Inadvertent perioperative hypothermia is associated with complications such as impaired coagulation and increased blood loss,1 increased cardiac morbidity2 and higher incidence of wound infections.3 Negative aspects such as patient discomfort have also been demonstrated.4 Prevention of perioperative hypothermia has become a major topic of care for anaesthesiologists, surgeons and nurses. Publication of a new German guideline for prevention of perioperative hypothermia,5 together with the British National Institute for Clinical Excellence guideline6 have given it further emphasis.
Adequate measurement of the body core temperature is a prerequisite in detecting and preventing perioperative hypothermia. The best site for temperature monitoring is still under debate.
Pulmonary artery catheters (PACs) are still considered as the ‘gold standard’ in body core temperature monitoring;7 however, they are invasive, dangerous, expensive and inappropriate for most routine surgery. Nasopharyngeal measurement has been shown to be a reliable method for core temperature monitoring, but it is not applicable in awake or head-injured patients.8 Sublingual temperature has been demonstrated as a good ‘near to core’ measurement both in awake and anaesthetised patients.9
Another comfortable and noninvasive method for core temperature assessment follows the ‘zero heat flux’ (ZHF) principle, which was first described in the early 1970s.10 The new 3M SpotOn sensor works by insulating the site of measurement, the forehead, against surrounding influences. It allows the skin underneath to heat up to body core temperature by creating an isothermic tunnel. At the point of ‘zero heat flux’, when the transfer of heat from the core to the forehead skin is complete, the temperature under the adhesive sensor reflects body core temperature.11
The aim of this study was to compare the 3M SpotOn ZHF sensor with the sublingual and nasopharyngeal probes with respect to accuracy, precision and correlation in patients undergoing elective surgery. We hypothesised that the ZHF sensor was adequate for temperature monitoring in the clinical setting to within a value of 0.5°C for accuracy and precision.
Materials and methods
This study was approved by the Institutional Review Board on 23 July 2013 (D 477/13) and registered with ClinicalTrials.gov (NCT02031159).
Men and women undergoing elective trauma or gynaecology surgery under general anaesthesia gave written consent and were enrolled in the study between October 2013 and January 2014. Patients with unexpected blood loss, haemodynamic instability determined by the need of continuous norepinephrine infusion or needing postoperative ventilation were excluded from this study.
All patients received a premedication of oral midazolam (0.07 to 0.1 mg kg−1, maximum 7.5 mg) 30 min before induction of anaesthesia. Standard monitoring including ECG, pulse oximetry, noninvasive blood pressure (3 to 5 min intervals) and neuromuscular monitoring was established. General anaesthesia was induced with sufentanil and propofol, and maintained by sevoflurane (0.7 to 1 MAC). An endotracheal tube or laryngeal mask airway was inserted according to applied standard operating procedures of the department of anaesthesiology. Patients were actively warmed during surgery using a forced air warming blanket device (3M Bair Hugger Therapy, St. Paul, Minnesota, USA), which was set to a working temperature of 43°C.
Body core temperature was measured at 15, 45 and 75 min after induction of anaesthesia from the nasopharyngeal probe (Adult temperature probe, D-OS4; Exacon Scientific A/S, Roskilde, Denmark) and sublingual probe (SureTemp Plus; WelchAllyn Inc. Corporate Headquarters, Skaneateles Falls, New York, USA) and also, at the same time, from single-use sensors (3M SpotOn, St. Paul, Minnesota, USA) attached to the forehead using the ZHF method.11 Sublingual temperature was monitored in the posterior sublingual pocket by lifting the tongue. For nasopharyngeal temperature, the sensor was carefully inserted through a nostril just posterior to the soft palate.
To estimate the sample size needed to detect a significant difference in mean temperatures (paired Student's t-test corrected for multiple testing), we used data from our previous temperature study.9 The analysis was performed with G*Power 220.127.116.11 using the following values: mean difference = 0.15°C, SD = 0.333, power (β) = 95% and α (two-sided) = 5%, giving 77 patients. Due to the evaluation of a new thermometer device with an unknown reliability and dropout rate, we decided to include 120 patients.
Statistical analysis was performed using GraphPad Prism 5.0 software (GraphPad Software, San Diego, California, USA). Correlation between sublingual and ZHF, and nasopharyngeal and ZHF was evaluated by Spearman's rank correlation analysis.
Bland–Altman statistics were used for interpretation of accuracy (mean difference between methods = bias), precision [standard deviation (SD) of difference] and limits of agreement (1.96 SD). From previous studies, we determined that a value of 0.5°C for accuracy and precision is considered as clinically sufficient.8 A P value of less than 0.05 was considered statistically significant.
Of 120 patients recruited, 37 patients were excluded for different reasons (Fig. 1). For 19 patients, sublingual temperature could not be obtained at 45 min of surgery time because of calibration failure of the sublingual probe. For another 12 patients, surgery time was less than 60 min, for four there were signal errors with the SpotOn sensor and two patients opted for spinal anaesthesia shortly before surgery, leaving 83 for analysis. Patient characteristics are presented in Table 1.
