Secondary Logo

Journal Logo


Warming before and after epidural block before general anaesthesia for major abdominal surgery prevents perioperative hypothermia

A randomised controlled trial

Horn, Ernst-Peter; Bein, Berthold; Broch, Ole; Iden, Timo; Böhm, Ruwen; Latz, Svea-Kristina; Höcker, Jan

Author Information
European Journal of Anaesthesiology: May 2016 - Volume 33 - Issue 5 - p 334-340
doi: 10.1097/EJA.0000000000000369
  • Free



Epidural analgesia (EDA) is a standard procedure in various abdominal surgical procedures to improve intraoperative and postoperative pain management. However, in patients receiving combined general and epidural anaesthesia, intraoperative core temperature decreases more rapidly than in patients receiving general anaesthesia alone.1,2

Neuraxial analgesia reduces the vasoconstriction threshold and causes a functional lower-body sympathectomy which results in a core-to-peripheral redistribution of body heat followed by core temperature decrease.3 Prewarming, meaning forced-air warming before induction of general and/or regional anaesthesia or analgesia, markedly increases peripheral tissue heat content.4

It has been demonstrated that prewarming with forced-air warming systems is effective in reducing hypothermia in surgical patients.5 However, the effects of prewarming before and after EDA prior to major abdominal surgery have not been investigated systematically. In pregnant women scheduled for caesarean section under EDA, prewarming was found to be effective as body core temperature was significantly higher in the actively warmed group compared with the unwarmed group.6

Current clinical guidelines for prevention of inadvertent perioperative hypothermia recommend prewarming for both neuraxial anaesthesia and general anaesthesia.7,8 Recently, we demonstrated that only 20 (or even 10) min of prewarming mostly prevented patients from perioperative hypothermia when they underwent surgery of 30–90 min under general anaesthesia.9 However, to our knowledge, no published study has investigated the effects of prewarming on the incidence of postoperative hypothermia in combined general anaesthesia and EDA.

We hypothesised that warming patients before and after the insertion of an epidural catheter to provide analgesia for major abdominal surgery would decrease the incidence of postoperative hypothermia. The aim of this study was to prevent hypothermia because of EDA implemented for major abdominal surgery by warming the patients during two periods, before and after the insertion of an epidural catheter.


This study was approved by the Institutional Review Board of the Christian-Albrechts-University Kiel (A133/12) (Chair: Professor M.H. Mehdorn) on 4 September 2012 and conducted between January 2013 and August 2014. After obtaining written informed consent, we included 99 adult patients scheduled for elective major abdominal surgery under combined general anaesthesia and EDA with an expected duration of surgery of at least 120 min. Patients were excluded if they were under 18 years of age, classified as American Society of Anesthesiologists’ physical status 4 or higher or if patients refused EDA.

The primary outcome measure was the incidence of postoperative hypothermia defined as a body core temperature less than 36°C on arrival at ICU. The secondary outcome measure was the incidence of shivering. In addition, the perioperative change in the body core temperature was recorded.

On the day of surgery, patients were premedicated with midazolam 3.75 to 7.5 mg orally at the discretion of the attending anaesthesiologist. About 60 min before the expected start of surgery, patients were transferred to the preoperative care unit where baseline measurements were obtained before administration of midazolam. A cannula was inserted in a peripheral vein in the back of the hand or elbow flexure and an infusion of Ringer's solution was started at a rate of 500 ml h−1. All intravenous fluids were warmed by an active fluid warming device (Level 1, Hot Line, Smiths Medicals, Rockland, Massachusetts, USA) to 41°C up to the end of surgery. Routine haemodynamic monitoring including ECG, SpO2 and noninvasive blood pressure measurement at 3 to 5-min intervals was established at this point.

A 16-gauge Tuohy needle was inserted between the eighth and ninth (T8/9) or the ninth and tenth (T9/10) thoracic interspaces with the patient in the sitting position. The loss-of-resistance technique was used to identify the epidural space. A catheter was inserted 4 to 5 cm into the epidural space, and 2 to 3 ml of ropivacaine 0.75% were injected as a test dose to exclude intrathecal positioning of the catheter. Following a 10-min observation period, an additional bolus of 6 to 8 ml of ropivacaine 0.2% was injected through the epidural catheter, depending on the height of the patient.

Patients were randomly assigned to one of three treatment groups: passive insulation but no active warming of the skin before the start of the surgery (‘no warming’), active preoperative forced-air warming for 15 min after epidural catheter insertion and application of the test dose but before injection of 6 to 8 ml of ropivacaine 0.2% (‘warming after epidural catheter placement’) or active preoperative forced-air warming for 15 min before insertion of the epidural catheter and for 15 min after insertion of the epidural catheter and administration of the test dose but before injection of 6 to 8 ml of ropivacaine 0.2% (‘warming before and after epidural catheter placement’).

