Secondary Logo

Journal Logo

Obstetric Anesthesiology: Original Clinical Research Report

Core Temperature Monitoring in Obstetric Spinal Anesthesia Using an Ingestible Telemetric Sensor

du Toit, Leon MBChB*; van Dyk, Dominique FCA(SA)*; Hofmeyr, Ross MMed, FCA(SA)*; Lombard, Carl J. PhD(SA)†,‡; Dyer, Robert A. FCA(SA), PhD*

Author Information
doi: 10.1213/ANE.0000000000002326

Perioperative hypothermia (core temperature <36.0°C) is associated with increased postoperative morbidity and length of hospital stay in adults undergoing noncardiac surgery.1,2 The evidence to support any link between maternal perioperative hypothermia and patient outcomes is sparse in the obstetric anesthesia literature.3 The practical difficulty in monitoring core temperature during cesarean delivery (CD) under spinal anesthesia (SA) has been a barrier to investigation of hypothermia and implementation of interventions to ensure normothermia during this procedure. The literature suggests that body temperature is often poorly monitored during SA due to the lack of an accurate, noninvasive, commercially available monitor.4,5 Current knowledge of the thermal insult of obstetric SA is limited to small randomized control trials evaluating snapshots of participant temperatures.3

We expected that SA for elective CD is associated with significant thermoregulatory insult. The study utilized a continuous core temperature recording system to generate high-resolution data on the thermal insult and thermal recovery associated with the procedure.


This was an observational study designed to describe the continuous variable: temperature-over-time. The primary outcome variable was the maximum core temperature decrease after spinal anesthetic injection. A 1.0°C decrease was regarded as clinically significant. The secondary outcome variables were the lowest absolute core temperature, time to lowest temperature, time to recovery of core temperature, hypothermic exposure1 (degree-hours below 37.0°C), and time-weighted hypothermic exposure. An ingestible telemetric temperature sensor was used to record the intestinal temperature of each participant throughout the day of her CD. Before enrollment of participants, approval of the protocol was obtained from the Human Research Ethics Committee of the University of Cape Town and the hospital management of Mowbray Maternity Hospital, Cape Town (the study location). Written informed consent was obtained from all participants the evening before their CD. The manuscript adheres to the STROBE guidelines.

Women scheduled for elective CD were recruited between July 14 and September 11, 2015. Women were screened by direct questioning and review of their medical records for the presence of exclusion criteria. Eligible women were approached consecutively to identify the first willing participant per study day. The temperature recording system allowed for 1 participant to be studied at a time. Exclusion criteria were as follows: known or suspected gastrointestinal obstruction, a history of ongoing swallowing difficulty, known esophageal disease, previous gastrointestinal surgery, any disorder of intestinal motility, expected need for magnetic resonance imaging before the sensor is excreted, presence of a cardiac pacemaker or other implanted electromedical device, American Society of Anesthesiologists Physical Status classification greater than 2, body mass index (BMI) <18 kg·m−2 or weight <37 kg, BMI >40 kg·m−2, a history suggestive of symptomatic thyroid disease, expected major blood loss, and nonconsent.

Participant intestinal temperature was measured as a surrogate for core temperature. An ingested telemetric temperature sensor (CorTemp; HQ Inc, Palmetto, FL) was used to continuously measure intestinal temperature. The CorTemp system was designed and validated to monitor humans performing strenuous activities. It received US Food and Drug Administration clearance as a General Household and Personal Use Device in 1988. The system has a reported accuracy of 0.1°C. The recorder was calibrated by the manufacturer before the study. The single-use, silicone-coated, temperature sensor (size: 22 × 9 mm) was swallowed by each participant before transfer from the ward to the operating room. Although this system is commonly referred to as measuring intestinal temperature, the exact position of the sensor was not assessed in this study and may have been prepyloric at the time of spinal anesthetic injection. The reusable external recorder placed over the right flank of the participant logged the sensor data for later download. The recorder links with the sensor using near-field magnetic induction communication. The recorder was programmed to record sensor temperature every 10 seconds. Recording was continuous for at least 8 hours after spinal anesthetic injection. The sensor was not retrieved after transiting the gastrointestinal tract. During temperature data collection, the recorder data were not visible to the investigator or clinicians.

A standardized patient data collection form was used to record participant demographics, event details, and event timing. Spinal anesthetic levels were determined by the investigator using ethyl chloride spray to elicit cold discrimination. Ambient temperature was recorded with a data logger (iMini; Cryopak, Edison, NJ). Thermal comfort was measured using the Fanger 7-point scale ranging from −3 (blue, cold), through 0 (gray, neutral), to +3 (red, hot). The scale has been adopted by the American Society of Heating and Air-Conditioning Engineers and is also known as the ASHRAE thermal sensation scale. On the evening before their CD, participants were educated on the use of the scale. The assessment was performed at regular intervals before, during, and after SA.

