Women undergoing cesarean delivery are a vulnerable but often overlooked population in guidelines for perioperative temperature management. Inadvertent perioperative hypothermia, defined as the unintentional cooling of core temperature to below 36°C during surgery,1 has detrimental physiological effects that have been well studied in the nonpregnant population. These include increased blood loss,2 higher wound infection rates,3 immune function suppression,4 prolonged drug action,5 , 6 increased duration of recovery stays7 and increased hospital stay,3 increased costs,8 shivering9 , 10 and, importantly, discomfort. Impacts on neonatal outcomes, such as temperature at birth,11 umbilical vein,11 arterial pH,12 and Apgar scores12 have been demonstrated in some studies as well as a relationship between neonatal hypothermia and hypoglycemia.13 Hypothermia is often undetected until the postoperative phase, causing significant disruption to postoperative care, as well as maternal-newborn bonding and feeding, while rewarming is applied.
Rates of perioperative hypothermia among women undergoing cesarean under spinal anesthesia have been estimated as being as high as between 32%14 and 80%.15 In addition, perioperative hypothermia appears to be exacerbated by intrathecal morphine.14 , 16–18 Since, in clinical practice, spinal anesthesia, commonly utilizing intrathecal morphine, often comprises standard care for this population, it is important that health care providers establish preemptive measures to reduce the occurrence of hypothermia, shifting the emphasis from treatment to prevention for all women undergoing caesarean delivery.
Guidelines for the general adult population advise 30 minutes of preoperative warming.1 A shorter period may be more clinically acceptable and practical, while still reducing intraoperative core temperature decline. Horn et al19 tested passive warming versus 10, 20, or 30 minutes of preoperative forced air warming, in a randomized controlled trial of 200 patients undergoing laparoscopic surgery under general anesthesia, finding that 10 minutes of preoperative warming resulted in significantly improved core temperature. An optimum warming period of 20 minutes was recommended where clinically possible.19 Fifteen minutes of preoperative warming before induction of epidural anesthesia, plus continuation of forced air warming during surgery, has also shown efficacy at reducing hypothermia in a population of women receiving epidural anesthesia but who did not receive opioids.11
This single-blinded, randomized controlled trial compared the effect of a period of 20 minutes of preoperative forced air warming alongside intraoperative intravenous (IV) fluid warming with usual clinical care (IV fluid warming and no preoperative forced air warming) in women receiving intrathecal morphine during elective cesarean delivery on the primary outcome of maternal temperature change from baseline to the end of the procedure. Secondary outcomes—for exploratory analysis only—included temperature decline assessed over time, hypothermia, maternal thermal comfort, mean arterial pressure (MAP), shivering, agreement between aural canal and bladder temperature measurements, neonatal axillary temperature at birth, Apgar scores at 1 and 5 minutes, skin-to-skin contact at birth, breastfeeding at birth and upon discharge from hospital, and incidence of wound complications.
Women with singleton pregnancies booked for elective cesarean delivery at term under spinal anesthesia with intrathecal morphine were enrolled in this pragmatic, single-blinded randomized controlled study, after hospital and university ethics approval, and informed consent. Exclusion criteria included known allergy to morphine, known impaired thermoregulation or thyroid disorders, vascular disease, or poor cutaneous perfusion, American Society of Anesthesiologists score >II, history of preeclampsia or eclampsia, planned intensive care unit admission, tympanic membrane/aural canal that was not visible on otoscopy and baseline temperature ≥37°C. The study was registered on the Australia and New Zealand Clinical Trials Registry (Trial No: 367160, registered at http://www.ANZCTR.org.au/ on October 10, 2014 by the principal investigator Judy Munday).
After informed consent, and otoscopy, participants were randomly assigned to either the control or the intervention group. The randomization schedule was computer-generated, utilizing fixed-size blocks (at http://www.randomisation.com) of 5 per block and placed within sequentially numbered opaque envelopes. An independent coordinator generated the allocation sequence, and allocation to groups was concealed from the blinded outcome assessor.