Patient recruitment was achieved during the preoperative visit by the attending anaesthesiologist. Randomisation was performed by an uninvolved nurse on arrival at the preoperative care unit by rolling a dice; a roll of 1 or 4 resulted in enrolment to the ‘no warming’ group, 2 or 5 meant allocation to the ‘warming after EDA’ group and 3 or 6 resulted in allocation to the ‘warming before and after EDA’ group.

Prewarming was performed at the preoperative care unit with a forced-air blanket (Level 1 Snuggle Warm Upper Body Blanket, Smiths Medicals, Rockland, Massachusetts, USA) positioned over the whole body of the patient and covered by a cotton blanket. A Level 1 Equator warmer (Smiths Medicals, Rockland, Massachusetts, USA) was set to ‘high level’ (44°C) during the warming period for both the study groups receiving active warming. To guarantee correct duration of the warming period, a countdown timer (Type 8G1HZ1A, EverFlourish Europe, Friedrichsthal, Germany) was used.

During the warming procedure, patients were asked every 5 min about their thermal comfort; if they felt overheated, the warmer was lowered to 40°C. When active warming was stopped, the blanket was left on the patient's skin without blowing air. Preoperative, intraoperative and postoperative ambient temperature was maintained near 23°C.

After the warming procedure, patients were transferred to the operating theatre. General anaesthesia was induced using propofol 1.5 to 2.5 mg kg−1 and sufentanil 0.2 μg kg−1, and was maintained with sevoflurane (0.7 to 1.0 minimum alveolar concentration) by an anaesthesiologist blinded to the warming randomisation. Atracurium (0.5 mg kg−1) was used for muscle relaxation and an endotracheal tube was inserted. A fresh gas flow of 0.5 l min−1 was established with a minimum inspired oxygen concentration of 40% and an end-tidal CO2 concentration of 4.7 to 5.3 kPa (35 to 40 mmHg). An electric GemStar pump (Hospira Inc, Lake Forest, Illinois, USA) was connected to the epidural catheter for continuous epidural administration of ropivacaine 0.2% at a rate of 6 ml h−1 until the end of surgery. If signs of hypovolaemia in combination with unstable haemodynamics such as tachycardia (heart rate >100 bpm) or low mean arterial blood pressure (MAP <60 mmHg) were observed, the epidural infusion was stopped at the discretion of the attending anaesthesiologist.

In all groups, the patients’ upper bodies were actively warmed during the surgical procedure using a Level 1 Equator warmer (Smiths Medicals, Rockland, Massachusetts, USA) which was set to ‘high level’ (44°C).

During the observation period, peripheral oxygen saturation, heart rate and MAP (IntelliVue MP50; Philips, Boeblingen, Germany) were recorded. Core temperature was continuously measured at the tympanic membrane using a tympanic temperature sensor (YSI 400; Smiths Medicals, Rockland, Massachusetts, USA). The aural probes were inserted by the patients until they felt the thermocouple touch the tympanic membrane. Appropriate placement of the sensor was confirmed when patients easily detected a gentle rubbing of the attached wire. Then, the aural canal was occluded with cotton wool and taped in place. The first tympanic membrane temperature was assessed after an equilibration period of 5 min in accordance with current guidelines.1,7,8,10

Core temperature was recorded at defined time points: at baseline, 15 min after the beginning of warming (if applicable); after positioning of the epidural catheter, 15 min after the second period of warming (if applicable); at the beginning of surgery and then once every hour and on arrival at the ICU.

Mean skin temperature was calculated from measurements at the chest, arm, thigh and calf, and recorded at baseline, 15 min after the first and second warming periods and at the end of surgery on ICU arrival using an infrared device (Infrared Temperature Scanner, Model Dermatemp DT-1001; Exergen Corporation, Watertown, Massachusetts, USA).11

Postoperatively, all patients were admitted to the ICU. When core temperature was above 35.5°C, vital signs and haemodynamics were stable without vasoactive drugs and normal spontaneous breathing was observed, the trachea was extubated directly after the end of surgery. Patients presenting with hypothermia were neither awakened nor extubated but continuously ventilated under controlled conditions on the ICU, and active skin surface warming was started with the equipment mentioned earlier. In these patients, the trachea was extubated when normothermia was achieved.