Anesthetic technique was standardized as per institutional practice. An intrathecal injection of 10 mg hyperbaric bupivacaine plus 10 μg fentanyl (2.2 mL total volume) was administered. Left lateral tilt at 20° was used to minimize aortocaval compression. A coload of modified Ringer’s lactate solution 15-20 mL·kg−1 taken from an IV fluid warming cabinet was infused. The warming cabinet temperature was set at 38.0-40.0°C. Management of spinal hypotension was according to conventional institutional practice. IV bolus phenylephrine was the first-line and ephedrine the second-line agent for treatment of spinal hypotension, whenever systolic blood pressure decreased by 15%. Forced air warming was used at the discretion of the primary anesthesiologist. Surgical irrigation is not used at the study center. The investigator was a second anesthesiologist not responsible for patient management. The protocol allowed for shivering to be treated with IV pethidine (25 mg) once the baby was delivered. Intra- and postoperative analgesics were administered at the discretion of the primary anesthesiologist. Typically, this included titrated IV morphine in the operating room or recovery room. Discharge from the postanesthetic recovery room was at the discretion of the primary anesthesiologist and the recovery room staff. The participant could not be discharged from the recovery room while the level of SA remained above T8. The thermal comfort score was not part of the discharge criteria. Patients were allowed oral fluids after discharge from recovery room to the ward. No patient had a solid meal for the 8-hour duration of the study.

Statistical Analysis

Data Handling.

Data from the CorTemp recorder and the iMini data logger were downloaded using software provided by the manufacturers. Before statistical analysis, CorTemp data were processed to remove artifacts. The custom autofilter function was used to identify and delete temperature points that differed more than 0.2°C from the preceding point. This was selected on the principle that a core temperature change of 0.3°C or more in 10 seconds is biologically implausible. Less than 2% of data points were removed by processing before statistical analysis. Visual comparison of raw and processed temperature versus time data plots confirmed that no critical information was lost during processing. CorTemp data were combined into a single plot using the median spline smooth function.

Statistical Methods.

Patient demographics, case characteristics, and temperature recordings were summarized using descriptive statistics. Baseline temperature was defined as the mean temperature for the 5-minute period immediately before transfer onto the operating room table. To quantify hypothermic exposure for each participant, we calculated the area bounded by the 37.0°C line above and the temperature-versus-time curve below. This was expressed as degree-hours of hypothermic exposure. To ensure accuracy, the calculations for hypothermic exposure were performed in duplicate using a custom function for the area under the curve and by performing a pharmacokinetic analysis. Since the exposure duration was a random variable, we also expressed the hypothermic exposure as a time-weighted value by dividing each participant’s area under the curve by the duration below 37°C. Ninety-nine percent and 95% confidence intervals (CIs) were estimated for the primary and secondary outcomes. The large sample approximation method of Hollander and Wolfe6 was used to calculate the CIs for median values.

Microsoft Excel (Professional Plus 2010, version 14; Microsoft Inc, Redmond, WA) was used for data handling and processing, with Stata (version 13; StataCorp LP, College Station, TX) used for median spline smooth and graphical representations. For additional information on the data processing strategy refer to Supplemental Digital Content 1-4, Figure S1,, Figure S2,, Figure S3,, and Figure S4,

Sample size was estimated for a descriptive study of continuous variables. To estimate a mean change in temperature with acceptable accuracy, specifically a 99% CI width of 0.4°C and an expected standard deviation of 0.4°C, would require a sample size of 27 participants. For example, the 99% CI would be 0.8 to 1.2°C for an observed mean change of 1.0°C. The expected standard deviation of 0.4°C for the temperature change was based on a randomized control trial investigating the effect of forced air warming on oral temperatures during CD.6 In the literature on perioperative outcomes, a range of core temperature decreases from 0.4°C to 3.0°C has been associated with adverse patient outcomes.2 Perioperative hypothermia is defined as a core temperature below 36.0°C; this cutoff has been used in perioperative hypothermia research1 and associated with adverse patient outcomes.2 It has been incorporated into published clinical standards and practice guidelines for the prevention of hypothermia.7,8 Accepting that normal core temperature is 37.0°C, a decrease of 1.0°C was chosen as the level of clinical significance for this study. The absolute temperature measurement was not used as the primary outcome, as it is affected by the anatomical site and the timing of temperature measurement, making it difficult to compare results between studies. Specifically, intestinal temperature, as measured in this study, is consistently higher than oral and rectal temperatures in basic physiology and exercise literature.9,10


