
                     there was a moderate correlation between each of the current
                     devices and core temperature, but no agreement between core tem-
                     perature and either device. Recommendations were made to use
                     just one device throughout the organization, or to use the device
                     used on admission throughout the hospitalization.

                     Keywords: perianesthesia care, temperature measurement, re-
                     search, oral thermometry, axillary thermometry, core temperature.
                          Ó 2008 by American Society of PeriAnesthesia Nurses.

MEASURING patient body temperature is                                  temperature. Along with other clinical data,
a part of routine nursing care and is impor-                           body temperature is also used to determine
tant to the detection and treatment of hypo-                           the presence of an illness and as a means of
thermia and hyperthermia in all patients. In                           evaluating patient responses to treatments1,2;
most cases, this basic but important piece                             it is also useful in detecting immunological
of data influences clinical decisions and, be-                          reactions and hypermetabolic states.3 Body
cause of that influence, every effort should                            temperature can also uncover noninfectious
be made to ensure the accuracy of that                                 causes of hyperthermia such as drugs, tu-
                                                                       mors, blood products, and biosynthetic
   Georgita Tolbert Washington, RN, MSN, CCRN, CCNS, is
Director of Education at Johnson City Medical Center and
Chair of the Research Council for Mountain States Health
                                                                       Patients in the PACU are monitored to ensure
Alliance, Johnson City, TN; and Joetta Lynn Matney, RN, BSN,           that as they wake from anesthesia, their body’s
MBA, is Resource Staff Nurse at Mountain States Health Alli-           normal temperature control mechanisms are
ance, Johnson City, TN.
   We have received no benefits in any form from any commer-
                                                                       intact and functioning. The purpose of this
cial party that is either directly or indirectly related to the sub-   study was to compare the oral or axillary tem-
ject of this article. The thermometers investigated in this study      peratures obtained from postsurgical patients
were devices already in use at our facilities.                         using the 3M Tempa-DotÔ Single-Use chemi-
   Address correspondence to Georgita Tolbert Washington,
RN, MSN, CCRN, CCNS, Mountain States Health Alliance, 400              cal thermometer (3M, St Paul, MN), and the
North State of Franklin, Johnson City, TN 37604; e-mail                Alaris TURBO)TEMPÔ (Alaris/Cardinal, Dub-
address:                                        lin, OH) electronic thermometer to determine
   Ó 2008 by American Society of PeriAnesthesia Nurses.
                                                                       which compared more favorably with the core
   doi:10.1016/j.jopan.2007.10.001                                     temperature.

36                                                                     Journal of PeriAnesthesia Nursing, Vol 23, No 1 (February), 2008; pp 36-48
TEMPERATURE MEASUREMENT                                                                           37

Background                                          quences and complications of hypothermia
                                                    can be costly to health care institutions. The
Temperature measurement is important to the         patient length of stay in the PACU is increased,
care and management of the perioperative/           the flow of patients from the holding area
perianesthesia patient. Hypothermia places          through to the PACU is hindered, and these
these patients at increased risk for coagulation    complications can also increase hospital
and cardiac problems, and postoperative in-         length of stay.2,7 The adverse affects of hypo-
fections.4-6 Because of heat loss during anes-      thermia can also result in the increased use
thesia, the human body tends to acclimate to        of blood products and mechanical ventilation,
environmental temperature; therefore, air-          posing significantly increased physiological
conditioned ORs are likely to induce hypother-      risks to the patient. Hypothermia and its
mia in surgical patients. Surgical exposure,        cumulative adverse outcomes are estimated
infusion of room temperature intravenous            to cost health care institutions an extra $2,500
fluids, certain drugs, and low ambient temper-       to $7,000 per patient for each 1.5 C (2.7 F)
ature also contribute to the risk of hypother-      decrease in temperature.2,5,11
mia. Adding to this risk is the fact that general
anesthesia and paralytics do not allow the          Our health care system provides two devices
patient’s normal heat-preserving mechanisms         to measure patients’ body temperature; the
to be activated. These mechanisms include           Alaris TURBO)TEMPÔ (Alaris/Cardinal) elec-
increased metabolic rate and shivering.5,7          tronic thermometer and the 3M Tempa-DotÔ
                                                    Single-Use chemical clinical thermometer
Accuracy and consistency are also essential to      (3M). Although patient care staff in various
the measurement of body temperature. Darm4          clinical areas anecdotally noted discrepancies
and Cory8 estimated that 60% to 80% of all          of one to two degrees between the two
patients arriving in the PACU are hypothermic       devices, the PACU RNs across the system were
and are subject to the many adverse affects         the most vocal because the patient flow
of hypothermia, including the discomfort of         through the PACUs was dependent on patients
shivering and chilling. Hypothermia can also        reaching a body temperature of 97 F. The
increase cardiac irritability, alter cardiac out-   differences led them to wonder if they were
put, and cause myocardial depression, leading       discharging hypothermic patients or keeping
to several other complications.4,9                  patients in the PACU who had in fact reached
                                                    97 F (36.1 C). The nurses wanted to know
When the body is cold, its immunologic activ-       which thermometer to use to safely discharge
ity is decreased, thereby increasing the poten-     patients back to their rooms.
tial risk of infection.2,4 Hypothermia decreases
platelet function and coagulation factors,          In keeping with our shared governance
thereby increasing the risk of bleeding. De-        model, the PACU RNs brought their concerns
creased wound healing occurs as a result of         and observations to the patient care practice
immunological suppression and increased             council. Collaboration with the research coun-
protein breakdown, and urinary nitrogen loss        cil resulted in a research study to investigate
in hypothermia.8,10 Hypothermia also results        the discrepancies in temperature measure-
in altered drug metabolism, which can lengthen      ment devices. Nurse Managers of the PACUs
the duration of some anesthetic agents’ seda-       recruited RN team members to participate in
tive effects.7,9                                    the study by acting as data collectors. The re-
                                                    search team consisted of the principle investi-
Hypothermia also has some indirect conse-           gator who was an RN, 13 RN data collectors,
quences, and the management of these conse-         and one RN co-investigator.
38                                                                    WASHINGTON AND MATNEY