ZHF, sublingual and nasopharyngeal temperature measurement was performed without any noticeable side effects. Data sets were analysed separately to ensure independent data. ZHF temperatures measured by the 3M SpotOn sensor were minimally higher than nasopharyngeal temperatures (mean difference 0.07°C) (P = 0.1424) and slightly lower than sublingual temperatures by 0.35°C (P < 0.0001).
Analysis of Spearman rank correlation between ZHF and nasopharyngeal pairs gave a correlation coefficient (r) of 0.85 [95% confidence interval (95% CI) 0.78 to 0.90] at 15 min, 0.88 (95% CI 0.81 to 0.92) at 45 min and 0.89 (95% CI 0.83 to 0.93) at 75 min after induction of anaesthesia (Fig. 2).
At the same times, ZHF and sublingual pairs showed r-values of 0.69 (95% CI 0.55 to 0.79), 0.76 (95% CI 0.64 to 0.84) and 0.84 (95% CI 0.75 to 0.90), respectively (Fig. 3).
For ZHF vs. nasopharyngeal group, Bland–Altman analysis revealed a bias (SD) [95% limits of agreement] of 0.07°C (0.22) [−0.38 to 0.51] at 15 min, a bias of 0.05°C (0.22) [−0.39 to 0.48] at 45 min and a bias of 0.10°C (0.18) [−0.25 to 0.46] at 75 min after induction of anaesthesia (Fig. 4). The ZHF vs. sublingual group showed a bias of −0.37°C (0.30) [−0.95 to 0.22], −0.36°C (0.30) [−0.95 to 0.23] and −0.33°C (0.27) [−0.84 to 0.19] (Fig. 5).
Our main findings in this study can be summarised as follows: In comparison with the nasopharyngeal probe, the 3M SpotOn sensor shows a bias of no more than 0.1°C, which can be interpreted as negligible.8 With sublingual, we measured a bias of up to −0.37°C. The SpotOn sensor correlates well with the nasopharyngeal (sublingual) probe, with a coefficient of up to 0.89 (0.84). Compared with the nasopharyngeal probe, the ZHF sensor almost perfectly represents body temperature, but shows slightly lower values than sublingual measurement (P < 0.0001). In comparison with the sublingual and nasopharyngeal probes, we found the SpotOn sensor adequate for clinical pracice in terms of correlation, accuracy and precision, in haemodynamically stable patients without unexpected blood loss.
Obtaining accurate measurement of core body temperature during an entire perioperative period is still a challenge, and it remains the basis for the prevention and detection of perioperative hypothermia. Ideally, the recommendation of the new German S3 guideline should be followed and a single method and site of measurement should be used throughout the entire perioperative period to avoid a comparison of different techniques. In the absence of a PAC, sublingual temperature monitoring is accepted as a reliable alternative and can be regarded as a ‘near to gold’ standard.12 Interestingly, this applies to both awake and intubated patients.13 Nasopharyngeal temperature monitoring is also well accepted and, according to the German guideline, is a valid technique,5 but adverse events have been reported including severe nasal bleeding.14
Nasopharyngeal temperature probes are unsuitable in the awake patient. Sublingual devices can be used but seem bulky in the perioperative setting with a cover that has to be replaced for hygiene reasons every time a probe is taken out of the mouth. The patient must also be able to breath through the nose, and temperature differences of up to 0.9°C have been found with probes placed incorrectly, for instance behind the front teeth.15
Infrared thermometers are commonly used but have failed to show sufficient agreement with established methods such as rectal thermometry,16 and therefore, should not be used for temperature assessment.5
There are no hygienic concerns with the ZHF forehead sensor used in this study. It allows continous temperature monitoring with the same device throughout the perioperative period and therefore does not impair data interpretation.
In our study, were measured a negligible difference of 0.07°C in mean temperature between ZHF and nasopharyngeal probes and a slight but significant difference of 0.35°C between ZHF and sublingual temperatures. However, the clinical relevance of these differences is questionable. The tolerable accuracy and precision of temperature measurement is still in discussion. In 2008, Sessler8 regarded an inacurracy in body core temperature of up to 0.5°C as tolerable for clinical routine because adverse effects were not observed for less than 0.5°C. But modern temperature monitoring equipment should be able to measure more exactly, and within a range of ±0.1 to 0.2°C. Therefore, a mean deviation of 0.35°C, as found in our study, should represent the maximum tolerable difference. However, in the absence of a real gold standard in our study, we are unable to define the true core temperature.
Our study may be criticised for the high number of patients who failed to reach the final analysis. Technical problems with the sublingual probe calibration were responsible for 19 of these. Nevertheless, we decided to follow a strict regimen regarding data analysis, which resulted in 3 x 83 data sets. There may have been a possible bias arising from the fact that sublingual and nasopharyngeal temperature monitoring devices are widely used in our hospital, although the SpotOn sensor is a new and unfamiliar device. To keep this to a minimum, only one person, familiar with the relatively easy SpotOn sensor technique, was responsible for all measurements in the operating theatre, so bias was little or absent altogether.