Participants in the control group received usual care consisting of no active warming during the admission and preoperative period. Participants in the intervention group received 20 minutes of full-body preoperative warming in which perioperative midwives independent of the study applied a forced air warming device (Cocoon; Care Essentials Pty Ltd, North Geelong, Victoria, Australia) set to 43°C in the preoperative waiting area, before entering the operating room (OR) for induction of spinal anesthesia. The investigator remained in the operating theater and did not access the preoperative waiting area to ensure blinding. A delay of more than 20 minutes between the end of the preoperative warming and transfer to theater was considered a protocol deviation. Patients were monitored during the intervention to assess for adverse side effects related to warming, such as diaphoresis or nausea and vomiting.
All women received IV fluid warming (compound sodium lactate) warmed to 38.5°C (via Biegler fluid warmer; Biegler GmbH, Mauerbach, Austria), were covered with a warmed cotton blanket and surgical drapes, and received standardized intraoperative anesthetic medication and IV fluids. After induction of spinal anesthesia, a temperature sensing indwelling urinary catheter (Mon-a-Therm; Medtronic, Minneapolis, MN) was inserted. All patients received spinal anesthesia (or combined spinal-epidural anesthesia with no opioids via the epidural catheter) in the sitting position at the L3-4 interspace, with 2.2 to 2.4 mL hyperbaric 0.5% bupivacaine, intrathecal morphine 100 μg, and intrathecal fentanyl 15 to 20 μg. Block height was tested using ice, and the procedure commenced once a sensory block above T4 was achieved. IV carbetocin 100 µg was administered at delivery. Rectal paracetamol 1 g and diclofenac 100 mg were administered at the end of the procedure. Variations to the protocol were documented and recorded. Ambient preoperative holding bay and OR temperature was recorded via thermostat. At the end of the procedure, all patients were covered with a warmed cotton blanket and a reflective foil blanket, before transfer to the postanesthesia care unit (PACU). If temperature decline, or temperature ≤ 35.5°C (as per institutional guidelines), shivering, or cold discomfort was experienced in PACU, further warmed blankets were offered and/or forced air warming commenced as per routine care.
Maternal temperature was measured using both a calibrated Genius aural canal thermometer (Medtronic, Minneapolis, MN) (cited as reading a mean of −0.4°C less than pulmonary artery measurement)20 and Mon-a-Therm indwelling urinary catheterization (cited as providing accuracy to within 0.1°C of pulmonary artery measurement)21 at the following time points: baseline, prespinal, postspinal, every 15 minutes, and at the end of the procedure, on arrival to PACU, then every 15 minutes until ready for discharge from PACU. Maternal thermal comfort was measured using a 100-mm visual analog scale (VAS), used in a number of studies measuring patient thermal comfort.22–26 Shivering was assessed via a 3-point scale used in previous studies in this population,27 , 28 in the absence of a validated shivering scale. MAP was measured at baseline, prespinal, postspinal and at the end of the procedure, however only baseline, prespinal and postspinal measurements were analyzed due to the individual difference in the use of vasopressors in response to clinical need, which was not specified in the anesthetic protocol. An independent midwife assessed neonatal axillary temperature, and Apgar scores, at 1 and 5 minutes after birth. Duration of skin-to-skin at birth, feeding intention, breastfeeding, and timing of feed at birth were recorded, as well as breastfeeding at 10 days postnatally which was determined retrospectively from the Universal Postnatal Contact Survey. Wound infection and dehiscence upon hospital discharge, and patient concerns with the postnatal wound (at 10 days) were also determined via chart review. Demographic data collection included maternal age, parity, and gravidity. Surgical variables such as intraoperative blood loss, volume of IV fluid infusion, anesthetic medication (including any that deviate from the agreed protocol) duration of procedure, preoperative, and OR ambient temperature were also recorded. This article adheres to the CONSORT criteria for the reporting of RCTs.29
Descriptive statistics were generated to summarize sample characteristics and hypothermia prevalence. Data are expressed as means and standard deviations, median and range, or as frequencies and percentages as indicated. A general linear model was used to assess the primary outcome of aural temperature change between groups, with adjustment for baseline temperature and surgery duration.