Theoretical Framework                                for accuracy and they require meticulous
                                                     cleaning.13 Glass mercury thermometers are
Thermometry is the science and practice of           also extremely hazardous to patients and the
measuring the level of thermal energy, known         environment.3 Tympanic thermometry can
as temperature.12 The term temperature is            be used as an indication of core temperature
used to describe the measurement of energy           because the tympanic membrane is perfused
flow in the form of heat from one entity to an-       by the internal carotid artery, which supplies
other. Heat transfer is a dynamic process mea-       the temperature control center in the hypo-
sured with thermometers.3 Assessing body             thalamus. Temporal artery thermometry is
temperature is important, and the thermome-          used to reflect core temperature because the
ters used must be accurate, reliable, linear,        temporal artery arises from the carotid ar-
and hold calibration. If used properly, they         tery.14 Chemical thermometers contain heat-
should not cause the patient discomfort or           sensitive chemicals that change color when
inconvenience.3                                      exposed to heat. They change in intervals of
                                                     0.1 C, with each dot having a different level
Normal body temperature is a balance be-             of sensitivity. The temperature is recorded
tween heat production and heat loss.13 The           according to the last dot to change color.16
temperature of deep central body tissue is           Although these sites are used as alternatives
referred to as ‘‘core temperature,’’ and is stable   to core temperature, there is a lack of evi-
at 37.0 C (98.6 F) 6 0.6 C (1.08 F). Although    dence as to their accuracy as an alternative.14
core temperature can be maintained in this
narrow range, surface temperatures cannot.           Thermometry is also affected by the site of the
Surface temperatures change with the environ-        temperature measurement because body re-
ment, which can range as much as 12.8 C to          gions are perfused differently. Peripheral tem-
60 C (55-140 F). Core temperature is more          peratures are obtained orally, rectally, and via
consistent and accurate, whereas peripheral          the axilla. Areas exposed to the ambient tem-
temperature influenced by the environment             perature will be more likely to take on that
can be unstable and less reliable.14                 surrounding temperature. Physiological con-
                                                     ditions such as infections and its responses,
Although core body temperature may be mea-           and hemodynamic status may also affect tem-
sured at different places in the body under          perature readings in different parts of the
different circumstances,3 the thermistor on          body.3
the pulmonary artery catheter is considered
to be the gold standard for measuring core           Several research studies have demonstrated
body temperature.1,15 Other anatomical loca-         that females, the very young, and the very
tions include the distal third of the esophagus      old are more vulnerable to developing hypo-
using an indwelling probe, and the tympanic          thermia.4,8,9,17 Others have indicated that the
membrane’s anterior inferior quadrant using          types and methods of anesthesia contribute
infrared ear thermometers. Patient body tem-         to the incidence of hypothermia.7,9,10,18-20
peratures obtained with indwelling rectal            The type and length of surgery, and the use
and bladder thermistors are also considered          of preoperative, intraoperative, and postoper-
core temperatures.13                                 ative warming have also been shown to affect
                                                     the development and degree of hypothermia
Thermometry is accomplished using several            in postsurgical patients.8,18
other types of thermometers to measure
patient peripheral temperature. Glass ther-          For measurement accuracy in thermometry,
mometers using mercury are still in use even         devices must be calibrated and that calibration
though they require longer time to equilibrate       maintained. The user must be sure that the
TEMPERATURE MEASUREMENT                                                                            39

