It is difficult to pinpoint the lowest temperature elevation considered to be definitely abnormal for all children under all circumstances Some children normally have rectal temperatures as low as 36.2°C (97.2°F) or as high as 38°C (100.4°F) Children, like adults, also have diurnal variations in temperature, with the peak usually occurring between and PM Factors such as excessive clothing, physical activity, hot weather, and ovulation can raise temperature in the absence of disease For the appropriately dressed child who has been at rest for 30 minutes, a rectal temperature of 38°C (100.4°F) is defined as fever for this discussion Using the proper technique to record temperature is important for maximum accuracy Rectal thermometry combines attributes of being the least invasive way to most approximate core temperature compared to other invasive methods, such as esophageal or bladder thermometry Optimal technique for rectal thermometry includes appropriate positioning and restraint in infants (prone, supine, or on the side with hips slightly flexed), depth of insertion (about to cm), and time for equilibration (2 to minutes with glass thermometers or several seconds with electronic digital probes) The thermometer should not be placed directly into a fecal mass because the fecal temperature may not have equilibrated with rapid fluctuations in core temperature and thus may be falsely low as temperature rises rapidly Other noninvasive methods have varying accuracy and precision with most being shown to overestimate at lower body temperatures and underestimate at higher temperatures Oral and axillary temperatures are usually about 0.6°C (1°F) and 1.1°C (2°F) lower than rectal temperatures, respectively More recent attempts to measure temperature with less invasive techniques include temperature-sensitive pacifiers and forehead strips, both of which have been found to be unreliable Infrared tympanic membrane (aural) thermometry, based on the premise that the tympanic membrane shares vascular supply with the hypothalamus, has also shown inadequate accuracy and precision in both afebrile and febrile children and is not useful in health care settings Temporal artery thermometry, which uses similar infrared technology to measure heat produced by the temporal artery, has more recently been studied The temporal artery technique has been shown to more closely approximate rectal temperatures; however, it has still been found to underestimate high temperatures and overestimate low ones Temporal artery thermometry may be useful as a noninvasive screening tool in children where fever is not a major concern (e.g., trauma), but should be used cautiously in children where accurate temperature measurement is important As even a low-grade fever may be clinically significant in young infants