The largest volume of research identified during the literature search addressed various aspects of temperature measurement. These studies highlight the large range of methods and body sites that are used for the measurement of temperature (see table three). Because of the volume of research, comparisons of different temperature measurement methods will be summarised as a separate systematic review. Summarised in this practice information sheet are studies addressing aspects of oral, rectal axillary and tympanic temperatures.
While much attention has focused on measurement accuracy, one study evaluated touch as a screen for fever and found that while mothers and medical students overestimated the incidence of fever when using touch, they rarely missed its presence in a child. The results of this study perhaps challenge the current focus of research on the accuracy of measurements using tenths of a degree, when simple touch is an accurate measure for fever. The use of temperature as a discharge criterion for an ambulatory surgical unit has been studied, but results suggest it is not useful in neither rapid or deep differentiating readiness for Oral Temperatures Studies evaluating measurements from the different areas of the mouth recommend using either the right or left posterior sublingual pocket, as these result in higher recorded temperatures.
Evaluation of the impact of oxygen therapy on oral temperatures have reported contradictory results regarding its statistical significance, however no study reported a clinically significant effect.
Similarly, different rates of oxygen flow, from 2 litres to 6 litres per minute, and warmed or cooled inspired gas, were found not to have an influence on oral temperature measurements. Two studies found that rapid respiratory rates had a small influence on oral temperatures, but these results were contradicted by another study that found breathing, alone or in combination, had any significant effect on oral temperatures.
Studies have shown that drinking hot or cold water has a significant impact on recorded oral temperatures, and it has been suggested waiting 15 to 20 minutes after drinks to ensure accuracy. Smoking does not change oral temperature measurements.
Researchers have evaluated the of the temperature accuracy time required for mercury thermometers to accurately record the person's oral temperature. One study found that with healthy adults, using a two minute insertion time resulted in 27% of the temperature readings having an error of at least 0.3°C. A study assessing thermometer insertion time in afebrile and febrile adults, suggested a six minute insertion time as a compromise between optimal time and clinical practicality while another recommended a seven minute insertion time to ensure the majority of afebrile and febrile temperatures are correctly recorded. However, a survey of nurses showed that most left the mercury thermometer in the mouth for less than 3 minutes.
There has been only limited re- search focusing on axillary temperatures. One study evaluated axillary temperature measurements in elderly females, and found great variation between individuals.
While the mean axillary temperatures were approximately 36°C, the wide range of tem- peratures encountered prevented the identification of a single figure that could be considered the "normal" axillary temperature. Another study evaluated the influence of intravenous infusions, via upper limbs of neonates, on axillary temperatures and found there was little significance in terms
There has been considerable research addressing tympanic temperature measurements ranging from the influence of infection and cerumen on measurement accuracy, to optimal technique. Studies have evaluated the impact of otitis media on tympanic temperatures and suggest it has little effect. While some studies have reported a statistically significant difference in tympanic temperatures between ears in people with unilateral otitis media, this difference was approximately 0.1°C and so of little clinical importance. The presence of cerumen does influence tympanic temperature readings, and while results are variable, they suggest a significant proportion of the temperature readings taken from the occluded ear will be more than 0.3°C lower than the ear that is not occluded.
Studies evaluating technique suggest an ear tug should be used during the measurement of tympanic temperatures, as this is reported to straighten the external auditory canal. Failure to use the ear tug means infrared thermometers are only partially directed at the tympanic membrane. The tug technique in adults has been described as pulling the pinna (auricle of ear) in an upward and backward direction, and in infants it is pulling the pinna in a backward direction.
Evaluations of the impact of ambient temperatures on tympanic temperatures suggest that while a hot environment can significantly affect readings, cold appears to have little effect.
Cost analyses of the different temperature measurement methods suggest infrared measurements may be the most cost effective despite the greater initial costs. These savings are the result of the rapid reading capabilities of these instruments, and the labour cost savings that result.
Many studies have compared the different methods of temperature measurement, and commonly rectal temperatures are used as the standard comparison. However, these studies will be summarised in a separate systematic review. The most common reported issue related to rectal temperature measurement is that of rectal perforation, which appears to be a risk primarily for the newborn and very young. Other reported complications include peritonitis secondary to rectal perforation, and one case of intra-spinal migration of a rectal thermometer in a two year old, which broke during routine rectal temperature measurement. A ten year review of hospital records identified 16 children admitted to a surgical unit with broken or retained rectal thermometers. In response to this problem axillary temperature measurements have been recommended in preference to the rectal measurements. With the advent of infrared tympanic thermometers, these complications are likely to become less common.