thermometer is giving an accurate reflection          when compared with tympanic thermome-
of the patient’s true body temperature. Any          ters. Potter and associates26 also concluded
deviation from the standard reference will           that an electronic device should be used to val-
result in erroneous temperatures,8 which may         idate the chemical device in orally intubated
be used to make critical clinical decisions. It is   patients. Compared with the pulmonary
also imperative that whichever device is used,       artery readings, however, Farnell and associ-
it is used correctly.14                              ates18 concluded that tympanic and chemical
                                                     readings were associated with erroneous mea-
Review of the Literature
There are many published research studies            Van den Bruel et al25 compared a chemical
that compare different methods of obtaining          thermometer to a mercury thermometer in
body temperature in different patient popula-        oral and axillary temperatures and found the
tions, including perioperative/perianesthesia        Tempa Dot (3M) to be reliable. Erickson
patients. These studies also compared temper-        et al28 determined that chemical thermome-
atures obtained by different routes, the differ-     ters are useful for screening or when infection
ent methods of thermometry to each other,            and safety are issues. They also recommend
and the different methods to core tempera-           that electronic or mercury thermometers be
ture.1,4,15,16,21-28                                 used for confirmation of temperature when
                                                     critical clinical decisions are needed.
Hooper and Andrews14 conducted an integra-
tive review of studies that compared invasive        Jensen et al23 stated in their findings that when
and noninvasive temperature measurement              comparing digital tympanic, oral, axillary, and
methods. The articles reviewed included              rectal thermometers to rectal mercury temper-
oral, tympanic, temporal, and oral/tympanic          atures, electronic rectal temperatures are the
comparison studies. Their review determined          most accurate reflection of core temperature.
that oral temperatures were the most accurate        Fisk and Arcona15 compared tympanic and pul-
assessment of temperature in an adult, acutely       monary artery temperatures and concluded
ill patient population. Analysis of 23 articles      that tympanic thermometry was not an accu-
meeting inclusion criteria determined that oral      rate representation of core temperature.
measurements taken in the right or left buccal
pocket is an accurate indication of core tem-        This review of the research literature reports
perature. In the studies reviewed, all tempera-      varying conclusions related to the site, type
tures were obtained with an electronic digital       of device used, and which temperatures mea-
thermometer. Oral electronic temperatures            surement sites were compared. All devices
were also found to be a reliable replacement         demonstrated some statistically significant dif-
for core temperatures in the conclusion of           ferences. Each facility should review the clini-
a study by Giuliano et al comparing pulmonary        cal significance between devices and decide
artery, tympanic, and oral methods.1                 what is best for patients in their facility.

Potter and associates26 studied 85 adult pa-
                                                     Research Aims
tients, comparing chemical dot and electronic
thermometers, and Farnell and associates18 ex-       Nurses throughout our system, especially in
amined 160 patient temperatures, comparing           our PACUs, observed discrepancies in pa-
chemical and tympanic with the pulmonary             tient temperatures when obtained with the
artery temperature, and determined that the          Alaris TURBO)TEMPÔ electronic thermometer
chemical thermometer was useful, accurate,           (Alaris/Cardinal), and the 3M Tempa-DotÔ
and reliable for measuring body temperature          Single-Use chemical clinical thermometer (3M).
40                                                                           WASHINGTON AND MATNEY

The electronic thermometer was anecdotally               Methods
observed to typically measure lower than the             Subjects and Setting
chemical thermometer. Therefore, when pa-
                                                         After Institutional Review Board (IRB) ap-
tients’ temperatures were obtained with the
                                                         proval, this study took place in the PACUs of
electronic thermometer, it took longer for their
                                                         a 1,999-bed health care system. Thirteen
temperatures to reach 36.1 C (97 F) in the
                                                         PACU nurses were recruited by their clinical
PACU. This led to increased PACU length of
                                                         managers and, after completing the required
stay, which also delayed admissions from the
                                                         human subject research training, served as
surgical suites, thereby impacting the surgery
                                                         data collectors for this convenience sample.
schedule. This also affected critically ill pa-
                                                         They recorded the temperatures of 752 physi-
tients admitted to intensive care units directly
                                                         ologically stable patients admitted to four
from surgery, who may have been cold as a re-
                                                         PACUs. Using power analysis, this sample size
sult of surgery or whose primary condition in-
                                                         provided a large effect .0.8.29 Core tempera-
cluded hypothermia.
                                                         tures were measured in the OR by the anesthe-
The research aim of this study was to deter-             siologist using an esophageal temperature
mine if there was a relationship between                 probe. These probes were also placed by the
the temperatures obtained with the Alaris                anesthesiologist and their locations were out-
TURBO)TEMPÔelectronic thermometer (Alaris/               side the span of control of these investigators.
Cardinal), the 3M Tempa-DotÔ Single-Use
chemical clinical thermometer (3M), and the              All postsurgical patients admitted to the PACU
last OR core temperature taken. We also                  from the surgical suites between November
wanted to determine if there was sufficient               2004 and January 2006 were included in the
agreement between each device and the last               study. They were patients of all ages, anesthe-
OR core temperature to allow the devices to              sia types, and undergoing various surgical pro-
be used interchangeably.                                 cedures (Table 1). Excluded were patients

                                 Table 1. Descriptive Statistics (n 5 752)

       Descriptive Variable                 Result (%)             Descriptive Variable         Result (%)