Postoperative pain relief was achieved using a ‘patient-controlled epidural analgesia’ (PCEA) system with an electric GemStar pump (Hospira Inc., Lake Forest, Illinois, USA). The pump was programmed to deliver a continuous flow rate of 6 ml h−1 of ropivacaine 0.2%. In addition, the patient controlled a bolus function of an additional 6-ml ropivacaine 0.2% followed by a 20-min lockout interval so the upper limit of ropivacaine administration was set to 48 mg h−1.

Sample size calculation for the study was based on an expected treatment effect of 0.5°C on the postoperative core temperature because 0.5°C is the smallest difference that has been shown to be associated with hypothermia-induced complications.12

A sample size of 99 patients, divided into three groups, was estimated to provide 80% power for detecting a statistically significant difference at an α level of 0.05.

Statistical analyses were performed by using the statistics software GraphPad Prism 5.0 (GraphPad Software, San Diego, California, USA) and R 2.11.0 (R Foundation for Statistical Computing, Vienna, Austria). Peripheral oxygen saturation, MAP and heart rate were first averaged over time for each patient. These values were subsequently averaged among the patients in each group.

Kolmogorov–Smirnov testing showed no deviation from a normal distribution for tympanic temperatures, age, height, weight and BMI. These continuous, normally distributed variables were analysed using one-way analysis of variance and Scheffé's F test. Differences between variables were compared with paired or unpaired two-tailed Tukey–Kramer's t test or χ2 tests. To determine the interaction between time and prewarming, a repeated measures analysis of variance was performed with ‘time’ as the repeated measure and ‘prewarming’ as the factor, followed by Bonferroni correction. This analysis was repeated without the nonprewarmed control group to investigate possible differences between the two prewarming periods according to the procedure proposed by Pandit.13 Data are expressed as mean ± SD or median (IQR [range]), respectively. P less than 0.05 was considered significant.


All 99 patients were investigated up to the end of the study protocol (Fig. 1). Patients tolerated preoperative warming well and none asked to lower the warming device temperature to 40°C. As a result of the randomisation procedure, the number of patients between the treatment groups was slightly different.

Fig. 1
Fig. 1:
Flow diagram of study. EDA, epidural analgesia.

Patient characteristics, duration of surgery, dose of epidural ropivacaine, intravenous infusion volume, urinary output volume, room temperature and surgical procedures in each study group are shown in Table 1.

Table 1
Table 1:
Patient characteristics, duration of surgery, amount of epidural ropivacaine, intravenous infusion volume, urinary output volume, room temperature and surgical procedures in each study group

In the ‘warming before and after EDA’ group, there were more male patients (17 of 34) compared with the ‘warming after EDA’ group (nine of 33); in this group, the volume of infused Ringer's solution was higher compared with the group without warming (Table 1). During the observation period, peripheral oxygen saturation, heart rate and MAP did not differ among groups.

Changes in core temperature in each group are shown in Fig. 2. Preoperatively, baseline core temperature was near 36.5°C in 92 patients and did not differ among groups. Four patients in the ‘warming after EDA’ group and three patients in the ‘warming before and after EDA’ group were hypothermic (<36.0°C) at this time.

Fig. 2
Fig. 2:
Tympanic temperatures (T tymp) in patients who received passive insulation (no warming), 15 min of active warming after epidural blockade or active warming for 15 min both before and after epidural blockade. O, no warming; ▴, warming after EDA; •, warming before and after EDA. Data are presented as mean ± SD. § P < 0.05 versus baseline values, * P < 0.05 versus ‘no warming’ group, P < 0.05 versus ‘warming before and after EDA’ group. EDA, epidural analgesia.

In patients without warming, mean core temperature was 0.9°C lower compared with baseline values on arrival at ICU. Seventy-two percent of these patients were hypothermic. In the ‘warming after EDA’ group, core temperature on arrival at ICU was not significantly different from the baseline value and 1.0°C higher than in the patients without warming. Two patients (6%) were hypothermic at the end of surgery. In the ‘warming before and after EDA’ group, core temperature on arrival at ICU had increased by 0.7°C compared with the baseline value and was significantly higher than in the unwarmed patients (+1.5°C) and in patients receiving warming following EDA only (+0.5°C). None of the 34 patients in the ‘warming before and after EDA’ group suffered from hypothermia on arrival at the ICU.

Mean skin temperature of the patients calculated by averaging four skin temperatures were comparable at the beginning (baseline) of the observation period and showed values near 32.5°C (Table 2). In unwarmed patients, skin temperature at the end of surgery was significantly lower (−1.1°C) compared with the baseline value. In contrast, when warming was started before and/or after initiation of EDA, skin temperature at the end of surgery was significantly higher (+2.0 and +2.1°C) compared with the baseline value and with patients without prewarming (Table 2).