An exploratory analysis of secondary outcomes was undertaken, using linear mixed-model analysis (to allow for fixed effects of baseline temperature, time and group, and a random intercept for repeated measures) for aural temperature decline from immediately after spinal insertion until the end of the procedure. Linear mixed-model analysis was also used to assess thermal comfort between groups at repeated time points. Pearson χ2test of independence with continuity correction was used to analyze hypothermia incidence, shivering and neonatal outcomes, with the Mann-Whitney U test used for nonparametric MAP data. Bland-Altman analysis (using MedCalc; MedCalc Software, Ostend, Belgium) examined agreement30 between aural canal and bladder temperature, and to provide a means to establish the accuracy of the aural canal measurements used for the primary analysis. SPSS software (version 22) was utilized for all other data analysis: P < .05 was considered statistically significant for the primary outcome, and P < .01 for the secondary outcomes.
All analyses were performed on the intention-to-treat population, which included all participants in the groups to which they were assigned, irrespective of protocol deviations.
A required sample size of 15 participants in each group was calculated, based on a repeated measures design with the initial temperatures being the same and the temperature decline being 0.4°C greater in the unwarmed group than the warmed group when measured 45 minutes after commencement of surgery. A standard deviation of 0.4°C was used in the calculation, based on the data reported by Chung et al.31 A type I error rate of 0.05 and a power of 90% were specified. The sample size was inflated from a total of 30 to a total of 50 to allow for attrition.
Patients were enrolled in the study between February 2015 and February 2016. All 50 patients completed the study (Figure 1), however, there were 13 protocol deviations: 7 in the preoperative warming group and 6 in the control group. Three patients in the preoperative warming group had suspected bladder injury and received methylene blue dye; from the point of this occurrence, bladder temperature for these patients was disregarded. Maternal baseline characteristics, as well as surgical and anesthetic variables, were similar across treatment groups apart from baseline temperature (Table 1). In the warming group, 4 patients experienced sweating. Due to this, 1 patient ceased the warming period 2 minutes early by request. No other adverse events related to the warming intervention were reported.
Intention-to-treat analysis revealed no significant difference in aural temperature change from baseline to the end of the procedure between groups: F (1, 47) = 1.2, P = .28, partial eta squared = .03) (Table 2).
Although the preoperative warming group experienced higher intraoperative mean temperatures, from the insertion of spinal anesthesia until 30 minutes, this was not statistically significant and by 45 minutes temperatures in both groups were the same, when analyzed using linear mixed-model analysis, and controlling for baseline temperature (Figure 2). There were no statistically significant differences in hypothermia incidence between the groups (see Table 2).
Maternal thermal comfort did not differ between groups at any time point (Table 3). There was no clinically significant differences in MAP between groups or differences in postoperative outcomes. No patients experienced wound infection or dehiscence, assessed at discharge, in either group. On follow-up, 1 patient in the control group had a postnatal wound infection (10 days postpartum). Neonatal outcomes were also similar between groups.
Bland-Altman analysis indicated that, apart from one outlier, differences between aural canal (Genius) and bladder (Mon-a-Therm) temperature measurement devices appear to be consistent as temperature changes. The mean difference between devices was 0.04°C (SD 0.25). The limits of agreement ranged from 0.93°C to 0.86°C, however, only 2-paired measurements exceeded a difference of 0.5°C, conventionally cited as a clinically acceptable measurement variation (Figure 3).
Twenty minutes of full-body preoperative warming, before spinal anesthesia with intrathecal morphine for cesarean delivery, does not result in a significant decrease in intraoperative maternal temperature decline. Despite the decreased core to periphery heat gradient that is proposed to result from preoperative warming,11 by the end of the procedure, both groups experienced temperature decline with similar end of procedure temperatures.