Age                                                               Surgical procedure
 Birth–11 mo                                    0.4                 Abdominal                      53.1
 12 mo–3 y                                      0.3                 Cardiovascular                  5.0
 4 y–6 y                                        0.1                 Neurological                    7.0
 7 y–11 y                                       0.4                 Orthopedic                     11.4
 12 y–18 y                                      0.9                 Genitourinary                  15.1
 19 y–29 y                                      7.8                 Oral                            0.3
 30 y–55 y                                     54.3                 Ear-nose-throat                 3.1
 56 y–88 y                                     34.9                 Plastics                        1.1
 $89 y                                          0.4                 Thoracic                        1.6
Gender                                                              Chest/mastectomy                1.2
 Male                                          25                   Endocrine                       0.1
 Female                                        75                   Excision lipoma                 0.1
Warming                                                             Anesthesia
 Preop warming (warm blanket)                  32                   General                        93.8
 Postop warming (warm blanket)                 93                   Spinal                          4.5
                                                                    Epidural                        0.3
                                                                    Block                           0.1
                                                                    Sedation only                   0.3
TEMPERATURE MEASUREMENT                                                                          41

who were admitted from the OR directly to an       after all the final revisions had been completed.
intensive care unit. Patients were admitted to     Once a month, the PI randomly either ob-
two of the PACUs after general surgical proce-     served each data collector or verbally assessed
dures and to the other two PACUs after primar-     proper instrument use to ensure continued
ily obstetrical and gynecological procedures.      correct use of each device. The observations
The IRB allowed patient consent to be waived       made were objective; each RN data collector
because there was no delay in patient assign-      had to read the digital numbers from the elec-
ments or any routine treatments, and there was     tronic thermometer, and the dots on the chem-
no deviation from any standards of care for        ical thermometer. Each then only recorded
these patients. If an abnormal temperature         their own patients’ temperatures on the collec-
was obtained, the patients received treatment      tion tool. During the study, the Alaris TUR-
according to PACU protocol. When patients          BO)TEMPÔ thermometer (Alaris/Cardinal)
met PACU discharge criteria, their role in the     was calibrated according to recommenda-
study was complete and no further data were        tions, and the 3M Tempa-DotÔ Single-Use
collected.                                         chemical clinical thermometers (3M) were
                                                   checked periodically for expiration dates.
Each site had adequate supplies of both the        Procedure
Alaris TURBO)TEMPÔ thermometer (Alaris/            An important factor in this study or any
Cardinal) with probe covers, and the 3M            study that involves the use of equipment is
Tempa-DotÔ Single-Use chemical clinical ther-      the assurance that the equipment is func-
mometers (3M). Immediately before the study,       tioning properly and that it is used correctly.
each electronic thermometer had its calibra-       The Alaris TURBO)TEMPÔ thermometers
tion verified by the engineering department         (Alaris/Cardinal) were calibrated by the bio-
according to manufacturers’ recommenda-            medical engineer immediately before the
tions. Both the chemical and electronic device     study and periodically throughout the study
probe covers were replenished from the sys-        according to the manufacturer’s recommen-
tem supply department when needed.                 dations. The 3M Tempa-DotÔ Single-Use
                                                   chemical clinical thermometers (3M) were
The data collection tool was developed from        checked for expiration dates and correct
reports of data collected in other studies in      storage during the study. The data collectors
the literature, and it was further refined in       were also monitored monthly to ensure
meetings with PhD faculty from the local col-      proper technique for each device. The accu-
lege of nursing. The clinical manager of the       racy of the 3M Tempa-DotÔ Single-Use
PACUs and the Magnet Coordinator reviewed          chemical clinical thermometer is clinically
the tool and made further recommendations.         certified to 60.1 C (60.2 F) for temperature
Finally, the RN data collectors working in the     ranges 35 C and 40.5 C (96 F and 104.8 F)
PACUs reviewed the tool for content validity       per The American Society for Testing and Mate-
and ease of use. The principal investigator        rials (ASTM) standard.30 The difference is
(PI) also reviewed the tool to assure the elimi-   10.2 C (60.4 F) for temperatures 35.8 C to
nation of all protected health information not     36.9 C (, 98 F) and 39.1 C to 40.4 C
necessary for the study.                           (.102 F).31 The accuracy of the Alaris
                                                   TURBO)TEMPÔ thermometer is clinically
Using the educational materials from each          certified at 60.1 C (6 0.2 F).32
company, all RN data collectors were in-ser-
viced by the PI on the proper use of both de-      Each patient admitted to the PACU was as-
vices. Data collectors were also in-serviced       signed to one RN. When that assigned RN
on the correct use of the data collection tool     was a data collector, the patient was entered
42                                                                   WASHINGTON AND MATNEY

