Doctors who use in-ear infrared devices to take a child's temperature may be missing fevers that a standard rectal thermometer would have detected.
The British meta-analysis included 31 comparisons of rectal and ear temperature in 4441 children. Although the average difference between the two readings was small, some wide differences raise the concern that ear thermometry could both over- and underestimate core body temperature. The differences were also unpredictable, making it difficult to provide useful advice.
These findings suggest that doctors who use infrared thermometry may not only fail to detect fever in infants, but may also find it impossible to take accurate readings where body temperature needs to be precise (Lancet, 2002; 360: 603-9).