The results of our study contrast with Horn et al’s findings that 15 minutes of upper body preoperative warming 43°C, continued intraoperatively, resulted in over 1°C difference between control and intervention group at the end of surgery, in favor of warming.11 However, ambient temperature was higher in Horn’s study, and surgical duration was slightly less than our study (Table 1). In addition, their population received epidural anesthesia with no opioids, which may contribute to the marked differences between their warmed and unwarmed groups.11Similarly, De Bernardis et al32 also found temperature declined less when women received preoperative warming that continued intraoperatively (versus the control group receiving IV fluids warmed only to 37°C). All patients received spinal anesthesia with 80 µg intrathecal morphine.
When considered in conjunction with the results from other comparable studies,11 , 31 , 32 several key variations appear important: the use (and dose) of intrathecal morphine, surgical factors including ambient temperature and surgical duration, and the use of preoperative strategies that are both multimodal and continued intraoperatively. Although it has been proposed that increased heat loss may occur with intrathecal morphine due to cephalic spread decreasing the temperature set point, the reasons for this remain unconfirmed. Given current evidence, it cannot be said with certainty that intrathecal morphine blunts the response to warming.
Both groups in our study received IV fluid warming (as per National Institute for Health and Care Excellence guidelines that fluids of ≥500 mL should be warmed to 37°C or more),1 in the form of crystalloid coloading at the time of spinal anesthesia, as is usual care in our institution. This may help to maintain temperature during the period of intravascular volume shift that occurs during spinal anesthesia.17 However, it is evident that IV fluid warming alone is not sufficient to prevent hypothermia in most patients, as indicated by the incidence of hypothermia in the control group in this study, again further suggesting that multimodal interventions are likely to be of the most benefit.17
Both researchers and clinicians have questioned whether forced air warming is tolerable or practical for obstetric patients.33 , 34 Although this study did not assess tolerability in any meaningful way beyond recording adverse events related to warming, or patient symptoms of sweating, nausea, or discomfort, it appears that patients largely found the duration and 43°C setting tolerable. Only 1 patient asked to cease the intervention 2 minutes early, which compares favorably with results from Fallis et al’s23 study of upper-body intraoperative forced air warming, where 14 patients decreased the temperature of the forced air warmer from 43°C to a lower setting. Research into obstetric patients’ preferences for warming interventions may be warranted.
The intensity and incidence of shivering may indicate the severity of hypothermia. In our study, no preoperatively warmed patients, as opposed to 3 patients in the control group, experienced severe shivering. Warmed IV fluids were found to be effective at reducing shivering in recent meta-analysis.35 Nonthermogenic factors, such as catecholamines resulting from pain or anxiety, may also contribute to shivering,36 , 37 and larger studies of the impact of combined warming strategies incorporating preoperative warming upon shivering are warranted.
This study was designed to test a pragmatic approach to warming by using a short preoperative full-body warming regime, based on evidence of the optimal duration of effective preoperative warming.11 , 19 Warming was applied in the preoperative waiting area before women entered the OR. Our study protocol specified no greater than a 20-minute time delay between the end of the warming regime and entry to the OR, but some participants experienced longer delays, which reduced power of the study to detect a difference between groups. The benefits of preoperative whole-body warming may be evident if warming is continued into the OR, through induction of neuraxial anesthesia, to the commencement of the surgical skin preparation.11 , 31
The use of aural canal thermometry is not without controversy, and disagreement exists as to the accuracy of this method. However, this method is not invasive and therefore may be more acceptable to patients. Our study used measures to assess and increase the reliability of aural canal thermometry, including checking the visibility of the tympanic membrane via otoscopy, using 1 outcome assessor, and using an additional measurement of bladder temperature (cited as providing an acceptable near-core measurement). Temperature decline was assessed until the end of the procedure, while other studies also report temperature in PACU.38 Temperatures measured after arrival in the PACU were not analyzed because some patients received postoperative warming interventions; any measurements beyond the arrival temperature into PACU would therefore be confounded.
In conclusion, based on the intention-to-treat results of this study, a short period of preoperative forced air warming, in conjunction with intraoperative IV fluid warming, is not effective at preventing temperature decline in women that receive intrathecal morphine for cesarean delivery. These results do not correspond with the benefits reported for women undergoing cesarean delivery who have received preoperative warming that continues intraoperatively or have not received intrathecal opioids. However, because intrathecal opioid administration is a common practice in many institutions, effective methods of preventing perioperative hypothermia in this population warrant further exploration; combined warming interventions are likely to be of the most benefit.