into the study, and only that RN obtained and      Data Analysis
recorded that particular patient’s tempera-        The data were entered into the SPSS program
tures. Upon patient arrival in the PACU, the       (SPSS, Inc, Chicago, IL) and analyzed with
data collectors obtained the patient’s tempera-    the assistance of the statistician from the re-
ture twice. The data collector placed either       search department of the local college of nurs-
a 3M Tempa-DotÔ Single-Use clinical chemical       ing. Descriptive and inferential statistics, as
thermometer or the Alaris TURBO)TEMPÔ              well as the Bland-Altman33 method of analysis,
thermometer orally or axillary, using the          were used to analyze the data.
proper procedure. The temperature was ob-
tained, read, and recorded. Then, using the        Results
same route, the temperature was obtained           The oral or axillary chemical and electronic
with the device not used the first time, and        temperatures and the core temperatures of
that temperature was obtained, read, and re-       752 patients were obtained and recorded.
corded on the collection tool. The device          The data were cleaned and recoded and exam-
used first for each patient was alternated          ined for outliers. For the cleanest analysis
so as to avoid bias towards any one device,        possible, and the results being similar, the
and to compensate for the time it took to          five highest and lowest outliers for each OR
switch devices. Each patient’s last OR core        core, electronic, and chemical temperature
temperature was also recorded on the data col-     readings were removed. This resulted in the
lection tool. The choice of oral or axillary was   removal of 25 patients’ temperatures, leaving
a clinical decision determined by the patient’s    a sample size of 727 patients as subjects,
ability to hold either thermometer in the          with a large effect of .0.80 (Table 1).
                                                   Because there was no statistically significant
As an example, for the first patient admitted to    difference between the mean of the oral tem-
a data collector, that RN might use the Alaris     perature and the mean of the axillary temper-
TURBO)TEMPÔ thermometer to obtain an               ature, they were combined for these analyses.
oral temperature. When that registered and         Paired sample t-test determined that the Alaris
was recorded, the RN data collector would          TURBO)TEMPÔ thermometer measured ap-
then use the 3M Tempa-DotÔ Single-Use              proximately one-half degree lower than OR
chemical clinical thermometer to also obtain       core temperature, and the 3M Tempa-DotÔ
an oral temperature. This process would            Single-Use chemical clinical thermometer
continue using the same site until one device      measured about one-half degree higher than
measured 36.1 C (97 F), the temperature          OR core temperature. This resulted in the
required for discharge from the PACU. Only ad-     chemical thermometer measuring one degree
mission and discharge temperatures from both       higher than the electronic thermometer
devices were recorded for the study and used       (Table 2). Correlations and scatter matrices
for analysis.                                      (Fig 1) illustrate a moderate linear relationship
                                                   between the last OR core temperature and
For that data collector’s next patient, the 3M     both the Alaris TURBO)TEMPÔ thermometer
Tempa-DotÔ Single-Use chemical clinical ther-      and the 3M Tempa-DotÔ Single-Use chemical
mometer would be used first, obtaining and          clinical thermometer (Table 3).
recording temperature in the same manner,
followed by the Alaris TURBO)TEMPÔ ther-           The final analysis conducted was based on
mometer. The patients who remained in the          a method suggested by Altman and Bland34
PACU across a shift change and were unable         and Bland and Altman,33,35,36 and has been
to be assigned to another RN data collector        cited frequently in the literature when com-
were not included in the study.                    paring two methods of obtaining clinical
TEMPERATURE MEASUREMENT                                                                                                         43

 Table 2. Paired Sample t-test (Bias Variables)                                 ature and the 3M Tempa-DotÔ Single-Use
                                                                                chemical clinical thermometer was –0.57 F;
                                                                                the limits of agreement, 1.21 and –2.36.
                                     t-test   Significance Difference

Difference OR           11.846                  0.001         0.48858
  core temperature—
  electronic admission                                                          This study consisted of 727 patients, and the
  temperature ( F)                                                             recording of a total of 3,635 electronic, chem-
Difference OR          215.998                  0.001       –0.57470            ical, and core temperatures. This patient sam-
  core temperature—
                                                                                ple size is by far the largest found in our
  chemical admission
                                                                                literature    review,1,15,18,19,23,24,28   giving
  temperature ( F)                                                                                 28
                                                                                a power of 0.995. According to Cohen,29
                                                                                power is the ability of a test to detect an effect
data.1,15,19,26 The mean difference between                                     given that the effect actually exists, indicating
the OR core temperature and the Alaris                                          that the sample size was large enough to be
TURBO)TEMPÔ thermometer was 0.48 F;                                            representative of the population. This effect
the limits of agreement, 2.55 and –1.57. The                                    size is more likely to yield significant results.
mean difference between the OR core temper-                                     The average patient in this study sample was
          Last OR core temperature
          Admission Electric Temp
          Admission Chemical Temp

                                                  Fig 1.   Correlation of scatter plot matrices (in  F).
44                                                                      WASHINGTON AND MATNEY