Table 2
Table 2:
Skin temperatures at different time points in each group, the incidence of hypothermia on arrival on ICU, number of ventilated patients and duration of ventilation in the ICU

Postoperative shivering was observed in two patients in the ‘no warming’ group but in none of the two warmed study groups (P > 0.05).

On arrival at the ICU, 11 of 32 patients (34%) without warming were mechanically ventilated because of postoperative hypothermia. The duration of postoperative mechanical ventilation was 36 ± 60 min with a maximum of 180 min. No patient in either warmed group required mechanical ventilation postoperatively.


The results of the present study showed that warming patients for 15 min before and after epidural catheter placement was sufficient to prevent hypothermia in all patients. Furthermore, prewarmed patients presented significantly higher skin temperatures up to the end of the surgical procedure. Despite the differences in the incidence of hypothermia, the incidence of shivering did not differ significantly among groups. Without prewarming, 72% of the patients undergoing major abdominal surgery under combined general anaesthesia and thoracic EDA became hypothermic at the end of the surgical procedure.

The value of EDA for major abdominal surgery was recently described in a systematic review.14 According to our study protocol, all patients received active air skin surface warming of their upper body during the surgical procedure and all fluids were actively warmed to 41°C. Therefore, the high incidence of postoperative hypothermia in the unwarmed patients may be surprising. The rapid decrease in body core temperature during EDA in combination with general anaesthesia is the result of sympathetic impairment, which causes pooling and redistribution of blood into the lower extremities. The vasoconstriction threshold of these patients is decreased and their core cooling rate is increased during the combination of general and epidural anaesthesia compared with general anaesthesia alone.15 Because of inadequate temperature monitoring and management during neuraxial analgesia, hypothermia may be underestimated by many anaesthesiologists.16

However, the vasodilatation and blood redistribution induced by neuraxial anaesthesia can be useful. After warming of the patients’ skin before the initiation of the procedure, inevitable vasodilatation and blood redistribution will not cause temperature loss because warm blood in the peripheral cutaneous vessels will be redistributed to the body core. Therefore, neuraxial analgesia may even help to prevent perioperative hypothermia if adequate prewarming is performed. However, this remains speculative because no control group without an epidural catheter but with a warming procedure was analysed.

In our hospital, the trachea is not extubated following major abdominal surgery if body core temperature is below 35.6°C. In these patients, mechanical ventilation on the ICU is continued until normothermia is reached, indicated by a core temperature of at least 36°C. Current guidelines present no clear threshold of the core temperature at which patients can be awakened and the trachea extubated safely. The common recommendation is that ‘patients should be extubated when they are normothermic’.1,7 Therefore, 34% of our patients in the ‘no warming’ group were mechanically ventilated postoperatively with an average duration of 36 min, up to a maximum of 3 h in one patient. Active skin surface warming was performed during this period. The trachea of all patients was extubated when the body core temperature reached 36°C.

The strengths of the present study are that we conducted a well controlled prospective randomised trial including patients scheduled for major abdominal surgery with a duration of anaesthesia and surgery expected to render them hypothermic. There were no dropouts.

Our study may be criticised because many patients did not undergo open surgery but laparoscopic procedures. However, the incidence of perioperative hypothermia in laparoscopic surgery has been found to be comparable to that in patients undergoing open abdominal surgery.17 In a retrospective study including 58 814 patients, Sun et al.18 recently demonstrated an incidence of hypothermia 3 h after induction of anaesthesia of about 35%. This is lower than our results in the ‘no warming’ group. However, it is not known how many patients in the study by Sun et al. received an epidural catheter and infusion. In addition, that investigation included a huge nonhomogeneous number of patients, a variety of surgical procedures and different outcome measures and is therefore not really comparable with our study. According to our study design (warming before induction of general anaesthesia), patients could not be blinded even though this factor is unlikely to have influenced our results. The higher volumes of crystalloids infused in the ‘warming before and after EDA’ group than the ‘no warming’ group are not expected to influence the incidence of hypothermia between the groups. This is supported by the higher temperatures in patients in the ‘warming after EDA’ group versus the ‘no warming’ group for similar volumes of crystalloids. Urinary output was comparable in all the groups. The interpretation of our findings may be limited because we did not measure the actual ‘heat content’ of patients. However, core temperature is the main outcome measure in clinical practice. As we did not examine the effectiveness of the epidural block, the possibility of a partial or complete failure of some epidural blocks in any of the groups studied cannot be excluded. Nevertheless, the infused volume of ropivacaine was comparable among the groups. For the sample size calculation, we used perioperative change in the body core temperature and not the incidence of hypothermia. Finally, our results refer to epidurally administered local anaesthetics but not to a combination of local anaesthetics and opioids. As opioids may aggravate blood redistribution by functional sympathectomy, further investigations are needed for clarification.