Thirty-two women were recruited for this study. One participant was unable to swallow the sensor. All other participants reported it easy to swallow the sensor with a small amount of room temperature water (<50 mL). Of the 31 participants who swallowed the sensor, 3 were excluded from analysis. In 1 case, the spinal level was inadequate for surgery, and general anesthesia was required. A second case was delayed to allow for investigation of suspected preeclampsia first detected on arrival in the operating room. In the third case, the recorder did not log temperature data due to battery failure; this was only detected on retrieval of the recorder. All statistical analysis was based on the remaining 28 participants. For further case details of the excluded data sets, see Supplemental Digital Content 5 and 6, Figure S5,, Figure S6,

Demographic data are summarized in Table 1, and case characteristics appear in Table 2. Ambient temperature data from the iMini data logger are presented in Figure 1. Thermal comfort scores, as assessed by the Fanger 7-point scale, are reported in Table 3. Figure 2 combines all the CorTemp data in a median spline smooth of intestinal temperature versus time.

Table 1.
Table 1.:
Study Population Description
Table 2.
Table 2.:
Case Characteristics (n = 28)
Table 3.
Table 3.:
Thermal Comfort Scores (−3 to 3)
Figure 1.
Figure 1.:
Boxplot of ambient temperature data. The boxplot presents the ambient temperature for the 4 stages of monitored perioperative care as percentiles (minimum, p25, p50, p75, maximum). The dots outside the boxes each represent a participant that experienced a mean ambient temperature outside the expected distribution for that stage.
Figure 2.
Figure 2.:
Median spline smooth of CorTemp data. The plot combines the intestinal temperature data of the participants for a restricted time interval from 180 minutes before to 540 minutes after the time of spinal anesthetic injection at 0 minutes.
Table 4.
Table 4.:
Measured Outcome Variables

Temperature outcomes are reported in Table 4. The participants experienced a mean (standard deviation) decrease in intestinal temperature of 1.30°C (0.31); 99% CI, 1.14–1.46; median (interquartile range [IQR]) 1.37°C (1.12–1.48) after spinal anesthetic injection. The temperature nadir was reached after a median (IQR) of 0.96 (0.73–1.32) hours (95% CI, 0.88–1.22). In 21 participants (75%), intestinal temperature continued to decrease after leaving the operating room. However, in all cases the temperature nadir was reached before discharge from the anesthetic recovery area. The median (IQR) lowest intestinal temperature was 35.98 (35.76–36.23)°C. Fourteen of 28 (50%) participants had intestinal temperatures below 36.0°C after spinal injection. Participants spent a median (IQR) of 3.20 (2.35–5.40) hours below the 37.0°C line (95% CI, 2.58–4.53). Twenty-three participants (82%) were discharged from the anesthetic recovery area before their intestinal temperature returned to 37.0°C. In 8 participants (29%), intestinal temperature did not recover to baseline during the monitored period after spinal anesthetic injection. A median (IQR) of 4.59 (3.38–5.92) hours (95% CI, 3.45–5.90) was required for recovery to baseline intestinal temperature in the remaining 20 patients. There was a wide range in time to recovery of intestinal temperature. Participants sustained a median (IQR) of 1.97 (1.00–2.68) degree-hours of hypothermic exposure (95% CI, 1.23–2.45). The median (IQR) number of degrees below 37°C per hour was 0.45 (0.35–0.60) (95% CI, 0.36–0.58).


The performance of intestinal temperature monitoring has been documented in circadian physiology research and during exercise for several decades.9,11 However, it has not been used in the setting of neuraxial anesthesia. Using an intestinal telemetric sensor acceptable to awake patients, this study documented perioperative thermal insult and recovery with high resolution. Although the device carries minimal risk, considerations of cost may limit application to research and specific clinical situations where unpredictable thermoregulatory changes are anticipated and prolonged perioperative monitoring is desirable. While auricular temperatures are frequently measured in the clinical setting using infrared thermometers, it is accepted that these devices do not have the requisite accuracy and precision for research purposes.12,13 Bladder temperature measurement is hampered by the practice of insertion post-SA and the exposure of the bladder during surgery to the operating room environment. All participants experienced a decrease in intestinal temperature after spinal anesthetic injection. In 86% of participants, intestinal temperature decreased by more than 1.0°C. For the primary outcome (mean decrease in temperature), the lower bound of the 99% CI was 1.14°C, which is above the clinically relevant decrease of 1.0°C. The result is in keeping with the effect seen in the control arm (n = 15) of a trial of active patient warming.14 These investigators used serial oral temperature measurements during CD and reported a mean (standard deviation) decrease of 1.3 (0.4)°C after spinal anesthetic injection. Ambient temperature varied between 21.5°C and 23.2°C (Table 2; Figure 1) and was unlikely to have affected the outcome.