               Table 3. Correlation                    Both the 3M Tempa-DotÔ Single-Use chemical
                                                       clinical thermometer and the Alaris TURBO)
                            Last OR Core Temperature
                                                       TEMPÔ thermometer had moderately strong
Admission electronic                  0.537            positive correlations with the OR core temper-
  temperature ( F)                                    ature, with an approximately 0.5 F degree
Pearson correlation                   0.001            difference between each device and the core
  significance (2-tailed)                               temperature, and the correlation matrices
Admission chemical                    0.607            indicate the same relationship (Fig 1). With
  temperature ( F)                                    this correlative relationship, the clinician
Pearson correlation                   0.001
                                                       would have to remember those differences be-
  significance (2-tailed)
                                                       tween each device and the core temperature
                                                       when using the temperature to make clinical
female, between the ages of 30 and 55 years,           decisions. This is consistent with the findings
and was pre-warmed before having abdominal             of Farnell and associates24 for chemical ther-
surgery under general anesthesia (Table 1).            mometers (Table 2).
Patients having abdominal surgery are more
likely to be pre-warmed because of the                 These differences in our study may have oc-
amount of body heat that can be lost from an           curred because the core temperature was
open abdominal incision.9                              taken in the OR and the admission temperature
                                                       was taken in PACU. Thus, there was opportu-
Factors that may have influenced the out-               nity for some temperature change during the
comes for this sample include age, type of             transfer. Also, the 3M Tempa-DotÔ Single-Use
anesthesia, type of surgery, and pre-warming           chemical clinical thermometer dots each
in the preoperative holding area. Independent          have different chemical mixtures that change
t-test for equality of means determined that           temperature at specific temperatures in 0.1 C
type of anesthesia—general or nongeneral—              (0.2 F) increments. This may have allowed
had no effect on the last OR core temperature          for measurement of smaller changes in temper-
(t 5 0.764, P 5 .445, 95% CI). The type of sur-        ature than with the chemical device.
gery—abdominal or nonabdominal—did influ-
ence the last OR core temperature (t 5 3.712,          Although correlation represents a relationship
P # .001, 95% CI). When type of surgery was            between two variables, Bland and Altman33
held constant, there was a weak negative cor-          state that agreement is different from correla-
relation between last OR core temperature and          tion. Agreement refers to how well, on aver-
age. When pre-warming was held constant,               age, two methods measure the same parameter
there was a weak negative correlation between          and can be used interchangeably. The need is
last OR core temperature and age (Table 2).            to know how much of a disagreement there
                                                       is between methods to determine which is
Age was most likely not a factor in the last OR        the closest to the standard, and whether the
core temperature because the majority of our           differences are significant enough to cause
patients were within a 15-year age span (Table         clinical problems in patient care and manage-
1).Those at extremes of ages—the very young            ment.35,36
and the very old—are more affected by heat
loss and more likely to develop hypother-              Bland and Altman examine the amount of dis-
mia.4,8,9,17 In our study, there was no statistical    agreement between two methods of obtaining
difference between the temperature sites—              the same measurement. They term this mea-
axillary or oral; therefore, these measurements        sure of disagreement the bias, which is esti-
were analyzed as electronic or chemical. This          mated by calculating the mean difference
finding is consistent with Erickson et al.28            between the two methods. It is the difference
TEMPERATURE MEASUREMENT                                                                                                                                                 45

        Table 4. Limits of Agreement                                                                                     between the two methods. The scatter plots
                                                                                                                         also give a visual indicator of agreement or dis-
                                                                                   Upper Lower
                                                                                                                         agreement. If the limits of agreement are small,
                                                                        Bias       Limit       Limit Difference
                                                                                                                         the two methods are considered to be in agree-
Last OR core temp   0.4886 2.55 21.57                                                                      4.12          ment and perhaps interchangeable. If the bias
  and electronic                                                                                                         is close to zero, on average, the two methods
  thermometer ( F)                                                                                                      agree (Table 4).15
Last OR core temp 20.557 1.21 –2.36                                                                        3.57
  and chemical                                                                                                           The Bland-Altman scatter plots (Figs 2 and 3)
  thermometer ( F)
                                                                                                                         demonstrate much scatter, biases significantly
                                                                                                                         different from zero, and large limits of agree-
between temperatures on the same subject                                                                                 ment. The bias, or the amount of disagree-
taken with two different types of thermome-                                                                              ment, for both devices is approximately
ters.35 The bias is used because there may be                                                                            0.5 F, as was the t-test of difference in means.
a tendency for one method to exceed the                                                                                  The limits of agreement are a difference of
other, and using the mean difference takes                                                                               4.12 F for the electronic device, and 3.57 F
this into account.33 The limit of agreement is                                                                           for the chemical device. This large range indi-
the range where 95% of the differences be-                                                                               cates less agreement between each device and
tween the two measurements are located35                                                                                 the core temperature, and that the devices are
and is defined as the bias 6 two standard devi-                                                                           not interchangeable in reproducing the core
ations. Bland and Altman35 suggest that it may                                                                           temperature. These findings were inconsis-
be more informative to plot the differences in                                                                           tent with Giuliano et al,1 but consistent with
measurements between the two methods                                                                                     Farnell,24 Fisk,15 and Jensen.23 Clinically, the
against the mean of those measurements                                                                                   disagreement found between the methods in

         Difference OR core temp-Electronic admission temp

                                                                     +1.96 SD (2.55)


                                                                     Mean difference (.49)


                                                                     -1.96 SD (-1.57)



                                                                       93.0             94.0        95.0          96.0       97.0        98.0           99.0   100.0
                                                                                                    Mean OR core-Electronic admission

                                                                                    Fig 2. Bland-Altman scatter plot—electronic thermometer (in  F).
46                                                                                                                                             WASHINGTON AND MATNEY

          Difference OR core temp-Chemical admission temp


                                                                    +1.96 SD (1.21)


                                                                    Mean difference (-.56)


                                                                    -1.96 SD (-2.36)



                                                                      94.0              95.0            96.0           97.0           98.0              99.0   100.0
                                                                                                     Mean OR core-Chemical admission

                                                                                   Fig 3.    Bland-Altman scatter plot—chemical thermometer (in  F).

