Palpation as a method of fever determination in Malawian children who are less than 5 years old: how reliable is it?

Fever is a common occurrence in children who are < 5 years old and palpation of the forehead may or may not be a reliable method for determining fever in such children. In a study of 1120 Malawian children of this age attending outpatient's clinics, each child's mother and a clinical officer (CO) were asked to palpate the child's forehead and decide whether the child was febrile (felt warm or very warm) or afebrile (felt normal). The rectal temperature of each child was then taken using a thermometer and the child considered febrile if this temperature was > or = 38 degrees C. Using palpation, mothers judged 973 (86.9%) of 1120 children to be febrile and CO judged 565 (50.4%) of 1118 to be febrile, whereas thermometer readings indicated 410 (36.7%) to be truly febrile. False-positives (i.e. afebrile children judged to be febrile by palpation) accounted for 574 (59.0%) of the 973 children who were considered febrile by their mothers and 228 (40.4%) of the 565 children so considered by CO; mothers reported significantly more false-positives than CO (P < 0.05). False-negatives (i.e. febrile children judged to be afebrile by palpation) accounted for 11 (7.5%) of the 147 children who were considered afebrile by their mothers and 73 (13.2%) of the 553 children so considered by CO; CO reported significantly more false-negatives than mothers (P < 0.05). Overall, mothers were as likely as CO to misjudge a child (721/1120 v. 781/1118; P > 0.05). Although the sensitivity of mothers and CO in determining fever was similar (97.3% v. 82.2%; P > 0.05), CO gave a higher degree of specificity than the mothers (67.8% v. 19.2%; P < 0.000001). Although the present results indicate that palpation is not a reliable method of determining fever in children who are < 5 years old, caregivers should continue to use palpation as a useful first step in deciding when a child needs to be referred.