In summary, the present study under the described experimental design shows that warming of patients scheduled for major abdominal surgery not only before initiation of general anaesthesia but also before and after initiation of EDA is capable of preventing postoperative hypothermia. Our findings have an impact on the routine anaesthetic practice to prevent perioperative hypothermia. We conclude that EDA with its sympathetic interaction remains underestimated in routine clinical practice when it comes to its capability of (co-)inducing hypothermia and impairing thermoregulatory mechanisms.19 These results should be confirmed when epidural opioids are administered in combination with local anaesthetics.

Acknowledgements relating to this article

Assistance with the study: we would like to thank Rita Kipf, RN, and Ingeborg Meyer, RN, for their assistance with the study.

Financial support and sponsorship: none.

Conflicts of interest: BB has received consulting fees and/or travel support from CSL Behring, Air Liquide, 3 M, GE Healthcare and Deltex. He is member of the 3 M, Pulsion Medical Systems, Orion Pharma and Ratiopharm advisory boards. JH has received payments for lectures from 3 M, The Surgical Company, Moelnlycke and Abbvie.

Presentation: preliminary data from this study were presented at the German Anaesthesia Congress (DAC), 7 to 9 May 2015 in Düsseldorf, Germany.


1. NICE, 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), London. 2008. [Accessed 25 September 2015]
2. Frank SM, Beattie C, Christopherson R, et al. Epidural versus general anesthesia, ambient operating room temperature, and patient age as predictors of inadvertent hypothermia. Anesthesiology 1992; 77:252–257.
3. Matsukawa T, Sessler DI, Christensen R, et al. Heat flow and distribution during epidural anesthesia. Anesthesiology 1995; 83:961–967.
4. Hynson JM, Sessler DI, Moayeri A, et al. The effects of preinduction warming on temperature and blood pressure during propofol/nitrous oxide anesthesia. Anesthesiology 1993; 79:219–228.
5. de Brito Poveda V, Clark AM, Galvao CM. A systematic review on the effectiveness of prewarming to prevent perioperative hypothermia. J Clin Nurs 2013; 22:906–918.
6. Horn EP, Schroeder F, Gottschalk A, et al. Active warming during cesarean delivery. Anesth Analg 2002; 94:409–414.
7. Forbes SS, Eskicioglu C, Nathens AB, et al. Evidence-based guidelines for prevention of perioperative hypothermia. J Am Coll Surg 2009; 209:492–503.
8. Torossian A, Bein B, Bräuer A, et al.. AWMF S3 guideline on avoidance of perioperative hypothermia, 2014. Available at: [Accessed 1 March 2015].
9. Horn EP, Bein B, Bohm R, et al. The effect of short time periods of preoperative warming in the prevention of peri-operative hypothermia. Anaesthesia 2012; 67:612–617.
10. Hooper VD, Chard R, Clifford T, et al. ASPAN's evidence-based clinical practice guideline for the promotion of perioperative normothermia. J Perianesth Nurs 2009; 24:271–287.
11. Ramanathan NL. A new weighting system for mean surface temperature of the human body. J Appl Physiol 1964; 19:531–533.
12. Winkler M, Akca O, Birkenberg B, et al. Aggressive warming reduces blood loss during hip arthroplasty. Anesth Analg 2000; 91:978–984.
13. Pandit JJ. The analysis of variance in anaesthetic research: statistics, biography and history. Anaesthesia 2010; 65:1212–1220.
14. Popping DM, Elia N, Van Aken HK, et al. Impact of epidural analgesia on mortality and morbidity after surgery: systematic review and meta-analysis of randomized controlled trials. Ann Surg 2014; 259:1056–1067.
15. Joris J, Ozaki M, Sessler DI, et al. Epidural anesthesia impairs both central and peripheral thermoregulatory control during general anesthesia. Anesthesiology 1994; 80:268–277.
16. Arkilic CF, Akca O, Taguchi A, et al. Temperature monitoring and management during neuraxial anesthesia: an observational study. Anesth Analg 2000; 91:662–666.
17. Stewart BT, Stitz RW, Tuch MM, Lumley JW. Hypothermia in open and laparoscopic colorectal surgery. Dis Colon Rectum 1999; 42:1292–1295.
18. 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.
19. Ozaki M, Kurz A, Sessler DI, et al. Thermoregulatory thresholds during epidural and spinal anesthesia. Anesthesiology 1994; 81:282–288.
© 2016 European Society of Anaesthesiology