Participants reached their intestinal temperature nadir after a median of 1 hour. A recent study on the hypothermic effect of intrathecal morphine during CD reported a mean (standard deviation) time to nadir of 59.5 (17.6) minutes in the control arm (n = 30).15 In the current study, the majority of participants experienced a continued decrease in intestinal temperature after leaving the operating room. Fourteen of 28 participants became hypothermic (core temperature <36°C). Most were discharged from the anesthesia recovery area before their intestinal temperature returned to 37.0°C. Maternal temperature was not part of the discharge criteria, although the protocol asked care providers to consider thermal comfort scores when making the decision to discharge participants. Despite the significant decrease in core temperature, the median thermal comfort score remained zero (neutral) at all but the late afternoon measurement time (Table 3).

There is limited literature investigating recovery from hypothermia in the postoperative period. One retrospective study in 58, 814 patients undergoing general anesthesia for nonobstetric surgery of more than 60 minutes’ duration reported a hypothermia rate of 64% and monitored recovery by accessing electronic medical records. There was a linear correlation between increasing hypothermic exposure and adverse patient outcomes (specifically transfusion requirements and length of hospital stay).1 The recovery pattern to normal core temperature after SA has not previously been described. Although CD is typically of short duration, SA in the present study was associated with hypothermic exposure of 1.84 degree-hours. Several hours were required for intestinal temperature to recover to baseline, and 8 participants did not recover to baseline temperature during the 8-hour monitoring period. These findings challenge current normothermia guidelines and standards recommending active warming only for patients undergoing procedures exceeding 30 minutes (NICE CG65, UK) or 60 minutes (SCIP Inf-10, United States), since the parturient undergoing CD will be subject to a decreasing core temperature for approximately 60 minutes before thermal recovery begins.

Mean maternal rectal temperature at the start of labor has been recorded as 37.1°C.16 At 37.32°C, the mean baseline temperature in the present case series was slightly higher than expected. In healthy volunteers, a separate study found intestinal telemetric temperature to be consistently 0.2°C higher than rectal temperature.9 The higher baseline temperature in the present study may be explained by the temperature measurement site used in this study.

Choice of vasopressor and method of administration could affect thermoregulatory changes. In our study, bolus doses of ephedrine and phenylephrine were used at the discretion of the primary anesthesiologist. Prophylactic phenylephrine infusions may be used to limit temperature loss associated with SA.17

There is a paucity of data on the correlation between temperature measurement at various sites in the presence of a neuraxial block. A limitation of this study is the absence of a second core temperature measurement site for comparison. This study excluded parturients with a BMI above 40 kg·m−2. The mean BMI in this study was 30 kg·m−2. Obesity is known to protect against perioperative hypothermia.18 One investigation in nonobstetric abdominal surgery reported a lower risk of perioperative hypothermia in obese patients with a mean BMI of 31.6 kg·m−2 compared to nonobese patients with a mean BMI of 22.9 kg·m−2.18 Our results may not be generalizable to populations with a BMI that differs significantly from that in our study. We did not investigate whether analgesic agents affected thermoregulation in the study; these data may not be generalizable to settings that use greater doses of intrathecal bupivacaine, preoperative paracetamol, epidural anesthesia, or neuraxial morphine. Additionally, the use of pethidine, indomethacin, and forced air warming was not standardized. Finally, we used IV crystalloid coloading. The fluid was taken directly from the fluid warming cabinet (set at 38–40°C) at the time the participant was moved into the operating room. We did not use in-line fluid warming and cannot report on the actual temperature at which fluid was infused into the participants. Results may differ if different volumes or temperatures of IV fluid are used. Meta-analyses suggest that fluid warming is an effective method for decreasing perioperative hypothermia during CD.3


In this study, the utility of telemetric intestinal temperature monitoring was demonstrated in awake patients undergoing regional anesthesia. Elective CD under SA with warmed fluid coloading was associated with a significant decrease in core temperature, and 50% of parturients became hypothermic (<36.0°C). Core temperature derangement extended several hours beyond the surgical time. Future studies should investigate the value of temperature recovery patterns to predict patient outcomes. Outcomes related to maternal wound healing, neonatal well-being, and readmission rates may be of specific interest in the obstetrics population.


Name: Leon du Toit, MBChB.

