Background: The mercury thermometer (MT) was considered the reference standard for the evaluation of body temperature; however, since April 2009, it has no longer been available in Italy.
Objective: This study aimed to evaluate the accuracy of digital thermometer (DT) and galinstan thermometer (GT) in comparison with the MT.
Methods: We prospectively recruited 284 children (age, 1 month to 17 years; mean, 8.5 years) seen in the emergency department of a tertiary pediatric hospital between November and December 2010. For each patient, body temperature was measured sequentially in the right axilla in a randomized fashion using DT, GT, and MT. Fever was defined as an axillary temperature of 37.5°C or greater. The temperature readings with DT and GT were compared statistically with those of MT (reference standard).
Results: No statistically significant difference in mean temperature was found between MT and GT in pairwise comparison (P = 0.06), whereas significant differences were found between MT and DT (P < 0.001) and between GT and DT (P < 0.001). Sensitivity and specificity of DT were 67.5% and 98.0%, respectively. When considering children with a temperature of 39°C or higher, the false-negative rate was 65.4% with DT and 30.8% with GT.
Conclusions: Although both DT and GT had good specificity and positive predictive value compared with MT, GT had higher sensitivity and a lower rate of false-negative rates. Galinstan thermometer is more accurate in the measurement of body temperature compared with DT or MT.