This prospective study assessed the agreement between TTP and esophageal or rectal temperature measurements in infants and toddlers. We confirmed the clinical accuracy of TTP as an alternative to esophageal temperature measurements in pediatric patients within the clinically acceptable temperature range. We further demonstrated that TTP was more accurate with esophageal temperature than with rectal temperature. In the future, the results of this non-invasive temperature measurement will contribute to perioperative temperature measurement and management in pediatric patients, especially infants.

Esophageal temperature reflects core body temperature in pediatric patients most accurately and is widely accepted in perioperative clinical practice [12]. However, an ideal temperature measurement system should be able to measure core body temperature, even in awake patients. TTP is non-invasive because it can measure core temperature from the body surface and allows continuous monitoring of temperature in patients of any age or level of consciousness. No studies have evaluated non-invasive core temperature measurements in infants. Therefore, TTP can be considered a clinically useful perioperative core body temperature monitoring device because of its high accuracy in measuring esophageal temperature, not only in adults but also in infants and toddlers.

Nemeth et al. demonstrated an acceptably accurate TTP in pediatric patients compared with Tesoph [20]. They revealed a mean difference of  – 0.07 °C with 95% limits of agreement of  – 1.00 °C to 0.85 °C in pediatric patients with a mean age ± SD of 34.8 ± 25.2 months (range, 5 days to 6.9 years), a wider age range than our study population [20]. Sang et al. also investigated the accuracy of non-invasive zero-heat-flux thermometer similar to the TTP sensor, SpotOn™ sensor (3 M™ Bair Hugger™ sensor, 36,000, 3 M Medical, USA), compared with Tesoph in non-infant pediatric patients with mean age ± SD of 45 ± 16 months (range, 1–8 years) [26]. They showed a bias of -0.07 °C and 95% limits of agreement of  – 0.41 to 0.28 °C [26]. Our study also showed that the bias of each measurement was 0.09 °C with 95% limits of agreement within  – 0.53 to 0.70 °C and the proportion of error within ± 0.5 °C was 92.2% in younger pediatric patients (mean age ± SD of 14.0 ± 9.5 months). Therefore, we confirmed the clinically acceptable accuracy of TTP compared with esophageal temperature in pediatric patients, especially in infants and toddlers.

Zeiner et al. assessed the agreement between non-invasive thermometry, using a Tcore™ (Dräger, Drägerwerk AG & Co. KG, Lübeck, Germany) and Trect in pediatric patients with a mean age of 26.7 months [27]. They indicated a bias of 0.41 °C and limits of agreement of  – 0.74 to 1.57 °C between the two measurements [27]. Our study showed that the bias of each measurement was 0.41 °C with 95% limits of agreement within  – 0.35 °C to 1.17 °C and the proportion of error within ± 0.5 °C was 65.7% between TTP and Trect. In this study, TTP was more accurate with esophageal temperature than rectal temperature. Rectal temperature is less reliable because it lags significantly behind the core temperature in response to rapid thermal changes owing to poor perfusion [25]. In addition, our results of the temperature accuracy under laparoscopic surgery in pediatric patients showed that the mean difference between TTP and Tesoph was 0.06 °C, which was within the clinically acceptable range. However, the bias between TTP and Trect was 0.50 °C, which is difficult to be accepted clinically. Therefore, it was suggested that TTP may have a broader clinical utility.

Nemeth et al. reported the occurrence of a minor skin lesion (superficial epidermal excoriation) owing to accidental removal of the sensor [20]. In our study, even in infants, only a few cases of temporary skin redness were noticed after the TTP sensor removal, similar to the ECG sensor; however, no skin lesions required intervention. The TTP sensor is considered safer than probes inserted into body cavities such as the esophagus or rectum. Furthermore, zero-heat-flux thermometer, SpotOn™, includes a heating component, which is initially preheated to equilibrium with the skin surface temperature. Therefore, manufacturers warn not to apply the sensor to fragile skin. TTP sensors, however, do not actively heat, but rather utilize local temperature and heat flow readings to estimate the core temperature. Therefore, we assume the TTP to be safer in infants, especially in awake conditions.

This study had several limitations. First, the study was conducted under general anesthesia. General anesthesia causes iatrogenic vasodilation, which restricts the thermal response of vasoconstriction in the periphery and leads to the redistribution of heat from the core to the periphery [28]. Additional studies are required to evaluate whether TTP is accurate in pediatric patients with several perioperative conditions, such as during awake state. However, our unpublished data compared TTP and axillary temperature in a small number of awake pediatric patients (n = 9) in the intensive care unit. The results showed that the bias and SD between both measurements was -0.07 °C in a mean age of 1.6 ± 1.8 months, and TTP may be clinically useful even under awake conditions in intensive care units. Second, the esophageal temperature probe was placed in the distal esophagus, and the specific insertion length was not defined. Zhong et al. showed that nasopharyngeal thermometers accurately measured core body temperature when the probe was inserted at the optimal location of an appropriate insertion length, depending on age [29]. However, in this study, the accuracy of the TTP and Tesoph measurements was clinically acceptable. Third, in this study, body temperature changes during surgery were small owing to strict temperature control with forced-air warming system. Therefore, we were unable to evaluate the tracking ability between each measurement method. In the future, it is necessary to evaluate the tracking ability in situations in which the body temperature changes significantly, such as before and after cardiopulmonary bypass. Finally, in this study, most pediatric patients had APA-PS1, which indicated good general conditions, including growth and development. Further research is required to determine whether core temperature measurements using the TTP sensor can be applied clinically in neonates and premature infants.

To summarize, this study demonstrated that the core body temperature evaluated using TTP in infants and toddlers had higher accuracy with esophageal than rectal measurements within the clinically accepted body temperature range. Further studies are warranted in extreme temperature ranges and changes to allow for more clinical applications of TTP in pediatric patients.