In this study of 12 term gestation infants, approximately 5
to 9 weeks of age, the gradient and relationship between
abdominal skin temperature and axillary temperature were
highly variable within individual subjects. While the mean
difference between temperatures was 0�C for the sample as a
whole, aggregating the data in this fashion obscures
individual differences. The standard deviation of the
difference between abdominal skin and axillary temperature was
>/=0.5�C in 4 of 12 subjects. Thus the gradient between
abdominal skin and axillary temperatures is not a consistent
value, rather the difference between abdominal skin and
axillary temperature appears to fluctuate across time within
some subjects and to vary across subjects. Ideally the
gradient between comparable temperature sites should be
consistent. Within subjects, the correlation between abdominal
skin and axillary temperature was also variable. Two of 12
subjects had r values >/= 0.85 while 4 subjects had
r values < 0.5. Comparable measures should be
accurate and reproducible. Finally the limits of agreement
determined from a random sample of the data indicate that
approximately 95% of all differences between abdominal skin
and axillary temperature would lie between �0.77 and 0.93�C, a
wide spread, which may not be acceptable for some clinical
applications.
These findings suggest that abdominal skin and axillary
temperature may not consistently track each other over time.
The graphed data (Fig 1) illustrate this point. The first
portion of the recording shows very close correspondence of
axillary and abdominal skin temperature, compared with the
later portion. There are three likely contributors to the
correspondence of abdominal skin and axillary temperature.
First, both sites reflect skin blood flow, which may vary
based on the infant's thermoregulatory status. Second, the
degree of probe insulation will alter the recorded
temperature. While insulated, reflective probe covers were
used in data collection, no control of clothing or covering
was implemented. Further, adduction of the arm, another form
of insulation, was not controlled, possibly influencing
axillary temperature. Third, probe adherence, is a concern.
While probes generally remained in place throughout the data
recording period, even minor movement of the probe can alter
the temperature recorded. These three factors should be
included in future studies of abdominal skin and axillary
temperature.
Evidence strongly supports the relationship between infant
temperature and mortality.[20] How, then, should
study results be applied to practice? The current study does
not answer questions regarding the superiority of either
abdominal skin or axilla as an estimate of core temperature.
The data do provide information regarding the comparability of
these two sites in a small sample of term infants, however,
caution is warranted in generalizing beyond the study sample.
Evidence regarding comparability of temperature monitoring
sites is relevant to clinical decision making. It is not
reasonable to assume that the gradient between abdominal skin
and axillary temperatures is consistent across time nor that
changes in temperature recorded from these sites are entirely
parallel.
Whether abdominal skin or axillary temperature is selected
for continuous monitoring of an infant, findings suggest that
the type of site should be used consistently. Switching
between axillary and abdominal skin sites may increase
measurement error. The results also suggest that choice of
temperature site should be individualized. Factors such as
infant size and amount of tissue insulation, external
insulation, positioning needs, and skin integrity may
influence the comparability of abdominal skin and axillary
temperature.[15,21,22] Further, probe adherence and
consistent insulation are critical for accurate continuous
skin temperature measurement.
When using either abdominal skin or the axilla to
continuously monitor temperature, it is important to recall
that neither site is an ideal estimate of core temperature.
Additionally, several factors contribute to measurement error
in skin temperature. Continuous recordings of temperature
should be evaluated for trends and pattern. The goal is a
picture over time, rather than one isolated measure in time.
Although the temperature of infants monitored continuously is
typically documented in the patient record, graphing the
temperature not only helps visual identification of infant
temperature pattern, but can also help diagnose problems with
the probe itself. The monitor display of abdominal skin or
axillary temperature is one of numerous pieces of data used in
clinical decision making. Nursing judgment remains a
key element in thermal care. (My
emphasis)
Interesting: from Medscape: only page 5 of 5. MM
