Reliability of Four Types of Electrothermometer Thermocouples

Abstract

0132 Thermocouples are used with electro thermometers to measure tissue temperature during therapeutic modality research. One manufacturer reports their thermocouple measures are reliable within ± 0.1°C, but we could find no independent verification. Purpose: To examine measurement variability between different models of type- T canstantan-copper thermocouples and the variability within a 3-level type-T multisensor. Methods: 1st Exp: Pairs of 3 types of thermocouples (IT18, IT21, st PT6) were connected to 6 randomly selected channels in 2 electrothermometers and immersed in a 16.3°C water bath. Thermocouples were attached to a different channel within a machine for each of 6 trials. 2nd Exp: Four 3-level multisensors were nd connected to a single electrothermometer. Probes within a single multisensor were connected to sequential channels. The thermocouples were immersed in an 18.4°C water bath. Four trials were performed allowing all multisensor probes to occupy each channel set. Measurements for both experiments were taken every 10 sec for 3 min. Electrothermometer measurements were compared to a mercury thermometer graded in 0.1°C increments. Absolute differences between visual and electronic thermometers were compared with ANOVA's and post-hoc tests. Results: 1st Exp: There was a st difference between probes, all recorded less than the mercury thermometer. The PT6 < was closest (−.08±.05°C) followed by IT18 (−.12±.04°C) and IT21 (−.14±.04°C). 2nd Exp: Probes within the multisensors differed from each other (−.18±.03°C,−.17±.02°C,−.16±.02°C, respectfully). Conclusion: Temperature measure reliability differs between types of thermocouples. Surface probes (PT6) appear to be within manufactures and electrothermometer specifications (±0.1°C). Single intramuscular thermocouples (IT18 & IT21) were less reliable and multisensors were not within advertised specifications. Reports of changes in intramuscular temperatures of less than 0.2°C, although statistically significant, must be viewed with skepticism.