Contribution: This author helped design and conduct the study, collect and analyze the data, and prepare the manuscript.

Conflicts of Interest: L. du Toit has a speaker agreement with 3M South Africa as a topic expert in perioperative thermoregulation.

Name: Dominique van Dyk, FCA(SA).

Contribution: This author helped design and conduct the study, collect the data, and prepare the manuscript.

Conflicts of Interest: None.

Name: Ross Hofmeyr, MMed, FCA(SA).

Contribution: This author helped design and conduct the study, collect and analyze the data, and prepare the manuscript.

Conflicts of Interest: None.

Name: Carl J. Lombard, PhD(SA).

Contribution: This author helped design the study, analyze the data, and prepare the manuscript.

Conflicts of Interest: None.

Name: Robert A. Dyer, FCA(SA), PhD.

Contribution: This author helped design and conduct the study, collect and analyze the data, and prepare the manuscript.

Conflicts of Interest: None.

This manuscript was handled by: Jill M. Mhyre, MD.


1. Sun Z, Honar H, Sessler DI, et al. Intraoperative core temperature patterns, transfusion requirement, and hospital duration in patients warmed with forced air. Anesthesiology. 2015;122:276–285.
2. Doufas AG. Consequences of inadvertent perioperative hypothermia. Best Pract Res Clin Anaesthesiol. 2003;17:535–549.
3. Sultan P, Habib AS, Cho Y, Carvalho B. The effect of patient warming during caesarean delivery on maternal and neonatal outcomes: a meta-analysis. Br J Anaesth. 2015;115:500–510.
4. Frank SM, Nguyen JM, Garcia CM, Barnes RA. Temperature monitoring practices during regional anesthesia. Anesth Analg. 1999;88:373–377.
5. Arkiliç CF, Akça O, Taguchi A, Sessler DI, Kurz A. Temperature monitoring and management during neuraxial anesthesia: an observational study. Anesth Analg. 2000;91:662–666.
6. Hollander M, Wolfe E. Nonparametric Statistical Methods. 1973.New York, NY: John Wiley.
7. Scott AV, Stonemetz JL, Wasey JO, et al. Compliance with Surgical Care Improvement Project for Body Temperature Management (SCIP Inf-10) is associated with improved clinical outcomes. Anesthesiology. 2015;123:116–125.
8. Torossian A, Bräuer A, Höcker J, Bein B, Wulf H, Horn EP. Preventing inadvertent perioperative hypothermia. Dtsch Arztebl Int. 2015;112:166–172.
9. Edwards B, Waterhouse J, Reilly T, Atkinson G. A comparison of the suitabilities of rectal, gut, and insulated axilla temperatures for measurement of the circadian rhythm of core temperature in field studies. Chronobiol Int. 2002;19:579–597.
10. Waterhouse J, Drust B, Weinert D, et al. The circadian rhythm of core temperature: origin and some implications for exercise performance. Chronobiol Int. 2005;22:207–225.
11. Byrne C, Lim CL. The ingestible telemetric body core temperature sensor: a review of validity and exercise applications. Br J Sports Med. 2007;41:126–133.
12. Imamura M, Matsukawa T, Ozaki M, Sessler DI, Nishiyama T, Kumazawa T. The accuracy and precision of four infrared aural canal thermometers during cardiac surgery. Acta Anaesthesiol Scand. 1998;42:1222–1226.
13. Lawson L, Bridges EJ, Ballou I, et al. Accuracy and precision of noninvasive temperature measurement in adult intensive care patients. Am J Crit Care. 2007;16:485–496.
14. Butwick AJ, Lipman SS, Carvalho B. Intraoperative forced air-warming during cesarean delivery under spinal anesthesia does not prevent maternal hypothermia. Anesth Analg. 2007;105:1413–1419.
15. Hui CK, Huang CH, Lin CJ, Lau HP, Chan WH, Yeh HM. A randomised double-blind controlled study evaluating the hypothermic effect of 150 microg morphine during spinal anaesthesia for caesarean section. Anaesthesia. 2006;61:29–31.
16. Schouten FD, Wolf H, Smit BJ, Bekedam DJ, de Vos R, Wahlen I. Maternal temperature during labour. BJOG. 2008;115:1131–1137.
17. Ro Y, Huh J, Min S, et al. Phenylephrine attenuates intra-operative hypothermia during spinal anaesthesia. J Int Med Res. 2009;37:1701–1708.
18. Fernandes LA, Braz LG, Koga FA, et al. Comparison of peri-operative core temperature in obese and non-obese patients. Anaesthesia. 2012;67:1364–1369.

Supplemental Digital Content

Copyright © 2017 International Anesthesia Research Society