this study is too large when it may be neces-                                                                          eter would be used only for patients requiring
sary to use the temperature to make treatment                                                                          isolation.
                                                                                                                       The limitations of this study can be used for fu-
Implications and Limitations                                                                                           ture research studies. We did not record the
The purpose of this study was to determine                                                                             ambient temperature of the OR or that of the
the device that measures closer to core tem-                                                                           PACU during each patient’s stay, which may
perature. As a result of the study, it has been                                                                        have influenced any changes in temperature
determined that our health care alliance can                                                                           when transferring from the OR to the PACU.
continue to use both devices, with the realiza-                                                                        The study could not address the linearity of
tion that the 0.5 F differences may make a clin-                                                                      either device because we were unable to mea-
ical difference in treatment and outcomes,                                                                             sure core temperature upon patient discharge.
especially in the hypo- or hyperthermic pa-                                                                            It would have been impractical and uncom-
tient. Each clinician must take into account                                                                           fortable to the patient to leave or replace the
this average difference when assessing patient                                                                         esophageal temperature probe to measure
temperatures. Another option is to use which-                                                                          core temperature on discharge. Therefore
ever device is used initially throughout the pa-                                                                       we do not know if either device may have
tient’s hospital stay. This would require some                                                                         been in agreement at hypothermic levels or
method of documentation of the device to en-                                                                           at higher levels once the patient warmed to
sure the continuity of measurement. A third                                                                            normal temperature. We also did not have con-
option is for the alliance to provide only one                                                                         trol over the position of the probe in the
device to assess patient temperature. The pa-                                                                          esophagus by the anesthesiologist. An agree-
tient care practice council recommended                                                                                ment with the anesthesiologist about consis-
that the electronic thermometer be used rou-                                                                           tent placement would have strengthened the
tinely for all patients. The chemical thermom-                                                                         study.
TEMPERATURE MEASUREMENT                                                                                                                 47

The extensive education and observations of                            used the results of this study to improve pa-
the data collectors could account for the lack                         tient care by using the proper equipment
of difference between oral and axillary tem-                           with the correct technique to help make
peratures taken. The average bedside clinician                         sound clinical decisions. When purchasing
may not take such care with each and every                             patient care equipment, nurses should have
temperature obtained. These patients were                              input to evaluate clinical effectiveness, effi-
also artificially cooled and warmed, and under                          cacy, and patient safety issues, especially when
general anesthesia, which affects the body’s                           two methods of measuring the same physio-
thermoregulation. This study may need to be                            logical parameter will then be available.
replicated on nonsurgical patients to assess
what happens with patients with intact ther-
moregulatory centers, and who have other                               Acknowledgments
conditions that may induce extremes of tem-
                                                                       We acknowledge with grateful appreciation the enthusiasm and
peratures.                                                             sustained efforts of our data collectors who completed the fed-
                                                                       eral requirement to participate in human research and spent 14
This article is the report of a research study ini-                    months collecting data for this study. The following RNs were
                                                                       invaluable data collectors in the completion of this study: Patty
tiated by a question from bedside nurses that is                       Altman, RN; Judy Anderson, RN; Rhonda Dingus, RN; Cathy Dot-
very relevant to their nursing practice of car-                        son, RN; Kathy Gresham, RN; Larissa ‘‘Lisa’’ Griffith, RN; Pamela
ing for patients. These nurses also participated                       S. Johnson, RN; Chana Kirby, RN, CPAN; Sharon Merryman, RN;
in the research study by collecting the data                           Gail Perry, RN; Judith Moody Rosenmeier, RN; and Michael
                                                                       Stevens, RN. We also extend a special thank you to Deborah
and completing the IRB requirements for                                T. Pfortmiller for her assistance with statistical analysis of these
human research. Our health care system has                             results.