The authors would like to acknowledge the assistance of Annie McArdle (Registered Midwife, Mater Health Services) for organizational assistance before data collection, Dr Simon Maffey for assistance in developing the anesthetic protocol, and the perioperative midwives and staff of Mater Health Services, Brisbane.
Name: Judy Munday, PhD.
Contribution: This author is identified as the principal and corresponding author, based on the criteria specified by the International Committee of Medical Journal Editors and was responsible for initiation of the project, study design, data collection, data analysis, write up of results and writing of the final manuscript.
Name: Sonya Osborne, PhD.
Contribution: This author helped with the research protocol, study design, and review of final manuscript.
Name: Patsy Yates, PhD.
Contribution: This author helped with the research protocol, study design, and review of final manuscript.
Name: David Sturgess, MBBS, PhD.
Contribution: This author helped with the study protocol and review of results and of the final manuscript.
Name: Lee Jones, AStat, BSc, Hon.
Contribution: This author helped with the statistical review and review of the final manuscript.
Name: Edward Gosden, MSc.
Contribution: This author reviewed the protocol and helped with the statistical review and review of the final manuscript.
This manuscript was handled by: Jill M. Mhyre, MD.
1. NCCNSC Clinical Practice Guideline. The Management of Inadvertent Perioperative Hypothermia in Adults. 2008.London: National Institute for Clinical Health and Excellence.
2. Rajagopalan S, Mascha E, Na J, Sessler DI. The effects of mild perioperative hypothermia on blood loss and transfusion requirement. Anesthesiology. 2008;108:71–77.
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. 1996;334:1209–1215.
4. Beilin B, Shavit Y, Razumovsky J, Wolloch Y, Zeidel A, Bessler H. Effects of mild perioperative hypothermia on cellular immune responses. Anesthesiology. 1998;89:1133–1140.
5. Heier T, Caldwell JE. Impact of hypothermia on the response to neuromuscular blocking drugs. Anesthesiology. 2006;104:1070–1080.
6. Leslie K, Sessler DI, Bjorksten AR, Moayeri A. Mild hypothermia alters propofol pharmacokinetics and increases the duration of action of atracurium. Anesth Analg. 1995;80:1007–1014.
7. Lenhardt R, Marker E, Goll V, et al. Mild intraoperative hypothermia prolongs postanesthetic recovery. Anesthesiology. 1997;87:1318–1323.
8. Mahoney CB, Odom J. Maintaining intraoperative normothermia: a meta-analysis of outcomes with costs. AANA J. 1999;67:155–163.
9. Liu WH, Luxton MC. The effect of prophylactic fentanyl on shivering in elective caesarean section under epidural analgesia. Anaesthesia. 1991;46:344–348.
10. Roy JD, Girard M, Drolet P. Intrathecal meperidine decreases shivering during cesarean delivery under spinal anesthesia. Anesth Analg. 2004;98:230–234.
11. Horn EP, Schroeder F, Gottschalk A, et al. Active warming during cesarean delivery. Anesth Analg. 2002;94:409–414.
12. Yokoyama K, Suzuki M, Shimada Y, Matsushima T, Bito H, Sakamoto A. Effect of administration of pre-warmed intravenous fluids on the frequency of hypothermia following spinal anesthesia for cesarean delivery. J Clin Anesth. 2009;21:242–248.
13. Baker B, Lawson R. Maternal and newborn outcomes related to unplanned hypothermia in scheduled low-risk cesarean delivery births newborn and Infant Nursing Rev. 2012;12:75–77.
14. Hess PE, Snowman CE, Wang J. Hypothermia after cesarean delivery and its reversal with lorazepam. Int J Obstet Anesth. 2005;14:279–283.
15. Chakladar A, Dixon M, Harper C. Warming mattress to prevent inadvertent perioperative hypothermia and shivering during elective caesarean section. Br J Anaesth. 2011;107:290P–291P.