We chose to exclude urgent and unstable patients from our study and concentrate on those scheduled for elective surgery. Elective patients were preferred because they reflected daily clinical routine and provided better standardisation. Thermoregulation is subject to greater disturbance in noneuvolaemic patients, and there may be a need for vasoactive agents, making data incomparable. More studies are needed to elucidate this.
In one recently published report, a prototype SpotOn sensor was tested against the PAC in patients undergoing elective cardiac surgery under cardiopulmonary bypass.17 This provided good evidence that the new device will work well in haemodynamically unstable patients, with an overall temperature difference of −0.23°C. Intraoperatively, Eshraghi et al.17 found an average temperature difference of −0.08°C, which is comparable to our bias of 0.07°C (SpotOn vs. nasopharyngeal). The authors conclude that core temperature can be measured noninvasively via the prototype SpotOn sensor even though precision was slightly worse than the designated 0.5°C limits compared with PAC measurements.
The SpotOn sensor had a high failure rate in comparison to the nasopharyngeal sensor in our study, but it performed better than the sublingual probe. We found a SpotOn failure rate of 3.3%, while Eshraghi et al.17 found a rate of 1.9%, which seems comparable. Future software enhancements may be able to eradicate this problem.
In summary, we found the 3M SpotOn sensor using the ZHF method to be an adequate, noninvasive, single-use device for temperature monitoring during the perioperative period in haemodynamicly stable, noncritically ill patients. Correlation, accuracy and precision were on a high level in comparison to nasopharyngeal measurement and indicated acceptability for clinical use in comparison with sublingual measurement.
Acknowledgements relating to this article
Assistance with the study: none.
Financial support and sponsorship: the SpotOn temperature sensors were provided by 3M (St. Paul, Minnesota, USA).
Conflicts of interest: none.
Presentation: preliminary data were presented at the German Anaesthesia Congress (DAC), 8 to 10 May 2014.
1. Rajagopalan S, Mascha E, Na J, Sessler DI. The effects of mild perioperative hypothermia on blood loss and transfusion requirement. Anesthesiology
2. Frank SM, Fleisher LA, Breslow MJ, et al. Perioperative maintenance of normothermia reduces the incidence of morbid cardiac events. A randomized clinical trial. JAMA
3. Kurz A, Sessler DI, Lenhardt R. Perioperative normothermia to reduce the incidence of surgical-wound infection and shorten hospitalization. Study of Wound Infection and Temperature Group. N Engl J Med
4. Lista F, Doherty CD, Backstein RM, Ahmad J. The impact of perioperative warming in an outpatient aesthetic surgery setting. Aesthet Surg J
5. Torossian A, Bein B, Bräuer A, et al. AWMF S3 guideline on avoidance of perioperative hypothermia, 2013. Available at: http://www.awmf.org/leitlinien/detail/ll/001-018.html
(last accessed 01/07/14).
6. Clinical Practice Guideline. The management of inadvertent perioperative hypothermia in adults. National Collaborating Centre for Nursing and Supportive Care commissioned by National Institute for Health and Clinical Excellence (NICE): April 2008. Available at: http://guidance.nice.org.uk/CG65
(last accessed 20/07/14).
7. Krizanac D, Stratil P, Hoerburger D, et al. Femoro-iliacal artery versus pulmonary artery core temperature measurement during therapeutic hypothermia: an observational study. Resuscitation
8. Sessler DI. Temperature monitoring and perioperative thermoregulation. Anesthesiology
9. Hocker J, Bein B, Bohm R, et al. Correlation, accuracy, precision and practicability of perioperative measurement of sublingual temperature in comparison with tympanic membrane temperature in awake and anaesthetised patients. Eur J Anaesthesiol
10. Fox RH, Solman AJ, Isaacs R, et al. A new method for monitoring deep body temperature from the skin surface. Clin Sci
11. Teunissen LP, Klewer J, de Haan A, et al. Noninvasive continuous core temperature measurement by zero heat flux. Physiol Meas
12. Torossian A. Intraoperative temperature management. Anasthesiol Intensivmed Notfallmed Schmerzther
13. Hooper VD, Andrews JO. Accuracy of noninvasive core temperature measurement in acutely ill adults: the state of the science. Biol Res Nurs
14. Wass CT, Long TR, Deschamps C. Entrapment of a nasopharyngeal temperature probe: an unusual complication during an apparently uneventful elective revision laparoscopic Nissen fundoplication. Dis Esophagus
15. Erickson R. Oral temperature differences in relation to thermometer and technique. Nurs Res
16. Craig JV, Lancaster GA, Taylor S, et al. Infrared ear thermometry compared with rectal thermometry in children: a systematic review. Lancet
17. Eshraghi Y, Nasr V, Parra-Sanchez I, et al. An evaluation of a zero-heat-flux cutaneous thermometer in cardiac surgical patients. Anesth Analg