   1. Giuliano KK, Scott SS, Elliot S, et al. Temperature measure-        10. Frank SM, Nguyen JM, Garcia CM, et al. Temperature
ment in critically ill orally intubated adults: A comparison of pul-   monitoring practices during regional anesthesia. Anesth Analg.
monary artery core, tympanic, and oral methods. Crit Care Med.         1999;88:373-377.
1999;27:2188-2193.                                                        11. Agrawal N, Sewell DA, Griswold ME, et al. Hypothermia dur-
   2. American Society of Perianesthesia Nurses. Clinical guide-       ing head and neck surgery. Laryngoscope. 2003;113:1278-1282.
line for the prevention of unplanned perioperative hypother-              12. McGee TD. Principles and Methods of Temperature
mia. Available at: Accessed September            Measurement. New York. NY: JohnWiley & Sons, Inc; 1989.
19, 2004.                                                                 13. Roberts JR, Custalow C, Hedges JR. Temperature. In
   3. Holtzclaw BJ. New trends in thermometry for the patient          Roberts JR: Clinical Procedures in Emergency Medicine. 4th ed.
in the ICU. Crit Care Nurs Q. 1998;21. Available at: http://www.       Available online at: Accessed July 7, 2006.                         80306883-3/0/1193/13.html?tocnode 552353055&from
   4. Darm RM, Hecker RB, Rubal BJ. A comparison of noninva-           URL513.html#4-u1.0-B0-7216-9760-7..50005-7- -cesec53_59.
sive body temperature monitoring devices in PACU. J Post               Accessed October 23, 2007.
Anesth Nurs. 1994;9:144-149.                                              14. Hooper VD, Andrews JO. Accuracy of noninvasive core
   5. Mahoney CB, Odom J. Maintaining intraoperative normo-            temperature measurement in acutely ill adults: The state of
thermia: A meta-analysis of outcomes with costs. AANA J.               the science. Biol Res Nurs. 2006;8:24-34.
1999;67:155-163.                                                          15. Fisk J, Arcona S. Comparing tympanic membrane and pul-
   6. Bitner J, Hilde L, Hall K, et al. A team approach to the pre-    monary artery catheter temperatures. Dimens Crit Care Nurs.
vention of unplanned postoperative hypothermia. AORN J.                2001;20:44-49.
2007;85:921-929.                                                          16. Mazur LJ, Yetman RJ, Chan W. Temperature comparisons
   7. Al-Qahtani AS, Vessahel FM. Incidence of intraoperative          with four thermometer types. Ambul Child Health. 1997;3:21-26.
hypothermia. Adopting a protocol for its prevention. Saudi                17. Backlund M, Lepantalo M, Toivonen L, et al. Factors asso-
Med J. 2003;24:1238-1241.                                              ciated with postoperative myocardial ischaemia in elderly
   8. Cory M, Fossum S, Donaldson K, et al. Constant tempera-          patients undergoing major non-cardiac surgery. Eur J Anaesthe-
ture monitoring: A study of temperature patterns in the post-          siol. 1999;16:826-833.
anesthesia care unit. J Post Anesth Nurs. 1998;13:292-300.                18. Arkilic CF, Akca O, Taguchi A, et al. Temperature monitor-
   9. Forstot RM. The etiology and management of inadvertent           ing and management during neuraxial anesthesia: An observa-
perioperative hypothermia. J Clin Anesth. 1995;7:657-674.              tional study. Anesth Analg. 2000;91:662-666.
48                                                                                           WASHINGTON AND MATNEY

   19. Catta neo CG, Frank SM, Hesel TW, et al. The accuracy            27. Prentice D, Moreland J. A comparison of infrared ear ther-
and precision of body temperature monitoring methods during           mometry with electronic predictive thermometry in a geriatric
regional and general anesthesia. Anesth Analg. 2000;90. Avail-        setting. Geriatr Nurs. 1999;20:314-317.
able at           28. Erikson RS, Myer LT, Woo TM. Accuracy of chemical dot
938. Accessed October 23, 2007.                                       thermometers in critically ill adults and young children. Image J
   20. Ozaki M, Kurz A, Sessler DI, et al. Thermoregulatory           Nurs Sch. 1996;28:23-28.
thresholds during epidural and spinal anesthesia. Anesthesiol-          29. Cohen J. Statistical Power Analysis for the Behavioral
ogy. 1994;81:282-288.                                                 Sciences. 2nd ed. Hillsdale, NJ: Lawrence Erlbaum Associates;
   21. Hecker RB, Brownfield RM, Rubal BJ. Bayesian analysis of        1988.
noninvasive versus oral temperature measurements to deter-              30. ASTM. Temperature measurement. In: Annual Book of
mine hypothermia in postoperative patients. South Med J.              ASTM Standards. West Conshohocken, PA: ASTM; 2001. p. 373.
1996;89:71-77.                                                          31. 3M. Technical information. 3M Tempa-Dot Single-Use
   22. Fallis WM, Christiani P. Neonatal axillary temperature         Clinical Thermometers. Available at:
measurements: A comparison of electronic thermometer pre-             healthcare. Accessed August 13, 2004.
dictive and monitor modes. JOGNN. 1999;28:389-394.                      32. Alaris Turbo Temp Thermometer. Alaris Medical Systems.
   23. Jensen BN, Jensen FS, Madsen SN, et al. Accuracy of dig-       Available at: Accessed August 13,
ital tympanic, oral, axillary, and rectal thermometers compared       2004.
with standard rectal mercury thermometers. Eur J Surg. 2000;            33. Bland JM, Altman DG. Statistical methods for assessing
166:848-851.                                                          agreement between two methods of clinical measurement. Lan-
   24. Farnell S, Maxwell L, Tan S, et al. Temperature measure-       cet North Am Ed. 1986;1:307-310.
ment: Comparison of non-invasive methods used in adult criti-           34. Altman DG, Bland JM. Measurement in medicine: The
cal care. J Clin Nurs. 2005;14:632-639.                               analysis of method comparison studies. The Statistician. 1983;
   25. Van den Bruel A, Aergeerts B, DeBoeck C, et al. Measuring      32:307-317.
the body temperature: how accurate is the Tempa-DotÒ? Tech-             35. Bland JM, Altman DG. Measuring agreement in method
nol Health Care. 2005;13:97-106.                                      comparison studies. Stat Methods Med Res. 1999;8:135-160.
   26. Potter P, Schallom M, Davis S, et al. Evaluation of chemical     36. Bland JM, Altman DG. Applying the right statistics: Anal-
dot thermometers for measuring body temperature of orally             ysis of measurement studies. Ultrasound Obstet Gynecol. 2003;
intubated patients. Am J Crit Care. 2003;12:403-408.                  22:85-93.

To top