16. Halloran OJ. Warming our cesarean section patients: why and how? J Clin Anesth. 2009;21:239–241.
17. Cobb B, Cho Y, Hilton G, Ting V, Carvalho B. Active warming utilizing combined iv fluid and forced-air warming decreases hypothermia and improves maternal comfort during cesarean delivery: a randomized control trial. Anesth Analg. 2016;122:1490–1497.
18. 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.
19. Horn EP, Bein B, Böhm R, Steinfath M, Sahili N, Höcker J. The effect of short time periods of pre-operative warming in the prevention of peri-operative hypothermia. Anaesthesia. 2012;67:612–617.
20. Robinson J, Charlton J, Seal R, Spady D, Joffres MR. Oesophageal, rectal, axillary, tympanic and pulmonary artery temperatures during cardiac surgery. Can J Anesth. 1998;45:317–323.
21. Russell SH, Freeman JW. Comparison of bladder, oesophageal and pulmonary artery temperatures in major abdominal surgery. Anaesthesia. 1996;51:338–340.
22. Buggy DJ, Crossley AW. Thermoregulation, mild perioperative hypothermia and postanaesthetic shivering. Br J Anaesth. 2000;84:615–628.
23. Fallis WM, Hamelin K, Symonds J, Wang X. Maternal and newborn outcomes related to maternal warming during cesarean delivery. J Obstet Gynecol Neonatal Nurs. 2006;35:324–331.
24. Fossum S, Hays J, Henson MM. A comparison study on the effects of prewarming patients in the outpatient surgery setting. J Perianesth Nurs. 2001;16:187–194.
25. Wilson L, Kolcaba K. Practical application of comfort theory in the perianesthesia setting. J Perianesth Nurs. 2004;19:164–173.
26. Sessler DI, Ponte J. Shivering during epidural anesthesia. Anesthesiology. 1990;72:816–821.
27. Woolnough M, Allam J, Hemingway C, Cox M, Yentis SM. Intra-operative fluid warming in elective caesarean section: a blinded randomised controlled trial. Int J Obstet Anesth. 2009;18:346–351.
28. Saito T, Sessler DI, Fujita K, Ooi Y, Jeffrey R. Thermoregulatory effects of spinal and epidural anesthesia during cesarean delivery. Reg Anesth Pain Med. 1998;23:418–423.
29. Ioannidis JP, Evans SJ, Gøtzsche PC, et al; CONSORT Group. Better reporting of harms in randomized trials: an extension of the CONSORT statement. Ann Intern Med. 2004;141:781–788.
30. Kirkwood B, Sterne J. Essential Medical Statistics. 2003.2nd ed. Malden, MA: Blackwell Science Ltd.
31. Chung SH, Lee BS, Yang HJ, et al. Effect of preoperative warming during cesarean section under spinal anesthesia. Korean J Anesthesiol. 2012;62:454–460.
32. de Bernardis RC, Siaulys MM, Vieira JE, Mathias LA. Perioperative warming with a thermal gown prevents maternal temperature loss during elective cesarean section. A randomized clinical trial. Braz J Anesthesiol. 2016;66:451–455.
33. Petsas A, Vollmer H, Barnes R. Peri-operative warming in Caesarean sections. Anaesthesia. 2009;64:921–922.
34. Chakladar A, Harper CM. Peri-operative warming in caesarean sections: guidance would be NICE. Anaesthesia. 2010;65:212–213.
35. Munday J, Hines S, Wallace K, Chang A, Gibbons K, Yates P. The clinical effectiveness of interventions to assist perioperative temperature management for women undergoing cesarean section: a systematic review. JBI Database of Systematic Reviews and Implementation Reports. 2013;11:45–111.
36. Chan VW, Morley-Forster PK, Vosu HA. Temperature changes and shivering after epidural anesthesia for cesarean section. Reg Anesth. 1989;14:48–52.
37. Alfonsi P. Postanaesthetic shivering. Epidemiology, pathophysiology and approaches to prevention and management. Minerva Anestesiol. 2003;69:438–442.
© 2018 International Anesthesia Research Society
38. 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.