Comparing liquid crystal thermometer readings and mercury thermometer readings of infants and children in a traditional African setting. Implications for community-based health.

Abstract

Liquid crystal thermometer (LCT) readings of skin temperatures were compared with mercury thermometer (MT) rectal temperature readings to assess the reliability of LCTs. Temperatures of 498 children were measured at two points in time. LCT skin temperature readings of children 0 to 52 months were on average 0.50 degrees C and 1.97 degrees C lower than MT rectal temperature readings. A strong correlation between temperature differences and LCT readings indicated that the greatest differences occurred at the lower LCT readings. These conclusions indicate LCT skin readings undermeasure temperature. Some of these differences were due to MTs not measuring temperatures below 35 degrees C. Children under 1 year of age had significantly greater differences than any other age group. Their LCT readings were, on average, 1.65 degrees C lower than their MT readings. Using MTs as a standard, LCTs were 100% sensitive and 92% specific for detecting children with hypothermia. LCTs were 38.5% sensitive and 100% specific for detecting fevers. These results suggest that LCTs leave undetected a large proportion of children who have fevers. However, they are sensitive for identifying children with hypothermia. A knowledge, attitude and practice (KAP) study indicated that local mothers can be identified who understand principles and procedures of LCTs, and accept them for health care of their child.